Blanksida
Electronic Availability of the Book
through the Association for Information Systems
It is a great pleasure for us that this book is made available electronically
through the Association for Information Systems, as the first volume in the
AIS Virtual Bookstore. We are most grateful to Gordon Davis and Detmar
Straub, who were instrumental in making this possible.
The electronic version of the book is complete and identical to the print
version, with the sole exception of the text on this page (omitted in the
print version). The book is originally published by the Economic Research
Institute, Stockholm, and can be referenced as a regular printed volume:
Sundgren, B., Mårtensson, P., Mähring, M. & Nilsson, K. (Eds.) (2003).
Exploring Patterns in Information Management: Concepts and Perspectives for Understanding IT-Related Change, Economic Research Institute,
Stockholm School of Economics, Stockholm.
We encourage inclusion of Internet reference information (URL) whenever
the book is referenced.
It is our hope that electronic availability of the book through AIS will lead
to an extended readership and stimulate further discussion and development of the ideas expressed in the book.
The Editors
Exploring Patterns in
Information Management
EFI Mission
EFI, the Economic Research Institute at the Stockholm School of Economics, is a scientific
institution which works independently of economic, political and sectional interests. It conducts theoretical and empirical research in management and economic sciences, including
selected related disciplines. The Institute encourages and assists in the publication and distribution of its research findings and is also involved in the doctoral education at the Stockholm School of Economics.
EFI selects its projects based on the need for theoretical or practical development of a
research domain, on methodological interests, and on the generality of a problem.
Research Organization
The research activities are organized in twenty Research Centers within eight Research
Areas. Center Directors are professors at the Stockholm School of Economics.
ORGANIZATION AND MANAGEMENT
Management and Organisation; (A)
Prof Sven-Erik Sjöstrand
Center for Ethics and Economics; (CEE)
Adj Prof Hans de Geer
Center for Entrepreneurship and Business Creation; (E)
Prof Carin Holmquist
Public Management; (F)
Prof Nils Brunsson
Information Management; (I)
Prof Mats Lundeberg
Center for People and Organization; (PMO)
Prof Jan Löwstedt
Center for Innovation and Operations Management; (T)
Prof Christer Karlsson
ECONOMIC PSYCHOLOGY
Center for Risk Research; (CFR)
Prof Lennart Sjöberg
Economic Psychology; (P)
Prof Guje Sevón
MARKETING
Center for Consumer Marketing; (CCM)
Acting Prof Magnus Söderlund
Center for Information and Communication Research; (CIC) Adj Prof Bertil Thorngren
Marketing, Distribution and Industrial Dynamics; (D)
Prof Björn Axelsson
ACCOUNTING, CONTROL AND CORPORATE FINANCE
Accounting and Managerial Finance; (B)
Prof Lars Östman
Managerial Economics; (C)
Prof Peter Jennergren
FINANCE
Finance; (FI)
Prof Clas Bergström
ECONOMICS
Center for Health Economics; (CHE)
Prof Bengt Jönsson
International Economics and Geography; (IEG)
Prof Mats Lundahl
Economics; (S)
Prof Lars Bergman
ECONOMICS STATISTICS
Economic Statistics; (ES)
Prof Anders Westlund
LAW
Law; (RV)
Prof Erik Nerep
Chairman of the Board: Prof Håkan Lindgren. Director: Associate Prof Bo Sellstedt
Address
EFI, P.O. Box 6501, SE-113 83 Stockholm, Sweden • Internet: http://www.hhs.se/efi/
Telephone: +46(0)8-736 90 00 • Fax: +46(0)8-31 62 70 • E-mail: [email protected]
Exploring Patterns in
Information Management
Concepts and Perspectives for
Understanding IT-Related Change
In honour of
Professor Mats Lundeberg’s
60th birthday
Edited by:
Bo Sundgren
Pär Mårtensson
Magnus Mähring
Kristina Nilsson
© 2003 by EFI and the authors
ISBN: 91-7258-631-1
Keywords: information management, information systems, informatics,
information technology, information, data, knowledge, organisational change,
change management, models, frameworks, theory, methodology, epistemology,
learning, knowledge management, e-business
Cover: Design and manipulated photograph by Christofer Tolis
Published and distributed by: The Economic Research Institute (EFI),
Stockholm School of Economics, Box 6501, SE-113 83 Stockholm, Sweden.
Internet: www.hhs.se/efi
Printed by: Elanders Gotab, Stockholm, 2003
Electronic version: www.hhs.se/im/exploringpatterns
We gratefully acknowledge the generous contributions
towards the publication of this book from:
Capio Diagnostik
Handelsbanken
Innovare Corporate Adviser
Pantor Engineering
Process Management Consulting
Red Lemon Datakonsult
Blanksida
Acknowledgements
When we began this book project in honour of Mats Lundeberg, we thought
we were in for a challenge: After all, our plan was to invite a selection of
scholars in the field of information management – all very busy people – to
write this book with us. We decided to invite people who we knew had
worked closely with Mats over the years, including those who had completed their dissertations under his supervision. Even if we had a sneaking
suspicion that people would try their best to join the project, it would have
been impossible to anticipate the overwhelmingly positive response: People
who really did not have the time to write a chapter for this book somehow
managed to find the time – simply because it was for Mats!
As editors, we would like to express our sincere gratitude to all authors for
their positive response to our invitation, for their excellent contributions
and for their friendly cooperation in the process of completing this book.
We are particularly grateful to one person who was very influential in the
formative years of Mats’ research career: Professor Emeritus Börje Langefors, who has written his personal preface to this book.
As it turned out, coming up with the list of contributors was the real challenge: It would have been possible to produce a very long list of people
who work or have worked with Mats. We focused on some of Mats’ longtime research partners, aware that it would be close to impossible not to
miss someone. Our apologies to anyone we might have failed to invite.
The entire Department of Information Management at the Stockholm
School of Economics has been involved in the production of this book. All
doctoral students have contributed in discussions on various chapters as
well as in the editing work. Our sincere thanks to Martin Andersson, David
Blank, Magnus Bratt, Niklas Källberg, Lasse Lychnell, Anders Mårtensson, Susanne Ohlin-Kjellberg, Björn Thodenius, Christofer Tolis, Frank
Ulbrich and Pablo Valiente. In particular, Christofer Tolis has contributed
generously and extensively throughout the final production process.
On behalf of all authors we hope that you will find the book interesting.
Stockholm, June 4th 2003
Bo Sundgren
Pär Mårtensson
Magnus Mähring
Kristina Nilsson
viii
Exploring Patterns in Information Management
— Preface —
In Honour of Mats Lundeberg’s
60th Birthday
Börje Langefors
Now that Mats Lundeberg is celebrating his 60th birthday it is natural for
me to remember when he first joined our department Administrativ Informationsbehandling (Business Information Processing). This was at a time
when most of the interest in the discipline moved around computer technology and “software” had begun to mean computer programs and programming. When it came to problems in applications, it was generally
assumed that the systems work had to be done by systems analysts who
had to ask the users about the requirements. Mats’ group at our institute,
the “ISAC Group”, were among the first to recognize that it had to be the
users themselves who must do the main systems design work. The analysts
would merely assist in this work. It then became clear that there were
many functionalities to be specified for the system that could not be identified as pre-existing requirements, but, rather had to be created during the
systems work.
The need to create the requirements to be specified, introduced organization theory into the information systems area, in addition to computer science aspects – “datalogical” aspects had to be supplemented with “infological” aspects. Mats Lundeberg became a leader in this development.
Mats’ recent work on e.g. Handling Change Processes seems to me very
interesting and important to the field of IT applications. The introduction
of new information handling resources is likely to entail changes in the
organization and how to handle these becomes an important question.
Mats has shared with me a deep interest in the problem of how information
is related to data and to knowledge. This I have found very stimulating.
Many persons have emphatically insisted that information is not knowledge. Of course, such statements are rather pointless as long as no one
knows what knowledge really is. During these last few days, in thinking
around the stimulating work I have been doing together with Mats over the
years past, there has come to my mind a rather simple analysis of the relation between Information and Knowledge. It is often pointed out that a
simple collection of separate fact or information messages does not con-
x
Exploring Patterns in Information Management
stitute knowledge. With this I agree, but the interesting observation here is
that a separate fact statement does not provide an information message
unless it “links” to existing pre-knowledge, as the infological equation
indicates. Thus while one may think of separate statements, separate information messages do not exist.
Now this linking implies that inferences may be drawn from the message
through pre-knowledge, and the possibility to make inferences from a message seems to be a very reasonable qualification for knowledge. I said that
this rather sudden insight has struck me in these few days but it agrees
with an earlier intuition (THAIS)1 that information could be seen as increments of knowledge.
As I came upon the thought that it can be so simply demonstrated that
information is knowledge, while thinking over my work together with
Mats over many years, I find it natural to feel that this is one more influence from him – though I am myself to blame for any critique it may generate, of course.
To finish, I want to congratulate Mats on his birthday and I am very
pleased to see the very interesting contents list for this book in his honour.
Börje Langefors
1
Langefors, B. (1966, fourth edition 1973) Theoretical Analysis of Information Systems,
Studentlitteratur, Lund, & Auerbach, Philadelphia, PA.
Table of Contents
Acknowledgements ............................................................................vii
Preface: In Honour of Mats Lundeberg’s 60th Birthday .....................ix
Börje Langefors
1. Introduction ..........................................................................................1
Bo Sundgren, Pär Mårtensson, Magnus Mähring & Kristina Nilsson
PART ONE: FUNDAMENTAL CONCEPTS
2. Information Systems for Concerted Actions ......................................11
Bo Sundgren & Gösta Steneskog
3. IT: An Ambiguous Technology? ........................................................39
Michael J. Earl
4. The Paradox of Perfect Knowledge ....................................................49
Alexander Verrijn-Stuart
5. Patterns of Change and Action:
A Socio-Pragmatic Perspective on Organisational Change ................63
Göran Goldkuhl
PART TWO: REFLECTIONS ON IT-RELATED CHANGE
6. Change Work in Organisations:
Some Lessons Learned from Information Systems Development ......83
Anders G. Nilsson
7. Patterns in Change Projects: Typical Traps ......................................101
Pär Mårtensson
8. Errors Help Users Learn? .................................................................117
Alf Westelius
PART THREE: MODELS AND FRAMEWORKS
FOR IT-RELATED CHANGE
9. IT Projects and the X Model .............................................................133
Erling S. Andersen & Åge Sørsveen
10. Implementation of eBusiness Models – the MTO-Framework ........147
Niels Bjørn-Andersen, Helle Zinner Henriksen & Michael Holm Larsen
xii
Exploring Patterns in Information Management
11. On Interpretation of Strategic Knowledge Creation
in a Longitudinal Action Research Project .......................................165
Pentti Kerola, Tapio Reponen & Mikko Ruohonen
12. Patterns in Information Management:
A Multi Level Analysis of Swedish Companies ...............................193
Kristina Nilsson
13. Some Issues in the Evolution and Use of Conceptual Frameworks:
A Commentary on the Lundeberg Framework .................................209
Magnus Mähring
14. Steps to an Ecology of the Multilevel Approach
to Information Management .............................................................229
Hans-Erik Nissen
15. Information Management:
Defining Tasks and Structuring Relationships .................................249
Dietrich Seibt
PART FOUR: DEVELOPING THE FIELD
OF INFORMATION MANAGEMENT
16. Building an International Academic Discipline
in Information Systems .....................................................................273
Gordon B. Davis
17. Users Matter – A Long Term Perspective ........................................291
Rolf Høyer
18. Building and Testing Theory on New Organizational Forms
Enabled by Information Technology ................................................305
Allen S. Lee
19. Choosing the Problem:
Information Technology versus Information Systems Phenomena ....321
Ron Weber
Contributing Authors ........................................................................333
Ordering Information ........................................................................339
—1—
Introduction
Bo Sundgren
Pär Mårtensson
Magnus Mähring
Kristina Nilsson
What is the identity of the field of information management?1 What makes
it unique and different from other disciplines? This is a question that has
challenged Mats Lundeberg during his research career, and it is a question
to which he has given several important answers through his research.
This book contains contributions from a number of researchers, colleagues
of Mats Lundeberg, who in various ways have addressed the question of
what the field is about, devoting time, energy and creativity to subjects
they feel are important research topics in information management. Many
of the contributors in this volume also play and have played important
roles in the shaping of the field.
We believe that the collection of chapters in this volume outlines part of
the identity of the field of information management. Our first and foremost
aim with this volume, however, is that you, the reader, will gain new
insights that you find valuable in your role as a student, reflective practitioner, or researcher.
The occasion and direct reason for coming together and writing this book
is to celebrate Mats Lundeberg on his 60th birthday. Each contributor (or
team of contributors) has been free in choice of topic, but as editors, we
have been confident that the selection of people (friends of Mats) and the
occasion (Mats’ birthday) would result in a coherent collection of topics.
We were right. The title of this book contains a number of keywords that
we may group into several themes that are important to Mats Lundeberg in
his research, to the contributors to this book, and to the book itself. In
addition to these themes, the development of the academic field of information management, in which Mats Lundeberg has played a significant
role, is also of importance.
1
Also information systems or informatics.
2
Exploring Patterns in Information Management
The Concepts of “Information” and “Information Systems”
As suggested by Ron Weber, as well as by Bo Sundgren and Gösta Steneskog, and by Allen Lee in their respective contributions, the concepts of
“information” and “information system” are two of the most central concepts, if not the most central concepts in the discipline of information
management. Lee defines information as the knowledge that a person
forms from data, knowledge being the understanding that a person has, or
the mental model of reality that Sundgren and Steneskog write about.
The failure to distinguish between information and data, and between
information systems and data processing systems, which is so common not
only among practitioners, but also – unfortunately – among researchers,
has serious consequences. Even if we perceive the same reality and use the
same data and data processing systems, we cannot be sure that we interpret
data and understand reality in the same way. At the same time we need to
co-operate in the societies where we live, and in the organisations where
we work. However, if we are aware of the problems, we can do something
about them, for example by constructive socialization (interacting and
communicating with each other), as suggested by several contributors to
this book, and by systematically associating data with metadata in data
processing systems, as Sundgren and Steneskog suggest.
Patterns and Perspectives
Nobody knows exactly how human beings form concepts and create comprehensible mental pictures of the real world, but it seems that from the
very second when we enter this world we start seeing regularities and patterns. By interacting with other human beings we tune the patterns we perceive with theirs. In his research, Mats Lundeberg has devoted a lot of
energy to finding patterns and demonstrating the power of patterns when
analysing and designing information systems and IT-related change. Patterns, when first experienced, may be rather vague. An important task for
the researcher is to describe the patterns more precisely. When this is done,
we often refer to the patterns as “models”. A number of important patterns/models, also used by many practitioners in the field of information
management, carry the signature of Mats Lundeberg, including the models
of the ISAC approach to information systems development, the X-model,
the Y-model, the V-model, and the multi-level model. In their contribution
to this book, Erling Andersen and Åge Sørsveen use the X-model (which
they developed in collaboration with Mats Lundeberg) as a general project
evaluation model. They do so in two ways: in analysing a particular project
Sundgren, Mårtensson, Mähring & Nilsson
3
as well as in analysing a population of projects on an aggregate level,
thereby identifying recurring patterns. In her contribution, Kristina Nilsson
uses the multi-level model as a framework for analysis in discussing current practice in information management in a selection of major Swedish
companies. Magnus Mähring discusses the evolution of several of these
models and their uses, pointing to avenues for further development, particularly in research studies.
As researchers taking active part in practical projects, we may contribute
substantially by introducing powerful patterns into the work. At the same
time we should keep our eyes open for patterns that we have not recognised before. This is what Anders Nilsson, Pär Mårtensson, and Alf
Westelius have done in their essays. Anders Nilsson writes about “lessons
learned”; about how lessons from information systems development can
inform the practice of managing change. Pär Mårtensson identifies “typical
traps” in change projects, using models to detect recurring patterns of
behaviour and thinking that may misdirect change efforts. Alf Westelius
puts focus on the opportunities of learning from different kinds of errors
and malfunctions that occur in organisations in general and in relation to
information systems in particular. What all three authors do is point to, and
describe with practical illustrations, a number of typical patterns, the
knowledge and understanding of which may help us diagnose problems in
change projects, or even better, to avoid certain types of problems.
Sometimes one may get the feeling that the discipline of information management has produced too many “models”, that there is an overflow of
models that make us slightly overwhelmed. To some extent this “pollution” problem is worsened by the requirement from practitioners that a
model should be accompanied by a tool. Unfortunately such tools sometimes tend to become ends in themselves, and the users of the tools may
even forget what were really the good ideas behind the models behind the
tools.
There are different strategies for tackling this problem. One approach is to
view the different models, with associated tools, as tools in a toolbox or
components in a library of models. If this is going to improve the situation
it is necessary to make sure that the tools and components fit together. In a
sense there must be some kind of “super-model” that connects the individual tools and models. This is well in line with the striving for generality
that is typical for science and research.
On the other hand we must be on our guard against unrealistic and even
harmful pretensions, both from researchers and from commercial software
4
Exploring Patterns in Information Management
providers, to have the “complete model” or “final solution” within reach.
Such generality does not exist.
The idea of perspectives, often advocated by Mats Lundeberg, offers a
good balance between an unstructured chaos of incompatible models
(“apples and oranges”) and the general model that covers everything. For
example, Dietrich Seibt in his article identifies four dimensions (task,
technology, organisation, person) and then elaborates how this structuring
may help in the development of information systems supporting business
solutions. Seibt also discusses the value of more abstract patterns and suggests that “maybe we need ... a continuum of patterns, which can be concretised step by step”.
A slightly different structuring of perspectives, some version of which is
used by many authors, comes from distinguishing between strategic
aspects (pragmatics – why are we doing something and for whom?), conceptual aspects (semantics – what are we doing?), and operations aspects
(syntactics – how are we doing it?) Modelling techniques developed and
used in information management often focus on one of these aspects, e.g.
value modelling focused on strategic issues, process modelling focused on
operations, and conceptual modelling focused on information contents.
Applying different perspectives is not only a way of making an analysis or
a design task more manageable – this is in itself very important, of course
– but it is also an efficient way to discover incompletenesses, complexities,
inconsistencies, and even conflicts. “Two descriptions are better than one”,
as Hans-Erik Nissen says, quoting Bateson. No analyst and no project
manager can know everything that is relevant for a certain task, or be a
specialist in everything, but by systematically applying different perspectives and modelling methods, he or she may decrease the risks of omitting
something that is important, or underestimating the complexity of the task
at hand.
Understanding IT-Related Change
Ron Weber, in his contribution, points out that human use of a certain
technology does not usually lead to a new scientific discipline or even to
new theories. Michael Earl, in another article in this book, similarly
observes that sociologists and economists have not felt the need to make a
distinction between information technology and other technologies. Yet
the use of information technology in businesses is certainly a major issue
for many researchers in information management, including Mats Lundeberg. How could we explain this?
Sundgren, Mårtensson, Mähring & Nilsson
5
Most other technologies that have so far been developed and used by human
beings have aimed mainly at supporting and amplifying physical capabilities; consider for example tools used by craftsmen, machines used by workers, and cars used by people who want to transport goods and themselves
from one place to another. In contrast, information technology to a great
extent aims at supporting and amplifying our mental and intellectual capabilities: e.g. our abilities to observe and obtain information about reality, to
remember and process this information in view of other knowledge that we
are already in possession of, to share our knowledge with others, and to plan,
execute, and evaluate individual or concerted actions.
The use of information technology certainly involves ergonomic and psychosocial problems that also occur in use of other technologies, and there
is hardly a need to establish a new discipline for the study of those problems in connection with information technology. By being an extension of
the human mind rather than of the human body, information technology
introduces new classes of problems and opportunities. Earl, for example,
discusses the learning process. Learning to use a tool or a machine supporting physical work is not likely to affect us very much as persons,
whereas the use of information technology supporting intellectual work
may have a considerable impact on our understanding of the world around
us. When we use information systems supported by information technology, there is a unique and urgent need to understand the complex relationships between reality (as perceived by different people), data, information,
and knowledge (personal and organisational).
For ages, philosophers have studied questions of knowledge acquisition (or
epistemology) and sociologists and organisation theorists have developed
important theories of socialization and organisational knowledge and
behaviour. We should certainly take advantage of the scientific results from
these and other areas when we study information systems and IT-related
change. But we also have a lot to add. This is demonstrated in the essays in
this book by Göran Goldkuhl, by Hans-Erik Nissen, and by Pentti Kerola,
Tapio Reponen, and Mikko Ruohonen.
Usage of information systems and information technology has an interesting feature that it shares with science in general. It is often driven by
human curiosity and planning; prognosticating is not always very meaningful. The development may sometimes take a route that is completely
unexpected even for the originator of the development. Earl mentions Tim
Berners Lee as an example: did he foresee the amazing enabling scope of
the world wide web? The computer itself is another example. As can be
concluded from its name in English, it was originally intended for mathe-
6
Exploring Patterns in Information Management
matical computations. Experts in many countries, including Sweden and
the United States, who were asked by politicians how many computers
would be required in their country, came up with answers like “one, two,
or possibly three”.
In line with these observations one must consider both planned and
unplanned changes when new information solutions are introduced in an
organisation. Everything cannot be pre-planned. It is common in information management literature to distinguish between continuous improvements and radical changes (cf. the chapter by Goldkuhl). It is easy to
believe that whereas the former may occur more or less spontaneously in
an organisation, the latter will require carefully planned projects. The
examples just given show that even radical changes may occur spontaneously; usage of the Internet gained momentum and established itself as a
fundamental infrastructure throughout society in just a few years.
On the other hand it can be debated whether everything that seems new is
really new and requires new theories and methods. Niels Bjørn-Andersen,
Helle Zinner Henriksen, and Michael Holm Larsen in their contribution
suggest that e-business projects require new implementation models, and
they propose a new framework for that purpose. Allen Lee, on his part,
questions in what way and to what extent phenomena like “virtual work
teams”, “virtual libraries”, “virtual markets”, and “virtual corporations”
really are new phenomena, and if electronic commerce really is a new
form of social organising.
The difficulties of clearly seeing and foreseeing even very important
effects, positive and negative, of an IT-related change, make it extremely
difficult to calculate investments in information systems and information
technology in the same way as calculations are made for business investments of more conventional type. Information technology is an ambiguous
technology, says Earl, referring to uncertainties in several dimensions.
Alexander Verrijn-Stuart in his article shows that these uncertainties imply
a fundamental impossibility to make precise cost/benefit analyses (as are
often required), but argues that managers may to some extent compensate
the lack of perfect knowledge with sensible and alert reactions to signals
from within the organisation and its environment.
Kristina Nilsson notes in her study that “it is normally the amount required
for an investment that decides whether or not IT/IS-related issues are
treated in the Executive Group”. Together, these statements vividly illustrate that the current practice of strategic planning and control for ITrelated initiatives leaves ample room for improvement.
Sundgren, Mårtensson, Mähring & Nilsson
7
Developing the Field of Information Management
Several contributions in this book touch upon the history of information
management. In particular, Gordon Davis discusses the international development of the discipline and Rolf Høyer summarises part of the history of
the so-called “Scandinavian School” in information management. Obviously, Mats Lundeberg has contributed to both developments and is part of
both histories.
Høyer points to the sustained importance of key areas in the field, thus
reminding us – as does Lee – that this field is not dependent on catering to
current trends to stay relevant, although our relevance may also benefit
from providing perspectives on current issues in the world of organisations. Davis argues for further development of the field that harnesses and
advances the unique core of the field, while also exploiting the opportunities provided by the many interfaces that the field has to other disciplines.
This book arguably demonstrates this strategy: while paying considerable
attention to the core of the field, several chapters also illustrate how the
field reaches out into other, related fields. Mats sometimes quotes Ashby’s
law of requisite variety to argue for the virtue of flexibility. It would seem
that the field of information management, by combining a core with the
ability to change and adapt, might possess this virtue.
Taken together, the chapters in this book thus address the question of the
identity of the field. Given the above strategy, any answer to this question
must be tentative, but it is precisely through our repeated, tentative
answers to persistent questions that we most advance our knowledge.
8
Exploring Patterns in Information Management
Blanksida
PART ONE:
FUNDAMENTAL CONCEPTS
Blanksida
—2—
Information Systems
for Concerted Actions
Bo Sundgren
Gösta Steneskog
Introduction
It is often said that we are entering the Information Society. But hasn’t
man has always been forced to obtain information about what is going on
around him just to stay alive? Didn’t the members of the hunting pack
communicate with each other during concerted actions to kill their quarry?
Being able to use languages was a giant step forward in communication.
Vast amounts of information were coded into some sentences, transmitted,
and decoded by the receiver. The next invention was writing. It allowed
the sharing of information over time and space. The invention of printing
machines enabled mass-production of information in a way that really
changed the world into the “Gutenberg Galaxy” (McLuhan, 1962).
Electronic communication facilities with the telegraph, telephone, radio,
TV, IT, and now the Internet have radically shortened the time required to
communicate over the whole world. We have got an Internet Galaxy.
Still – the purpose of all these exciting developments has been to further
improve man’s inherited abilities to get informed and to communicate with
others in order to “get things done, to achieve goals beyond the reach of
the individual” (Scott, 1998). Our approach here is to see our world as an
Information Galaxy, or a Cyberspace, where human beings process information, communicate and use different tools in order to coordinate their
thinking and actions. That is why we believe that the study of information
management is so important.
Information and Data
A major problem in the present thinking about information is that the
words “data” and “information” are often used indiscriminately and inter-
12
Exploring Patterns in Information Management
changeably, without reflection. Or they are defined in terms of each other.
There is also a certain terminological inflation here: what was originally
called “data” is now called “information” or even “knowledge”.
As mentioned, human beings have always created and processed information and knowledge in their minds. They have also, long before computers
were invented, produced and processed data. Actually human beings use
two kinds of data. On the one hand, a human being is able to perceive the
outside world through her senses: sight, hearing, feeling, taste, and smell.
The senses, applied to reality, produce data, perception data, which the
human brain is able to interpret.1 Through these data the human mind will
step by step be able to create a mental picture, or model, of the external
world.
On the other hand, human beings sometimes try to represent part of their
mental models of the real world by means of data – symbolic data. These
data symbolise some selected part of the human being’s perception and
understanding of the real world. Some symbolic data (iconographic pictures, onomatopoetic sounds) are very easy to interpret. Other symbolic
data, for example a word in a language, written in a certain alphabet, are
based on some kind of coding convention, and anyone who wants to interpret such data must know the coding convention.
Information
Mental models
Real World
Phenomenon
“A car”
Direct
Data
Symbolic
Data
Figure 1. Data and information
Perception data may be called direct data, since they directly reflect the
real world. Symbolic data are indirect data in the sense that they are (a) the
1
The human being is also able to perceive signals from her own body like hunger,
thirst, pain, etc.
Sundgren & Steneskog
13
result of a human, creative act, which is then (b) (re)interpreted by (possibly) another human being.
Perception data reach a person via her senses, and the person then interprets the data into an understanding of the current situation, using concepts
and frameworks that already exist in her mind – the person’s frame of reference. Concepts and frameworks are the results of a life-long learning
process. A major part of that learning is influenced by the social environment where we live, so our frameworks, and hence also our interpretations,
are socially based.
Philosophers have always discussed how we obtain information and
knowledge about the real world. There are certainly different views, or
paradigms, about how the knowledge formation process actually works,
but most thinkers seem to agree that we form some kind of mental pictures, or conceptual models, of an outside world inside our own minds, and
that these conceptualisations constitute a basis for our understanding of
reality and our actions vis-à-vis this reality.
As suggested by Ogden’s triangle (Ogden & Richards, 1956), shown in
Figure 2, (symbolic) data can be seen to represent, or stand for, a realworld phenomenon, but this relationship is only indirect, since it depends
on the real-world phenomenon first being mentally conceptualised by a
human being.
sy
to
mb
o
ers
lis
ref
es
INFORMATION
DATA
stands for
REAL-WORLD
PHENOMENON
Figure 2. Ogden’s triangle
Since symbolic data are themselves a part of reality, they may again be
perceived by human beings, and re-interpreted into concepts and information. The interpreter may be the person who originally stored the symbolic
data, but it may also be somebody else. In the former case, the data may
remind the person about something that he or she has already forgotten,
14
Exploring Patterns in Information Management
and in the latter case the symbolic data may be part of a communication
process between the two persons involved.
When a person stores symbolic data in some physical form, or medium,
outside the human mind, e.g. on a stone, a piece of paper, or in a computer,
the person uses the medium as an extension and amplifier of the memory
capacity of her own mind. Similarly, a person may use symbolic data and
data processing tools as extensions and amplifiers to her own information
processing capabilities. Consider for instance an engineer analysing and
solving construction problems by creating and manipulating mathematical
symbols and models, supported by instruments like pencil and paper or
software-supported computers.
When symbolic data are used for storing and communicating information
over time and space, the storage and communication processes may be far
from perfect. In fact, one can never know if one person interprets the same
data in the same way as another person. One cannot even be sure that the
same person will interpret the same data in the same way at different
points of time. Different persons, and the same person at different times,
will have different frames of reference, and this is one important reason
why the interpretations are likely to be different.
Langefors (1995) describes the mental process of interpreting data into
information by means of the infological equation
I = i(D, S, t)
where
•
•
•
•
I is the information contents obtained by a human being
i is the process of interpretation and creation of meaning
D is the received data
S is the frame of reference, or accumulated knowledge, used by the
interpreter
• t is the time used for interpretation
So far we have analysed how an individual human being may form concepts and information, and how a person may use data and man-made tools
in order to amplify her own mental capabilities and communicate with
other individuals. But the human being is, to a higher or lesser degree, a
social creature. We do things together, both because we like to do things
together, and because we need to do them together. In a modern society we
are in fact extremely dependent on each other, and it is hard to imagine
that anyone of us would survive particularly long, if we were left alone in
the world. We will return to this.
Sundgren & Steneskog
15
Reality as a Mental Model and Social Construction
Does reality exist as something objective, independently of human beings
perceiving and conceptualising it? This question has been debated by philosophers through ages. An extreme position is that reality exists only in
my mind; thus the objective reality is only an illusion; if I cease to exist, so
does reality.2 Most people today are probably convinced that there is an
external reality that existed before me and that will continue to exist after I
have passed away. But how “objective” is this reality, and how independent is it of the mental models in our minds? There are certainly physical
things like mountains, trees, buildings, etc, that seem to have an objective
and independent existence; these are things that we can perceive through
our senses. But what about an abstract entity like a business or an organisation? We can perceive a lot of entities that are associated with a business, e.g. buildings, equipment, staff, etc, but we cannot perceive the business as such. Yet the business seems to be much more important than for
example the building where the business is located. Let us assume that
there is a bad business located in a bad building and a good business
located in a good building. Even if these two businesses switch buildings,
the bad business will probably continue to be a bad business, whereas the
good business will by and large remain a good business. If all people
would disappear from earth, all businesses would certainly disappear with
them. A business seems to be a part of reality that exists in our minds –
and only in our minds. Thus a business seems to be an example of what
Berger and Luckmann (1966) call a social construction.
The existence of social constructions does not necessarily imply that the
whole of reality, or all its parts, is a social construction. But there is often
an element of social construction even in concepts that we normally regard
as very “objective”. For example, consider the concept of a table or a chair
(Based on a discussion in Flensburg, 1986). Most of us would probably
claim that we have no problem distinguishing between tables and chairs;
furthermore we might even claim that we are able to provide definitions of
the two concepts that would make the distinction between them clear. But
if we try to formulate these definitions, we will all the time find that there
are chairs that fulfil proposed definitions of tables and vice versa. We cannot rely on semantic aspects only. We must at least introduce pragmatic
aspects as well. For what purposes do we use tables, and for what purposes
2
Note the distinction between the position that “my reality” is the only reality that
exists, and the position that “my reality” is likely to be different from “your reality”, even if we live in “the same reality”.
16
Exploring Patterns in Information Management
do we use chairs? As human beings being part of a certain type of society,
we typically use chairs for sitting at a table, where we have placed things
that we are using for whatever we are doing, alone or together with other
people.
Information Processing and Data Processing
The human brain is perfectly capable of processing information without
any assistance of external tools. We have already discussed how the
human mind interprets perceptions from the external world in order to
form concepts and information. The interpretation process is partially controlled by earlier interpretations, residing in the human being’s mind. Even
if a person is born with a blank mind, or an almost blank mind, her frame
of reference will be updated all the time through new perceptions and
interpretations. A human being will also reflect upon the information that
is already in her mind, and this process will again result in new or modified concepts, new or modified knowledge, and a new understanding of
herself and the world around her. We do not know exactly how a human
being “digests” information, but there are certainly elements of both
induction and deduction. “Intuition” is used as a term for describing a certain type of mental processes that we cannot really analyse.
It is important to realise that information (a) cannot be stored, at least not
in a literal sense, outside the human mind where it has been created; (b)
cannot be communicated to other people, at least not without more or less
serious, and more or less unknown, distortions. Still all humans have
always wanted to do exactly these things: we want to store information
outside our minds, using external storage media as an extension to and
strengthening of our own capabilities of memorising and remembering,
and we want to exchange information with other people, for both personal
and social reasons. In order to do these things we have certain proxy processes at our disposal, given by God, developed by chance and genetic
mechanisms (cf. Darwin), or invented by man. For example, we use spoken and written messages in different languages, and we use computers for
processing symbolic data.
Figure 3 gives an overview of some basic interactions between
• information and information processes in the human mind
• data representing information outside the human mind
• the real world, reality
Sundgren & Steneskog
17
The figure gives a dynamic view of the world. There is a basic “reality circle” where
1. A person perceives reality using her senses and possibly some manmade instruments
2. The mind interprets perception data, using her existing frame of reference
3. The mind digests existing concepts and information
4. A person decides to do something, e.g. change reality or create data,
and acts accordingly
5. Reality is being changed or changes itself.
3
Mind
In
te
rp
re
t
digest
2
1
Pe
rc
ei v
De
4
Body
cid
e
5
Reality
changes or is being changed
e
thermometer
thermostat
Ac t
Symbolic Data
changes and is processed
Figure 3. Interactions between human beings, data, and reality
By combining these basic operations into sequences and iterations we may
describe more complex phenomena. Note also that man-made instruments
and tools may be used in many processes. For example, a pair of glasses or
a microscope may be used when making observations of reality, and pencil
and paper may be used when creating data representations.
Reality may also seem to produce data by itself, but the process is actually
enabled by man-made instruments, e.g. a thermometer that registers temperature. In that sense all data are dependent on the existence of a human
mind. Also when data “automatically” changes reality, e.g. in the case of a
18
Exploring Patterns in Information Management
thermostat, there is a human mind behind the design and construction of
the tool (cf. Langefors, 1995).
Reality
changes or is being changed
thermometer
thermostat
Symbolic Data
changes and is processed
Figure 4. Automatic control loop
Human Interaction:
Communication, Co-operation, and Conflicts
We have already noted that the human being is a social creature. We need
to communicate In order to co-operate and to achieve certain goals, and,
what is maybe even more important for our behaviour, we want communication and co-operation for its own sake. Even hermits need a certain
amount of social contact. All human interactions are not idyllic. Sometimes we run into disagreements, conflicts, and wars, but even in such
situations it seems to be a natural human struggle to find ways out, through
negotiations and compromises, i.e. through information processing and
communication.
Human beings seem to have lived in collectives and societies in all times,
that is, they have organised their lives together to some extent. Families,
households, villages, tribes, and nations are examples of different kinds of
more or less “natural” organisations. The members of a group or a society
co-operate in many different ways. Sometimes a task is simply too big for
a single person to manage. In other cases specialisation and division of
labour turns out to be rational for achieving individual as well as collective
goals.
Many societies, e.g. the Vikings, found it worthwhile to establish contacts
with others, driven maybe by curiosity as much as by a desire to reach
material advantages through trade and conquering.
In later times the human drive to co-operate in order to achieve goals
beyond the reach of individuals has translated into the formation of organisations for specific purposes, e.g. business companies, but also churches,
trade unions, government agencies, hospitals, etc. Today organisations
Sundgren & Steneskog
19
become more and more sophisticated and are themselves organised into
higher-level organisations and networks. Whereas an organisation of traditional type usually has a hierarchical structure, networks have more complex mechanisms for control and co-operation. Markets represent yet
another form of organised human interaction.
People are the driving force of an organisation. A concerted action requires
communication between the participants. Each participant must have a
clear understanding – information – about the current situation and what is
expected from him or her. This in turn requires every participant to gain
sufficient knowledge about professional and business-related frameworks,
as well as an understanding of languages and other ways of communicating.
Earlier in this article we used the infological equation and a graphical
model to clarify the distinction between information and data and to
explain the importance of this distinction for understanding a human
being’s usage of information and data. In order also to cover concerted
human actions in order to achieve goals beyond the reach of individuals,
we need to elaborate these models.
Direct and Indirect Communication
Figure 5 shows models where communication and co-operation between
people have been included. Using these models, we can analyse in some
detail what happens when a person A communicates with a person B. We
may distinguish between direct and indirect communication.
Direct communication may mean that Person A decides to represent some
piece of information in her mind with some data (e.g. a spoken statement)
that can be perceived and interpreted by Person B. Next step is the inverse
process: i.e. Person B replies.
Other situations that can be described by the same pattern are telephone
calls, fax transmissions, exchange of surface mail and e-mail, etc. The
communication may not always be instantaneous, but it will still be
regarded as direct as long as it is sent directly to some identified person(s).
Another important type of direct communication is a non-verbal one. We
are part of each other’s reality. Movements, postures and other signals
provide rich and varied forms of direct communication. In spite of
broadband, TV-conferences, teleworking and telecommuting, face-to-face
communication will remain to be of great importance. It is easy to forget
that in these days, “Man is man’s joy” (Havamal).
20
Exploring Patterns in Information Management
Mind
Mind
Interpret
Decide
Mind
Mind
Interpret
Decide
Act
Act
Perceive
Perceive
Symbolic Data
Symbolic Data
input-process-output
Figure 5. Direct communication using data (left), and indirect communication with
intermediate processing (right)
Indirect communication is characterised by the fact that data produced by
A is stored and possibly processed by a man-made data processing system
from where it can be retrieved by B and other people, not necessarily
known to A, possibly long after A stored the original data. This is shown
to the right in Figure 5.
Languages using symbolic data have tremendously increased man’s abilities to communicate both directly – the message is delivered unchanged –
and indirectly – the message is processed and combined with stored date
before being perceived by the receiver.
Communication over Time and Space
Due to the shortcomings of human memory, man has been using different
techniques for a long time to improve the preservation of information.
Nowadays paper and pencils (and electronic equivalents) are important
tools for storing and retrieving data of a not-so-well-structured type.
Storing and retrieving data could be looked upon as communication
between people over time. Our abilities to remember have grown. In the
old days a human being’s mental capabilities to remember was almost the
only way for carrying the heritage from generation to generation. Sophisticated methods were developed to ensure this – our ancestors were much
more skilled than we are.
Sundgren & Steneskog
21
Our abilities to eliminate the time gap by storing and later retrieving data
have increased substantially by the use of written languages. It has been,
and still is, a major tool for accumulating human experiences over the generations. Dramatic improvements in this respect have taken place in the
“Gutenberg Galaxy” and now in the Cyberspace.
Communication over longer distances was once solved by the use of couriers (Marathon). This was later improved by the use of written messages carried by a messenger. In order to decrease the delivery times, flagstaffs and
smoke-puffs were tools used to communicate over a distance. Recent inventions are the telegraph, telephone, and radio. Now we live in cyberspace,
where huge amounts of data are available immediately and everywhere.
Other technological developments have also substantially improved our
abilities to expand the richness of our data by the use of photos, pictures
and other iconographic data.
Data processing systems can be seen as offering proxy processes for human
exchange of information over time and space. If A is an archivist, and B is a
researcher using data archived by A, A and B may not know each other, and
they may not even live during the same century. Yet, thanks to the stored
data, there may be some kind of communication between A and B. Obviously, this communication will not be perfect – there are many sources of
error in the communication process – but there are ways to improve the
quality of this kind of communication. However, such improvements require
a good understanding of the distinction between information and data.
Sharing of Data and the Need for Metadata
We have just described a situation where people share data (over time and
space). Sharing of data is actually a proxy process for sharing of information; as we already know, sharing of information is fundamentally impossible. We can only do our best to improve the chances that different persons sharing the same data will interpret them in the same, or at least
similar, ways. How can we do that?
A person’s interpretation of data depends on the person’s frame of reference, which consists of concepts and information in the person’s mind. If
two persons have the same, or at least compatible3, frames of reference, it
seems likely that they will interpret the same data in similar ways.
3
Two frames of references are compatible (for a certain purpose) as long as they
do not (severely) contradict each other in relevant parts.
22
Exploring Patterns in Information Management
But how do we know that two frames of reference are compatible? They
cannot be inspected or compared with each other, at least not directly,
without intermediary data processes that will anyhow introduce uncertainties and errors into the comparisons. A frame of reference is the product of
life-long learning, driven by the person’s perceptions and reflections. Thus
one thing that would increase chances that two persons have compatible
frames of reference is that they share similar experiences. If this is not the
case, we may try to make the frames of reference more compatible in some
other way, e.g. by documenting essential parts of the respective frames of
reference and making these documentations available to the persons who
need to be able to communicate and share data. This again means representing information by data. Since these data represent background information needed for proper interpretation of other data, we call them metadata; data about data. The communication of metadata is subject to the
same fundamental difficulties as the communication of the basic data that
they describe, but even so, adequate metadata will reduce the range of possible interpretations of the data that they describe, and thus improve the
chances of different persons making similar interpretations of the same
data. In other words we increase the intersubjectivity of the data.
Concerted Actions Towards Goals
Adequate information management is especially important when we try to
co-ordinate the actions of individuals and groups in complex and unpredictable situations, e.g. in a military campaign, or when managing a multinational enterprise.
An organisation’s total information system is made up from the mental
concepts and frameworks of the participants in the organisation, the data
passed to and between the participants (processed along the road), and the
resulting individual perceptions and understandings of the situations leading to individual actions. These actions are expected to lead to the fulfilment of shared goals.
Concerted actions towards goals are facilitated by
•
•
•
•
•
common understanding of goals (some of which may be conflicting)
good communication
common culture, languages, coding conventions
compatible frames of reference
common data and metadata
Sundgren & Steneskog
23
Individual actions and, even more so, concerted actions by people in cooperation often benefit from a certain amount of planning. In a collective
planning process, the participants develop shared descriptions of
• the present situation
• a desirable future situation
• possible ways of getting from the present situation to the desirable
situation
The descriptions can be seen as models, and this is an example of how data
in the shape of models can be used as instruments for people (alone or in
co-operation) to control an external reality.
Human Beings and Computers
Computers have become very important tools in this world of information
and data but the relationships between humans and computers have varied
over the years. When computers first came around they were literally
computers; they helped human specialists, physicists and engineers, to
make computations in a fast and accurate way. The computers were more
advanced and more efficient than the calculators that had existed before,
but they were calculators. They were extensions to the human specialists
that used them, and they amplified the computational capabilities of their
users. The humans were still in full control of their new tool.
At this time the governments in many countries, including Sweden and the
United States, appointed committees, where highly qualified experts were
asked to estimate the number of computers needed in the future. The prognoses typically indicated that a country would need one, two, or possibly
three computers. It should be noted the even the biggest and most powerful
computers in those days had much less capacity than the simplest PC has
today.
How could the experts be so wrong in estimating the need for computers?
There were several reasons. The computers were still very expensive, and
it was hard to justify the purchase of a computer. The experts did not foresee the rapid technical progress that would make computers dramatically
more powerful and less expensive in a near future. Neither did they take
into account that, once computers existed, they would not only solve
existing mathematical problems more efficiently, but they would also
make it possible to solve new categories of more complex mathematical
problems that had not even been possible to consider before.
24
Exploring Patterns in Information Management
However, most importantly, the experts were not imaginative enough to
realise that computers had the potential to be useful for many other categories of data processing tasks than solving complex mathematical problems.
But, as is often the case, when a technical novelty gets used by more and
more people, quite unexpected things start happening. Many innovations
are driven by curiosity and even by mistakes.
Thus it was soon discovered that computers could do much more than
computing. Computers turned out to be well suited to assist human beings
with tedious and error-prone tasks like accounting, invoicing, production
of statistics, and even quite trivial data shuffling. New categories of people
were introduced to computers, and these users had a need for more complex man/machine interfaces than the mathematicians, who just fed the
computer with some equations and a few numbers and often expected only
a few numbers in return.
Computers were still big, monolithic4, and very expensive creatures, so
all users had to share the same computer. Obviously all of them could not
be in the computer room, at least not at the same time. New professions
were created: operators, systems programmers, data entry staff, and
application programmers, who all acted as middle-persons between the
real users and the computer. The real users were now called “end-users”,
since they were at the end of a long chain of computer servants. In order
to clearly demonstrate their newly acquired status, and not to be unduly
disturbed by the end-users, the computer operators and systems programmers, headed by a manager of computer operations, took on white
coats, like doctors, and locked the door to the computer room, which now
became a so-called closed shop. The application programmers and the
data entry staff, to say nothing of the real users, who were now almost
forgotten, had to abide outside the closed door, waiting patiently for their
turn, which, with some luck, could come approximately once per day.
When this moment occurred, the end-user should have submitted a consistent and complete requirement specification, handed it over to an
application programmer, who then converted the specification into a
computer program in some user-friendly, high-level programming language, like COBOL5, after which a data entry operator converted the
COBOL program into a bunch of punched cards, and put it in a queue
outside the door to the closed shop. The compilation (translation) of the
COBOL program on the card deck into machine-readable code would be
4
5
Monolithic, from Greek “one stone”.
COmmon Business Oriented Language.
Sundgren & Steneskog
25
done by the computer during the following night, and in the morning the
application programmer would have to search through a heap of listings
outside the closed shop, in most cases just to find that he or she had made
some syntactic errors that had prevented the compiled code from being
executed. When the programmer had managed to eliminate all syntactic
errors, the computer would be able to run the program, but the results
would probably be wrong anyhow, because of some logical errors in the
program. After some further debugging, testing, and running, the enduser would finally get the results, possibly only a few weeks after the day
he or she had submitted the specification to the programmer.
The real users were separated from the computers by staff and technology.
Then there was a revolution. On-line terminals appeared which made it
possible for users outside the closed shop to be in direct contact with the
computer. Note that in this context the user was initially not the real user,
the end-user, but the application programmer, who was now also regarded
as a user, since he or she was sitting outside the closed shop. After some
further efforts, the application programmers learnt how to develop interactive applications for the end-users, who were then also able to communicate directly with the computer behind the locked door from their terminals. Note how well the words “end-user” and “terminal” describe the real
position of the users/customers who were supposed to be served by the
computer and the computer servants. It is an indication of a computer-centred world view or “Weltanschauung”.6
After yet another decade there was another revolution – and this time it
was a real revolution, since it implied a shift of power from the computer
servants in their white coats to the real users, the end-users. We refer to the
introduction and striking success of personal computers, or microcomputers as they were called by the technicians. Now it was clearly demonstrated what it could mean for development and progress, if a new technology becomes available for everyone. Many people start using the new
gadget like a toy. But even if only some small fraction of the usage of a
gadget leads to something useful, if only by chance, mass usage often
results in important and unexpected innovations. What the technical
designers of a new tool have intended is one thing, what the users do and
request may be something quite different. But the users may not have been
able to specify their needs and requests, before there existed something
that could possibly satisfy them, if only partially and imperfectly to begin
with.
6
The overall perspective from which one sees and interprets the world.
26
Exploring Patterns in Information Management
The personal computers enabled human beings to regain control over
computers and computerised data processing systems. They could once
again have immediate access to computers and computer-controlled
resources without any “white coat” intermediaries. When the personal
computers became connected with each other and with common databases,
new and more powerful forms of organised work became possible – with
individual persons and collectives of persons in control.
The important effects of the PC revolution were of organisational nature.
Power structures changed. This has not yet been fully understood and
accepted by some surviving “white coats”, who now and then try to get
their revenge. “It should not matter where the resources are, as long as they
are available as if they were at your finger-tips” is a common type of
argument in favour of a more centralised technical organisation. The
argument is seductive, and it might even be correct – if nothing would ever
go wrong. But things go wrong, and in some rare cases you may even have
good reasons to break the rules and do things in another way than according to centrally imposed procedures.
We will now analyse computer-supported human activities from the perspective of human beings, that is, we put the human in focus rather than
accepting the view of humans as an end-user to whom information system specialists pay lip services but still too often keep away from real
power.
Implications for Computerised Information Systems
We have made it clear in this article that information and information systems (proper) can only exist in human minds. However, many people, who
use the term “information system” are actually referring to, what we would
call here data processing systems, or possibly computer-supported or computerised information systems: information systems and data processing
systems in interaction.
Figure 6 illustrates a number of human beings (with information in their
minds) interacting with a shared data processing system. The data processing system supports the information processing capabilities of humans
in the sense that it stores and processes data that represent and can be
(re)interpreted into information in human minds. The storage and processing functions of a data processing system are proxies of memory and
information processing functions of human minds.
Sundgren & Steneskog
27
THE WORLD OF INFORMATION
INFORMATION
STORAGE AND
PROCESSING
INFORMATION
STORAGE AND
PROCESSING
INFORMATION
INFORMATION INFORMATION
STORAGE AND
STORAGE AND STORAGE AND
INFORMATION
PROCESSING
PROCESSING PROCESSING
STORAGE AND INFORMATION
PROCESSING STORAGE AND
PROCESSING
HUMAN INTERACTION WITH THE DATA PROCESSING SYSTEM
REALITY
DATA
PREPARATION:
CODING,
EDITING &
CORRECTION
OBSERVE
MEASURE
REGISTER
INPUT
DATA
STORE &
PROCESS
CLEAN DATA
RETRIEVE &
ANALYSE
DATABASE
PRESENT,
INTERPRET &
ACT
OUTPUT
DATA
THE WORLD OF DATA
Figure 6. Computer-supported or computerised information system: data processing system and information systems in interaction
Computers have certain advantages over the human brain, mainly by being
faster and more accurate in certain types of operations. Because of this,
data processing systems may be used for amplifying certain human capabilities. On the other hand, human beings have also certain advantages
over computers. For example, humans have creativity, imagination, and
intuition, and are capable of contextual thinking and unexpected associations.
Human beings, equipped with suitable, computerised data processing systems, can obviously achieve much more than human beings alone or in cooperation with each other. But these “bastard systems” are not without
problems. We have already discussed how important it is that the human
users of data processing systems are in full control, preferably without any
intermediaries. Only then can people feel that they have a really efficient
tool that fits into their hands, or rather their minds; the computerised system becomes a natural extension to the human mind.
28
Exploring Patterns in Information Management
Another class of problems have to do with the distinction between information and data that we have also discussed. Data in data processing systems are never perfect representations of anyone’s information. Nor are
they objective in the sense that they will automatically be interpreted in the
same way by different people. One remedy to this problem that we have
mentioned is metadata, “data about data”.
Metadata
Let us take a simple example to show the importance and roles of metadata. Consider a statistical table. We look into the table, and somewhere in
it we see a cell containing the figure 12345. This figure is data. Seen in
isolation it tells us nothing. However, if we know something about statistical tables, we know that there are labels in so-called stubs and headings
that briefly describe the intended contents of the cell. For example, we
may find out that 12345 is supposed to be “the average income during the
year 2002 of people living in city C”. This gives us an idea of the meaning
of the data in this table cell. But there are still many uncertainties. What is
meant by “income”? Is it income from regular work only, or does it
include extra incomes, income from capital, pensions, allowances, ...? And
who were the people living in C during 2002? Everyone living there some
time of the year, and if so, how has their income been made comparable
with the incomes of those who have lived in C for the whole year? And
how precise is the figure 12345? Is it based on data about all persons in C,
or is it based on a so-called sample survey, which implies a certain sampling error? Have the data been obtained by means of a questionnaire, and
if so, have all respondents understood the questions properly, and have
they returned complete data? Have they answered the questions truthfully,
or have they had reasons to overestimate or underestimate their incomes,
implying a so-called bias?
In order to be able to answer questions like these, we need metadata
together with the data. Metadata have a similar relation to data, as the
frame of reference (in the human mind) has to perception data entering the
human mind. Furthermore both data processing systems and humans
require metadata in order to be able to process data.
Metadata may have several roles. They may describe the (intended) meaning
of data, the precision of data, the origin of data, the format of data, etc. Very
often it is not enough to describe the data as such, information about the processes behind the data is also needed. Let us return to our income example
above. It makes a difference in many respects, if income data come from a
Sundgren & Steneskog
29
survey, where the respondents are anonymous, or whether they come from
an administrative system managed by a taxation authority. And we need to
know what efforts the data producer has made in order to check the quality
of the data and investigate suspicious data (possible errors).
Naturally metadata can never be perfect. They can never completely bridge
the gap between data and information, and they cannot ensure that different users of the data will interpret them in exactly the same way, even less
ensure that the data are interpreted in a “correct” way. But metadata can
reduce the discrepancies between different users’ interpretations and
improve the conditions for constructive communication between people,
without too many misunderstandings – provided of course that the persons
communicating want to understand each other.
It should also be noted that some tasks performed by humans and organisations are more demanding in terms of “information harmony” between
people than others. If we are conducting research, or if we are going to
make an extremely important decision with implications for many people
for a long time, we need to be much more rigorous in our communication
and information management than if we are engaged in a casual discussion
at a dinner party.
The scope of communication and co-operation must also be taken into
account. A prime minister speaking to voters with widely varying backgrounds and mental frames of reference has another problem than people
working and living together in a small organisation, e.g. a local company,
or even a household.
A collective of people who are working together need to share conceptual
frameworks and a communication language in order to co-operate efficiently and effectively. These frameworks and the terms of the language
and their meaning may be more or less unique for the organisation. The
more they are adapted to the task of the organisation, the more unique they
are and the less understandable they are for outsiders. Jargon within the
guild may be very effective for the insiders but excludes the outsiders. This
also strengthens the development of group feelings and of feeling of
belonging, but it may also induce destructive thinking: “it is us against
them; we are good, and they are stupid”.
Data Systems as Business Infrastructure
Information, information and data processing, and data systems are integral parts of a business. A lot of information processing in the daily life of
30
Exploring Patterns in Information Management
an organisation may not even be recognised as information processes, e.g.
informal communication during coffee-breaks and gossiping in the corridors. The more structured and formalised the information and the information processing is, the more likely it is that it is explicitly recognised and
handled by some kind of consciously designed and computer-supported
data system. Formalised information systems may or may not be more
efficient than informal exchange of information. For example, it is often
efficient for the organisation to make sure that important knowledge about
key processes are not dependent on the presence of individual persons, but
is well documented and easily available from a common knowledge base –
as we discussed in connection with “tacit knowledge”. On the other hand
“management by coffee-drinking” is often a more efficient way of influencing the employees of an organisation than written orders or formalised
information meetings.
We may look upon all the information and data systems of a business
together as a system, or network, where people are the main components,
processing information by their minds, possibly assisted by computerised
tools.
Computer-supported data systems nowadays cover most areas of a business. They are more advanced, better integrated, and easier to use than
they used to be in the past, and they are becoming an indispensable part of
the business infrastructure.
Figure 7 visualises the data systems infrastructure of a business. It indicates that the infrastructure consists of a network of loosely coupled business-internal and business-external data systems.
ANALYTICAL
SYSTEMS
OPERATIVE
SYSTEMS
OTHER
BUSINESSES
INTERNAL SYSTEMS
EXTERNAL SYSTEMS
OFFICE SYSTEMS
HOUSEHOLD
SYSTEMS
Figure 7. The data systems infrastructure of a business
GOVERNMENT
SYSTEMS
Sundgren & Steneskog
31
The operative systems are the traditional data system applications of a
company: order management, production control, inventory, customer
management, accounting, personnel. Today many of these applications
may be covered by a so-called enterprise system. The applications focus
on the management of individual objects and transactions like specific orders, customers, suppliers, employees etc. Since the data directly affect
individual cases of individual people and enterprises, it is important that
the data in these systems is correct and up-to-date. This kind of data and
information is called operative data and information, object-specific data
needed by basic operations in the business.
The analytical systems manage data needed in the evaluation of business
processes on different levels and in high-level planning and decisionmaking. This kind of data, analytical data, often consists of statistics and is
also called directive data. Analytical or directive data systems often rely on
data input from operative data systems. However, analytical data usually
need not be as precise and up-to-date as operative data; it is enough if it
has good statistical quality. The purpose of directive data is to improve the
quality of decisions.
Office systems, including personal systems, are systems that the employees
of an organisation use for managing their own daily work, including communication with others, inside and outside the organisation. Office systems
also facilitate for the employees to provide the administrative systems of
the organisation with necessary input. Examples of office systems are
word processors, spreadsheets, calendars, time accounting systems, knowledge and contents management systems, etc. The office systems may also
provide entries to other internal and external systems that are part of the
data and communication systems infrastructure. The organisation’s intranet is typically used for this purpose.
It is becoming more and more common for organisations to make external
systems more or less integrated parts of their own data systems infrastructure. For example, links to websites of other businesses and government
organisations may be of interest for business operations as well as administration (e.g. travel management) and knowledge retrieval. Web links are
examples of loose integration. The co-operation with suppliers, customers,
and other partners may benefit from stronger forms of integration, e.g. socalled extranets.
Household systems, or home systems, may also become part of the data
systems infrastructure of a business. Employees may sometimes work
from home and may need access to at least some parts of the data systems
32
Exploring Patterns in Information Management
infrastructure of the business. Furthermore, households and individuals
may also be customers, and as such they may prefer to maintain their relations with the business via the Internet (e.g. electronic banks and other ebusinesses).
Many tasks in business processes will require smooth interactions between
the different categories of data systems visualised in Figure 7. There is no
sharp borderline between what belongs to different systems.
One desirable feature of the data systems infrastructure of a business is
that the different systems that are part of the infrastructure should be easy
to integrate with each other. Furthermore, it should be easy to add other
internal and external data systems in the future. Thus the infrastructure
should be an open network of co-operating systems.
Operative and Directive Data Systems
Operative information is necessary for the basic operations of an organisation, “the daily business”. For example, a retailer must know the prices of
the products to be sold, a library must know who has borrowed their books,
an airline company must have a reservation system in order to be able to
book passengers, etc. Without all necessary operative information, a business will stop working more or less immediately. Directive information, on
the other hand, is used as a basis for non-routine, managerial decisions.7
Examples of such decisions are
• whether the company should invest in a new production plant, and if so,
where to locate it
• which products to focus on in a forth-coming marketing campaign
• whether information services offered on the company’s web-site should
be free of charge
• whether the company should put a bid on another company, and if so,
what price should be offered
Directive information is also used in the evaluation of business processes
on different levels as well as in research and development processes.
Directive data often appear in the form of statistics, presented in tables or
graphs. However, it should also be noted that a lot of information that is
7
Here “directive information” should be interpreted as “information that gives
direction or guidance”. It should not be mixed up with directives in the sense of
(e.g. military) orders or commands.
Sundgren & Steneskog
33
used for managerial decision-making is of a more or less informal nature,
based upon the decision-maker’s personal intuition.
Directive data are not necessary, strictly speaking, for the daily business of
an organisation. However, they are expected to improve the quality of
planning and decision-making. Directive data may be of critical importance for the survival of the business, especially if it operates in a competitive environment.
It is relatively easy to find out which operative data are needed by a certain
business. In principle you identify the information that is needed by the
basic business processes of the organisation. Once you have identified all
this information, there is no need to argue about whether data representing
this information are needed or not. They are needed. If you think they are
expensive to produce, you still cannot avoid it, but maybe you can find a
more efficient way to produce them.
The situation is quite different for directive data. In a technical sense,
many managerial decisions can be taken without any computer-produced
data at all. As an example, suppose your business is going to invest in a
new factory somewhere. You have a choice between two sites, A and B.
As a serious, rational manager, you will probably ask for a lot of data, before you make the decision. But why not toss a coin instead? It will save
you a lot of time and a lot of work as well.
There are at least two reasons why a typical western business manager
would collect and evaluate data to become informed before an important
decision is taken. The first reason is of course an ambition to be rational.
Many of us are convinced that a decision will be more rational, in the
sense that it will lead to results of higher quality, if we behave like “the
economic man”:
•
•
•
•
identify decision alternatives
collect data about the alternatives
evaluate the alternatives
choose the best alternative
The second reason is an ambition, typical for our culture, to be rational, or at
least pretend to be rational, even in situations where we have de facto already decided what to do. Even in such situations we often want to present
the decision as if it had been prepared according to the “economic man”
model mentioned above. In other cultures there may be opposite preferences
about how to present decisions. There may be a dictator who may want the
decision to look like an act of God. Nevertheless, a clever dictator in such a
34
Exploring Patterns in Information Management
culture may secretly collect and evaluate data and information in order to
“help” God (or an oracle) to come to “the right” decision.
Another example, which has been an object of debate, is whether experts
do better than monkeys on the stock market, that is, whether a data-based
placement strategy will beat a strategy based upon a random number generator (or a monkey’s random choices).
It is sometimes debated among business managers, which kinds of directive
data and information, and how much of it, would be optimal. Obviously it
takes time and resources to collect and process directive data, so there is a
balance to be struck between the costs and the benefits of such information.
One extreme view on this was expressed by the managing director of a
major Swedish bank, who stopped all production of management reports in
his organisation. The production of such reports would be resumed, only if
there were strong and well motivated requests for them. The same managing
director also claimed that budgets and prognoses are useless.
Typical tasks for operative and directive data systems are listed in Table 1.
OPERATIVE DATA SYSTEMS
DIRECTIVE DATA SYSTEMS
Automating or supporting manual and,
to a large extent, repetitive processes
Supporting planning and control processes, which are, to a great extent, of
a non-repetitive character
Supporting repetitive processes within a
function, e.g. personnel administration
Supporting decision-making ad hoc
Taking note of regular events (transactions, operative decisions)
Supporting non-routine strategic decisions
Supporting a business process initiated
by a customer until it is completed
Supporting research and development
activities
Table 1. Typical tasks for operative and directive data systems
Real world data systems often support a combination of operative and
directive tasks. For example, a personnel management system, or a customer management system, may support both routine and non-routine
decisions and actions. Another example is a banker, who is handling a loan
request from a customer; the banker may use a directive data system in
order to determine whether the request should be granted, and if the
request is granted, the banker may use an operative data system in order to
settle the details of the loan business between the bank and the customer.
In business processes where operative and directive tasks appear closely
together, it may be clarifying to analyse the operative and directive sub-
Sundgren & Steneskog
35
systems separately. Table 2 contrasts some typical properties of operative
and directive data systems.
OPERATIVE DATA SYSTEMS
DIRECTIVE DATA SYSTEMS
Users and usages known at systems
development time
Users and usages partially unknown at
systems development time
Provide data that is necessary for
operative processes; the information
must be provided despite costs
Provide data that improves the quality
of directive processes; there is a tradeoff between value and cost
Repetitive usage
Ad hoc usage
Data collection is well planned and
forms an integral part of the system
Combine available data from different
sources
Strong connections between collection
and use of data
Data are used for different purposes
than those originally intended
Users know the meaning and quality of
data relatively well
Metainformation has an important role:
information about definitions and quality
Table 2. Typical properties of operative and directive data systems
A directive data system should serve situations, which can only partially
be foreseen at systems development time. When a concrete, directive data
need becomes manifest, for example when a decision-maker is going to
make a concrete decision, there is seldom time to change the data system,
or even to collect new data. Thus the user must use existing systems and
existing data. On the other hand, in an operative data system the usage
situations are repetitive and can often be described with good precision at
system development time.
In an operative data system there are often close connections between collection and usage of data. An order receptionist, for example, adds new
data to the order management system in the same process as he or she uses
data from the same system. One good effect of such close connections
between data collection and usage is that the user will gain a good understanding of the meaning and quality of the data in the system, i.e. is able to
create good information.
In a directive data system the connections between collection and usage of
data are much weaker. Data often come from several other data systems,
and formalised, computerised data must often be combined with informal
data from other sources, including information from the user’s own memory and judgement. In order for the user to be able to interpret the meaning
and relevance of data that has been collected elsewhere and for other pur-
36
Exploring Patterns in Information Management
poses, the data must be accompanied by some kind of “quality declaration”, or metadata.
Requirements on directive data systems are not as precise and stable as
requirements on operative data systems. A directive data system must not
only be able to adapt to changes in transaction volumes and other technique-related changes; it is constantly confronted with new data and processing requirements. For such a system one can never “freeze” a requirement specification; on the contrary, the system must be planned for ever
on-going changes in user needs and business environment conditions.
Information and Data System Tools
ICT –information and communication technology – is constantly providing
us with new tools for supporting our management of information and data.
It started with the telegraph then was followed by the telephone, the radio,
the TV and very complex network technologies such as fibre and satellites.
For processing, storing and retrieving data we got mainframe computers
and personal computers. Presently, there is a flow of new products such as
mobile phones, hand-held computers and combinations thereof.
Our use of all these new possibilities allowsour information potential to
grow rapidly. People are communicating directly and indirectly, face-toface or not, in ways never seen before. Projects are run with participants in
different countries in more or less constant contact with each other and
sharing documents in real time. Ad hoc meetings and demonstrations are
quickly organised by established Internet communities. Football teams and
military forces are using more or less complex networks for communication and data management to co-ordinate the actions of the individual persons.
Conclusion
The present evolution of information and communication technology and
its applications creates new possibilities for people to work together and to
co-ordinate their activities in order to achieve personal and common goals.
To understand and possibly control these developments, it is important to
take an information and data view on the different types of enterprises we
work with: societies, markets, companies, groups of people, and so on.
People and computers are the nodes in such networks, where data are
flowing between the nodes, and information processing in people’s minds
Sundgren & Steneskog
37
is leading to concerted action towards common goals. As man is the main
component in these networks too much focus on data processing systems is
sub-optimal. To refer to man – the main component – as the “end-user” at
the “terminal” (the end of the real world – the data processing system) is a
dangerous misconception of the world of information and data.
A human-centred approach requires knowledge from a number of disciplines, primarily systems theory, data systems development, information
science, cognitive psychology, and philosophy. It is not enough to have a
deep but narrow competence in one area – a wider perspective is necessary
for the efficient management of information and data.
References
Berger, P.L. & Luckmann, T. (1966) The Social Construction of Reality: A Treatise in the Sociology of Knowledge, Anchor Books, Garden City, New York.
Flensburg, P. (1986) Personlig databehandling; introduktion, konsekvenser,
möjligheter (In English: Personal Computing; introduction, consequences,
possibilities), doctoral thesis, University of¨Lund, Lund, Sweden.
Langefors, B. (1995) Essays on Infology: Summing up and Planning for the
Future, Edited by Bo Dahlbom, Studentlitteratur, Lund, Sweden.
McLuhan, M. (1962) The Gutenberg galaxy: the making of typographic man,
Toronto University Press, Toronto, Canada.
Ogden, C.K. & Richards, I.A. (1956) The Meaning of Meaning: A study of the
Influence of Language upon Thought and of the Science of Symbols, Harcourt
Brace, New York.
Scott, W.R. (1998) Organizations: Rational, Natural, and Open Systems, 4th ed.,
Prentice Hall International, Upper Saddle River, New Jersey.
38
Exploring Patterns in Information Management
Blanksida
—3—
IT: An Ambiguous Technology?
Michael J. Earl
A Simple Challenge
During one of my early visits to the Stockholm School of Economics, Mats
Lundeberg invited me to conduct a seminar for his MBA and doctoral students on “managing information technology”. Somewhat unnerved by this
rather broad topic, I asked Mats what particular question I might address.
He replied in an impromptu sort of way – of course it was probably far
from impromptu – why don’t you tackle the question what is different or
special about IT?
Like many of Mats’ questions, this apparently simple challenge was quite
demanding. Once we had exhausted the coward’s (or perhaps humble) tactic of asking the students themselves that question, I had to suggest some
ideas.
IT is a high expenditure activity
IT is critical to many organisations
IT has become a strategic weapon
IT is needed by our economic context
IT is affecting all levels of management
IT may mean a revolution for management information systems
IT involves many stakeholders
IT matters do matter
IT management makes the difference
Table 1. Statements on IT in Business in 1989
In the late 1980s, for that was when the seminar took place, IT was perceived in business as an exciting, important and “can’t be dismissed any
more” technology; in some ways it was seen as a phenomenon – perhaps
not yet fully understood. So I tackled the question at a phenomenological
level, I drew on the opening chapter of my book of that time (Earl, 1989)
40
Exploring Patterns in Information Management
and proposed a descriptive list of attributes. These are reproduced in
Table 1 and we developed in the seminar the scale, implications and
management imperatives of each. It was in retrospect an exercise in
pragmatism.
Today, although IT has been deployed in business for over 50 years, the
question, “what is different about IT?” occasionally raises its head,
among sceptics as much as the curious, in both academe and management. I am not sure we have an adequate answer yet, but an essay like
this is perhaps the place to develop some thoughts. Since Mats Lundeberg challenged me with it – not in a sceptical way, but as a topic for
debate – perhaps he deserves over a decade later a more reflective
response. For Mats’ questions tend to be testing and instructive – and
trouble you for some time.
Some Existing Perspectives
In an underrated article, Curley and Pyburn (1982) distinguished between
industrial and intellectual technologies. The key learning challenge of the
former was how to use them – if you like the craft-like challenge of training. The key learning challenge of the latter was understanding what the
technology could do. Pyburn and Curley agreed that IT was of this sort.
This rings true in at least two ways.
The experiential learning or stage models of IT, for example of Nolan
(1973) and several derivative articles (for example Nolan, 1979) suggest
that in applying and managing IT, we tend to learn in an evolutionary way
by doing. We find it difficult to appreciate the scope of IT and to recognise
the management implications because IT is neither straightforward nor
simple. Perhaps this is because, as computer scientists have pointed out, it
is a general-purpose technology, where uses are not tightly prescribed.
Most industrial technologies – lathes, automobiles or drilling machines –
are specific or single purpose which are narrow in scope.
Sociologists for the last century, with some notable exceptions, have not
differentiated between technologies. In their concern over the impact of
technology on work and workers, they have argued over whether technology – usually industrial technologies – enhances or degrades jobs.
Braverman (1974) leads the perhaps more dominant degradation school,
where technology is seen to be deskilling and a means to allow employers
(and thus capital) to exert further control over labour. The opposing school
often based on workers’ attitudes and behaviour in the workplace tends to
Earl
41
argue that jobs are improved by technology, but more because it takes the
burden out of work rather than enriches it. Bright (1958) and Schrank
(1978) have argued this way.
Interestingly, however, as technologies have become more sophisticated,
or “intellectual”, some scholars, for example Blauner (1964), have suggested that work can be more satisfying in terms both of being less routinized and more sociable. Then when obviously information technologies
have been studied, further evidence to this effect is available. For example,
Zuboff (1988) in her landmark study of computer-based technologies in
both the factory and the office documented persuasive evidence that while
IT may displace physical effort and operational know-how, it also may
stimulate reskilling, in particular providing opportunities for workers to
deploy knowledge and more intellectual skills. Zuboff distinguished
between “automating” and “informating” work; the latter enabled development and use of “intellective” skills.1
Economists, like sociologists, often aggregate technologies, particularly in
searching for macro-level generalisations. Technology is seen as an
exogenous variable which may stimulate product or process innovation. It
is only when the black box of the firm, or industries, is opened that
endogenous processes of innovation, learning and adaptation are
addressed.
