Lithics 21, 2000, 45-51
QUARTZ TECHNOLOGY IN SCOTTISH PREHISTORY
Alan Saville and Torben Bjarke Ballin
Historic Scotland and the National Museums of Scotland agreed to fund the initial stage jointly.
After discussion of the selection criteria for the
pilot project and following examination of material
in the care of the National Museums, a number of
quartz assemblages were accepted as suitable for further research. As most published quartz assemblages
are of Bronze Age date, it was decided to focus on
Mesolithic assemblages, and Kilmelfort Cave, Argyll
(Coles 1983), Lealt Bay, Isle of Jura (Mercer 1968),
and Shieldaig, Wester Ross (Walker 1973) were
those selected for initial study. A specialist report (by
Torben Ballin) was produced on each of the three
chosen assemblages, together with a project report
(Ballin 2001), which included recommendations for
the proposed main project.
The main aim of the quartz project is to shed
light on quartz variability, that is, to define how
quartz assemblages in different periods and areas of
the Scottish quartz province (N and NW Scotland)
differ. With the general variation defined, it will be
attempted to explain the variation. It is assumed that
most differences between assemblages will be due to
differences in chronology, regionality (territoriality),
access to resources, or site activities. The prime purpose of the pilot project was to explore the prospects
of a full-scale quartz project and assess the problems
associated with the analysis of quartz assemblages.
INTRODUCTION
It is generally recognized that, in the Stone and
Bronze Ages of Scotland, numerous different lithic
raw materials were used for the manufacture of tools
(Saville 1994). While there are considerable variations at local level (e.g. involving bloodstone: Wickham-Jones 1990 or pitchstone: Haggarty 1991, 91),
taking Scotland as a whole the most common raw
materials exploited appear to be flint, chert and
quartz. The true importance of quartz has to a degree
been obscured by research bias since, probably because of difficulties associated with identifying tools
in quartz, many archaeologists have shunned this raw
material (cf. Lindgren 1998). As a result, many publications of Scottish quartz assemblages tend to be
characterized by lack of enthusiasm, detail and precision (Table 1). From the literature there is an impression that quartz was a raw material only used in a
few remote corners of prehistoric Scotland, whereas
in fact it was a major lithic resource, or at least an
important supplement, in large parts of the country
over several millennia. The purpose of the project
'Quartz Technology in Scottish Prehistory' is to increase awareness of the significance of quartz in earlier prehistory.
PROJECT STRUCTURE AND AIMS
A draft project proposal was produced by Torben
Ballin and discussed with Alan Saville and Patrick
Ashmore (Historic Scotland). A two-stage project
structure was suggested, with a pilot project to be
completed in the financial year 2000101 and a main
project to be carried out in 2002/03, founded on the
results and recommendations of the pilot project.
METHODOLOGY
To ensure comparability between the selected assemblages, it was important to develop a standard methodology for description and discussion of the involved assemblages. The first element of the chosen
45
Quartz Technology in Scottish Prehistory
Assemblage and associated inferential problems
Date
Amount of quartz
Mesolithic?
668
Lealt Bay, Jura (Mercer 1968)
Quartz and flint assemblages not kept separate in presentation
Mainly Mesolithic
'12 lb of milky quartz'
Lussa River, Jura (Mercer 1971)
Quartz and flint assemblages not kept separate in presentation
Mainly Mesolithic
'72lb of milky quartz'
N Cam Bay, Jura (Mercer 1972)
Quartz and flint assemblages not kept separate in presentation
Mainly Mesolithic
'5 lb 10 oz of milky quartz'
Kilmelfort Cave, Argyll (Coles 1983)
Terminological problems*
Mesolithic?
331
Carding Mill Bay, Argyll (Finlayson 1992)
' ... sorted from only a sample of the shell material'
Early Neolithic
(Obanian)
990
Loch Olabhat, North Uist (Finlayson forthcoming)
Unpublished
Neolithic
Hundreds
Stanydale, Ness of Gruting, Gr. School, Wiltrow (Calder 1956)
Material mainly excavated before 1950
Neolithic/Early B.A.
?
Scord of Brouster, Shetland (Whittle 1986)
Terminological problems*
Neolithic/Early B.A.
'Nearly 8,000'
Callanais, Lewis (Ballin forthcoming b)
Late Neol.lEarly B.A.
234
Lairg, Sutherland (Finlayson 1996)
c.2,000 pieces of c.1 0,000 analyzed
Late Neol.lBronze Age