At the macro level, economists often are building on Schumpeter’s (1934)
theory of economic development, linking firms’ entrepreneurial behaviours and new paradigm technologies through processes of “creative
destruction”. The real rise of IT (i.e. the convergence of computing and
telecommunications) in the 1980s coincided with interest in longwave or
“kondratiev” cycles of economic activity due to technological discontinuity. This was in the Schumpeterian tradition and Freeman’s (1982) work in
particular concluded that economic restructuring does arise from such
cycles.
Interestingly, Nolan and Croson (1995) built on the concept of “creative
destruction” by recasting the “stages theory” of IT assimilation and nesting
it in Schumpeter’s original work to advance a six stage model of organisational transformation. In that the “stages theory” is premised on managerial
processes of organisational learning, we see links between both econo1
In my case study on Shorko Films SA and a subsequent paper based on this case
on Knowledge Management (Earl, 1994) I observed both the potential and
practice of “informating” in a manufacturing plant. Here there was clear
investment in “upskilling”.
42
Exploring Patterns in Information Management
mists’ and management scholars’ views of the challenge of new technologies and firms’ responses to them. Both have explicitly or implicitly
emphasised learning, once micro-level studies and analyses are embraced
(for example, Loveridge, 1990).
In the case of information or “intellectual” technologies, it is perhaps the
increased demands on learning that distinguish them from industrial or
automating technologies. And this is one reason why sociologists may
have discerned differences in their respective impacts on work. The learning challenge, to repeat, is both about how to use them and what they can
do. The “what” scope is what makes IT in neo-Schumpeterian terms a discontinuous technology.
In the last few years, another framing or “theory” has caught managers’
and strategy academics’ attention. Implying discontinuity, it is the concept of “disruptive technologies” (Christensen, 1997). Borne out of the
business schools’ management of technology subject area, it is perhaps a
conceptualisation about new technologies in general and attracted special
interest during dotcom mania because it helped explain or prescribe how
managements should respond to technologies which threatened existing
markets and business models. It shed light on questions about market
cannibalisation, timing of adoption and organisational responses. To use
Christensen’s term this “innovator’s dilemma” was both described and
analysed and undoubtedly the notion of disruptive technology advanced
our thinking.
The model has appealed to some observers of IT; however I have reservations about whether it either describes or explains what is special or different about information technologies. For Christensen, disruptive technologies may cover both “industrial” and “intellectual” technologies and he
probably did not set out to answer our question “what is different about
IT?” So to IT scholars, practitioners and managers I suggest that the “disruptive technology” model is instructive but it does not capture three critical characteristics of IT. With this in mind, I venture to propose another
label “ambiguous technology”.
Ambiguous Technology
The adjective “ambiguous” may not be perfect. Commonly it is used to
describe double meaning or doubtful classification, but also the Oxford
English Dictionary (and what other dictionary dare I use?!) suggests “of
uncertain issue”. It is this aspect of IT which should not be underestimated,
Earl
43
namely uncertainty. (“Uncertain Technology” therefore could be a preferred label, but it does sound rather prosaic.)
In Figure 1 I suggest that there are three essential uncertainties about IT;
enabling, commissioning and impact. These uncertainties may be stronger
in the case of new or emerging information technologies; and new or
emerging may be assessed relative to a technology’s adopting context as
much as to its arrival in the marketplace.
Enabling
Uncertainty
Impact
Uncertainty
Commissioning
Uncertainty
Figure 1: IT as “Ambiguous Technology”
Enabling Uncertainty is about scope: what can a particular information
technology do? This is the practical question that arises from IT being a
general-purpose technology. Inventors and developers of an information
technology may have a view, vision or theory about its application, but the
history of IT shows that it is quite normal for information technology forecasters to be short on prescience about the killer application or the timing
of their predicted revolutionary impact. For example, did Tim Berners Lee
foresee the amazing enabling scope of the world wide web? Or who predicted that SMS text messages would be a killer application of the mobile
‘phone? In short, a mix of imagination, economic need, and experiential
and accidental learning are as likely to help us discover what IT enables us
to do as are the claims, visions and design features that IT developers and
manufacturers promote.
Such enabling uncertainty has at least three managerial implications. If we
want or expect IT to enable new ways of doing business or discovery of
new business opportunities, we need to embrace experimental models of
strategy-making rather than just the more conventional and rather analytical, linear models which seek alignment with existing business strategies
and known business needs.
44
Exploring Patterns in Information Management
Second, we cannot leave this strategic exploration to technologists.
Experimentation and discovery has to include front-line managers and
users who suggest, try and assess new applications in diverse contexts.
Third, the performance metrics of such voyages of discovery are not concentrated on return on investment, savings and immediate benefit. They
should include what did we learn, what new ideas emerged and what
unanticipated benefits arose.
Commissioning uncertainty is about the obvious and most-emphasised anxiety of IT: will a particular technology work? Put more graphically, is this
latest technology another example of “vapourware” or “snake-oil”? Or more
managerially, is this at the risky end of “bleeding edge” where examples of
successful adoption so far are rare? We should not dismiss this uncertainty.
There are enough examples of software products that are withdrawn (after
all, the IT industry coined the unfortunate and paradoxical word “de-commit”), database techniques that over-promised and application systems that
are aborted because they are not reliable. To be sure there are “industrial
technologies” that soon get replaced by better versions, automobiles that get
recalls and drug products that are withdrawn. But it is the ex ante anxiety
about “will it work” that stands out about information technologies because
they are complex, often quite innovative, dependent upon human-created
software and sometimes difficult to test.
Three managerial implications arise from commissioning uncertainty. First
adoption of a new information technology may well have considerable
technical challenges (often called simply technology risk). If these are perceived as high, then we should demand that there is a strong business case
for adopting the technology – even if the agreed business case is that a
voyage of enabling or application discovery seems to make sense in that
the particular technology has the potential to be a “killer technology” for
our organisation.
Second, it pays to adopt risk- or uncertainty–reducing tactics. These include
importing someone with previously acquired know-how on making it work,
including personnel from the technology provider or vendor, in the team,
and creating incentives for the vendor to make it work.
Third, the important performance metric here is have we transferred and
captured the learning on how to make it work.
Impact Uncertainty is about impact and implementation. One obvious
question is whether users – employees, customers, citizens, other businesses… – will adopt and use a new technology or system. Another is
Earl
45
whether a system will work in its intended context. A vital one is the effect
on user behaviour, work practices, organisational decision-making and so
on.
At least three managerial implications arise here. First is the question does
the technology fit the context or do we have to adjust the context to the
priorities and potential of the technology. In both cases we have to recognise that we are dealing with socio-technical systems and without examination of social realities the system or technology is destined to fail.
Second, if uncertainty on this dimension is high, it is imperative that users
(and today this can be customers, consumers, allies etc) are involved in
specifying, designing and introducing the application. Indeed, this is the
domain of prototyping – in its true sense of live trialling in use (Earl,
1978).
Third, the performance metrics or evaluation schemas are clear. We have
to measure the operational and social outcomes as well as the technological and economic results. This is where we realise that “IT is more than
IT”.
Ambiguous Technology in Practice
Two different but contemporary IT themes provide a stage for assessing
this framework or model. In the early days of e-commerce (Earl and Khan,
2001), enabling uncertainty was often handled by embracing short horizon
rolling plans or strategies and new venture capital approaches to managing
projects. Commissioning uncertainty was embraced by accepting volatile
and disposable front-end layers of a three-tier architecture. Impact uncertainty was addressed by adopting launch and learn tactics and making lessons in use the top priority for further systems development. And overall,
multidisciplinary teams were assembled recognizing that e-commerce was
not just a technology play.
In contrast, enterprise resource planning (ERP) systems are less uncertain,
today, in terms of the enabling and commissioning dimensions. They do
not lack these uncertainties but today businesses are nowadays clearer
about their ERP goals and the commonly deployed application suites are
not leading edge. It is the impact uncertainty which really distinguishes
ERP projects, where questions such as process fit or process change, data
cleaning and data standards, local working practices and cultural diversity
and so on are raised. Implement ERP systems without examining and fac-
46
Exploring Patterns in Information Management
ing up to such social and operational realities and failure is just around the
corner.
So What?
The reason for developing this framework of ambiguous technology is that
most available alternative models do not capture the “uncertainty of issue”
that IT demonstrates in practice. Indeed, there is a tendency by IT vendors
and practitioners to be quite unambiguous in their rhetoric and over-certain
in their actions.
Those of us in academe have a duty to explain what is different or special,
if anything, about IT – and especially to be articulate about both the
promise and reality of IT.
Technology developers should not be daunted by the ambiguity framework, but they might avoid technological determinism in their pronouncements and recognise and embrace elements of ambiguity in at least the
three dimensions I have emphasised. Equally, they should retain their
excitement and enthusiasm about IT because the corollary of each ambiguity or uncertainty is that “you just never know”!
Those applying and managing IT might assess the “ambiguous technology” framework and ask do they assume too much certainty and knowledge when they formulate IT strategies, develop information systems and
evaluate their benefits. Or should they incorporate more experimentation,
more learning and even more uncertainty reduction in these activities?
Scholars who have conflated information technologies with industrial
technologies (plus those who have posited differences arising from the
intellectual content of the former) might consider whether ambiguity or
uncertainty are important differentiators. If so, there may be quite a
research agenda to work on.
And Mats Lundeberg should keep on asking questions, for the role of academics in what Mats often calls “our subject” – as elsewhere – is to ask
good questions as well as to seek good answers. Indeed, better questions
may lead to better answers.
References
Blauner, R. (1964) Alienation and Freedom: The Factory Worker and His Industry, University of Chicago Press, Chicago, Illinois.
Earl
47
Braverman, H. (1974) Labour and Monopoly Capital: The Degradation of Work
in the Twentieth Century, Monthly Review Press, New York.
Bright, J. (1958) Automation and Management, Harvard Business School Press,
Boston, Massachusetts.
Christensen, C.M. (1997) The Innovator’s Dilemma: When New Technologies
Cause Great Firms to Fail, Harvard Business School Press, Boston, Massachusetts.
Curley, K.F. & Pyburn, P.J. (1982) “‘Intellectual’ Technologies: The Key to
Improving White Collar Productivity”, Sloan Management Review, Vol. 24,
No. 1, pp. 31-39.
Earl, M.J. (1978) “Prototype Systems for Accounting, Information and Control”,
Accounting, Organizations and Society, Vol. 3, No. 2, pp. 161-170.
Earl, M.J. (1989) Management Strategies for Information Technology, Prentice
Hall, Hemel Hempstead, England.
Earl, M.J. (1994) “Knowledge as Strategy: Reflections on Skandia International
and Shorko Films”, in Ciborra, C. & Jelassi, T. (Eds.) Strategic Information
Systems – A European Perspective, John Wiley & Sons, Chicester, England.
Earl, M.J. & Khan, B. (2001) “E-Commerce is Changing The Face of IT”, Sloan
Management Review, Vol. 43, No. 1, pp. 64-72.
Freeman, C. (1982) Unemployment and Technical Innovation, Frances Pinter,
London.
Loveridge, R. (1990) “Incremental Innovations and Appropriate Learning Styles”,
in Loveridge, R. & Pitt, M. (Eds.) The Strategic Management of Technological
Innovation, John Wiley & Sons, Chichester, England.
Nolan, R.L. (1973) “Managing the Computer Resource: A Stage Hypothesis”,
Communications of the ACM, Vol. 16, No. 7, pp 399-405.
Nolan, R.L. (1979) “Managing the Crises in Data Processing”, Harvard Business
Review, Vol. 57, No. 2, pp. 115-126.
Nolan, R.L.& Croson, D.C. (1995) Creative Destruction: A Six-Stage Process for
Transforming the Organization, Harvard Business School Press, Boston, Massachusetts.
Schrank, R. (1978) Ten Thousand Working Days, the MIT Press, Cambridge,
Massachusetts.
Schumpeter, J. (1934) The Theory of Economic Development: An Inquiry into
Profits, Capital, Credit, Interest, and the Business Cycle, Harvard University
Press, Cambridge, Massachusetts (originally published 1911).
Zuboff, S. (1988) In the Age of the Smart Machine: The Future of Work and
Power, Basic Books Inc., New York.
48
Exploring Patterns in Information Management
Blanksida
—4—
The Paradox of Perfect Knowledge
Alexander Verrijn-Stuart
Introduction
“Gallia est omnis divisa in partes tres” is the famous opening line of Julius
Ceasar’s De Bello Gallico. As such, it often serves as a metaphor for wellordering a resume, a line of reasoning or a presentation. In his satirical
variation on this theme, Pierre Daninos (1954) expressed the variety of
views in the same geographical area by having his protagonist begin “La
France est divisé en 48 millions Français”1. The latter characterization
may be most appropriate for describing the abundance of approaches and
lack of consensus regarding the role of information and information systems in any context, world wide. Yet, in spite of the many – often petty –
differences of opinion, the international academic community has benefited hugely from the – always friendly and stimulating – exchanges in
such gremia as IFIP TC8 and its derivatives.
In this paper, I should indeed like to address three issues:
• the existence of ‘information systems’, as such,
• the value of ‘information’, as such,
• the persistence of ‘information’ such as we define it…
These thoughts are motivated by the apparent common belief that information systems are intrinsically beneficial and the even more common
misconception that refining them must lead to perfect knowledge. Alas,
few situations permit anything like it. Understanding evolves piecemeal,
by introspection, observation, comparison, debate.
A contribution to the information systems community at large, on the
occasion of Mats Lundeberg’s 60th anniversary, is offered with the greatest
of pleasure.
1
The quote was made from memory and may be incorrect.
50
Exploring Patterns in Information Management
Do ‘Information Systems’ Exist?2
The ‘existence’ of things in the ‘real world’ has been an entertaining topic
for philosophers through the ages. Intuitively, we feel that whatever manifests itself in our environment does exist, in a true sense. It may be argued
that a person’s perception of some phenomenon cannot be exactly identical
to that of someone else, but that solid ‘scientific’ observation should lay
the foundation for irrefutable ‘knowledge’ about it. In a way, such a statement already takes us out on slippery ground. Will everybody repeat every
experiment or do we just trust that our precursors have done a good job?
Are we going to read all relevant scientific papers so as to convince ourselves that the original studies have come up with the ‘truth’? Of course
not.
Even in the hard sciences we just trust that results obtained so far are consistent. We accept summaries and global explanations. We ‘believe’, as we
must, for even if we wanted to duplicate all those studies, we could not
possibly find the time to do so. However, as long as we adopt a critical
attitude we may carry on with confidence (meaning that we accept that
those equipped to delve deeper into the domains in question will continually query all previous findings and properly come up with improvements
of falsified theories).
Things are somewhat worse in the ‘softer’ life sciences, and very much
more so in the social sciences. There, the problems are compounded by the
inherent variability of phenomena, the impossibility of truly replicating
experiments and the fact that the observer is part of the domain under
investigation. Strictly speaking, that is also the case in physics, where the
influence of the observer gives rise to the uncertainty principle. However,
its recognition permits us to formulate ‘exact’ theories with precise statistical inference options. In that sense, the life and social sciences may be
considered just more difficult empirical research domains.
The deeper problem for all sciences is the metaphysical one of understanding and relating the vast amount of well researched material. Philosophically, that means what our findings are really about. But equally, we
have the practical problem how we can meaningfully represent things and
2
This section owes much to the author’s contacts in TC8 through the years, but
especially to the CRIS (Comparative Review of Information System methodologies) conferences in the 1980s and the FRISCO (Framework of Information System COncepts) task group in the 1990s. The influence of Langefors’ Theoretical
Analysis of Information Systems (Langefors, 1966/74) remains undeniable.
Verrijn-Stuart
51
use our knowledge. What about the world of abstraction, where we talk
about things rather than observe them? Where we reduce a multitude of
recognizable attributes by constructed summaries, groupings or qualitative
labels? Where the name of an author (say, Wittgenstein, Langefors, Lundeberg) immediately conjures up an entire view about a wide subject.
Where, more mundanely, the concept ‘stock’ is understood as representing
a quantity of identifiable goods. Where we count, relate and classify
things. We may be precise in our abstractions, but must admit that the
underlying models leave out a lot. We state that we have captured the
‘essential’, but our essence may not be that of someone else.
In short, whenever we represent things and communicate by exchanging
the resulting representations, we are in fact negotiating so as to arrive at a
common view. This is done subconsciously in everyday conversation and
generally accepted when the problem is addressed seriously. However, it
applies to all ‘information’ streams in society, be they ‘informal’ (as in
telephone or face-to-face conversations in the office) or ‘formal’ (as part of
the procedures of what we call ‘information systems’). No matter how
strictly defined the latter, there are good arguments to say that they are no
more than a small portion of what goes on in the organization, in society.
If the recognized systems actually constitute the tip of the informational
iceberg, what then is that iceberg? Hence, the rhetorical question: do
‘information systems’ exist?
A balanced system analysis should never start from the point of view of
what information is required. A better insight is obtained by asking for the
full characteristics or the organization as a whole. This is where the concept ‘system’ comes into play. The term system may be loosely defined as
a collection of elements that display coherence, either as components-andtheir-interactions (CI model) or states-and-transitions (ST model). In
either form, the definition remains typically subjective, in that what is covered by it depends on some personal choice. Although common usage has
made us react with confidence to statements regarding ‘systems’, it cannot
be said that you will recognize a system when you see one. A more useful
definition is:
System ::= perceived domain, with at least one ‘systemic property’
not possessed by any of its sub-domains, and seen as distinct from
its ‘environment’
This definition implies the cohesion (because it is a domain with a recognizable environment) of a number of components (because it is a domain
that may have sub-domains), strengthened by the special joint characteris-
52
Exploring Patterns in Information Management
tic (the systemic property or properties) that makes it stand out and causes
any sub-domain to be a mere component, even if it might have some more
restricted systemic characteristics itself. A traditional example of a system
is a motor vehicle with components, engine, body, wheels, etc. The engine
consists of a carter, cylinders, valves, a carburetor and so on. Often, it is
said to be a sub-system of the vehicle. However, the overall systemic
property of the latter is the transportation capability, whereas the engine
just provides the propulsion (which might have served in other contexts, as
well). The engine just adds something and only in this particular context.
’Environment’
’Organization’
Org.sub-system
Material / Informational flow
Machine/device
Figure 1. Intuitive description of an organization.
The same treatment may be applied to some ‘organization’, say an enterprise or a public service. As a whole, it is recognized because it has a name
and a number of qualities. Typically, the description does not make
explicit what constitutes the ‘environment’. Nor does one normally sum up
every component, action and interaction, in detail. However, a global
description as in Figure 1 is immediately recognized as a set of ‘organizational units’ that interact by messaging, passing physical streams between
each other and the environment, using devices for doing so. These devices
may be mechanical (sausage makers, packing machines, trucks) or com-
Verrijn-Stuart
53
munication equipment (telephones, faxes and computers). The streams
may be mechanical (sausage ingredients, packing material, finished tins) or
informational (orders, instructions, bills, delivery messages). The description may restrict itself to mainstream activity (acquisition, manufacturing,
sales), subsidiary (storage, transportation, administration, human
resources, ICT), or any further detail.
The full description evidently covers a system. Within it, however, there
are a number of less well defined sub-systems. The most important one is
the collectivity of all messaging, formal and informal, which may be called
the Information System in the Broader Sense (ISB). A large part of the
messaging is semi-formal, e.g. the contents of documents that flow through
the organization, the thrust of telephone conversations (and even the gossip
exchanged in the cafeteria). Semi-formal, in that the exchange is about
recognizable things and situations, trigger agreed actions and so on. Without them, the organization would not function as it does, but many of the
informational components and flows are not recognized in explicit protocols and procedures. More easily distinguishable are the computer-based
applications, which constitute setups that usually are referred to as ‘information systems’, but deserve to be viewed as something far more
restricted. These ought to be called Information Systems in the Narrower
sense (ISNs).
In this view, every organization has one overall ISB, within which one will
find a number of ISNs. Naturally, one desires the best imaginable supporting services from well designed, robust and maintainable computer
applications. However, one should be aware of the way in which they constitute a minor part of all information related activity. The ISB does not
manifest itself as a defined system – it very much depends on how it is
viewed in the organization. Furthermore, it does no more than representing
perceived reality: the ‘information system’ does not ‘exist’ in any true
sense…
Now, does this matter? Up to a point, it does not. However, catering for
the ever changing needs of the organization requires proper and flexible
embedding of computerized systems. It is no accident that Software
Requirements Engineering of the 1980s as a practice was succeeded by
Business Process Reengineering of the 1990s. Yet, most of the analysis for
this was implied in Mats Lundeberg’s ISAC methodology (as first suggested in his PhD thesis; Lundeberg, 1976). The ‘Analysis of Change’ is
the key. The consequences will be discussed below, in the context of
information persistence.
54
Exploring Patterns in Information Management
Meanwhile, let it suffice to reiterate that the ‘existence’ of information
systems ought to be treated with caution, even if their usefulness is indisputable.
Does Information have a ‘Value’?3
If one must avoid being too apodictic about the concept ‘information system’, this applies even more so regarding the underlying concepts. Any
participation in a multi-disciplinary study group will reveal fundamental
philosophical divides between participants with different academic backgrounds. When it comes to practical cooperation in the ICT field, this is
similarly fraught with conflict due to the hidden agendas of the interest
groups concerned, such as software engineers, project managers, information users and top management. The absence of a common line of reasoning and the varying interpretations of the key concepts ‘knowledge’ and
‘information’ are the main reason for this and, hence, the ineffectiveness of
billions of investment. Now, a useful characterization is:
• Knowledge is individual, but may be ‘shared’ in the sense of agreeing
(after negotiation) within a community.
• Information (in connection with ‘information systems’) should best be
viewed as any increment in (personal) knowledge, for decision-making,
reassurance, entertainment.
Given this approach, we can answer the question of the ‘value’ of information by a circuitous argument. When one asks what value should be
attached to ‘information’ in some specific context, the answer is almost
always positive, but without any true quantification. Analysis of ICT usage
usually turns around two aspects, (1) the cost savings achieved by computerised task performance and support, and (2) the greater ease of obtaining
relevant information. The first are only semi-quantifiable, the second
statement is probably valid, but utterly qualitative.
For an individual, the cost specification is limited to the purchase of a
PC, the service provider subscription and the telephone charges associated with internet access. The benefits are mostly vague – text processing
and email rank high; other advantages are mixed. The case of a business
enterprise is more complex. Although reasonably clear on the cost side
3
The origin of this section lies in lectures on ‘information quantification’ given at
Leiden University in the 1980s. This text is a transcription of a section in the latest
draft of the forthcoming Revised FRISCO Report (2003).
Verrijn-Stuart
55
(but see below), the revenue aspects are often matters of faith in the market it is engaged in. That last aspect is even harder to define for nonprofit bodies and government agencies. However, all may certainly be
viewed as ‘organizations’. In various degrees, they provide a suitably
representative range of study objects. Firstly, they do not constitute
homogeneous groups of cooperating persons, but diverse structures.
There is top management, line management, staff management and there
are individual workers. At each level there are different ‘information’
requirements and responsibilities (for action and decision, including a
variety of calls on ICT based support).
For organisations, the cost side is difficult to specify precisely. Obviously,
there is the hardware (computers and networks), and the system and application software. But both are subject to curious economics. For instance,
over what period should hardware be written off, given that its economic
life is much shorter than its technical one? When an attractive new model
is on offer? When desirable new capabilities become available? Even this
relatively tangible aspect is hard to quantify. The software case is harder
still, for current applications will continue to function without degradation
as long as the platform supports it. Other costs must also be recognised,
many of which are staff related. Examples are the training of non-specialists and the salaries of specialists (with the difficult choice “internal or
external?”). There is the cost of documentation (in-house libraries and help
desks, or reliance on external on-call support and specially ordered
research reports). There is also the cost of security (either by investing in
prevention or, upon errors and losses, in rebuilding one’s knowledge base).
And finally, there is the cost of communication, that is to say, of the formal
and informal message streams through the organizations, which always
existed, but are handled quite differently in a computerised setup. We shall
revert to these issues, but already note at this point that they apply equally
– in analogous form – to the case of the ‘individual’.
On the revenue side, few people (other than gurus and daring innovators)
will make explicit forecasts, except in cases of entirely ICT dependent
services. But even there, one faces the normal uncertainty whether the
market will take to a new product or not. However, any change in one’s
overall way of working – as a switch to or extension of ICT support is –
must be evaluated both on costs and benefits. Vague references to
improving the quality of the information flow may help sell new
approaches, but should be met with critical analysis. We shall demonstrate
that significant comparisons may be made. But first, we turn to the core
problem, namely that of the value of ‘information’.
56
Exploring Patterns in Information Management
Intuitively, information is linked to knowledge and more than that, it is
equated with an increment in knowledge. Value of information must therefore be associated with the advantage of better decision making. This reasoning may be applied to an individual person, the group to which that
person belongs (the ‘organisation’) as well as to a society as a whole.
Interestingly, one encounters cases with positive, negative and neutral
value correlations, as follows.
• Positive – Cases where providing knowledge to an individual
increases the value to society as a whole because others can continue
to benefit in equal measure as before (examples: a road map enables
route finding, the more people who have access to it, the better all can
find their ways; books containing useful knowledge; even books for
enjoyment)
• Negative – Cases where providing (disclosing) knowledge diminishes
someone’s ‘value’ – and possibly society’s as well; namely when that
person’s ability to act effectively is hampered or destroyed by inadvertent or malicious sharing of knowledge (examples: stolen access code
to secure location, losing control over an organisation’s special knowledge through industrial espionage, etc.)
• Mixed – Cases where a shift in value comes about; say privileged
knowledge is advantageous until it is shared (examples: private knowledge of a profitable chemical formula or manufacturing process;
advantage disappears when disclosed – however, value to society may
remain unchanged; similarly, natural beauty does not diminish in value
when knowledge about it is shared, up till some limit, when overcrowding spoils it for everybody…)
• Neutral – Copyrighted material is disclosed, but only available to others
against payment; economic value remains unchanged until lapse of
copyright
• Time-dependent – Within organisations, spreading ‘good practices’ or
other useful knowledge is to the benefit of the organisation as a whole,
until acquired by all members, when it no longer adds value. Time table
knowledge is useful – acquiring it constitutes information, but it will be
of no value when the train in question has already departed.
• Quantity-dependent – External knowledge may be of value when introduced into an organisation, but too much (‘overload’) may lead to confusion; likewise, making available knowledge to outsiders (PR, advertising, etc.) may be useful, unless done to excess.
Verrijn-Stuart
57
These examples illustrate why ‘information’ cannot be treated as an economic good in the traditional sense. Quantification must be done indirectly, by means of a relevant model. Rather misleading is the reference to
the term ‘information’ in connection with the DNA and RNA steering
mechanism determining what organic molecules will form or degrade,
which inspired the idea that the process uses them as instructions. More
appropriate is its use in the ‘Theory of communication’ originated by
Shannon. Encoded messages may constitute the results of a (quantitative)
experiment which (given sufficient scope and possible repetition) answers
the question what the state is of some area under observation – in that case
the outcome of the experiment helps resolve prior uncertainty – the theory
also looks at possible loss in the channel via which the package of symbols
passes (from sender to receiver) and provides measures for the number of
times a message may need to be repeated to achieve some degree of statistical assurance of correctness; because the number of encoded symbols of
the messages play the central role in the theory, all measures related to
them are quantifiable (although some only in a statistical sense); the traditional terminology for this kind of uncertainty-removing message quantification (entropy) has been adopted as denoting some ‘amount of information‘.
Similar to Shannon’s transmission of encoded messages, ‘data processing’
consists of recording symbols on some medium (paper tape and punched
cards when the terminology first arose, currently on various electronic and
solid state devices). The problems of storage capacity (memory), processing speeds (computer) and band width (transmission) are all quantitative;
thus one might speak of an ‘amount of data’, but only in a limited and
remote sense is that measure related to the meaning that is to be associated
with it. The popular saying ‘a picture says more that a thousand words’ is
no more than evident if one realises that 1000 words require about 6Kb of
representation and a modest JPEG file easily ten times that number – on
the other hand, a simple 1 Kb icon file may pack a very effective message.
Thus, the number of recorded symbols is a poor measure of the increase of
knowledge or even of the strength of a reminder.
In view of these three established and widespread uses of the term ‘information’ as a quasi-quantifiable concept, it is even more important to clarify the ephemeral nature of the situations in which it does have the deeper
meaning introduced above, namely that of increment of knowledge in the
context of individual and organisational decision making.
A proper model for ‘information quantification’ for an organisation would
detail the description of Figure 1 and contain the following elements:
58
Exploring Patterns in Information Management
• all actors (persons, departments, devices)
• all actands (material, informational)
• all activities (production, control, coordination, including external contacts)
• all physical streams (input, throughput, output, creation and consumption of physical goods)
• all message streams (input, throughput, output, creation and deletion of
data)
• all interrelationships between these, and
• all cost and benefit data related to the above (cash flows with origins
and destinations, etc.)
Obviously, such a complete description would be overly detailed for our
purpose, but it constitutes a basis for a thought experiment: one assumes
that an overall model is available (including the economic valuations) and
subsequently poses questions of what the result would be of extending or
reducing the model. For instance, if a particular coordinating activity is
partially based on forms and partially computer-based, what would be the
implications of automating further portions of the coordination task support? In other words, one performs marginal analyses on the model with
respect to local modifications.
In practice, the base model for a specific exercise would be a local model.
By projection, the relevant features with their marginal costs and benefits
should give an important indication of the economic consequences of any
proposed changes. The only additional requirement would be that one
investigates the prime coupling to the overall organisation, in so far as
applicable. It is in the latter that significant repercussions occur to which
one must be alert (e.g. expenditure in one department may lead to benefits
in another department, and vice-versa). Among the aspects deserving special attention are security matters (as in the first four cases mentioned
above); for ‘robustness’ of information systems a price must be paid!
This kind of analysis is not dissimilar to investment studies in cases of
factory expansion, new market entries and company mergers. The psychological weaknesses in those are the need for completeness of the number of elements to be considered (and general lack thereof), and the interpretation of factors that can only be judged qualitatively (i.e. optimistically
or pessimistically). The strength lies in the fact that one only needs to
specify a number of contributory factors rather than come up with one
Verrijn-Stuart
59
overall (probably dubious) value. Thus, the question of ‘the value of a particular piece of information‘ is replaced by a balanced analysis of the many
factors that may give rise to it.
The resulting knowledge may never be ‘perfect’, in the sense of providing
the basis for uniquely optimal decision-making, but we can at least associate a relative price with it.
A ‘Real-time’ Information System Language?4
Any specification is a description, preferably in precise terms, intended as
a basis for constructing some artefact. If the artefact is to be used in a ‘realworld’ environment, somehow its components must be selected from what
is available already, be it in material or abstract form, or constitute links to
or mappings from that environment.
The argument may be turned around. If one wishes to arrive at a specification to fit into that ‘environment’, what about just describing it, including
the desired artefact and then strip off the description of the environment
apart from the description of the artefact? This way, the specification
results from the complete situation after a kind of ‘projection’. Formally,
this means that one first describes an overall Business System (BS),
including all ‘informational’ activities, next projects it to the ISB level
(after suitable analysis and redefinition of those I-activities) and finally
projects it to the ISN level (again having analyzed and decided on the most
desirable computerized sub-systems):
BS | Projection(BS ISB) | ISB | Projection(ISB ISN) | ISN
Any language would suffice for these descriptions, but obviously one that
contains appropriate abstractions is preferable. From the descriptive point
of view, the classes GeneralUnit (external/internal unit), ResourceUnit
(actor/operand), Actions, Tasks would be all that one needs, given to a
number of relationships and restrictions. GeneralUnits and ResourceUnits
are ‘things’, Actions and Task constitute ‘activities’. Together, they combine to an overall concept ‘system’:
S = <GU, RU, AC, TK>
4
The ideas giving rise to these views result from research work done in Leiden in
the 1990s, culminating in Guus Ramackers’ Thesis (1994), a series of contributions to CAiSE, EJC, ISCO, WG8.1 and other conferences, and recently presented
at the Colloquium of the Institute for Logic Language and Computation (ILLC) at
the University of Amsterdam.
60
Exploring Patterns in Information Management
While this formulation looks very abstract, in fact, it is easy to implement
as a computerized ‘information planning’ system, where the various
activities are represented by net diagrams (for which the so-called Coloured Petri Net happens to be very useful). Moreover, each view on the
overall system may be represented by process, data and event models such
as those used in all IS design methodologies and currently unified by the
UML.
The essential point of this argument is that a properly chosen description
of an organization and its environment provides the means of performing
any continuous ‘Analysis of Change’ and, furthermore, implies all specifications for the appropriate computerization one might wish. Linking them
through an effective iCASE tool one might create all actual computerized
sub-systems at the same time.
In principle, this is precisely what is done by modern (usually proprietary)
development tools, but actually filling in all detail obviously remains a
considerable task. However, there are various attractive derivatives that are
worth considering. Firstly, the fact that one has described the organization
(or at least a sub-system of it) formally means that both the static and
dynamic features of the system are part the overall model, say, the data
structures, processes and events. These imply the way of working, such as
the input, output and storage formats as well as the triggers regarding user
intervention. Thus, if the description is captured by some modelling tool,
the so-called ‘system documentation’ is implied as well. Consequently,
that documentation, i.e. the user and operators manuals are – in principle –
capable being generated automatically. In this fashion, a description language might be defined which not only permits the formulation of expressions at BS, ISB and ISN level, but also the code and documentation for
each version, that is to say:
description of the organization ↔ computerized information systems
Now, an even more daring prospect offers itself. Would it be possible to
change the computerized systems ‘on-the-fly’, in other words:
∆ description ↔ ∆ information systems (dynamically, including
population)
If that were possible, expensive parallel runs of the updated system might be
avoided, general consistency would be achieved and past data would remain
available. The first evident obstacle is the potential need to change data
types and document formats. The problem is known in the areas of word
Verrijn-Stuart
61
processors (where previous formats are only usable up to a point) and image
processors (where change of colour depth or file format generally result in
loss of quality). However, given the same kind of reservations, a suggestion
is presented, that should be applicable at a high level of ‘information system
planning’, i.e. for maintaining the most up to date view of general information use (the ISB) and the system-level computerization (the ISNs, as such,
without their detailed specification and implementation). Just add one more
class to the system concept, CT (calendar time):
S = <GU, RU, AC, TK, CT>
where CT = { status, DT | status={def,undef}, DT=date-time } and all subsystems, structures and protocols are similarly ‘date-time-stamped’. Any
updating (‘Analysis of Change’ at ISB level and at ISN level) would be
conducted in the ‘undefined’ status, while the current use of the previous
defined-status version would continue. Once the desired new version has
been accepted, the status would change to defined and that new version
kept.
The advantage would be that during any later use of data from earlier versions flags might be shown, indicating potential incompatibilities or other
restrictions. A disadvantage would be that the entire system would grow
into one huge ‘historic’ database application. In theory usable for the
information requirements of an organization, in practice certainly feasible
for smaller off-line computerization and, as mentioned above, for ‘information system planning’.
Yes, ‘information’ might be made persistent, in that previously recorded
representations of knowledge may remain available longer than the life
cycle of the project in which it was collected, but this requires much care
and dedication. Alas, the recording media may not change as much as the
software systems that run on them, but new technologies will come about
relentlessly. After the clay tablet and the cave drawings, we invented papyrus and paper. The quill pen was replaced by the punch card, key board,
voice input, solid state devices and what not. Whereas hard media based
documents may often be preserved and available for study by historians,
digital data do not last much beyond the ‘next generation’ of equipment.
Changing ideas of how to keep statistical data may be annoying to those
wishing to incorporate those from previous periods; when printed, they can
always be accessed. It is alright to permanently store electronically
recorded data, but one must save a working copy of the appropriate playback device along with it. Now, there is a challenge for maintaining mankind’s knowledge!
62
Exploring Patterns in Information Management
Envoi
As an old acquaintance and friend of Mats Lundeberg, I would like to wish
him many years of enjoyable research, education, project work, participation in sometimes boring but often stimulating international committee and
working group activity. We first met when I was honoured to be the Faculty Opponent for his Doctoral Examination. The not yet polished propositions of 1976 became stimulating views to many of us. No mean
achievement.
All the best, even if perfect knowledge will forever remain a ‘contradictio
in terminis’, wisdom a beacon on the horizon, but friendship a persistent
intangible!
References
Daninos, P. (1954) Les carnets du major W. Marmaduke Thompson, Hachette,
Paris.
Langefors, B. (1966/74): Theoretical Analysis of Information Systems, 4th ed.
Studentlitteratur, Lund, Sweden and Auerbach, Philadelphia.
Lundeberg, M. (1976) Some Propositions Concerning Analysis and Design of
Information Systems, Doctoral dissertation, Trita-IBADB, No. 4080, Royal
Institute of Technology, Stockholm.
Ramackers, G.J. (1994) Integrated Object Modeling: An Executable Specification
Framework for Business Analysis and Information Systems Design, Thesis
Publishers, Amsterdam, The Netherlands.
—5—
Patterns of Change and Action:
A Socio-Pragmatic Perspective on
Organisational Change
Göran Goldkuhl
Introduction: Action and Change
Action means a change in the world. The notion of action implies that an
actor brings about some change. The world will be changed as a result of a
successful action. Actions performed in organisations imply changes, but
not all such actions can be called organisational change. This essay is an
investigation into this seeming paradox:
• All action implies change1
• Not all organisational action implies organisational change
My purpose is to investigate organisational change through an understanding of organisational actions. By looking at organisational actions,
patterns of change and non-change will emerge. Learning about patterns of
organisational change is a way of improving our capability to deal with
such changes. This is a belief that I share with Mats Lundeberg. “You can
improve your ability to handle change processes in business by learning to
recognize patterns” (Lundeberg, 1993, p. x).
Above I described action as change in the world. When we act we intervene in the world in order to change it in some way. Action means making a difference. For example when a firm manufactures goods, the
employees are acting upon some material in order to create valuable
products for customers. This is an interventionist view of action. This
view must be supplemented in several ways. We do not only act in order
to change something out there, in the external world. Very often we act
in order to change ourselves, for example to improve our knowledge.
We investigate the world directly or through mediating sources in order
1
The case of omissions will be commented below.
64
Exploring Patterns in Information Management
to learn more about the world. In these cases we intentionally try to
change ourselves; to improve our knowledge. Such action I distinguish
from interventionist action. I call such action interpretive and such
action will be performed with an inquiring purpose. These two types of
action (interventionist and interpretive) both imply change. From the
actor perspective interventionist action is intended to change his external
world.
Changes in the external world can be material changes or semiotic
changes. In the first case you do something with a clear material purpose;
e.g. chopping wood. You transform the wood into firewood. In the second
case you present signs, as when asking someone to chop firewood. It
would be possible to misinterpret the first type of action to be a non-social
action, while the other one is a social action. It is clear that the second type
of action – presenting signs – is a social action. It is aiming at a social
influence. The first type of action is primarily aiming at a material change.
Such material actions can however also be seen as social actions. If the
chopping is a response to a request for chopping (i.e. there are social
grounds for that action), and if the purpose thereby is to deliver some firewood to another person (i.e. there are social purposes), such action should
be seen as a social action2. Interventionist action can thus be a material
action or a communicative action.
An interventionist action is directed towards the external world and aims
at making external changes. It is however important to acknowledge that
such action is nearly impossible to perform successfully without simultaneously interpreting the external world. In order to make a proper intervention, there is a need for a prior apprehension of the situation (Mead,
1938). The direct intervention is usually performed together with a continual monitoring and awareness of the situation. The actor will in this
way learn about his own action through interpreting preconditions, performance and effects. Giddens (1984) speaks of this learning aspect as
the reflexivity of action. The actions have repercussions back on the
actor.
Interpretation is thus an integral part of interventionist action. It serves
intervention. It is however important to recognise that interpretive actions
can be performed on their own, without any parallel intervention. It is also
important to see that in some cases intervention is subordinate to interpre
2
This follows the analysis of Weber’s (1978) notion of social action made in
Goldkuhl (2001).
Goldkuhl
65
tation and inquiry. When you make an experiment you make some external
changes in order to investigate, observe and learn about the world. Intervention is in this case a means to observational ends. Intervention serves
interpretation.
An actor can purposefully reflect upon his own knowledge. This can be
done in order to articulate tacit knowledge, draw conclusions, construct
categories, arrive at new insights, and shape new ideas or other creative
and knowledge developing acts. This kind of internal transformation is
conceived as action when it is made with some deliberation and endeavour. I call it reflective action.
One more supplement can be made to the action notion. There is human
behaviour, which is not oriented towards change, but we still call it action.
In action theory, the human omission to act is also considered an action; an
omission act (von Wright, 1963). Not all human “non-behaviour” is
viewed as omission action. We call something an omission act only when
the actor had an apprehended possibility to act and he avoided making
such an interventionist action.
Characteris- Direction
tics
Action preconditions
Primary
change
(intentional)
Possible side
effects
Type of action
Interventionist action
Outward
External world
(towards
to influence
external world)
External influence (material
or social)
Reflexive
feedback
(internal
change)
Interpretive
action
Outward
External world
(towards
to observe
external world)
Internal
change
(improved
knowledge)
External
change of an
inquiry can
occur
Reflective
action
Inward
(towards
knowing)
Knowledge to
be reflected
upon
Internal
change
(improved
knowledge)
Omission
action
None
External world
to be left
unaffected
No intentional
change
Table 1. An action classification
Changes can
occur without
influence of
the actor
66
Exploring Patterns in Information Management
These four types of action3 can be seen as pure types (ideal types). Many
performed actions in real life can, as indicated above, be combinations of
these different types. The four types of action are described in a table
(Table 1) with the purpose of characterizing and comparing them.
In this essay I focus mainly on interventionist action but partly also on
interpretive and reflective action as preparatory actions for interventionist
action. My primary interest is action aiming at change.
Organisational Action
Organisational Action as Change in the World
I stated above that interventionist action in an organisation is oriented
towards change of the world through influencing material or communication. Only part of such action means organisational change. We change
something in the world but this does not mean that the phenomenon we
call organisation is changed.
Many ordinary business actions performed in an organisation are directed
towards creating value for the customers4. A main organisational purpose
is to make a difference to its customers. This socio-pragmatic view implies
also a view on organisations as actors. An organisation is a unity and with
a capability to act. It can however not act by itself. Human actors perform
actions in the name of the organisation. Humans act as representatives of
the organisation (Ahrne, 1994; Taylor & Van Emery, 2000; Goldkuhl &
Braf, 2001).
Let us look closer at different organisational actions directed towards the
customers. There are actions aimed at catching the customer’s attention
of the organisation’s capability and products. There are actions of offer,
using sales proposals to influence the customer to buy products. If the
3
This socio-pragmatic framework has been more thoroughly elaborated in other
publications; cf. e.g. Goldkuhl (2001; 2002), Goldkuhl & Röstlinger (2002) and
Goldkuhl & Ågerfalk (2002).
4
I will use commercial organisations as the prototype case when discussing
organisational change and action. I think that much of what I say may also be relevant for non-commercial settings. A consequence of using commercial organisations as prototypes is that I use the word “business” instead of more general terms,
like e.g. workpractices. Confer Goldkuhl & Röstlinger (2002) for an analysis of
the workpractice concept.
Goldkuhl
67
customers order products, there may be confirming actions with commitments to deliver a product to the customer. Such commitments need to be
fulfilled. The demanded products are produced and delivered. The customer will be exhorted to pay when presented an invoice. This description of business action follows a generic business logic; confer Goldkuhl
(1998).
All these actions aim at making changes; making difference in the world.
The organisation tries to make changes in customer’s attention, and tries
to influence the customer to buy products. When making a delivery
promise the relationship to the customer is changed; the organisation
commits itself to future actions towards the customer. The delivery
promise is expressed so the customer can count on product delivery.
When producing and delivering products (goods or services) changes are
made in the external world. Organisational action is about change; influencing and changing the world.
Making business means to a great extent coordinating actions between
supplier and customer as implied above. A supplier also needs to coordinate the different actions within the organisation itself. Different persons
from different functions in the firm must cooperate and coordinate their
different actions in order to create value to the customer. Such coordination means communicative action with the purpose of making the different
actions of different persons organisationally congruent.
Institutions Governing Organisational Action
There is a recurrent performance of the kind of business actions
described above. To be competitive in a market there is a need both to
use a minimum of resources and to adapt to customer needs and
demands. It is not possible or economically proper to invent new ways of
performing business on every occasion. There is a great power of repetition and routine. The infrastructure of the organisation is used over and
over again. The actors can perform the same types of actions over and
over again. Many actions will be of routine character. There will be
institutionalised ways of performing business. Of course there will be a
natural variation of actions within such a social institution. Different
problematic situations arise, which must be treated in ways deviating
from the normal way. Different customer preferences give demands for
modified action and results.
Organisational institutions explain routine and stability in an ever-changing world. If there were no organisational institutions there would not be
68
Exploring Patterns in Information Management
any recurrent typical organisational actions. I turn to the concept of institution, as a force of preserving stability and order in the organisation, in
my quest to understand organisational change. An institution describes and
prescribes the way things are done. Institutions describe what to do and
how to do it and sometimes also why to do it (Berger & Luckmann, 1967;
Giddens, 1984).
I do not claim that there are detailed rules for all kinds of organisational
actions. Many actions are only governed by vague knowledge and situational characteristics have a strong impact. There are also differences
between organisations. Some organisations have operations of a more
routine character, while other have high fluctuations in customer
demands and are very knowledge intensive. In such organisations the
institutions often tend to be weaker and the power of each individual is
greater.
The notion of social institution is well described by Berger & Luckmann
(1967). They describe how institutions arise through processes of habitualisation and typification. Habits are abstracted and typified to action patterns, which later on function as rules for conduct. When followed in
actions, institutions are continuously expressed, and thus reinforced. Institutions reside in inter-subjective knowledge about the social and material
world and how to act within it. Institutions have therefore a capacity to
preserve social order and stability. An institution is however dependent on
the actors’ recognition of it. If the actors change their collective conceptions, institutions will change accordingly. If actors change their ways of
conduct, institutions will change.
Several scholars use the notion of institution in order to describe and
explain organisational action; see e.g. March & Olsen (1989), Powell &
DiMaggio (1991) and Scott (1995). Organisational institutions are
described as collective and regulative knowledge governing and framing
organisational action. In doing this there is a stronger bond to history than
to the future (which is acknowledged by March & Olsen, 1989). The intersubjective knowledge basis is evident for institutions. Institutions exist and
proliferate through inter-subjective knowledge. But is this the whole picture? In a socio-pragmatic spirit I would like to adopt a more comprehensive view of organisational institutions. In order to do this I first turn to the
etymological origin of the word “institution”. It originates5 from the Latin
verb “instituere” (composed of in- + statuere) with the meaning of “set up”
and “establish”. An institution is thus something, which is set up with the
5
See e.g. Merriam Webster’s Collegiate® Dictionary; http://www.m-w.com.
Goldkuhl
69
purpose to give some stability. The word “establish” (having a similar
meaning as institute) has its origin in the Latin word “stabilis” meaning
stable. A socio-pragmatic (re-) interpretation of the institution concept
gives the following meaning: An institution is the result of institutionalising acts and it has the function of preserving stability in future actions; see
also Giddens (1984) about the duality of his prominent concept “structure”.
Carriers of Organisational Institutions
Scott (1995) uses the notion of carrier when describing institutions. This
notion seems to be a way to escape a too limited cognitive view of institutions6. There may be different carriers of an institution7. Inter-subjective
knowledge is one carrier of an institution, and this is an indispensable carrier. Without any knowledge (explicit or tacit) there would not be any
actions in compliance with the rules of the institution. There may, however, be other carriers of organisational institutions. Institutional knowledge may be expressed linguistically and recorded in documents. Such
documents will have functions of instructing and reminding people in the
organisation about the institution. When employees are uncertain about
their expected conduct they can inspect manuals and other documents in
order to obtain guidance.
Parts of institutions may also be manifested in material artefacts. Using
artefacts (like production technology, information technology) is not only
done with reference to economic considerations of replacing people with
equipment. To implement artefacts is also a way of enforcing designed
procedures on the organisation. A computer-based information system (IS)
is a good example of an externalised institution. Rules are programmed
into the artefact. The rules are followed when the artefact is executed and
used. This is not only the case with the automatic parts of the IS. Also
interactive8 parts of the IS, when the user and the IS interactively perform
6
I have borrowed the concept of institutional carrier from Scott (1995). I have
defined other carriers than Scott.
7
In Goldkuhl (2002) I have described the concept of multi-existing phenomena;
i.e. social phenomena which at the same exist in different realms of the world; for
example in cognitive, semiotic and material realms. An organisational institution
is such a typical multi-existing phenomenon.
8
Confer Goldkuhl & Ågerfalk (2002) about automatic vs. interactive use-situations of information systems. In line with Latour (1992) I give artefacts a prominent place on the organisational scene.
70
Exploring Patterns in Information Management
some actions, will have an enforcing power on the organisation to comply
with the institution. Artefacts will usually bring restrictions to the actors’
way of performing actions. The artefact, as an instrument, will not only
support human actions, but also direct and constrain the actions
(Engeström et al., 1999). Artefacts will have an institutional power on the
organisation.
There may be conflicts between the different institutional carriers, i.e.
between the knowledge of different actors and different recorded descriptions and different artefacts (Goldkuhl & Braf, 2001). Such conflicts and
incongruencies may be a source for organisational change (ibid.).
I define an organisational institution (as part of an organisation) in the
following way: An organisational institution comprises prescribed ways of
interpreting, conceptualising and conducting organisational work and thus
making such interpretation and conduct similar and congruent over time
and between actors. An institution is manifested in different carriers; i.e.
in inter-subjective knowledge of organisational actors, in documented
descriptions, instructions and assignments, and in material artefacts with
capabilities of performing or supporting actions.
Inter-subjective, practical
knowledge
Norms/rules/
assignments
Action
Material
instruments
Operating
instructions
Figure 1. Different carriers of an organisational institution affecting
organisational action
Figure 1 is an illustration of different carriers of an organisational institution and that these institutional carriers affect organisational action. It is
Goldkuhl
71
important to recognise that there is a social9 basis for all carriers; for the
cognitive, semiotic and material carriers.
An organisational institution (as inter-subjective knowledge) involves different types knowledge: For example categories, conceptions, values, preferences, role definitions, action rules, standards for action results. Institutional knowledge resides in both, what Giddens (1984) calls, practical and
discursive consciousnesses. An institution involves a meaning-universe
with both coherence and tension.
Organisational Change Actions
Organisation Change Actions vs. Normal Business Actions
Organisational life is not possible without routine, repetition and institution. But on the other hand organisations cannot survive if such institutionalised patterns do not change in line with changes in preferences and
demands from the environment.
Actions that are performed with a consequence of changing some institution of the organisation I will call an organisational change action. Most
actions of an organisation are not directed towards a change of institutions.
They are performed according to institutions and with the purpose of
making differences in business directed towards its customers. I call such
actions normal business (NB) actions. In my conceptual determination of
organisational change (OC) action above, I did not write “actions performed with the intention to produce changes in some institution”. I wrote
“actions that are performed with a consequence of changing some institution”. Of course many OC actions are performed with the primary purpose
of changing the organisation (its institutions).
Some actions in the organisation are thus intentionally oriented towards
changing other actions (the NB actions). The way to do this is to change
the institutions governing the NB actions (Figure 2). The domain of OC
action is other actions. The purpose is to change such actions. OC actions
aim at modifying, obliterating or creating new actions. OC action is about
9
The social character has not been made explicit in Figure 1. This illustration
should however be interpreted as an institutionally focused model of
organisational action derived from the more exhaustive model of social action
found in Goldkuhl & Röstlinger (2002 p 18).
72
Exploring Patterns in Information Management
composing other actions. Organisational change is action oriented towards
other action, thus action of second order. It can be called meta-action.
Institutions
New institutions
Organisational
change
Normal
business actions
New normal
business actions
Figure 2. Organisational change
Project-Based Organisational Change
Organisational change is often performed on a project basis. One creates a
separate arena for discussing and designing alternative ways of NB action.
This is a common approach to organisational change (Figure 3). In a project there is a clear distance to ordinary business. The project members are
gathered to reflect on the ways of performing business. A project arena
hinders them from being drowned in ordinary work, and this arena may
afford a mental possibility to reflect on the ordinary work. Through the
project work (the OC actions) new ways for NB actions are suggested. The
quality of such redefinition of NB work is dependent on
• knowledge of current praxis
• innovativeness in design
• competence in performing organisational change
Parts of the knowledge of the current situation can be tacit; i.e. be part of
“the practical consciousness” (Giddens, 1984). There may be a need to be
articulate and reconstruct such tacit knowledge; i.e. to make it part of discoursive consciousness.
If new ways of working are decided upon, then the NB actions are to be
modified according to these proposals. This is the problem of implementation of change, which is well known. The NB actors can be partly others or
totally others than those who designed the new principles for action. There
must be an organisational authority to claim the new way of working and
the NB actors must comply with this if new ways of action are to be established.
Goldkuhl
73
If this is the case, the NB actors will try out new ways of action. Different
proposals for NB action can have different levels of detail concerning prescriptions. Sometimes such proposals leave (intentionally or by accident)
much room for action design made by the actors themselves.
are
studied
Inquiry, reflection
and design
New institutions
Institutions
New proposals
Normal
business actions
implementation
as
Transformation and
re-institutionalisation
New normal
business actions
Figure 3. Project-based organisational change
Change of Organisational Institutions
If the actors are performing these new and different ways of action, the
institutions will gradually change. The new ways of action will be incorporated in the inter-subjective knowledge of the organisation. This can be
seen as a process of re-institutionalisation. If the proposed ways are
rejected there will be no new institution and the organisational change will
fail. An institution can be enforced on the organisation through the use of
artefacts. Material arrangements may compel certain behaviour. Artefacts
and written assignments and instructions will in many cases have a power
to create a modified conduct (cf. Latour, 1992) and new inter-subjective
knowledge may arise which is a foundation for institutions to survive.
An adaptation during change implementation and establishment will
probably make the institutionalised way of working (at least partly) different from the ways proposed by the project. There are difficulties in
designing all actions and action aspects on the sketch-board and this is
often not even desirable. People want degrees of freedom for their actions
and often dislike overly detailed prescriptions.
A change project must not have a strict separation and sequencing of, on
the one hand, reflection and design, and on the other, trying out and implementing new ways of action. Experimentation and testing (like proto-
74
Exploring Patterns in Information Management
typing) can be made in alternation and close cooperation with a more
abstract design. There are different change strategies for project work and
there must not be a strict linear way (design → implementation) as
described above. When experimenting with new ways the process of reinstitutionalisation will start at the same time.
Evolutionary Organisational Change
Not all organisational changes are performed in this intentional and
designing way using a separate project arena. All organisations change
gradually without explicit change projects. Not all issues are important
enough to emanate in a project.
Institutions not only arise from conscious design. They arise also from
evolution and habitualisation of action (Berger & Luckmann, 1967). People change their actions gradually. They adapt to new situations. When a
situation is conceived as problematic, this is a trigger to perform action in
an alternative non-standard way (Dewey, 1938). Such a new single action
will however not lead to organisational change if no other conditions exist.
If the new action is a response to a new demand or situation that is recurrent, this type of action will probably be repeated and then it will possibly
be habitualised. The way of dealing with such a situation must be deployed
to others in the organisation in order to be institutionalised.
Even if there is not a new challenge in a situation, a new conduct may arise.
An actor may discover better ways to respond to a common situation. In that
case the new type of action must prove to be successful. It must be considered successful by several organisational actors, who also must be prepared
to relearn. Institutions have great power on thinking and acting (Berger &
Luckmann, 1967; Giddens, 1984). Every time an NB action is performed in
the institutionalised way, this reinforces the institution and makes it still
stronger. “This is the way to perform business” (Figure 4). Institutions have
a sustaining power, which sometimes must be violated.
New ways of action will however be incorporated continuously in the
inter-subjective body of knowledge in the organisation (i.e. the institution).
New situations and more successful performances may give rise to institutional shifts. In such situations the organisational change is however not
the primary purpose of the action performed. There are NB actions performed in partly new ways. As a consequence (not a deliberate intention)
the institution is gradually changed (Figure 5). The new or modified
actions will have repercussions on the organisations such as gradually
modifying its institutions.
Goldkuhl
75
Institutions
governing
Unproblematic
situation
reinforcing
Normal
business actions
Figure 4. “Business as usual”
Usually one single person does not have the power to change an institution
governing the work of many persons. The new ways of action must be
distributed among the colleagues. This is often a process of mutual influence and adaptation. The process may not include verbal instructions. It
can be limited to imitation of others persons’ actions serving as exemplars.
Institutions
governing
Problematic
situation
(new demands
and ideas)
Possibly
changing
Normal
business actions
(new ways)
Figure 5. Evolutionary organisational change (work-integrated)
Continuous Improvement as Organisational Change
I have distinguished between two types of organisational changes: projectbased organisational change and organisational change performed directly
in the running business. Are there no other alternatives? Design vs. evolution should not be considered as disjunct categories. It could be seen as a
continuum, with several possible and identifiable categories. Between
project-based design and running adaptation a distinct category of change
can be identified. It is what many people call continuous improvement.
Continuous improvement (CI) is an integral part of the change approach of
Total Quality Management; see e.g. Rao et al. (1996). CI is not usually
made on a project basis. It is performed rather closely to the “production
arena”. After each execution of a business process, the staff within that
76
Exploring Patterns in Information Management
process should reflect on the process and try to improve it and its action
constituents.
Continuous improvement is not performed directly in a running business.
The actors take “one small step away” from the NB actions. They assess
what has been performed and try to improve it. This approach has resemblances to the project-based development since it involves reflection, conscious design and implementation of new ways of working. Such ways
must be institutionalised in order to be permanent. It differs from projectbased development since it is not performed within a separate change
organisation (project). It is performed in close connection with daily work.
In this sense it resembles running adaptation.
Is continuous improvement really performed continuously? Running
adaptation can be seen as a case of organisational change that is performed
continuously in the business whenever a need arises. I would like to contest that continuous improvement is performed continuously. A more
appropriate way to describe it is to say that it is performed recurrently. We
do not perform such changes all the time. It is rather performed recurrently
on certain occasions.
Design vs. Evolution
– A Typology of Organisational Change
The described change strategies will probably have different magnitudes of
change. Running adaptation will probably involve small changes. Continuous (recurrent) improvement can involve larger changes. Still farreaching changes can be obtained in project-based design.
Of course there can be different magnitudes of change even in different
project-based approaches. The Business Process Reengineering (BPR)
concept emphasises changes of great organisational impact (Hammer &
Champy, 1993; Davenport, 1993). Such a project should have an innovative nature. On the other hand, a common change project will probably
involve changes of more moderate scope.
The smallest change, made running directly in NB action (i.e. work integrated), I call adaptation. The next level I would rather like to call refinement than (continuous) improvement in my typology. I save “improvement” to the next level, which I call partial improvement. Thus continuous
improvement will be renamed recurrent refinement. The BPR case I call
radical renewal (or innovation).
Goldkuhl
77
This typology, with categories from evolution to design, involves thus the
following four categories (Figure 6):
•
•
•
•
running adaptation
recurrent refinement
partial improvement
radical renewal
Organisational
change
Change without
separate change
organisation
Running
adaptation
Recurrent
refinement
Project-based
change
Partial
improvement
Radical
renewal
Figure 6. A typology of organisational change
Davenport (1993) has made a similar division into continuous improvement, project-based improvement and radical innovation. These categories
seem to be equivalent to the three last categories in my taxonomy. Davenport does not ground his conceptual division in pragmatic theory. This is
perhaps one reason why he does not identify running adaptation as one
category of organisational change.
Organisational change is performed through the change of organisational
institutions. To call something an organisational change, and thus distinguishing it from normal business action (and change within such action),
there must be a change of the organisational pattern of action. All organisational action is aimed at change. Only those actions which have an orientation (directly or indirectly) towards change of other organisational
action are called organisational change. Continuous improvement/recurrent
refinement and project-based change have a clear intention and therefore a
direct orientation towards organisational change. Running adaptation is
performed within normal business action and therefore only has an indirect
78
Exploring Patterns in Information Management
orientation towards organisational change; i.e. the organisational change is
consequential rather than intentional.
References
Ahrne, G. (1994) Social organizations. Interaction inside, outside and between
organization, Sage, London.
Berger, P.L. & Luckmann, T. (1967) The social construction of reality, Doubleday
& Co, Garden City.
Davenport, T.H. (1993) Process innovation. Reengineering work through information technology, Harvard Business School Press, Boston, Massachusetts.
Dewey, J. (1938) Logic: The theory of inquiry, Henry Holt, New York.
Engeström, Y., Miettinen, R. & Punamäki, R-L. (Eds.) (1999) Perspectives on
activity theory, Cambridge University Press.
Giddens, A. (1984) The constitution of society. Outline of the theory of structuration, Polity Press, Cambridge.
Goldkuhl, G. (1998) The six phases of business processes – business communication and the exchange of value, Accepted to the 12th Biennial ITS conference
(ITS´98), Stockholm.
Goldkuhl, G. (2001) “Communicative vs material actions: Instrumentality, sociality and comprehensibility”, in Schoop, M., Taylor, J. (Eds.) (2001) Proceedings of the 6th International Workshop on the Language Action Perspective
(LAP2001), RWTH, Aachen.
Goldkuhl, G. (2002) “Anchoring scientific abstractions – ontological and linguistic determination following socio-instrumental pragmatism”, in Proceedings of
European Conference on Research Methods in Business, Reading.
Goldkuhl, G. & Braf, E. (2002) “Organisational Ability – constituents and congruencies”, in Coakes, E., Willis, D., Clarke, S. (Eds.) (2002) Knowledge Management in the SocioTechnical World, Springer, London.
Goldkuhl, G. & Röstlinger, A. (2002) “Towards an integral understanding of
organisations and information systems: Convergence of three theories”, in
Proc of the 5th International Workshop on Organisational Semiotics, Delft.
Goldkuhl, G. & Ågerfalk, P.J. (2002) “Actability: A way to understand information systems pragmatics”, in Liu, K. et al. (Eds.) (2002) Coordination and
Communication Using Signs: Studies in Organisational Semiotics – 2, Kluwer
Academic Publishers, Boston.
Hammer, M. & Champy, J. (1993) Reengineering the corporation. A manifesto for
business revolution, Nicholas Brealey, London.
Goldkuhl
79
Latour, B. (1992) “Technology is society made durable”, in Law (ed.) (1992) A
sociology of monsters: Essays on power, technology and domination, Routledge & Kegan Paul, London.
Lundeberg, M. (1993) Handling change process. A systems approach, Studentlitteratur, Lund, Sweden.
March, J.G. & Olsen, J.P. (1989) Rediscovering institutions. The organizational
basis of politics, Free Press, New York.
Mead, G.H. (1938) Philosophy of the act, The University of Chicago Press, Chicago, Illinois.
Powell, W.W. & DiMaggio, P.J. (Eds.) (1991) The new institutionalism in organisational analysis, University of Chicago Press, Chicago.
Rao, A., Carr, L.P., Dambolena, I., Kopp, R.J., Martin, J., Rafii, F. & Shlesinger,
P.F. (1996) Total Quality Management: A cross functional perspective, John
Wiley, New York.
Scott, W.R. (1995) Institutions and organizations, Sage, Thousand Oaks.
Taylor, J. & Van Every, E. (2000) The emergent organization. Communication at
its site and surface, Lawrence Erlbaum, London.
Weber, M. (1978) Economy and society, University of California Press, Berkeley.
Von Wright, G.H. (1963) Norm and action, Routledge & Kegan Paul, London.
80
Exploring Patterns in Information Management
Blanksida
PART TWO:
REFLECTIONS ON
IT-RELATED CHANGE
Blanksida
—6—
Change Work in Organisations:
Some Lessons Learned from
Information Systems Development
Anders G. Nilsson
Change Work in Organisations
In change work we have the ambition to improve or enhance different
activities within a specific situation or context. We can think of e.g.
changes in society, in organisations or in family life. In this case the focus
will be on change work in organisations; private companies as well as in
public services. Change work implies a purposeful growth and development of organisations. This development work can be performed by operating in networks (inter-organisational change) or accomplished by undertaking individual measures (intra-organisational change).
Business Development
We will use business development as an overall concept for change work
in organisational contexts. Business development generally consists of
different tasks which can be collected into some appropriate levels (Lundeberg, 1993). We can recognise three levels of development work in
practice with a distinct scope and focus (cf. Österle, 1995; Nilsson,
1999):
• Strategy development; focusing on corporate strategies for improving
the relationships between our company and the actors in the market
environment, e.g. customers, clients, suppliers and business partners
(cf. Ansoff, 1990; Porter, 1980; 1985).
• Operational development; focusing on how to make the business
operations more efficient within our company. The workflow between
different functions or processes in the organisation is designed in a new
and better way (cf. Davenport, 1993; Rummler and Brache, 1995).
84
Exploring Patterns in Information Management
• Information systems development; focusing on how support from
information systems (IS) can be useful resources and efficient enablers
for running the business operations more professionally and strengthening the competitive edge of our business achievements (cf. Avison
and Fitzgerald, 2003; Fitzgerald et al., 2002).
Information systems development is regarded as an essential part of business development. It should be in harmony with the efforts taken in strategy and operational development. In today’s business world, information
support has become a more integrated part of business operations and, in
many cases, a vital part of the business mission itself. In fact, the information systems can also create new business opportunities for companies to
reinforce their competitive edge in the market place.
There is not a need to work in a “top-down” fashion from strategy development through operational development down to information systems
development. In a real case we can start at a certain development level and
let the outcome of this work trigger some other levels upwards and downwards, often in several rounds. We can therefore regard the development
levels as essential inquiry areas during a whole change process in organisations. In many cases development of corporate strategies, business
operations and information support are often carried out as separate change
measures and as independent projects in organisations. The challenge is to
have a proper organisational co-ordination and timing between the three
development levels.
Information Systems Development
By information systems development we mean analysis, design and
implementation of useful IT artifacts to support some kind of business in
organisations (Orlikowski and Iacono, 2001). By IT artifacts we mean
the use of hardware and software solutions to improve the business
activities within and between organisations. The IT artifacts can be of a
varied character – for example we can create information systems in
organisations by using bespoke (tailor-made) software, application packages or component-based solutions. We are here focusing on computerbased systems for developing and changing the situation in concrete
business cases.
Research on information systems development (ISD) has its roots back in
the mid 1960s. Scandinavian researchers have had a great influence on the
evolution of information systems as an academic discipline (see Iivari and
Lyytinen, 1998). Personally, I had the privilege of being a member of the
Nilsson
85
Scandinavian school and tradition of information systems development
(Langefors, 1973; 1995). My main experiences are based on working with
the ISAC approach for requirements specifications (Lundeberg et al.,
1981), the SIV method for purchasing standard application packages (Nilsson, 1991; 2001) and the Business Modelling framework for studying
method combinations (Nilsson et al., 1999). After practising in the ISD
area for 30 years, as a researcher/teacher (for academia) and advisor/counsellor (to industry), I feel a great need to offer some reflections on
my findings.
Some Lessons Learned
My perception of the ISD field can be described in many different ways.
In this case, I have chosen to present my strongest impressions from
working with information systems development as lessons learned for
change work in organisations. This an attempt to explore essential patterns or fundamental principles for business development grounded in
theory and practice. The lessons learned are summarised in 10 points and
the order between them is approximately how they appeared to me over
the years.
Lesson 1: Proceed from User Needs
Change work in organisations can be seen as a social field of forces
between different interest groups or stakeholders such as general managers, business people and systems designers. There exist from time to time
communication gaps or misunderstandings when people from these interest groups try to deal with development matters (Nilsson et al., 1999).
Therefore it is important to find ways to bridge the communication gaps
between key actors during change work or business development.
From the ISD area we have learnt the lesson to proceed from the user
needs, requirements and terms during the development work. The simple
argument is that there are the real users (or business people) who in their
daily work should live with the proposed changes, e.g. new information
systems. The principle of user orientation goes back to professor Börje
Langefors’ infological approach to information systems development. The
theory of infology states the significance of designing and operating
information systems from a user point of view in order to achieve desired
results in organisations (Langefors, 1973; 1995). The ISAC approach for
systems development was strongly built on a user perspective for change
86
Exploring Patterns in Information Management
work (Lundeberg et al., 1981). Professor Mats Lundeberg has, later on in
his X model approach for change processes, suggested combining people
(user) and task issues in order to achieve a successful development work
(Lundeberg, 1993). The principle behind socio-technical design also
emphasises the importance of integrating people and technical matters
during systems work (Mumford, 1971). An interesting observation is that
the same reasoning lies behind the well-known success formula created by
Likert (1961, p. 212):
Degree of success in change work = f (Quality x Acceptance)
The success formula states that to attain a successful result, we must have
both sufficient quality in the designed solutions (e.g. the IT artifacts), and a
good acceptance among the co-workers (users) to give them a motivation
for using the solution. A low value in either quality or acceptance will lead
to an unsuccessful result – hence the multiplication sign in the formula!
Lesson 2: Apply Methods and Models
Change work in organisations usually implies a comprehensive and complex task in dealing with for the above mentioned stakeholders or interest
groups. We need to make many different decisions on a huge number of
issues during development processes. We need to take care of a variety of
users and their mental models (images) of business performance. Business
and systems development often involve people from different application
areas (such as production, marketing, accounting) and they have various
perceptions of the present situation – they may stress several desires and
requirements that can be overlapping or contradictory in character. We can
also have communication problems between business people (users) and
systems people (designers) during the development work. How can we
professionally handle such problems?
One lesson learned from the ISD field is that development work is performed more efficiently with support from formal methods and models. In
this sense a specific method (or approach) can be a useful tool in creating a
common language between general managers, business people and designers. Another basic principle in the infological approach to systems development postulates that applying methods and models is a very sharp way
of representing users’ needs and requirements (Langefors, 1995). The
ISAC approach was one of the first methods in the world that tried to show
a systematic method from problem capturing (so-called change analysis)
down to creation of technical solutions (data/program design). See Lundeberg et al. (1981).
Nilsson
87
By method, we mean concrete guidelines or prescriptions for a systematic
way of working with development tasks in organisations. It is possible to
distinguish between three main constituents of a method (Nilsson, 1995):
Perspectives (basic principles and assumptions), Work Model (steps and
documentation) and Interest Group Model (stakeholders and collaboration
forms). There has been much debate over the years about the actual effects
of method use in practice. Below we summarise some essential needs for
applying methods and models to support development work in organisations:
• Requirements specifications; the ultimate need to produce an exact,
consistent and complete requirements specification for designing the
future business and information operations.
• Explain IT possibilities; the need for explaining how new IT possibilities (e.g. e-business, Internet) can enhance business operations and
sharpen corporate strategies.
• Describing business flow; the need for describing and coordinating the
complex nature of material flow, service flow, information flow and
cash flow in organisations.
Applying methods and models for a systematic way of working with
development issues should be adapted to the special situation for change in
organisations. There is a trend today to use tool-boxes or tool-kits consisting of a various selection of methods for different situations (Nilsson,
1999). We have not yet found a “super methodology” in order to attack or
handle all possible development tasks during change work. Therefore it is
important to try to combine separate methods from different fields like
accounting (Samuelson, 1980), service marketing (Edvardsson et al.,
2000), strategic management (Kaplan and Norton, 1996) together with our
ISD methods (Avison and Fitzgerald, 2003; Andersen, 1994).
Lesson 3: Consider Different Perspectives
Change work in organisations will give better results if we turn the
descriptions or models of the business operations over in our minds. By
changing perspective and observing our operations from many different
angles, we will gain a deeper understanding of the underlying mechanisms
in the organisation. Thus by doing so, we will have a more solid base from
which to suggest vigorous changes in the business operations. It is therefore important to consider different perspectives when we try to understand and change the business situation in our organisations. Lundeberg
88
Exploring Patterns in Information Management
(1993) proposes a combined perspective approach to manage change processes in business.
Another lesson learned within the ISD area is that methods for systems
work should emphasise different perspectives or aspects when describing
business operations and their supporting information systems (cf. Olle et
al., 1991; Sowa and Zachman, 1992). Requirements specification is an
instrument for accurate descriptions of the contributions and effects a
specific IT artifact needs to provide to the business operations. The
specification should illustrate different users’ demands on the new business and systems solutions. One problem with requirements specifications has been that they are one-dimensional in character, illuminating
only a very limited perspective of business and information modelling. In
development work we need to describe business and information operations from many different perspectives, such as (Nilsson, 1995; 2001):
•
•
•
•
Intentions, concerning goals, visions, problems and strengths, etc.
Activities, concerning functions, processes, and tasks, etc.
Resources, concerning data, concepts, components and objects, etc.
Behaviours, concerning events, rules, actors, and force fields, etc.
In the early days of the ISD field, different method schools were in competition with each other. Each of them represented only one of the four
above mentioned perspectives. Today we more and more seek suitable
combinations for development work. In real life projects these four basic
aspects need to complement each other. Hence we need an appropriate
mixture of perspectives depending on the situation at hand (cf. Yourdon,
1993; Jacobson et al., 1994; Booch et al., 1999).
Lesson 4: Understand the Current State
– Change to a Better Situation
Change work in organisations is often performed in parallel with the dayto-day operations in a going concern. An understanding of the current state
of business gives a more stable base to create realistic changes for achieving a better situation in the future. Lundeberg (1993) proposes to use time
frames when working with change processes in business. He distinguishes
between observing the organisation from a past situation, to a present
situation and into a future situation. In practice a historical review of earlier situations or milestones when developing organisations is seldom
done. At best we take care of the past experiences when analysing the present situation. A classical problem in change work in organisations is
illustrated by Figure 1 and explained below:
Nilsson
Present situation
89
Next situation
Problem analysis
Problems
Measures
Strong points
Better points
Strength analysis
Figure 1. From present situation to next situation.
We are often focused, in change work, on practical problem solving or
“troubleshooting” when it can be better to look for strong points in our
business situation. We are educated to make a professional analysis of the
problem complex and then to find clever business solutions or smart
development measures. But in many cases the problematic situation may
represent only 25% of our current state of business. The mistake is that we
don’t focus on strengths and opportunities to make an even better situation
for future operations. The strong points should be taken care of and sharpened and improved before entering the next situation.
A lesson learned from the ISD field is the importance of starting the development work with a careful pre-study to identify problems and strengths in
the present situation as a platform for suggesting appropriate development
measures – it could be information systems solutions or other kinds of business changes. The ISAC approach became an image or ideal type for other
ISD methods as regards separating the development work into two main
phases: first a change analysis stage before starting up an information systems project (Lundeberg et al., 1981). The business process management
movement of today also focuses on different states or situations (cf.
Rummler and Brache, 1995). These approaches separate between making
process models for the “is-state” and the “should-state” for our business. By
making “is-models” you are able to understand the organisation in its current
situation before making “should-models”, in order to be competent to
change to a better situation (Tolis and Nilsson, 1996).
Lesson 5: Use Appropriate Enablers
Change work in organisations can be driven by using different opportunities, or limited by various constraints. Human and technology factors can
be both opportunities and constraints for change, depending on the situation. Davenport (1993) uses the term “enablers” for possible drive forces
for change. The human factors represent the capabilities offered by
90
Exploring Patterns in Information Management
knowledge, skill and motivation of the co-workers in the whole organisation. The technology factors represent the capabilities offered by human
information, computer solutions, software applications, telecommunications, etc. Personal knowledge development and technological innovations
can be good enablers when changing the business situation in organisations. Ploom (1988) describes a change model where an organisation goes
through three consecutive phases: (1) the efficiency phase, (2) the integration phase and (3) the transformation phase. The phase of transformation
represents the most challenging process where we use different enabling
factors to obtain a strategic position for the company in the value chain on
the market.
One lesson learned from the ISD area is that systems development efforts
have changed focus over the years from designing information systems in
order to support business operations (resource approach), to a position
where we design information in our computer-based systems to create new
business opportunities for the organisation, and hence strengthen the competitive edge on the market (enabling approach). In the first approach the
information systems are regarded as resources in change work. Starting
with the needs of the users, a business operations specification is made
which provides both content and structure requirements on the information
systems. In the second approach the information systems are regarded as
enablers for change. Here the focus is on the potential that a new information system represents for the organisation (see Nilsson and Pettersson,
2001). The information system becomes an enabler for renewing the business. New technological innovations in multimedia (cf. Packer and Jordan,
2001), the Internet and electronic commerce (Earl and Khan, 2001) have
become new value-adding enablers to the business.
Lesson 6: Time is Critical for Change
Change work in organisations is often restricted by time in one sense or
another. In other words, time is a critical factor for handling change processes. As mentioned above (under Lesson 4), Lundeberg (1993) proposes
to use time frames when working with change processes in business, i.e.
distinguishing between past situations, present situations and future situations. During change work in practice we have also to balance between
quality and temporal issues in order to gain successful results in time and
budget. In change projects we have to deliver acceptable results at agreed
deadlines. An interesting observation is that time can be perceived somewhat differently by various stakeholders. For example, there has been
noticed a phenomenon in practice that can be framed as “threshold levels”.
Nilsson
91
This means that a user demands some specific messages from a reporting
system or a data retrieval system at a certain point in time – in this case
he/she wants this information neither earlier nor later in time.
In the ISD field we have learnt the lesson that time is very important when
developing information systems. It is not an understatement to say that
Langefors (1973) in his theoretical and empirical work has “reinvented”
the significance of the time concept for successful systems design. In the
infological approach to information systems development we can find
three circumstances where time has to be considered in a clear and explicit
way (Langefors, 1995). Firstly, we have to strive for executive optimal
solutions or sufficiently good information systems with regard to the user
needs, together with time and cost limits for implementation. It is in this
sense a trade-off between information needs and time restrictions. Secondly, we have to consider the infological equation where the time component is essential for a user to be able to interpret personal information
from a given set of data. The infological equation states: I = i (D, S, t)
where “I” is the information conveyed, “i” is the interpretation process,
“D” is the data at hand, “S” is the pre-knowledge, frame of reference or
mental structure of the user, and “t” the time required or available for the
process. When a user needs more time for interpretation it could mean
losses in efficiency. Thirdly, we have to consider how messages should be
designed for a better understanding and communication. An elementary
message (e-message) represents the smallest information unit in a system
and is defined as the following triplet: object, time, property. It says that
such an elementary message describes a property (e.g. price) for a specific
object (e.g. article) at certain time (e.g. year-month-day). The reasoning
behind this principle for systems design is that it is urgent with time
stamps for messages in order to avoid confusion in operating future information systems.
Lesson 7: Pay Attention to a Life Cycle Philosophy
Change work in organisations goes through a life cycle with sequential,
parallel and/or iterative phases. It is the same way with change processes
as with e.g. product and market development processes. A life cycle can be
partitioned in a number of phases or areas. According to Nilsson (2003),
on a crude level a development process can consist of phases for change
analysis (with enterprise models), formulation (of requirements specification), implementation (of business solution) and after some time assessment (review of business operations). These phases or areas focus on different kinds of problems and demand various bodies of knowledge and
92
Exploring Patterns in Information Management
competence. What pattern lies behind a life cycle philosophy? Development work can be seen as a form of decision-making activity. Simon
(1965) states that all kinds of decision-making go through three phases:
intelligence (I), design (D) and choice (C). When we come to the situation
to carry out or execute a decision it is according to Simon again a decisionmaking activity (with its own IDC triplet). Lundeberg (1993) describes a
general model for change processes (based on IDC) comprised of three
recurrent and overlapping phases: planning (goals), operation (activities)
and evaluation (evidence).
What we have learnt as a lesson from the ISD area is that it is fruitful to
consider a system’s life cycle consisting of phases for acquisition, use,
maintenance and phasing-out. Strictly speaking, by information systems
development we mean the acquisition phase including steps for analysis,
design and implementation of IT artifacts (cf. Andersen, 1994; Hawryszkiewycz, 2001). The life cycle for creating and managing information
systems in organisations has over the years shown to be an essential and
valid concept, and therefore it forms an important basis for construction of
methods for systems work. In 1967, professor Langefors worked out and
presented an original proposal for partitioning of the system’s life cycle.
The result was four classical problem areas which have had a great impact
on subsequent development of ISD methods and approaches (Langefors,
1974): (1) object system analysis and design, (2) information analysis, (3)
data system architecture and construction, and (4) realisation, implementation and operation. The two first areas treat infological or user-oriented
problems, while the two last areas treat datalogical or technical problems.
The traditional ISAC approach was built on these four classical problem
areas within information systems development (Lundeberg et al., 1981).
Lesson 8: Reuse Successful Solutions
Change work in organisations should be more effective if we can gain
access to past experiences formalised in e.g. best practice models, application templates and/or standardised solutions. These represent generalised
experiences of a certain business or application domain. As a concrete
example we can mention the RP model as a framework for designing
accounting information systems (Samuelson, 1980). Experience of good
ideas and strong points from successful business cases should be taken
care of when designing for future situations. A possible alternative is to
reuse knowledge by purchasing requirements specifications from outside
instead of acquiring ready-made or pre-specified solutions. In this case, we
build our competence for carrying out development work on a higher level
Nilsson
93
of abstraction – established and proven knowledge – rather than on fixed
or “quick and dirty” attitudes. During change work we can use or create
solutions with different degrees of “pre-specification”. By this we mean
how complete specifications we have in advance when the development
work starts (Nilsson, 1995). A development situation which gives us an
opportunity to gain a higher degree of pre-specification facilitates the possibilities for reuse of successful solutions!
From the ISD field we have learnt a lesson to use standard application
systems (Nilsson, 1991; 2001) and ERP-systems or enterprise systems
(Davenport, 1998; 2000) as an efficient way of reusing successful solutions. The degree of success for implementation of such kind of readymade software as IT artifacts in organisations depends on how well-prepared the managers and users are for this new business challenge. A careful vendor assessment is an important work task during the acquisition
process. Obtaining new package releases from the vendor are critical
issues for the work with maintenance management. Another essential trend
in the ISD field is the phenomenon labelled object-oriented or componentbased systems development (Jacobson et al., 1993). Business objects or
components here represent application parts in miniature. This way of
working with objects/components is an approach for reuse in a small scale
compared to the situation with standard application systems which is
reusing solutions in a larger scale. Today, traditional ERP vendors try to
renovate and reconstruct their old packaged software using object or component development techniques as a competitive weapon.
Lesson 9: Discover Reality by Prototyping
Change work in organisations should in practice be carried out with the
help of some systematic model of planned activities (see Nilsson, 1995).
The initial change models for development work, presented in the market,
were sequential in nature. They were often labelled “waterfall models”
meaning that a certain phase must be finished before the next phase can
start. But change work is seldom strictly sequential or linear in character.
Therefore new change models were presented in the market proposing that
development work should be performed in a number of partly overlapping
phases. They were labelled “sliced models” meaning that certain phases
can be done in parallel. But change work in practice is not so often purely
sequential and/or parallel in character. Therefore new change models were
presented again in the market proposing that development work can be
iterative in nature. They were labelled “prototyping models”, meaning that
certain phases should be recurrent as new knowledge appears during later
94
Exploring Patterns in Information Management
work. By making rapid prototypes of a desired future business situation in
reality or every-day life, the various interest groups have the possibility to
discover the effects of introducing new types of solutions (e.g. IT artifacts). The stakeholders can react to the prototype solutions and give valuable feedback for further specification of the new business situation. A
prototype gives a concrete picture of a business solution and implies a rich
learning environment for the managers and users who participate in and
run the change processes.
Already at an early stage in the evolution of the ISD area we learned a lesson that users need to experience a prototype or realistic systems sketch
before they could describe the exact requirements on a new information
system (Bally et al., 1977). Prototypes can be designed in various forms,
all the way from simple “tear and wear” solutions to more advanced pilot
systems expandable to “full-scale” solutions (cf. Buddy et al., 1992). The
argument for prototyping in ISD work is that an IT artifact is perceived as
a rather complex phenomenon. We can not therefore always plan for new
information systems in a strictly analytical way (in the model world) but
we need also to do some practical experiments (in the real world) gradually. In analytical systems development the requirements specification
from the users needs to be complete and “frozen” before an implementation of the information system may begin. In experimental systems development (prototyping) there is an interplay between the work with specification and implementation, sometimes in several rounds. The prototyping
approach has not always produced the desired effects in organisations,
since we seldom are capable of supplementing the experimental work with
a solid evaluation phase. Close to the ISD field is the multimedia area.
When developing multimedia systems, the prototyping approach gives the
stakeholders a deeper understanding of how the interactive media product
would work in practice (cf. Elin, 2001; Packer and Jordan, 2001).
Lesson 10: Promote Business in Manageable Steps
Change work in organisations means that we are advancing the business
towards some concrete visions or goals (cf. Lundeberg and Sundgren,
1996). There are many different types of change programs in practice.
Business process reengineering (BPR) implies work with radical changes
to achieve dramatic improvements in business performance (see e.g. Davenport, 1993). In other words, we strive for quantum leap process
improvements with immediate results. Total quality management (TQM)
implies work with incremental changes to gradually achieve better results
in business performance (see e.g. Ishikawa, 1985). In other words, we
Nilsson
95
strive for continuous process improvements from time to time. We can
also think of mixed forms between these two extreme types of change programs. One possible example of such a change program can be called
business process elevation (BPE). Here we try to make business improvements in distinct and manageable steps (see e.g. Nilsson et al., 1999). The
size of changes required in business operations depends on the specific
situation. In other words, we strive for promoting the business (“verksamhetslyft” in Swedish) on a regular basis in order to achieve our visions
and goals. Below we launch a change model for business promotion in line
with a BPE philosophy (see Figure 2).
Goal Analysis
Strength
Analysis
Implementation
Assessment of
Next Situation
Promote
Business
Improvement
Analysis
Problem
Analysis
Stakeholder
Analysis
Assessment of
Present Situation
Figure 2. A change model for promoting business – The Clock Model
A lesson learned from the ISD field is the significance of starting up development work from a change analysis which builds a platform for further
development of e.g. information systems. The model for promoting business operations is based on the change analysis method in the traditional
ISAC approach (Lundeberg, et al., 1981). The change model can be
regarded as a clock starting with a goal analysis for the organisation (12
o’clock). We then move on with strength analysis, problem analysis and
stakeholder analysis, i.e. people (users) who are affected by the problems
and strengths. These analyses build a platform for assessing the present
situation before making a “brain-storming” session with an improvement
analysis where we generate appropriate change measures. Again we make
an assessment but now for the next situation for the organisation. Thereafter it is time for the implementation phase when we introduce the desired
business changes in daily work. After a period of time we start a new
change program for business promotion according to the clock model.
96
Exploring Patterns in Information Management
References
Andersen, E.S. (1994) Systemutveckling: Principer, Metoder, Tekniker {Information Systems Development: Principles, Methods and Techniques}, Studentlitteratur, Lund, Sweden.
Ansoff, I. & McDonnell, E. (1990) Implanting Strategic Management, 2nd Edition,
Prentice-Hall, New York.
Avison, D.E. & Fitzgerald, G. (2003) Information Systems Development: Methodologies, Techniques and Tools, 3rd Edition, McGraw-Hill, London.
Bally, L. & Brittan, J. & Wagner, K.H. (1977) “A Prototype Approach to Information System Design and Development”, Information & Management, Vol.
1, pp. 21-26.
Booch, G. & Rumbaugh, J. & Jacobson, I. (1999) The Unified Modelling Language User Guide, Addison-Wesley, Reading, Massachusetts.
Budde, R. & Kautz, K. & Kuhlenkamp, K. & Züllighoven, H. (1992) “What is
Prototyping?”, Information Technology & People, Vol. 6, No. 2-3, pp. 89-95.
Davenport, T.H. (1993) Process Innovation: Reengineering Work through Information Technology, Harvard Business School Press, Boston, Massachusetts.
Davenport, T.H. (1998) “Putting the Enterprise into the Enterprise System”, Harvard Business Review, July-August, Vol. 76, No. 4, pp. 121-131.
Davenport, T.H. (2000) Mission Critical: Realizing the Promise of Enterprise
Systems, Harvard Business School Press, Boston, Massachusetts.
Earl, M. & Khan, B. (2001) “E-Commerce Is Changing the Face of IT”, MIT
Sloan Management Review, Fall 2001, pp. 64-72.
Edvardsson, B. & Gustafsson, A. & Johnson, M.D. & Sandén, B. (2000) New
Service Development and Innovation in the New Economy, Studentlitteratur,
Lund, Sweden.
Elin, L. (2001) Designing and Developing Multimedia: A Practical Guide for the
Producer, Director and Writer, Allyn and Bacon, Boston.
Fitzgerald, B. & Russo, N.L. & Stolterman, E. (2002) Information Systems Development: Methods in Action, McGraw-Hill, London.
Hawryszkiewycz, I. (2001) Systems Analysis and Design, 5th Edition, Prentice
Hall/Pearson Education Australia, Sydney.
Iivari, J. & Lyytinen, K. (1998) “Research on Information Systems Development
in Scandinavia – Unity in Plurality”, Scandinavian Journal of Information Systems, Vol. 10, No. 1-2, pp. 135-186.
Ishikawa, K. (1985) What is Total Quality Control?: The Japanese Way, PrenticeHall, Englewood Cliffs, New Jersey.
Nilsson
97
Jacobson, I. & Christerson, M. & Jonsson, P. & Övergaard, G. (1993) Object-Oriented Software Engineering: A Use Case Driven Approach, 4th revised printing, Addison-Wesley, Wokingham, England.
Jacobson, I. & Ericsson, M. & Jacobson, A. (1994) The Object Advantage: Business Process Reengineering with Object Technology, Addison-Wesley, Wokingham, England.
Kaplan, R.S. & Norton, D.P. (1996) The Balanced Scorecard: Translating Strategy into Action, Harvard Business School Press, Boston, Massachusetts.
Langefors, B. (1973) Theoretical Analysis of Information Systems {THAIS}, Studentlitteratur, Lund and Auerbach, Philadelphia.
Langefors, B. (1974) “Information Systems”, IFIP Congress 74, Stockholm,
North-Holland Publishing Company, Amsterdam, pp. 937-945.
Langefors, B. (1995) Essays on Infology: Summing Up and Planning for the
Future, Studentlitteratur, Lund, Sweden.
Likert, R. (1961) New Patterns of Management, McGraw-Hill, New York.
Lundeberg, M. (1993) Handling Change Processes: A Systems Approach, Studentlitteratur, Lund, Sweden.
Lundeberg, M. & Goldkuhl, G. & Nilsson, A.G. (1981) Information Systems
Development: A Systematic Approach, Prentice-Hall, Englewood Cliffs, New
Jersey.
Lundeberg, M. & Sundgren, B. (Eds.) (1996) Advancing Your Business: People
and Information Systems in Concert, Stockholm School of Economics and
Studentlitteratur, Lund (also published electronically on the Internet:
http://www.hhs.se/im/efi/ayb.htm).
Mumford, E. (1971) Systems Design for People: Economic Evaluation of Computer Based Systems, The National Computer Center (NCC), Manchester,
England.
Nilsson, A.G. (1991) Anskaffning av Standardsystem för att Utveckla Verksamheter: Utveckling och Prövning av SIV-metoden {Acquisition of Application Packages for Developing Business Activities: Development and Validation
of the SIV Method}, Stockholm School of Economics, Stockholm (in Swedish
with English Summary).
Nilsson, A.G. (1995) “Evolution of Methodologies for Information Systems
Work: A Historical Perspective”, in Dahlbom, B. (Ed.) The Infological Equation: Essays in Honor of Börje Langefors, Gothenburg Studies in Information
Systems, Göteborg University, Göteborg, Sweden, pp. 251-285.
Nilsson, A.G. (1999) “The Business Developer’s Toolbox – Chains and Alliances
between Established Methods”, in Nilsson, A.G. & Tolis, C. & Nellborn, C.
(Eds.) Perspectives on Business Modelling: Understanding and Changing
Organisations, Springer, Berlin, pp. 217-241.
98
Exploring Patterns in Information Management
Nilsson, A.G. (2001) “Using Standard Application Packages in Organisations:
Critical Success factors”, in Nilsson, A.G. & Pettersson, J.S. (Eds.) On Methods for Systems Development in Professional Organisations: The Karlstad
University Approach to Information Systems and its Role in Society, Studentlitteratur, Lund, Sweden, pp. 208-230.
Nilsson, A.G. (2003) “Business Modelling as a Base for Multimedia Development: Concerning Strategic, Process and Systems Levels in Organisations”, in
Burnett, R. & Brunström, A. & Nilsson, A.G. (Eds.) Perspectives on Multimedia: Communication, Media and Information Technology, Wiley, London
(forthcoming).
Nilsson, A.G. & Pettersson, J.S. (Eds.) (2001) On Methods for Systems Development in Professional Organisations: The Karlstad University Approach to
Information Systems and its Role in Society, Studentlitteratur, Lund, Sweden.
Nilsson, A.G. & Tolis, C. & Nellborn, C. (Eds.) (1999) Perspectives on Business
Modelling: Understanding and Changing Organisations, Springer, Berlin.
Olle, T.W. & Hagelstein, J. & Macdonald, I.G. & Rolland, C. & Sol, H.G. & Van
Assche, F.J.M. & Verrijn-Stuart, A.A. (1991) Information Systems Methodologies: A Framework for Understanding, 2nd Edition, Addison-Wesley, Wokingham, England.
Orlikowski, W.J. & Iacono, C.S. (2001) “Research Commentary: Desperately
Seeking the ‘IT’ in IT Research – A Call to Theorizing the IT Artifact”, Information Systems Research, Vol. 10, No. 2, pp. 121-134.
Österle, H. (1995) Business in the Information Age: Heading for New Processes,
Springer, Berlin.
Packer, R. & Jordan, K. (Eds.) (2001) Multimedia: From Wagner to Virtual Reality, W.W. Norton & Company, New York.
Ploom, A. (1988) “Information Technology and the Manufacturing Enterprise”,
NordDATA 88, Helsinki, Band 2, pp. 275-281.
Porter, M.E. (1980) Competitive Strategy: Techniques for Analyzing Industries
and Competitors, The Free Press, New York.
Porter, M.E. (1985) Competitive Advantage: Creating and Sustaining Superior
Performance, The Free Press, New York.
Rummler, G.A. & Brache, A.P. (1995) Improving Performance: How to Manage
the White Space on the Organization Chart, 2nd Edition, Jossey-Bass Publishers, San Francisco.
Samuelson, L.A. (1980) Models on Accounting Information Systems: The Swedish
Case, Studentlitteratur, Lund, Sweden.
Simon, H.A. (1965) The Shape of Automation for Men and Management, Harper
and Row, New York.
Nilsson
99
Sowa, J.F. & Zachman, J.A. (1992) “Extending and Formalizing the Framework
for Information Systems Architecture”, IBM Systems Journal, Vol. 31, No. 3,
pp. 590-616.
Tolis, C. & Nilsson, A.G. (1996) “Using Business Models in Process Orientation”,
in Lundeberg, M. & Sundgren, B. (Eds.) Advancing Your Business: People
and Information Systems in Concert, Chapter VIII, Stockholm School of Economics and Studentlitteratur, Lund, Sweden (also published electronically on
the Internet: http://www.hhs.se/im/efi/ayb.htm).
Yourdon, E. (1993) Yourdon Systems Method: Model-Driven Systems Development, Yourdon Press, Prentice-Hall, Englewood Cliffs, New Jersey.
100
Exploring Patterns in Information Management
—7—
Patterns in Change Projects:
Typical Traps
Pär Mårtensson
Introduction
The aim of this chapter is to investigate change processes and to reveal
patterns in terms of typical traps. There are two underlying reasons for this
aim: first, to increase our knowledge about change processes, and second,
to offer people working with change processes in practice ideas for how
their work could be improved.
I will start out by addressing three fundamental underlying views for the discussion. The first is that I view reality as socially constructed (cf. Berger and
Luckmann, 1966). The second is that there is an underlying systems
approach in the discussion (e.g. Langefors, 1966). The third is related to the
view of reality and concerns the value of different perspectives. Given the
assumption of a social construction of reality and that reality is a mental
phenomenon, the value of including different perspectives is significant (cf.
Lundeberg, 1993). By finding ways of including different perspectives, one
can increase our ability to perceive different aspects of reality.
These underlying views taken together imply that my view of change
processes in the discussion to follow is in line with Mats Lundeberg’s
work presented in “Handling Change Processes: A Systems Approach”
(1993). This view suggests that it is possible to improve the ability to
handle change processes in a business context by learning to recognize
patterns.
The chapter is structured in the following way. After the introduction I discuss some theoretical aspects of change. Then follows a section on change
from a practical perspective, where I describe the empirical foundation for
the patterns, which are described in the next section in form of seven typical traps in change projects. Then there is a discussion and some practical
implications, where I address people working as project leaders in practice
102
Exploring Patterns in Information Management
(cf. Robey and Markus, 1998). The chapter ends with some concluding
remarks.
Change in Theory
The amount of previous research on different aspects of change is extensive (e.g. Lewin, 1947; Watzlawick et al., 1974; Lundeberg, 1993; Kotter,
1996). Literature includes many different aspects of change, and in the
following I briefly touch upon five themes: contexts of change, communication for change, levels of change, leading change and models for change.
I do not include what one needs to communicate in order to accomplish
change, like for example the need to communicate a change vision, which
is discussed in detail elsewhere (e.g. Kotter, 1996).
Contexts of Change
The systems approach, or systems thinking as suggested by Checkland
(1981), offers a broad view for understanding a context. The importance of
viewing development processes in their wider contexts has been emphasized in theories on information systems development (e.g. Lundeberg et
al., 1981). Here, perceived needs in business activities could be seen as
goals for development efforts regarding information systems, which in turn
could be seen as means to fulfill these business needs. In later theories on
change processes, business needs in turn have been put in context in terms
of people involved and their intentions (Lundeberg, 1993).
Applying the systems approach can help place change processes in their
wider contexts and help understand the complexity involved, by offering
ways to view the complexity in terms of different parts with relationships
among the parts. Abstraction is one way for the human mind to deal with
complexity (Argyris, 1982). Thinking in abstractions, for example in the
form of levels of abstraction (Bateson, 1972), can help describe how different contexts are interrelated. When using levels of abstraction it is
important to bear in mind that there are no “true” levels, but one chooses
what to view as levels. This is in accordance with the systems approach,
where one chooses what to view as a system (e.g. Churchman, 1968;
Checkland, 1981).
Communication for Change
In the discussion on communication for change I will point to three
aspects: meeting the other persons where they are (the congruence aspect),
Mårtensson
103
allowing for a requisite variety in the communication (the flexibility
aspect), and seeing the situation at hand through different frames of references (the reframing aspect).
The congruence aspect of communication deals with the need to meet
other persons where they are when establishing any form of communication. This is described and discussed in terms of the congruence model
(Andersen et al., 1994; Lundeberg, 1993), where one key point is to establish a communication process with another person by starting a discussion
about something that the other person is willing to discuss. The congruence can also be expressed in terms of communicative steps: inform each
other, exchange opinions, make use of each other’s opinions, and create
new ideas together (Andersen et al., 1994). The initial phase of a communication process builds a basis for further communication and helps create
conditions for achieving intended results from the communication process,
for example in the form of changes.
The flexibility aspect of communication deals with the need to be flexible
in a communication process. This aspect is related to the first aspect
about congruence, where there is a need to meet a person where he or she
is. The theoretical background to the flexibility aspect can be found in
the area of cybernetics and industrial process control. Ross Ashby
formulated the “Law of Requisite Variety” (Ashby, 1956), which basically says that “only variety can destroy variety” (ibid., p. 207). This may
sound a bit destructive, but transferred to a communicational situation it
means that the ability to be flexible, or show variety, is a key to successful communication. A greater degree of flexibility is more likely to result
in intended ends.
The reframing aspect of communication is the importance of framing and
reframing the situation where any communication is used to achieve
changes (Watzlawick et al., 1974). Reframing may be described here as
changing the “conceptual and/or emotional setting or viewpoint in relation
to which a situation is experienced and to place it in another frame which
fits the ‘facts’ or the same concrete situation equally well or even better,
and thereby changes its entire meaning” (Watzlawick et al., 1974, p. 75).
Reframing could also be expressed in terms of bisociation where a situation is not only associated to one context, but bisociated with two
(Koestler, 1964). One example of reframing, or bisociation, is humor
where the punchline often presents the facts through a new frame (ibid.).
In a similar way it is important to reframe situations and view situations
through different frames, in order to see new patterns, when communicating to achieve change (Watzlawick et al., 1974).
104
Exploring Patterns in Information Management
Levels of Change
Change efforts can be seen as being of different orders: where changes of
the first-order take place within a system, and second-order changes are
when changes of the system take place (Watzlawick et al., 1974). The two
orders could be seen as taking place on two different logical levels
(Bateson, 1972).
The two fundamentally different types of change have become most wellknown in learning contexts through the concepts of single-loop and double-loop learning (Argyris and Schön, 1974). That is, learning within given
settings and frames, versus learning by changing the setting and moving
beyond the frames.
One way to capture these opportunities for learning is through reflection
on action (e.g. on change efforts). In order to enhance the possibilities of
learning in relation to change processes there is a need for the ability for
“reflection-in-action” (Schön, 1983).
Leading Change
Many research efforts have focused on aspects of leaders’ roles in
change processes and how change could be handled. Much of the
results of these efforts have been published with a practical and partly
normative stance (e.g. Kotter, 1996). There are many different types of
processes included on the managerial agenda (e.g. Mårtensson, 2001)
and the reframing activities discussed above can help reveal driving
forces for including change efforts on the managerial agenda. By asking how a particular change effort is handled and how this could be
perceived from other perspectives, different framings of the change
effort could be revealed. Framing and reframing of change processes
can also be seen in the context of attracting managerial attention (or
not) to a change effort. People can for example try to “sell” an issue to
managers by framing the issue in a certain way (cf. Dutton and Ashford, 1993).
Change efforts are not only handled, they are also sometimes mishandled.
Watzlawick et al. (1974) have suggested three basic ways of mishandling
change, as illustrated below in Figure 1.
Mårtensson
105
Necessary:
No
Taken Action:
Yes
A
No
Yes
B
C
Action taken at
the wrong level
Figure 1: Three Ways of Mishandling Change
(based on Watzlawick et al., 1974, p. 39).
The three ways of mishandling change could be described as: (A) action is
necessary but is not taken, (B) action is taken when it should not be, and
(C) action is taken at the wrong level. Action taken at the wrong level
could mean that there is need for second-order change activities, but
efforts are only made in terms of first-order changes. In such a situation
more effort will not help, instead there is a need for a shift in focus to
change at another level. Underlying this view of change one could trace
the theories of logical levels. It is worth noting that the expression “necessary” could be challenged in terms of the clarifying question “according to
whom and by what criterion” (Lundeberg, 1993).
Models for Change
A number of models have been developed for use in change processes (e.g.
Lundeberg, 1993). Here I will briefly introduce two such models: the Xmodel and the Y-model.
The X-model is a general model for describing task and person (relationship) aspects of processes (Lundeberg, 1993). In the X-model two fundamental levels related to persons and task are described. The model basically says that all processes include both person-related and task-related
aspects. There is input to the process in the form of person preconditions
and task preconditions. The process in itself consists of behavior aspects
and task-related aspects. There is output from the process in form of person outcomes and task outcomes. Figure 2 illustrates the X-model.
106
Exploring Patterns in Information Management
Input
Process
Output
Relationship
Person
Preconditions
Behavior
Person
Outcomes
Task
Task
Preconditions
Task
Processes
Task
Outcomes
Figure 2: The X-model (Lundeberg, 1993, p. 15).
The Y-model is a general model for describing five different process
focuses (Lundeberg, 1993). The model basically says that all these process
focuses are important when handling change processes. The five focuses
are: Current Situation (to find out where one is); Intended Future Situation
(to choose where one wants to be); Need for Changes (to select what to
achieve to get from where one is to where one wants to be); Change Alternatives (to find out different ways to achieve this); Outcomes (to act in
order to get from where one is to where one wants to be).
It is worth noting that the different process focuses in the model are not
presented to be sequential. Figure 3 illustrates the Y-model.
Current
Situation
Need
for
Changes
Change
Alternatives
Intended
Future
Situation
Figure 3: The Y-model (Lundeberg, 1993, p. 19).
Outcomes
Mårtensson
107
Change in Practice
The patterns discussed below are based on more than 40 change projects
that have been carried out in Scandinavia between 1995-2002 where all
projects have lasted for about one year. In terms of different types of
change efforts, the change processes could be described as project-based
improvement (cf. Davenport, 1993). My own role in relation to the projects is that I have been a coach to the project leaders. This means that I
have had a good insight into the projects, but have not been directly
involved in the projects myself. The project leaders have been working in
various industries, and most of them have been between 30 and 45 years
old. Some of them have been experienced project leaders, while others
have had less experience.
The projects have been carried out as a part of the Executive MBA program at the Stockholm School of Economics and have followed a format
where the first part of each project has been to carry out change studies (cf.
Lundeberg, 1993). Basically these change studies aim at analyzing the
situation in terms of the Y-model described above. The second part of the
projects has been to implement the solutions suggested in the change
studies. During the process, groups have been formed with four to five
projects in each group, where the project leaders have met on a regular
basis in order to share experiences and to help each other in the projects.
When these groups met, the project leaders had prepared written reports of
the situation in the project to share with the other people in the group.
Seven Typical Traps in Change Projects
After having coached a number of change projects, I began to recognize
some patterns. In the following paragraphs I will present some of the patterns that have emerged, and I will do so in the form of seven typical traps
that I have seen project leaders fall into.
The Jeopardy Trap is one where the answer is given from the beginning of
the process. Even if there is a change study carried out first, the answer is
already clear to the project leader and it is only a matter of finding a suitable question to the answer (just like in the TV-show “Jeopardy”). This
means that the process of carrying out the change study, and investigating
the different process focuses in the Y-model, becomes a meaningless exercise since the result of the change study is given from the beginning.
108
Exploring Patterns in Information Management
A risk with the Jeopardy Trap is that the project leader is committed to a
certain solution from the beginning and is not open for alternatives. This
means that there may be more suitable solutions that the project leader
cannot see, or does not want to see.
The Neutron Bomb Trap is where people seem to be extinguished from the
change project. (The expression Neutron Bomb is used to illustrate something that wipes out human life, but leaves the rest. It is worth noting that I
use the expression only as an illustration of a phenomenon, without
detailed knowledge about neutron bombs.) Often there are clear task-oriented descriptions and logical lines of argument in the projects, but people
are not included at all. In terms of the X-model described above, this
means that the entire focus is on the task-oriented level.
A risk with the Neutron Bomb Trap is that the change project is planned
with too much focus on task and not enough attention on the person-oriented level. This means for example that the project may face difficulties
in the implementation phase if people and their driving forces are not
included in the analysis of the situation.
The Confusion Trap is where different process focuses are considered in
the change project, but these focuses are not coherent. The project leader
may have worked through all parts of the Y-model and described the different focuses, but the different parts are not related to each other. There
may for example be a description of a current situation related to one part
of the organization and a description of a future situation of another part,
etc. Some parts may also be missing. The confusion can furthermore be
related to the described deliverables from the change project, where the
intended effects are not coherent with the intended deliverables.
A risk with the Confusion Trap is that the project leader does a good job
analyzing a situation, but does not pay enough attention to the importance
that the different parts are coherent. This means that there may be an
extensive analysis as a basis for the planned actions, but this analysis does
not capture the actual situation.
The Bad-Good-Improve Trap relates to the precision in descriptions and
communication in the change project. The trap is named after the simplistic formulation of the situation analyzed, where the current situation is
described as “bad”, the intended future situation is described as “good”,
and the need for changes is described as “improve”. The use of these three
expressions is a simplification intended to illustrate too simplistic descriptions of the different process focuses in the Y-model.
Mårtensson
109
A risk with the Bad-Good-Improve Trap is that the shallow descriptions of
the situation in the change project may cause misunderstandings. They
may also cause a lack of understanding of the root causes for the change
project.
The Poker Trap is where the project leader keeps the cards secret from
others. He or she discloses as little as possible about the project. Other
project members or people in different types of reference groups, etc, are
kept as uninformed as possible. Information about the project is presented
bit by bit, when the project leader sees necessary. Of course there may be
pieces of information that the project leader does not need to share with
other people. The trap is to illustrate that too limited communication about
the project may cause problems.
A risk with the Poker Trap is that the project leader does not open up the
information about the project enough to allow for useful feedback regarding the project which can help the project leader to see unexpected possibilities. It may also cause misinterpretations about the project and make
people more suspicious than necessary about the project.
The Chameleon Trap concerns the written presentations of the project and
is where descriptions of the project are made to meet requirements from all
different target groups. The project leader may need to report about the
project to various people and in order to do this in an efficient way he or
she prepares one report to meet the needs of different groups. (The word
Chameleon is used to illustrate something that is intended to change form
and shape in order to meet different situations.)
A risk with the Chameleon Trap is that the project leader prepares something that is not useful to anyone. The presentations do not meet any target
group. Instead of being efficient the project leader rather may be the opposite.
The Bravery Trap means that the change project grows and that “everything” is linked to the project and included in it. The project scope just
keeps expanding. The project leader wants to do a good job, and there are
good intentions behind the situation. The problem is that there is the intention to solve all possible problems within the scope of the project. The
“best project” becomes the opponent of the “good project”.
A risk with the Bravery Trap is that the scope of the project grows so
much that the project cannot be completed. There is also a risk that this
growth is not identified since it takes place gradually.
Table 1 below summarizes the seven typical traps.
110
Exploring Patterns in Information Management
Trap
Description
The Jeopardy
Trap
The answer is given from the beginning and one tries to find a
question.
The result of the change study is given from the beginning.
The Neutron
Bomb Trap
People seem to be extinguished.
The Confusion
Trap
Every part of the Y-model is included, but the parts are not
coherent. (Or some parts may be missing.)
Often there are clear task-oriented descriptions and logical
lines of arguments, but people are not included.
and/or
Deliverables and effects are described, but they are not
coherent.
The Bad-GoodImprove Trap
Current Situation = Bad.
Intended Future Situation = Good.
Need for Changes = Make Better.
The Poker Trap
The cards are kept secret from others. As little as possible
about the project is disclosed.
Information about the project is presented gradually bit by bit.
The Chameleon
Trap
The descriptions of the project are made to meet requirements
from all different target groups.
The Bravery
Trap
Everything is linked to the project and included in it.
The project just keeps growing.
There is a real good intention to solve all possible problems
within the scope of the project.
The best project becomes the enemy of the good project.
Table 1: A Summary of Seven Typical Traps in Change Projects
Discussion
The seven typical traps in change projects described in the previous section
illustrate an array of potential difficulties that a project leader can face.
Some traps concern the ability to deal with different contexts of the change
process (e.g. the Neutron Bomb Trap and the Chameleon Trap). A lack of
ability to think in abstraction may explain some of these difficulties (cf.
Argyris, 1982; Lundeberg, 1993). Here, it may be of importance to find
suitable ways of seeing one’s own project through different frames (cf.
Mårtensson
111
Koestler, 1964). Difficulties related to communication in change processes
explain some traps (e.g. the Poker Trap and the Chameleon Trap).
Focusing change efforts on the wrong level (cf. Watzlawick et al., 1974)
can help explaining some traps (e.g. the Jeopardy Trap and the Bravery
Trap). There are also links between aspects of leading change and some
traps (e.g. the Confusion Trap and the Bravery Trap). Finally, some traps
more directly concern models for change (e.g. the Confusion Trap and the
Bad-Good-Improve Trap). Models may be used with the intention to
improve the quality of the change project, but for various reasons the result
may not always be good. One reason may be a lack of precision in
descriptions and an insufficient understanding of the models.
In the following paragraphs I will discuss how each trap maybe can be
understood seen through a theoretical lens.
The Jeopardy Trap, where the goal for the project is decided from the outset of the project, relates to leading change. The project leader knows, or
wants to know, the answer from the beginning, and may want to reduce the
uncertainty in the change process. This could for example result in a situation where unnecessary actions are taken (cf. Watzlawick et al., 1974), as
the change study is never allowed to analyze the situation at hand. By
deciding the answer of the analysis from the beginning, there is also the
risk that there will be a focus on a first-order change even if there is a need
for a second-order change (cf. ibid.).
The Neutron Bomb Trap links to the use of models (X-model) for change,
where there is a need to cover both task-oriented and person-oriented
aspects of the change process. Omitting persons from the analysis of the
situation is also an example of a narrow view of the context of the change
process (cf. Lundeberg, 1993). A possible reason for the trap may be the
desire to reduce complexity or simply a tradition to focus on task-oriented
aspects of change.
The Confusion Trap relates to the use of models (Y-model) for change,
where different process focuses are not coherent. The use of models can
help identify and clarify possible misunderstandings in change processes
(cf. Lundeberg, 1993) and help further the understanding of the context
of the change. The project leader faces challenges when leading the
change process, and the trap may indicate that there are different types of
mishandling the change processes (cf. Watzlawick et al., 1974) in terms
of, for example, necessary actions not being taken due to confusion.
112
Exploring Patterns in Information Management
The Bad-Good-Improve Trap relates closely to the use of models (Ymodel) for change. The trap also concerns communication for change,
especially internal communication within the project. By reducing the precision in descriptions of the project, the project leader reduces the possibilities for successful communication. This may be the result of a too simplified change analysis and not enough effort being put in the analysis process. There may also be political reasons behind the trap, i.e. an unwillingness to clarify the situation. The congruence aspect of communication (cf.
Andersen et al., 1994) is described as crucial, and this trap may cause difficulties when aiming at congruence.
The Poker Trap mainly involves communication aspects of the change
project as it deals with the secrecy of project leaders. This trap also
relates to leading change. Secrecy may derive from an uncertainty perceived by the project leader. The trap may inhibit reflection on the
actions in the project and thereby reduce the possibilities for learning (cf.
Schön, 1983).
The Chameleon Trap is also highly related to communication aspects of
change. Here the written reports are in focus, and especially flexibility (or
rather lack of flexibility) (cf. Ashby, 1956). This trap may come from a
willingness to reduce the efforts spent on reporting about the project, and a
limited understanding of the information needs of the different groups in
the context of the project, as discussed above in relation to the congruence
aspect of communication (Andersen et al., 1994).
The Bravery Trap is a result of good intentions, but where the project
leader may face difficulties in dealing with the complexity involved in
the context of the change (cf. Argyris, 1982). The trap can sometimes be
explained by, not only the project leader increasing the scope of the project, but also people trying to “sell” issues for the project leader to
include in the project (cf. Dutton and Ashford, 1993). By moving the
change efforts from a first-order change to a second-order change (cf.
ibid.), the context of the change could be seen in another frame (cf.
Koestler, 1964).
Practical Implications
“Ok, so what can I do if I am a project leader?” In this section I will
address this question and allow myself to be more normative and give
some pieces of practical advice. If you are a project leader for a change
Mårtensson
113
project, the following are some ideas about how to avoid the traps discussed above.
You can avoid:
• the Jeopardy Trap by being open to varied results from the change
study. Allow yourself to not know everything from the very beginning,
and accept a certain amount of uncertainty in early phases of the change
project.
• the Neutron Bomb Trap by including people in the change study. If you
normally do not describe much about people, but focus on task-oriented
matters instead, try to do things differently this time. Try to find out
more about the driving-forces and individuals for (or against) the
change efforts.
• the Confusion Trap by making sure that all parts of the Y-model, as
well as deliverables and effects, are described, and that they are
coherent. Instead of trying to include everything in one single Ymodel, notice that you may find it more fruitful to present the
situation in multiple Y-models.
• the Bad-Good-Improve Trap if you perceive the precision in the
descriptions as important. You need to spend enough effort on precision
in order to get a sufficient level of details. Especially, the needs for
change are of vital importance to pinpoint.
• the Poker Trap if you share information with other people. This does of
course not mean that you should share everything with everyone, but
dare to open up. If it is a lot of information to grasp for people, choose
ways to present it with clarity.
• the Chameleon Trap if you think through various target groups and how
they can be reached. Remember that what looks like a short cut often is
the longest way round. If you try to reach everyone with the same document, you may not reach anyone.
• the Bravery Trap if you delimit the change study in its final phases.
Allow for openness in the early phase (cf. the Jeopardy Trap) and then
be realistic in the action plan. Take some time to reflect on alternative
dimensions for delimitations of the project, and look for different types
of phases in the project.
In Table 2 practical implications of the traps are summarized.
114
Exploring Patterns in Information Management
Trap
Practical Implications
The Jeopardy
Trap
Be open for a range of results in the change study.
The Neutron
Bomb Trap
Include people in the change study.
The Confusion
Trap
Make sure that all parts of the Y-model, as well as deliverables
and effects, are described and that they are coherent.
Accept a certain amount of uncertainty in the beginning.
Find out about driving-forces for (or against) change.
Notice that there may be different Y-models for the interested
parties.
The Bad-GoodImprove Trap
The precision in the descriptions is important.
The Poker Trap
Share information with other people.
Needs for change are of vital importance.
If there is much information: choose ways to present it with
clarity.
The Chameleon
Trap
Think through various target groups and how they can be
reached.
The Bravery
Trap
Delimit the change study in its final phases.
Look at the feasibility in the action plan.
Reflect on alternative dimensions for delimitations.
Table 2: A Summary of Practical Implications from the Seven Typical Traps
Concluding Remarks
The aim with this chapter has been to reveal patterns in order to increase
our knowledge about change processes, and to offer people working with
change processes ideas for how they can improve their work. It is worth
highlighting that falling into traps most often is a result of good intentions.
Project leaders want to achieve good results in their projects, but on the
way it is easy to fall into different types of traps. One should bear in mind
that there may be rational and logical reasons behind the traps. In this
chapter I have presented some patterns that have emerged from a number
of change projects.
My intention with the chapter has been to illustrate some typical patterns
in change projects, which hopefully can contribute to our understanding of
challenges in change processes. The form of seven typical traps should be
seen as seven opportunities to learn about change projects and how typical
traps could be avoided. The traps are not the seven deadly sins, but merely
Mårtensson
115
seven areas to pay attention to, if you are interested in change projects
from a practical perspective.
Mats Lundeberg (1993) has defined three core subprocesses in handling
change processes: to perceive reality as it is, to make use of the freedom of
action that you have, and to learn from the consequences of what you do.
Hopefully, the seven typical traps presented in this chapter can be a tangible help when dealing with these three subprocesses.
References
Andersen, E.S., Baustad, I. & Sørsveen, Å. (1994) Ledelse på norsk: Prinsipper,
arbeidsmåter og resultater {Management In Norwegian: Principles, Ways of
Working and Results}, Ad Notam Gyldendal, Oslo, Norway.
Argyris, C. (1982) Reasoning, Learning, and Action: Individual and Organizational, Jossey-Bass Publishers, San Francisco.
Argyris, C. & Schön, D.A. (1974) Theory in Practice: Increasing Professional
Effectiveness, Jossey-Bass Publishers, San Francisco.
Ashby, W.R. (1956) An Introduction to Cybernetics, Chapman & Hall, London,
England.
Bateson, G. (1972) Steps to an Ecology of Mind, The University of Chicago Press,
Chicago, Illinois.
Berger, P.L. & Luckmann, T. (1966) The Social Construction of Reality, Anchor
Books, Doubleday, New York.
Checkland, P. (1981) Systems Thinking, Systems Practice, John Wiley & Sons,
Chichester, England.
Churchman, C.W. (1968, second edition 1979) The Systems Approach (second
edition), Dell, New York.
Davenport, T.H. (1993) Process Innovation: Reengineering Work through Information Technology, Harvard Business School Press, Boston, Massachusetts.
Dutton, J.E. & Ashford, S. J. (1993) “Selling Issues to Top Management”, Academy of Management Review, Vol. 18, No. 3, pp. 397-428.
Koestler, A. (1964) The Act of Creation, Arkana Penguin Books, London, England.
Kotter, J.P. (1996) Leading Change, Harvard Business School Press, Boston,
Massachusetts.
Langefors, B. (1966, fourth edition 1973). Theoretical Analysis of Information
Systems (fourth edition), Auerbach Publishers, Philadelphia, Pennsylvania
(also published by Studentlitteratur, Lund, Sweden).
116
Exploring Patterns in Information Management
Lewin, K. (1947) “Frontiers in Group Dynamics”, in Lewin, K. (1997). Resolving
Social Conflicts & Field Theory in Social Science, American Psychological
Association, pp. 301-336, Washington, DC.
Lundeberg, M. (1993) Handling Change Processes: A Systems Approach, Studentlitteratur, Lund, Sweden.
Lundeberg, M., Goldkuhl, G. & Nilsson, A. (1981). Information Systems Development: A Systematic Approach, Prentice-Hall, Englewood Cliffs, New Jersey.
Mårtensson, P. (2001) Management Processes – An Information Perspective on
Managerial Work, Economic Research Institute (EFI), Stockholm School of
Economics, Stockholm.
Robey, D. & Markus, M.L. (1998) “Beyond Rigor and Relevance: Producing Consumable Research about Information Systems”, Information Resources Management Journal, Vol. 11, No. 1, pp. 7-15.
Schön, D.A. (1983) The Reflective Practitioner: How Professionals Think in
Action, Basic Books, HarperCollins Publishers, New York.
Watzlawick, P., Weakland, J.H. & Fisch, R. (1974) Change: Principles of Problem Formation and Problem Resolution, W.W. Norton & Company, New
York.
—8—
Errors Help Users Learn?
Alf Westelius
Errors – A Problem or an Opportunity
Are errors always something negative? Ought an IT application that is
being installed always be error-free when it is implemented? Or could
there be any positive aspects of the existence of errors?
In information systems literature, errors are seen as something to be
avoided, and something that can be reduced by proper planning. In combination with communication, errors are not discussed as something positive
either. When errors and communication appear together, it is typically a
question of designing communication to reduce errors in that communication, even when the framework is one of organisational learning (e.g.
Salaway, 1987). Other literature stresses that errors make us rethink, affect
sensemaking and action. Oatley and Johnson-Laird (1987) identify five
basic emotions, and propose that they are evoked when our attempts to
reach a goal are hindered – for example when errors stop us from completing the task at hand. They view the purpose of emotions as directing
attention and provoking changes in priorities. Simon (1945, 1997, pp. 9091) notes that we can not handle tasks and problems in parallel if they
really demand our attention; we are forced to deal with them one by one,
serially. He then sees the same connection between interruptions, emotions
and attention. Interruptions evoke emotions, and emotions are needed to
direct, interrupt and change the priorities in the information processing we
humans engage in. Weick is on a similar track regarding interruptions,
emotions, and thinking, stating:
The reality of flows becomes most apparent when that flow is interrupted. An interruption to a flow typically induces an emotional
response, which then paves the way for emotion to influence sensemaking. (Weick, 1995, p. 45).
Thus, errors are interesting because they provoke emotions that direct our
attention to the solving of the problem that caused the emotion to arise.
Could it be that they prompt communication – raising our attention to the
118
Exploring Patterns in Information Management
point where we start to overcome the barriers of communicating with people we do not know? The malfunction in an information system would
keep someone from reaching the goal of completing a task. The attention
directed at solving the problem could lead the person experiencing the
problem to contact others, who are seen as responsible for the problem or
able to solve or help solve it. Could errors be beneficial in that they lead to
the development of networks?
A Strategic IT Platform Project
In the European part of an international, industrial group, some centrally
placed managers in one of the companies and at corporate headquarters
decided to increase efficiency by increasing the use of IT.1 Implementing the
same ERP system in all the European subsidiaries in a unified manner would
provide a platform for administrative savings, and increased connectivity to
better be able to serve international customers. The implementation would
be expensive. Some of the required functionality was not available from any
Business Suite vendor and would have to be developed. But the direct
administrative savings and the potential marketing and customer service
benefits would quickly ensure a healthy payback… So they thought.
The Project and its Effects
The project turned out to be the largest the group had ever undertaken.
Seven years and much error-fraught work later, the ERP system was
implemented across Europe. Sales volume had increased substantially, but
no one claimed that the sales volume increase was caused by the Business
Suite implementation. Despite the increase in sales volume, there had been
little increase in back office personnel. However, it was debated whether
or not this apparent increase in productivity could be attributed to the use
of the Business Suite. ERP proponents said that it was obvious. Critics said
that with a proper ERP system, the ratio between sales and back office personnel could have been much greater. Contacts with the actual users indicated yet another possible explanation: the pace of work had increased
1
This account is based on a study in an international company at division management and sales and service company level. More than 70 interviews and meetings have been conducted in four companies and at division headquarters in the
period 2001 to 2003. We have also had access to a substantial amount of project
documentation. The research has been made possible by a grant from VINNOVA,
Swedish Agency for Innovation Systems.
Westelius
119
substantially, transforming the work climate from pleasant and communicative to stressful and strictly task-oriented. Clerks claimed to be
exhausted when they went home in the evenings.
Beneficial Networks Developed
Whatever the actual connection between the ERP implementation and efficiency, there was one point on which everybody seemed to agree: networks
had developed between people, especially at middle and lower hierarchical
levels within the European division as a consequence of the project and the
subsequent attempts to understand and cope with the ERP system. These
networks were not part of the planned benefits. They were rather a surprise.
Of course, a pan-European project would call for contacts across company
borders. Of course, training and supporting users would call for the design of
a training program and the establishment of a support structure. However,
that these instrumental contacts, together with other contacts that developed
spontaneously, rather than being designed, would take on an importance
over and above the purpose for which they were designed, and be viewed as
assets for the organisation as such – that was a surprise.
Should it have been?
Interacting to Sort Out Errors
In an organisational setting, contact between people is needed to sort out
errors. A well-functioning organisation can develop a skill at effectively
dealing with errors. The string quartet is an example of an organisation
where error handling plays a prominent part.
Two Levels of Errors and Two Modes of Error Handling
The smooth enactment of error handling is one of the hallmarks of a
well-functioning string quartet. Flow in rehearsing, with flow in error
handling, provides enjoyable and efficient rehearsals and learning. Errors
in error handling, on the other hand, take time, cause inefficiencies, cause
irritation and strained relationships (Westelius, 2001). Weick’s
commentary to Westelius (2001), centred on error handling and dealing
with interruptions at two levels: errors that interrupt the playing, and
errors that interrupt the flow of the repair work. Interruptions that thwart
the attainment of a goal cause agitation. Repeated interruptions could
then make learning more difficult. Achieving flow of repair work that
consists of frequent interruptions is therefore not trivial, he noted. I
120
Exploring Patterns in Information Management
would suggest that in order to achieve flow of such repair work, you
have to view the detection of first-level errors and subsequent halts in the
execution, not as interruptions, but as an objective of the work. In quartet
playing, this is the case during rehearsals. During performances, the view
of – and the handling of – such errors is quite different.
In the string quartet, members are physically close to each other and therefore have a good chance to detect signals from each other and communicate.
Detection of errors in the playing, and handling them through repetition and
correction, is part of what you expect to do when you rehearse. Error handling during performance, however, centres on avoiding errors, minimising
and covering up those that do occur, and recovering as unnoticeably as possible, should some major problem occur. Designing fault avoidance and fault
recovery strategies can then also be part of rehearsing. Giving cues – emphasis on a certain note or beat, a nod, a motion, … these small indicators can
help uphold co-ordination and confer a sense of security and control: we
know that we are together. Should we lose co-ordination, this is a recognisable passage were we can reassemble, alternatively imperceptibly restart
from (or jump to) point X at the sign of a designated leader. The fluent and
improvised leadership during rehearsals is then replaced with predetermined
leadership and cues that have been explicitly agreed upon.
In a similar manner, we can expect that the attitude towards errors connected
with information systems use, and the behaviour when and if they occur,
would differ according to organisational role and view of your job. We
could probably find people who are predominantly “performers”, and others
who mostly think and act as “rehearsers”? In finance, we could expect to
find rehearsers. Part of the job in a finance department is to identify data that
appears to be inconsistent or questionable. Whether you contact someone
who has caused the error, or you just try to correct it, depends on how you
view your job. In a sales department, we are more likely to find “performers” than “rehearsers”. Here, the data that you enter into or receive from the
administrative computer application are just means to perform your task. If
there are errors in the data, it is an interruption, an embarrassment, and
something that you do not want to let interrupt the “real” work. You could
even expect people in sales to disregard or find workarounds to deal with
erroneous data, rather than spend time and effort on trying to correct it.
Proximity and Communication
In an organisational setting, where you interact with others at a temporal or
spatial distance, co-operation poses somewhat different problems than in
Westelius
121
the close-quarter interaction setting of the string quartet. It has been
claimed that networks mainly develop spontaneously when the distance
between people is less than 50 paces. In Moberg’s study of organising in
flexible offices, physical proximity seemed to be very important to communication patterns within a department (Moberg, 1997), but could pose
obstacles (because of disturbance and loss of privacy) as well as facilitate
contact. Bergum’s studies of managers of remote workers (Bergum, 2000)
indicated an extreme emphasis on travelling (by the managers) and telephone conversations to keep frequent and rich contact with the employees
in the group. That is also an example of the importance of proximity to a
network – proximity, when not a natural part of the daily work, has to be
created if people are to form a network. But why should people using an
administrative piece of software form a network? One answer would be “to
facilitate the handling of errors”. The administrative software, such as a
Business Suite or an ERP system, is intended to facilitate communication
and co-ordination in the organisation. It is then also likely to increase the
dependence between people, and increase the need for standardisation. In
addition, using new, complex software causes a need to learn how to operate it. If it is applied across a large organisation, it is likely that knowledgeable users are to be found in other parts of the organisation, rather
than next to you. If errors and interruptions in the form of low-quality data
or systems-use problems appear repeatedly, established relationships with
people who can help sort out – and maybe even prevent – these first-level
errors, could diminish the risk for second-level errors, interruptions of the
first-level error handling.
What evidence is there that such network formation takes place?
Examples of Errors Leading to Contacts
– and Sometimes on to Networks
Among the cases described in Westelius (1996), there are several examples
of information system problems leading to contact. The meeting held by the
production chief accountant and the project manager in case G, to handle the
sniping against the ABC accounting, is one (pp. 168-169). The chief
accountant in case H, travelling around to show local accountants how they
should enter data into the new accounting application, is another (p. 195). In
neither of these examples do these error-prompted contacts lead to the construction of networks of any permanence. One reason seems to be that these
contacts solve the problems (at least as viewed from project and systems
management). The discontented in G have been listened to and been given a
chance to speak up, and hence stop complaining about errors in the ABC
122
Exploring Patterns in Information Management
accounting. The accountants in H have learned to some extent how to work
with the new accounting system, and manage to enter that year’s budget.
However, it is interesting to note the further development in H. The local
accountants feel they need to learn more about the new principles of
accounting and the proper use of the accounting system. They do not desire
closer contact with the accounting manager (who they experienced as talking
down to them). Instead, they lobby for a meeting where they can get to know
each other and then help each other master the new accounting system (pp.
197-198). Thus they, as problem owners, desire and form a network to help
them solve future errors and interruptions connected with the use of the new
accounting principles and accounting system.
Two Types of First-level Errors
Let us look a bit closer at the types of first-level errors that can occur. One
type of error that can arise is that I do not understand how to use the system in my daily work. I can then experiment on my own, try to use my
existing personal network and rely on people I have confidence in, or try to
establish contact with new people to learn from or with, to solve my problem.
The other type of error is that I receive incorrect data from someone. I can
try to correct it myself. I may also, sooner or later, decide to contact that
individual – and perhaps build a relationship. If I do, it will become natural
to think about that other individual and consider relationships between his
or her work and my own. This will amount to learning about my job in a
larger context (viz. including these other people I have come to contact and
the work they perform).
Problems with Running the System
Let us return to the strategic ERP implementation case. Users in a small
subsidiary (Denmark) were trained by the central project’s trainers, and put
in contact with users and managers in a big subsidiary (England) – people
who knew more about the business and new ways of offering and delivering service (functionality rental instead of product sales and service contracts). When asked to seek help in using the application from the central
helpdesk, these users found it far easier and more rewarding to continue
asking people in the large subsidiary for help. People in the large subsidiary offered assistance, but also were reinforced in their belief that they
were more advanced users than the other subsidiaries, and that they had the
least to gain from the common application of the new ERP system. The
Westelius
123
people in the small subsidiary found the help from the large subsidiary
very useful – help not just regarding the use of the application, but also of
the underlying business concept and that business process.
It seems the Danes started communicating with the English, not because
they were designated as helpers as much as because they actually had
answers, while the system supplier’s international designated helpdesk did
not. Thus, the Danes learned about the business processes as well as about
how to set up and use their ERP installation by keeping a close communication with the English. The English were somewhat flattered, but also felt
that they were giving without receiving – that they were doing tasks that a
well-organised support organisation should have been able to handle. In
that sense, their part of the communication was organisational error handling – handling the malfunctioning support organisation by doing their
work for them.
Problems with Incorrect Data
With an administrative piece of software, you influence others, and are
influenced by them, at a distance. Westelius and Westelius (1990, pp. 120
ff.) talk of the degree of integration with and dependence on others in
terms of using data produced by others, and producing data that others will
use. The implementation of an ERP system or a Business Suite typically
increases both types of dependencies. You come to depend on data from
others and risk causing problems for others by producing data for them,
data that is flawed in relation to the use they want to make of it. When
using an integrated Business Suite or ERP system, these data are directly
available in the computerised system, without the manual filters and controls that you have if you use printouts or other output from another
department as part of the basis for your actions and decisions.
Petri (2001) explored the question of making the effort to produce data that
others will use – what will motivate you to prioritise that ahead of other
tasks when time is limited? One of Petri’s conclusions was that the incentive system needs to be designed to encourage the registration of good
quality data, which was often not the case in his studies. But he also provided examples of the importance of the informal and emotional side. If
you care, if you feel a social bond and obligation, then you will prioritise.
Following up on that lead, the dependence on shared data that accompanies the use of an ERP system then has to be matched by “real” encounters. In order to care about each other and how they affect each others’
work, users would need to have both heart and mind in the contact – to get
to the point where the others are faces and real people, not just anonymous
124
Exploring Patterns in Information Management
role holders. When the previously anonymous “other” receives a face, you
begin to understand your interaction, not only cognitively, but also emotionally. You begin to care.
Returning to the strategic ERP implementation and one of the large subsidiaries, the people working in Finance found that they became the hub of
the data flows. It was in the finance modules of the application that data
from different parts of the organisation came together, and inconsistencies
became visible. All sloppily entered or erroneous transactions ended up
there sooner or later. As one key user said, Finance became the error-handling station and the data laundry of the organisation. The key user realised
that it would be an impossible task for the people in Finance to be reactive
error correctors. In a proactive manner, she started contacting and visiting
those she could trace to be the sources of different data quality problems,
and made them aware of the problems their present use of the system was
causing. Most of these users in other departments came to adopt a new
view of their system’s use, and the Finance key user became a personification of other users who could be affected by heedless use of the ERP system. But to a large extent, this was a network that revolved around the
Finance key user, described by a manager as “the fount of all knowledge”.
After a year, she had to temporarily transfer out of her department in order
to, by her absence, force the development of new and more multilateral
relationships and routes of contact.
This example shows how contacts can develop out of error handling, but it
is not obvious that they will. In another subsidiary, a centrally placed user
in Finance also reported being the one who detected errors in the transactions coming from other parts of the company. But in her case, the main
error handling method was to learn more about what people in other
departments were doing. Once she reached sufficient understanding of
their jobs, and the parts of the application that they were running, she
could correct the errors they entered. She would have wanted to contact the
people directly, but did not have access to the translation between user
codes and actual identities. In serious cases, she could ask a superior with
access to the codes to tell her who the “culprit” was, so she could contact
them. But in the majority of cases – the minor, everyday mistakes – she
did not see the cumbersome identification and contact process as a viable
option. This shows how fragile the network-building process is.
Different Views on What is Correct Data
Another example from the same company illustrates how error handling can
deepen contacts within a group to the point where the contacts remain after
Westelius
125
the task has been solved. The group charged with specifying the product and
article codes for global use in the system encountered unanticipated problems in agreeing on a common list, leading to a heated debate and long
negotiations. Their initial views on what was to be considered “correct” differed substantially. Implementing the set of codes they had agreed on also
turned out to be difficult, with differences between local lists and differences
between codes in the system, in the warehouses, on forms and in people’s
minds, etc. All these errors led to communication, and in turn to the gradual
development of a shared appreciation of the value of standardisation among
the people in the code group. The people who once formed that group still
find it easy to contact each other regarding diverse work-related matters,
although they are now back in line positions in their respective country.
They believe that had their joint task been easy and uncontroversial, these
strong relationships would not have formed.
Emergent and Designed Networks
So far, the focus has been on emergent networks, but networks develop in
different ways. Drawing an analogy to the analysis of change management
by Orlikowski and Hofman (1997), we could expect to find planned development of networks, emergent networks, and improvisational, opportunitybased development. Support organisations are often designed to become
planned networks. An example given above is the ERP supplier’s international helpdesk, which was intended to be a central node in a network of
users seeking assistance. In that example, it did not manage to fill its
intended role. If we look at actual support-seeking practice, we find a large
portion of emergent networks. Most of the examples above illustrate such.
They may appear to be promising developments, but typically they could
have benefited form organisational support. The improvisational, opportunity based development suggested by Orlikowski and Hofman is a strategy
of identifying and supporting positive developments that take place, and
trying to turn them into organisational practice. Not only allowing, but
actually supporting and sponsoring the initiative of local accountants in
case H above, could be such an example. Had opportunity-based development of networks been a strategy in that industrial group, the initiative
could have been picked up and turned into action much sooner, avoiding
some of the frustration and furthering the concurrent but uncoordinated
sensemaking that was going on.
Had opportunity-based development of networks been a strategy in the
company implementing the ERP system, the proactive network building by
the key-user in Finance could have been identified and made a model (or
126
Exploring Patterns in Information Management
source of inspiration, at least) in other subsidiaries. This would probably
have benefited the subsidiary where the user in Finance was hampered by
user anonymity.
An example of opportunity-based development that did take place in that
industrial group was the establishing of IS co-ordinators – a network of
people who are charged with the task of supporting the co-ordination
between different departments and the diffusion of good ideas and examples. So far, it can be noted that the establishing of this network takes time
and is very dependent on the individuals it consists of, and of their managers’ willingness to commit resources to it. The transition from a fire-fighting mission to the constructive, forward-looking network it was intended
to become, is still largely something to be accomplished.
New and Existing Networks
In the strategic ERP implementation case, the development of valuable
networks was an unexpected bonus. With another frame of mind, it would
have been a central objective. Actor-network theory (ANT) tends to look
at the degree of success of planned change in terms of network development, where the intention of the designers of the change, the inscription of
the ERP system, for instance, is a central concept of the planned network.
A number of writers (for example Latour, 1995; Monteiro & Hanseth,
1995; Hanseth & Braa, 1998; Gäre, 2003) have proposed ANT as a lens –
and used it – to analyse the success or lack of success of information systems implementations. The implementation is viewed as a struggle
between the inscription and programs to promote that inscription, and
translations and counter programs modifying or thwarting the adoption of
the information system (at least in the intended way). Thus, ANT could be
said to combine a planned development and an emergent development perspective. But ANT also proposes that existing networks are important and
play a substantial role in forming the adoption (or non-adoption) of the
intended change.
From the local view of a user charged with using the new ERP system,
coping becomes central. The main question the user poses is not: how has
the designer intended that I should use this information system? It is
rather: how do I learn to cope with this new information system? The other
users close by are one source of inspiration. Trainers who have shown
competence, and with whom one has developed a social bond, or at least
an acquaintance, are another source; the trainer, although it is not her or his
job to answer questions once the training session is over, will often do so.
Westelius
127
Asking Those You Know, and the Problem of Escaping Being
Regarded as Knowledgeable
Who do you ask? It is far easier to ask the competent person than to look
in the manual or try to talk to someone who is charged with answering
questions but is less accessible, is not a personal acquaintance, has
unknown competence… Schultze and Leidner (2002) suggest that “knowledge is a double-edged sword: while too little might result in expensive
mistakes, too much might result in unwanted accountability” (p. 213).
Continuing along that line of argument, displaying knowledge can make
you appreciated and sought after, but such attention can also become a
problem. The key-user in Finance, discussed above, had to transfer out of
her department to escape. At Linköping University, the students who have
taken responsibility for keeping the student lab network and hardware
operational can hardly enter the computer labs without being asked a host
of questions – not because it is their job to answer such questions, or
because the answer cannot be found in the documentation, but because
they are identified as knowledgeable and are accessible. I myself have at
times posed questions to the most knowledgeable in the IT staff or to
someone I have come to know, rather than finding out who may really be
responsible for answering the type of question I want to ask. The social
bonds, the easy way for the one asking, the reluctance of the knower to
appear rude, … they all combine to keep unlicensed routes of communication open.
Inefficient Problem Solving in Existing Networks
Error handling can lead to unexpected solutions – and perhaps unwanted
ones. In the company implementing the ERP system, the spare parts order
receivers had problems communicating properly with packing: on the
screen, they had a field where they could write comments, but after some
time they learned that the packers could only see the first 50 characters.
(Order receivers and packers were almost next door to each other, but did
not typically meet in person.) Attempts to write comments in the address
field led to strange address labels on the packages shipped. Walking over
to the packers and writing on their whiteboard had the disadvantage that it
had no obvious connection to the computer system, where the order
receivers noted, and the packers received, the packing instructions. Finally,
the order receivers managed to develop a communication that served their
needs. Lotus Notes was used for sending free-format instructions concerning shipments, and the comment field in the administrative application
was used to indicate that there was a Lotus Notes message to read.
128
Exploring Patterns in Information Management
Here, order receivers, who felt responsible for correctly dispatching the
spare parts that had been ordered, devised a solution to the problem, using
the channels and tools that they knew of. Had they instead used the formal
communication channel for problems with the ERP system, and had this
channel worked smoothly, the simple solution would have been to change
the parameter controlling the number of characters displayed to the packers. However, this formal communication channel did not work smoothly.
The order receivers had learned that errors and problems reported through
that channel were not very likely to get solved, at least not at short notice.
In this particular instance, it was also known to application designers and
developers at the ERP supplier, that the field was controlled by a parameter that is set at the installation of the program, and that is easy to change.
However, that knowledge was not present at the local level of the support
organisation. Thus, the error (packing information not reaching the packers) set the order receivers looking for a solution, using the ways they
knew and the tools they found that they could influence.
This is an example of an existing network being stronger than the one that
is planned by the project managers and intended for use in a case like this.
The designers of the new support network have not managed to implement
it in such a way that it supplants the existing networks and becomes the
preferred network of the users.
Another example of strong, existing networks is provided in Buck et al.
(2001). They investigated internal communication in a paper and pulp
group, where the new group manager was trying to implement a knowledge management inspired culture of knowledge sharing and communication within the group. The manager also tried to facilitate communication
and knowledge sharing through increased use of IT. Buck et al. noted that
the paper mill operators rather asked their friends in other companies (outside the industrial group) for help and advice when they encountered
problems, than asking unknown (or known) people in their own industrial
group. From a headquarters perspective, this was contrary to organisational
norms, but from an individual perspective it made sense.
Thinking in terms of information system development and information
system implementation projects, rather than in terms of actor-network
building, more easily leads to fragmented cultures (isolated actor-networks) that rely to the greatest part on existing networks and their previous
communication and action cultures. The existing networks provide a
known and developed mode of error handling, and can be expected to
occasion fewer second-level errors and interruptions of the error handling,
than a new and untried network. This results in re-enactment of the exist-
Westelius
129
ing structure and translation of the new tools aimed at minimal change of
the present structure, rather than modification of the existing structure to
accommodate (and explore) the new impulses and ideas. Thus, when we
try to create rational communication channels, support functions, we often
underestimate the strength of extant networks and the importance of trust.
Perhaps it would make more sense to figure out what networks are already
in existence, and try to supplement them with some new nodes and supply
them with relevant knowledge and resources, rather than try to build
entirely new networks. And perhaps it makes sense to try to channel the
attention and energy provoked by first-level errors into such network
improvement efforts. Perhaps, after all, errors can be useful in helping
users to learn.
References
Bergum, S. (2000) Managerial communications in telework, Linköping Studies in
Science and Technology, Thesis; 807, Linköping University, Linköping, Sweden.
Buck, E., Castevall, J., Dunér, M., Eding, D. & Eklöw, M. (2002) Rottneros
interna kommunikationsstrategi ur ett KM-perspektiv, report in MSc program
in course Strategic Applications of IT, Sektionen för industriell ekonomi, Linköpings Tekniska Högskola 2001-09-28 http://i98daned.island.liu.se/TDEI55/,
accessed 2001-09-29.
Gäre, K. (2003) Tre perspektiv på förväntningar och förändringar i samband med
införande av informationssystem, Linköping Studies in Science and Technology, Dissertation; 808, Linköping University, Linköping, Sweden.
Hanseth, O. & Braa, K. (1998) “Technology as traitor: emergent SAP infrastructure in a global organization”, Proceedings of the Nineteenth International
Conference on Information Systems, Helsinki, December 13-16, 1998, pp. 188196.
Latour, B. (1995) “Social theory and the study of computerized work sites”, in
Orlikowski, W., Walsham, G., Jones, M.R. & DeGross, J. (Eds.) Information
Technology and Changes in Organizational Work, Chapman and Hall, London.
Moberg, A. (1997) Närhet och distans: studier av kommunikationsmönster i satellitkontor och flexibla kontor, Linköping Studies in Science and Technology,
Dissertation; 512, Linköping University, Linköping, Sweden.
Monteiro, E. & Hanseth, O. (1995) “Social Shaping of Information Infrastructure:
On Being Specific About the Technology”, in Orlikowski, W., Walsham, G.,
Jones, M.R. & DeGross, J. (Eds.) Information Technology and Changes in
Organizational Work, Chapman and Hall, London.
130
Exploring Patterns in Information Management
Oatley, K. & Johnson-Laird, P. (1987) “Towards a cognitive theory of emotions”,
Cognition and Emotion, Vol. 1, No. 1.
Orlikowski, W.J. & Hofman, J.D. (1997) “An Improvisational Model of Change
Management: The Case of Groupware Technologies”, Sloan Management
Review, Vol. 38, No. 2, pp. 11-21.
Petri, C.-J. (2001) Organizational Information Provision: Managing mandatory
and discretionary utilization of information technology, Linköping Studies in
Science and Technology, Dissertation; 720, Linköping University, Linköping,
Sweden.
Salaway, G. (1987) “An Organizational Learning Approach to Information Systems Development”, MIS Quarterly, Vol. 11, No. 2, pp. 244-265.
Schultze, U. & Leidner, D.E. (2002) “Studying Knowledge Management in Information Systems Research: Discourses and Theoretical assumptions”, MIS
Quarterly, Vol. 26, No. 3, pp. 213-242.
Simon, H. A. (1945, 1997) Administrative Behavior: A Study of Decision-Making
Processes in Administrative Organizations, 4th Edition, Free press, New York.
Weick, K. (1995) Sensemaking in organizations, Sage, Thousand Oaks, California.
Weick, K. (2001) “Commentary”, Reflections – the SoL Journal on Knowledge,
Learning and Change, Vol. 2, No. 4, pp. 55-56.
Westelius, A. & Westelius, A-S. (1990) Decentraliserade informationssystem, två
fallstudier inom ekonomistyrning, EFI, Stockholm School of Economics,
Stockholm.
Westelius, A. (1996) A study of patterns of communication in management
accounting and control projects, EFI, Stockholm School of Economics, Stockholm.
Westelius, A. (2001) “On classical music and business – listening, leading, learning”, Reflections – the SoL Journal on Knowledge, Learning and Change, Vol.
2, No. 4, pp. 47-55.
PART THREE:
MODELS AND FRAMEWORKS
FOR IT-RELATED CHANGE
Blanksida
—9—
IT Projects and the X Model
Erling S. Andersen
Åge Sørsveen
IT Project Failures
Many IT projects fail. We read about failures all the time, in newspapers,
trade magazines, and scientific papers. Research undertaken by Standish
Group International (1999) concluded that out of 23,000 projects, 28 percent failed completely, 46 percent were characterized by cost and time
overruns and only 26 percent succeeded.
Ewusi-Mensah (1997) points out that the cancellation of projects can be
attributed to a combination of several factors, including the following:
• Projects goals: lack of general agreement on a well-articulated set of
project goals and objectives
• Project team composition: weak or problematic project team
• Project management and control: bad management, poor decisions, lack
of IS to measure progress and identify risks
• Technical know-how: team not capable of the task, lack of expertise
and experience, not relevant application-domain knowledge
• Technology base or infrastructure: the current infrastructure is not satisfactory for the kind of project
• Senior management involvement: monitoring of progress and making
decisions are deferred to technical experts
• Escalating project cost and time of completion: not addressed before
crisis stage
Conradi (1997) has estimated that failed IT projects gave the Norwegian
government a loss of 2,500 million NOK during the first half of the 1990s.
The most spectacular failure project was TRESS-90, an IT-system for
Norwegian Social Security (Rikstrygdeverket), which was delayed for five
years and lost approximately 1,200 million NOK.
134
Exploring Patterns in Information Management
An investigation project (Pitfalls and Success Criteria of Large Public IT
Projects) was set up as a consequence of all the failures. Its task was to
collect the experiences from large public IT projects. It should draw lessons from the failures and propose actions to prevent similar problems in
the future.
The report (Statskonsult 1998) summarized the problems the following way:
• The project is not tied to a business plan or an IS/IT strategy
• Unrealistic goals and too ambitious; too little focus on one’s own ability to implement
• Unclear lines of responsibility
• Too big and complex a project; the project is not divided into smaller
deliverables
• Poor project management and control; lack of resources and abilities to
handle unforeseen events
• Poor contracts and contracts are not used as a management tool
• Plans and estimates made on a poor basis, combined with high risk
• Too much focus on technology; underestimating the importance of
organizational and competence development
• Not the right choice of technology: too new or too old
• Lack of line or top management attention
• Large changes in specifications of requirements all the time, without
any procedures for handling changes
• Too little attention paid to the possibility of outsourcing; if used, not
careful enough
Existing Approaches Project Control
We see that project success is not easy to achieve. It would be very helpful
for project managers in critical stages of their work to be able to assess
whether they are on the right track and, ideally, to receive early warning
signals if the wrong course has been chosen.
Traditional project control is focused on performance, costs and quality,
and instruments such as Earned Value Analysis and Critical Ratio are
Andersen & Sørsveen
135
important tools for the project manager (Meredith and Mantel 1995,
Fleming and Hoppelman 1996). This type of control will continue to be
imperative, but the problem is that they are based on retrospective information; all data tell us what happened yesterday.
The Project Implementation Profile (PIP) of (Pinto and Slevin 1988) made
a valuable contribution to the field of project management insofar as they
demonstrated how to use critical success factors to diagnose a project’s
status. PIP or similar approaches should be used as a supplement to traditional monitoring of projects (Andersen and Jessen 2000).
Requirements for a General
Project Evaluation Model
The weakness of most existing models for project control is that the model
itself determines what type of problems you focus on. As a consequence,
the model does not give the complete picture of the project, nor an unbiased starting point for analysis. Models for cost control naturally focus on
describing and discussing costs; quality assurance models concentrate on
how good the quality is, and so on. However, as shown, research has identified very mixed reasons for why projects fail.
Consequently, there is a need for a general model, which does not predetermine the type of problem the project is facing. The model should allow
for descriptions of all aspects of a project and in that sense provide a solid
base for further analyses of the situation. Every model must of course be
based on some kind of structure, but the structure should not decide which
problems the project has to tackle.
A good way to ensure that all the important aspects of a project will be
scrutinized is to have project team members and users describe the actual
situation of the project without any restrictions as to what they are allowed
to bring up. This means that the modelling technique should be so simple
that all involved should be able to make a their own description after a
short introduction.
This paper will present a model (fulfilling the requirements above) that can
be used for describing and analyzing the present situation of a project. We
will first present the model itself and then give examples of its use, both on
the individual project level and on a more aggregated level.
136
Exploring Patterns in Information Management
The X Model
The candidate for a general evaluation model is called the X-model (Figure
1) (Andersen et al. 1994). It consists of five elements: Personal inputs,
Factual inputs, Work processes, Personal outputs, and Factual outputs. The
name reflects its shape.
Personal
inputs
Factual
inputs
Work processes
Personal
outputs
Factual
outputs
Figure 1: The X Model
The personal inputs and outputs are the members of the project organization and their attitudes, needs, knowledge, skills, experience and relations to others. This is what may be called the “soft part” of the project
organization. The inputs refer to the situation at the start of the project or
a previous situation, while the outputs are related to the present situation.
The factual parts of the model focus on the more formal or structural part
of the project. The inputs may describe the tasks to be performed, the
problems to be solved or the challenges to be met, the project plans and the
formal organization. The outputs should show what the project has
achieved so far and what has not been accomplished.
The work processes are the project activities (in groups, meetings or individually), the decision processes, the communication processes and the
general working climate. Processes integrate both personal and factual
aspects and it is meaningless to make a distinction between the two in a
description of the present status of a project.
The project manager may at certain stages of the project have an X
model created. The best result is achieved when several participants first
Andersen & Sørsveen
137
make their own independent descriptions of the project and decide which
aspects of the project they will focus on. The participants then co-operate
in combining the individual descriptions into a common description. This
model is then used for analyses of the connections between outputs, work
processes and inputs. This insight into the project situation will help the
project manager and the project team to decide how to proceed to better
the functioning of the project. The subsequent actions might be quite different from the results of a traditional project control.
The Theoretical Justification of the X Model
The X model is based on two theoretical approaches, which may help us
understand how an organization functions: systems theory and socio-technical theory. The X model combines these two approaches in a single and
consistent framework.
Systems theory focuses on an entity and its parts and helps us to clarify
how the different parts are interrelated (Langefors 1966). An organization
may be regarded as a system with inputs, transformation processes and
outputs as the main parts of the system. Seiler (1967) used systems theory
in this way to design an elementary framework for diagnosing human
behavior in organizations. His organizational system distinguished between Inputs, Actual behavior and Outputs. However, he did not develop
this framework into an easy-to-use modeling technique.
The input-transformation-output model is a causal model (Emery and Trist
1965). The arrows of the X model imply causal relationships. The outputs
are the results of the transformation processes. They are based on the
inputs available.
The socio-technical school (Mumford and Weir 1959), also inspired by the
systems theory, divides the organization into a social and a technical subsystem. The concepts of social and technical subsystems are broadly
defined: technical subsystems also cover economic and commercial
aspects; social subsystems incorporate every human aspect that may be of
interest to a person in a work situation. These concepts are similar to what
is called factual and personal in the X model. The main idea of the sociotechnical school is that the social and technical subsystem cannot be
regarded as isolated systems in a study of an organization; we have to consider the relationships between them.
138
Exploring Patterns in Information Management
An X Model for an IT Project
The purpose of the X model is firstly to describe the present situation of a
project. An example, derived from our collected sample of X models, is
shown as Figure 2. It is a very simple X model made by one of the project
team members. The X model is made after the project has reached its first
phase of deliveries.
Personal inputs
Factual inputs
• Strongly committed team
• Common understanding of
project’s mission
• Different attitudes to time
schedule and quality
• Not very good competence
in project work
•
•
•
•
•
Work processes
• PM not always present
• PM vague in
communications
• Weekly meetings
• Increased knowledge of
project work
• Increased experience in
co-operation
co• Frustration and
resignation among team
members
Personal outputs
•
•
•
•
Unrealistic time schedule
Goals not measurable
No steering committee
Bad project start -up
External PM
• No teambuilding
• No focus on quality
• Decisions taken without
participation of whole team
• Tools not used for project
control
Phase 1 fulfilled
Delays
Quality not as expected
Costs overrun
Factual outputs
Figure 2: The X Model of an IT project
We would claim that the X model gives an interesting view of the present
situation of the project and highlights aspects that would not have been so
obvious if only traditional control methods have been used.
The X model would secondly be the starting-point for an analysis of the
project. We see that the actual status of the presented project is such that
actions are necessary. The X model should help us to identify the causalities between the personal and factual outputs on the one hand and the work
processes and the personal and factual inputs on the other (Figure 3).
Andersen & Sørsveen
Personal inputs
Factual inputs
• Strongly committed team
• Common understanding
of project’s mission
• Different attitudes to time
schedule and quality
• Not very good
competence in project work
• Unrealistic time schedule
• Goals not measurable
• No steering committee
• Bad project start -up
• External P M
Work processes
• PM not always present
• PM vague in
communications
• Weekly meetings
• Increased knowledge of
project work
• Increased experience in
co-operation
• Frustration and
resignation among team
members
Personal outputs
139
• No teambuilding
• No focus on quality
• Decisions taken without
participation of whole team
• Tools not used for project
control
• Phase 1 fulfilled
• Delays
• Quality not as expected
• Costs overrun
Factual outputs
Figure 3: An X Model with Causal Relationships
The X model shows that the quality is not as good as expected (factual
output). We may hypothesize that this is caused by the lack of focus on
quality and that all project members are not involved in the decisions
(work processes). This might be caused by the attitudes to quality (personal input). When conducting a further causal analysis there may be a
need for more detailed descriptions of the conditions presented by the
model.
The results of the description of the present situation and the causal analysis provide the background for actions to be taken to correct or better the
present situation. An action list has to be worked out. Later descriptions of
the present situation will demonstrate if the actions were adequate.
The Norwegian Project Scene Based on X Models
Participants in the part-time Master’s program in project management at
the Norwegian School of Management BI have to write a thesis as a pro-
140
Exploring Patterns in Information Management
gram requirement. The thesis should be based on their observations of a
real project over a period of nine months. Many students use the X model
to evaluate their project, and in this way we have been able to collect X
models for 74 different Norwegian projects. The projects are not restricted
to IT projects, even if many belong to this category.
The students are employees from companies, non-profit organizations and
government agencies. Usually three students write their thesis together and
they typically choose to observe a project among their own companies.
Sometimes one student is also a participant in the chosen project or the
project manager. They collect information about the project by observing
and interviewing project team members and people affected by the project
and by studying plans and other written material. This means that they
have good access to information. The students use all their information
about the project to establish an X model, where the present situation usually refers to a point in time before the project is finished.
The empirical data will allow us to study which aspects of the personal and
factual inputs, the work processes and the personal and factual inputs are
regarded as the most important in order to understand the status of the
project. All the collected X models enable us to get a comprehensive picture of Norwegian projects, as described and experienced by people
familiar with the projects. The X model contains descriptions made by
people in their own words, but in order to have them processed we have to
perform a coding of all the statements.
Two persons coded the X models independently of each other, using a
coding scheme covering many possible statements from the respondents.
The coders had to decide which code best covered the intention of the
original statement and whether it was complied with completely, partly or
not at all. If there was a difference of coding between the two, a third person acted as judge and decided on the final coding.
Table 1 shows which aspects of the project situation were most focused on.
For each of the five elements of the X model we present the four most frequently mentioned aspects. We may say that these are the aspects of project work that project members are most focused on when they are told to
explain the situation of a project. The majority agrees that it is very
important to have as the focus of attention the motivation of the team
members, the clarity of the goals, the feedback given in the communication
process, if the motivation is upheld and if the mandate seems to be fulfilled.
Andersen & Sørsveen
141
No. of
responses
Yes
Partly
No
(%)
(%)
(%)
Strongly motivated for the project (PI)
67
70.1
16.4
13.4
Good knowledge of subject area of the
project (PI)
60
56.7
26.7
16.7
Good knowledge of project work and project
methods (PI)
51
54.9
29.4
15.7
Extensive experience with project work,
methods and tools (PI)
49
24.5
32.7
42.9
Clearly expressed project objectives/goals
(FI)
56
30.4
19.6
50.0
Appropriate organization/clear lines of
responsibility (FI)
53
18.9
24.5
56.6
Good plans/clear time schedule/fixed milestones (FI)
45
26.7
15.6
57.8
Enough resources allocated for the project
work (FI)
45
26.7
15.6
57.8
Good feedback (WP)
60
28.3
48.3
23.3
Good co-operation between project team
members (WP)
53
47.2
24.5
28.3
Good management/leadership (WP)
49
34.7
26.5
38.8
Good project control (WP)
40
20.0
22.5
57.5
Strongly motivated for further project work
(PO)
48
27.1
27.1
45.8
Increased knowledge in general (PO)
37
73.0
13.5
13.5
Increased knowledge of subject area of the
project (PO)
34
79.4
8.8
11.8
Increased knowledge of project work and
project methods (PO)
31
54.8
12.9
32.3
The mandate/charter/contract are fulfilled
(FO)
39
59.0
12.8
28.2
Completion as scheduled (FO)
38
21.1
10.5
68.4
Plans are followed/milestones or phases
are achieved as planned (FO)
36
25.0
5.6
69.4
Balanced results achieved (FO)
31
12.9
19.4
67.7
Table 1: The most frequently used statements of the five elements of the X model
The main results from Table 1 are also depicted as an X model (Figure 4).
142
Exploring Patterns in Information Management
Table 1 and Figure 4 tell us how Norwegian projects work in general: they
have motivated and well-informed team members who sometimes lack
project experience. Only 30% of the projects have clear goals. For the
majority of the projects there are no clear lines of responsibility, the plans
are inadequate and not enough resources are allocated to the project. The
feedback to project team members is not good enough, but for the most
part the co-operation between them functions well. The quality of the leadership differs and project control is rather weak. People learn a lot from
project work, but their motivation for further project work seems to be
reduced. The task is completed as described by the mandate or project
charter, but most projects are not completed on time and within the budget.
The results are more focused on technical than social factors.
Personal inputs
Factual inputs
• Strongly motivated for the
project
• Good knowledge of subject
area of the project
• Good knowledge of project
work and project methods
• No extensive experiences
with project work, methods
and tools
• Project objectives/goals not
clearly expressed
• No appropriate
organization/no clear lines of
responsibility
• No good plans/no clear time
schedule
• Not enough resources
allocated
Work processes
• Medium feedback
• Good co-operation between
project team members
• Not strongly motivated for
further project work
• Increased knowledge in
general
• Increased knowledge of
subject area of the project
• Increased knowledge of
project
Personal outputs
• Not so good management /
leadership
• Poor project control
• The mandate / charter /
contract are fulfilled
• Completion not as scheduled
• Plans are not
followed/milestones or
phases are not achieved as
planned
• Not a balanced pso result
Factual outputs
Figure 4: The most typical situation for Norwegian projects
One of the main intentions of the X model is to perform a causal analysis,
which means investigating why the results are as they are and what can be
done in order to improve the situation. Our data give us an opportunity to
Andersen & Sørsveen
143
study which input factors and work processes influence the different output factors. Figure 4 looks at the personal and factual output factors, which
were most focused on. The table shows which input and transformation
factors are significantly correlated with these output factors.
Strongly motivated for
further project work
(PO)
The mandate/charter/
contract is fulfilled
Correlated with
Correlation
Sig.
N
Strongly motivated for project
work (PI)
.309
.044
43
Budget/costs determined (FI)
.843
.000
14
Good co-operation among project
team members (WP)
.426
.012
34
Meeting schedules are followed/meetings are conducted in
a good way (WP)
.475
.022
23
Good feedback (WP)
.360
.019
42
Good information (WP)
.529
.020
19
Extensive experience with project
work, methods and tools (PI)
.391
.048
26
Strongly motivated for project
work (PI)
.449
.048
26
Clearly expressed project objectives/goals (FI)
.379
.042
29
Good plans/clear time schedule/fixed milestones (FI)
.578
.008
20
Good co-operation among project
team members (WP)
.448
.037
22
Good co-operation with base
organization/users (WP)
.722
.005
13
Good management/leadership
(WP)
.438
.032
24
Good working processes (WP)
.611
.016
15
Table 2: The two most frequently used output statements and their correlations to
other statements
Table 2 shows that different factors affect the two chosen output factors.
Motivation for further project works depends on how well the team members are kept informed and given feedback as well as the quality of the
project meetings and the co-operation within the team. Strong motivation
at the start and predetermined budget help to keep up the motivation.
144
Exploring Patterns in Information Management
Motivation is also of importance in fulfilling the mandate. Other important
factors are good co-operation within the team and with the base organization, good management, experienced people, good plans, and clear goals.
Conclusions and Further Work
This paper has presented the X model. It is a tool designed primarily for
evaluating (describing and assessing) an individual project. We would
claim that the X model should be part of the toolbox of every project manager. Further research should be conducted to determine the strength and
weaknesses of the X model compared to other tools for project control or
evaluation.
We have been able to collect a substantial number of X models for Norwegian projects. The data show that many projects have serious shortcomings
from the start such as vague goals and unclear lines of responsibility. Only
about 20% are completed as scheduled. Our data show which factors to
concentrate on to obtain better project results, and to achieve good results
on several output factors a broad range of factors have to be improved.
A Personal Note of Acknowledgement
The X model as a concept was born in 1977 when Mats Lundeberg and the
authors of this article were closely associated in the development work
performed at the Institute for Individual and Organizational Learning in
Oslo. Responsible for the progress of developing the model into a practical
tool have been Åge Sørsveen and Ingeborg Baustad, assisted by Erling S.
Andersen.
Important stages in the development during the 80s and 90s have taken
place in Stockholm, where Mats Lundeberg with great hospitality has
placed himself and his institute at our disposal. To have such a place of
sanctuary is invaluable.
References
Andersen, E.S., Baustad, I. & Sørsveen, Å. (1994) Ledelse på norsk. {Leadership
– The Norwegian Way}, Ad Notam Gyldendal, Oslo.
Andersen, E.S. & Jessen, S.A. (2000) “Project Evaluation Scheme”, Project Management, Vol. 6, No. 1, pp. 61-69.
Andersen & Sørsveen
145
Conradi, R. (1997) “A Revised Agenda for SW Process Support” European Conference of Object-Oriented Programming (ECOOP’97), Jyväskylä, Finland, 9
-11 June.
Emery, F.E. & Trist, E.L. (1965) “The Causal Texture of Organizational Environments”, Human Relations, Vol. 18, pp. 21-35.
Ewusi-Mensah, K. (1997) “Critical Issues in Abandoned Information Systems
Projects”, Communications of the ACM, Vol. 40, No. 9.
Fleming, Q.W. & Hoppelman, J.M. (1996) Earned Value Project Management,
Project Management Institute, Upper Darby, Pennsylvania.
Langefors, B. (1966) Theoretical Analysis of Information Systems, Studentlitteratur, Lund, Sweden.
Meredith, J.R. & Mantel Jr,, S.J., (1995) Project Management. A Managerial
Approach, John Wiley & Sons, Inc., New York.
Mumford, E. & Weir, M. (1959) Computer Systems in Work Design – the ETHICS
Method, Associated Business Press, London.
Pinto, J.K. & Slevin, D.P. (1988) “Critical Success Factors Across the Project Life
Cycle”, Project Management Journal, Vol. 19, No. 3, pp. 67-75.
Seiler, J.A. (1967) Systems Analysis in Organizational Behavior, Richard D. Irwin
and Dorsey Press, Homewood, Illinois.
Standish Group International (1999) CHAOS: A Recipe for Success, Downloaded
from http://www.standishgroup.com/sample_research/index.php
Statskonsult (1998) Erfaringer fra store statlige IT-prosjekter. Vurderinger og
mulige tiltak {Experiences from large public sector IT projects. Evaluation
and possible actions} Oslo.
146
Exploring Patterns in Information Management
— 10 —
Implementation of eBusiness Models
– the MTO-Framework1
Niels Bjørn-Andersen
Helle Zinner Henriksen
Michael Holm Larsen
Introduction
Venture capitalists typically require that you can explain the business
model in the time it takes the lift to get to the tenth floor. Implementation
typically takes years. There is a disproportionately large amount of focus
on what constitutes an innovative new business model compared to implementation, since most e-business failures are attributed to failures in
implementation.
Few researchers in the Nordic countries have been as influential as Mats
Lundeberg when it comes to business transformation using IT, especially
the problem/business-oriented side of IS/IT. From the early 70s, Mats
Lundeberg has contributed to theoretical developments as well as to
empirical implementations. Think for example of the ISAC method (Lundeberg, 1971; Lundeberg & Andersen, 1974), but also of much more practical guidelines like his textbook ‘Handling Change Processes’ (1993). It
was therefore natural for us, in honour of his large contributions within this
field, to choose the topic of implementation of e-business models.
The purpose of this paper is to develop an integrated approach for implementation of eBusiness models based on a taxonomy including four very
different approaches to eBusiness implementation/adoption. These
approaches are:
• Traditional IS/IT implementation insights especially as these were conceived until the late 80s.
1
The research leading to this framework has been coordinated as part of the
‘eFactors Network of Excellence’ funded under the FP-6 of the EU contract number IST 2001-34868. Please consult www.e-factors.net for further information.
148
Exploring Patterns in Information Management
• Business process reengineering guidelines for implementation dating
from the early to mid 90s.
• Technology diffusion and adoption theory starting with the earlier work
of Rogers, but updated in the late 90s.
• Venture capital guidelines for eBusiness ventures from late 90s.
It is quite clear that these approaches are very different. There is also evidence that none of them is sufficient to secure a successful implementation. But it is our contention that each of them holds a piece of the truth.
Accordingly, the paper will contrast the four bodies of knowledge and
develop a multi-perspective taxonomy for implementation of eBusiness.
The organization of the paper is as follows. First, definitions of the two
key terms, eBusiness models and implementation, are presented. The subsequent sections contain a presentation of our design approach for the
implementation framework, an outline of the key implementation factors
derived from reviews of literature and our implementation framework for
eBusiness models, and a discussion of the validity and applicability of the
eBusiness implementation framework. The final section presents the conclusion and recommendations for future research.
Basic Definitions
This section presents our basic understanding of eBusiness models and
implementation.
e-Business Models
Whether the company is a new venture or an established player, a good
business model is essential to every successful organization (Magretta
2002).
In this paper we adopt the following definition of an eBusiness model: “An
architecture of product, service, and information flows, including a
description of the various business actors and their roles; as well as a
description of the potential benefits for the various business actors, and a
description of the sources of revenues”, cf. Timmers (1998:4). This definition frames the discussion of implementation factors and initiatives.
An underlying assumption of this paper is that the characteristics of eBusiness models call for research on the rethinking of the basis of implementation. Compared to earlier information systems, there are a number of rea-
Bjørn-Andersen, Henriksen & Larsen
149
sons why traditional implementation models, theories, and methodologies
do not hold any longer, and a multi-perspective is called for:
• New stakeholders. Venture capitalists rather than internal business
management make decisions about new systems, and their decisions are
based on very different criteria from the traditional business unit manager due to limited knowledge of market, wish for fast exit strategies,
etc.
• Interorganizational nature of eBusiness. eBusiness systems are interorganizational, covering multiple organizations where there are no ultimate decision maker. This requires much more comprehensive analyses
of competition, markets, value chains and networks, collaboration etc.
It is not enough to have the commitment of a dedicated member of topmanagement.
• Time compression. Traditionally it could take years to develop IS/IT
systems. Today many eBusiness systems are developed in time-slots of
weeks or months.
• Interactive systems development. Waterfall and modified versions of
waterfall models have given way to much more iterative systems
development relying a lot on prototyping, testing and continuous development.
An illustration of the need for multi-perspective business analysis and that
more stakeholders may influence the decision making of the implementation process is presented in Figure 1.
Venture Capitalist
Developer
Top Management
Business partners
Process manager
Operator
Client
Figure 1. Overview of implementation actors during the life-cycle of an eBusiness
system
150
Exploring Patterns in Information Management
Assuming that the horizontal axis represents a time scale, the figure also
illustrates that the different actors in principle have different and/or overlapping time-periods in which they are actively involved in the design and
implementation process.
Definition of Implementation
Rogers (1995) argues that the implementation stage ends when the new
idea becomes an institutionalized and regularized part of the adopters
ongoing operations. Any systems development project may be seen as consisting of three rather different sets of activities: requirement specifications, design and implementation. But implementation is not a particular
stage occurring after a design stage. Instead we subscribe to the view that
implementation is a set of activities starting almost at the very beginning
of any eBusiness project and continue as Rogers suggests above, until the
solution has been adopted and fully integrated not just in the target organization developing the eBusiness solution, but also for everybody else in
the value network related to and affected by the solution. This is illustrated
in Figure 2.
Demand specification
Development
Implementation
Figure 2. The Process of Organizational Implementation
The horizontal axis represents time, whereas the vertical axis represents
the amount of efforts dedicated to particular activities, i.e. demand specification, development and implementation.
Common for almost all conceptualizations of the term implementation is
that some degree of organizational action has taken place. This requires
different degrees of commitment and a large variety of actions until the
intended benefits are realized as a successful implementation (Gottschalk,
1999).
DeLone and McLean (1992) found that the most common IS implementation success factors were system usage and user satisfaction, but these are
clearly too limited when considering eBusiness systems, where adoption
Bjørn-Andersen, Henriksen & Larsen
151
by other organizations in the value network is of key importance. Indicators like number of visits to web-sites, revenue, execution etc. are other
key performance signs crucial to implementation success in relation to
eBusiness models.
Linton (2002) found in his review of ten years of implementation literature
that implementation success could be traced back to five factors: organizational structure, technology, project management, divisibility, and social
interactions. Social interaction is especially important and has been centre
stage for many implementation researchers. For example Tornatzky and
Fleischer (1990) claim that implementation success can be assessed by the
degree of interaction with other technologies within the organization. That
is especially the case when considering implementation of eBusiness models in existing organizations, where the integration of eBusiness models
with existing ERP-systems are of great importance if the organizations are
to derive benefit from their eBusiness solutions. Furthermore, the integration with systems in other organizations and between organizational units
in the different organizations is absolutely necessary for the success of an
eBusiness.
Methodology for Developing the Implementation
Framework
The relevant implementation factors are derived through a literature review
of the four influential areas of expertise/research: venture capitalists experiences, business process reengineering (BPR), diffusion theory, and systems development. From articles within these four areas, the individual
implementation factors are selected if they have been found to have a normative value, and if they are cited within the body of knowledge of the
specific expertise/research area.
The implementation factors are then compared across the four areas, and
identical factors are eliminated in order to appear only once. All factors are
then clustered into groups of factors with similar characteristics in order to
provide an overview.
Hence, the factors are shown in a matrix presenting the various expertise/research areas. This matrix provides the basis for the resulting implementation framework where technological, organizational and market
clusters represent our clustering of the factors. This framework is presented in the following section where the four reference disciplines are
presented side by side and structured according to the three clusters.
152
Exploring Patterns in Information Management
Contributions of Implementation Factors
from the Literature Review
Contributions of Implementation Factors
from Venture Capitalists Experiences
The contribution from venture capitalists to eBusiness projects has flourished especially during the dot-com period, and have increased dramatically in numbers during the last 5 years. Venture capitalists (VC) have
always belonged as an integrated part of the financial sector which is
reflected in the implementation factors they emphasize.
Little research literature prescribes the prerequisites demanded from the
venture capitalists, although general guidelines on what incubators and
venture capitalists may offer in services is found in literature, cf. Hansen et
al. (2000). Hence, insight from venture capitalists is found from alternative
sources.
A typical example of a VC inspired insight was provided by T. Forcht
Dagi (2001), MD of Cordova Ventures, and professor at The Georgia
Institute of Technology, argued that the pitfalls that led many start-up
companies to fail were numerous and caused by wrong interpretations and
insufficient planning and analysis of the eBusiness model. Among others,
the true costs of starting and running business were not understood, especially since the eBusiness market suffered from weak barriers to entry. The
Internet does not, per se, provide any sustainable competitive advantage,
and customers may to a large extent be reluctant to purchase due to security issues, over-exposure, and unpleasant customer experience. Furthermore, revenue models were often flawed, cash flow from financing
eclipsed cash flow from operations, and business plans had poor strategic
vision.
Noble (1999) suggests a model that divides an implementation into four
stages. The stages are pre-implementation, organizing the implementation
effort, managing the implementation process, and maximizing cross-functional performance. The focus of the model is on cross-functional issues and
dynamics. This is why it is relevant to think in relation to implementation of
eBusiness models, which contain the same characteristics. The research of
Noble (1999) provides critical success factors (CSFs) for each implementation stage from a managerial point of view. The “managerial levers”, cf.
Noble (1999:25), as the CSF’s are named, provide insight from a research
conducted through executive interviews and middle manager surveys with
Bjørn-Andersen, Henriksen & Larsen
153
respect to goals, organisational structure, leadership, communications, and
incentives. A key learning point of this framework is that the management of
these factors changes through the implementation stages.
Lazer & Livnat (2001) suggest a five-step evaluation process of eBusiness
models that is materialised in specific questions regarding the economic
viability of the eBusiness model. These are:
• What market failures and transaction costs are addressed by the business model?
• How effective can the e-commerce firm be in reducing the market failures or transaction costs?
• Will the e-commerce company be able to expropriate benefits from
customers?
• What are the necessary resources to conduct the business?
• Can competitors erode profits?
De et al. (2001) suggest a micro economic perspective on evaluating
eBusiness models emphasizing traditional areas such as transaction costs,
switching costs, network externalities and product versioning. In addition
to this the authors suggest that successes and failures of eBusiness models
also need to be evaluated based on infrastructure investment models, user
experience models, and models for revenue generation in order to reveal
the inherent complexity of conducting electronic business.
Contributions of Implementation Factors from BPR
The concept of Business Process Reengineering (BPR) was originally
coined by Hammer (1990). The focus was very radical, emphasizing radical organizational changes through obliteration of activities instead of (or
before) automating the activities. The approach was later softened and
focused on redesign of processes as the driving force in reengineering
projects, cf. (Hammer 1996; 1999). The reengineering activities during the
early 90s primarily addressed intra-organizational initiatives, cf. e.g.
Hammer & Stanton (1995) and Keen (1997), but were later extended in
scope to network redesign by focusing on inter-organizational redesign
projects, e.g. Keen & McDonald (2000) and Hammer (2001).
One of the key issues in implementing BPR projects according to Hammer
& Champy (1993) is that reengineering success depends on addressing the
full business system diamond (Hammer & Champy 1993; Champy 1995).
This identifies the relationship between business processes, jobs and struc-
154
Exploring Patterns in Information Management
tures, management and measurement systems, and values and beliefs. When
restructuring the business process, the content of jobs and of organisational
structures changes for all employees. Changing jobs and structures requires
changes in management principles and performance measurement systems.
These new management principles and performance measurement systems
induce change in values and beliefs, which in turn enable the new business
processes. Consequently, reengineering is not complete until all elements of
the business system diamond have been changed and aligned (e.g. Larsen &
Leinsdorff 1998), which is a process that may be undertaken iteratively in
order to gain the buy-in, acceptance and appreciation from the employees
involved (Larsen & Bjørn-Andersen 2001).
Moreover, alignment of the business processes with the business strategy
is considered important (Tinnilä 1995; Clemons et al. 1995; Sarkis et al.
1997; Lockamy & Smith 1997) as well as alignment with the information
technology strategy. Hence, recruitment of the necessary skill-base and
training are vital for BPR-project success, cf. Bashein et al. (1994) and
Martinez (1995). In addition to this, scoping the BPR-projects (Hall et al.
1993), assuring learning processes (Galliers 1997) as well as shared values
(Grover et al. 1995) are crucial for obtaining radical results.Change management emphasizing communication, training and handling of political
controversies is important in order to maneuver in a highly political landscape of a BPR project (McElrath-Slade 1994; Taylor 1995; Davenport
1995; Homa 1995). Finally, most authors agree that all BPR efforts are
unlikely to reach success unless the top management is committed, supported and engaged in the activities (e.g. Davenport & Short 1990; Bashein
et al. 1994; Willcocks & Smith 1995).
Contributions of Implementation Factors from Diffusion Theory
The diffusion theory is not specifically targeted at adoption, implementation, and diffusion of IT. However, the theory is relevant to any technological innovation, and researchers within MIS (e.g. Premkumar et al.,
1994; Ramamurthy et al., 1999; Ramamurthy & Premkumar, 1995; Cooper
& Zmud 1990) have often used this perspective when defining normative
guidelines for successful implementation of IT. These sources are used as
guidance in the description of the key factors influencing successful IT
implementation in organizations from a diffusion perspective.
According to the diffusion school of thought, implementation is when a new
practice is put into use (Marble, 2000). Implementation therefore involves
behavior change in the organization (Rogers, 1995). A more specific defini-
Bjørn-Andersen, Henriksen & Larsen
155
tion targeting organizational IT implementation is given by Kwon and
Zmud, who claim that organizational IT implementation is “the managerial
concerns focusing on the effective diffusion of information technologies into
organizations, business units, and work groups” (Kwon & Zmud, 1987).
Cooper and Zmud (1990) defined IT implementation as “an organizational
effort directed toward diffusing appropriate information technology within a
user community.” The means for “the diffusion of information technology”
according to this line of thought are presented in the following.
The factors influencing implementation represent a broad variety of
themes. Researchers within diffusion theory have presented useful classifications of the numerous factors (e.g. Kwon & Zmud, 1987; Tornatzky &
Fleischer 1990; Premkumar & Ramamurthy, 1995). In this context the
Kwon and Zmud (1987) taxonomy identifies the following factors: 1)
Characteristics of the user community influencing implementation, e.g.,
commitment to change, education, social approval, degree of understanding of the technology. 2) Characteristics of the organization influencing
implementation, e.g., organizational structures, management support,
organizational compatibility. 3) Characteristics of the technology influencing implementation, e.g., degree of complexity, compatibility, standards. 4) Characteristics of the task to which technology is applied influencing implementation, e.g., task uncertainty, responsibility, task variety.
5) Characteristics of the organizational environment, e.g., uncertainty,
dependence, and power.
Contributions of Implementation Factors from System Development
The system development perspective typically sees implementation as the
last step in the development life cycle. It is “the conversion and installation
of newly developed systems” (Marble 2000). From the system development perspective systems success can be measured by four parameters
(Coe, 1996): 1) use of the system measured by intended or actual use of
the system; 2) favourable attitudes toward the system on part of users; 3)
degree to which the system accomplishes its original objective, and 4)
payoff to the organization. These measures are closely related to successful
implementation. However, as pointed out by Coe (1996), numerous
implementation efforts related to information systems are technical successes but at the same time organizational failures.
An organizational oriented view on systems development is put forward by
Eason (1988) and by Tornatzky and Fleischer (1990). According to them
focus is on the organizational change caused by information technology.
156
Exploring Patterns in Information Management
The organizational perspective to system development suggests that certain issues are crucial in implementing IT in organizations such as: testing
and validating the technical system; organizational change; acceptance of
change; integration with other systems and, training and support. These
issues indicate that implementation of IT is an organizational adaptation
and learning process where the significance of technology is de-emphasized in favor of human/ organizational aspects.
Another view of implementation from a system development perspective,
which is more focused on the system, is presented by Dahlbom & Mathiassen (2000). They suggest that a set of quality parameters concerning the
fulfilment of users’ objectives is necessary for implementation. These
parameters include: correctness, reliability, efficiency, integrity, and
usability. This view of system efficiency as a parameter for successful
implementation is also suggested by Coe (1996) who argues that system
failure can be avoided by observance of five efficiency measures related to
systems delivery: implementation process owner, training, front line support, explication of efficiency measures, and effective communications.
Finally, Iversen et al. (2001) advocate the importance of risk management
during the implementation process.
Construction of the Implementation Framework
This section describes how the factors of the different perspectives are
clustered and the basis for this clustering.
Based on the literature reviews of the four reference disciplines mentioned
above, all factors were individually put on a blackboard in order to get an
overview. Then identical factors within the same reference discipline were
eliminated. A clustering of the factors was then undertaken through a
iterative process of identifying a common denominator of the clusters. The
final clustering process resulted in three clusters:. Technological factors,
Market factors and Organizational factors – in short the TMO-framework.
The table in appendix is the result of the clustering process.
As illustrated all four disciplines have a strong emphasis on organizational
factors. Some of these factors which are represented in all four disciplines
are management support and organizational structure directly related to
commitment to change. The diffusion theory and systems development literatures do not focus on marked factors in particular, which has been a
source of recent criticism (Lyytinen & Damsgaard, 2001; Kurnia &
Johnston, 2000) since it narrows the scope of diffusion theory in relation to
IOS. The technological cluster is represented in all four disciplines. How-
Bjørn-Andersen, Henriksen & Larsen
157
ever, focus is very different depending on the discipline. Whereas diffusion theory and system development literature focus on manifest attributes
of the technological artefact, VC and BPR focus on more abstract characteristics related to the capabilities of the technology.
Discussion
The validity and applicability of a framework is of particular importance as
it is formulated to serve as recommendations and guidelines for future
implementations of eBusiness models. Hence, the robustness of the proposed framework is discussed in the following.
As demonstrated above, literature suggests a huge number of factors that
may affect implementation of information systems and eBusiness models
in particular. In our synthesis and presentation of relevant factors we have
only selected those factors which were found as being important in more
than one source.
Since almost all eBusiness models are encompassed in the selected definition of Timmers (1998), we believe that the framework represents the
gross list of potentially relevant factors. However, the actual list of implementation factors for a particular e-business model may provide some
variation in the final outcome of the framework.
Our large survey of implementation factors within the four perspectives
resulted in a large number of factors which meaningfully could be grouped
into the three clusters: technology, market, and organization – the TMOmodel. Depending on the theoretical perspective, emphasis varied on the
three dimensions. However, given our multi-disciplinary approach it is
concluded that a feasible model for eBusiness model implementation has
to embrace the three dimensions.
Space does not provide the possibility of validating the framework here.
We have done that elsewhere, both theoretically and empirically, through
the application of the framework on a specific case of Haburi.com. Interested readers are referred to Larsen et.al. (2002).
Conclusions and Future Research
Our paper argues that the prerequisites for implementation of eBusiness
models, compared with traditional information systems, are changed due to
causes like new stakeholders, need for multi-perspective business analysis,
158
Exploring Patterns in Information Management
time compression of development time and changed development methods. Hence, it is relevant to suggest a framework highlighting important
implementation factors derived from various relevant disciplines.
The specific research questions investigated in this paper are: 1) What are
the key IT implementation factors in different perspectives? 2) How
should the key IT implementation factors be classified in order to provide
a coherent framework for eBusiness Model implementation?
The key IT implementation factors of each of four influential areas of
expertise/research – i.e. venture capitalists experiences, BPR, diffusion
theory, and systems development – were presented based on a literature
review. The implementation factors were classified in a framework – the
TMO-model – that identified the technological, the organizational, and the
market related factors relevant for implementation of eBusiness models.
The next step in the research is to test the framework on eleven already
available (successful) e-business cases from several European countries.
Following a revision, it is expected that the framework will be further
detailed and populated with a number of illustrative cases and made available for a wider audience.
References
Bashein, B.J., Markus, M.L. & Riley, P. (1994) “Preconditions for BPR Success –
And How to Prevent Failures”, Information Systems Management, Spring, pp.
7-13.
Champy, J. (1995) Re-engineering Management. Nicholas Brealey, London.
Clemons, E.K., Thatcher, M.E. & Row, M.C., “Identifying Sources of
Reengineering Failures: A Study of Behavioral Factors Contributing to
Reengineering Risks”, Journal of Management Information Systems, Vol. 12
No. 2, pp. 9-36.
Coe, L.R. (1996) “Five Small Secrets to Systems Success”, Information Resources
Management Journal, Vol. 9, pp. 29-38.
Cooper, R.B. & Zmud, R.W. (1990) “Information Technology Implementation
Research: A Technological Diffusion Approach”, Management Science, Vol.
36, No. 2.
Dagi, T.F. (2001) “E-Business and the Venture World of 2001. Cordova Ventures
and The Georgia Institute of Technology”, Presentation at the Global E-Management MBA Seminar, Atlanta, Georgia.
Bjørn-Andersen, Henriksen & Larsen
159
Dahlbom B. & Mathiassen L. (2000) Computers in Context – The Philosophy and
Practice of Systems Design. Blackwell Publishers Inc., Cambridge, Massachusetts.
Davenport, T.H. & Short, J.E. (1990) “The New Industrial Engineering: Information Technology and Business Process Reengineering”, Sloan Management
Review, Vol. 31, No. 4, pp. 11-27.
DeLone, H.W. & McLean, R.E. (1992) “Information Systems Success: The Quest
for the Dependent Variable”, Information Systems Research, Vol. 3, pp. 60-95.
De, R., Mathew, B. & Abraham, D.M. (2001) “Critical Constructs for Analyzing
E-Businesses: Investment, User Experience and Revenue Models”, Logistics
Information Management, Vol. 14, No. 1/2, pp. 137-148.
Eason, K. (1998) Information Technology and Organisational Change. Taylor &
Francis Inc., London.
Galliers, R.D. (1997) “Against Obliteration – Reducing Risk in Business Process
Change”, in Sauer, C. & Yetton, W. (Eds.) Steps to the future: fresh thinking
on the management of IT-based organizational transformation, Jossey-Bass,
San Francisco, California, pp. 169-186.
Gottschalk, P. (1999) “Strategic Information Systems Planning: The IT Strategy
Implementation Matrix”, European Journal of Information Systems. Vol. 8,
No. 2, pp. 107-118.
Grover, V., Jeong, S.R., Kettinger, W.J. & Wang, S. (1995) “The Implementation
of Business Process Reengineering”, Journal of Management Information
Systems, Vol. 12, No. 1, pp. 109-144.
Hall, G., Rosenthal, J. & Wade, J. (1993) “How to Make Reengineering Really
Work”, Harvard Business Review, November-December, pp. 119-131.
Hammer, M. & Champy, J. (1993) Reengineering the Corporation – A Manifesto
for Business Revolution, Nicholas Brealey Publishing, London.
Hammer, M. & Stanton, S.A. (1995) The Reengineering Revolution, Harper Business, New York.
Hammer, M. (1996) Beyond Reengineering, Harper Business, New York.
Hammer, M. (1990) “Reengineering Work: Don’t Automate, Obliterate”, Harvard
Business Review, July-August, pp. 104-112.
Hansen, M.T., Chesbrough, H.W., Nohria, N. & Sull, D.N. (2000) “Networked
Incubators: Hothouses of the New Economy”, Harvard Business Reivew, September-October, pp. 74-83.
Homa, P. (1995) “Business Process Reengineering – Theory- and Practice-based
practice”, Business Process Reengineering & Management Journal, Vol. 1,
No. 3, pp. 10-30.
Iversen, J., Mathiassen, L. & Nielsen, P.A. (2001) “Risk Management in Process
Action Teams”, in Mathiassen, L., Pries-Heje, J. & Ngwenyama, O. (Eds.)
160
Exploring Patterns in Information Management
Improving Software Organizations – From Principles to Practice, AddisonWesley, Boston, Massachusetts.
Keen, P.G.W. & McDonald, M. (2000) The eProcess Edge: Creating Customer
Value & Business in the Internet Era, McGraw-Hill, Berkeley, California.
Keen, P.G.W. (1997) The Process Edge Creating Value Where It Counts, Harvard
Business School Press, Boston, Massachusetts.
Kurnia, S. & Johnston, R.B. (2000) The Need of a Processual View of Interorganizational Systems Adoption, Journal of Strategic Information Systems,
Vol. 9, pp. 295-319.
Kwon T.H. & Zmud R.W. (1987) “Unifying the Fragmented Models of Information
Systems Implementation”, in Boland, R.J. & Hirschheim, R.A. (Eds.) Critical
Issues in Information Systems Research, John Wiley & Sons Ltd, pp. 227-251.
Larsen, M.H. & Bjørn-Andersen, N. (2001) “From Reengineering to Process Management – A Longitudinal Study of BPR in a Danish Manufacturing Company”, Proceedings of the 34th Hawaii International Conference on System
Sciences (HICSS 34). January 3-6, Island of Maui, Hawaii.
Larsen, M.H., Henriksen, H.Z. & Bjørn-Andersen, N. (2002) “The MTO framework
for Implemenation of eBusiness Models”, ICEB 2002 conference, Taiwan.
Larsen, M.H. & Leinsdorff, T. (1998). “Organisational Learning as a Test-Bed for
Business Process Reengineering”, Proceedings of the 31st Hawaii International Conference on System Science (HICSS’98), January 6-9, The Big Island
of Hawaii, Hawaii, Vol. 5, pp. 343-354.
Lazer, R. & Livnat, J. (2001) E-Commerce: Business Models that Make Economic
Sense, Working paper, New York University, Leonard N. Stern School of
Business, New York.
Linton, J.D. (2002) “Implementation Research: State of the Art and Future Directions”, Technovation, Vol. 22, pp. 65-79.
Lockamy, A. III & Smith, W.I. (1997) “A Strategic Alignment Approach for
Effective Business Process Reengineering: Linking Strategy, Processes and
Customers for Competitive Advantage”, International Journal of Production
Economics. Vol. 50, No. 2-3, pp. 141-153.
Lundeberg, M. (1993) Handling Change Processes – A Systems Approach, Studentlitteratur, Lund, Sweden.
Lundeberg, M. (1972) Information analysis, ISAC Instruction Report. TRITAIBADB-4401, Royal Institute of Technology, Stockholm.
Lundeberg, M. & Andersen, E.S. (1974) Systemering – Informationsanalys, Studentlitteratur, Lund, Sweden. (Also available in Norwegian: Lundeberg, M. &
Andersen, E.S. (1974), Systemering – Informasjonsanalyse, Tanum, Oslo.)
Lyytinen, K. & Damsgaard, J. (2001) “What’s Wrong with the Diffusion of Innovation Theory?”, in Diffusing Software Product and Process Innovations. Pro-
Bjørn-Andersen, Henriksen & Larsen
161
ceedings at IFIP TC8 W.G 8.6, fourth working conference, April 7-10, Banff,
Canada, pp. 173-190.
Magretta, J. (2002) “Why Business Models Matter”, Harvard Business Review.
May, pp. 86-92.
Marble, R.P. (2000) “Operationalising the Implementation Puzzle: An Argument
for Eclecticism in Research and in Practice”, European Journal of Information
Systems, Vol. 9, pp. 132-147.
Martinez, E.V. (1995) “Successful Re-engineering demands IS/Business Partnership”, Sloan Management Review, Summer, pp. 51-60.
McElrath-Slade, R. (1994) “Caution: Re-engineering in Progress”, HR Magazine,
Vol. 39, No. 6, pp. 54-57.
Noble, C.H. (1999) “Building the Strategy Implementation Network”, Business
Horizons, November-December, pp. 19-28.
Premkumar, G. & Ramamurthy, K. (1995) “The Role of Interorganizational and
Organizational Factors on the Decision Mode for Adoption of Interorganizational Systems”, Decision Sciences, Vol. 26, pp. 303-336.
Premkumar, G., Ramamurthy, K. & Nilakanta. (1994) “Implementation of Electronic Data Interchange: An Innovation Diffusion Perspective”, Journal of
Management Information Systems, Vol. 11, pp. 157-177.
Ramamurthy, K. & Premkumar, G. (1995) “Determinants and Outcomes of Electronic Data Interchange Diffusion”, IEEE Transactions on Engineering Management, Vol. 42, pp. 332-351.
Ramamurthy, K., Premkumar, G., & Crum, M. (1999) “Organizational and
Interorganizational Determinants of EDI Diffusion and Organizational Performance: A Causal Model”, Journal of Organizational Computing and Electronic Commerce, Vol. 9, pp. 253-285.
Rogers, E.M. (1995) Diffusion of Innovations. The Free Press, New York.
Sarkis, J., Presley, A. & Liles, D. (1997) “The Strategic Evaluation of Candidate
Business Process Reengineering Projects”, International Journal of Production Economics, Vol. 50, No. 2-3, pp. 261-274.
Timmers, P. (1998) “Business Models for Electronic Markets”, in Gadient, Y.,
Schmid, B.F. & Selz, D., EM – Electronic Commerce in Europe, EM – Electronic Markets, Vol. 8, No. 2.
Tinnilä, M. (1995) “Strategic Perspectives to Business Process Redesign”, Business Process Re-engineering & Management Journal, Vol. 1, No. 1, pp. 44-59.
Tornatzky L.G. & Fleischer M. (1990) The Process of Technological Innovation.
Lexington Books, Lexington, Massachusetts.
Willcocks, L. & Smith, G. (1995) “IT-enabled Business Process Reengineering:
Organizational and Human Resource Dimensions”, Journal of Strategic Information Systems, Vol. 4, No. 3, pp. 279-301.
162
Exploring Patterns in Information Management
Appendix: Summary of Implementation Factors
TECHNOLOGICAL FACTORS
Venture
Capitalists:
Focus on technology suppliers and
partners
Focus on incubating environments
for basic ICT-support.
BPR:
Diffusion Theory:
Focus on information technology in
support of business process
effectiveness
Complexity
Accurate data
System
Development:
Efficiency
Interaction
Maintaining the
integrity of
throughput
Compatibility
Reliability
Standards
Correctness
IS infrastructure
Integrity
Integration
Out-of-the-box
thinking
Extensive project
definition and
planning
Experience with IT
IT design
Recognise the
potential of IT
Inductive thinking
instead of deductive thinking
Integration
Usability
Understanding
existing data,
applications and
databases
IT capability
Information gain
instead of technology costs
Collect data from
source
Managing IT is
culturally dependent
MARKET FACTORS
Venture
Capitalists:
Other investors
Market analysis
Sustainable competitive advantage
Exit opportunities
Barriers to entry
Customer experiences
Strategic vision
BPR:
Diffusion Theory:
Alignment of business processes
and strategy
N.A.
Customer focus
Customer value
definition
Definition of customer performance
measures
System
Development:
N.A.
Bjørn-Andersen, Henriksen & Larsen
163
ORGANIZATIONAL FACTORS
Venture
Capitalists:
Management
The board
Products and/or
services
Revenue model
Organizational
structure
True costs of
starting and running the business
Growth is obtained
organically or by
acquisition.
Focus on goals
Leadership
Communication
Incentives
BPR:
Diffusion Theory:
Top management
commitment
Job tenure
Process orientation
Scoping of BPR
projects
Clean sheet principle
Education
Resistance to
change
Appropriate userdesigner interaction and understanding
System
Development:
Acceptance of
change
Training and support
Job redesign
Organizational
change
Holistic redesign of Commitment to
business system
change
Organizational
redesign
Performance
based incentive
structure
Plan implementation process
Skill-base and
training
Definition of (non)
value adding
activities
Performance
measurement
Learning
Recognition and
management of
diverse vested
interests of IT
stakeholders
Social approval
Communicability
Individual learning
Organizational
learning
Shared values
Innovation champion
Communication
Specialization
Training
Centralization
Handling of political controversies
Formalization
Top management
support
Compatibility with
organizational
tasks
Relative advantage
Cost
Profitability
Divisibility
Trialability
Observability
Internal need
Human resource
development
Understand innovation
Measure effectiveness
164
Exploring Patterns in Information Management
— 11 —
On Interpretation of Strategic
Knowledge Creation in a Longitudinal
Action Research Project
Pentti Kerola
Tapio Reponen
Mikko Ruohonen
Introduction and Research Idea
Knowledge creation is inevitable in the Information Resource (IR) strategy
process. Yet all the stakeholders need to commit themselves to joint
objectives, to finding out linkages of business and IR opportunities, then to
deciding on IR investments along the business strategy and finally to
evaluating the outcomes. This demands a new breed of managers who create for themselves a new way of thinking the role of information resources
in the context of business (see Dickson et al. 1984; Brancheau and Wetherbe 1987; Niederman et al. 1991).
The objective of this study is to get a deep understanding on how an interactive strategy generation (e.g. Sanchez and Heene, 1997) should be carried out to meet the integration goals. The research work has been realized
in an empirical context, where researchers have been involved in a development process over fifteen years. This is an exceptionally deep and long
research period.
Action research has been used as a research methodology, but its features
have also been used in the planning process. We interpret action research
as Hult and Lennung (1980) have defined it:
Action research simultaneously assists in practical problem-solving
and expands scientific knowledge, as well as enhances the competencies of the respective actors, being performed collaboratively in
an immediate situation using data feedback in a cyclical process
aiming at an increased understanding of a given social situation,
primarily applicable for the understanding of change processes in
166
Exploring Patterns in Information Management
social systems and undertaken within a mutually acceptable ethical
framework.
This definition also describes very much the nature of our research work.
The research questions are the following:
• How do we understand the multifaceted knowledge creation process of
IR strategy process?
• How do we understand the problems of commitment and implementation and their solutions in an IR strategy process?”
For tackling the first problem area we have developed a framework to
guide the operative process, called Evolutionary Model of Information
Resources Strategy (EMIS-model, Reponen, 1994). In our research
Nonaka and Takeuchi (1995) framework (NT-theory) of knowledge creation is used as a theoretical framework in order to focus and analyze this
longitudinal action research. Kerola and Reponen (1996) emphasize the
role of both managers’ and researchers’ joint knowledge creation process
in strategy creation.
The paper proceeds as follows. The second section defines and clarifies
our conceptual and theoretical basis by situating it in basic concepts of the
EMIS-model and in the fundamentals of the NT-theory. The third section
concentrates on the description of the case organization and its business
strategies, and on some interpretations by the NT-concepts. The fourth
section is focused on describing and interpreting the multiple phases of IR
strategy generation and implementation. The fifth section gives the major
findings, conclusions and recommendations.
Conceptual Background and Framework
Theoretical Foundations and Nature of the EMIS Model
The foundations of the EMIS model are in the discussion on the competitiveness of business companies (Porter 1980, 1985) and in the role of
Information and Communications Technology (ICT) in the competition
(McFarlan & McKenney 1983). The aim of the approach is to take a business view on the use of ICT and to promote interaction between different
stakeholders in order to support the implementation of an IR strategy
(Reponen 1994; Galliers 1991).
Kerola, Reponen & Ruohonen
167
The model has two very different objectives. On one hand it is a normative
framework to guide the strategy development process, and on the other
hand it is also a way to increase interaction of stakeholders in the target
company to generate double-loop learning and to increase mutual understanding on ICT issues. The approach tries to combine systematic planning
with intuitive thinking. The EMIS model offers enough structure for the
strategy creation process, but emphasizes also qualitative and implicit considerations.
The main contribution of the EMIS model is the early adaptation of interactive learning in the strategy generation process. Instead of “top-down” or
“bottom-up” thinking it represents “inside-out” thinking (Ein-Dor and
Segev 1981, Hirschheim 1982, Galliers 1987). Trying to collect insights of
different people on the potential use of ICT, creates the essential contents
of the strategy process (Lederer and Sethi 1996). The objective of creating
shared vision was intuitively present in the early stages of using the EMIS
model, but has become more explicit over time.
This approach has resemblance with Mintzberg’s concept of emergence
(Mintzberg 1994) and promotes interaction of people and information
systems (Lundeberg & Sundgren 1996). Emerging strategy is something
that evolves from different stakeholders’ ideas and thinking, not necessarily from the formal planning process. In the EMIS model an effort is made
to collect the emerging ideas from different people in the organization.
R
PA
TIC
IPA
NT
S
Co
C O NT E NT S
ns u
Us
l ta n
D ev
ts
el o
AD
per
P
Ma
s
nag
Lin
em
e nt
Ma
e
nag
S
em
e nt
M a enio
r
nag
em
e nt
e rs
B e n efits
Inv e st m e n ts
A rc h ite c tu re
O rg a n isa t ion
C o mp eti tio n
St r ate g y
L e c t u re s
M e e t in g s Te a m -w or k Int e rv ie w
E x p ert
R e p or ti n g
V is io n
D ra w i ng u p
M E TH O DS
o f Pl a ns
D e cis io n
D e sign
D e fi n it i on
S u rve y i ng
C o m m e n c e me n t
STAGES
Figure 1. EMIS Knowledge Space of Information Resource Strategy
168
Exploring Patterns in Information Management
Information collection in the EMIS model can be illustrated with a fourdimensional space of knowledge (Figure 1). Its dimensions are different
stakeholders, different stages of the process and different working methods. The main idea is that with multiple methods and with multiple participants it is possible to create an environment where mutual understanding
increases and the proposal has wider intellectual background than otherwise.
The EMIS process consists of the following stages: commencement, survey, definition, design and decision. The commencement stage is where we
create as common a language as possible within the participants of the
strategy process. Interactive meetings of those people involved in the planning process can do this. Lectures and discussions on integrating business
and ICT are important parts of this stage.
Surveying means charting the development ideas of different people, collecting facts and designing different development alternatives. This can be
done with interviews, project group meetings, expert reports, by collecting
earlier reports and by idea generation.
In the definition stage the decisions on the general objectives of the strategy will be made. The key findings from the survey will be presented to
the decision-makers to decide guidelines for the planning of future alternatives. This stage is highly interactive process between decision-making
bodies and the planning team. The objective of this interaction is to
increase the knowledge of top management on the use of ICT in the processes of the organization.
In the design stage the plans for different sub-areas of the strategy will be
created. It is important that there is a wide representation of different business expertise involved. Teams may be nominated for different sub-areas
to define the alternative solutions for each of them.
In the decision stage the final decisions are made on objectives and general
lines of the strategy. An IR strategy is a wide and holistic concept including the following elements: the competitive objectives of using ICT, a plan
for maintaining present systems and building new ones, a rough overall
architecture, an organizing of information management, an investment
proposal and expected benefits/risks of the strategy. A strategy is a plan for
all these sub-areas; the emphasis depends on the situation of the case company.
An IR strategy as a plan is a written document for future development and
use of information systems. The real strategy i.e. perspective exists in the
Kerola, Reponen & Ruohonen
169
minds of people who have participated into the strategy process. This
understanding may be supported with written reports, but the implementation is very much based on the internal views of people in the organization.
From this viewpoint all interaction between different stakeholders is
extremely important.
The role of researchers as outside facilitators is important in charting the
internal views of stakeholders. By interviewing people the researchers can
collect both common and conflicting opinions of ICT usage. Thus the final
strategy is a combination of both internal and external expertise.
Integrated Individual-, Team- and Organization-Oriented Knowledge
Creation Model – Nonaka’s Generic Concepts
The purpose of this section is to represent some fundamental and basic
concepts of the Nonaka and Takeuchi theory. Those are then utilized in the
third and fourth chapters. Why have we selected this theory for retrospective interpretations? The main reason has been in the analogies between
the EMIS model and the NT-theory, concerning the objects of interest,
aims and substructures.
Nonaka considers “knowing” as a dynamic human process of justifying
personal belief towards the “truth”. In a strict sense, only individuals create
knowledge. However, the main essence of Nonaka’s theory of organizational knowledge creation is the integration of individual, team and organization orientations (Nonaka 1994, Nonaka and Takeuchi 1995).
Nonaka and Konno (1998) emphasize the specific contextual nature of
knowledge creation. They introduce the Japanese concept “Ba” and its
interpretation as a context which harbors meaning. Their most fundamental
idea is that knowledge is created through the interaction between tacit (T)
and explicit (E) knowledge.
Tacit knowledge is ultimately personal, context-specific and therefore hard
to formalize, making it difficult to communicate or share with others. It
requires no specific language for communication. Subjective insights,
intuitions, hunches, and personal bodily skills fall into this category of
knowledge.
Explicit or “codified” knowledge can be expressed in words and numbers,
and it is more shareable in the form of data, scientific formulae, law, rules,
specifications, manuals and the like. This category of knowledge can be
transmitted between individuals formally and systematically utilizing natural, figurative, systematic and formal languages.
170
Exploring Patterns in Information Management
The NT-theory postulates four different modes and subprocesses of
knowledge conversion: socialization T->T’, externalization T->E, combination E->E’ and internalization E->T. They link these concepts into the
different “bas” as described in Figure 2.
Figure 2. Different Ba’s of Knowledge Creation in SECI-model and foci in action
research
Socialization (from T to T’) is a process of sharing experiences (between
people) and thereby creating tacit knowledge as mental models and/or
technical skills that can be called “sympathized knowledge”. Tacit knowledge is exchanged in practice through joint activities – such as being
together, spending time or living in the same environment – capturing
knowledge through physical proximity. Co-experiences allow people to
become aware about others’ way of feeling and thinking – in order to
empathize with others. Therefore they are more receptive and capable of
acquiring new tacit knowledge utilizing the existing one as basis.
Socialization is implemented in the context of “Originating Ba”. It is the
primary context and basis from which the knowledge-creation process
begins. It includes synchronizing behavior, improvisation and face-to-face
possibilities. From it emerge care, trust, love and commitment. Organizational issues that are closely related to it are culture and knowledge vision.
Kerola, Reponen & Ruohonen
171
Externalization (from T to E) is a collective reflection process articulating
tacit knowledge into explicit concepts. Yet it is a critical process, as
explicit knowledge takes the shape of metaphors, analogies, concepts,
hypotheses or models that results in the creation of “conceptual knowledge”. It requires the expression of T and its translation into comprehensible forms that can be understood by others who have the necessary language skills. During the externalization phase an individual commits to the
group and thus becomes one with the group. The articulation of T into E
involves techniques that help to express one’s ideas or images as words,
concepts, visuals, metaphors, analogies and narratives. Dialogue, listening
and contributing to the benefit of others, strongly supports externalization.
The “Interacting Ba” (Figure 2) is more consciously constructed, as compared to originating one. Selecting people with the right mix of specific
knowledge and capabilities is critical. Two sub-processes would operate in
concert: individuals strive to share the mental models of others, but also
reflect and analyze their own. In dialogue people engage jointly in the
creation of meaning and value.
Combination (from E to E’) involves the conversion of explicit knowledge
into more complex sets of existing E. It is a team interaction process of
systemizing concepts into a knowledge system. Individuals exchange and
combine knowledge through such media as documents, meetings, telephone conversations or computerized communication networks, in a way
that can lead to new explicit “systemic knowledge”. In the combination
process, justification – the basis for agreement – takes place and allows the
organization to take practical concrete steps.
“Cyber Ba” is a context and place of interaction in a virtual world instead
of real space and time. Combination is most efficiently supported in collaborative environments utilizing ICT technology.
Finally, internalization (from E to T) is a process of converting the organization’s explicit knowledge into the individuals’ new tacit knowledge.
This requires that the individual identify the knowledge relevant for one’s
self within the organizational knowledge. That again requires finding one’s
self in a larger entity. Learning by doing, training, re-experiencing and
exercising allow the individual to access the knowledge realm of the team
and the entire organization. The output of this process might be termed
“operational knowledge”.
“Exercising Ba” supports the internalization phase involving both human
and ICT environments. Especially focussed training with senior mentors
and colleagues is recommended over teaching-based learning.
172
Exploring Patterns in Information Management
Our Approach for Interpretations of Real World Situations
In principle, all the modes of knowledge conversion are in continuous
and dynamic complementary interaction with each other. The whole
knowledge creation process is as good as its weakest sub-process. Therefore, it is very essential to assess the balance between the different subprocesses. Because of the inherent tensions of knowledge creation the
balance is not easy to achieve. Dynamically, however, NT-theory emphasizes the spiral evolution of knowledge creation: from socialization to
externalization, from externalization to combination, from combination
to internalization and from internalization to socialization (SECI-cycle)
etc. The interaction between T and E will gradually grow in scale, starting at the individual level and moving up through expanding communities of interaction that cross sectional, departmental, divisional and
organizational boundaries. Now we mainly concentrate on the organizational growth of knowledge.
We interpret the theoretical sub-processes SOC, EXT, COM and INT
(Figure 2) in real life situations as follows:
• SOC = socialization in focus; other sub-processes exist, but they are
subordinated to SOC Seci-cycle in the context of Orig-ba (size of
letters refers to the focus in the cycle)
• EXT = externalization in focus; other sub-processes exist, but they are
subordinated to EXT sEci-cycle in the context of Orig-ba and IntAba
• COM = combination in focus; other sub-processes exist, but they are
subordinated to COM seCi-cycle in the context of Orig-, IntA- and
Cyb-bas
• INT = internalization in focus; other sub-processes exist, but they are
subordinated to INT secI-cycle in the contexts of Orig-, IntA-, Cyband Exc-bas.
The main reason for this practical real-life interpretation is that when real
people interact they always generate both tacit and explicit knowledge
interactively. This type of foci specification could be continued by enumerating all the double foci alternatives (e.g. SEci-cycle etc), all the triple
alternatives (e.g. SEcI-cycle etc.) and finally two quadruples seci- and
SECI-cycles.
In the EMIS model the sub-processes commence, survey and definition
have foci in socialization and externalization within their responsive con-
Kerola, Reponen & Ruohonen
173
texts. The design and decision sub-processes have more foci in combination and internalization within their responsive contexts.
Stakeholders inside and outside of the company define the strategy and
therefore the real nature of its planning is different from the rational planning models. Learning and knowledge creation during the process is
extremely important. Therefore, Nonaka’s knowledge creation sub-processes would help to understand strategy creation and implementation processes. People use their tacit knowledge in their actions. Knowledge in
documents such as strategy plans is explicit knowledge. The main problem
is therefore to explain, convert, and express tacit knowledge in explicit
concepts and terms. We can say that challenge of management is to try to
explicate their “tacitness”.
Externalization provides symbols and icons for the organization, which
you can rely on and work with. The result of this process is conceptual
knowledge. Any written or encoded document is externalized from tacit
knowledge of process participants or is converted from other externalized
knowledge. The problem is, however, commitment. Many strategy plans
are never implemented due to low commitment, which is, in our view,
basically due to misunderstandings and errors in this knowledge conversion. With systematic knowledge created through a combination it is possible to describe instructions, tools and systems for the organization. This
means an effective combination of plans and budgets with technology
development investments.
Operational knowledge is the result of internalization, i.e. the use of documents and artefacts for transformed routines and processes. In the case of
ICT deployment it means that people should adopt new procedures, work
tasks and business processes enabled by strategy-directed technology.
However, this is the main problem of organizations. It has been fashionable to discuss the productivity paradox and service quality, which in our
view is a problem of effective internalization. Internalization is also critical
for the “next circle or spiral” of organizational knowledge creation. Without a new understanding of business models and processes enabled by ICT
architectural solutions there is less possibility to upgrade the use of ICT.
In the following sections we describe and interpret two different, but
essentially interrelated areas of strategic knowledge creation, at first concerning development of business strategy, and then development and
implementation of IR strategies as one essential sub-area of the business
strategy implementation.
174
Exploring Patterns in Information Management
Introduction to the Case Enterprise
and its Business Strategies
This section at first describes the case enterprise and then tells a story how
its strategies were developed and implemented. The concepts of NT-model
have been used in the description and interpretation.
Case Company’s Business Environment
S-Group is the second largest Finnish wholesale-retail chain with a market
share of over 30 % (in daily goods). Its total annual sales are near 6,5 billion euros and it employs almost 20 000 people. The group consists of different businesses like super- and hypermarkets, department stores, discount shops, agriculture markets, hotels, restaurants, car dealerships, service stations and some special shops and boutiques. The multidimensionality of the company has increased over the last ten years making the strategic planning and decision making very complex.
The cooperative structure of the S-group is such that the country has been
divided into 23 districts where all have their own independent area cooperatives. They have a national central body called The Finnish Center of
Cooperatives (SOK). Earlier SOK was the wholesaler for all the cooperatives, but now its operations have been re-engineered from the wholesaling
function to the information and logistics center. S-Group is the joint name
for all area cooperatives, companies owned by cooperatives and their central cooperative SOK.
All the businesses are now logistical supply chains from suppliers to the
customers. SOK itself runs some of the chains such as department stores,
agriculture markets, some restaurants and hotels. However, area cooperatives manage and operate most of the supermarkets, local shops, fuel and
service stations, car dealerships and some restaurants and hotels. Although
the ownership is somewhat diverse the objective of the whole S-Group is
to strengthen the chains, and the traditional wholesale function is diminishing. Business chain operation model means that many of the decisions
on selections and assortments are made centrally.
The strategic challenge of the S-Group is to combine the effectiveness of
centralized decision making to the expertise of local area cooperatives. The
management structure of the company is such that the highest decision
making body is the administrative council where the local cooperatives
have their representatives. The council nominates the board of directors
that makes the actual planning and implementation work. This model is
Kerola, Reponen & Ruohonen
175
replicated in each area level. The council controls and coordinates the
work of the board and managing director. In addition to this, all business
chains have their operative boards which decide on assortments and logistics.
The nature of the cooperative organization is such that families may join as
members of an area cooperative and become owner-customers. This will
cost a certain amount of money and give all a share of the cooperative. All
members can then participate in the decision making of the local cooperative by electing representatives to decisive bodies of the organization
(board of representatives, supervisory board and the board of directors).
However, all cooperatives have professional managers for running business operations. The total number of these owner-customers is currently
over one million households. The number has been increasing for several
years, primarily due to active development work in the S-Group and
recently also because of the growing importance of customer management
in Finnish retailing industries.
Knowledge Creation in Business Strategy Development
Next follows a short description on the strategic development of S-Group
comparing its decisions to the general management trends. This section
offers an overview of the business development in the case enterprise during the years of this longitudinal action research. The objective is to
describe to the readers the business moves of S-Group in order to give
background understanding for the decisions in the IR management.
The traditional operation model in retailing businesses had been centralized and therefore inefficient, due to the time-lapse between the finished
manufacture of goods and their retail, during which time capital lies inactive. The role of middlemen such as wholesalers was seen as threatened
and difficult. The industry had plans to start direct supplies to customers or
retailers; the idea of electronic commerce was introduced and discussion
started on decreasing excessive steps in the logistical supply chain. There
was even incipient conflict between industry and wholesalers about who
controlled the material and information flows.
The main competitor of S-Group in Finland established area distribution
centers to overcome excessive centralization. S-Group had its traditional
model of wholesaling which was not very competitive. Its main competitive edge was the group itself because it was very difficult for the area
cooperatives to buy from other sources. But this model did not guarantee
price competitiveness.
176
Exploring Patterns in Information Management
In 1987 a new managing director was appointed from inside the Group. In
his earlier position he had already been responsible for business strategy
development and one of his first actions was to finalize a new strategy for
the group. The challenges described above were evident, and new operating strategies were needed. The socialization process was in the strategy
generation process within the top management of SOK. They visited several other countries to find ideas for innovative solutions. They looked for
new development paths both with external influences and internal consideration. The ideas matured with both interaction and personal thinking. In
this stage the methods of socialization were internal discussions and
meetings. The action researchers were not yet present.
On the basis of these idea-generating processes the management became
convinced that the traditional wholesaling should be replaced with nationwide chains. These strategic plans were generated mainly within the top
management of SOK with only some participation from the area cooperatives. Therefore, the suggested model was very centralized, where the
decisions about assortment and supply channels were made in the chain
management. The objective was to create a very cost effective distribution
model to compete with the existing structures.
Another idea was to link customers closely to the area cooperatives. Some
pioneering examples of customer bonus systems already existed and in
some cases they had contributed to increasing market share. S-Group was,
however, a very early adapter of this thinking. In a cooperative this is a
very natural way of operating, as the members of the cooperatives are also
owners of the organization. The objective was set to strengthen the customer links.
In retrospect we can notice that business strategy generation was a SOKcentered socialization process and SEci-cycle by its nature. Top management of SOK was a key actor and player in the strategy process. Consequently the management at the highest level of the Group was involved
and committed to new strategic lines, but at the area cooperative level the
knowledge of strategic objectives was limited. Management of area cooperatives was not present in the Orig- and Int-ba’s of the knowledge creation process, leaving them outsiders of the inner circles of the planning
process.
The socialization and externalization processes resulted in strategic plans
following the SEci-cycles. The role of wholesaling would consequently
change dramatically: there would be fewer steps in the delivery chain,
operations would be faster and customer contacts would be closer than
Kerola, Reponen & Ruohonen
177
earlier. The basic ideas of the strategy were based on the earlier experience
of the management, outside influence, and interaction.
From a knowledge creation viewpoint, tacit strategic information was
shared within a limited group. As the socialization process was only partial, externalization to governing bodies and area cooperatives was a challenging task. This may, however, have been an intentional choice of the
managing director since the new strategy proposed a radical business
transformation from traditional wholesale-retail model to “business chain”
strategy. We can raise the question: when conflicting goals exist how
should socialization be realized? Presenting premature thoughts and plans
may result in early rejection.
This strategy was decided early 1988 and implemented over the following
ten years. The managing director clearly had a vision of how wholesaling
should be developed to meet the requirements of the future. Naturally
implementation requires making the plans explicit and acceptable. Because
of the limited nature of knowledge sharing in the business strategy development process, the following IR strategy generations were essential in
implementing the new business strategy. The two cornerstones of the strategy have proven to be essential for the whole business: integrated logistic
wholesale-retail chains and customer bonus systems (Figure 3). In Figure
3-Figure 6 we utilize the pictorial structure of seci-cycle where the upper
left rectangle refers into the s-focussed sub-process, the upper right rectangle into the e-focussed sub-process etc.
Innovative ideas for
transformation; visit to UK
- Before 1987
Early introduction of business chain strategy to area
cooperatives (“towards
business chain”)
- From 1988-1996
Looking for next
business challenges?
- Beyond 2000?
Positive business results
convince area cooperatives; commitment
increases
- Since 1996-
Figure 3. Transforming knowledge to a new business model – from the wholesaler
model to a business chain model, described as a macro SECI-cycle.
178
Exploring Patterns in Information Management
The main problem with implementing the business strategy was the discussion about the decision authority in different parts of the organization.
The main idea of the new operating system was a relatively centralized
decision concept, which was difficult or impossible to accept by many of
the area cooperatives. Some discussion had been going on regularly about
the right balance between local and central decision making. In the implementation of the model the degree of centralization has somewhat
changed.
Since then the strategy implementation process has continued along similar
lines. The concepts of market-oriented chains and customer bonuses have
been developed in the spirit of the earlier strategy, but always considering
the balance between the economies of scale and local expertise. The business strategy calls for proper information systems to meet all the business
requirements. Thus in 1988 an information resource strategy creation process was carried out for the first time. IR strategy process was clearly used
as part of the externalization and internalization process of the business
strategy. We can observe similar examples in several business cases:
information systems development is the concrete way of implementing
business objectives.
Recently S-Group has gained a greater market share than competitors and
it has been very profitable after many recession years. It is still a cooperative, with customer ownership, but its business operations are very modern
and far from what might expect of a traditional, inefficient cooperative
company. This change has been received with a broad-minded strategy and
operations. The S-Group has turned this type of ownership into one of its
strengths.
Knowledge Creation and Interpretation in the IR
Strategy Creation and Management Processes
Over the last fifteen years there have been principally four different
stages in the IR strategy process of S-Group, but all of them have been
based on the work done during the first stage in 1987-88. These stages
and their essential features are reported below in chronological order.
The business development described above forms the basis for ICT plans
and decisions. One of the main targets of the IR strategy projects was to
ensure the coordination between business and ICT utilization. Interpretations through theoretical lenses of Nonaka and Takeuchi will be provided.
Kerola, Reponen & Ruohonen
179
Creating a New IR Strategy
As expressed in the business strategy, a task force was nominated in 1987
to create the information resources strategy for the whole group. The
assignment of the task force was “…to improve the competitiveness of SGroup by directing ICT development projects to support business strategy”. The goals of the strategy were the following: “The final result of the
project should be a development program on ICT utilization for the years
1989-1992, total ICT architecture and a cultural change to increase the
commitment of all parties into the new strategy”. The IR strategy project
had to take care of the overall internalization of the S-Group business
strategy.
With these objectives in mind the strategy creation process was carried out
in 1987 as a participative learning process in accordance with the principles of EMIS-model. The following knowledge subtopics were covered:
the business strategy implications to the ICT, business information needs,
new software needed, the status of the existing software, ICT architecture,
organizing IS function and an investment analysis. In the planning organization there was a high level of representation from different parts of the
S- Group. Additionally, most of them were business representatives, not
from the data processing unit.
The parent-organization was interested in creating synergistic IS activities
in the area of customer and marketing databases and logistical activities.
This objective indicates the need for joint network architecture for the
whole group. Business chains and their stores were investing in European
Article Numbering (EAN) based Point-of-Sales (POS) terminals and backoffice information systems, in order to integrate with the basic ICT architecture of the company. The retailing business in general was in a turnaround situation (McFarlan 1984).
The process concluded with the proposal of an IR strategy which included
the following issues:
Competitive objectives. The main objective of the information resources
strategy was to support the change from wholesale structure to market oriented chains. The existing software had been built for wholesale operations
and logistics. The new, intended operating model was very much different
and required new software generation, which would support market-oriented chains. The challenge was to create the technological base for
implementing new business structures. This process can be interpreted as a
seCI-cycle for developing business strategy knowledge.
180
Exploring Patterns in Information Management
ICT architecture. The strategic objective, a new simplified operating
model was created and new marketing and customer-oriented chains were
designed. The business model was operationalized with a totally new
logistical architecture which included manufacturer alliances and new
logistical arrangements. In order to develop customer-related management,
a customer card and bonus information system was designed for the whole
S-Group nation-wide. This was a pioneering system in Finland and very
advanced thinking globally. Through NT-lenses it was the seCi-cycle for
developing IR strategic knowledge.
Applications portfolio. A proposal was made to direct the new software
development towards strategic chains, owner-customer marketing and
logistical systems. The ICT architecture was proposed to be decentralized
in a communications network with a multivendor policy. This suggestion,
to gradually move from a mainframe architecture to a more decentralized
model, was the result of high business management involvement in the
planning process. Most of the data-processing professionals were skeptical
about this move. Our interpretation of this issue is that it fits the seCi-cycle
for IR strategy knowledge development.
The strategy proposal was presented to the Board of S-Group in December
1988. The proposal was accepted with some minor changes. The objective
of the CEO of S-group was to create a modern and competitive nationwide business model, implemented with modern ICT technology. But its
realization required support from all the area cooperatives. Some of them
thought, however, that the IR strategy had been designed mainly for the
central organization, rather than originating in their own local interests.
Naturally there were both socialization and internalization problems
(Figure 4).
The progress of the strategy was to be evaluated by the following adpcentered measures: the productivity of data processing should be at the
level of the competitors, EDI should cover 90 % of product range, customer bonus registration should work in all units, teleshopping should be
in test use, and electronic mail should cover all the users of the organization. These measures cover some of the main targets of the strategy either
directly or indirectly. They were concrete and therefore usable. During the
planning process it became evident that it would be difficult to find or create operational and concrete measures for IR strategy development. The
measures decided describe, however, the progress of the strategy in such a
way that it could be followed.
Kerola, Reponen & Ruohonen
181
Implementing the Strategy
After the strategy was decided in December 1988, its implementation
started immediately. Coordination of the strategy was located in the corporate management where the information systems manager was responsible
for carrying out the plans. A small information systems department of a
few employees was founded. Its main role was to implement the IR strategy mainly by buying services from multiple suppliers. Much of the systems planning was, however, made within the S-Group’s organization.
The new applications were built in the following order:
• Cashier register systems for the stores (1989-1991)
• Department store systems (1989-1995)
• Customer bonus systems (1990-1991).
One of the most important questions asked at the beginning of the strategy
planning process was: “What will the future IS organization and ICT
architecture be?” (Figure 4)
A strategy perspective
Introducing an IR strategy
making process for all key
plan 1988
stakeholders
Externalizing a plan
- creating shared views
Building an IS
architecture
?
-?
Implementing strategy with
an IS architecture
framework:
- POS system -1991
- department store IS -1995
- customer bonus IS -1991
Figure 4. IR strategy development described as a macro level SECi-cycle
An interesting form of outsourcing, starting a new company as a joint
venture, was introduced in the implementation stage. This was a somewhat
exceptional form of outsourcing, but in Finland there were several very
similar cases in the mid-1980s. This indicates some kind of bandwagon
effect for following other companies’ examples.
The Data Processing Company was growing to almost 200 people with
orders from both the S-Group and outsiders. Its strategy was to concentrate
182
Exploring Patterns in Information Management
on the mainframe applications and network management. At that time
SOK had only one manager coordinating its information systems services,
while services were bought outside. In the new strategy a decision was
made to decentralize the architecture. This goal conflicted somewhat with
the strategy of the data processing company (Reponen, 1997).
Because of the new business strategy the S-Group had to reflect further on
the new role of IS services. Information systems played an important role
for business and that demanded new qualities from the IS/IT staff. Some
external changes also increased the speed of transformation (i.e. merger in
in-bound logistics, joint-manufacturing agreements, decrease in “internal
invoicing”).
Updates of the IR Strategy
The first update in 1990. As there were some delays in building and
implementing the new software, regional cooperatives presented their
doubts on the whole strategy. Therefore, a decision to update the strategy
was made in 1990 to make sure that it met the changing requirements.
However, implementation according the IR strategy 1988 continued. The
same “control act” happened again in 1992 when a special report was
ordered from an outside consultant on the status of the IR strategy implementation and on the service level of the information systems. After
carefully going through the situation, the consultant came to the
conclusion that the strategy plan was quite appropriate and up-to-date.
His view was that implementing the IR strategy would be extremely
important for the future success of the S-Group. This gave support to the
IS professionals to continue their work, but there still remained some
conflicts between central and local decision making. All these updates
we interpret as examples of the sECI-cycles (Figure 5).
The reason for the revisions was uncertainty among management on the
relevance of the plans made a few years earlier and on the relationships
between area cooperatives and SOK. Economic recession had also started
and some screening of the investment program had to be made. The objective to decrease total costs by re-engineering operative processes was
emphasized. The EMIS approach was used in trying to find out business
management’s opinions and attitudes towards the IS function. This was
done by university facilitators interviewing twenty upper management
level persons in the organization.
Kerola, Reponen & Ruohonen
183
A strategy perspective
Introducing an IR strategy
making process for all key
plan created in 1988
stakeholders
Externalizing a plan
Auditing IR
- creating shared views
Redesigning
strategy
IS service
developorganization
ment by the
Local concerns
consultant
From internal IT
- criticism towards
“headquarter” orientation
- low rate of localization
- implementation speed
department via focused IT
company to totally
outsourced IS services
Figure 5. Internalization problems resulted in strategy revisions
In the new strategic plan a great deal of emphasis was put on the ICT
architecture of the new software generation. The plan was a technical
advancement of the earlier strategy. The business objectives remained
almost the same as earlier, but the role of operational chains was crystallized. The needs of regional cooperatives were taken more into consideration, which caused changes in the software design. On the overall corporation level the process can be interpreted as the complementary SECi-cycle.
The second update in 1993. The second revision of the IR strategy was
made in 1993. Again the new strategy built on the earlier plans and their
realization. The main reason for updating was that there were concerns
about the implementation of the new software. In the planning process the
integration between business and ICT worked very well, but during the
implementation phase the interaction was not high enough. The business
managers did not take the ICT questions enough into their agenda of
important decisions. The support of top management was, however,
always evident and made the implementation easier.
Both business and IS management felt that the progress was not fast
enough, and something should be done. The reasons for this situation were
not clear, therefore a new project was needed. The regional co-operatives
also felt that the systems development had been done in a too centralized
way.
EMIS approach was used in this stage in two ways. The researchers interviewed top management to find out their views on how strategy implementation could be improved. They also interviewed managers of area
184
Exploring Patterns in Information Management
cooperatives to chart their opinions on the role of the information systems
from their perspective.
Has the perspective
changed and should it be
altered?
A written IR strategy plan
created at the end of 1988
and updated 1993
Rethinking premises
Trying to localize and
take account of areas
- Problems with
and criticism to
“headquarters” orientation
- Expectations on faster
application development
and progress
From internal IT
department via focused IT
company to totally
outsourced IS services
Figure 6. Reconsidering the basic assumptions of the strategy,
based on sceptical user attitudes.
On the basis of these interviews and the discussion in the project group
more emphasis was put on the development of applications needed in the
local supermarkets and department stores. The quality and costs of the
project work was emphasized. Also an integrated approach to the earlier
decentralized information systems was introduced. The following applications were developed:
•
•
•
•
•
•
management information systems
chain management systems (1993)
home shopping test system (1993)
logistical systems (1994)
management accounting systems (1995-1997)
office systems (1997).
During the implementation process there has been a lot of discussion on
how S-Group’s information systems serve users and how competitive they
are compared to other systems. Business managers have not been totally
convinced of the quality of the information systems’ work. With an interactive planning process it was, however, possible to create a good starting
environment for the new ICT investments.
With NT-lenses this update could be seen as the iteration of the sECIcycle.
Kerola, Reponen & Ruohonen
185
Summary of Knowledge Creation Interpretations
In the following there is a summarizing interpretation of the whole IR
strategy creation and implementation process over a decade. The purpose
is to offer a holistic view of the strategy development and use.
Socialization. A socialization process was used most clearly and strongly
in the early stages of the strategy development, in clarifying the business
requirements for the use of ICT. At the commencement and surveying
stages a wide participation was aimed at. Numerous stakeholders had an
opportunity to influence the planning process. Socialization was also twodirectional so that information was both collected and offered in the planning sessions. The tools of socialization were interviews, teamwork, lectures, discussions and meetings. The objective was to obtain relevant and
business-related internal guidelines for ICT utilization.
Socialization of strategy issues was supported with a series of seminars
during the whole process. Those seminars tried to gather key business and
ICT people from all over the organization and provide key concepts and
“language” for their development efforts. Seminars were organized both at
headquarters and in other more locally based regions.
The originating bas and their updates have not been explicit in the whole
process. In principle, the organizational form and the business idea of the
case enterprise are highly sensitive and challenging for selecting originating bas. In this case the natural internal tensions between the key stakeholders were purposefully left out of the consideration, especially in the
early phases of business strategy development.
Externalization. Based on the surveying stage, the project group planned
directions for the use of ICT. These directions were still on a very general
level and therefore it was easy for everybody to accept these objectives.
The real test was later, when the plans were implemented and the changes
were made to the earlier operations. The CEO clearly used ICT as one of
the tools to make relatively radical structural changes in the S-Group.
Some of them were implemented with attraction of technology, but also
with clear background reasoning on what to do.
The goals of the IR strategy are actually far more demanding than just creating a plan. Deriving an applications development portfolio is seemingly
externalization of the strategy vision. This was achieved through expert
discussions and with a large comment-and-revision round after strategy
workshops. The process was more or less a committee-type of work during
186
Exploring Patterns in Information Management
which the wordings and structures of the strategy document was evaluated.
However, no assessment of understanding of strategy intentions was made.
The definition of application portfolio was more ICT centered, and the ICT
manager was one of the key persons. The project group followed and
supervised the development; the actual planning work was made by their
own and outside ICT experts. The final result of this definition stage was
an IR plan with an implementation schedule.
The interacting bas and their updates included the essential co-effort
between enterprise personnel and academic facilitators. This decision especially affected interaction between the CEO and the leader of the research
team. This included also the decision to utilize the EMIS-model.
Combination. The next step, after externalizations in the large, was to
develop concrete action plans for the explicit objectives of ICT business
use. This included a combination of stages to make the changes happen.
Architecture planning requires a stable view of the grounding of IS operations, i.e. long-term development of information systems. The ICT
department started to prepare IS infrastructure in order to add and develop
strategy-based applications. This was rather complicated due to a dispersed
regional decision-making structure. The SOK could not command all their
regional cooperatives to invest in new integrated information systems
while all of them had independent area activities. Selling the idea of the
new architecture and software generation to all stakeholders was the challenge. This was done with a combination of the influencing power of the
CEO and discussions and seminars.
The situation was problematic because of the different maturity of existing
information systems through the S-Group. Some business chains naturally
preferred to use their own systems and the others had their own IS culture.
It meant that some of the chains had to unlearn and relearn some IS features and some had to learn both new business and IS activities (see more
Ruohonen 1991).
The period after accepting the first IR strategy was clearly dominated by
ICT experts who built a new business chain oriented ICT architecture in 56 years. This included a number of difficult phases. Some application
development projects failed, some totally new start-ups had to be made,
and some technological platforms had to be changed due to the evolution
of ICT. However, the key ideas of the IR strategy plan remained and made
it easier to stay on the development path in a coordinated way.
Kerola, Reponen & Ruohonen
187
The highest risk in long-term development was to forget or neglect previous strategic thought just for the case of new technological innovations.
Diffusion of any technological innovation takes years and therefore it is
important to ground the development ideas carefully before starting.
This summary of combinations is highly adp-centred in its content. The
cyber-bas were not significant in the whole process.
Internalization. The third objective, a cultural change, was the most farreaching as it included learning of the social system of the S-Group. This
was not really clear to S-group people since human resource development
was centralized and developed training and learning programs for daily
business needs. However, some ICT-oriented issues were included in the
programs. Some of these challenges were noted during the strategy process
workshops. Key managers were invited to discuss and clarify their views
in executive meetings. Executive interviews also revealed some of these
internal, cultural contradictions. However, more work should have been
done to prepare the organization for cultural change, which was waiting
after ten years of development time. We must remember that the final outcome of the strategy was cultural change, and not necessarily the first draft
of the strategy plan.
In addition there was a learning challenge for regional cooperatives, too.
They had to rethink their own IS activities and interfaces with the joint
ICT architecture. This part of the development was partly neglected due to
the independent nature of these organizations. Some of the regional cooperatives started quickly, some were skeptical and careful in their development, and some resisted the Group activities. This was naturally reflected
in IS education and recruitment activities in the field. The Group offered
IS education and consultation services, but these were not heavily used due
to skepticism towards new IS ideas.
As a summary, the internalizational activities were the most underdeveloped in the whole process and the exercising bas were not transparent.
Results, Conclusions and Evaluation
The research and practical results of the longitudinal action research effort
are twofold and intertwined. We summarize the research results first, then
the practical ones.
188
Exploring Patterns in Information Management
The retrospective and multiple use of theoretical NT-lenses has ‘opened
the eyes’ of researchers and produced some crystallized findings. The
major results are as follows:
• Thinking knowledge-creation in seci-cycles produces holistic views.
Especially, it clarifies the neglected and underdeveloped sub-processes
and supports the assessment of balanced development
• Think more about combination during socialization and externalization
processes. In the vision or perspective creation phase, more ideas
should be developed to reflect the desired architecture of the organization. Even rough illustrations of forthcoming architectures help strategy
process participants see the combination of business and IS activities.
This is difficult when the main effort is focused on plan creation, i.e.
writing and rewriting the strategy document. However, when the strategy document is “ready” the human resources implications should be
regarded and preferably included in the IR strategy. This will help
internalization of new business and IS practices.
• Think more about internalization during externalization process. Usually after strategy creation, IS/IT experts start to develop systems and
final users are “left in peace” to do their business and wait for new
innovative systems. The problem is that learning of new ICT-enabled
information systems demands time and increased internal motivation.
Therefore human resources experts should be linked more tightly to IR
strategy processes.
From a practical point of view the longitudinal case described above is an
example of successful ICT planning and implementation, resulting in
increasing competitiveness. In the case company ICT has been used intentionally to open new roads and to redesign processes. Internalization of
new objectives and procedures has been, on the one hand one of the main
objectives, but on the other hand it has been one of the most difficult
problem areas. Interactive strategy generation process has been seen as one
of the mediators in creating shared values for ICT utilization.
We can show some of the results with facts, like an increased market
share in slowly growing markets, improved profitability, and an
increased number of members of the cooperative. We can also see the
radical development in the ICT sector, such as a very advanced customer
bonus system, new point-of-sales systems in the markets, redesigned
logistics, and early trials with e-commerce. S-Group has developed
highly integrated information systems with flexible reporting features. It
has also been an early adapter of outsourcing ICT services. Operations
Kerola, Reponen & Ruohonen
189
have been so convincing that some competitors have imitated them. The
use of different bonus systems has increased and chain models became
stronger than earlier. S-Group has, with its pioneering example, strengthened these trends.
We have also observed a change in thinking about the role of ICT in business operations. We have interviewed managers and personnel in different
parts of the group to find out their attitudes towards using ICT in competition. We have clearly noticed that their conceptual thinking has changed
and understanding has increased in integrating business objectives and ICT
potential. It is, however, difficult to show and to prove this change in
internal “tacit” thinking. The changes happen slowly and people do not
necessarily notice them by themselves. An interactive strategy generation
and implementation process has clearly contributed to the learning effect
of ICT’s potential.
ICT has been used in S-Group to foster significant changes in operational
models. Based on these changes its competitive position has improved in a
remarkable way and it has gained market share. As a matter of fact ICT is
frequently used in breaking down barriers and implementing changes. This
often causes resistance to accepting new technology, since ICT is regarded
as the reason for changes, not as a tool of implementing new thinking. In
this case we do not know what would have happened with some other
strategy, but it is evident that continuing the old operating models would
have resulted in market failure.
In implementing the new software there have been conflicts and disagreements, mainly concerning decision power and the degree of decentralization. The socialization process of implementation was not as impressive as
in the planning stage. An interesting question can be raised, how should
the interaction in the implementation stage be organized?
Behind the success of the strategies there is enough shared vision about the
objectives and measures to obtain them. In the beginning of the process the
use of the EMIS model increased belief as to the possibilities of ICT in
supporting business. At the same time the shared objective to utilize ICT
was accepted. Using an action research approach, the researchers have had
an opportunity to influence the decision making in the case enterprise, and
to experience its changes, conflicts and learning themselves. The utilization of NT-interpretations was tested and retested. With this first paper we
are able to share early experiences with other researchers – and some
practitioners, too.
190
Exploring Patterns in Information Management
References
Brancheau, J.C. & Wetherbe, J.C. (1987) “Key Issues in Information Systems
Management”, MIS Quarterly, Vol. 11, No. 1, pp. 23-45.
Dickson, G.W., Leitheser, R.L., Wetherbe, J.C. & Nechis, M. (1984) “Key Information Systems Issues for the 1980’s”, MIS Quarterly, September, Vol. 8, No.
3, pp. 135-159.
Ein-Dor, P. & Segev, E. (1981) A Paradigm for Management Information Systems, Praeger Publications, New York.
Galliers, R.D. (1987) “Information Systems Planning in the United Kingdom and
Australia – A Comparison of Current Practice”, Oxford Surveys in Information
Technology, 4, 223-255.
Galliers, R.D. (1991) “Strategic Information Systems: Myths, Reality and Guidelines for Successful Implementation”, European Journal of Information Systems, Vol. 1, No. 1, pp. 55-64.
Hirschheim, R. (1982) Information Management Planning in Organizations, London School of Economics, Working paper, London.
Hult, M. & Lennung, S. (1980) “Towards a Definition of Action Research: A Note
and Bibliography”, Journal of Management Studies 17(2), 241-250.
Kerola, P. & Reponen, T. (1996) “Action Research as an Intermediary Knowledge
Work Method between Information Systems Science and Practice – A Case
Study on Information Management Strategy Planning”, Publications of the
Turku School of Economics and Business Administration, Series Discussion
and Working Papers 7.
Lederer, A.L., & Sethi, V. (1996) “Key Prescriptions for Strategic Information
Systems”, Journal of Management Information Systems, Vol. 13, No. 2, 35-62.
Lundeberg, M. & Sundgren, B. (1996) (eds) Advancing Your Business –People
and Information Systems in Concert, EFI, The Economic Research Institute,
Stockholm School of Economics, Stockholm.
McFarlan, F.W. & McKenney, J.L. (1983) Corporate Information Systems Management – Issues Facing Senior Managers, Dow Jones Irwin, Homewood, Illinois.
McFarlan, F.W. (1984) “Information Technology Changes the Way You Compete”, Harvard Business Review 62(3), pp. 98-103.
Mintzberg, H. (1987) “The Strategy Concept I: Five P’s for Strategy”, California
Management Review, Vol. 30, No. 1, pp. 11-24.
Mintzberg, H. (1994) The Rise and Fall of Strategic Planning, Prentice Hall.
Niederman, F., Brancheau, J.C. & Wetherbe, J.C. (1991) “Information Systems
Management Issues for the 1990s”, MIS Quarterly, Vol. 15, No. 4, 475-500.
Kerola, Reponen & Ruohonen
191
Nonaka, I. (1994) “A Dynamic Theory of Organizational Knowledge Creation”,
Organizational Science, Vol. 5, No. 1, pp. 14-37.
Nonaka, I. & Takeuchi, H. (1995) The Knowledge-creating Company, Oxford
University Press, New York.
Nonaka, I. & Konno, N. (1998) “The Concept of ‘Ba’: Building a Foundation for
Knowledge Creation”, California Management Review, Vol. 40, No. 3, pp. 4054.
Porter, M.E. (1980) Competitive Strategy, Free Press, New York.
Porter, M.E. (1985) Competitive Advantage, Free Press, New York.
Reponen, T. (1994) “Organizational Information Management Strategies”, Journal of Information Systems, Vol. 4, pp. 27-44.
Reponen, T. & Auer, T. (1997) “Information Systems Strategy Formation Embedded into a Continuous Organizational Learning Process”, Information
Resources Management Journal, Vol. 10, No. 2, pp. 32-43.
Ruohonen, M. (1991) “Stakeholders of Strategic Information Systems Planning –
Theoretical Concepts and Empirical Examples”, Journal of Strategic Information Systems, Vol. 1, No. 1, pp. 15-28.
Sanchez, R. & Heene, A. (1997) “Reinventing Strategic Management”, European
Management Journal, Vol. 15, No. 3, pp. 309-316.
192
Exploring Patterns in Information Management
Blanksida
— 12 —
Patterns in Information Management:
A Multi Level Analysis of Swedish
Companies
Kristina Nilsson
Introduction
This chapter presents a number of patterns that surfaced during a series of
interviews around the subject of Information Management, held in 17 large
Swedish multi-national companies the year 2002 (Ulbrich & Nilsson,
2002). The content is mainly empirical but the findings are analysed
according to one of Mats Lundeberg’s frameworks: different levels of
abstraction (Lundeberg, 1993). Besides presenting the empirical findings,
the purpose with the chapter is to picture how this model can be used in
practice and to discuss what kind of patterns may occur while using it.
The Theoretical Part – the Model
The model that is used to analyze the empirical material is a situation
framework. Figure 1 shows the model.
The model contains seven different abstraction levels and focuses on a
situation chosen by the user of the model. It can for example be used to
analyze a perceived current situation from different individuals’ perspectives. The logic of the model goes like this: “There are a number of persons (stakeholders or interest group members) with certain characteristics.
These persons behave in different ways in order to influence the achievement of certain results. In order to achieve results in business activities are
carried out. These activities need information. The information is provided
by information systems, which in turn interact with the environment in
order to operate” (Lundeberg, 1993, p. 5). The reasoning can also be done
the other way around: that is, the environment provides information systems with information that influences and is needed in the activities, which
leads to the results, and so on. Another key feature, beyond the different
194
Exploring Patterns in Information Management
parts or levels, is that the model stresses the influence between the levels;
they do not exist in a vacuum.
Persons
Behaviour
Results
Activities
Information
Info. System
Environment
Figure 1 The situation framework – seven different levels of abstraction
There are no universally true levels; different levels are important in different situations and different persons stress different levels, more or less.
Or more precisely, their fields of interest, day to day problems and challenges, will relate to different levels in the model. A technician working
with IS development will probably express more concern that relates to the
IS/IT level than a marketing person working in the same company. The
latter’s daily work will probably address the levels of results or activities.
People will also be more or less aware of the influence between the different levels or areas, how for example one single action within marketing
affects the results of the company, or how the executives’ statements and
goals for the company will influence day-to-day activities in business processes.
The Use of the Model in This Particular Study
The multi-level model may be used in many different situations or with
different purposes. In analyzing empirical material, it may for example
be used in order to see or find different patterns. It can be used in a single
interview, to see if an individual expresses problems or issues relating to
Nilsson
195
a certain area, or likewise important to highlight if one or several levels
are missing. This could, for example, help us to: understand the problem,
the interviewed person’s view of reality, or the investigated business in
general. But it could also tell us that we need to gather more information
before we are able to draw any conclusions about patterns or the problem. Material from different interviews can be compared, and then yet
another pattern may occur on an aggregated level. This may hopefully
give us an even richer picture of the studied area. The model is based on
thoughts that different levels are important per se and that they influence
each other, but all taken together – they give a holistic picture. This will
in turn give more information than the single parts (cf. systems thinking,
Lundeberg 1993, p. 51-).
There are, as stated earlier, no universally true levels. If you then use them
as a tool to get a better understanding of a chosen part of reality – who can
then judge whether or not the pattern or the picture given by the model
corresponds to reality? We all have different perspectives on this reality.
One important checkpoint is of course the persons interviewed. They have
to recognize the description. On the other hand, the picture may give them
new insights, things they haven’t recognized before. The model, in this
case, adds new knowledge. Sometimes they cannot be expected to understand the tool the researcher is using; then it primarily gives the investigator new insights. Therefore one important prerequisite for using the model
is to be clear about what perspectives or whose perspective you are using
when applying it. In addition, we must consider the purpose, what do we
intend to achieve by using it?
This particular study was done from an Information Management (IM)
perspective. The researchers performing the interviews were from the IMarea. The mission was mainly to discuss issues relating to this area. The
interviewees were not from the IM area; most of them were working
within management accounting, finance and control. Several of them were
members of the group executive board and as such company leaders.
However, a majority also have to manage complex issues relating to
Information Management. For example, a Chief Financial Officer may be
the purchaser of an ERP-system which will influence major business processes.
The empirical material is presented below according to the different
abstraction levels. The section starts with a brief description of the background of the study, its purpose, the sample of companies and interview
method.
196
Exploring Patterns in Information Management
The Empirical Part – the Interviews
The sample companies and interview persons were predefined. The companies belong to a network initiated by a group of professors representing
Management Accounting and Control and Information Management at
Stockholm School of Economics (SSE). Professor Mats Lundeberg was
one of initiators. The group has met regularly at SSE since 1993 and the
intention with the network was to create a forum for the exchange of ideas
and experience between industry and academia, and between individuals
working in different industries facing similar problems and challenges.
Normally the largest companies in a wide variety of different industries
were invited to join the network. Only one company from the same industry should be represented in the group because the intention was to create
an open non-competitive climate among the participants. In 2002, seventeen companies were represented in the network by either the Chief Financial Officer (CFO), Chief Information Officer (CIO) or Chief Group Controller. The majority of the companies were multi-national and their businesses world-wide.
Purpose
The purpose of investigation was to gather information and input for forthcoming meetings in the group. Several participants were, at this point in
time, newcomers; they had taken over the membership from a former colleague, but there were also a few new companies in the group. The main
idea with the interviews was to gather information about the present situation in the companies. For example challenges, opportunities, problems
and threats facing the company and the individuals as professionals in their
work, and within the Information Management area. The interviews should
also cover major running or planned change and/or business development
projects within the organization.
The empirical material was gathered through semi-structured interviews
which were sent back to the interviewees afterwards. The interviews were
done with the representative in the group at SSE or another person he or
she recommended. The latter could for example be the Chief Accountant
or the Group Controller. All interviewees were working at the head office
or company group office. In several cases the interviewed persons also
belonged to the Executive Group in the company. Twenty individuals were
interviewed in total, i.e. in some cases more than one person participated in
the interview or was interviewed at another point in time. All interviews
were done between February and May 2002.
Nilsson
197
The interviews took approximately 1.5 hours and the individuals were
asked to speak freely around a couple of pre-defined areas, primarily
relating to their work and secondarily to the area of Information Management. The sessions were always initiated by asking the individual to
describe their area of responsibility and what they presently worked with.
This chapter will present the findings in line with Lundeberg’s levels of
abstraction as described above. The findings are presented under the
headings: Persons and Behaviour, Results, Activities, Information, Information Systems and Environment.
Person and Behaviour
“Person” refers to individuals’ beliefs, values, knowledge, and “behaviour”
to the actions taken by the individuals. In this presentation the description
of “person and behaviour” is done together. The material did not offer
enough details to do a meaningful, separate description of each level.
Executives’ Knowledge of IS and IT
The interviews showed that higher executives feel great respect for and
have a good understanding of the business application of IT. For example;
how IT can improve and change business processes, refine products and
services but also be a source for creating new business ventures. This
insight and understanding did not include technical knowledge and competence. These seemed to be rather low. One interviewee implied that technical competence is low by definition since the executives are not educated
within the field. For example, they cannot judge whether an IT-related idea
needs a week, a month or several months to be implemented.
It is noteworthy that several interviewees mentioned that the technical
competence seemed to be higher in executive groups with members who
have experience from working or living in the USA. One person thought
that the executives’ technical competence is higher in American companies
compared with Swedish.
However, one could argue that lack of technical competence is not an issue
that needs attention because these questions should not be handled by top
executives, except for cases when the CIO is part of the group executive
board (which did not seem to be the case here). A majority of the companies have a CIO and or an IT-council. This person or group analyzes technical issues and prepares information material to the executives for deci-
198
Exploring Patterns in Information Management
sion making. This person or group therefore complements the executive
group with technical competence.
In the investigated companies the CIO’s or IT-council’s role seemed to be
counselling and control. The CIO should for example see that decisions and
activities regarding IT-issues were in line with company-wide guidelines. In
decentralized companies he or she was the communication link between the
business units and the group company office. In these cases the CIO’s major
assignment was to investigate different business units’ information and IT
needs and communicate them to a company group level.
What kind of IT-related questions are, then, the executives expected to
handle and understand?
In the investigated companies these questions seemed mainly to be strategic. The reason for being discussed and decided upon in the Executive
Group is not primarily that the issues relate to IT, but rather that the
investments exceeds a certain amount. This requires a decision at the top
level. Questions that have been handled by the Group Executive Board are,
for example: whether or not to invest in an Enterprise Resource Planning
System (ERP); whether or not to outsource IT-operations; whether or not
to invest in Electronic Business (B2C or B2B); but also policy questions
on what kind of information should be spread over the Internet about the
company, its products and services.
Results
This level of abstraction refers to the results that people try to achieve. It is
also the outcome of the business activities, or the goal of these activities.
In this investigation the result refers to business strategy and IT, valuation
and evaluation procedures of IT-projects and outsourcing as an optional
way to increase the efficiency of the IT-function.
Strategies
A majority of the 20 interviewees did not know if they had a company
wide IT-strategy. Several of them also questioned if this was necessary.
They meant that the overall business strategy included the IT-strategy. The
need for using IT is a consequence of the business and therefore the ITstrategy should be embedded in the business strategy. If a separate ITstrategy is needed it should contain, for example, constraints and policies
for the organization such as the infrastructure regarding operating system,
Nilsson
199
software and hardware investments. A couple of companies were preparing Internet-strategies, which also could be seen as business strategies, in
which IT is used as a tool for implementing the strategy.
Investment Control
There seemed to be a large spread among the companies when it came to
monitoring and following up the cost for planned and implemented IT-projects. Some companies followed up their projects thoroughly with dedicated controllers, others did nothing like it. The latter is common because
the evaluation is forgotten about when, for example, a system is up and
running. The individuals are engaged in new projects. The pre-planning
process requires a lot of time and effort, there is little energy left for
evaluations, and as long as no one asks for it, it is not done.
Previously all IT-investments were handled as direct costs and affected the
result directly. The bookkeeping rules have changed now and companies
are forced to treat IT investments like other investments, and gradually
depreciate the amount spent. Some of the interviewees thought this is good
because it gives a more relevant picture of the companies’ IT-portfolio and
its value. Others believed that this may lead to overspending because the
units do not have to face the total cost at once. It seemed however, that a
common theme in the companies was to write off the investments as fast as
possible.
Outsourcing
All interviewees mentioned that outsourcing is treated as an optional way
to go when discussing IT-issues and -investments. However, in reality, it
seemed almost only be done for IT-operations. Outsourcing can for example be to manage and support hard- and software.
Outsourcing was also seen as challenging from a cost perspective. The
interviewees thought that it was difficult to monitor and follow up the
costs and evaluate whether outsourcing really resulted in the promised
resource and cost savings.
Activities
This level refers to activities taken by individuals to reach certain results.
In a company setting it may, for example, refer to business activities performed to reach the intended results of the business.
200
Exploring Patterns in Information Management
Current Activities and Projects in the Companies
In this investigation, “activities” mainly refers to a number of ongoing projects that relate to information management. The ordinary business activities
were described by each interviewee, but it goes beyond the purpose of this
chapter to describe these for each one of the companies.
E-business, E-procurement and E-invoices
A number of interviewees stressed that electronic business is affecting or
going to affect their way of doing business in the short or the long run.
However, it turned out that our interviewees mainly referred to electronic
invoicing as electronic-business and as an ongoing project within the company. They were convinced that the use of electronic invoices would
increase a lot. Several of the participating companies were already using it:
the suppliers demanded it or they used it to increase the process efficiency
internally and externally, i.e. among the business units or by forcing the
customers to accept electronic invoices.
The system support for e-invoices seemed to vary a lot. Some companies
scan the paper based invoiced and distribute it electronically afterwards,
some integrate the scanned invoice with a Workflow Management System
to make the distribution easier. A few companies use more advanced systems for e-procurement which cover the whole procurement process.
Improved Information Support through Data Warehouse
An implementation of a Data Warehouse (DW) was planned in several
companies. In a Data Warehouse the data or information is gathered from
different sources or pre-systems and stored in a common database. It eases
the access to important information for decision makers such as executives
and controllers. It often contains key figures such as management
accounting information, employee numbers, production statistics and customer satisfaction indexes. The content may overlap with information supplied by management accounting information systems or ERPs, but sometimes not. One important reason for implementing a DW is that it is seen
as more accessible and flexible. It offers a more resource effective way to
produce ad-hoc reports with short lead times.
Improved Information Infrastructure
Approximately 50 percent of the companies in the investigation used or
were planning to implement an Enterprise Resource Planning System to
Nilsson
201
improve the information handling and information systems integration in
the company. The rest had no such plans.
The interviewees were aware of that the most convenient way, from a
company wide perspective, would be to use one common system in the
whole organization. However, it was not possible. The reasons were different standards, rules and regulations for companies around the world,
and that system suppliers and support also differ in different parts of the
world. These circumstances made it more difficult for different units to use
the same system as the mother company. The goal was therefore to minimize the number of ERPs on a global basis. The companies tried to identify and gather around one common system for a specific part of the world
or country. It was common that a company could have fout to ten different
ERP-systems operating around the world.
All interviewees stressed the large amount of work, time and cost related
to implementing ERP-systems. A long pre-planning process is required to
ensure that the company, units or functions gain the most from the investment. The planning process also entails whether or not they should invest
in a full-scale version and/or in one or a few modules. The most commonly
used modules were finance, personnel and controlling. No company had a
full version implemented. The reason for not choosing a full scale implementation was the level of costs and complexity. None of the visited companies chose a “big-bang” implementation but rather an incremental, stepwise one.
Shared Services
A common project that was under investigation and implementation in
several of the companies was the creation of a unit or function for shared
services. This means centralizing activities that are common for different
business units such as accounts payable, accounts receivable and/or
invoicing. But there could also be other common services within a company. The services provided by this centre are normally only offered internally but they can also be offered to customers outside the company. The
name used for this function or service differs among companies, for example “internal service provider” or “advanced service administration provider”.
The purpose of a shared service centre is mainly to support the business by
providing high quality services, cheaper and more efficiently compared to
when all individual business units handle it separately. Some of the interviewees questioned the idea. They had not experienced the promised
202
Exploring Patterns in Information Management
advantages or cost savings. Another disadvantage mentioned was that the
responsibility for the relevant areas would be taken away from the business
units which might lower the quality, since they could not be held responsible for errors in this area anymore. Other interviewees stressed the advantages: such as that the change requires definition of common business processes in the company and this may increase quality and decrease lead
times significantly and lead to a “best praxis”.
Information
This level of abstraction refers to information individuals and business
processes use.
Executive Information – Lead Times for Recurrent Reports
In this investigation the questions regarding information refer to research
done in the same companies between the years of 1991-1995 (Nilsson,
1999). The research covered recurrent management accounting reports, the
reports’ frequency and lead-times.
In 1993 a survey was sent to the top 210 largest companies in Sweden. The
purpose was to gain a picture of recurrent management accounting reports,
the length of the lead-time for monthly and quarterly recurrent reports, and
the receivers’ (CFOs or their equivalents) opinion of the lead-time. The
latter refers to whether the executives thought the lead-time should be
compressed, if it was reasonable at is was, or if it should be lengthen.
Since almost all the companies in the network at SSE also participated in
the investigation in 1993 (16 out of 17 answered the survey of 1993), the
opportunity was used to follow up the previous study results. The intention
was to investigate what the changes were almost 10 years after the initial
study.
In year 1993, 147 companies (of 210) distributed a monthly recurrent
management accounting report to the executive group. The average leadtime, after the end of the calendar month, was 13 working days (n= 147 of
210, standard deviation 6.3 and the spread from 3 to 35 working days). 41
% of the respondents thought that the lead time should be compressed by
an average of four working days within 12 months. 57 % thought that the
lead time was good enough and 2 % thought that it should be increased.
In 2002, the average lead-time for the group of sixteen companies that participated in both surveys was nine working days after the end of the month.
Nilsson
203
In 1993 the lead time for the report in this group corresponded to an average of thirteen working days. This means that the lead time has changed by
an average of four working days in nine years.
The results show that the group that expressed no need for changes in 1993
had compressed the lead-times by more days compared to the group that
expressed a wish to compress it. The “no change” group has on average
compressed the lead time by seven working days (six of the sixteen companies). The “change” group has on average compressed it by five working
days since 1993.
Quarterly Reports
These reports are compulsory for companies noted on the Swedish Stock
Exchange in Sweden and are, in contrast to the monthly reports, public and
official information. These reports are more extensive than the internal
since a full balance and income sheet is required. The internal report may
be extensive too but in these cases the information need of the executives
decides the content of the report.
The public reports get a lot of attention in media. Just before they are
released, analysts speculate about their content. Once the information is
released, this may affect the stock rates to a smaller or larger degree. It is
important to minimize the risk that the information leak out before the
publication. One way is to compress the lead-time, the more time available
for consolidation of the material the greater is the risk that it will leak out.
But the pressure from investors and the stock exchange to get more frequent and fast financial information has also forced the companies to compress the lead-times further.
In the investigation of 1993, 106 out of 210 companies reported that they
publish quarterly recurrent financial reports. The average lead-time was 33
working days after the end of the third month (standard deviation 10.7, a
spread from 12 to 60 working days). In this case only 23 % thought that
the lead-time should be compressed with in average five working days
within twelve months. In the current investigation 2002, fifteen companies
published the report and the average lead-time was 19 working days after
the end of the reporting period (spread 15 to 28 working days). The conclusion is that there is still a huge spread in lead-times, but still, the leadtimes have changed a lot compared to 1993. One explanation given to the
little interest in compressing the lead-times 1993 was that it was mainly
the available dates for the Board of Directors to meet that, in the end,
decided when it would be published. This is probably still relevant to some
204
Exploring Patterns in Information Management
extent, but the pressure from investors, banks and the stock exchange is
much bigger today. If a company wants to meet their present and future
investors’ needs they will have to adjust to their demands towards current
and frequent information.
Information Systems
This abstraction level refers to the information systems that collect and
process information to and from the business environment.
Information Systems in This Study
This level does not contain any empirical findings. One explanation is that
the area studied is information management which means that several IS
related issues are placed on the “activities” level. In a more ordinary business analysis these projects could probably fit the IS level instead.
Environment
This level of abstraction refers to the immediate business environment.
This environment interacts with stakeholders within the company, puts different kinds of constraints on it, or is a source for new business opportunities.
The Studied Companies’ Use of IT versus the Competitors’
There was a great variety among the companies when it came to their
views of their positions and use of IT as a competitive tool when compared with the competitors. The majority thought that they were somewhere in between, not leading nor lagging behind. Only one of the interviewed persons thought that the investments they had done lately would
give them a competitive advantage in a short time span. There also
seemed to be some differences among industries. The automobile and
bank industry were, for example, seen as being in the forefront compared
to other industries in their use of IT in business processes and as a competitive tool.
The interviewees compared their companies with European counterparts
when asked this question. If the comparison was extended to include the
US market, the picture changed. The interviewees thought that the application of IT was more extensive in American companies and compared to
Nilsson
205
these, Swedish companies were lagging behind. An example mentioned
was extensive use of hand-held computers for sales support.
Another important issue that relates to the competitiveness is mergers and
acquisitions. The lead time, the time it takes to enfold new units into the
business and organizational structure, affects the competitiveness. Takeovers and mergers are rather frequent and resource intensive. The companies need to be able to manage them fast and smooth to not lose customers
and competitiveness. Some companies had strict procedures on how to go
about it, others not.
Discussion and Conclusions
In this section the use of the model is discussed. This is followed by a
summarized presentation of the findings, i.e. the patterns that occurred
while using the model.
There was no intention at the start of the study to use the model; still, the
application of the model shows that almost all of the levels are present. In
this case, when focusing on IM, no empirical material fits into the information systems level (including IT). This seems reasonable, since the chosen focus area is Information Management and the business in this case,
i.e. the abstraction level “activities”, covers ongoing projects within the
field of IM. We could also conclude that the interviewed persons view
activities and projects with an IS or IT component from a business perspective – they are business driven not technology driven. The interviewees are not IS nor IT-specialists, they can not be expected to express issues
in IS-terms. This line of reasoning is supported by the statements done
around the strategies – IT-strategies are embedded in the business strategy.
There is no separate IT-strategy in a majority of the companies. This is in
line with the way ongoing IM-projects are handled in the model.
Looking closer at the mentioned projects there are only two projects, Data
Warehouses and ERP-systems, that are technology focused, at least more
compared to the other mentioned projects. If we look at the purpose with
the projects they are business based. For example, a company is implementing a Data Warehouse to improve information support to executives.
The reason is that the executives need the information in their line of work.
If we look at the relationships between the different levels, a consistent
logic occurs: the leaders are interested in IT and know about its relevance
for business, they do not treat it as a technical matter (person and behaviour). They evaluate IT in terms of how it will increase efficiency and
206
Exploring Patterns in Information Management
effectiveness, lower cost, sustain or create new competitive advantages.
The business strategy directs the IT-activities and IT-investments (results).
The ongoing activities are in line with the strategy and intended results, for
example to invest in an ERP-system in order to increase the efficiency in
business processes (activities). Investments are done in ERP-systems and
or Data Warehouses to improve information processes – the companies for
example want to compress lead times and distribute current, frequent and
high quality information to investors and executives (information). The
environment, in terms of immediate competitors, forces the companies to
continually evaluate their business processes and use IT to gain or sustain
competitive advantages (environment).
One important question is whether or not the use of the model makes any
difference to the outcome or conclusions, as compared to the result when
the model is not used? In the first analysis of the results (Ulbrich & Nilsson, 2002), the findings were grouped according to five different themes,
which in turn reflected the questions that were asked during the interviews.
The themes were; 1) lead times for recurrent management accounting
information, 2) IT-issues at the executive level 3) current projects in the
organization 4) valuation of IT 5) the competitive position. The empirical
findings are the same, here analysed and presented in a different way. The
major difference between the two approaches is that the initial report lacks
consistency in presentation. The multilevel analyses highlight the different
parts (different levels) and that these influence each other. It also gives us
an overall picture of the area when all levels are added together. The
analysis shows that Swedish executives in a wide variety of industries tend
to evaluate IT from a business perspective. The analysis also shows that
the interviewees are consistent when they make statements about the topics. The model adds structure, and the evolved patterns may be more difficult to identify when the model is not used. Of course, another model may
have added this insight too, but that goes beyond the purpose of this chapter.
Finally, when discussing the results and patterns, we have to take the
researchers into account, i.e. the persons who apply the model and sort the
material into the different levels. Of course, this may influence the way the
model is handled. It is important to be frank and clear about the chosen
perspective and the reasons behind using the model in a certain way. The
intention was to apply the model to empirical findings which described a
number of companies’ activities within the Information Management area.
The analysis has shown that this is possible and useful. Of course, one
Nilsson
207
should be careful with extending the results outside the studied companies
and area.
Conclusions – Patterns in Short
On a general level the material tells us that Swedish companies treat and
view IT from a business perspective. IT-investments are evaluated in terms
of business use; business development and new business ventures. The
business strategy enfolds IS and IT activities. Executive groups complement their lack of technical knowledge by the help of a CIO or an IS/ITcouncil. The investigation shows that a number of large Information Management projects are in progress, that are intended to alter, extend and or
improve the business processes. The need for these changes seems to be
found in changing business needs, not primarily in obsolete technology.
In short, the following patterns evolved from the interviews:
• Higher executives have an interest in information technology and a
good understanding of its significance for business and business development.
• Higher executives have less technical knowledge regarding IS/IT. This
is knowledge is provided by a CIO or an IT-council who prepares information and decision material around IT-questions.
• It is normally the required amount for investments that decides whether
IT/IS-related issues are treated in the Executive Group or not.
• IT-related issues are embedded in the business strategy. It is less common with a separate IT-strategy.
• Follow-ups and evaluations of IT-projects are detailed and thoroughly
done in some companies but in others very little are done in this sense.
• Outsourcing is seen as a potential alternative when deciding on ITactivities in general, but in reality, it seems to be relevant only when it
concerns IT-operations.
• During 2002, several companies were implementing ERP-systems,
establishing electronic business and processes, and one common project
regarded electronic invoicing. Another project was to improve information support to executives and controllers by implementing Data
Warehouses. Yet another project was to evaluate whether or not to
implement shared services for common functions among business units.
208
Exploring Patterns in Information Management
• The information support to executives and investors has gradually
improved over the years. Internally distributed recurrent monthly management accounting reports are on average completed less than two calendar weeks after the end of the month. Public quarterly reports are
distributed to the market on average 3.5 calendar weeks after the end of
the reporting period.
• Swedish companies tend to equate their application of IS/IT to a competitive tool to their European and Swedish counterparts and competitors. They seem to think that US based companies are leading.
References
Nilsson, K. (1994), Rapportering i svenska storföretag 1993 – rapporteringstider
för kvartals/tertialsrapporter, delrapport 2, (Recurrent management accounting
reporting in Swedish large companies 1993 – lead times for quartly reports,
part 2), EFI Research Report, SSE, Stockholm.
Nilsson, K. (1999), Ledtider för ledningsinformation, (Lead times for Executive
Information), Doctors Dissertation, EFI, SSE, Stockholm.
Lundeberg, M. (1993), Handling Change Processes – A Systems Approach, Studentlitteratur, Lund, Sweden.
Ulbrich, F. & Nilsson, K. (2002), Frågor kring ämnet Information Management i
svenskt näringsliv våren 2002, (Information Management Related Issues in
Swedish Industry, the Spring 2002), EFI, Electronic working paper series,
SSE, Stockholm.
— 13 —
Some Issues in the Evolution and
Use of Conceptual Frameworks:
A Commentary on the Lundeberg
Framework1
Magnus Mähring
Introduction
This chapter discusses a conceptual framework for management of ITrelated change, developed by Mats Lundeberg (1992; 1993; 1995; 1996;
2000). While focusing this specific framework, the chapter also addresses
issues that pertain to the development, adoption, adaptation and use of
conceptual frameworks in general.
The Lundeberg framework has been used extensively in teaching as well
as in research, in particular at the Stockholm School of Economics, for
more than a decade. Over time, models in the framework have been
modified several times and its areas of intended and actual application
have expanded. In my view, these developments suggest at least three
reasons for providing a commentary on the framework, its evolution and
use:
First, the generality and versatility of the framework has proven to be a
major benefit to its potential and actual usefulness, while its concomitant
complexity can be an adoption threshold. A commentary on the framework
may provide additional avenues to understanding and using the framework.
1
The author wishes to thank Allen S. Lee, Christofer Tolis and Alf Westelius for
valuable comments on an earlier version of this chapter; the Computer
Information Systems Department and the Electronic Commerce Institute, Georgia
State University, for providing a stimulating environment in which to pursue the
ideas contained in this chapter; the Sweden–America Foundation and the Carl
Silfvén Scholarship Fund for research funding.
210
Exploring Patterns in Information Management
Second, while the framework has found a variety of uses, potential uses
remain unexplored and the use of the framework to date has not been
reviewed. By discussing current and potential uses of the framework, I
hope to provide input to further use as well as development of the framework.
Third, while frameworks are not uncommon within information management, issues pertaining to the evolution and use of frameworks are seldom
discussed. A discussion of this particular framework may serve to shed
light on some issues of general interest concerning conceptual frameworks.
I depict this chapter as a commentary because I do not intend to provide
anything near a complete description or discussion of the entire Lundeberg
framework. Rather, this is a short and personal selection of views and
comments on the framework and on its use, offering the reader either an
introduction to the framework or an opportunity to reflect on her/his view
of it. Correspondingly, the chapter offers observations and pointers regarding approaches to use and areas of use for conceptual frameworks in the
area of information management.
Below, I summarize briefly some of the framework’s key models and their
characteristics, as well as comment on intellectual roots and some underlying principles of the framework. After providing this basis, I comment
on drifts in purposes over time, review how the framework has been used
and discuss avenues for further development and application of the framework. Finally, I point to some considerations in the development of
frameworks as well as to applying frameworks in research within information management.
A Framework Apart:
Characteristics, Roots and Mechanisms
The core model of the Lundeberg framework is the levels of abstraction
model. This model builds on principles from mathematics and formal logic
(Whitehead & Russell, 1910), systems theory (Churchman, 1968), information theory (Shannon & Weaver, 1949), information systems theory (Langefors, 1973/1966) and theories on communication, learning and epistemology
(Watzlawick et al., 1967; Bateson, 1972, 1979). The basic idea of this model
(see Figure 1 for two typical generic versions) is that a certain situation (i.e.
a selection of phenomena situated in time) can be classified into process and
structure aspects and that these aspects can be “sorted” in accordance with
their “level of abstraction” (Lundeberg, 1993; 1996).
Mähring
211
A key claim for this model is that through sorting, different aspects of a
complex situation and the interrelationships between aspects can be better
understood. Execution of the sorting requires identification of entities as
well as determination of how they fit into the pre-determined framework.
Persons
Individual
Perspectives
Behavior
Results
Intersubjective
Conceptions
Activities
Represented
Information
Information
People
in Action
Business
Activities
Information
System
Physical
Data
Information Systems
Environment
Figure 1. Two Levels of Abstraction Models (from Lundeberg 1993; 2000)
While the framework is in practice open for redesign to fit a certain problem area, the standard presentation of the levels model emphasizes predetermined levels (Lundeberg, 1993). An example of use of the predetermined levels to structure a problem area can be found in Nilsson (2003).
An example of how the abstraction levels model can be adapted is shown
in Figure 2.
Individual
Perspectives
Behavior
Strategy
Methods
Improvement
Methods
Strategy
Processes
Improvement
Processes
Figure 2. Adapted Levels of Abstraction Model (from Andersson, 2003)
A central tenet of the levels of abstraction model is the interdependency of
structure and process and the repetition of structure–process pairs, which
212
Exploring Patterns in Information Management
arguably can be expanded infinitely. Another key characteristic is the
flexible or fluid nature of the relationships between levels. The relationship
between two levels is often denoted “influence”, without definition of what
that influence entails (cf. Lundeberg, 1993). Important, however, is that
influence is dual: the abstraction levels do not as such depict unidirectional
relationships or stipulate that influence is asymmetrical. However, many
examples used to illustrate the levels model have a bias towards top–down
influence, starting with the intentions, goals and priorities of individuals
(e.g. in Lundeberg, 1993).
This provides versatility but also reduces guidance regarding how to
apply the framework, especially regarding how to adapt the framework
for use in a specific context. The guidance provided is that levels should
be alternating “process” and “structure” levels and that the “level of
abstraction” determines the vertical placement of phenomena in the
model.
This raises the question of how to determine whether one entity is “more
abstract” than another entity, as well as what makes for a meaningful
sequencing of abstraction levels. Guidance for determining these issues
can primarily be found in examples and several examples suggest that
actorhood, ownership (hierarchy) and economic value can guide the
sequencing of abstraction levels (Lundeberg, 1993; 1996).
For example, Lundeberg (1996) structures a problem relating to telecommunication technology acquisition to include the hardware, communication requirements, information access requirements, business process
design, key performance indicators (lead time), and ultimately business
value expressed in the priorities and concerns of top management. Here,
the construction of a “problem space” seems to be guided by economic
value and by ownership over the business (with executives as agents acting
on behalf of principals).2
While the levels of abstraction model is non-temporal, it is closely linked
to the so-called X-model, which targets analysis of processes. The Xmodel does so by focusing process prerequisites (or inputs), process attributes and process outcomes (Lundeberg, 1993). Important for the usefulness
of the X-model is the structuring of a process into one or several sequences
of pre-conditions, process and outcomes (ibid.). A generic X-model is
shown in Figure 3.
2
Here, the levels model also resembles an ends–means hierarchy.
Mähring
Input
Process
Output
Relationship
Person
Preconditions
Behavior
Person
Outcomes
Task
Task
Preconditions
Task
Processes
Task
Outcomes
213
Figure 3. The X-model (from Lundeberg, 1993)
The linkage between the levels model and the X-model is that each stage
in the X-model can be described with guidance from the levels model. The
two categories that compound the levels model in the overview version of
the X-model can be called relationship and task issues, “soft” and “hard”,
or behavioral and non-behavioral aspects of the studied process. The guidance from the levels model for this analysis varies depending on which
version of the framework is used. Early versions of the framework use all
levels for each state, whereas later version separate structure and process
levels and thus vary what levels pertain to structure versus process states.
Another model in the Lundeberg framework is called the Y-model (Figure
4). In its basic structure, this model seems to resemble a “rational” decision model, distinguishing between actual and desired state, the problem
(the gap between actual and desired states, or change requirements), alternative solutions and action plans (cf. Simon, 1977; also operationalizations
e.g. in Hill, 1981). Used as an analysis tool, however, the Y-model does
not assume a sequential process through “stages”, a significant difference
towards a rational (as well as bounded-rational) decision model.
Current
Situation
Need
for
Changes
Change
Alternatives
Outcomes
Intended
Future
Situation
Figure 4. The Y-model (from Lundeberg, 1993)
214
Exploring Patterns in Information Management
Below the surface, the Y-model can be seen as reframing the abstraction
level model in the form of an ends–means hierarchy. Conceptually, goals,
change needs and change alternatives can be seen as a potentially infinite
ends-means chain (cf. Lundeberg, 1996; 2003). In this use, the focus
becomes one of selecting appropriate problem focus and problem contexts.
Like the levels model, this model leaves considerable discretion to the
user(s) regarding how the model is applied.
While there are other models in the framework, these three models can be
seen as the most central and other models do not substantially differ from
these in their characteristics, albeit in foci, which includes individual and
shared perspectives and a model depicting iterative action towards goal
attainment (similar to a cybernetic model).
Another type of characteristic that is important for the understanding of the
Lundeberg framework – and quite different from the aspects discussed
above – is how the framework is presented in writing. A brief look at the
use of language in most presentations of the framework serves to illustrate
a central aspect of the framework itself as well as of the basis for understanding the framework. Here are a few examples: “The better you are able
to perceive reality, the better you are prepared to act” (Lundeberg, 1993, p.
1); “Person outcome: All executives agree on what we want to achieve”
(ibid., p. 17); “Reality in a business contains a number of concrete phenomena – for instance, different persons, things and activities, that you can
observe” (Lundeberg, 1996, p. I:9); “There are different transformation
processes (different observation, interpretation and coding processes)
between reality, perceptions of reality, and representations of perceptions
of reality” (Lundeberg, 1996, p. I:9-I:10); “This book is based on the idea
of social construction of reality” (Lundeberg, 1993, p. 75); “Reality is primarily a mental phenomenon” (ibid.); “… if you want to perceive reality
as it is, you must perceive other persons’ perceptions of reality” (ibid.).
The above sentences have been selected to illustrate the use of language
and coverage of topics of a whole text, not to test consistency between
these statements as such. As we can see, the ground covered in these presentations of the framework ranges from the concrete, practical, personal,
normative and action-urging to the abstract, complex, impersonal and theoretical. In fact, most presentations of the framework have to date been
geared primarily at a student/practitioner audience, with a secondary
research audience certainly not being neglected. That the language is in
large parts geared towards practitioners partly explains, I believe, some
inconsistencies between statements seemingly conveying epistemological
positions (as indicated above and as discussed by Nissen, 2003): To a cer-
Mähring
215
tain extent, some of the statements quoted above can be understood as
“interventions” or gentle provocations of a reader with practical interests
and purposes.
However, I stated above that the use of language not only signifies the
intended readership and the subtext of statements, it also reflects a fundamental aspect of the framework as currently designed and used: it has
come to encompass a multitude of audiences and purposes, whose interests
and priorities sometimes diverge. To understand this and the consequences
thereof, it becomes useful to assess drifts in purposes and uses over time.
Drifts in Framework Purposes and Uses over Time
A progression of interest by Lundeberg can be traced from information
analysis (Lundeberg & Andersen, 1974)3 and systems development methods (Lundeberg et al., 1978; 1981) to the herein discussed sustained interest in models and frameworks for change. Throughout this progression,
systems thinking (e.g. Checkland, 1981; cf. also the hierarchy of living
systems in Miller, 1978) seems to have been a constant intellectual basis.
The progression from information systems development methodology to
the Lundeberg framework can also be observed, I believe, in the use of
principles from the OSI (open systems interconnection) model,4 which
can be detected in subsequent developments of the ISAC approach (Lundeberg, 1983; Nilsson, 1988; 1991, pp. 48-53) as well as in the basic
structure of the levels of abstraction model (Lundeberg, 1993). The theoretical linkage is to logical levels and logical types, as mentioned above.
More importantly, the ISAC methodology for IS development (Lundeberg et al., 1978; 1981) incorporated business process design and aspects
of organizational change, preceding the widespread attention given to
this topic in the 1990s, starting with Hammer (1990) and Davenport &
Short (1990).5
3
The creation of I-graphs for determining information needs actually relies on a
principle similar to the levels of abstraction: information precedence graphs build
on a principle of alternate sequencing of information units and information processes (Lundeberg and Andersen, 1974, ch. 2).
4
The OSI model can be likened to a logical levels model of communication protocols (see Zimmerman, 1980; Miller, 1981).
5
For discussions of the ISAC methodology and its influence, see Nilsson (1995)
and Iivari & Lyytinen (1998).
216
Exploring Patterns in Information Management
Thus, highly consistent with the evolution from earlier work with information systems methodology and business process design, the initial purpose
of the Lundeberg framework was, in my view, primarily to provide tools
for analyzing situations and processes concerning focused and planned
organizational change, primarily related to information and information
technology (Lundeberg, 1993).
However, even from early on (ibid.), the framework also came to include
elements that focused self-reflection by the “user” of the framework, the
organizational actor(s) employing the models in the framework in organizational settings. As such, the framework is also positioned as (what I call)
“a cognitive toolbox” for IT-related change, with a strong focus on individuals and on processes of social interaction.
While not expressly positioned as such by the originator, models in the
Lundeberg framework were also soon to be used as analysis models in
research within information management (e.g. Mårtensson & Mähring,
1992). Although this may seem to be a purpose closely related to the original, the difference is in fact fundamental and largely ignored: It requires of
the framework as well as the researcher(s) to span the distance between
practicable action and academic research in one stretch. It also, most
likely, leads the researcher to use practical constructs, rather than research
constructs, in attempting to build theory (for a discussion on theory building, see Lee, 2003)6. This observation brings us to the uses of the framework and how these uses have evolved over time.
Examples of Uses
and Avenues for Further Development
The framework has been used in teaching and learning in academic environments for over 15 years. Several hundred majoring students have used
the framework in “real life” projects in organizations, as part of a majoring
course in information management at the Stockholm School of Economics.
Course ratings for this course have been consistently excellent and many
students have testified to the learning effects of using the framework.
Important in this context is that the framework has been learnt through
action-based learning, i.e. problem-based learning taking place in actual
organizational settings and with one of several purposes being to influence
actual organizational practices. The framework has also been used in
6
Lee (2003) uses the terms first-level constructs and second-level constructs.
Mähring
217
executive education and in corporate development programs in similar
ways and with similar results.
In this use of the framework, the focus is on developing personal and interpersonal skills related to analysis, intervention and change in organizations. It also incorporates personal reflection related to these tasks. This
use of the framework thus includes the “methodology use”, i.e. the framework as a collection of analysis models to be applied on documented
observations of circumstances and events in an organization. It also
includes the “cognitive toolbox” use, in which the framework serves as a
collection of structures for sorting and framing impressions of a complex
problem set, thereby focusing and guiding the attention of actors using the
framework. Through this, the use of the framework also guides and
restricts how a problem situation is conceptualized as well as the development of strategies for related actions/interventions.
The use of the framework in this way is facilitated by process-oriented
teaching and by rather intensive support of the learning process over time.
It is quite likely that this intensive support is a major factor behind the successful adoption and use of the framework in these settings. The question
remains how “distanced” learning, adoption and use of the framework
occur, i.e. the adoption and use based only on available documentation and
without the benefit of process-oriented learning methods.
As mentioned above, the Lundeberg framework has also been used in several research studies over time. This was – as I interpret events and writings – not an originally intended area of use, but nonetheless one that now
contributes to perceptions of the framework’s usefulness.
Perhaps the most common use of the framework in research contexts is a
selective use of one or two models as basis for construction of a proprietary framework of analysis. For example, Nilsson (1999) uses the X-model
in combination with the levels model to construct a framework that supports the basic structure of her presentation of research results as well as
the research model. The use of the model to structure the thesis provides
for a pedagogic, stepwise presentation of study results. In Nilsson’s model,
lead times for corporate financial reporting and attitudes towards lead
times (dependant variables) are explained using attributes of the reporting
process as well as attributes of process pre-conditions. Both these types of
attributes (independent variables) are structured in accordance with a levels model separating people, behavior, business results, operations, information, IT and environmental factors. While providing an informative
framework for presentation of the study and its results, the use of the X-
218
Exploring Patterns in Information Management
model in this study may also have limited the identification of causal relationships, especially since chains of X-models were not used. For example,
the X-model in this application does not readily help determine the impact
on actual lead times of executives’ attitudes towards lead times (since
attitudes are classified as a type of outcome), only the impact on lead times
of their interest in and demands on lead times (ibid., ch. 10).
In Mähring (2002), the levels model is related to the use of Strauss & Corbin’s (1990) concept “levels of analysis”. This concept is used to develop a
model of multiple levels of analysis for a case study of project governance.
The model distinguishes between the information system under development, project work, project management, project governance (the primary
level of analysis), corporate IT governance, the organization and the environment. Although inspired by the abstraction levels model, these levels of
analysis basically constitute a systems levels model, where the “degree of
abstraction” is of lesser importance.
Mårtensson (2001) also constructs a framework for analysis based on
models from the Lundeberg framework, in this case a model that classifies
the foci of management processes in terms of levels (person, business,
information) as well as process characteristics (preparing, performing,
evaluating). Like Nilsson (1999), this model is used to structure the presentation and analysis of data. Mårtensson (2001) also uses two levels of
change and learning as part of framing and explaining a key finding,
namely a distinct difference between execution-oriented management processes and development-oriented management processes (ibid., pp. 288291). Clearly an example of abstraction levels, this model is also highly
influenced by Argyris’ & Schöns’ (e.g. 1995) model of single-loop and
double-loop learning.
Westelius (1996) exhibits perhaps the most extensive use of models from
the Lundeberg framework in a research study to date. This study uses the
X-model to build an a priori overview model of a type of change process
(development and implementation of principles of managerial accounting
and control). This a priori model is then used to guide interviews. Furthermore, an X-model is used to depict the research process (ibid., p. 39).
Another model in the framework (a perspectives model not discussed
above) is used to build understanding of how project managers and other
actors utilize perspectives of different stakeholder groups in accounting
change. Further, a levels of abstraction model is used to target how
involved actors focus concrete output, such as measurements and descriptions, while paying less attention to the uses of measurements and descriptions (second level) and effects of the use (third level).
Mähring
219
Although not fully exploited as such, the X-model developed by Westelius
to depict the process of management accounting and control change
(Figure 5) could be used as basis for a process theory for this type of
change process. I will discuss this particular use of the framework further
below.
Idea
Anchoring
Initiation
Project
Formation
Implementation
Adjustment
“Theory”
Study
Continuous
Operation
Investigation
Review
Pilot
Project
Knowledge
Dissemination
Design
Termination
Figure 5. Phases in the Life Cycle of Principles of Management Accounting and
Control (from Westelius, 1996)
Neither of the above examples of research uses of the framework concern
the building of targeted frameworks for specific problem areas. One
attempt to build a substantive or specific framework on the basis of the
general Lundeberg framework is Andersson (2002; 2003). In this adapted
framework three interrelated models depict (1) change of a project process,
(2) change of project pre-conditions and (3) change in project improvement practices. This interrelation of three frameworks, in levels, again
relates to single-loop and double-loop learning (Argyris & Schön, 1995),
or perhaps better Bateson’s (1972; 1979) three-level model of learning.
From the above examples, we can distinguish a difference between using
the Lundeberg framework in the research design (as in all of the above
examples) or as a basis for research results (as e.g. in Westelius, 1996). A
third use of the framework is as a basis for development of specific or substantive frameworks. A problem with this type of research undertaking is
how to assess the research results, since frameworks cannot be readily
tested, e.g. subjected to falsification in the way theories can. Perhaps,
frameworks could be subjected to testing through evaluation of use in a
manner similar to IS development methodology evaluation (cf. Nilsson,
1991).
220
Exploring Patterns in Information Management
It might be that the use of the framework as a tool for theory building has
been less frequent and perhaps harbors the most potential for future
research undertakings. To discuss this, however, I first need to briefly convey what the words “theory”, “model” and “framework” are intended to
mean in this context.
Theory, here, means “a statement of relations among concepts within a set
of boundary assumptions and constraints” (Bacharach, 1989). Adding to
this, a theory would normally consist of a number of statements and the
statements would aspire to some degree of generality, e.g. being valid over
a certain range of specific instances (Sutherland, 1975; Weick, 1989).
The word model can denote as least two different things. It can either denote
a research model, often depicting causal relationships between phenomena,
but sometimes depicting e.g. a temporal/sequential ordering of phenomena
or events (cf. Mohr, 1982; Langley, 1999). A research model depicts key
aspects of a theory but does not by itself constitute a theory (it rarely captures all the relationships between concepts, nor a comprehensive view of
boundaries and constraints). A conceptual model, on the other hand, might
provide a view, a perspective on a certain type of phenomenon. It need not
directly depict a theory but can be related to one or several theories. Its aim
is often to provide guidance for practical action or to convey a worldview.
The Lundeberg framework consists of conceptual models, not research
models and in the above research examples, these models were often used to
code or structure data. (the other frequent use of models in the above examples was as a basis for creating a framework for analysis.)
Finally, I here see a framework as either a high-level model aimed at conveying a worldview or a set of conceptual models with similar purpose.
(Yes, this means that the boundary between higher-level conceptual models and frameworks is somewhat blurred.) For example, the MIT Management in the 90s research program (Scott Morton, 1991) used a framework
consisting of five concepts or entities: people, tasks, structure, technology
and management processes – all interrelated. Based on a model commonly
known as “Leavitt’s diamond” (Leavitt, 1965), this framework can be seen
as a guidance for the research program; part of a worldview (thus possible
part of the definition of boundaries of a theory) but not a theory or model
subjected to (or even possible to subject to) testing.
Unlike a theory or a research model, which can be subjected to scrutiny
e.g. through falsification, a conceptual model or a framework does not
constitute a well-defined knowledge claim, and thus cannot be subjected to
scrutiny, at least not in the same way as a theory and not with the purpose
Mähring
221
of directly advancing knowledge (cf. Popper, 1989/1935). A framework
can, however, be of use in building theory.
The potential use of the Lundeberg framework for the purpose of theory
building is dependent on the understanding of the framework as a metaframework. This is a fundamental difference compared to using the framework in practical contexts within an intended domain, in which case the
models should be applicable without adaptation. Using an analogy to
research on IT-mediated organizational communication (Yates & Orlikowski, 1992; Orlikowski & Yates, 1994), the use of the Lundeberg framework
to develop theory can be seen as akin to the use of the concept of “communication genres”: as a lens for conducting research (e.g. making sense of
observations) rather than as a substantive theory to be tested. In both cases,
the “lens” allows for building substantive theory but does not constitute it.
Using the framework as a meta-framework also has the potential of
addressing a problem related to constructs largely ignored in research uses
of the framework to date, but introduced above: If the models of the Lundeberg framework are used as substantive models, they are likely to
restrict theory building through their use of very general, practice-oriented
concepts (e.g. “results”, “persons”, “environment”). My concern here is
that use of the models in the framework for research purposes, for example
to sort research data or to develop propositions or research questions, is
likely to lead the researcher to framing her/his data as well as the research
questions using constructs that are intended for practical use. If, on the
other hand, the models are used as meta-models, as structures that can
facilitate and guide construction of substantive research models (i.e. models addressing a specific research topic), this enables the researcher to use
theoretical constructs (e.g. “cognitive dissonance”, “self-justification”,
“normalization of deviance”) instead of practical constructs in the construction and use of these substantive conceptual models (cf. Lee, 2003).
Although use of the X-model may (as discussed earlier) present some problems, there is also, as mentioned (cf. the discussion about Westelius, 1996,
above), evidence that it might be useful as a tool for building and presenting
process theory (cf. Mohr, 1982; Markus & Robey, 1988). For example, the
X-model could be linked to the punctuated equilibrium model of organizational change (Tushman & Romanelli, 1985; Gersick, 1991; Van de Ven &
Poole, 1995), which sees change as occurring through sudden periods of
revolutionary change interspersed with longer periods of relative continuity.
Similarly, Newman & Robey (1992) separate information systems development processes into “episodes” (periods of ongoing work) and “encounters”
(short event-like occurrences where e.g. conflicts are enacted). The X-
222
Exploring Patterns in Information Management
model, which could easily be used to depict both types of sequences mentioned above, would seem well suited as a tool for structuring a process theory in this vein. Use of the X-model would also facilitate increased attention
to “interfaces” between phases, e.g. between episodes and events.
While this example provides an instantiation of how theory building may
occur with the help of one conceptual model, the question remains how a
researcher can benefit from using a conceptual framework, such as the
Lundeberg framework, in building theory? The following steps indicate a
possible route of some generality:
• Determine areas where use of the framework is likely to support contradiction of existing theory. This requires a literature review of the
research area in question.
• On the basis of the generic Lundeberg framework, develop a “substantive” analysis framework for the specific problem area and support that
framework through use of reference theories (e.g. communication theory, systems theory, theories on learning) that are consistent with the
specific framework, as well as existing, proprietary theory from the
specific research area.
• Use the reference theories, proprietary theories and the substantive
framework to design a study that attempts to falsify proprietary theory
in the targeted topic area.
• Revise the substantive framework in accordance with findings and propose both the specific findings and the revised framework as research
contributions.
In the use of the framework for research purposes in general, and for theory building in particular, I would especially like to stress the importance
of linking specific instantiations of the framework extensively to theory.
While these are some ideas for developing the research use of the Lundeberg framework, there are surely other possible paths to do so. Herein,
however, the time has come to conclude the discussion.
Concluding Remarks
The above discussion reminds us that intellectual structures, including
models and frameworks, shape our thinking even as we reshape these
structures through applying them. Methods, models and frameworks thus
focus, guide and restrict our vision and attention, thereby having subtle and
Mähring
223
powerful influence over the perception and framing of situations, problems
and solutions (Boland, 1979; Yakura, 1992; Beath & Orlikowski, 1994; cf.
also Giddens, 1984).
Concerning the Lundeberg framework specifically, a key to furthering the
practical use of the framework seems to be increased knowledge of how
the framework is learnt and applied under different learning circumstances,
such as on the basis of written documentation only, or in combination with
traditional, non-intensive teaching methods. This knowledge would provide valuable input to any future work with redesign of the framework.
Another issue raised above concerns how to evaluate a conceptual framework, for example how to determine which of two versions of a framework
is superior. Above I mentioned the possibility of evaluating a framework
similarly to an IS development methodology: by evaluating a number of
projects in which the framework has been applied. This would largely be an
evaluation of usefulness (Alvesson & Sköldberg, 1994), which seems reasonable if practical use is the sole concern. In the same vein, perceived
meaningfulness, i.e. meaningfulness in the eyes of framework users, could
also be investigated. As indicated in the above, 15 years of use of the Lundeberg framework in organizations, as well as at least ten years of use in
empirical research, constitutes rich “data” for further developments of the
framework, but also for some types of evaluations of it.
For research purposes, meaningfulness can be indirectly assessed through
assessment of analyses or interpretations facilitated through use of the
framework. The third criterion for theory evaluation suggested by Alvesson & Sköldberg (ibid.), correspondence (of theory to observations/phenomena) is more difficult and would primarily apply to specific findings in
studies employing the framework (see the discussion above on theory versus frameworks).
Concerning research use of the Lundeberg framework, I have in the above
pointed to several key issues, such as how the understanding of the framework as a meta-framework might be central to its use in theory building,
how substantive frameworks can be evaluated. I have also offered an
exemplification of how specific models in the framework can be applied in
building theory, as well as suggested a possible general route for theory
building using a conceptual framework. As stated above, I view extensive
linking of an instantiated framework to substantive theory as essential.
While the framework has been revised quite a few times over the last decade (see references), there is an avenue for development that has not been
extensively explored and that might offer potential for additional develop-
224
Exploring Patterns in Information Management
ment: It might be valuable to develop versions of the framework, and
accompanying descriptions, that are targeted for specific purposes and
user/reader categories. This would provide opportunities to reduce complexity by reducing the need to “span” theory and practice within one
framework and one text, while simultaneously offering opportunities to
pursue specific (or “narrow”) issues further. A thereby reduced need for
tradeoffs between user/reader categories would likely lead to improved
chances to effectively reach a targeted user group with a specific text. This
approach need not be seen as a reductionist, but could well be viewed as a
stepwise systems approach to further framework development. Such targeted frameworks could include frameworks for IS research undertakings,
for use by change agents and for use in business process analysis and
design work. It could also include a dedicated theoretical discussion of the
framework, including its epistemological and ontological assumptions.
Is it then worthwhile to spend so much effort on the refinement and validation of a conceptual framework? Should we not, in the spirit of the early
21st century, speed up, declare victory, move on? I have, in fact, already
answered this question, when I stated that conceptual frameworks guide
and restrict our vision and attention and that they subtly and powerfully
influence how we perceive, frame, and approach situations, problems and
solutions. The power of abstraction thus depends on subtleness and nuance
– and this is something Mats Lundeberg has an acute understanding of.
References
Alvesson, M. & Sköldberg, K. (1994) Tolkning och reflektion: Vetenskapsfilosofi
och kvalitativ metod {Interpretation and Reflection: Philosophy of Science and
Qualitative Method}, Studentlitteratur, Lund, Sweden.
Andersson, M. (2002) “Developing Professional Services in Project-Intensive
Organizations: An Information Management Perspective”, unpublished dissertation proposal, Stockholm School of Economics, Stockholm.
Andersson, M. (2003) “PRIO: Verksamhetsutveckling i projektintensiva organisationer” {PRIO: Business Development in Project-Intensive Organizations},
SSE/EFI Working Paper Series in Business Administration, No. 2003:03,
Stockholm School of Economics, Stockholm. http://swoba.hhs.se/hastba/abs/
hastba2003_003.htm
Argyris, C. & Schön, D. (1995) Organizational Learning II: Theory, Method and
Practice, Addison-Wesley, Reading, Massachusetts.
Bacharach, S.B. (1989). “Organizational Theories: Some Criteria for Evaluation”,
Academy of Management Review, Vol. 14, No. 4, pp. 496-515.
Mähring
225
Bateson, G. (1972) Steps to an Ecology of Mind, Ballantine, New York.
Bateson, G. (1979) Mind and Nature: A Necessary Unity, Widwood House, London.
Beath, C.M. & Orlikowski, W.J. (1994) “The Contradictory Structure of Systems
Development Methodologies: Deconstructing the IS–User Relationship in
Information Engineering”, Information Systems Research, Vol. 5, No. 4, pp.
350-377.
Boland, R.J. Jr. (1979) “Control, Causality and Information Systems Requirements”, Accounting, Organizations and Society, Vol. 4, No. 4, pp. 259-272.
Checkland, P. (1981) Systems Thinking, Systems Practice, Wiley, Chichester.
Churchman, C.W. (1968) The Systems Approach, Dell, New York.
Davenport, T.H. & Short, J.E. (1990) “The New Industrial Engineering: Information Technology and Business Process Redesign”, Sloan Management Review,
Vol. 31, No. 4, pp. 11-27.
Gersick, C.J.G. (1991) “Revolutionary Change Theories: A Multilevel Exploration
of the Punctuated Equilibrium Paradigm”, Academy of Management Review,
Vol. 16, No. 1, pp. 10-36.
Giddens, A. (1984) The Constitution of Society: Outline of the Theory of Structuration, University of California Press, Berkeley, California.
Hammer, M. (1990) “Reengineering Work: Don’t Automate, Obliterate”, Harvard
Business Review, Vol. 68, No. 4, pp. 104-112.
Hill, P.H. (1981) Making Decisions: A Multidisciplinary Introduction, AddisonWesley, Reading, Massachusetts.
Iivari, J. & Lyytinen, K. (1998) “Research on Information Systems Development
in Scandinavia – Unity in Plurality”, Scandinavian Journal of Information
Systems, Vol. 10, No. 1-2, pp. 135-186.
Langefors, B. (1973/first edition 1966) Theoretical Analysis of Information Systems (Fourth Edition), Studentlitteratur, Lund, Sweden.
Leavitt, H.J. (1965) “Applied Organizational Change In Industry: Structural,
Technological And Humanistic Approaches”, in March, J.G. (Ed.), Handbook
of Organizations, Rand McNally & Company, Chicago.
Lee, A.S. (2003) “Building and Testing Theory on New Organizational Forms
Enabled by Information Technology”, in this volume.
Lundeberg, M. (1983) “Att utveckla verksamhet och system i samklang med varandra” {Developing Business Operations and Systems in Concert}, Institute V
report series, Report V-3301, Institute V, Stockholm.
Lundeberg, M. (1992) “A Framework for Recognizing Patterns When Reshaping
Business Processes”, Journal of Strategic Information Systems, Vol. 1, No. 3,
pp. 116-125.
226
Exploring Patterns in Information Management
Lundeberg, M. (1993) Handling Change Processes – A Systems Approach, Studentlitteratur, Lund, Sweden.
Lundeberg, M. (1995) “A Multilevel Approach to Information Management”, in
Dahlbom, B. (Ed.), “The Infological Equation: Essays in Honour of Börje
Langefors”, Gothenburg Studies in Information Systems, Department of Informatics, School of Economics and Commercial Law, Gothenburg University,
Göteborg, Sweden, Report 6, pp. 251-285.
Lundeberg, M. (1996) “Nivåsynsättet – En första introduktion”, in Lundeberg, M.
& Sundgren, B. (Eds.), Att föra verksamheten framåt – Människor och informationssystem i samverkan, Studentlitteratur, Lund, Sweden. Also published
in English as: “The Multilevel Approach – A First Introduction”, in Lundeberg, M. & Sundgren, B. (Eds.) (1996) Advancing Your Business – People and
Information Systems in Concert, Economic Research Institute (EFI), Stockholm, pp. I:1-I:24. http://www.hhs.se/im/efi/mle.pdf
Lundeberg, M. (2000) “Perspektiv på verksamhetsutveckling och IT” {Perspectives on Business Development and IT}, in Bengtsson, L., Lind, J. & Samuelson, L.A. (Eds.), Styrning av team och processer: Teoretiska perspektiv
och fallstudier {Management of Teams and Processes: Theoretical Perspectives and Case Studies}, Economic Research Institute (EFI), Stockholm, pp.
87-115.
Lundeberg, M. (2003) “Verksamhetsutveckling – en första introduktion av tre
arbetsmodeller {Business Process Development – A First Introduction of Three
Models}, SSE/EFI Working Paper Series in Business Administration, No.
2003:05, Stockholm School of Economics, Stockholm. http://swoba.hhs.se/
hastba/abs/hastba2003_005.htm
Lundeberg, M. & Andersen, E.S. (1974) Systemering – Informationsanalys {Systems Analysis and Design – Information Analysis}, Studentlitteratur, Lund,
Sweden.
Lundeberg, M., Goldkuhl, G. & Nilsson, A. (1978) Systemering {Information Systems Development}, Studentlitteratur, Lund, Sweden.
Lundeberg, M., Goldkuhl, G. & Nilsson, A. (1981) Information Systems Development – A Systematic Approach, Prentice-Hall, Englewood-Cliffs, New Jersey.
Lundeberg, M. & Sundgren, B. (Eds.) (1996) Advancing Your Business – People
and Information Systems in Concert, Economic Research Institute (EFI),
Stockholm.
Miller, J.G. (1978) Living Systems, McGraw-Hill, New York.
Miller, L.J. (1981) “The ISO reference Model of Open Systems Interconnection”,
in Levy, B., Proceedings of the ACM ‘81 Conference, pp. 283-288.
Mähring, M. (2002) IT Project Governance, Economic Research Institute (EFI),
Stockholm.
Mähring
227
Markus, M.L. & Robey, D. (1988) “Information Technology and Organizational
Change: Causal Structure in Theory and Research”, Management Science, Vol.
34, No. 5, pp. 583-598.
Mårtensson, P. (2001) Management Processes: An Information Perspective on
Managerial Work, Economic Research Institute (EFI), Stockholm.
Mårtensson, P. & Mähring, M. (1992) Information Support from Staff to Executives – An Explorative Study, Economic Research Institute (EFI), Stockholm.
Mohr, L.B. (1982) Explaining Organizational Behavior, Jossey-Bass, San Francisco.
Newman, M. & Robey, D. (1992) “A Social Process Model of User–Analyst Relationships”, MIS Quarterly, Vol. 16, No. 2, pp. 249-266.
Nilsson, A.G. (1988) “Information Systems Development – A Frame of Reference
and Classifications”, Institute V report series, Report V-4801, Institute V,
Stockholm.
Nilsson, A.G. (1991) Anskaffning av standardsystem för att utveckla verksamheter
– Utveckling och prövning av SIV-metoden {Acquisition of Application Packages for Developing Business Activities – Development and Validation of the
SIV Method}, Ekonomiska Forskningsinstitutet (EFI), Handelshögskolan i
Stockholm, Stockholm [In Swedish with English summary].
Nilsson, A.G. (1995) “Evolution of Methodologies for Information Systems Work
– A Historical Perspective”, in Dahlbom, B. (Ed.), “The Infological Equation:
Essays in Honour of Börje Langefors”, Gothenburg Studies in Information
Systems, Department of Informatics, School of Economics and Commercial
Law, Gothenburg University, Göteborg, Sweden, Report 6, pp. 251-285.
Nilsson, K. (1999) Ledtider för ledningsinformation {Lead Times for Management
Information}, Economic Research Institute (EFI), Stockholm.
Nilsson, K. (2003) “Patterns in Information Management: A Multi Level Analysis
of Swedish Companies”, in this volume.
Nissen, H.-E. (2003) “Steps to an Ecology of the Multilevel Approach to Information Management”, in this volume.
Orlikowski, W.J. & Yates, J. (1994) “Genre Repertoire: The Structuring of Communicative Practices in Organizations”, Administrative Science Quarterly,
Vol. 39, No. 4, pp. 541-574.
Popper, K. (1989/first edition 1935). Logik der Forschung (Neunte, verbesserte
Auflage) {The Logic of Scientific Discovery (Ninth, Improved Edition)}, J. C.
B. Mohr, Tübingen.
Scott Morton, M.S. (1991) The Corporation of the 1990s: Information Technology
and Organizational Transformation, Oxford University Press, New York.
Shannon, C.E. & Weaver, W. (1949) The Mathematical Theory of Communication, University of Illinois Press, Urbana, Illinois.
228
Exploring Patterns in Information Management
Simon, H.A. (1977) The New Science of Management Decision, Prentice-Hall,
Englewood Cliffs, New Jersey.
Strauss, A. & Corbin, J. (1990) Basics of Qualitative Research: Grounded Theory
Procedures and Techniques, Sage Publications, Newbury Park, California.
Sutherland, J.W. (1975). Systems: Analysis, Administration and Architecture, Van
Nostrand, New York.
Tushman, M.L. & Romanelli, E. (1985) “Organizational Evolution: A Metamorphosis Model of Convergence and Reorientation”, in Staw, B.M. & Cummings, L.L. (Eds.), Research in Organizational Behavior, Vol. 7, pp. 171-222.
Van de Ven, A.H. & Poole, M. S. (1995) “Explaining Development and Change in
Organizations”, Academy of Management Review, Vol. 20, No. 3, pp. 510-540.
Watzlawick, P., Bavelas, J.B. & Jackson, D.D. (1967) Pragmatics of Human Communication, Norton, New York.
Weick, K.E. (1989) “Theory Construction as Disciplined Imagination”, Academy
of Management Review, Vol. 14, No. 4, pp. 516-531.
Westelius, A. (1996) A Study of Patterns of Communication in Management
Accounting and Control Projects, Economic Research Institute (EFI), Stockholm.
Whitehead, A.N. & Russell, B. (1910) Principia Mathematica, Cambridge University Press, Cambridge.
Yakura, E.K. (1992) “Practices of Power: Meaning and Legitimation in Information Technology Consulting”, unpublished doctoral thesis, MIT Sloan School
of Management, Massachusetts Institute of Technology, Cambridge, Massachusetts.
Yates, J. & Orlikowski, W.J. (1992) “Genres of Organizational Communication:
A Structurational Approach to Studying Communication and Media”, Academy of Management Review, Vol. 17, No. 2, pp. 299-326.
Zimmerman, H. (1980) “OSI Reference Model – The ISO Model of Architecture
for Open Systems Interconnection”, IEEE Transactions on Communication,
Vol. 28, No. 4, pp. 425-432.
— 14 —
Steps to an Ecology of the Multilevel
Approach to Information
Management
Hans-Erik Nissen
Background
In his research Mats Lundeberg has focused on how people use information technology in business processes. (Lundeberg, 1995, 1996) One contribution from this research he calls “A Multilevel Approach to Information Management.” I will call this the “ML approach”.
In his presentations of the ML approach Lundeberg (1995, 1996) refers,
among other sources, to Argyris and Schön (1978), Bateson (1972, 1979),
Berger and Luckmann (1966), and Watzlawick et al. (1967). This indicates
his familiarity with other research traditions than the Anglo-Saxon, logical/empirical (LE), one. LE advocates an ideal of arriving at objective truth
of an observer-independent reality. The sources mentioned above all work
within a Continental, hermeneutic/dialectic (HD) tradition and beyond.
This tradition advocates an ideal of a community of observers/investigators
arriving at an intersubjective, coherent truth regarding some domain of
observation/interaction. LE traditions presuppose it is possible strictly to
separate theory from practice. HD traditions perceive theory and practice
as dialectically related. With respect to causation LE researchers look for
linear causal chains. HD researchers look for mutual and longer recursive
chains of determination. Nissen (2002) gives a brief overview of the two
traditions. It builds on an extensive study of their metascience in Radnitzky (1970).
Both traditions comprise a number of schools (Radnitzky, 1970). The most
extreme of these Lakoff and Johnson (1980) have called objectivism and
subjectivism. All of the sources for the ML-approach mentioned above
stand far from subjectivism. The authors behind them even seem open to
go beyond the HD tradition. One emergent alternative intended to overcome the myths of objectivism and subjectivism has been presented briefly
230
Exploring Patterns in Information Management
in Lakoff and Johnson (1980) and more extensively in Lakoff and Johnson
(1999). They call their alternative an experientialist one.
Lundeberg in his presentations, surprisingly, only applies to a limited
extent what researchers from HD traditions write. Lundeberg (1995, 1996)
also sometimes writes in a way that leads readers to perceive him as
applying LE traditions in contradiction to HD traditions. Here I will argue
that the ML approach would gain from better coherence in what it fetches
from different traditions. This I will do by reflecting upon what I call the
ecology of the ML approach. Before I address that issue I will discuss two
other issues. First, I will discuss some problems encountered by researchers, who mix methods originally developed within different traditions.
Second, I will present how I have understood the ML approach by studying Lundeberg (1995, 1996).
Mixing Methods Developed in Different Traditions
Some of the presuppositions of the two research traditions more or less
contradict each other. Some researchers in the social sciences, working in a
HD tradition, hence avoid all methods developed outside of it. For
instance, they avoid entirely using quantitative methods. At the same time
functionalist schools of social sciences, which are the oldest and still in the
majority, follow LE traditions and freely use methods developed within
these.
Research guiding interests should lead the choice of methods of investigation and how to use what is found when applying them (Nissen, 2002, pp.
75-78). According to Radnitzky (1970) research activities can be studied in
broader and more limited contexts. LE traditions have focused on products
of research and how to safeguard their validity. The context of study has
been limited to the context of justifying its products. This precludes criticism coming from outside of the tradition itself. Radnitzky chooses to
study the context of discovery, growth of science, in society. This opens
for critical reflection about a research tradition from outside of itself.
For any research program, research-guiding interests become important.
Radnitzky distinguishes five such interests:
•
•
•
•
•
Technical (work)
Hermeneutic, mediating tradition(s)
Emancipatory
Improving the world picture
Improving reflection on existential themes.
Nissen
231
The first of these often entails an economic interest of those financing a
research program.
A research program might be guided by more than one interest. In some
situations two or more research guiding interests might lead to different
methods as most appropriate. Then it helps if the researchers have decided
beforehand the relative importance of their interests. In some cases available resources for a research program may allow the use of several methods. If so, the very differences of findings might indicate areas of further
inquiry (Bateson, 1979, chapter III).
Methods of investigation and analysis might be chosen freely, provided
they are applied coherently with the research guiding interests of a study.
Choosing basic presuppositions from different traditions, however, is a different story. In that case a researcher should avoid including contradictory
ones for a research program. Not to do so will only lead to inconclusive or
contradictory findings in the end.
In this paper I will follow a hermeneutic research interest. I will try to
understand what I have read about the ML approach. Moreover, I will
relate it to more texts by the authors mentioned above than those referred
to in Lundeberg (1995, 1996).
The ML Approach and its Presuppositions
In this section I will present my understanding of the ML approach and its
presuppositions. Here I will stop by raising some questions. To these I will
return in the section below on an ecology of the ML approach.
Presuppositions of the ML Approach
Lundeberg states the following five presuppositions for the ML approach:
• Persons in business firms are autonomous individuals with responsibility for their own actions
• Business units are means for persons to achieve personal goals and
fulfil personal strategies
• The purpose of using information technology in business units is to
support the business processes of the units
• Information is interpreted from data by using a specific frame of reference
232
Exploring Patterns in Information Management
• Each process in a business unit includes personal behaviour and task
activities in an inseparable whole (Excerpts from Lundeberg, 1995, 8485. Italics in the original.)
The first presupposition Lundeberg (1995) explicates with reference to
1
Berger and Luckmann (1966) . Lundeberg, by only stressing the subjective
reality of individual persons, largely misses what a study of Berger and
Luckmann (1966) could contribute to the ML approach. In order to argue
this conjecture I first illustrate the central thesis from their treatise on “The
social construction of reality”. (See Figure 1.)
Figure 1. Three dialectic moments in constructing social reality.
The text in Figure 1 follows Berger and Luckmann (1966). The dialectic,
however, also applies to parts of societies like business enterprises and
smaller business units. After presenting their fundamental dialectic the
authors write:
It may also already be evident that an analysis of the social world
that leaves out any one of these three moments will be distortive.
One may further add that only with the transmission of the social
world to a new generation (that is, internalization as effectuated in
socialization) does the fundamental social dialectic appear in its
1
The book by Berger and Luckmann originally appeared in 1966. It has later been
reprinted repeatedly. My page references go to a Pelican Book reprinting from
1984.
Nissen
233
totality. To repeat, only with the appearance of a new generation can
one properly speak of a social world. (Ibid., p. 79.)
A new business enterprise has its founders (producers). A new way of
conducting some task in a business unit will have some people who developed it. These producers generally perceive what they have created as a
human product. When mediated to new generations of people, not creating
new inventions, these become objective reality. As such they become
internalized by “new generations”. The internalization relevant in business
contexts is what Berger and Luckmann (1966) call “secondary socialization”. This demands learning from people who move to a new enterprise or
a different social group within an enterprise. In all three moments language
and the distribution of what people know-in-action constitute an important
part (ibid., pp. 157-166).
Explicating his second presupposition, Lundeberg stresses individual
persons behind the design of particular business processes. This focuses
attention on the history of an enterprise, an aspect stressed within HD
traditions. He also discusses different stakeholders as involved in the
design. He does not explicitly mention conflicting interests between
these. Timing for reconsidering the design he makes contingent upon that
the original designers are gone. This at least indicates some preference
for consensus over openly expressed conflict. To what extent is such a
position coherent with what Bateson (1979, pp. 77-100) writes on
becoming informed by differences? To what extent is it coherent with
ideas about double loop learning? See Argyris and Schön (1978) and
Argyris (1990).
About his third presupposition Lundeberg states that using information
technology is not an end in itself. However, he does not state explicitly
whose intentions he expresses. Many people and enterprises supplying
information technology promulgate it as an end in itself. This may look
like taking a very short-sighted position. Still, this myopia has lasted for
about fifty years now. A lasting uneven distribution of power and knowhow between suppliers and users offers one explanation (Nissen, 2002).
This presupposition, too, seems to avoid addressing issues of conflict.
His fourth presupposition Lundeberg grounds in Langefors´ infological
equation. By this he introduces an important distinction between information and data. This raises some questions. To what extent does Lundeberg
(1995, 1996) always uphold this distinction? As Langefors (1993, p. 151)
states, his infological equation “... appears to imply that it is impossible to
convey information, or knowledge, by the communication of data”. Lange-
234
Exploring Patterns in Information Management
fors then suggests a way out of this apparent dilemma. What does Langefors’ resolution of this dilemma entail? Does his resolution hang together
with other sources like Berger and Luckmann (1966)?
By his fifth presupposition Lundeberg (1995) distinguishes between
2
ongoing inseparable processes and partial descriptions of these . In this
context he also proposes a general model for describing “all processes in
business firms” (ibid., p. 86). Below I will discuss some implications of
this claim.
The Aims of the ML Approach
One aim of the ML approach is to reflect on uses of information technology in a broader and hence more meaningful context. By this the approach
“... deals with an important success factor in developing business processes
and using information technology: to consider technology in a meaningful
context.” (Lundeberg, 1996, I:2). Upholding a distinction between information systems and data processing systems might better fulfil this aim.
What does the context of business processes offered hide?
To guard against meaninglessness Lundeberg (1995, 1996) introduces and
advises to uphold logical levels of abstraction. This I perceive as another
aim. He supports his advice with references to Bateson (1979),
Watzlawick et al. (1967), and Whitehead and Russell (1910) Principia
Mathematica (PM). My reading of the former authors indicates that their
positions not are coherent with those of PM. They stress sticking to a clear
logical accounting, too. However, they take a different position when it
comes to handling paradoxes than the authors of PM. (See, e.g.,
Watzlawick et al. on paradoxes in human communication as opposed to
contradictions in data bases or computer programs, pp. 187-229.)
The ML approach also aims at bringing autonomous persons and their individual perceptions into focus. The step from individual perceptions and
actions seems only explained by Langefors´ (1993, p. 152) hypothesis that
there exists some intersubjectively shared “fact knowledge.” Taken for an
“as if hypothesis”, observers in many cases might find empirical support for
it. This, however, does not help to assess its scope of application. Moreover,
the very idea of fact knowledge indicates roots within the LE tradition and
2
Its wording seems strange to me. It separates “personal behavior” and “task
activities” from interacting people of flesh and blood. This seems to contradict the
main message of the fifth presupposition.
Nissen
235
3
particularly in the Vienna school of unified science . An alternative could be
to resolve Langefors´ dilemma by ideas from Berger and Luckmann (1966).
That might avoid assumptions which contradict each other.
Some Features of the Meta-model for the ML Approach
4
Lundeberg (1996) presents the ML approach as a meta-model of reality
(ibid., I:15). He also writes (ibid., I:9) that it “... builds on a presupposition
about a difference between reality, perceptions of reality, and representations of reality.” When Lundeberg (1995, pp. 86-98) builds the ML
approach in steps he stresses the importance of distinguishing between different logical levels. In his discussions of these he refers to Bateson (1972,
1979) and to Whitehead and Russell (1910). When Lundeberg (1995, p.
86) refers to Bateson (1972) he writes “...Logical levels…are about the
relationships between larger and smaller contexts, between classes and
their members.” Bateson generally writes about logical types and logical
typing. His position to logical typing in practice differs from Russell’s theory of types. Much of the argument in Lundeberg (1995, pp. 86-87) seems
to focus on Russell’s theory.
For a critical discussion of the meta-model for the ML approach I need to
introduce some ideas from Bateson (1979). First, however, let me indicate
in what sense Bateson’s view on logical typing goes beyond that of Russell. In a section on “Logic is a poor model of cause and effect” Bateson
(1979) writes:
When the sequences of cause and effect become circular (or more
complex than circular), then the description or mapping of those
sequences onto timeless logic becomes self-contradictory. Paradoxes are generated that pure logic cannot tolerate. ...” (Ibid., p. 70.)
His position on logical typing Bateson (1979) argues in a chapter called
“From classification to process.” Lundeberg (1995, p. 87) briefly refers to
this chapter. In order to be able to discuss how he uses Bateson’s ideas I
3
A group of scientists and mathematicians who, starting in the 1920s and under
the leadership of Moritz Schlick, met regularly to discuss the foundations of
mathematics and natural science. According to them, scientific knowledge is the
only kind of factual knowledge and all traditional metaphysical doctrines are to be
rejected as meaningless. (Also cf. Radnitzky, 1970, Vol. I, 22-24.)
4
Readers not familiar with the ML approach should acquaint themselves with it
by reading Lundeberg (1995, 1996).
236
Exploring Patterns in Information Management
have drawn Figure 2a and b. These I have adapted from figures in Bateson
(1979) on pages 209 and 213 respectively.
Figure 2a) logical levels of inquiry and theorizing, b) logical levels of control.
Figure 2a shows Bateson’s position to what I have called “Logical levels
of inquiry and theorizing.” Bateson uses for illustration his own research
on people in New Guinea. The fieldwork produces descriptions of processes. Reflecting on these results in classifications and other forms of
explanation. I have chosen to call work on explaining “theorizing.” Bateson started his fieldwork from preknowledge of classifications by other
researchers of individual people according to their characteristics.
Moving up from one level of process to the next a researcher also broadens
his domain of observation. In Figure 2a, the unit of investigation on the
lowest level was an individual person. On the next level Bateson’s unit of
investigation was an interacting couple of a woman and a man. On the
highest level he investigated classes of couples studied over a longer time
period. This enabled him to observe changes in patterns of interaction.
Nissen
237
Similarly the classifications in the left column of Figure 2a correspond to
different logical levels. Lundeberg (1995, pp. 97-98 in Figures 13 and 14)
presents logical levels of input, process, and output of a business unit. His
process means a process of largely material transformations, not one of
stepwise inquiry. His classification of input and output exhibits two heterogeneous sets of classes. Each class is based on observations in a different domain. According to Bateson (1979) this makes them belong to different logical types. However, it does not entitle them to classifications on
different logical levels. That presupposes the different domains of observation always will include the domains of observation on a lower level.
Bateson (1972, pp. 177-193) discusses different levels of abstraction in
communication. However, the different classes of input and output do not
both relate to each other as communication on an issue, and as communication on a meta-level about that issue.
Figure 2b illustrates logical typing in the context of control. Control and
coordination constitute important issues in the context of managing business activities. The figure only illustrates the issue of control. In any business there exist many activities to be controlled. Moreover, the activities in
many different business units need to be coordinated.
Figure 2b illustrates control of room temperature under the external load of
changing weather conditions. On the lowest level the system observed
comprises a room in which the temperature will be controlled. It also comprises a heating device and a thermostat. To simplify the illustration the
thermostat is supposed only to shut the heating device on and off. The
external load could exceed what can be compensated for when the heating
device is on all the time. Moreover, the external source of energy might
fail. If so, room temperature would fall outside its normal interval. This
type of control is called control by negative feedback.
The system thus delimited, when designed and implemented, is an electro/mechanical system. No interaction of any person is involved. Even on
this level the control system has constraints. Outside of these it cannot
function properly. A control system user should ask for these constraints.
Only by knowing them can he be aware of the limited protection it can
offer. Oscillations in these and similar mechanical control systems after a
disturbance might converge. However, sometimes they might gain in
amplitude until the system collapses. Bateson (1979, pp. 118-119) discusses this by means of early work on designing governors for steam
engines. For understanding dynamic, concrete systems the equations of
relationships between adjacent parts of the system are not enough. These
relations when embodied in machines work with time constants (delays)
238
Exploring Patterns in Information Management
not determined by the equations. This results in emergent properties within
the whole machine. The ML approach does not alert its users to the importance of time delays in the coupling of business processes.
Now let the domain of observation broaden to also include the person living in the room. She might perceive the room temperature as agreeable or
as too cold or too hot. In the first case she does not do anything about it. In
the other cases she might, provided she understands how to calibrate the
thermostat, change its calibration. This she does according to what she currently perceives as an agreeable room temperature. This Bateson (1979)
distinguishes as a type of feedback control on a higher level.
Broaden the observation over time of the person living in the room or over
a number of people. Then the concept of a personal threshold for the interval of an agreeable temperature emerges. The current threshold for a person’s limits of an agreeable room temperature Bateson looks upon as a
calibration on a higher level of a logical type. This makes him ask, “How
can a personal threshold become controlled?”
Bateson then introduces another level of feedback by observing a person
over a long period. Then there emerges a concept of a distribution of personal thresholds of agreeable temperature. This concept, however, can also
be applied synchronously, for instance, to the population in an office. In
that case an observer goes from a single person and her threshold to the
distribution of thresholds in a class of people. This means going from one
logical type to another. Even going from observing a person briefly to
observing her over longer time means a shift in logical typing.
A single person’s threshold of an agreeable temperature can be calibrated by
genetics or training. At least training can be controlled by social status.
Bateson’s illustration stops at this level of calibration. However, answering
the question, “What calibrates the social status of a person?” would broaden
the domain observed still more. Bateson (1979, p. 214) briefly indicates
changes in social status of a person resulting in a changed threshold for room
temperature. A person might, e.g., become a monk or a soldier.
Bateson concludes from his figure corresponding to Figure 2b (ibid., Figure 11, p. 213):
In other words, the feedbacks and the calibrations alternate in a hierarchical sequence. Note that with each completed alternation (from
calibration to calibration or from feedback to feedback), the sphere
of relevance that we are analysing has increased. ...
Nissen
239
... there is a change in logical typing of the information collected by
the sense organ at each level. (Ibid., pp. 214-215.)
The stepwise presentation in Lundeberg (1995, 1996) of the ML approach
certainly comprises a number of hierarchical levels. However, from them
and the accompanying text I feel unable to understand how he applies
Bateson’s logical typing. When does Lundeberg connect to Bateson’s
ideas of inquiry and theorizing, and when to his ideas on control?
On control hierarchies Bateson (1979) writes something worth reflecting
on in the context of business management. After discussing different levels
of calibration and control, when a policeman stops someone who drives
too fast, he writes:
Notice that within the system of police and law enforcement, and
indeed in all hierarchies, it is most undesirable to have direct contact
between levels that are non-consecutive. It is not good for the total
organization to have a pipeline of communication between the
driver of the automobile and the state police chief...The effect of any
such jumping of levels, upward or downward, is that information
appropriate as the basis for decision at one level will be used as
basis for decision at some other level, a common variety of error in
logical typing. (Ibid., p. 215.)
Bateson (1979, pp. 216-217) also argues that contexts of calibration and
control comprised of people also function as contexts of learning. Lundeberg (1996, I:17) only mentions that Bateson (1979) also discusses logical
typing in learning. One type of learning only demands that the learner goes
round the cybernetic circuit a number of separate times. As an illustration
he uses a marksman learning to shoot a moving target with a rifle. This he
contrasts to a man learning to shoot with a shotgun hidden under a table.
The latter “... must accumulate his skill, packing his successive experiences, like Chinese boxes, each within the context of information derived
from all previous relevant experiences” (ibid. p. 217). Bateson (1979) then
summarizes his position on logical typing:
From this paradigm, it appears that the idea of ‘logical typing’, when
transplanted from the abstract realms inhabited by mathematicological
philosophers to the hurly-burly of organisms, takes on a very different
appearance. Instead of a hierarchy of classes, we face a hierarchy of
orders of recursiveness. (Ibid., p. 217, italics in original.)
Hierarchically ordered recursive loops abound in adaptive organizations as
well.
240
Exploring Patterns in Information Management
The meta-model of the ML approach puts uses of information technology
in a broad and meaningful business context. It goes a long way compared
with a number of other methodologies in developing and assessing information systems. Still, I have raised some questions in this section. Starting
from these I will in the next section, suggest a supplementary generative
model. With it I intend to broaden the context still further. I will also suggest a different way to describe business activities, their control, and
development.
An Ecology of the ML Approach
Biologists coined the term “ecology” for dealing with mutual relations
between organisms and their environment. However, there exist lots of
mutual relations also between an organization, or some part of it, and its
environment. Bateson (1972, pp. 460, 483, 488-493) also uses the term in
connection with cities, ideas, computer programs, and technology. Here I
employ the term in this broad sense.
Today many organizations, such as business enterprises, government
agencies, and non-profit organizations, use information technology. Models of organizations in context hence will entail use of information technology. Stafford Beer’s Viable Systems Model (VSM) stands out as one
such model, which is coherent with Bateson (1979). Particularly, it focuses
attention on multiple recursions and on distinct levels of control. In Figure
3 I present a version of VSM adapted from Leonard (1994).
In some respects I have simplified the model as presented by Leonard. In
one respect I have, however, introduced a radical change. In Leonard
(1994, p. 354) she lets “Total environment” comprise only “Future environment” and “Present environment”. In contrast with this I let the total
environment comprise the organization studied, too. To include the organization studied in the total environment is coherent with second order
cybernetics. In that development of cybernetics observers include themselves in the domain of observation. Good reasons for my way of envisaging the total environment can be found in Varela (1984).
Broadening the concept of ‘total environment’ radically breaks with common sense as currently mediated. The mutual dependence of an organization and its environment has to be kept in focus. If so, an environmentorganization distinction might cause no trouble in many cases. For this
reason Figure 3 shows the future and present environments as distinct from
the organization. To remind the reader of this distinction as a human con-
Nissen
241
struct I have characterized them as perceived. Sometimes a perceived environment is encountered as imposing troublesome restrictions on an organization. Then, rejecting the "external" environment, given by company traditions, might open new opportunities. (Cf. Argyris and Schön, 1978, and
Argyris, 1990, on double loop learning.)
Figure 3. Beer’s Viable Systems Model. Adapted from Leonard (1994, p. 354).
Figure 3 shows many recursive loops at different levels. The organization
shown can, in the final analysis, not be separated from the total environment. Still, when I comment on it, I will write about what is “internal” and
“external” to the organization. In the case of a business enterprise, “OP A”
242
Exploring Patterns in Information Management
stands for a business activity. Such activities are recursively coupled to the
present environment by flows of persons, material, and data5. They are
6
also coupled to their respective management functions , albeit only by
flows of data. People in operations or management have to interpret these
data. In cases of IT-artefacts handling or producing such data, the interpretation has been performed beforehand by people developing the artefacts7.
An operation, its coupling to some relevant part of the present environment, and to its management function constitute the first level of recursion.
Seen from a higher level what has to be controlled is a number of operations working concurrently. Beer calls this system of operations and their
management “system 1”. However, these operations generally are, to some
extent, interdependent. This calls for a special control function, which I
called “coordination” above. This coordinating function Beer calls “system
2”. It is needed to ensure that the different elements of system 1 act in
harmony. This means that it prevents uncontrolled oscillations between
various operations in system 1. Still, the different operating units within
system 1 and their management are in the model perceived as largely
autonomous. They have their own relations with the “external” environment.
The control function called system 3 is ultimately responsible for the
“internal” stability of the organization. Many data flows connect it downward in Figure 3. By this I intend to indicate that system 3 is predominantly alert to what is shown below it in Figure 3. The more limited
exchange of data with higher policy levels prevents these from becoming
flooded with data. This system has to take up management tasks when
decisions have to be made from perspectives of two or more operating
units. Two-way communication about resource bargaining can furnish an
example.
5
In Figure 3 all transmission of what often is called “messages” has been rendered
as data. This I have done in order to focus on the fundamental importance of
interpretation before data can inform directly or mediated via a data processing
program. The commonly used conduit metaphor hides this fact. It makes people
wrongly believe information and knowledge can be sent like packages over
technical links. Lakoff and Johnson, (1980, pp. 206, 231) argue why the conduit
metaphor misleads.
6
A management function will exist even in the case an operation is performed by
a self-steering group.
7
What is said for interpretation of data as this level is valid for such interpretation
at higher levels, too.
Nissen
243
System 4 can be described as the intelligence system of an organization. It
focuses on the future. This means, e.g., searching, screening, and analysing
data from the environment, and constructing scenarios of alternative probable future environments. Such functions as market research and R&D
belong to the realm of system 4.
Finally, system 5 fulfils the task of maintaining the identity of the whole
organization. It aims at balancing its present and future driven efforts.
System 5 has overall responsibility for policy. This will often call for balancing conflicting “internal” and “external” demands.
My brief account of the generative VSM may leave the impression that it
represents a hierarchical model placing the board of directors at the top
and rank and file workers at the bottom. This would grossly misjudge
Beer’s intention with the model. Let me support this by quoting what
Leonard (1994) writes about system 5: “... System Five, which is assuredly not limited to upper echelon staff, is where the ‘corporate culture’
and its vision and values are embodied. It is a System Five role to be
reflective and to keep checking out the answers to perennial questions
such as, ‘what counts as success?’” (ibid., p. 353). Organizational culture
is embodied in all its stakeholders. These again are embedded in cultures
and subcultures of the surrounding society. Different groups of
stakeholders can be expected to hold conflicting sets of values and different power. For an organization to remain viable critical reflection has
to acknowledge this.
Calling the model in Figure 3 a generative one means the following. For
any particular application a set of specific VSMs has to be developed. How
this could be done is briefly indicated in Leonard (1994, pp. 349-354). A
good way to start can often be first to look at the exchange between the
organization and its “external” customers, and how it relates to the area of
concern. Starting inquiry at the system 1 level corresponds in a sense to
going up logical levels of inquiry and theorizing similar to what is shown
in Figure 2a. Such investigations will result in several VSMs. Parts of
these will show what also would be found by an existing ML approach.
The latter, along with constructing a VSM, induces the investigator to ask
questions. I presume that some of these will be very similar, if not the
same.
However, a VSM keeps recursive loops, variations in these, and time constants in focus. In contrast to this Figure 14 in Lundeberg (1995, 98) only
presents parts of these punctuated into a number of linear causal chains.
This and Bateson’s (1979) stress on the importance of recurrent loops are
244
Exploring Patterns in Information Management
my basis for the following conjecture. VSM descriptions have a potential
to focus on important contexts of organizational activities not easily disclosed by the current ML approach. If this conjecture can be corroborated,
what I have sketched in this paper can be perceived as some steps to an
ecology of the ML approach.
Summary and Conclusions
The existing models of the ML approach seem based on the well-known
idea of hierarchical input output models. (Cf. Lundeberg, 1995, Figures 3,
13, and 14.) To these it contributes the presupposition that a “real business
process” should be perceived as an inseparable whole. It also stresses the
inclusion of business processes in analysis and design of potential uses of
information technology. This too means an important contribution to
methodologies for information systems development. The ML approach
brings in contexts necessary to consider for success in developing such
systems.
This paper suggests broadening the context by taking some steps to an
ecology of the ML approach. This I have done by more or less briefly
8
introducing the following seven steps .
The two first presuppositions of the ML approach heavily focus on individual persons. In contrast to these organizational processes generally
involve several people. Moreover, the last presupposition stresses that
elements of an organizational process should be perceived as an inseparable whole. Ideas suggested by Berger and Luckmann (1966) could bridge
the apparent gap. By Figure 1 and my comments to it I have sketched the
direction of this first step.
Second, the existing ML approach seems to play down conflicts. This is
appropriate if limited to avoiding contradictions in databases and programs, building on first order predicate logic. In developing organizational
processes and organizational learning conflicting, views should openly be
acknowledged and handled. Bateson (1979) repeatedly argues that “two
descriptions are better than one” (ibid., pp. 77-100). For organizational
learning, suppressing open expressions of conflicting views is counterproductive. Many arguments and cases supporting this statement can be found
8
The steps might be seen as steps “outward”. The order follows the order of the
previous text. Hence their order should not be taken as an order to follow in any
particular situation.
Nissen
245
in Argyris and Schön (1978) and Argyris (1990). I offer openly expressing
and handling conflicting views as a second step to a broader context.
The ML approach presupposes organizations trying to make good use of
information technology put on the market. However, these organizations
have to do so in an environment of powerful suppliers with (partially)
other goals. The existing ML approach does not alert its users to this fact.
In Figure 3 information technology is only implicit by the fact that today
data processing and transmission often gets a lot of technical support. No
presupposition is made as to the purpose of this support. This I count as a
third step in broadening the context.
The ML approach explicitly stresses the need to distinguish data and
information in coherence with Langefors (1993). However, in Lundeberg
(1995, 1996) the author does not seem to follow this advice consistently.
Langefors’ resolution to avoid the trap of solipsism induces descriptions of
“reality” in a fact language of LE close to the ideas of unified science of
the members of the Vienna circle. Here, this paper offers the option to
resolve Langefors’ dilemma by building on Berger and Luckmann (1966).
Figure 1 and the comments to it indicate how this fourth step of broadening the context could avoid invoking contradictory presuppositions from
LE and HD.
Lundeberg (1995, p. 86, Figure 3) presupposes that “all processes in business firms can be described by this model.” His figure shows a (hierarchical) input-output model. His models, illustrating the existing ML approach,
apply this pattern. Of course a researcher may decide only to produce hierarchical input–output models for describing business processes. However,
I doubt the fruitfulness of always sticking to this kind of model. As a fifth
step to broaden the context I suggest other models and metaphors also
being applied. As an example I have sketched Beer’s VSM model. This
model also indicates some features hidden by hierarchical input-output
models. They hide recursive loops. They also hide, or at least do not put
any focus on, time constants important to avoid dangerous oscillations in
dynamic systems.
The ML approach repeatedly stresses distinguishing levels of logical typing. Only a hierarchy of members of a class, the class and of classes of
classes, etc. seems implied. Bateson’s (1972, 1979) distinction between
logical types generated by shifting the focus of observation is briefly
quoted (Lundeberg, 1995, p. 86) but then ignored. Bateson’s (1972, pp.
177-193), arguing about different abstract levels of communication and
meta-communication, presents these as another case of logical typing.
246
Exploring Patterns in Information Management
Finally, Bateson (1979, pp. 216-217) argues for a distinction between two
logically different types of learning. He subsumes all his examples of logical typing under one heading. This he does as part of his efforts to find
“patterns that connect.” He does not state them as the same phenomena.
Explicitly to recognize all three of Bateson’s examples of logical typing I
offer as a sixth step to broaden the context of the ML approach.
In presenting Beer’s VSM I broadened the concept of the total environment to include the organization studied, too. This I did to alert model
users that our common “internal” – “external” distinction might become
misleading. This I offer as a final and radical seventh step of broadening
the context of the ML approach.
This paper has offered and argued for a number of steps to broaden the
context of applying the ML approach. A number of supplementary perspectives have been offered viewing the context of organizations using and
supplying information technology. This I have done in the spirit of
Bateson’s (1979) argument that two descriptions are better than one.
References
Argyris, C. (1990) Overcoming Organizational Defenses: Facilitating Organizational Learning, Allyn and Bacon, Boston.
Argyris, C. & Schön, D. (1978) Organizational Learning: A Theory of Action Perspective, Addison-Wesley, Reading, Massachusetts.
Bateson, G. (1972) Steps to an Ecology of Mind, Ballantines Books, New York.
Bateson, G. (1979) Mind and Nature: A Necessary Unity, Wildwood House, London.
Beer, S. (1994) Beyond Dispute: The Invention of Team Syntegrity, John Wiley &
Sons, Chichester.
Berger, P. & T. Luckmann (1966) The Social Construction of Reality: A Treatise
in the Sociology of Knowledge, Doubleday Anchor Books, New York.
Dahlbom, B. (Ed.) (1995) The Infological Equation: Essays in Honour of Börje
Langefors, Department of Informatics, School of Economics and Commercial
Law, Göteborg, Sweden.
Dittrich, Y., Floyd, C. & Klischewski, R. (Eds.) (2002) Social Thinking – Software
Practice, The MIT Press, Cambridge, Massachusetts.
Lakoff, G. & Johnson, M. (1980) Metaphors We Live By, The University of Chicago Press, Chicago.
Lakoff, G. & Johnson, M. (1999) Philosophy in the Flesh: The Embodied Mind
and its Challenge to Western Thought, Basic Books, New York.
Nissen
247
Langefors, B. (1993) Essays on Infology, Gothenburg Studies in Information Systems Report 5, University of Gothenburg, Göteborg, Sweden.
Leonard, A. (1994) “The Very Model of a Modern System-General: How the
Viable System Model Actually Works”, in Beer, S. Beyond Dispute: The
Invention of Team Syntegrity, pp. 346-356, John Wiley & Sons, Chichester.
Lundeberg, M. (1995) “A Multilevel Approach to Information Management”, in
Dahlbom, B. (Ed.) The Infological Equation: Essays in Honour of Börje
Langefors, pp. 81-102, Department of Informatics, School of Economics and
Commercial Law, Göteborg, Sweden.
Lundeberg, M. (1996) “The Multilevel Approach – A First Introduction”, in Lundeberg, M. & Sundgren, B. (Eds.) Advancing your business: People and
Information Systems in Concert. EFI, Stockholm.
Lundeberg, M. & Sundgren, B. (Eds.) (1996). Advancing your business: People
and Information Systems in Concert, EFI, Stockholm.
Nissen, H.-E. (2002) “Challenging Traditions of Inquiry in Software Practice”, in
Dittrich et al. (Eds.) Social Thinking – Software Practice, pp. 69-89, The MIT
Press, Cambridge, Massachusetts.
Radnitzky, G. (1970) Contemporary Schools of Metascience: Anglo-Saxon
schools of metascience (Vol. I) and Continental schools of metascience (Vol.
II), Akademiförlaget, Göteborg. (Second revised edition comprising Volume I
and II in one book).
Varela, F. (1984) “The Creative Circle: Sketches on the Natural History of Circularity”, pp. 309-323, in Watzlawick (Ed.) The Invented Reality: How Do We
Know What We Believe We Know? Contributions to Constructivism, W.W.
Norton, New York.
Watzlawick, P., Beavin, J.H. & Jackson, D.D. (1967) Pragmatics of Human
Communication: A Study of Interactional Patterns, Pathologies and Paradoxes, W.W. Norton, New York.
Watzlawick, P. (Ed.) (1984) The Invented Reality: How Do We Know What We
Believe We Know? Contributions to Constructivism, W.W. Norton, New York.
Whitehead, A.N., & Russell, B. (1910) Principia Mathematica, Cambridge University Press, Cambridge.
248
Exploring Patterns in Information Management
Blanksida
— 15 —
Information Management:
Defining Tasks and
Structuring Relationships
Dietrich Seibt
Introduction
The appearance of “Information Management” (abbreviated “IM”) as a
management concept for planning, design, development, implementation,
controlling, etc. of “Information Systems” (abbrev. “IS”) can be seen as a
reaction to, or a consequence of, new “Information and Communication
Technologies” (abbrev. “ICT”), which arrived approximately at the end
of the Seventies (Synnott, 1981; Synnott & Gruber, 1981; Szyperski,
1981, Horton & Marchand, 1982). Interestingly enough the notion “IM”
has not found too much acceptance and usage since the beginning of the
Eighties. Other similar notions as “Information Resources Management,”
“Management of Information and Communication Technology and
Information Systems,” or “Information Systems Management” are also
quite common. Some writers concentrate on the strategic aspects of IM
only.
Most writers agree that IM has several dimensions. First it has a TASK
dimension: IM itself includes a large number of management tasks,
which have to be fulfilled to solve different kinds of problems associated with the usage of ICT to support organisations – especially
firms/companies – to achieve their goals and to do their business efficiently and effectively. On the one hand IM has to decide, which tasks
of the organisation should be supported by ICT and by the development
of IS. Secondly IM has a TECHNOLOGY dimension. In many companies IM is responsible for a huge number of technical questions. These
must be answered to ensure successful development, operating, administration and maintenance of the technical parts of all IS which are running within a company. On the other hand IM uses ICT to a large extent
to fulfil its own tasks. Thirdly IM has an ORGANISATION dimension.
IM has to decide on structures and processes of the organisation, which
250
Exploring Patterns in Information Management
fit best with the technological solutions for the benefit of the company:
i.e. for the sake of achieving company goals. In most organisations IM
acts as an organisational unit – as an institution – which is responsible to
fulfil the above mentioned IM-tasks. Fourthly IM has a PERSON
dimension. Not only organisation structures and processes, but also
individual persons, groups etc. must fit within ICT-solutions. Some
writers integrate “organisation” and “person” aspects under the dimension “organisation.”
“Information Systems” are complex phenomena. They will be successful
for their owners, if they are designed, implemented and managed as manmachine-systems – as socio-technical systems – consciously considering
their specific organisational contexts. Each organisation has many of such
IS in different stages of their “lives.” IM has to make sure that the four
above mentioned dimensions fit together for each IS during its whole lifetime (see Leavitt, 1965).
Lundeberg’s Frameworks for Perceiving
Business Reality for the Purpose of Information
Management and Business Reshaping
In his book “Handling Change Processes – A Systems Approach,” Mats
Lundeberg states: “Reshaping business processes is an area of current
interest in combination with information management. Persons working in
the area of information management and business reshaping often deal
with change processes on various levels without explicitly differentiating
between these levels” (see Lundeberg, 1993, pp. 225f). For his perception
of “information management” he cites the book “The Corporation of the
1990s: Information Technology and Organisational Transformation,”
edited by Scott Morton (see Morton, 1991).
Lundeberg presents a framework of “nine general levels in business” (see
Lundeberg, 1993, p. 224). From an information management und business reshaping perspective he specializes the last three levels (see Figure
1). This modified framework can be considered a specialisation of his
general framework with “seven typical levels in business,” which he uses
as a tool for “perceiving reality” (see Figure 1). Lundeberg emphasizes
that there are no universally true levels of abstraction. The levels chosen
are related to the situation one is focusing upon (Lundeberg, 1993,
p. 223).
Seibt
Nine General
Levels
In Business
Nine
Modified
Levels
Seven Typical
Levels
Persons
Persons
Behavior
Behavior
Behavior
Business Ideas
Business Ideas
Busines
Unit Evaluation
Business
Strategy
Studies
Business
Strategy
Fulfillment
Follow
Up
Studies
Change
Studies
V-Model for
Reshaping
Strategies
Strategies
Goals
Goals
Activities
Activities
Activities
Preconditions
Information
Information
Information Syst.
251
in Business
Persons
Support Functions
Busines
Unit
Review
Business
Process
Design
Results
Business
Processes
Follow
Up
Studies
Change
Studies
Information Syst.
ImpleMentation
Activity
Studies
Information
Studies
Realization
V-Model for
Developing
System
Design
Environment
Environment
Environment
Information
Systems
Figure 1. Lundeberg´s Framework for Reshaping Business Processes
and Developing Information Systems (see Lundeberg, 1993)
To be more specific about his levels of business Lundeberg differentiates
between three kinds of outcomes of change processes related to information management and business reshaping (Lundeberg, 1993, p. 226):
• Business Ideas
• Goals
• Information
Related to these three kinds of outcomes he emphasizes three levels of
abstraction:
• Strategies
• Activities
• Information Systems
Corresponding to the three levels of abstraction he differentiates three
types of change processes, related to information management and business reshaping (Lundeberg, 1993, p. 226):
• Redesigning business strategies
• Reshaping business processes
• Developing information systems
252
Exploring Patterns in Information Management
In parallel Lundeberg models the processes of change. He uses a model for
Reshaping Business Processes and a separate model for Developing
Information Systems.
The process of reshaping business processes is structured by a V-Model,
which uses seven phases (see Figure 1):
•
•
•
•
•
•
•
Business unit review
Business strategy studies
Change studies
Business process design
Follow up studies
Business strategy fulfillment
Business unit evaluation
Lundeberg uses another seven-phases-V-Model to structure the process of
information systems development (Lundeberg, 1993, pp. 227–31):
•
•
•
•
•
•
•
Change studies
Activity studies
Information studies
System design
Realisation
Implementation
Follow-up studies
This is only a selection of examples of Lundeberg’s “Change Process and
Systems Thinking.” He uses his frameworks to recognize patterns, which help
him handle change processes in business reality. The attraction of his frameworks mainly stems from the high degree of abstraction and from the flexibility by which these frameworks are adaptable to individual business situations.
Other advantages are the compactness, determination and inclusiveness of his
approach, and the consistency by which he puts the change processes into the
centre of his approach. His approach is challenging because it shows the long
way you have to go from strategy-motivated change studies to information
system realisation and implementation. From an IM point of view it also
shows the multi-dimensionality of tasks, which have to be fulfilled by IM.
Heinrich: Information Management
– Planning, Administration
In the centre of Heinrich’s IM approach are two concepts which he introduces as his specialities (see Heinrich, 2002, p. 8):
Seibt
253
• “Information-Function”
• “Information-Infrastructure.”
Each company has its individual “Information-Function,” i.e. the entirety
of tasks, which are related to information and communication as economic
goods. The concept of the Information-Function comprises not only management tasks but also service tasks and other tasks in the fields of ICTemployment, which can not be seen as management.
Additionally each company has its individual “Information-Infrastructure,”
i.e. the entirety of systems (hardware, software, etc.), resources, procedures, regulations etc. for the purpose of production, dissemination and
usage of information.
The Information-Function has potentials/capabilities which can be
employed to achieve the strategic goals of the company. Heinrich states
that the top goal of IM is to build up an Information-Infrastructure, by
which the potentialities/capabilities of the Information-Function can be
transformed into company success (Heinrich, pp. 20ff). Effectiveness and
efficiency are formal objectives, which should guide all IM-activities. Like
other areas of management (e.g. Human Resources Management, Logistics-Management, etc.), IM is a cross-functional type of management.
Heinrich, who describes his IM-approach as “management-centred,” differentiates three layers of IM (Heinrich, 2002, pp. 22–23):
• Strategic IM
• Administrative IM
• Operative IM
The three layers represent a hierarchical approach. The first layer “Strategic IM” concentrates on the holistic view of the Information-Infrastructure.
The second layer “Administrative IM” (tactically oriented tasks) engages
in the various components of the Information-Infrastructure. The third
layer “Operative IM” concentrates on operating and maintaining the
Information-Infrastructure. Most of the tasks of this third layer are not
management tasks but service tasks. Figure 2 shows the tasks of the three
layers (Heinrich, 2002). One difficulty is the fact that each of Heinrich’s
IM tasks is bound to only one of the three layers. In reality many IM tasks
have to be fulfilled across more than one layer, e.g. the tasks of planning,
controlling, auditing and the tasks of technology management, human
resources management and business process management.
The support of IM-tasks by methods, techniques and tools is important
enough for Heinrich to introduce the field of “Information Engineering” as
254
Exploring Patterns in Information Management
a special field of IM. The second half of his book on Information Management (Heinrich 2002, pp. 339 ff.) therefore is devoted to:
• “Strategic Information Engineering,”
• “Administrative Information Engineering,” and
• “Operative Information Engineering”
Strategic
Layer
Strategic Situation Analysis
Strategic Goals Planning
Development of Strategies
Planning of Strategic Measures and Actions
Quality Management
Technology Management
Controlling
Auditing
Administrative
Layer
Project Management
Human Resources Management
Data Management
Life Cycle Management
Business Process Management
Knowledge Management
Contract Management
Security Management
Backup Management
Operative
Layer
Production Management
Problem Management
User Service
Figure 2. Heinrich’s Three-Layer of IM (see Heinrich, 2002)
Here Heinrich actually describes 24 concrete methods/techniques, which
are spread in a surprising (if not arbitrary) manner over the three IM-layers. Heinrich describes the importance of these methods/techniques for the
fulfilment of certain IM-tasks. But this does not deliver additional arguments which would support his 3-Layer-Framework as a specific management approach with a clear top-down-orientation.
Figure 3 addresses the connection between Strategic IM and Information
Systems Planning (ISP). Heinrich explains (see Heinrich, 1999, p. 24)
that ISP is performed within a frame which is given by “Strategic Information Management.” ISP is the instrument by which the strategic IMgoals – at the same time strategic goals of the company – are transformed
into Information Systems as the most important components of the
Information-Infrastructure. The planning goals, which are generated by
Seibt
255
Strategic IM (Figure 3), comprise concrete orders for new Information
Systems or Project Portfolios (IM-tasks of Layer 2). Results of ISP steer
the realisation of new systems, which are introduced into the Information-Infrastructure in such a way that they can be used productively
(Heinrich 2002).
In fo r m a tio n F u n c tio n
p la n n in g
g o a ls
S tra te g ic
In fo r m a tio n M a n a g e m e n t
In fo r m a tio n
S y s te m s
P la n n in g
In fo r m a tio n In fra s tru c tu re
ANSW
p la n n in g re s u lts
BASY
PERS
SONST
ANSW
= A p p lic a tio n S o ftw a r e
BASY
= B a s ic S y s te m s
PERS
= P e rs o n e ll
S O N S T = o th e r In fr a s tr u c tu re s
Figure 3. Connections between Strategic IM and Information Systems Planning
(Heinrich, 1999, p. 24)
Wollnik’s Three-Layer-Model
of Information Management
Wollnik, who developed and published his IM-model in 1987/1988, also
presents a three-layers-model. He sees “Employment and Usage of Information” as “the upper level” of IM (see Figure 4). Processes of employment and usage of information are interwoven e.g. in the processes of
decision-making, reporting, coordination and communication, accounting
and auditing (Wollnik, 1988, p. 37). The “middle level of IM” comprises
the “Layer of Information and Communication Systems.” Wollnik uses
the short term “Information System” for a variety of objects:
256
Exploring Patterns in Information Management
• IS as socio-technical systems in the specific organisational context of a
company
• ICT-applications resp. ICT application systems
• well defined techniques, procedures, methods and practices (see Wollnik, 1988, pp. 37–38).
Information systems are built on “the bottom level” of IM, i.e. on the
“Layer of Infrastructures for Information Processing and Communication”
(see Wollnik, 1988, p. 38). Examples of his “bottom level” are
• computers (host computers, PCs, etc.),
• telecommunication systems (PABX)
• computer centers.
L a ye r o f In fo rm a tio n U s a g e
a n d E m p lo ym e n t
R e q u e s ts / O rd e rs
S u p p o rt E ffo rts
L a ye r o f In fo rm a tio n a n d
C o m m u n ic a tio n S ys te m s
S u p p o rt E ffo rts
R e q u e s ts / O rd e rs
L a ye r o f In fo rm a tio n In fra s tru c tu re s fo r
In fo rm a tio n P ro c e s s in g a n d
C o m m u n ic a tio n
Figure 4. Wollnik’s Three–Layer Model of IM (see Wollnik, 1988, p. 38)
Information systems are supported by information infrastructures, which
support information usage and employment processes. In most cases the
higher layers generate requests, which guide the activities of the lower layers. On each of these levels many efforts and capabilities are organized
Seibt
257
to solve layer-specific problems with layer-specific means (see Wollnik,
1988, p. 39). Although Wollnik speaks of “upper, middle, lower” layers,
he does not portray these attributes hierarchically.
Wollnik introduces empirically founded details by differentiating “seven
areas of action”, which he puts over the three layers (see Figure 5). Each of
the seven “area of action” comprises its own:
• planning tasks,
• organisation tasks and
• control tasks,
which Wollnik sees as “general directions of operations.”
Two areas of action are allocated to the upper layer (Information usage
level):
• The Management of internal information usage/employment, in Wollnik’s opinion are either “execution-oriented,” “knowledge-oriented” or
“decision-oriented”
• The Management of external information usage/employment, in Wollnik’s opinion are either “transaction-oriented,” “service-oriented” or
“product-oriented.”
Two of the seven areas of action are allocated to the middle layer (ISlevel):
• The Management of the Structures (technological and organisational)
of Information Systems,
• The Management of the Processes of Design, Development and Implementation of Information Systems.
Both areas include task-components, information-components, persons as
IS-components, hardware-components, organisational rules, and programs
(software-components).
The last three areas of action are allocated to the lower layer (ICT-infrastructures-level):
• The Management of Provision of Technologies and Information Resources
• The Management of Operating Technologies and Information Resources
• The Management of Application Development for Operating Technologies and Information Resources.
258
Exploring Patterns in Information Management
As far as tasks can be separated in practice, the Management of Infrastructures does not contain the execution but only the steering of tasks
(Wollnik, 1988, p. 42).
Process of
Information
Management of
Information
Usage &
Employment
Usage
Directon
of Operations
Purpose of Internal Information
Usage & Employment
Executionoriented
Knowledgeoriented
Decisionoriented
Purpose of External Information
Usage & Employment
Transaction- Serviceoriented
oriented
Productoriented
Planning
Organization
Management of Internal Information Management of External Information
Usage & Employment
Usage & Employment
Controlling
Systems &
Processes
Management of
Information
Systems
Directon
of
Operations
Structures of Information Systems,
Application Systems,
Communication Systems
Tasks Infor- Permation sons
Process of Systems Development
Hard- Soft- Orga. Tasks Infor- Perware ware Rules
mation sons
Hard- Soft- Orga.
ware ware Rules
Planning
Organization
Management of the Structures of
Information Systems
Management of the Process of
Information Systems Development
Controlling
Management of
Infrastructure
Handling
InfrastrucDirection tures
of Operations
Planning
Organization
Controlling
Provision, Supply,
Allocation of Systems
Management of Provision
of Technologies &
Information Resources
Operating, Steering,
Administrating of Systems
Application
Development
Management of
Management of Operating
Developing Application
Technologies &
Systems & Information
Information Resources
Resources
Figure 5. Wollnik´s Seven IM-Areas of Action (see Wollnik, 1988, pp. 39–43)
The bridges to real world phenomena remain on a global level. IM is not
understood as an instrument to revolutionalize the tasks and the task-relations in practice, but as an approach to improve the quality of the outcomes
of information processing and communication processes in real world
companies. Wollnik considers the three layers to be steps in a quality chain
(see Wollnik, 1988, p. 39). The quality of information usage/ employment
depends on the quality of Information Systems. The quality of Information
Systems depends on the quality of Information Infrastructures. Most
important outcomes of IM for the management of a company are information and information resources, which enable management on all hierarchical levels to make better decisions. Usage and employment of information for better strategic planning and as a “strategic weapon” against competitors on markets (see Wisemann, 1985; Mertens/Plattfaut, 1986) are
included.
Wollnik’s approach does not include concrete statements on the question
of integrating IM into a company organisation. He is not explicit about the
Seibt
259
relationships between IM and company management nor about the problems of evaluating IM-tasks to be strategic, tactical or operative tasks.
Neither is his approach specifically engaged in the problems of change,
e.g. in the process of “Information Systems Development” as the main
trigger to produce effective and efficient IS. Nevertheless Wollnik’s “areas
of action” are taken as a modelling basis in several German books and articles on Information Management (see e.g. Picot, 1996; Krcmar, 2000;
Voß/Gutenschwager, 2001).
Seibt’s Four-Columns-Concept
of Information Management
In their book of 1986 Marchand and Horton emphasize five historical
stages of IM. In the early Sixties IM started as “Management of Programming or Programmers” and as “Management of Paperwork.” In the late
Eighties – after less than 30 years – in their opinion IM had reached the
fifth stage, which they characterised as “Management of Information
Employment for Strategic Purposes” (Management of Strategic Information Usage) (see Marchand/Horton, 1986).
Most of the changes in the perception of IM in companies have been
evoked by technological enhancements and changes. ICT was the trigger,
which has brought about new ideas on how to employ ICT to support strategic planning and how to increase business success. Certainly the advent
of a whole bunch of Network-Technologies, combined with the Internet
and the technology of Personal Computers for decentralized as well as centralized ICT-solutions, had the greatest influence on the growth of support
capabilities for all kinds of managers, experts, organisation units etc. Other
ICT-based developments, which already arrived in the late Eighties, but
which have been improved continuously since that time are:
• New Computer Architectures and new Operating Systems (e.g. clientserver architectures, browser architectures)
• New Programming Languages (e.g. object oriented languages)
• New Methods and Tools for Software Design and Development (e.g.
modelling tools)
• New Architecture of Application Software Systems (e.g. standardized
application software architectures for certain business branches)
• Technology of Knowledge Based Systems (e.g. Technology of Artificial Intelligence Systems and Expert Systems)
260
Exploring Patterns in Information Management
• New Architectures for Data Banks (e.g. architecture of Data Warehouses, Content Management Systems.
This list is by no means complete. It gives only a subjective view of some
important areas of ICT-progress. In parallel the growing experience with
successes and failures of IS as socio-technical systems in a huge number of
organisations has also led to new concepts for the management of IS
development and implementation processes:
• Concept of IS Lifecycles – a lifecycle concept, which comprises not
only the processes of IS-(first)development and (first)implementation
but also the processes of maintenance, revision and continuous
improvement of each IS during its lifetime.
• Holistic Concept of IS development – a concept, which organizes ISprojects as multi-dimensional processes, which give attention and participation to all kinds of players (managers, users, ICT-professionals,
business and organisation consultants etc.), who expect benefits/advantages or disadvantages from the IS to be developed.
• Concept of perceiving the processes of IS-development etc. as political
processes with consciously acting proponents and opponents.
• Concept of Ensuring Fit between the main IS-components, i.e. tasks to
be fulfilled by the ICT- subsystems, people involved as players/parties,
and organisational structures and processes (see Seibt, 1991).
• Concept of Business Reengineering combined with the Concept of
Organisation Change, which secures the dominance of long-term
company goals and benefits in the design considerations of IS-development.
• Concept of conscious goals specification/revision and continuous control of goal fulfilment. Not only financial goals, but all kinds of goals
which are important for the company and for the parties involved in ISdevelopment, must be included.
This list again is not complete. It also gives only a subjective view of experiences which will enlarge the likelihood of IS-success if they are considered in IS-development processes.
The result of such experiences is an approach to IM that intends to bring
about some new patterns of IM which partly go beyond the patterns pronounced by the models of Lundeberg, Heinrich and Wollnik (see Figure
6):
Seibt
261
Information Management of the Company
Management of
Networks and
Computer Resources
Objects
Computers and Networks
Information and Communication:
Technologies as Resources
Tasks
Management of
Information System‘s
Life-cycle
Management of Information
and Knowledge Demand
& Supply
Objects
Information and Knowledge;
Information- and KnowledgeSupply-Processes
Objects
Information- und Communikationsystems as Man-Mashine-Systems
In Organizational Context
Tasks
Strategic Planning
of I&C-Technologies
Architecture-Management
(ICT-Architectures)
System Management
Computer and
Installation-Management
Network-Management
Security -Management
• Controlling the Success of
Computers & Networks Usage
Tasks
Architecture-Management I
(Architecture of Application Systems)
Strategic Planning of
Informationsystem-Projectportofolio
Architecture-Management II
(Architecture of Methods- and
Tools-Systems)
Management of Development, Maintanance
and continous Improvement of various kinds
of Information Systems
• Controlling the Success of Development and
Maintanance Processes
• Controlling the Success
Informationsystems
Management of Enlarging
Company Success and
Enhancing Technologies
As Success Factors
Objects
ICT Based Enlargement of
Organization Success; New ICT-Based
Products and Services
Tasks
Strategic Planning of Informationand Knowledge-Supply in the
Organization
Strategic Success and
Potentials Planning
Company-Modelling
Business Process-Modelling
Strategic Planning of ICT Based
Products and Services
Global Data-, Functionsand Processes-Modelling
Starting New Production-Lines
Data-Management
Organizational Development
Standards and Procedures
Planning and Steering
Company-Changes
Knowledge-Management
• Controlling Success of
Information and Knowledge
Supply & Employment Processes
• Controlling Organization Success
and Enhancement of ICT-Based
Succes-Factors
Figure 6. Seibt’s “Four-Columns-Concept” of IM (see Seibt, 1993, p. 16)
Idea 1
Strategic planning is needed not only as a global strategic planning of
company goals, but has to be done by all organisational units which in
principle act as independent “players” – as entrepreneurs in market places
– delivering various kinds of services and components to various “clients”
within one company. This is true for the four sectors which are shown in
Figure 6 as “columns of IM.” Today in many situations the four sectors act
as independent outsourcers, consultants etc although they may have started
in the past as dependent departments within the company:
• Management of Networks and Computer Resources (column 1)
• Management of Information Systems Lifecycles (column 2)
• Management of Information and Knowledge Demand and Supply (column 3)
• Management of Enlarging Company Success and Enhancing ICT as a
Success Factor (column 4).
Idea 2
The classical dichotomy of the “Business Domain” and the “Technology
Domain,” within/or of one company (see Parker et al., 1988) should be
262
Exploring Patterns in Information Management
substituted by a more flexible, more adaptable pattern. The four sectors
can be seen as being interdependent but also independent elements of a
value chain. Internal units, which have been responsible for many years
within one company, will be substituted by external “players” if the offerings of the external players (outsourcers, consultants etc.) help to reduce
costs and/or generate higher benefits for the company.
Idea 3
Comparable to strategic planning as part of IM is the situation in the field
of architectures. Architectural Decisions and Management of Architectures
have become very important tasks in all columns of IM. Computer- and
network-architectures on one hand and standard application software
architectures on the other hand are at the same time interdependent and
independent factors. Both set long timeframes for top management decisions. The same can be true for implementation and usage of Data Warehouses, Knowledge Management Systems, and various kinds of data models, functional and process models, which belong to the third column and
which condition the processes of information and knowledge supply
through the whole company.
Idea 4
Each IM-column has its own controlling problems and tasks. The objects
of the first column are computers, networks, all kinds of ICT-equipment
etc. Control in this column has to concentrate on technical goals which are
pursued by configuration- and installation-processes, and by running technical subsystems. The objects of the second column are IS in the specific
organisational context of a company or of a department in a company.
Controlling in this second column has to concentrate on the economic,
organisational, personal, etc. aims of the users of one specific IS, concerning the question: does this IS support its users to fulfil their specific
local tasks in their company. Additionally controlling in the second column has to concentrate on the effectiveness and the efficiency of the lifecycle-processes of one specific IS. The objects of the third column are the
processes of information and knowledge supply for all kinds of users in the
company. Controlling in the third column means controlling the success of
these supply processes. Different criteria with different measures and different assessments have to be combined to achieve the controlling purposes in each of the four columns.
Seibt
263
Idea 5
The core of “Management of IS-Life-cycles” is supported by a “Concept
of Project Phases.” It stands at the centre of the process of Steering and
Controlling IS-Development. Many authors limit their analyses and recommendations to the process of application software development and to
the implementation of technical components. They neglect the tasks of
planning, designing, developing, and implementing the organisational and
personal components of a specific IS. An Information System is a sociotechnical system (people to be integrated with machines), which has to be
embedded in a specific organisation. This can only be successfully
achieved by performing a concurrent and equal process of implementing
all components from the very beginning to the end of the life-cycle of the
Information System (see Seibt, 1997, pp. 431 ff.). From this view several
consequences should be drawn for the definition of IM tasks in IS-development (see Figure 7):
Process of Steering and Controlling IS-Development
Process of Quality Management
Process of equal and concurrent implementation of system components
Preliminary Project
Main Project
Design
Organizational
CompoNents
(detailed)
of
Initialization
P1
Feasi-
Holistic
bility
System
study
Design
P2
M1
Detailed
Holistic
System
Design
M2
Realization
of ORGComponents
M3
M6
Design
of
Software
CompoNents
(detailed)
RealiZation
of
Software
Components
Integration
of
System
Components
Holistic
Testing &
Consolidation
of Integrated
System
M4
M7
M9
M10
Design of
PersonellCompoNents
(detailed)
Realization of
Personell
Components
System
Delivery
to
Operations
P = Preliminary
Project Phases
M = Main
Project Phases
Figure 7. Seibt’s Concept of Project Phases for IS-Development
(see Seibt, 1997, p. 433)
Idea 6
The stream of IM-activities designed to perform operative system development should be structured in at least a “preliminary project” and a “main
264
Exploring Patterns in Information Management
project,” in order to separate feasible and wanted projects from non-feasible and/or unwanted projects. Only those projects, which are technically,
organisationally and personally feasible, and at the same time wanted by
(top) management – and also wanted by enough “players” in the company
who will benefit from the IS to be built – should enter the stage of a “main
project.”
Idea 7
Project phases should be considered as blocks of work which can be performed sequentially or concurrently. The concept of project phases should
be combined with mile-stone-logic. Mile-stones can be introduced phaseindependent. As many mile-stones can be inserted in the course of phases
as are needed to achieve consensus between the parties (“players”), participating in the process of system development.
Idea 8
Concepts of project phases should be custom-tailored to each specific ISdevelopment. The technological, organisational and personal conditions of
an IS-development-process always constitute a specific environment, to
which a specific configuration of IM-tasks must be found and adapted.
This form of custom-tailoring is not restricted to the first phase of a project
but is a continuing process of adjusting IM-activities to the ever changing
external and internal conditions of each specific IS-development.
Idea 9
IS-development starts with global holistic design considerations and proceeds to a detailed holistic design of the IS. This generates the “bracket,”
from which the detailed design activities for technical, organisational and
for the personnel-components of the IS can be performed. After the realisation of all components, integration of all components should take place
(see Figure 7). Before delivery of the new IS to routine operations it has to
be tested and consolidated very carefully as totally integrated system.
Idea 10
The concept of project phases gets consistency by three cross-sectional,
respectively “cross-phases”, types of activities:
Seibt
265
• Process of Steering, Coordinating and Controlling IS-Development
• Process of Quality Management (analytical and constructive QM)
• Process of System Implementation (embedding the new system in an
existing environment).
Steering, Coordinating and Controlling are the dominating IM-activities
which penetrate and wrap up all activities of IS-development.
Krcmar’s Concept of Information Management
Krcmar takes Wollnik’s object-oriented “Three-Layers-Concept” as a
starting point for his IM-model (see Krcmar, 2000; Wollnik, 1988). The
“upper level” is named “Management of Information Economics”, concentrating on “Supply and Demand and Usage/Employment of Information.”
The “middle level” is named “Management of Information Systems,”
which is exactly the same name and meaning as Wollnik’s “Layer of
Information and Communication Systems.” The “lower level” is named
“Management of Information and Communication Technology” instead of
“Layer of Information Infrastructures.” Krcmar’s approach concentrates on
“Management of ICT.”
Management Tasks
of IM
Management of the
Information Economy
in the Company
- Strategy and IM
- Organization of IM
Management of
Information Systems
- Supply
- Demand
- Usage
- Data
- Processes
- Application Life Cycle
- Personel of IM
- Controlling as
IM-Tasks
Management of
Information - and
Communication -Technology
- Storage
- Processing
- Communication
- Bundles of ICT
Figure 8. Krcmar’s Model of IM (see Krcmar, 2000)
266
Exploring Patterns in Information Management
Interestingly enough, Krcmar additionally adds a column to his three layers (see Figure 8) This column comprises the “Management Tasks of IM,”
which in his opinion are cross-sectional tasks and spread across the three
horizontal layers (Krcmar, 2000, p. 34):
•
•
•
•
Company Strategies and IM
Organisation of IM
Management of IM-Personel
Management of IM-Controlling.
IM for Krcmar is at the same time a management and a technology discipline. On one side, for him IM belongs to the elementary components of
management of the company (see Teubner/Klein 2002). On the other side,
he mostly concentrates on technological questions.
In his last chapter he summarizes four cross-sectional aspects/topics,
which obviously could not easily be aligned within his three layers plus
one colum approach of IM (Krcmar, 2000, p. 322 ff.):
• Business Process-Orientation, Business Process Reengineering, Process
Design
• Security in various forms (e.g. data security) including Management of
Security
• Standardisation and distribution of business activities and systems
• Synchronisation of speeds of developing systems and components in
parallel.
Krcmar’s picture of IM tasks does not have the same rigidity and strictness
as the models of for example Lundeberg or Wollnik. On the other hand his
framework is rigid and consistent enough to deliver valuable patterns for
descriptions and analyses of real life cases. Krcmar is successful in
describing and analysing a large number of complex cases based on the
business realities of various companies (see for example Krcmar, 2000, pp.
89 ff., 100 ff., 144 ff., 201 ff.).
Laudon and Laudon’s Integrated Framework for
Describing and Analysing Information Systems
The book of these two well known American authors has the title “Management Information Systems – Managing the Digital Firm” and is now in
its 7th edition (see Laudon and Laudon, 2002). Throughout the book the
authors avoid the term “Information Management,” although there is no
Seibt
267
doubt that they write a book on Information Management of or within
firms. The book has five parts:
•
•
•
•
•
Organisations, Management, and the Networked Enterprise
Information Technology Infrastructure
Building Information Systems in the Digital Firm
Management and Organisational Support Systems for the Digital Firm
Managing Information Systems in the Digital Firm
Plan product
offerings
Establish sales
targets
Develop change
strategy
Business Challenges
M
Management
Coordination problems
from rapid growth
Manual process
New initiatives from
competitors
Private Web sites
Desktop
computers
Apparel Buying
Network
T
Technology
Suppliers
Retailers
Retail
customers
Employees
O
Organization
Information
IS
System
Business
S
Solution
Order mechandise
Increase service
Track orders
Reduce costs
Access benefit plans
Increase revenue
Purchase supplies
Broadcast messages
Review designs
Figure 9. Laudon and Laudon’s Integrated Framework for Describing and Analysing Information Systems (example here: GUESS. GUESS Annual Report March
2000; see Laudon and Laudon, 2002, p. 3)
In their introduction the authors explain that they use an “Integrated Framework for Describing and Analyzing Information Systems.” This portrays IS
as always being composed of “Management,” “Organisation,” and “Technology” components (see Laudon and Laudon, 2002, p. XXV). The authors
(p. 78) cite Leavitt (1965) and his model of four components/dimensions of
an IS. They stress that, according to this model, in order to implement (business) change all four components must be changed simultaneously. Their
“Organisation” pattern does include structural as well as human aspects.
They use hundreds of real world examples: i.e. company cases, illustrating
the management, organisation, and technology issues in their chapters. Not
only the “Window On Boxes Technique”, but also their “Management
268
Exploring Patterns in Information Management
Wrap-up Overviews of Key Issues” at the end of each chapter, take these
patterns as a point of focus to make the reader aware of what Laudon and
Laudon feel is most important. As Figure 9 shows, the emphasis is on the
problems of building information systems. Existing business challenges are
causes to develop new IS, which enable new business solutions. Management, Technology and Organisations are the driving factors for new IS,
which enable more efficient and effective business solutions, which again
cause new (positive or negative) business challenges.
Part Two of Laudon and Laudon’s book is devoted to “Information Technology Infrastructure” with three chapters on “Managing Hardware Assets,
Software Assets and Data Resources.” Part Four concentrates on “Management and Organisational Support Systems for the Digital Firm,” with
two chapters on “Managing Knowledge: Knowledge Work and Artificial
Intelligence” and “Enhancing Management Decision Making.” This part is
clearly engaged in IM-tasks, which Wollnik and Krcmar would assign to
the “Layer of Information and Knowledge Usage and Employment.”
Some Conclusions
Comparing the six concepts of IM, which have been described here, the
following conclusions can be accentuated:
• The concepts have different points of emphasis but they are not contradictory. In this respect the concepts complement each other.
• Different points of emphasis evolve, since the authors of the concepts
want to achieve different results. Some examples: One of Lundeberg’s
concerns is to apply IM not only to the development of IS but also to
(re-) shaping business. One of Seibt’s concerns is to stress the value
chain character of the relationships between the four columns of IM and
to open the view for a broad acceptance of outsourcing as part of IM.
• Without exception, all concepts are suited to serve as patterns to analyse and to model the reality of Information Management in companies.
Most of the concepts “prove” this ability by having been used to analyse and model real world IM-cases in companies, which are described
in books and articles.
• The power of a concept for a binding (re-) construction and (re-) shaping of the “Gestalt” of Information Management Structures in a real
world company is much harder to measure and to evaluate. Binding
construction and shaping of IM-Structures mostly will be the result of
Seibt
269
professional consultancy activities. Normally these results are not published because of confidentiality.
• One open question is: how important is the degree of abstraction of patterns for the purpose of pattern-usage? Is it better – for the purposes of
analysing and modelling reality – to have and to use more abstract patterns?
• Combined with this question comes an idea: maybe we need – especially for the definition and construction of IM-patterns – a continuum
of patterns, which can be systematically concretised step by step from
abstract to concrete “stamps” or instances/attributes.
References
Benjamin, R.I., Dickenson, C. & Rockart, J.F. (1985) “Changing Role of the Corporate Information Systems Office”, MIS Quarterly, Vol. 9, No 3, pp. 177–99.
Heinrich, L.J. (2002) Informationsmanagement. Planung, Überwachung und
Steuerung der Informationsinfrastruktur, 7th ed., Oldenbourg, Munich and
Vienna.
Heinrich, L.J. & Burgholzer, P. (1987) Informationsmanagement. Planung,
Überwachung und Steuerung der Informationsinfrastruktur, 1st ed.,
Oldenbourg, Munich and Vienna.
Horton, F.W. & Marchand, D.M. (1982) Information Management in Public
Administration. An Introduction and Resource Guide to Government in the
Information Age, Information Resources Press, Arlington, Virginia.
Krcmar, H. (2000) Informationsmanagement, 2nd ed., Springer, Berlin.
Laudon, K.C. & Laudon, J.P. (2002) Management Information Systems: Managing the Digital Firm, 7th ed., Prentice Hall, Upper Saddle River, New Jersey.
Leavitt, H.J. (1965) “Applied Organizational Change in Industry: Structural,
Technology and Humanistic Approaches,” in March, J.G. (ed.) Handbook of
Organizations, Rand McNally, Chicago, pp. 1144–1170.
Lundeberg, M (1993) Handling Change Processes – A Systems Approach, Studentlitteratur, Lund, Sweden.
Marchand, D.A. & Horton, F.W. (1986) Infotrends. Profiting From Your Information Resources, Wiley, New York.
Mertens, P. & Plattfaut, E. (1986) “Informationstechnik als Strategische Waffe,”
Information Management, Heft 2, pp. 6–17.
Parker, M.M., Benson, R.J. & Trainor, H.E. (1988) Information Economics.
Linking Business Performance to Information Technology. Prentice Hall,
Englewood Cliffs, New Jersey.
270
Exploring Patterns in Information Management
Picot, A. & Franck, E. (1992) “Informationsmanagement”, in Frese, E. (ed.),
Handwörterbuch der Organisation, 3rd ed., Poeschel, Stuttgart, pp. 135–36.
Rockart, J.F. (1998) “The Line Takes the Leadership – IS-Management in a Wired
Society,” Sloan Management Review, Vol. 29, No. 4, pp. 57–64.
Scott Morton, M.S. (ed.) (1991) The Corporation of the 1990s: Information Technology and Organizational Transformation, Oxford University Press, New
York.
Seibt, D. (1990) “Informationsmanagement und Controlling,” Wirtschaftsinformatik, Vol. 32, pp. 116–26.
Seibt, D. (1991) “Informationssystem-Architekturen,” Innovations- und Technologie-Management, Müller-Böling, D., Seibt, D. & Winand, U. (eds.), Stuttgart, pp. 251–84.
Seibt, D. (1993) “Begriff und Aufgaben des Informationsmanagement – ein Überblick”, in Preßmar, D.B. (ed), Informationsmanagement, Gabler, Wiesbaden,
pp. 3–30.
Seibt, D. (1997) “Vorgehensmodell”, in Mertens, P. et al. (eds.), Lexikon der
Wirtschaftsinformatik, 3rd ed., Springer, Berlin, pp. 431–34.
Strassmann, P.A. (1988) “Justification of Investments in Information
Technology,” Second International Management Symposium ´Erfolgsfaktor
Information’ (Diebold und Wirtschaftswoche), Frankfurt, pp. 403–48.
Synnott, W.R. (1981) “Changing Roles for Information Managers,” Computerworld, No. 38, pp. 18–28.
Synnott, W.R., Gruber, W.H. (1981) Information Resource Management –
Opportunities and Strategies for the 1980s, Wiley, New York.
Szyperski, N. (1981) “Strategisches Informationsmanagement im technologischen
Wandel,” Angewandte Informatik, Vol. 22, No. 4, pp. 177–95.
Teubner, A. & Klein, S. (2002) “Informationsmanagement (Vergleichende
Buchbesprechung),” Wirtschaftsinformatik, Vol. 44, No.3, pp. 285–99.
Voß, S. & Gutenschwager, K. (2001) Informationsmanagement, Springer, Berlin
etc., pp. 72–79.
Wiseman, C. (1985) Strategy and Computers: Information Systems as Competitive
Weapons. Dow Jones-Irwin, Homewood, Illinois.
Wollnik, M. (1987) “Aktionsfelder des Informationsmanagement,” Jahresbericht
der GMD, pp. 148–66.
Wollnik, M. (1988) “Ein Referenzmodell für das Informations-Management,”
Information Management, No. 3, pp. 34–43.
PART FOUR:
DEVELOPING THE FIELD OF
INFORMATION MANAGEMENT
Blanksida
— 16 —
Building an International Academic
Discipline in Information Systems
Gordon B. Davis
Introduction
Most academic disciplines within the broad field of management or economic sciences developed within the context of a country or a region.
Examples are accounting, marketing, and industrial relations. They are
working to be international. The academic discipline of information systems (or whatever name may be used in different universities or different
countries) became international very quickly. Several conditions facilitated
this development, and it has been remarkable in its scope and impact. Mats
Lundeberg is one of the academics who have nurtured the international
discipline of information systems. It is appropriate to honor that contribution on this occasion.
I have been fortunate to have been a part of many of the developments that
helped the formation of an international community of information systems scholars. This article describes my perception of some of the critical
decisions and events that helped build the international network. Since it is
also a personal journey, I often mention personal involvement and personal
experiences. I recognize that the description is limited by my own experience. It is not complete; I may have missed some critical contributions.
Rather than being a complete historical account, this article is my view.
The article discusses the name issue and why it took time for information
systems to develop an identifiable, well-defined international community.
It then focuses on seven critical events or developments that made it possible to have an international academic discipline for information systems.
These are the development of computing devices, the use of English as the
common language for computing-related disciplines, the formation of the
International Federation for Information Processing (IFIP) and its Technical Committee 8 (Information Systems), international efforts by scholars in
several countries, locating the IFIP TC8 working conferences internationally including the Manchester Conference sponsored by WG8.2, the found-
274
Exploring Patterns in Information Management
ing of the International Conference on Information Systems (ICIS), and the
founding of the Association for Information Systems (AIS) with an international governance structure.
The Name of the Systems, the Business Function,
and the Academic Discipline
In organizations, the term Information System or some equivalent refers to
both:
• The systems that deliver information and communication services
• The organization function that plans, develops, and manages the information systems
The name for the academic discipline more or less mirrors the organization
use. Some of the names that are used illustrate the common theme:
•
•
•
•
•
Information Systems
Management Information Systems
Information Management
Management of Information Systems
Informatics (usually modified by organization, administration, or similar terms)
Informatics has some appeal. It appears to have originated with the French
informatique. It was not used in the US because it was a copyrighted term
for a business (that later went bankrupt).
The term Management Information Systems or MIS reflected the strong
theme that the function and the academic field were most concerned about
the new, powerful uses of computers to change the information presented
to management and the analysis for management decision making. Transaction processing was, in the early years, not considered very interesting.
Over time, there has been a trend to employ the simple term, Information
Systems, in referring to the academic discipline, but there are still many
variations in practice. Within the academic discipline, many use terms that
reflect the management or administrative use of computers. For example,
the Stockholm School of Economics (Handelshögskolan i Stockholm) unit
for information systems refers to itself as the Department of Information
Management.
It is interesting to note the problem of terminology is found elsewhere. A
simple, non-encompassing term of computer is used for complex comput-
Davis
275
ing and communications and processing devices. The academic field is
termed Computer Science. There are other alternatives. For example, the
Swedish term dator for computer seems to reflect an emphasis on data
processing rather than computation.
Why it Took Longer to Build an International
Information Systems Academic Community Than a
Computing Community
The short review that follows of the international development of computing machinery shows the significant time lag between formation of an
international community for the overall field of computing and the formation of an identifiable, well-defined community of scholars in an academic
discipline for information systems. Some key dates I recognize (and I may
have missed some important ones) in the evolution toward an international
academic discipline illustrate the time lag:
1954 First business use of computers
1958 First speculation of importance to business of computers in Harvard
Business Review
1960 Forming of International Federation for Information Processing
(IFIP)
1965 Börje Langefors appointed as professor (joint chair at the Royal
Institute of Technology and the University of Stockholm) in Information Processing, with special emphasis on Administrative Data
Processing
1968 First formal MIS academic degree programs in the US (M.S. and
Ph.D.) at University of Minnesota
1968 Establishment of organization for information system executives
(CIOs); first called Society for Management Information Systems
and now Society for Information Management (SIM)
1976 Establishment of IFIP technical committee on information systems
(TC8)
1977 The journal MIS Quarterly started at the University of Minnesota
1980 First International Conference on Information Systems (ICIS)
1994 Formation of Association for Information Systems (AIS) as an international academic organization with an international governance
276
Exploring Patterns in Information Management
structure. Merger in 2001 with ICIS as world conference for AIS.
Alliances with regional conferences in Europe, Asia, and America
(ECIS, PACIS, and AMCIS)
In my view, the time lag was caused by three major factors: the time lag
between the introduction of computers and the recognition of an interesting, important business function and related interesting, important research
issues; the diverse backgrounds of academic researchers and conflicting
loyalties with existing organizations; and existing conferences and journals
that accepted IS research results.
Time Lag between Introduction of Computers, Business Function
and Research Issues
The time lag between the introduction of computers and the recognition of
an interesting, important business function and related interesting, important research issues: Punched card data processing and the related use of
business machines were not an interesting academic subject, either for
teaching or research. Even faculty and students in accounting viewed the
subject as not suitable for the curriculum. Since early use of computers
focused on simple transaction processing, it did not look interesting. What
sparked interest very early was the possibility of improved analysis,
improved reporting, and improved decision making.
As business developed and implemented computer-based data processing
systems, it became apparent that there were many interesting problems
ranging over topics such as requirements determination, development
methodologies, implementation, design of work systems, and evaluation. It
became interesting, but this emerged slowly. Why so slowly? Probably
because the cycle of technology innovation is very short; the cycle of process and system innovation is much longer (say roughly two to three years
for technology and six to nine years for process and systems).
Diverse Background of Researchers
The diverse backgrounds of academic researchers and conflicting loyalties
with existing organizations: Early academic researchers came from a variety
of backgrounds such as management, accounting, computer science, and
management science. Doctoral students in the 1960s who were interested in
information systems took doctorates in these subjects but researched interesting problems in information systems. Although information system
research was emerging rapidly, it was not until 1968 that the first formal
Davis
277
doctoral program in information systems in North America was established
at the University of Minnesota (along with an MIS research center).
An important consideration for researchers in schools of administration or
business (especially in North America) was the fact that their colleagues,
with academic training and traditions that did not incorporate information
systems and the information systems function, often did not understand or
appreciate the importance of the new technologies, systems, and organization function. A safe way to maintain academic credibility was to fit within
an existing management, accounting, marketing, or operations research
discipline. Some of the most prestigious universities were slow to recognize the new realities of the computer age.
Existing Conferences and Journals
Existing conferences and journals that accepted IS research results: Given
the diverse backgrounds of researchers and the diverse department affiliations, the early researchers looked to their home discipline for opportunities to present and publish their work. Several organizations formed special
interest groups around the issues of information systems. Examples were
SIGMIS by ACM and College on Information Systems by Management
Science (now INFORMS). Journals such as Management Science and
Communications of the ACM published MIS research. SIGMIS created
Database to cover the topic. There were journals in Europe that focused on
information systems research, but these had not yet become key outlets for
the entire community. IFIP was important, especially its technical committee on information systems, in sponsoring conferences and publishing
conference proceedings during the development period.
Given the three issues discussed above, the time delays in formalizing the
international community of scholars in a new academic discipline are
understandable. However, it should be recognized that the formal developments such as conferences and organizations were the result of informal
networks of scholars that developed rather quickly and inputs from forward-looking practitioners who recognized the need for good research.
Critical Development 1:
Development of Computing Devices
After World War II, there was interest in many universities around the
world in the design and development of computing machinery. Wellknown efforts took place at the University of Pennsylvania and Massachu-
278
Exploring Patterns in Information Management
setts Institute of Technology in the United States and Manchester University in the UK, but there were similar research efforts in Sweden, Switzerland, and other countries. The community of researchers shared designs
and experiences, so the development of computing machinery can be considered an international effort.
Building a computer in a university laboratory is one thing; building a
commercially viable computer is another. As frequently happens, a sponsor emerged. The United States government was interested in a computer
to use for tabulating the 1950 census of people in the US. There was an
historical precedent. Punched card data processing had been invented for
use in the 1890 US census. During the next 60 years, it grew to dominate
processing of simple transactions and preparation of fairly simple reports
for business and government. IBM was the leader in this industry. IBM
was not, however, the developer of the first commercial computer. It was
done by a start-up company, the Eckert and Mauchly company, headed by
two of the developers from the University of Pennsylvania. Their product
was the UNIVAC I.
The business use of computers happened in 1954 in two countries. The
first business use was in the UK by the Lyons Tea Company, owner of a
chain of tea shops. They commissioned Manchester University to develop
a business computer. It was named LEO for Lyons Electronic Office. In
the US, General Electric began a business use of a UNIVAC I for payroll
at its Louisville Appliance Park. The Harvard Business Review published a
futurist article by Leavitt and Whisler (1958), “Management in the 1980s”
that forecast large changes in organization structure and management in
the next 30 years based on the availability of computers.
Computer Science developed very early as an international community
even though each country tended to have its own organization. In the US,
the Association for Computing Machinery was founded in 1947 when
research was being done on new devices for computation. The first Computer Science departments with formal degree program in the US were
started at Stanford University and Purdue University in 1962. Research on
computer science issues resulted in doctorates earlier than this but not from
formal degree programs.
Why didn’t Information Systems begin at the same time as business use
began? As mentioned earlier, the problem was that data processing, as
typified by card punch equipment and various business machines, was not
academically interesting. The systems were easily learned by observation;
the processes were applied mainly to simple transactions and simple
Davis
279
reports. Applying computers to the same processes did not create an academic interest.
There was interest in computers in schools of administration and management, but the focus was mainly on the computer as a tool for analytical
models and sophisticated analysis. Almost every major management
school had one or more faculty who taught some elements of computer
technology. I observed that development first hand. My first book, Introduction to Electronic Computers (1965), was directed not at a new discipline but the general business students. I believed they should understand
something about computer technology and its use in business. Langefors
began as Professor of Information Systems in Sweden in the same year and
published his Theoretical Analysis of Information Systems (1966). My second book, Computer Data Processing (1969) had much more emphasis on
how the computer is used for data processing and other business applications. Note these books were 10 to 15 years after the first use of business
computers.
Critical Development 2:
The Use of English as the Common Language for
Computing-Related Disciplines
A common language is very important in building an international community of scholars in a discipline. Greek, Latin, German, and French have
provided such a common language for various communities at different
times in history. The development of computers, although occurring in different countries, had major developments in the US and the UK. This
encouraged the use of English as the language for the computing field. As
will be noted later, English was adopted as the language for the International Federation for Information Processing (IFIP). At the same time,
there was a general recognition by scholars and business leaders of the
value of an international language. English became the common language
of international commerce and of research and education in many fields.
The Netherlands and the Scandinavian countries had taught English as an
important second language; the English emphasis was further increased
during the period when computing was developing and the field of information systems was beginning to emerge.
The common language of English has meant that international conferences
on computing and information systems can be held at almost any location
in the world, research is freely exchanged across boundaries, and text-
280
Exploring Patterns in Information Management
books and trade books are made available internationally. For example, the
Swedish ISAC methodology developed by Mats Lundeberg and others was
published in English, Information Systems Development – A Systematic
Approach (1981). Information systems instruction and research in Sweden
illustrate the importance of a common international language. Any English
language book needed for instruction in Sweden can be used without
translation.
I was fortunate to write a significant book in the field in 1974 with second
edition in 1985 (with Margrethe Olson). Many rank the book as a defining
book for the field. It was used throughout the world by scholars who now
form the nucleus of the discipline. The book, Management Information Systems: Conceptual Foundations, Structure, and Development (1974, 1985),
outlines the major concepts employed in the field and their relationship to
the structure of systems and management of the function. A revision today
would add concepts and modify some of the structure that is defined, but it
has been noted as a classic textbook in the field. Similarly, Börje Langefors
Theoretical Analysis of Information Systems (1966) was important for the
development of the discipline in the Scandinavian countries.
Critical Development 3:
The Formation of the International Federation for
Information Processing (IFIP) and its Technical
Committee 8 (Information Systems)
In the early development of computing and its use in organizations,
national organizations were forming, but there was no accepted international forum. The United Nations provided the impetus for the formation
of an international information processing organization. UNESCO sponsored the first World Computer Conference in 1959 in Paris (five years
after the first business uses). This was followed by the organization in
1960 of the International Federation for Information Processing (IFIP).
IFIP is a non governmental, nonprofit umbrella organization for national
societies working in the field of information processing (essentially a society of societies).
Technical work, which is the heart of IFIP’s activity, is managed by a
series of Technical Committees (TCs). Each member society (usually
identified with a country) may appoint a representative to the governance
committee for each technical committee. There are currently 12 technical
committees. Each technical committee forms working groups. Individuals
Davis
281
throughout the world may be members of a working group by demonstrating interest and continuing activity in the work of the group.
The IFIP technical committee of interest in this view of the development
of an international academic discipline is TC8 (Information Systems). It
was established in 1976. Its aims are to promote and encourage the
advancement of research and practice of concepts, methods, techniques,
and issues related to information systems in organizations. Note that it was
formed 22 years after the first use of computers in business. It currently
has seven working groups.
•
•
•
•
•
•
WG 8.1 Design and evaluation of information systems
WG 8.2 The interaction of information systems and the organization
WG 8.3 Decision support systems
WG 8.4 E-business: multidisciplinary research and practice
WG 8.5 Information systems in public administration
WG 8.6 Diffusion, transfer, and implementation of information technology
• WG 8.8 Smart cards, technology, applications & methods
TC8 was important in helping to build an international community. Its first
chairman was Börje Langefors of Sweden. It started as somewhat Europecentric but rapidly expanded to worldwide participation. I personally
observed the building of that community through the TC8 national representatives and the meetings of the working groups. I was the second US
representative to TC8 and remained in that position (and as chairman) for
20 years. Mats Lundeberg was already the Swedish representative to TC8
when I was appointed, so he was part of this important early nurturing of
the field. The structure of IFIP and TC8 was a limiting factor that prevented it from becoming the focus of an emerging international community for an information systems discipline. Those limiting factors are
another story; however, its role in the early development was important.
Critical Development 4: International Efforts by
Scholars in Several Countries
It is difficult and somewhat dangerous to start mentioning specific names of
important innovators and contributors. Even a casual reading of the history
of inventions shows again and again that important innovations are “in the
air”. Several people are working on the same problem and coming to the
same solutions, but one or only a few are recognized as the inventors. We all
thought of Eckert and Mauchly as the inventors of the first computer, but
282
Exploring Patterns in Information Management
another won a lawsuit establishing prior invention. In fact, scientists and
engineers in many countries were working on the problem and converging
on a solution. Given this caveat, I recognize a few critical innovators who
helped in the formation of the international discipline. The Association for
Information Systems has recognized some of these by giving them the LEO
award for lifetime exceptional achievement in information systems.
Colleagues at the University of Minnesota
The founding of the Minnesota academic programs in MIS in 1968 was a
joint product of Professor Tom Hoffmann, Professor Gary Dickson, and
myself. Tom was chairman of the Management Sciences Department to
which we were finally attached. Gary was tireless in building the program.
He started the MIS Quarterly and nurtured it through the growing pains of
its first six years. He was one of those who started the first International
Conference on Information Systems in 1980.
Also important to mention is Janice DeGross, my administrative assistant
starting in 1977 and now the Production Editor of the MIS Quarterly. She
has edited and prepared for publication numerous publications in the IS
field including several of my books, ICIS proceedings since 1987, several
IFIP WG8.2 conference proceedings, and the MIS faculty directory. She
has been a resource for the field.
Early Academic Innovators in Information Systems
as an Academic Discipline
This list is representative and illustrative; it is certainly not exhaustive.
• J. Daniel Couger from the US (LEO). He was one of the important
providers of information on information systems education. He traveled
the world to communicate on IS curricula and his own research.
• Börje Langefors from Sweden (LEO). First information systems professor in Sweden. He not only contributed to the early conceptual literature; he also was doctoral supervisor for many of the important contributors in the field in Scandinavia, including Mats Lundeberg.
• Enid Mumford from the UK (LEO). She brought to the information
systems field the perspectives of organization behavior and especially a
socio-technical view. Her participation in IFIP WG8.2 gave international exposure to this approach to system design.
Davis
283
• Jay F. Nunamaker Jr. from the US (LEO). Innovator in system development methodologies, decision support systems, group systems, and
so forth. Leader in model curriculum development.
There were other early, important leaders in the development of the international academic discipline of information systems. I mention a few who
influenced me. Many have been recognized as AIS Fellows for their contributions to the development of the discipline.
• Niels Bjørn-Andersen of Denmark (AIS Fellow). Active in Denmark,
Europe, and internationally (brought ICIS to Europe). Second president
of AIS (1996)
• Pentti Kerola of Finland. Active in building the Scandinavian contributions and building the international community.
• William King of the US (AIS Fellow). Participant in most of the
important developments in the field. Editor-in-Chief of the MIS Quarterly (1983-1985), and principal architect and first president of AIS
(1995).
• Ephraim McLean of the US (AIS Fellow). Active in building the field
through ICIS and AIS. Currently Executive Director of AIS.
• Richard O. Mason from the US (LEO). A scholar with a broad background, he has added to the intellectual quality of the discourse about
information systems.
Researchers in Other Disciplines Who Contributed to the
Intellectual Foundation of Information Systems
Some not in the information systems academic discipline who directly
contributed to its intellectual foundations:
• Robert N. Anthony from the US. His 1965 Harvard University Press
monograph, Planning and Control Systems: A Framework for Analysis
(1965), was one of the most cited publications in the early MIS literature because it provided a basis for the structure of an organization
information system.
• Peter Checkland from the UK. His work on a soft systems approach
made him one of the most cited authors (Systems Thinking, Systems
Practice, 1981) in the early European MIS literature.
284
Exploring Patterns in Information Management
• C. West Churchman from the US (LEO). He was one of those who
clearly laid out the systems approach that underlies the systems work of
information systems designers and developers.
• Herbert A. Simon from the US (Nobel Prize). He contributed to system
concepts (The Sciences of the Artificial, 1969), cognitive science, artificial intelligence, and administrative science. He was perhaps the most
important conceptual person for decision-making concepts that provided a basis for systems to support decision making in organizations
(The New Science of Management Decision, 1960).
I could mention many others who were pioneers in building the discipline
(in the US: McFarlan, McKenney, Rockart, Scott Morton, Emery, Kriebel,
Zmud, Teichroew, Keen, Lucas, etc.) and some of the early doctorates
from Minnesota (Benbasat, Ives, Olson, Weber, etc.), but the above list
illustrates the diverse group from several countries who have built the current international discipline of information systems. I have tended to list
the “old timers”, so many not-so-old leaders and recent contributors are
not mentioned.
Critical Development 5: Locating the IFIP TC8
Working Conferences Internationally Including the
Manchester Conference Sponsored by WG8.2
I have noted the importance of IFIP in nurturing the emerging field of
information systems by the technical committee on information systems.
The working group conferences became a vehicle for building an international network of scholars, both by the subjects of the conferences and the
locations. This has been especially true of working group 8.2 on information systems and organizations. It is the group I worked with most, so my
view is biased. This group now has an equal number of European and
North American members. The conference venues rotate in order to
involve more researchers.
A very important conference in building the international community was
the IFIP WG8.2 1984 Manchester Conference on information systems
research methods (E. Mumford, R. Hirschheim, G. Fitzgerald, and T.
Wood-Harper, eds. Research Methods in Information Systems, North
Holland, Amsterdam, 1985). This conference was a landmark. The plan is
to have a second Manchester conference in 2004 with new proceedings
plus a reprint of the 1985 book.
Davis
285
The reason I count this conference as very important is its role in opening
up the discussion of the different research paradigms. Most of the
researchers in North America at that time tended to emphasize a positivist
approach to research with experiments, surveys, hypothesis testing, and so
forth. Many of the Europeans were doing post-positivist, interpretive
research. The conference opened the minds of many of the conferees and
helped open the field of information systems to a variety of research paradigms. Currently, there is reasonable acceptance of the following:
• Positivist, hypothesis testing, data-based research
• Interpretive research including research based on case studies
• Design science research
The IS research literature clearly defines the first two; the third is less well
defined. Design science research (the term used by Smith and March) is
based on the research paradigms of engineering and Computer Science. In
design science, designing and building a new, novel artifact such as a
computer application program, development methodology, or model is a
contribution to knowledge. In general, information systems research publications have expected that an artifact will not only have been built but
will also be tested to demonstrate proof of concept or value of the artifact.
Critical Development 6:
The Founding of the International Conference on
Information Systems (ICIS)
As mentioned previously, early researchers in information systems had
disciplines to which they belonged. Their conferences often provided
opportunities to present information systems research. This was especially
true of management science, operations research, and decision sciences.
The IFIP working groups on information systems focused on information
systems but tended to be around narrow topics. There was no general,
well-accepted, high quality information systems conference.
The first Conference on Information Systems (later renamed as the International Conference on Information Systems or ICIS) was held in 1980 in
Philadelphia (hosted by the Wharton School at the University of Pennsylvania). The second was held in Boston hosted by Harvard and MIT. A
major sponsor was the Society for Information Management, a society for
CIOs. The conference included a doctoral consortium. ICIS began as a
North American conference but grew quickly to a high quality international conference. It was held in Copenhagen in 1990 and has been held
286
Exploring Patterns in Information Management
four times outside North America in the past eight years. A major feature
is a high quality, invitational doctoral consortium with a mix of doctoral
students from different countries.
ICIS is high quality based on acceptance rates of about 15 percent. Printed
proceedings were produced from 1980 until 2000 and on CD-Rom from
1996 through 2000. Starting with 2001, conference proceedings are only
available online. Searchable past proceedings are available to all members
of AIS from www.aisnet.org.
There has existed a very open attitude at ICIS to subgroups within the
field. Several subgroups hold conferences immediately preceding or
immediately following ICIS. Examples are the Workshop on Information
System Economics (WISE), the Workshop on Information Technology
Systems (WITS), IFIP WG8.2, and several others.
Critical Development 7: The Founding of the
Association for Information Systems (AIS) with an
International Governance Structure
From the time of the first ICIS in 1980, there had been discussion of a new
international organization devoted exclusively to the academic field of
information systems. A poll of those attending ICIS showed that academics were about evenly split on the issue. It became more and more evident
that the lack of a single organization resulted in a lack of a strong voice in
matters affecting the field.
The need for such an organization was first spelled out in a March 1993
editorial in the MIS Quarterly. The editorial was co-authored by the current and four past editors-in-chief (Gary Dickson, William King, Warren
McFarlan, James Emery and Blake Ives). Bill King was the leader among
many key persons who helped in establishing the Association for Information Systems. It was formally established in 1995 with Bill as its first
president. The governance structure was designed to create a truly international organization. The position of president rotates among three regions:
Americas, Europe-Africa, and Asia Pacific Area. The term of presidents is
now from mid-year to mid-year. The presidents since inception have all
been leaders in the field:
1995
1996
1997
Bill King
Niels Bjørn-Andersen
Ron Weber
Americas
Europe-Africa
Asia Pacific
Davis
1998
1999
2000-2001
2001-2002
2002-2003
2003-2004
Gordon Davis
Robert Galliers
Michael Vitale
Blake Ives
Philip Ein-Dor
K.K. Wei
287
Americas
Europe-Africa
Asia Pacific
Americas
Europe-Africa
Asia Pacific
Since its inception, AIS has grown to include close to 50 percent of faculty
members worldwide (I estimate 6,000 IS faculty worldwide). Attendance
at ICIS is 1,000 to 1,200 indicating about 15 percent of faculty attend the
annual conference. This represents significant participation.
AIS has proved to be all that those of us who promoted its founding hoped
it would be. It has allowed the field to concentrate and rationalize many of
its resources. There has been an amalgamation of ICIS into AIS. It has
taken over responsibility for preexisting assets of the field such as the
Directories of IS Faculty, the past proceedings of ICIS, doctoral dissertation lists, survey of salaries for new hires, etc. It has created chapters and
special interest groups. It maintains loose ties with many conferences and
organizations that existed prior to its formation. AIS provides sponsorship
support and doctoral consortia support for the three regional IS conferences.
One of the issues in creating AIS was that of a journal. Given the existing
journals, AIS decided to create two e-journals: one was to contain a variety
of communications about pedagogy, curriculum, and issues in the field
(Communications of the AIS) and the other to be a high quality academic ejournal (Journal of the AIS). Both are operating. With respect to print journals, AIS adopted an interesting strategy. It first offered a choice of a journal from a short list of IS research journals; this was changed to offer
members discounted subscriptions to many of the top-rated IS research
journals published in both North America and Europe.
The discounts for journals continue as a benefit of AIS membership; a new
initiative was to offer electronic access to one print journal as a part of
membership. Since only one of the top three print journals was published
by an academic institution (MIS Quarterly), a proposal was made to partner with the MIS Quarterly and offer members online searchable access to
current and past issues. A governance structure was established to allow
AIS to have some influence on the MIS Quarterly.
Information systems as an academic discipline clearly began in the developed countries. Many in the field have been concerned about reaching out
to developing countries. IFIP has sponsored conferences in developing
288
Exploring Patterns in Information Management
countries. AIS has initiated programs to make conferences available and
less costly to faculty from developing countries. Since the cost of journals
is a major impediment to developing countries, AIS has an outreach program that provides access to its e-journals, its proceedings, and the MIS
Quarterly at a very nominal cost.
Summary and Conclusions
It has been a very interesting saga of development of an international academic discipline of information systems. There is much yet to do; there are
some countries that are under represented. However, the momentum has
been established. There is an incredibly high rate of participation in AIS,
the international academic organization. Members have access to a menu
of high quality, useful resources.
It has taken almost 50 years from the first business use of computers to
reach this point. However, 1954 is probably not the starting point, since
business use of processing devices was not interesting for research until
systems became more complex and management of the systems required
special skills. Therefore, I count the beginning of the journey from the
mid-1960s with the 1964 IBM System/360, the appointment of Langefors
as Professor of Information Systems in 1965, the first academic program at
Minnesota in 1968, or from the widely used MIS conceptual foundations
book in 1974. Others may count from the first ICIS in 1980. Depending on
when one starts counting, it has taken anywhere from 23 to 39 years for us
to come from a fragmented group to a fairly cohesive international field.
Issues that remain revolve around the boundaries of the field. Some scholars argue we will continue to find interesting research opportunities at the
intersection of information technology and other fields; other scholars say
we work at the boundary only if the resulting research can speak to issues
in the field. Still others, such as Weber, argue we need to pay more attention to the core of the field (that which is not “owned” by any other discipline). Perhaps there is a middle position: pay more attention to the unique
aspects but feel free to work at the intersection with any other field as long
as the result can be applied to the design, use, and management of information systems.
Coming back to the occasion that prompted this story of the journey of the
field (and my journey along with it), the Swedish participants in the journey have had significant impact. Mats Lundeberg has been one of these.
He has been an active participant in IFIP TC8, ICIS, AIS, the MIS Quar-
Davis
289
terly, and other developments I haven’t remembered. He has helped his
doctoral students to be part of the worldwide community of scholars. His
contributions are well worthy of celebration on this occasion.
References
History of the IS Field
Culnan, M.J. (1986) “The Intellectual Development of Management Information
Systems, 1972-1982: A Co-Citation Analysis”, Management Science, Vol. 33,
No. 5, pp. 156-172.
Dickson, G.W. (1981) “Management Information Systems: Evolution and Status”,
in Yovits, M. (Ed.) Advances in Computers, Academic Press, Vol. 20, pp. 1-37.
Taxonomy for the IS Field
Ives, B., Hamilton, S., & Davis G.B. (1980) “A Framework for Research on Computer-Based Management Information Systems,” Management Science, Vol.
26, No. 9, pp. 910-934.
Mason, R.O. & Mitroff, I.I. (1973) “A Program for Research on Management
Information Systems”, Management Science, Vol. 19, No. 5, pp. 475-487.
Design Science in IS
March, S.T. & Smith, G.F. (1995) “Design and Natural Science Research on
Information Technology”, Decision Support Systems, Vol. 15, No. 4, pp. 251266.
The Core of the IS Field
Davis, G.B. (2000) “Information Systems: Conceptual Foundations: Looking
Backward and Forward,” in Baskerville, R., Stage, J., & DeGross, J.I. (Eds)
Organizational and Social Perspectives on Information Technology, Kluwer
Academic Publishers, Boston, Massachusetts.
Davis, G.B. & Olson, M.H. (1974, second edition 1985) Management Information
Systems: Conceptual Foundations, Structure, and Development, McGraw-Hill,
New York.
Langefors, B. (1966, fourth edition 1973) Theoretical Analysis of Information
Systems, Studentlitteratur, Lund, & Auerbach, Philadelphia, PA.
Weber, R. (1987) “Toward a Theory of Artifacts: A Paradigmatic Base for Information Systems Research”, Journal of Information Systems, Vol. 1, No. 2, pp.
3-19.
290
Exploring Patterns in Information Management
Other References in This Chapter
Anthony, R.A. (1965), Planning and Control Systems: A Framework for Analysis,
Harvard University Press, Boston, Massachusetts.
Checkland, P. B. (1981) Systems Thinking, Systems Practice, Wiley, Chichester.
Davis, G.B. (1965) An Introduction to Electronic Computers, McGraw-Hill, New
York.
Davis, G.B. (1969) Computer Data Processing, McGraw-Hill, New York.
Lundeberg, M., Goldkuhl, G. & Nilsson, A.G. (1981) Information Systems Development: A Systematic Approach, Prentice-Hall, Englewood Cliffs, New Jersey.
Leavitt H.J. & Whisler, T.L. (1958) “Management in the 1980s”, Harvard Business Review, Vol. 36, No. 6, pp. 41-48.
Mumford, E., Hirschheim, R., Fitzgerald, G. & Wood-Harper T., (eds.) (1985)
Research Methods in Information Systems, North Holland, Amsterdam.
Simon, H.A. (1960, revised edition 1977) The New Science of Management Decision, Prentice-Hall, Englewood Cliffs, New Jersey.
Simon, H.A. (1969, third edition 1996) The Sciences of the Artificial, MIT Press,
Cambridge, Massachusetts.
— 17 —
Users Matter – A Long Term
Perspective
Rolf Høyer
Serving the Information Systems Users
During most of his long and productive research career, Mats Lundeberg
has served as a faithful gardener in the field of providing information system users with concepts, methodology and tools for analysis and design of
information systems. Nobody will refute that this gardener, indeed, has
presented significant contributions to the universal heritage of useful
knowledge and practical tools.
One may suggest that the underlying objective for this endeavour, above
all, has been to enhance the control information users can exercise upon
information technology by means of special user-adapted tools for analysis
and design. This activity has truly been a universal undertaking. Over the
years, the numbers of published titles by numerous authors suggesting still
more efficient tools may surely be counted in hundreds.
This development appears as a quest to provide information systems
users with intermediate technology between computers and programming
technology on one side, and human conceptions of the requirements to
new systems on the other. The core questions to be addressed has been
what systems should do, and above all, what information they should
provide.
An underlying assumption for this search for intermediate technologies is
that when applying such tools, information systems will be more useful
by serving more efficiently the entire work organisation in which they
are embedded. This fundamental assertion was initially advocated
already by Mats Lundeberg’s predecessor and teacher, Börje Langefors,
in his seminal work titled “Theoretical Analysis of information Systems”, first published in 1966 (Langefors, 1966). However, the importance of user involvement was developed in more detail in his following
292
Exploring Patterns in Information Management
book on systems for corporate control, published in 1968 (Langefors,
1968). This basic research laid the foundation for the subsequent research
contributions by Mats Lundeberg. From the onset his work aimed at
developing tools that might enhance and facilitate user involvement and
control in efficient ways, but at the same time also resulted in stringent,
perfectly documented models that directly might serve as blueprint to be
handed over to system engineers for subsequent programming and data
base construction.
The ideas cited above, concerning the need for user involvement and user
control of systems development, thus emerged remarkably early in the
history of management information systems. In the context of this very
special book, it may be appropriate to explore and sum up how these
ideas and imperatives were conceived in the first place, and gradually
came into use. Above all, it may be interesting to explore to what degree
users really have attained control of the production process of new systems. Are information systems users today really the kings and masters
of the development? And finally, are theories and methodologies developed during the last 40 years still relevant today, when applying a technology tremendously different from the time Mats Lundeberg started his
research career.
Two factors are of pivotal importance in shedding some light on the development. First and foremost, the technology itself is the driving force.
Information technology today has drifted far beyond the wildest fantasies
of the 1960s, facilitating development of systems that have changed
industry and our daily lives. Secondly, the user community and the role of
users have changed in significant ways. As will be shown, the concept of
“users” today has a much more pluralistic meaning than in the early history. The technological development itself can, to a very large degree
explain the changes in the role of users. But changes in the overall economy, new patterns of management practice, and general user competence
also play important roles.
This has led to a fundamental change in the requirements for intermediate
technologies, and in the conception of what a user is and her potential for
exercising control of an extremely pervasive technology which today constitutes an important part of everybody’s daily environment. But at the
same time, although there are enormous changes, in many ways some of
the most fundamental requirements to intermediate technology and concepts remain the same, because the human character of users, their cognition, remains unchanged.
Høyer
293
The Emergence of Concerns for Users
In order to shed some light upon the growth of the early, and pioneering
interest in the role of users, and development of theory and methodologies,
we will start with a short glance at the ways information systems first were
introduced in businesses and public agencies.
In the very beginning of the era of information systems design, user participation was definitely not an important issue. Information systems were
labelled “data processing systems”; being regarded as technical artefacts.
They were assumed to be designed and run by technicians. Their creators
were generally great enthusiasts, who, regarded as wizards, were able to
instruct computers to perform rather simple, well-defined data processing
tasks. Although simple, the construction appeared unintelligible. The
designers were those possessing the technical knowledge and skills to
master the technology. Most often their tasks turned out to be impossible
to combine with regular operative work within the organisation; hence
their profession quickly became totally devoted to the task of programming computers. The data system specialist’s role, as the master of information systems, was quickly born.
In this stage, the process of planning and definition of requirements was in
fact in most cases rather straightforward. Most computer based systems
were automated, machine based versions of existing manual and partly
mechanical systems, often appearing as sets of loosely coupled routines.
To a large extent, system development was an automatisation process,
often labelled ‘automatic data processing’. The designers simply would
copy the processes in use. Actually, methods labelled ‘systems mapping’
were common tools for the designers.
In system development processes of this character, user participation
would be limited to consultations about technical details of existing routines, and possible suggestions for minor improvements. On the other
hand, users may be said to have had a decisive influence on the systems
design, as they originally had participated in the gradual development of
the manual systems, so why should they bother by fumbling with the
automated versions?
When the systems portfolio grew in size, it became obvious that this was a
cumbersome way of developing systems. Without efforts for over-all system planning, one quickly ended up with totally unintegrated systems.
Furthermore, it became clear that investments in systems design had
unsatisfactory pay-offs. In many cases it was reported that the rate of
return of the investments was nil or even negative. It also became clear
294
Exploring Patterns in Information Management
that this kind of systems development incurred a very low level of real
innovation. The automated systems did not improve the operations of the
enterprise in significant ways.
Radical New Ideas Emerge – The Foundation Is Laid
This early development led to an interest in more sophisticated methods
for systems analysis and design. In Scandinavia, Langefors and Lundeberg
were the leading pioneers to address this challenge by launching fruitful
concepts and frameworks for subsequent development of useful methodologies.
Firstly, it seemed necessary to develop tools that might allow logical modelling of the information systems before any investment in the actual
design was made. Just as very few people would think of building a house
without some kind of architectural drawing, one realised that such a procedure would be required also for information systems. The merit of such
models would, above all, be that alternative designs might be suggested,
modified and rejected without the previous rigidity and prohibitive costs of
changing the finished, implemented systems.
Another important advantage of using logical models would be that they
might employ very simple and user-friendly symbolism, and thus allow
people without any special competence in programming or in the technical
aspects of data systems design to be involved in constructive ways in the
systems development process.
Börje Langefors very early on suggested and emphasised such general
principles as guidelines for method development. He also formulated several, explicit fundamental principles for systems analysis and design,
among which was the important imperative that system design methodology should address two different questions: WHAT systems should do, i.e.
which information should go in and out of the systems; and secondly,
HOW the systems should be constructed.
Langefors’ contribution was above all a of theoretical nature. Although it
was obvious that he had delivered an important foundation for a subsequent development of practical methods, this huge development task still
remained almost untouched. Very early in his career, while still a doctoral
student, Mats Lundeberg took up this challenge. In 1974, after several
years of research efforts, he published the first work in literature which
turned the basic ideas of Langefors into a consistent set of methods for
systems analysis and design (Lundeberg and Andersen, 1974). The book
Høyer
295
was titled “Informationsanalys” (Information Analysis) and was immediately incorporated into curricula in Scandinavian higher education. Gradually it also was adopted in industry, however frequently in modified and
simplified versions, adapted to perceived local needs.
It is important to acknowledge that this fundamental theoretical and practical oriented research was guided by a firm belief that in order to arrive at
efficient systems design, it was necessary to mobilise the information
users, giving them as much control as possible over the design process.
Lundeberg’s contribution was above all to harness the users with tools and
concepts opening up for such control.
One may assert that the important contributions of Langefors and Lundeberg, very crudely sketched out above, were rooted in a pragmatic ideology. Concepts and methods were suggested in order to help industry and
public agencies in the process of exploiting the new opportunities for
doing their business in more efficient ways, made feasible by the emerging
information technology. In order to arrive at a more constructive and innovative system design process, a strong user involvement was assumed to
play a decisive role.
Broadening the Perspective
However, the research society’s concern for users did not stop here. During the 1970s other researchers expanded the ideological foundation, by
pointing out that systems development ought to be seen as an organisation
development process. It was even advocated that systems development
might be regarded as an integral part of efforts for contributing to industrial democracy. These ideas were particularly developed in the Scandinavian countries, giving rise to the development of a set of ideas, concepts
and methods labelled as the Scandinavian tradition. This expanded ideological orientation served even more strongly to emphasise and magnify
the concern for the role of information systems users.
While reluctant to embrace industrial democracy and related ideas, the
organisation development challenge was incorporated constructively into
Lundeberg’s following theoretical works during the 1980s and 90s. Gradually his theories and models emphasised more and more that systems
development also should be conceived as people development, i.e. that
knowledge about the business situation and work organisation among the
people involved was an asset which ought to be built upon and improved
as an integral part of the system development process. This finally was
296
Exploring Patterns in Information Management
brought forward into a consistent methodology, as it appeared in the
important book “Handling Change Processes – A Systems Approach”,
published in 1993 (Lundeberg, 1993).
In this way Lundeberg’s work approached, and to some degree coincided,
with the rapid emerging new area in management literature labelled
knowledge management. This parallel development of organisation theory
was also oriented towards tendering the competence development of the
individuals, or the information users, as the individuals were labelled in the
information system world.
Technological Development Rolls on, Inducing
Changes in the User Role
During this mature period of method development the role of users in the
realm of technology gradually changed. Facilitated by advances in programming technology and continually improved hardware resources,
application systems got bigger and bigger. System architecture crossed
hierarchical structures in the organisation, involving the responsibilities of
several departments and divisions. Since this also implied that large numbers of potential system users were affected, the role of participating users
in the development process had to be dramatically simplified.
Above all, the increasing complexity of technology limited and crippled
the possibilities for realising the ideas of systems development as a vehicle
for industrial democracy. During the 1980s Scandinavian trade unions had
forced the introduction of special agreements to secure employees the right
to participate in system projects, but also to influence systems properties.
Norwegian legislation on the work environment (1977) actually introduced
a legal right for all the individuals who were affected by changes in their
work environment, clearly including systems changes, to have a say in the
change process. However, the changing character of further development
of large-scale systems severely limited the possibilities for realising the
intentions of the legislation and trade union agreements. At best, elected
representatives might act as spokesmen for the user community. Experience quickly showed that this turned out to be a very inefficient way of
realising the intentions of comprehensive user participation. Representatives were appointed on behalf of very many fellow employees, and consequently had great problems in communicating with their constituency,
the individual users. Another obstacle to proper functioning of the representative system was that user representatives either were not allowed time
and resources to function in other ways than formal hostages, or, if given
Høyer
297
resources, they quickly were adopted into roles as regular systems professionals.
It has also been suggested that the meticulously engineered methodologies,
with heavy emphasis upon strict and consistent formalism, simply
appeared far too complex for the user society, especially when it involved
large system projects. Undoubtedly, the mind-set of the developers of the
(presumably) user-friendly methods was probably quite different from that
of the average information system user. The overwhelming complexity and
strict formalism simply led to a situation of methodological over-kill.
At present, it seems that the whole idea of an ideologically rooted user
democracy, as introduced some twenty years ago, is fading away. Trade
unions are presently, wisely enough, more preoccupied with major, structural changes in the enterprise. Present experiences indicate that even in
such matters, unions seem to accept a rather moderate level of influence,
especially in times when the economy slows down. However, a pragmatically based user control is more important than ever. To the extent that
user involvement may enhance the quality of the final systems, or provide
opportunities for real innovation in the way the enterprise operates, user
involvement is still of crucial importance. It certainly is a paradox that due
to the ever increasing magnitude and complexity of systems, it seems as if
we are slowly returning to the early days of data processing, when only
specially assigned and competent persons were the only ones being given
responsibility for design and development of systems. Such persons may
also be external to the enterprise, acting as consultants on long time contracts, alien to the internal users. This may becomr still worse when the
company decides to acquire and use standard applications which are totally
foreign to the user environment. In this way, information systems tend to
have a character of standard commodities, thereby totally transforming and
potentially degrading the role of the information systems users.
The Arrival of Personal Computing
– Emergence of Systems Anarchy?
This description of current development may at first glance lead to a very
pessimistic conception of the current and future role of information users.
Are they really outmanoeuvred by the technological development? That is
certainly not the case. The rapid development of personal computers and
programming tools supporting personal computing, has led to a situation
where information users are empowered to design their own systems. By
means of templates and macros, even users in large and well-structured
298
Exploring Patterns in Information Management
bureaucracies develop personal routines within the standard packages for
word processing. They may customise their own versions of e-mail systems, and an ever increasing amount of analysis of business problems are
performed by means of spread-sheets using the extremely powerful capabilities of commonly available tools as Excel. It is also interesting to note
that both internal and external routines for business communication may
be improvised and sometimes formalised by the individual employee, and
is not limited to strict formalism and communication channels presented
by large, standardised systems that are part of the corporate infrastructure.
In many ways, this may seem as if a state of system anarchy is emerging.
To a large extent that is true. However, it may be a well-functioning
organisational system, directly serving the individual users. Numerous
small systems are born out of user needs, and die when the owner leaves
the job or the organisation. However, the costs and time required for
designing such routines, and even small systems, are so marginal, that this
practice may be superior to the complex and rigid institutional alternatives,
incurring great costs.
In such situations, the need for traditional system development methods
simply evaporates. There is no need for a repertoire of tools for system
analysis and for abstract models of future systems. In most cases, intuitively paper sketches will do reasonably well. When routines and small
systems grow and need revisions, they may easily be changed, and even
scraped, because the efforts of redesign are very moderate.
It is a paradox that this situation has a striking similarity with a particular
system design methodology suggested by the late Swedish professor Staffan Persson more than 20 years ago. Labelled “system design by system
sketches”, he demonstrated that non-trivial systems might be analysed and
designed through direct implementation by using the then extremely powerful programming language APL and awkward first generation personal
computers. For many reasons, this methodology never really was implemented by industry in general. It was obviously a generation ahead of its
time, because APL-programming skills were not very common. Today, an
increasing amount of systems development is in principle carried out this
way, however with a far superior technology and as supplementary systems to the main data processing infrastructure, having been designed by
means of traditional and complex methods.
Thus, one may conclude that within the company, the role of information
system users has changed significantly. Today, all are potential users, and
increasingly many may command a reasonable control of their immediate
Høyer
299
systems. Furthermore, an increasing fraction of employees is expected to
engage into various forms of personal system design, especially in the
emerging knowledge industry.
Emergence of a New Breed
of Information System Users
In the discussion above, we have applied a traditional conception of an
information system user, namely a person employed by a work organisation, rendering work and services to a principal. In order to serve the
organisation in an efficient way, the user is provided information delivered
by formalised systems, to which the user also is assumed to present information. We may name those internal users. However, management information systems always have served people external to the organisation,
both in the role as providers and receivers of information. It is interesting
to note that in the main bulk of literature on information systems, such
external users, until recently, have not been given much attention. They
have seldom been regarded as system users worth mentioning.
Today however, such users play an ever increasingly important role to the
organisation. The reason for this is that more and more of the economy
consists of businesses producing services where information is an important part of the service rendered. There is also a strong growth of businesses whose product is just information. In many cases the information
put on the market also has to be personalised according to specific
demands. Hence, the quality of the information itself and the way it is presented is a crucial property of the products. Consequently, issues related to
this changed industrial structure turn out to be of increasingly strategic
importance.
This is likewise true for public service organisations. The level of information interchange between public agencies and individuals as well as
businesses is steadily growing, creating a situation where the quality of
public service to an increasing degree is a question of information system
quality.
This trend has been tremendously accelerated by the fast emergence of
Internet based systems. Already the first generation of companies offering
services and products delivered by Internet based systems, discovered that
the success of their businesses and the competition power, to a large
extent, was determined not only by the properties of the objects or services
presented on the market. The ways the offerings were made, and how they
300
Exploring Patterns in Information Management
were presented and how easily the customer could interfere with the system, were also all of crucial importance.
Gradually, such observations have led to an emerging research interest in
improving the quality of the interaction between information systems and
customers. Initially, this research was rooted in marketing environments,
but gradually it also spread into the information system world. Only a few
years after this interest arose, we have seen a rapid development of more or
less powerful concepts and even emerging methodologies for designing so
called Customer Relationship Management Systems (CRM). This coincides with a more encompassing development of new ways of doing business, facilitated by Internet technology; the so-called e-business, that
already a few years after the term was coined appeared in the title of
numerous text-books, and rapidly penetrated the marketplace.
The e-business world is realised by information systems. These systems
connect customers in the role of information system users with the enterprise or the public agency. In the same way as for internal system users, it
is obvious that systems supporting e-business must meet the demands and
behavioural peculiarities of external users. This challenge however, represents quite a different task. While internal users were reasonably few and
were generally available not only for consultations, but also for mobilisation into systems planning, the external users are generally of a significantly different character.
In addition to this, the users are much more numerous. E-commerce companies may have thousands of customers; the largest global companies
count them in millions. Furthermore they are generally far more heterogeneous. They may be recruited from any part of society, and have most
divergent personal traits. The old concept of designing a user profile is in
many cases far more difficult than for internal system users, because the
latter group normally is well recorded.
Breakdown of Existing Paradigms
We are still living in some stage of first generation of e-commerce and
Internet based systems. This stage is characterised by enormously varying
quality of the information systems. Not only regarding technical systems
quality, but above all in the perceived quality of the systems as experienced by the external system user. Almost everyday we encounter ridiculous examples of unprofessional, clumsy dialogues, lack of overview, confusing instructions etc. For example; regular, state-of-the-art Norwegian
Høyer
301
personal banking systems request users to punch in up to 20-digit long,
continuous numerical strings as customer identification numbers for each
transaction. Most of us, in the role of system users, daily experience such
lack of understanding of elementary user problems.
Because the system quality, as experienced by the customers, plays a decisive role for the business success of e-commerce system; concerns for the
user may play an even more important role in systems development than
ever before, as the number of external system users increases. Obviously,
this problem has to be approached in quite different ways as it is for internal users. The emergence of research based methodology and insight will
be of crucial importance, representing a great challenge for the research
society.
It is also still a question of ideology. However, it is no longer an issue of
humanism, industrial democracy and trade union established rights. Now
service ideology is simply required as the dominant mind-set. For the systems designers it is important to understand and accept the new power base
of the systems users. Previously, designers and managers more or less
reluctantly might have given due consideration to the needs of internal
users and engage them in various levels of constructive co-operation. Now
the situation is turned around. The external system user is the supreme
King! If the customer doesn’t like the system, she is free to use the exit
option, that is leave the system – maybe forever. Thus, awareness and concern for customers is probably the main virtue and competence of the systems designers. This may be difficult to harmonise with the current mindset of many experienced systems designers, recruited from the internal
systems world. Hence, we may need a new breed of designers, raising a
great challenge for educators.
Experiences with first generation Internet based systems clearly demonstrate the inability of traditional systems designers to establish userfriendly interfaces and efficient communication patterns with users. In the
same way as one assumed that internal systems might be better designed
by mobilising the insight, experience and talent of end users, there is good
reason to conclude that external users may also contribute significantly to
the quality of the systems they are invited to use. This calls for methods for
mobilising the external users.
Many public service internet-based systems don’t allow the users the exit
option in the same way as they do in commercial systems. This is the case
when all citizens are forced by law to interact with systems, for example
by filing a mandatory income tax form. This of course reduces the user’s
302
Exploring Patterns in Information Management
potential for exercising power and influencing the systems. However,
today most public service systems that appear in an Internet version, are
required by law to offer a manual option. This gives the user a possibility
for rejecting the ‘modern’, but clumsy version. If such rejection is widespread, it will obviously represent a very significant feedback to the host
agency, indicating that the system quality is inferior, and calling for revision.
Meeting the New Challenges
How then may user viewpoints be elicited and channelled into the host
organisation? There are several possibilities. Some may be derived from
traditional marketing and product development. In this field, user panels
are frequently employed to solicit comment upon products and marketing
campaigns. User panels are carefully chosen subsets of the market being
addressed. Each subset is then invited to comment upon and give opinions upon marketing messages, as well as upon the products themselves.
Obviously, this has to be managed in systematic and detailed ways. Just
asking for comments from customers in general, in the form of putting
questions like “What do you think about our new web-pages?” will generally not channel much valuable information at all. However, in spite of
this, such more or less naïve invitations appear regularly in first generation
systems.
Many systems present FAQ-lists (Frequently Asked Questions) that are
assumed to present useful information for systems users, especially for
navigation in the maze of web pages that the users have to negotiate. The
bulk of FAQs usually represent useful information, but it is frequently
obvious that the help provided by individual FAQs simply works as a
cover for clumsy system design. When complaints or panel-derived suggestions are remedied by extensive lists of FAQs, the designers should also
rethink that very part of the system, looking for possible modifications
which simply would eliminate problems and confusion in the user community.
The emerging literature on design of Internet based systems has recently
started to pay attention to such matters. However, in the same way that a
need rose for basic concepts and methodology when the first generation of
data processing systems first emerged, a similar development of useful
tools is clearly needed as Internet based systems are getting more and more
Høyer
303
common. This seems to be a major challenge for current information systems research.
It is interesting to note that the same basic problems and dilemmas that
confronted systems designers some 30 or 40 years ago are conceptually
still there. In many ways they seem even more important to handle today.
Never before has it been more important to close the hiatus between system designers and system users. The reason for this is that today users are
“out there”; they are poorly defined and hard to communicate with. At the
same time they are of great strategic importance for the company because
they enjoy the powerful exit option in the transaction process.
A Heritage to Build Upon
One may conclude that the question of systems quality still has two different dimensions: 1. the technical quality of the implemented system; 2. the
quality of the system as experienced by the system user. In this way, it
seems that Langefors’ fundamental theorem on the HOW/WHAT dichotomy is seminal, and important as ever. Likewise, one may conclude that
further efforts for development of efficient methods for systems analysis
and design should be welcomed. However, as we have emphasised here,
they should reflect and be adapted to the current usage of technology.
Methods should be built upon grounded theory which helps us in understanding who the users are, what their true needs for information are, and
finally how they feel about using the information technology they meet in
their daily lives.
This is equally relevant for all categories of system users, whether they
arise as internal or external users.
The pioneering work of Mats Lundeberg back in the 1970s to develop the
first generation of comprehensive, useful tools and methods, has for many
years inspired new generations of researchers to invent and redesign methods and concepts adapted to contemporary technology. While the tools
presented for information analysis in 1974 may appear less attractive for
the common users, and hence of reduced importance today, Lundeberg’s
more recent theories for handling change processes and related organisational issues offer contributions to a fruitful knowledge base for further
endeavours in the new generation of information system researchers.
Because information technology continually changes our institutions and
our daily environment, the need to harness users with efficient intermediate technology is as urgent as ever. The garden first founded by Lundeberg
304
Exploring Patterns in Information Management
and his contemporary colleagues clearly needs to be maintained and cultivated by new generations of ambitious gardeners.
Because concern for user matters, obviously still matters.
References
Langefors, B. (1966). Theoretical Analysis of Information Systems, Studentlitteratur, Lund, Sweden.
Langefors, B. (1968). System för företagsstyrning {Systems for Corporate Control}, Studentlitteratur, Lund, Sweden.
Lundeberg, M. (1993) Handling Change Processes: A Systems Approach, Studentlitteratur, Lund, Sweden.
Lundeberg, M. & Andersen, E.S. (1974) Systemering – Informationsanalys {Systems Engineering – Information Analysis}, Studentlitteratur, Lund, Sweden.
— 18 —
Building and Testing Theory on
New Organizational Forms Enabled
by Information Technology
Allen S. Lee
Introduction
The advent of new information technologies has led some information
systems researchers to investigate the emergence of new organizational
forms. Often their usage of the term “virtual” serves to signal what they
see as new. Examples include virtual teams, virtual libraries, virtual markets, virtual communities, and virtual corporations. It is safe to say that any
particular virtual team, virtual library, virtual market, virtual community,
or virtual corporation did not exist twenty and even ten years ago; in this
sense, one can say that it is new. Yet this is different from saying that the
organizational form – of which a particular virtual team, virtual library,
virtual market, virtual community, or virtual corporation is an instantiation
– is new. This has led me to wonder: are “new” organizational forms necessarily new?
I am deliberately using the word instantiation with its database meaning: a
given database schema stays the same across time and across situations
while it is the data populating the schema, and not the schema itself, that
changes. The data populating a database schema at a single point in time is
an instantiation of the schema. Instantiations of the database schema come
and go. A recently appearing instantiation is what we would correctly perceive as new. On the other hand, the database schema or, in this analogy,
the organizational form, would stay the same. In this perspective, the
organizational form endures. If there is any merit to this line of thinking –
the thinking that organizations enabled by information technology do not
necessarily take new organizational forms, but are instantiations of old or
existing organizational forms – then we as scholars and practitioners can
enjoy a good measure of relief in realizing that what we have already
learned and theorized about organizations would still apply.
306
Exploring Patterns in Information Management
Often it seems that we as scholars and practitioners are witnessing a neverending onslaught of new technologies and organizational changes where
the result is that we feel both a sense of challenge (because there is so
much more to learn) and a sense of dread (because of the possibility that
we fail to learn enough to keep up with the onslaught of new technologies
and organizational changes). However, if it is only the instantiations of
organizational forms that are new and not the organizational forms themselves, then we can look to our past research and past experiences for lessons relevant to our understanding, managing, and theorizing about the
organizations we see today that are enabled by information technology.
In this essay, I will pursue this line of thinking to address these concerns:
building theory, what an organization is, what an information technology
is, what an information system is, and testing theory. The consideration of
whether new organizational forms are necessarily new compels a reexamination of these basic concerns.
Building Theory
The philosophy of science, history of science, and sociology of science
have offered numerous insights about how theorizing can, cannot, does,
and does not proceed. Reviewers and editors of my manuscript submissions to journals sometimes resist the insights that I take from these fields.
They and other scholars subscribe to a conception of science different from
the one I have learned from the philosophy, history, and sociology of science.
An aspect of science on which I disagree with many of my colleagues is
how to build a theory. The popular conception of how to build a theory
goes something like this: a researcher, who can be working alone, collects
data and then develops a theory based on the data. In this depiction of
building theory, data are the raw material and theory is the product. The
more data one collects, the better the resulting theory is. However, this
conception of how to build a theory is wrong because what it depicts is
infeasible. For an explanation of this, consider the following rows of data
(Figure 1):
2.1, 3.5, 4.3
1.9, 3.7, 4.1
2.2, 3.8, 4.4
Figure 1.
Lee
307
The data could be what a researcher inputs into a statistical software, such
as SAS, SPSS, Minitab, or even Excel, where each row represents a data
point. Next, consider the following theoretical propositions (Figure 2):
“EOU has a direct effect on U.”
“U has a direct effect on BI.”
“EOU has a direct effect on BI.”
Figure 2.
These are some propositions from the well-known technology acceptance
model, TAM. The variable U denotes the perceived usefulness of an information technology, the variable EOU denotes the information technology’s
perceived ease of use, and the variable BI refers to the behavioral intention
to use the information technology. If a TAM researcher were to give a
dataset (such as the one in Figure 1) and nothing else to another researcher,
could the latter researcher induce the TAM propositions from the data? In
other words, what mathematical, logical, or other formal procedure exists
that can transform the numbers in Figure 1 into the theoretical propositions in Figure 2?
The answer is that no such procedure exists. There is no way to induce,
generalize, or otherwise formulate theory from data or observations alone.
This illustration also shows that a researcher cannot even know what types
of data he or she will collect until he or she first has a good idea of what
the theory is in the first place. In a sense, the researcher creates the data by
instantiating the theory in this or that field setting, laboratory setting, or
sampling frame. Theory precedes data, whether the data are quantitative or
qualitative and whether the theory is positivist or interpretive. Analogously, in the world of information systems development, there is no
mathematical, logical, or other formal procedure by which to transform
data alone into a database schema.
If data come from theory, then where does a theory come from? In particular, where would a theory about new organizational forms come
from? To address this point, I choose my unit of analysis not to be the
individual researcher, but to be the community of scholars of which the
single researcher is a member. I frame any individual researcher as an
agent through whom the larger research community acts, where the
community infrastructure includes all the research theories, research
conventions, research journals, and other research entities that the com-
308
Exploring Patterns in Information Management
munity has developed accumulated. Any given researcher is not so much
an independent individual exercising free will in isolation as he or she is
an agent of this larger community. Any given researcher is the product of
the extensive socialization that she has experienced as a scholar, where
the socialization typically began in her doctoral program. There is also
the socialization reflected in the accolades that she has received from her
scientific community for her portfolio of publications. And there is the
socialization reinforced by the reputation that she has painstakingly built
up among her peers – a process involving ten, twenty, or more years as a
participant in career-shaping research institutions such as IFIP Working
Group 8.2, ICIS, ECIS, and the editorial boards of EJIS, MISQ, and
ISR.1
When a researcher is looking at what he or she assumes to be a new
organizational form enabled by information technology, does he formulate
a theory of new organizational forms based on his observations? Given the
impossibility of transforming data into theory, the answer is negative. The
researcher’s observations could very well be an input to his new theory,
but they would only be that: an input and, in my view, they would not be
determinative. Playing the major role would be the already existing body
of theory and research conventions that the larger research community has
built up over time. Through the individual researcher, this body of theory
and research conventions would be manifesting and instantiating itself
when he is observing one or another organization.
This line of thinking is not new. It is actually just an application of some
basic points about how to build a theory. I have developed my perspective on scientific research from the work of the historian of science,
Thomas S. Kuhn (1962, 1977). In my perspective, research is also a
social and political enterprise unfolding within its own contemporary and
historical context; a theory is not originated in the thought of an individual, much less an individual in isolation. What is called new theory is
normally old theory: in the research activity that Kuhn describes as
“normal science,” existing theory provides the lens through which a
given community of scientists makes observations in different empirical
settings. In the very arduous and demanding work of normal science, the
scientists see different empirical conditions across different empirical
1
The acronyms denote the International Federation for Information Processing,
Working Group 8.2; the International Conference on Information Systems, the
European Conference on Information Systems, the European Journal of Information Systems, Management Information Systems Quarterly, and Information Systems Research.
Lee
309
settings (such as different laboratory settings or different field settings).
The different empirical conditions challenge the existing theory by presenting puzzles that it cannot immediately or readily explain, whereupon
the community of scientists responds by articulating and refining the theory so that it eventually explains the puzzles. It is only in the situation
where a theory remains inconsistent with the evidence – despite the scientific community’s exhaustive attempts to refine and articulate the theory – that the scientific community replaces it with a new one through a
social and political process. This is the process that Kuhn calls “revolution science.” Revolutionary science is rare. Normal science characterizes most scientific activity.
This line of thinking about how to build a theory leads to the following
suggestion. When building a theory about new organizational forms, a
researcher should not focus first on collecting evidence about the organization(s) that manifest the apparently new organizational form(s), but
instead should look first to the body of current theory about organization
forms (i.e., existing theory pertaining to organizations that take forms
that are considered old or already known) and determine how well it fits.
If the evidence turns out to be consistent with current theory (i.e., if current theory can satisfactorily predict or explain the observed situation),
then there would be no need or justification for considering the organizational form to be new. If the evidence does not fit current theory, then the
researcher should initiate the process that Kuhn describes as “normal science” for refining current theory. The objective in normal science is to
refine or better articulate currently accepted theory so as make it consistent with the evidence. No theory is perfect and every theory can always
stand improvement. If no refinement is able to render the current theory
consistent with the evidence or if the current theory needs to be transformed so dramatically that it can no longer be considered the same theory, then the researcher would have justification either for concluding the
observed organizational form to be new. Moreover, in the end, it is not
the researcher alone but his or her scientific community that would
establish a new theory to be valid. Building a theory in this manner
would involve a process that is not only rational and cognitive, but also
social and political.
The basic points in my argument about how to build a theory pertain to
research in any discipline that aspires to be scientific. Aspects specific to
theorizing about organizations or organizational forms also need consideration.
310
Exploring Patterns in Information Management
What an Organization Is
Schools of business have paid much attention to a phenomenon that they
call “organizations.” Teaching and research in business schools deal with
business firms, which readily fall under the dictionary definition of the
word “organization.” Therefore it would logically follow that organizations, being their own category of phenomenon, would require their own
dedicated area of study, which in turn would mean that they also require
their own theories. Moreover, this conclusion would follow with greater
emphasis in situations involving new organizational forms that are seemingly appearing on the horizon. This, in a nutshell, is a conventional wisdom about organizational research.
I disagree with this instance of conventional wisdom. Consider what everyday people in the everyday world see in terms of their everyday common
sense. For instance, suppose that managers, executives, consultants, and
journalists see what they call “new types of firms” and “new kinds of
organizations.” Just because everyday people in everyday life see something that they consider to be new does not necessarily mean that scientific
theory should conceptualize the phenomenon in the same way.
Our research must, of course, account for the everyday meanings and beliefs
that everyday people have. Everyday meanings and beliefs play the role of
what the philosopher and phenomenologist Alfred Schutz (1962-66) calls
“first level constructs.” The theories that researchers create, following the
methodological rules of science, are what Schutz calls “second level constructs.” The second-level constructs making up a scientific theory need to
account for the first-level constructs of everyday understandings, but the
second-level constructs, in following the rules of science, need not be
beholden to, and may transcend, the everyday understandings of everyday
people in the everyday world. What everyday people, such as managers,
executives, consultants, and journalists see is not necessarily what scientific
researchers see. What is a “new” organizational form to a manager, executive, consultant, or a journalist need not necessarily be a “new” organizational form to a scientific researcher. What a native sees need not be what
the anthropologist sees. Indeed, for an illustration, I will momentarily
digress to an example involving ethnography. It is an illustration I use in a
doctoral seminar course that I teach annually.
In that course, my students and I cover numerous topics, one of which is
ethnography. I consider ethnography to be the most important qualitative
approach in business-school research. I assign to my students a short book
by Frederic O. Gearing (1970) that I first read when I was a doctoral stu-
Lee
311
dent more than 20 years ago. It is a book about the Fox Indians in Iowa,
which is a state in the central part of the United States. Gearing contrasts
and compares what he, from his perspective as a scientist, calls the white
man’s social structure and the Fox Indian’s social structure.
The Fox Indian’s social structure is different from the white man’s social
structure. What a white man calls “first cousins” is what a Fox Indian calls
“brothers and sisters.” This is because the white man’s social structure is
father-centric while the Fox Indian’s social structure is grandfather-centric.
In a society where the paternal grandfather provides the anchoring point in
the social structure, all the grandchildren bear the same relationship to him.
Hence, this calls for the same label to designate these grandchildren positions in the social structure (i.e., “brothers and sisters”), where any additional “first cousin” differentiation would make no sense.
In general, Gearing describes social structure as a more-or-less fixed
hierarchy of roles, where people move into and out of the roles over time.
And even though the people change, the social structure itself endures
and remains intact. Furthermore, each role has a set of behavioral rules or
norms attached to it which do not wholly determine how an occupant of
the role behaves, but nonetheless endow the role with certain opportunities and constraints that shape the actions and thoughts of the role’s
occupant. The social structure can and does change, but it changes more
slowly than the turnover of people in it. Social structure is a concept that
Gearing uses as a scientist. A companion concept to social structure is
“culture,” which Gearing describes as referring to the shared meanings,
shared codes, shared beliefs, or shared expectations that the Fox Indians
themselves have about the typical actions in which an individual Fox
Indian is allowed to engage in when he or she is occupying this or that
particular role in the social structure. For Gearing’s scientific concepts of
culture and social structure to be valid, the Fox Indians themselves need
not approve of them, nor even be aware of them in the first place.
One of the lessons I draw from Gearing is this: what everyday people see
is one thing, what scientific researchers see is another thing, and there need
not necessarily be any one-to-one correspondence between them. Everyday
managers can see what they think are organizational forms – old or new –
but researchers can choose to see these things in a different way. This lesson has relevance to my discussion about organizations and organizational
forms.
I have sometimes asked undergraduate and MBA students, “what is an
organization?” More often than not, they answer “an organization is peo-
312
Exploring Patterns in Information Management
ple.” My stock response to them is that an organization is not people. I
explain to them that an organization is what stays the same even when all
the people change. This conception of organization follows directly from
what I have learned about culture and social structure from ethnography.
What everyday people (undergraduates, MBA students, managers, executives, consultants, journalists, and other natives) see as an organization is
just another instantiation of “social structure” and “culture.” In this light, is
an organization necessarily a different phenomenon requiring its own theory? In other words, as for what undergraduates, MBA students, managers,
executives, consultants, journalists, and other natives consider to be new
organizational forms – are these necessarily new phenomena requiring new
theory or do they refer to instantiations of bodies of theory and streams of
research that already have a well developed presence in our community of
scholars?
I shall now explode (in the sense of exploding a process circle in a data flow
diagram) the example I have just given. In this example, I have shown how
the concepts, “organizations” and “new organizational forms,” can be considered to be instantiations of ethnographic theory. With a little imagination,
we can also see how “organizations” and “new organizational forms” can
also be considered instantiations of political theory, economic theory, psychological theory, information theory, sociological theory, structuration theory, and so forth. In other words, what everyday people see as “new organizational forms” need not necessarily be “new” to a researcher and therefore
need not necessarily call for the wholesale, custom development of new theory. This suggests that, perhaps, we can all take comfort in the strong possibility that what we already know still applies.
My argument has, so far, considered basic points about building theory and
organizational research. What additional considerations might there be for
a discipline that seeks to theorize about organizational forms enabled by
information technology? This depends on what the term “information
technology” means.
What an Information Technology Is
The conventional wisdom holds that technology – information technologies and other technologies – involves things that are mechanical or electronic. Digital computers are a form of information technology. At the
same time, digital computers are hardly the first form of information technology. We can seek lessons relevant to contemporary information technologies by returning to scholarly writings on older information technolo-
Lee
313
gies. An intriguing and, one of the oldest, examples of information technology is text.
It is likely that any person who reads takes the existence of text for
granted. One of the essays of the hermeneutic scholar Paul Ricoeur (1991)
served to open my eyes to the not-to-be-taken-for-granted properties of
text. In his essay, he points out that text is not just a written form of verbal
discourse. He explains, in his own words, that one might suppose that text
is simply the instantiation of speech, just as one might suppose that speech,
in turn, is simply an instantiation of language. Text can be the instantiation
of speech, but Ricoeur argues that text is something more than that, too.
First, text can be produced even when there is no speech. Text can be produced when there is no one else present with whom to speak or with whom
to communicate. And text can become separated from whoever its author
and original audience are. Ricoeur describes this as distantiation. Second,
text – unlike words spoken in real time – can become not only separate
from, but also independent of whoever it was that produced it; text can
continue to exist outside the presence of its creator and, in a sense, can
even come to say things that its creator never intended. Consider, for
example, what the Bible or the Koran means to the people who are reading
it today, in contrast to whatever these sacred texts originally meant to their
authors. Also consider what the US Constitution or any legal statute comes
to mean after years of successive interpretations of it by the courts, in contrast to whatever these legal texts originally meant to their authors. Therefore text is a tool that not only can do what its originators intended, but
also can lead to results that its originators never anticipated. Ricoeur
describes this as autonomization. Text can and does take on an autonomous life – a life of its own, a life independent of its author.
Third, text is not “inert”. It is not a sort of pipeline or conduit that somehow delivers packets of information to the reader. Instead, text is reactive.
It reacts with the reader. And the same text can react differently with different readers. A reader is not just a “blank slate” or empty vessel waiting
to be filled, but already has one or another “dictionary” in his or her head.
A reader is also someone who has already internalized the ways of this or
that society or organization, so that when she encounters a text, she is not
so much a recipient of information as she is externalizing meanings
already rooted in the dictionary and culture that she has previously internalized. Another way to say this is that the reader – a user of the non-electronic information technology that we are calling “text” – is, in a sense,
enacting the entire community, culture, organization, or society of which
he is a member and bringing that world to bear when creating his meaning
314
Exploring Patterns in Information Management
for the text. In this way, the reader (the technology user) takes over the
text, where the resulting meaning of the text can diverge extensively from
whatever was intended by the text’s author (the technology designer).
Ricoeur describes this as appropriation.
Text, as a form of information technology, has enabled the existence of
some organizations. The existence of text is a necessary condition for the
existence of some organizations, such as libraries and universities. For
certain other organizations, the existence of text might not be necessary,
but without text, these organizations could not operate in ways that we
would recognize; examples of such organizations are banks and courts of
law. I consider libraries, universities, banks, and courts to be examples of
organizations enabled by information technology, where the information
technology is text. When libraries, universities, banks, and courts first
appeared, they were certainly new instantiations of organizations, but
would their first appearance necessarily indicate the appearance of new
organizational forms? Also deserving consideration is the possibility that
they may be considered instantiations of old or already known organizational forms, whereupon old or already existing ethnographic theory,
political theory, economic theory, psychological theory, information theory, sociological theory, structuration theory, and so forth, would still
apply.
Do different organizations, in general, necessarily require their own separate and different theories? If we see them as examples of phenomena for
which we already have theories, then the answer would be that they do not
require new or different theories. At the same time, even though there
would be large, existing bodies of theory already available to us researchers, there would still be much work to be done. Earlier, I said that this is
want Kuhn calls “normal science”, where the new research would consist
of the arduous and challenging work of refining, articulating, and otherwise further developing existing theory so as to be able to explain organizations enabled by information technology.
My discussion of information technology now leads to the next concern:
what is meant by an information system and how is an information system
different from information technology?
What an Information System Is
I do not regard an information system to be the same entity as an information technology – whether the information technology is text or a digital
Lee
315
computer. Nor do I regard an information system to be the same entity as
an organization. In my view, an information system is an organization
enabled by information technology and, at the same time, an information
system is an information technology enabled by an organization. The two
thoughts may be merged as follows: an information system consists of an
organization and an information technology that so enable each other and
are so integrated into each other that neither could usefully function or
even exist without the other.
Where text plays the role of the information technology, we can say the
following: a library is an information system, a university is an information system, a bank is an information system, and a court of law is an
information system. Hence, even in the past when the information technology they used was not electronic, they were information systems even
then.
This conception of an information system involves a return to some basic
ideas in the information systems discipline. These basic ideas pose the following question about how to build a theory to explain so-called new
organizational forms: do libraries, universities, banks, and courts all require
the development of separate and unique theories to explain them, or might
they be well covered by existing ethnographic theory, political theory,
sociological theory, psychological theory, information theory, and so forth?
I am not concluding that there is no work to be done in building theories
about information systems. Rather, there is a great deal of work to be done
in the manner of “normal science”, where the work would largely be in
refining, articulating, and further developing existing theory so as to be
able to explain information systems.
In this essay, I have so far conveyed what I believe to be some basic points
about building theory, organizations, information technology, and information systems. Also requiring examination are some basic points about
information itself.
What Information Is
To characterize what information is, I will build on the example of text as
an information technology. I regard text – which I operationally define as
numbers and words in written form – as data. I define knowledge as the
understanding that a person has. (I also acknowledge the existence of what
many information systems scholars call “organizational knowledge,”
316
Exploring Patterns in Information Management
where I use the term “culture” to refer to this.) And I define information as
the knowledge that a person forms from data.
Even though the information systems discipline has made earnest
attempts to distinguish information, data, and knowledge from each
other, most of these attempts seem to have made no difference. Information systems researchers often appear to use the three terms interchangeably. My overall impression is that information systems researchers have a tendency to fall back on the idea that information is something
sent, received, processed, and stored, where the operant analogy compares information to a physical object. Indeed, a recent article by Elizabeth Davidson (1999) shows how the concept of data warehousing does
not clearly differentiate data, information, and knowledge. She reveals
the physically-oriented or physical-centric conception of data in data
warehousing, where data are seen and treated as if they were physical
inputs to a manufacturing process, the resulting products of which subsequently require storage and distribution.
Again, there are some basic points that are useful. As for data being a form
of text, I suggest that the large body of research in hermeneutics – which is
the academic field that devotes itself to the interpretation of text – promises to have a large repository of insights that are just waiting to be used by
the information systems discipline. As for what information is – or a person’s formation of meaning from data or other text – I suggest that psychology, symbolic interactionism, ethnography, and again hermeneutics,
all also have large repositories of insights that are just waiting to be used
by information systems researchers. As for knowledge – the understanding
that a person has – there is much that is still waiting to be applied from the
classic book The Social Construction of Reality, by Thomas Berger and
Peter Luckmann (1966). The subtitle of this book is, significantly, A Treatise in the Sociology of Knowledge. The phenomenological sociology of
Alfred Schutz – who was, by the way, the teacher of Thomas Berger and
Peter Luckmann – contains a treasure-trove of insights about knowledge
that knowledge-management research and information systems research in
general also have yet to use.
As members of the community of scholars, we have accessible to us a rich
infrastructure of theory about data, information, and knowledge, where this
is an infrastructure of theory that we do not need to re-invent. Instead we
can develop, articulate, and improve it as part of our larger effort in developing theories that explain the behavior of organizations enabled by
information technology.
Lee
317
Testing Theory
No discussion about theorizing and building theories is complete without
some commentary on how to test theories. In my experience as an editor,
reviewer, and reader of published research, I have seen that a basic point
about testing theory has been largely forgotten.
Many researchers proceed as if the validity of a scientific theory can be
properly established through induction, which refers to process of somehow inducing a theory from data. However, given the earlier lesson that
theory cannot be induced from data, induction is not an appropriate way to
test a theory. There is also the problem that induction allows the ad nauseam accumulation of consistent observations to support it – a situation
that Karl Popper (1965) dramatically illustrates for Adlerian psychology,
Marxist historiography, and astrology.
This directs our attention to the procedure of testing a theory deductively
instead of inductively. I conceptualize deductive testing as follows. Once a
theory has been formulated, a researcher can instantiate it in this or that
laboratory setting, field setting, or sampling frame. This means that the
theory, once instantiated, allows the researcher to deduce from the theory
what she should, and should not, observe in the given setting, provided that
the theory is correct. Positivist researchers would call these expected
observations “predictions”, but the positivist conception of this (“predictions”) is a special case, not the general case, of deductive scientific testing. Indeed, in one of his books, Michael Agar (1986) poses the device of
“strips” which I regard as an ethnographic manifestation of deductive
testing. The hermeneutic circle, as Klein and Myers (1999) explain in their
article, may also be argued to involve deductive testing.
Actual observations contradicting the expected observations would cast
doubt on the theory’s validity, whether the theory is positivist or interpretive. However, actual observations consistent with the expected observations would only be that: they would only be consistent with the theory and
could not definitively prove it to be true; at best, the theory could be said
to be true for the observed circumstances. Hence a researcher is allowed
only to accept a theory tentatively as “confirmed” or “corroborated”, but
never conclusively as “true”. In a way, the instantiation of a theory is a
particularization of the theory in a particular setting. This reasoning is
deductive in the sense that the researcher deduces statements (describing
details about what should or should not be observed) from a theory (when
applied in a particular setting).
318
Exploring Patterns in Information Management
Many information systems researchers, however, operate with the inductive belief that, in case research, the number of case sites must exceed a
certain minimum in order to validate the given theory, where a case study
involving a single site is considered weak or unacceptable. Acceptance of
the belief that scientific testing is deductive would displace this inductive
belief. I regard the persistence of this inductive belief as a sign of persistent resistance to deductive testing and persistent confusion over the difference between induction and deduction.
Illustration and Conclusion
In this essay, I have adopted the rhetorical device of presenting my argument in a way intended to provoke the reader. However, even if the provocative manner of the argument’s presentation were removed, the underlying message would remain intact. My argument about new organizational forms is not new. Certainly there are some organizational forms
enabled by information technology that are new. However, this essay has
given me a stage on which to make the point that not all that seems new is
new. It is difficulty to tell apart what is a new organizational form from
what is not.
To illustrate this, let me pose another analogy. Consider one of Sir Isaac
Newton’s theories, that “force is equal to mass times acceleration”, or
“f=ma”. Suppose that the community of scholars all accept that this theory,
“f=ma”, successfully explains the force of a steel ball that falls from the
top of a building. Now, suppose someone comes along and says: “I’m
observing a steel ball, but it isn’t falling from the top of a building. Instead,
it’s attached to the end of a string and I see it swinging back and forth in a
pendulum. This is obviously a new phenomenon, so the old theory,
“f=ma,” obviously no longer applies to explaining what the force is”.
For the steel ball that is part of a pendulum, are we necessarily encountering a new phenomenon so that “f=ma” does not explain its force, or do we
have a phenomenon where force can still be explained as “f=ma” so that it
is not necessarily a new phenomena after all? A student of first-year university physics knows that “f=ma” still explains the force in a pendulum,
where “a” or the acceleration is expressed with the help of differential calculus and sines or cosines. Hence the old theory, “f=ma” still applies, so
there is no necessity to invent a new theory for the steel ball that is part of
a pendulum. This is a different way of leading to a conclusion that I mentioned earlier: not everything that appears new is new.
Lee
319
The organizations that are emerging today and that are enabled by new
information technologies – are they necessarily new organizational forms,
or might they be like the steel ball that we happen to see, for the first time,
swinging in a pendulum, when previously we have only been accustomed
to seeing steel balls falling from the tops of buildings? How can we determine whether or not what we see is in fact a new phenomenon and, hence,
whether or not it requires a new theory? To answer this, I propose the following procedure.
• First, start with the premise that research on organizational forms
enabled by information technology can profitably begin with current
theory.
• Second, if current theory is true, then our instantiation of it in an actual
setting would lead us to expect to observe some things, but not others.
• Third, if the observations that we end up making do not match the
observations that the theory led us to expect, then the door would be
open to the possibility that current theory is wrong, incomplete, or otherwise deficient and that perhaps the organizational form is indeed
something new.
• Fourth, if eventually the community of researchers judges the existing
theory to be wrong, it would still be useful (and some would say, indispensable) for providing the needed basis or the starting point from
which to develop the new theory about the new organizational form.
Finally, if we were to give such a primary, foundational role to existing
theory in the way that I am suggesting, would this mean that the information systems discipline is subsidiary to the older, so called “reference disciplines”? To the contrary, in the same way that physics has contributed to
the engineering disciplines and in the same way that all the engineering
disciplines have developed their own scholarly research distinct from
physics, I see the following: I still see the older behavioral sciences and
design sciences as able to contribute to the information systems discipline,
but more importantly, I see that the information systems discipline is
already in the process of developing scholarly research distinct from the
older behavioral and design sciences, not only through the path of normal
science but also through the path of revolutionary science. For this reason,
I reject the term “reference discipline” and use the term “contributing discipline” instead. The scholarly study of information systems, originating
from the existing behavioral and design disciplines as its starting point, is
undergoing autonomization and is making contributions to theory transcending what the older disciplines have had to say.
320
Exploring Patterns in Information Management
In this essay I have considered the matter of building and testing theories
for new organizational forms enabled by information technology. Instead
of focusing on what might be new, I have returned to some old fundamentals about these basic points: building theory, what an organization is, what
an information technology is, what an information system is, and testing
theory. By taking these fundamentals seriously, we need not reinvent the
wheel when we proceed to develop better theory about new organizational
forms enabled by information technology.
References
Agar, M. (1986) Speaking of Ethnography, Sage Publications, Newbury Park,
California.
Berger, P. & Luckmann, T. (1966) The Social Construction of Reality: A Treatise
in the Sociology of Knowledge, Anchor Press, New York.
Davidson, E. (1999) “What’s in a Name? Exploring the Metaphysical Implications
of Data Warehousing in Concept and Practice”, Journal of End User Computing, Vol. 11, No. 4, pp. 22-32.
Gearing, F.O. (1970) The Face of the Fox, Aldine, Chicago, Illinois.
Klein, H. & Myers, M. (1999) “A Set of Principles for Conducting and Evaluating
Interpretive Field Studies in Information Systems,” MIS Quarterly, Vol. 23,
No. 1, pp. 67-93.
Kuhn, T. S. (1962) The Structure of Scientific Revolutions, University of Chicago
Press, Chicago, Illinois.
Kuhn, T. S. (1977) The Essential Tension: Selected Studies in Scientific Tradition
and Change, University of Chicago Press, Chicago, Illinois.
Popper, K. (1965) Conjectures and Refutations, Basic Books, New York.
Ricoeur, P. (1991) “The Model of the Text: Meaningful Action Considered as a
Text” in P. Ricoeur (Ed.) From Text to Action, Northwestern University Press,
Evanston, Illinois, pp. 146-167.
Schutz, A. (1962–66) “Concept and Theory Formation in the Social Sciences” in
Collected Papers (edited by M. Nathanson), M. Nijhoff, The Hague, pp. 4866.
— 19 —
Choosing the Problem:
Information Technology versus
Information Systems Phenomena1
Ron Weber
Introduction
Patterns! To see through the miasma and perceive the patterns in the
world. As researchers, this is our fundamental task. The types of patterns
we see, the ways in which we characterize and describe them, and the
quality of the theories we build to explain and predict them will largely
determine the contributions we make as scholars to our disciplines and to
knowledge more generally.
In this chapter of this book, which honors Mats Lundeberg as a researcher,
teacher, colleague, and friend, I want to focus on the sorts of patterns that I
believe lie at the core of the information systems discipline. Teasing out
and explaining patterns in information systems phenomena have always
been the focus of Lundeberg’s work – from his early work on the ISAC
methodology (Lundeberg et al. 1981) to his later work on business processes (Lundeberg 1992, 1993). In my own pursuit of patterns in information systems phenomena, Lundeberg often has reminded me astutely that it
is people’s perceptions of the world that ought to be the basis for our identifying, characterizing, and theorizing about the patterns that interest us in
our discipline.
Figure 1 depicts the fundamental argument I make in this chapter. Basically, I contend that the identification of novel patterns in phenomena provide the substance for articulating new, basic theories. These theories, in
turn, enable a discipline to establish its own separate, distinct identity or
place among other disciplines. Having a distinct identity contributes to the
longevity of a discipline.
1
I am indebted to my colleague, Paul Bailes, for helpful discussions on the subject
matter of this chapter.
322
Exploring Patterns in Information Management
Figure 1 Novel patterns, basic theory, and disciplinary identity
In the next section, I argue that patterns in phenomena are the foundation
of theory building and science. I distinguish between basic sciences and
applied or engineering sciences. I provide reasons for my wanting to
establish information systems as a basic science rather than an applied or
engineering science. Subsequently, I discuss the sorts of phenomena where
I believe novel patterns might be found to provide a basis for building
new, basic theory in the information systems discipline. In particular, I
argue that information systems-related phenomena and not information
technology-related phenomena will manifest these patterns. Finally, I present a brief summary of my arguments and some conclusions.
Why Are Patterns Important?
One way we can classify sciences is to use the two categories of basic sciences and engineering or applied sciences. Researchers in the basic sciences try to build fundamental theories to account for the behaviour of
things in the world – for example, atoms, cells, organs, people, organizations, societies, and economies. Researchers in the engineering or applied
sciences usually employ and adapt theories developed by researchers in the
basic sciences to solve practical problems. For instance, a researcher in the
field of aeronautical engineering might assemble a “package” of basic
theories from physics and psychology to produce a design methodology
for a plane that is more likely to meet the needs of its owners. Theories of
Weber
323
physics are needed to evaluate the likely flight performance under various
conditions of alternative designs for a plane. Theories from psychology are
needed to evaluate whether alternative designs for a plane are likely to
meet with the approval of passengers. In essence, the package of theories
is used to account for why planes that are designed according to the precepts of the methodology are likely to be more successful (at least in terms
of certain criteria).
In my view, the defining characteristic of a basic science is that its members have developed one or more powerful, general theories to account for
the patterns of behaviour in the things that are the focus of the science.
These theories must be substantive, original contributions. They cannot
simply be adaptations or extensions of theories from other disciplines. Of
course, what constitutes a substantive, original theoretical contribution is a
social and sometimes a political matter. The community of scientists in
general make a judgement. In due course, it acknowledges that a particular
discipline has “ownership” of a certain theory. Alternatively, it simply
ignores any aspirations of ownership that the members of a discipline
might hold for a theory. The community might conclude that the theory is
either not substantive or it is primarily an adaptation of a theory already
“owned” by another discipline. Judgements about the substance and ownership of theories are rarely, if ever, formal, overt affairs. Rather, they are
“observed” via the actions taken over time by the community of scientists
(primarily, I suspect, through the way researchers in other disciplines cite
the theory).
Note that basic science are often hybrid sciences in the sense they contain
both basic and applied-science elements. Researchers who work under the
ambit of the science have developed fundamental, basic theories to account
for some of the phenomena that interest them. At the same time, they borrow theories from other disciplines to account for other types of phenomena that interest them. Theories that are both intrinsic to and extrinsic to
the science are needed to account for the breadth of phenomena that command the attention of researchers who affiliate with the science.
For many scholars, whether they work within a basic science or an applied
or engineering science is unimportant. Both types of science clearly have
important roles to play in assisting humans to deal with the world. For
some of us, however, working with basic science is important. Inherently,
we find development and testing of basic theory to be more intellectually
satisfying than adapting or extending basic theory to an applied problem.
Some of us are also concerned about the longevity of the disciplines in
which we work. Applied sciences are “fragile” for a number of reasons.
324
Exploring Patterns in Information Management
First, they lack a distinct “identity”. As such, they often fall victim to politics or apathy when “turf wars” or resource battles occur in the organizational contexts in which they operate. Second, the applied problems that
are their focus may disappear or become relatively unimportant. For
example, technological advances may render the problems irrelevant or
uninteresting. Third, demonstrating progress within an applied or engineering science is often more difficult. Relative to basic sciences, therefore, members within them often experience more difficulty attracting
resources to support their research. Finally, some of us might hanker after
academic respectability. We value the “intellectually tough, analytic, formalizable, and teachable” subject matter often associated with the natural
or basic sciences (Simon 1981, p. 130).
Patterns are inextricably linked to judgements about the value of a theory
and the ownership of a theory. For example, one basis for evaluating the
value of a theory is the importance of the pattern it purports to explain or
predict. Patterns that are manifested in the behaviour of many things are
likely to be deemed more important (the value we ascribe to generality in
science). Thus, theories that provide powerful accounts of these patterns
are likely to be judged as valuable. On the other hand, patterns that appear
localized to only a few things in the world are likely to be deemed relatively uninteresting (although this is not always the case). Thus, theories
that account for these specific patterns are likely to be judged by scientists
as having low value.
The identification of new patterns in the behaviour of things is often the
precursor to the articulation of important theoretical work and ultimately
“property rights” in this work being assigned to the discipline whose
members undertake it. Moreover, new theories sometimes enable us to
“see” the patterns that have been our focus manifested elsewhere. They
open our eyes to phenomena that previously were hidden from us. As more
instances of the patterns are identified, the importance of the theory that
accounts for them will grow. As a result, the “identity” of the discipline
whose members developed the theory will become more firmly established.
In short, if we are seeking to establish the “identity” of a discipline by
establishing property rights to a powerful, general theory, our choice of the
phenomena on which to focus is critical. If we choose phenomena that can
be accounted for satisfactorily by theories already developed by other disciplines, our own discipline will remain an applied discipline – a discipline
that borrows theories from other disciplines. If, on the other hand, we
somehow manage to choose phenomena that are not well explained or pre-
Weber
325
dicted by theories developed by other disciplines, we lay a foundation for
establishing the identity of our own discipline. We must be able to see
patterns in the phenomena, however, that are manifested in the behaviour
of many other things (the requirement of generality). We must also be able
to develop novel, compelling, powerful theory to account for the patterns.
Identifying phenomena that manifest patterns that are amenable to building
powerful theories is the problem of the problem (Weber 2003).
Information Technology versus Information Systems
as the Source of Phenomena
Where ought we to look to identify phenomena that are not accounted for
well by theories developed in other disciplines? Our discipline clearly has
divergent views in relation to this question. Even though our discipline is
called the information systems discipline, some of our colleagues apparently believe the source of phenomena will be the things we call information technology. Others believe the source of phenomena will be information systems. I count myself in this latter group. In the subsections below, I
will argue why I believe information systems rather than information technology will be the source of the phenomena that allow us to build theories
that establish the identity of our discipline.
Information Technology as the Source of Phenomena
By information technology, I mean the artifactual resources we use to
develop, implement, operate, use, maintain, and manage an information
system. Often we classify information technology as hardware (e.g., computers, input/output devices, and network facilities like cables and
modems) or software (e.g., programming languages, CASE tools, enterprise resource planning systems, customer resource management systems,
and operating systems.)
What sorts of information technology-related patterns might be the focus
of researchers who seek to build basic theory and thereby to establish the
foundation for and identity of their discipline? One type of pattern is that
which arises in the behaviour of humans as they interact with information
technology – for example, a user of an information system as they work
with a particular kind of human-computer interface, a programmer as they
try to write program code using some kind of programming language, a
child as they try to master keyboard skills, or an employee trying to decide
whether to adopt a new kind of software package.
326
Exploring Patterns in Information Management
I have reflected on different patterns of this type at some length, and I cannot identify any that might provide the foundation for new theory. Existing
theories (e.g., psychological, social, and economic theories) seem to provide an adequate account for the patterns I have been able to identify. To
the best of my knowledge, research that examines patterns manifested in
human interaction with information technology uses existing theories or
adaptations thereof to account for the patterns. For example, witness the
extensive use of the theory of planned behaviour or theory of reasoned
action or adaptations of these theories to account for user adoption and
deployment of various forms of information technology.
Consider, also, other forms of technology that humans have developed –
for example, automobiles, electric toothbrushes, and mobile phones. To the
best of my knowledge, we have not had to develop new, basic theory to
account for patterns of behaviour associated with humans’ use of these
technologies. For example, we do not have a theory of the electric toothbrush – a new, basic theory that had to be developed specifically to
account for the patterns of behaviour that became apparent in human’s use
of their electric toothbrushes. In this regard, I have also questioned senior
colleagues in disciplines like sociology, anthropology, psychology, and
economics about whether they know of such theories. They have been
somewhat bemused by questions. Nonetheless, they have been unable to
point to a theory of the type I was seeking.
In essence, if we are to develop new, basic theory to account for the interactions of humans with information technology, we first have to identify
patterns of behaviour where extant theories fail – in other words, theories
from disciplines like psychology or sociology are unable to explain or predict satisfactorily the patterns of behaviour that are our focus. If we can
identify such patterns, of course they must also be “interesting” in the
sense we deem them important for some reason.
A second type of pattern associated with information technology that
might be our focus is one associated with the technology itself. Specifically, we might be concerned with giving a particular type of information
technology certain characteristics or properties so that it “behaves” in particular ways. For example, we may find that if we design an information
technology along certain lines, it works more effectively or efficiently or it
is more robust when component failure occurs.
I believe we have some notable examples of new, basic theories that have
been developed to account for this second type of pattern. Compiler theory
was developed to translate human-oriented languages into machine lan-
Weber
327
guages in effective and efficient ways. The need for compiler theory arose
because of the special computational properties of information technology
– properties that were not present (at least to the same extent) in prior
forms of technology. A programmer who knows compiler theory will be
capable of producing a much higher-quality compiler than a programmer
who has no knowledge of the theory. Admittedly, compiler theory is an
adaptation and extension of prior theories of computational linguistics that
have their home in the discipline of linguistics. The enhancements made to
these basic theories by researchers whose focus was compilers has been
sufficiently extensive, however, for “property rights” on the theory to be
ascribed to them rather than linguists.
Another example of a new, basic theory that has been developed to
account for this second type of pattern associated with the information
technology itself is the theory of normalization. Codd’s (1970) seminal
work on data normalization fundamentally changed how databases are
designed and implemented. It also laid the foundation for the development
of new types of information technology – namely, relational database management systems and relational database machines. Again, the theory of
data normalization was an adaptation and extension of the theory of relations developed within the discipline of mathematics. The enhancements
made to the theory of relations have been sufficiently substantive, however, that property rights to the theory of data normalization have been
ascribed to database researchers rather than mathematics researchers.
With this second type of pattern associated with information technology,
therefore, we have evidence that new, basic theories have been needed.
From one perspective, it might be argued that new patterns emerged as a
result of the special properties of information technology (computational
properties) relative to prior technological artifacts that humans had invented. From another perspective, it might be argued the patterns were already
present in phenomena associated with some previous forms of technology.
They became more salient with information technology, however, and thus
they commanded the attention of researchers in ways that had not occurred
before. Also, theoretical lenses that had been used or were being developed
to better understand and predict phenomena associated with information
technologies perhaps allowed researchers to see these patterns in richer,
more-perspicacious ways.
I believe we have evidence, therefore, of new, basic theory being needed to
account for novel (or perhaps more-salient) patterns manifested in phenomena associated with humans’ needs to make information technology
behave in certain ways. From our perspective as members of the informa-
328
Exploring Patterns in Information Management
tion systems discipline, however, we cannot lay claim to these theories.
The property rights associated with them belong elsewhere–specifically, in
the discipline of computer science. Thus, we cannot use these theories to
support a claim that we are a basic science. Nor can we use these theories
to articulate the nature of our own identify – an identity that is separate
from other disciplines.
In due course, perhaps other patterns manifested in phenomena associated
with humans’ needs to make information technology behave in certain
ways will become evident. If this outcome were to occur, I doubt, however, that information systems researchers will be the source of new, basic
theory to account for these patterns. For the most part, I believe we lack
the skills, knowledge, and background to develop the sorts of theories that
will be needed to account for such patterns. If history is any guide, they lie
more in mathematics and formal methods than in social-science methods.
Information Systems as the Source of Phenomena
Elsewhere I have attempted to define formally what I mean by an information system (Weber 1997). In essence, however, by an information system I mean a set of things that are coupled together to provide a representation of someone’s or some group’s perceptions of the states of and state
changes that occur in another system. For example, an order-entry system
is a set of coupled components (e.g., a program, a personal computer, a
server, and a network) that represents the states of and state changes that
occur in a customer. When a customer decides that she or he wants to purchase some good or service, an order-entry system represents the state
change in the customer by creating an order record for the customer. By
“observing” an information system, we obviate the need to observe the
represented system directly. For example, when we observe that an order
has been created for a customer in an order-entry system, we know a state
change has occurred in the customer to the effect that she or he wishes to
purchase a good or service that we provide. We do not have to physically
visit the customer and consult with her or him directly to find out whether
she or he wishes to buy something we provide to the marketplace. In this
way, our use of the information system conserves resources.
As with information technology, I believe two types of patterns that occur
in humans’ interaction with information systems might provide the foundation for building new, basic theory. The first arises in the behaviour of
humans as they interact with information system – for example, a user of
an information system examines different types of graphical and tabular
Weber
329
output to make a decision, a manager tries to use an information system to
give her or his organization a competitive advantage in the marketplace, or
an employee uses an information system in an effort to enhance her or his
power within an organization.
In the mid-1980s, Yair Wand and I reflected at length on patterns of this
type (Wand & Weber 1995). We were attempting to identify opportunities
to develop new, basic theory that might eventually establish an “identity”
for the information systems discipline. Eventually, we came to the conclusion that existing theories (e.g., psychological, social, and economic theories) seem to provide an adequate account for the patterns we were able to
identify as the focus of researchers in our discipline.
In this regard, consider any research that goes under the rubric of information systems research and try to identify a theory that is used in the
research that is not sourced from other disciplines. As Orkilowski and
Iacono (2001) have pointed out, much of the research that has been published in the major information systems journals pays only token attention
to information systems (or information technology for that matter). Similarly, Benbasat and Zmud (2003) have pointed out that much information
systems research is guilty of “errors of exclusion” and “errors of inclusion”. Errors of exclusion arise when the research includes neither the
information technology artifact nor constructs that are closely associated
with it. Errors of inclusion arise when the research includes constructs that
are “best left to scholars in other disciplines” because of their “significant
causal distance” from the information technology artifact.
In short, given the amount of research that has been undertaken already in
the information systems discipline on this first type of pattern, I am pessimistic about the likelihood of our discovering a pattern that will provide
the foundation for our developing new, basic theory. Nonetheless, in due
course I hope to be proved wrong in reaching this conclusion. If we could
identify such patterns, we would open opportunities for us to formulate
rich, basic theory that give our discipline its own identity.
The second type of pattern associated with information systems that might
be our focus is one associated with the information system itself. Specifically, we might be concerned with giving information technology certain
characteristics or properties so that it “behaves” in particular ways. Here,
Wand and I concluded there are opportunities for us to build new, basic
theory. In this regard, the patterns that have been our focus since the mid
1980s have been those that manifest information systems provide “good”
or “faithful” representations of someone’s or some group’s perceptions of
330
Exploring Patterns in Information Management
the system whose states and state changes the information system is supposed to track. Historically, these patterns have been the focus of researchers concerned with conceptual modelling – building graphical models of
an application domain as a basis for developing information systems to
support users who work in the domain.
While substantial work has been done on conceptual modelling methods,
much of it is atheoretical. Indeed, lack of theory resulted in conceptual
modelling research falling into disrepute – a problem that researchers on
conceptual modelling have found difficult to shrug off. Wand and I saw
the absence of theory, however, to be an important opportunity to build
theory in relation to the representational phenomena that we believe lies at
the heart of information systems. In this light, we have worked to articulate
and test basic theory about the nature of good or faithful representations.
We have used and adapted a theory of ontology developed by Bunge
(1977) as the basis for our work. In this regard, like other information systems research, we are borrowing a theory from another discipline (philosophy) to account for the phenomena that are our focus. In the case of conceptual modelling, however, we believe Bunge’s ontological theory will
have to be extended and adapted markedly for it to provide powerful
explanations and predictions of conceptual modelling phenomena. Prior
research on conceptual modelling suggests the sorts of ways that Bunge’s
theory needs to be extended and enhanced. Ultimately, we hope that the
“value-add” of the theoretical work done in the information systems discipline will lead other disciplines to ascribe ownership of the new theoretical
contributions to the information systems discipline. Of course, only time
will tell. For the moment, however, Wand and I believe that information
systems researchers have a rich agenda of work on conceptual modelling
that can be undertaken (Wand & Weber 2002).
As with patterns of the first type, perhaps patterns of the second type exist
that provide a basis for developing new, basic theory in the information
systems discipline. Once more, I hope this is the case because the identification of more patterns would lay the foundation for a richer, morediverse, more-interesting and potentially more-important discipline. For
the moment, however, I am unable to identify such patterns.
Summary and Conclusions
In the early 1990s, Mats Lundeberg wrote: “…my main message to you is
to perceive reality as it is. One strategy for doing so is to recognize patterns in reality by making use of some basic frameworks and frames”
Weber
331
(Lundeberg 1993, p. 1, my emphasis). As I indicated in the introduction to
this chapter, Mats has been a seeker of patterns in phenomena throughout
his entire career. The “frameworks and frames” he has articulated to try to
understand these patterns form the basis for developing representations or
models of a domain. These, in turn, form the basis for designing and
implementing information systems.
In this chapter, I have argued that identifying new patterns in phenomena
is an important way in which we establish a foundation to “see” new, basic
theory. I believe the development of such theory is a critical means by
which disciplines establish their own distinct identity. In the absence of
“owning” a basic theory (or theories), a discipline’s identity will be fragile
because the theories it uses to account for the phenomena that are the focus
of its members will be rooted in other disciplines.
As members of the information systems discipline, I believe we will have a
greater chance of establishing basic theory if we focus on information
systems-related phenomena rather than information technology-related
phenomena. Unfortunately, to date, the phenomena that have been our
concern have not motivated a need for new, basic theory. Instead, we have
been able to use theories borrowed from or adapted from other disciplines
to account for their behaviour. For the moment, I believe our best opportunities for developing new, basic theory lie in phenomena associated with
conceptual modelling – that is, building “good” or “faithful” representations of someone’s or some group’s perception of a real-world domain. We
need to tease out patterns that underlie representational phenomena to see
whether they provide the grist for the development of new, basic theory.
References
Benbasat, I., & Zmud, R.W. (2003) “The Identity Crisis Within the IS Discipline:
Defining and Communicating the Discipline’s Core Properties”, MIS Quarterly. Vol. 27, No. 2, June, pp. 183-194.
Bunge, M. (1977) Treatise on Basic Philosophy: Volume 3: Ontology I: The Furniture of the World, D. Reidel Publishing Company, Dordrecht, Holland.
Codd, E.F. (1970) “A Relational Model of Data for Large Shared Data Banks”,
Communications of the ACM. Vol. 13, No. 6, June, pp. 377-387.
Lundeberg, M. (1992) “A Framework for Recognizing Patterns When Reshaping
Business Processes”, The Journal of Strategic Information Systems. Vol. 1,
No. 3, pp. 116-125.
Lundeberg, M. (1993) Handling Change Processes: A Systems Approach, Studentlitteratur, Lund, Sweden.
332
Exploring Patterns in Information Management
Lundeberg, M., Goldkuhl, G. & Nilsson, A. (1981) Information Systems Development: A Systematic Approach. Prentice-Hall, Englewood Cliffs, New Jersey.
Orlikowski, W. & Iacono, S. (2001) “Desperately Seeking the “IT” in IT
Research–A Call to Theorizing the IT Artifact”, Information Systems
Research. Vol. 12, No. 2, June, pp. 121-134.
Simon, H.A. (1981) The Sciences of the Artificial, 2nd ed., The MIT Press Cambridge, Massachusetts.
Wand, Y. & Weber, R. (1995) “On the Deep Structure of Information Systems”,
Information Systems Journal, Vol. 5 No. 3, July, pp. 203-223.
Wand, Y. & Weber, R. (2002) “Information Systems and Conceptual Modelling–
A Research Agenda”, Information Systems Research. Vol. 11, No. 4, December, pp. 363-376.
Weber, R. (1997) Ontological Foundations of Information Systems, Accounting
Research Methodology Monograph No. 4, Coopers & Lybrand, Melbourne,
Australia
Weber, R. (2003) “The Problem of the Problem”, MIS Quarterly. Vol. 27, No. 1,
March, pp. iii-ix.
Contributing Authors
Erling S. Andersen, Professor of Information Systems and Project Management, Norwegian School of Management, Oslo, Norway. Dean of the
China activities of the Norwegian School of Management BI. Erling
Andersen has published a number of books and articles on information
technology, systems development, project management and management
in general, several of which have been published in several languages.
([email protected])
Niels Bjørn-Andersen, Professor of Information Systems, Director of
Center for Electronic Commerce, Copenhagen Business School, Copenhagen, Denmark. Niels Bjørn-Andersen's publication list includes 15 books,
more than 30 refereed articles, and about 100 other publications. He
organised the first ICIS conference outside North America, is former
president of AIS and has held numerous keynote addresses at conferences
around the world. He has received several academic awards, including
being named Fellow by the Association for Information Systems in 1997.
He is and has been the national leader in several ESPRIT research programs. ([email protected])
Gordon Davis, Honeywell Professor of Management Information Systems, Carlson School of Management, University of Minnesota, USA.
Internationally recognized as one of the founders of the academic field of
information systems, the writer of a foundational textbook and a leading
figure in several international academic organizations, including IFIP and
AIS. He is fellow of the Association of Computing Machinery and holds
honorary doctorates from the University of Zurich, the University of Lyon
III, and the Stockholm School of Economics. He was honoured with the
LEO award by the Association for Information Systems, for lifetime contributions to the field of information systems. ([email protected])
Michael J. Earl, Dean, Templeton College, Oxford and Professor of
Information Management, Oxford University. Formerly professor at London Business School, Michael Earl is extensively published in leading
journals, including Harvard Business Review, MIT Sloan Management
Review and MIS Quarterly. He has also written an influential textbook and
edited several important anthologies on information management.
([email protected])
Göran Goldkuhl, Professor of Information Systems Development at
Linköping University and Professor of Informatics at Jönköping International Business School. He is the director of the Swedish research network
334
Exploring Patterns in Information Management
VITS and is currently developing a family of theories founded on socioinstrumental pragmatism: Workpractice Theory, Business Action Theory
and Information Systems Actability Theory. Greatly interest in interpretive
and qualitative research methods, he has contributed to the development of
Multi-Grounded Theory and is active in international research communities such as Language Action Perspective and Organisational Semiotics.
([email protected])
Helle Zinner Henriksen, Assistant Professor, Received her Ph.D. from
the Department of Informatics at Copenhagen Business School, where she
currently works as an assistant professor. She is co-author of a book on
business-to-business e-commerce and has written several book chapters
and papers on this and other topics. Her main interests include business-tobusiness electronic commerce, EDI, adoption and diffusion of Interorganizational Information Systems, eGovernment, and regulation of electronic
commerce and eGovernment. ([email protected])
Rolf Høyer, Professor at the Centre for Media Economics, Norwegian
School of Management (BI), Oslo, Norway. He has previously been professor of Management Information Systems at the Universities of Gothenburg and Bergen, and also at BI. Rolf has authored and edited several
books on user participation and user influence, management of IT and
media management. ([email protected])
Pentti Kerola, Professor Emeritus of Informatics, University of Oulu Finland. After studies in applied mathematics, he worked during the sixties as
operations researcher in IBM and integration manager of ADP development in Stora Enso. He was then assistant professor of Informatics in the
Technical University of Helsinki and became professor in Oulu 1973. His
main research interests include macro-modelling of information systems
development, human–computer interaction and educational research of
information systems architects. He actively influenced the Nordic doctoral
education in information systems by establishing the IRIS conference
series. ([email protected])
Börje Langefors, Professor Emeritus of Information Processing, Royal
Institute of Technology and Stockholm University. Pioneer in the Swedish computer industry before becoming the first professor of information
systems in Sweden (1965). As one of the founders of the field, he has
had profound impact on Scandinavian IS research. He introduced the
term “information system” (in 1965), wrote the influential book Theoretical Analysis of Information Systems and chartered the infological
approach to information systems research. He was the first chair of IFIP
Contributing Authors
335
Technical Committee 8 and is a recipient of the LEO award from the
Association for Information Systems for lifetime contributions to the
field. ([email protected])
Michael Holm Larsen, Assistant Professor, holds a M.Sc. (Eng.) and a
Ph.D. from the Department of Manufacturing Engineering, Technical University of Denmark. Currently with the Center for Electronic Commerce
and the Department of Informatics at Copenhagen Business School. He has
published in journals such as Decision Support Systems, Computers in
Industry, International Journal of Intelligent Automation and Soft Computing and in various international conference proceedings such as HICSS,
ECIS, and in international conferences on production research, control and
automation, and intelligent manufacturing systems. His current research
interests are in entrepreneurship and business engineering, e-business
strategy and e-business models, business process reengineering and management, and distributed knowledge management. ([email protected])
Allen S. Lee is Professor of information systems and associate dean for
research and graduate studies in the School of Business at Virginia Commonwealth University. Former editor-in-chief of MIS Quarterly. Allen
Lee, a leading scholar in the field of information systems, is widely published and has delivered keynote addresses at numerous conferences. He
specializes in qualitative, interpretive, and case research on how information technology is used in organizations; as well as on information systems
implementation, electronic communication and research methodologies.
([email protected])
Magnus Mähring, Assistant Professor, Department of Information Management, Stockholm School of Economics and research associate at the
Institute for Business Process Development, Sweden. Visiting research
scholar, Computer Information Systems department, Georgia State University 2002-2003. Magnus has extensive teaching and consulting experience
and has worked with private sector, public sector and non-profit organisations. His current research focuses governance and management of IT projects and inter-organisational IT-related change. ([email protected])
Pär Mårtensson, Assistant Professor, Department of Information Management, Stockholm School of Economics, Sweden. He is also a research
associate at the Institute for Business Process Development (Institute V) in
Stockholm and Program Director for the International Teachers Programme. Pär teaches in undergraduate and executive education and is
internationally involved in work with management education and development. He also works with change and development projects in industry.
336
Exploring Patterns in Information Management
His current research focuses on management processes in business transformation and change. ([email protected])
Anders G. Nilsson, Professor of Information Systems at Karlstad University, Sweden. He has a Ph.D. in Information Management from Stockholm
School of Economics and is a research partner at the Institute for Business
Process Development (Institute V) in Stockholm. Anders has as a
researcher worked with the ISAC approach for information systems development, the SIV method for acquisition of standard application packages
and a Business Modelling framework for creating method combinations.
He has also been active as an advisor to many change projects in private
industry and the public sector. Anders is author/co-author of 14 books on
business and systems development. ([email protected])
Kristina Nilsson, Assistant Professor, Department of Information Management, Stockholm School of Economics, Sweden. She is Academic
Dean for Executive MBA and Program Director for Executive MBA Business Development & IT. She is also Program Director for the International
Teachers Programme running in Stockholm 2003-2005. Her present and
previous research has focused on executive information and information
support. She is also highly involved in pedagogical and educational development. ([email protected])
Hans-Erik Nissen, Professor Emeritus, Department of Information and
Computer Science, Lund University. Hans-Erik Nissen has 19 years of
industry experience with SCA, the last seven years organizing their first
computer centre. He is author of several books and articles on information
systems use, development, and IS research methodology. He was a member of the ISAC project during the 1970s and suggested including descriptions and analyses of business activities. He has been highly influential in
the groundbreaking research methodology workshops within IFIP Working
Group 8.2 during the 1980s and 1990s. ([email protected])
Tapio Reponen is Professor of Information Systems at the Turku School
of Economics and Business Administration, Finland. Currently he is the
Rector of the school. His research interests are: information management,
strategic information systems, organizing the IS function and knowledge
management. He has published, reviewed and edited articles and books
linked to these themes. He has also been in program and organizing committees of numerous international conferences. ([email protected])
Mikko Ruohonen holds a professorship at the University of Tampere and
a docentship at the Turku School of Economics and Business administration. He has worked in the field of information resources strategy since
Contributing Authors
337
1984 and his teaching and research focuses information strategies, electronic business, knowledge management and inter-organisational learning. He
is extensively involved in executive education, has published four textbooks and performs industry assignments. He is chairman of IFIP WG 3.4.
([email protected])
Dietrich Seibt, Professor, Director of the Department of Information Systems & Information Management, University of Cologne, Germany.
Dietrich Seibt has authored and edited several books on information management and written numerous book chapters and articles. His research
focuses areas that include management of information and information
systems, multimedia telecommunications applications, electronic commerce, and e-learning systems. ([email protected])
Åge Sørsveen, Consultant, teacher and researcher, Oslo, Norway. Åge
Sørsveen has decades of experience in organisational (including ITrelated) change and development of models, frameworks and learning
approaches in this field. He is author of Ledelse pa norsk (Norwegian
Leadership) with Erling Andersen and Ingeborg Baustad. (aage.sorsveen@
idrettsforbundet.no)
Gösta Steneskog, M.Sc., researcher and management- and IT-consultant.
He is involved in research, consultation, and knowledge transfer in the area
of Business Process Development, with a special focus on Process Management, Project Management and Information Technology. He has
worked in several positions at IBM and other corporations. Gösta has been
involved in European research projects and published in several books.
([email protected])
Bo Sundgren, Professor, Department of Information Management, Stockholm School of Economics, Sweden. He received his PhD from the University of Stockholm in 1973 with the thesis “An Infological Approach to
Data Bases”. His main research interests are conceptual modelling, analytical information systems, and metadata systems, and he has published
numerous books and papers on these topics. At present he shares his time
between academic research and a position as senior advisor to the management of Statistics Sweden, and he has also undertaken numerous tasks
related to statistical information systems on the international arena.
([email protected])
Alexander Verrijn-Stuart, originally a physicist, worked for the Royal/
Dutch Shell Laboratory and then had an international career with Shell in
computing and planning, before being appointed Professor of Computer
Science at Leiden University in 1970 (emeritus 1991). His publications
338
Exploring Patterns in Information Management
cover several areas in information systems, with particular emphasis on
core concepts and core aspects of the field. Founding member of IFIP
TC8 and WG8.1, succeeding Börje Langefors as TC8 chairman. Among
other things, served as faculty opponent at Mats Lundeberg’s doctoral
defence. Elected Member of the Royal Holland Society of Sciences and
Humanities (1975), of which he became Secretary for the Sciences in
1989. ([email protected]).
Ron Weber, Professor of Information Systems in the School of Business
and Research Director for the Faculty of Business, Economics and Law at
the University of Queensland, Australia. He teaches and researches in the
information systems and accounting areas, and he has published extensively in both Australian and international journals. He has also consulted
widely on information systems matters, especially the control and audit of
computer systems. Ron is a past senior editor for the MIS Quarterly and
the present editor-in-chief of MIS Quarterly (the first non-U.S. person to
hold that position). He has also been President of the Association of Information Systems. ([email protected])
Alf Westelius, Assistant Professor at the Department of Information Systems and Management, Linköping University, Sweden. He has a Ph.D.
from the Stockholm School of Economics. Much of his research concerns
organisational change ventures where IT is intended to play a large role.
People’s learning and (changing) conceptions of their work and the organisation are in focus. Research, teaching and advisor activities all deal with
business, and governmental agencies as well as non-profit organisations.
Alf is also a cellist and member of the board of a number of Swedish
music-related societies. ([email protected]).
Ordering Information
This book can be ordered directly from:
The Economic Research Institute (EFI)
Stockholm School of Economics
Box 6501, SE-113 83 Stockholm, Sweden.
Phone: +46-8-736 90 00 • Fax: +46-8-31 62 70
E-mail: [email protected] • Internet: www.hhs.se/efi
Download Information
All chapters in this book can be downloaded
in electronic form from:
www.hhs.se/im/exploringpatterns
EFI Publications
Information on the full range of publications from
the Economic Research Institute is available from:
www.hhs.se/efi
Blanksida
Blanksida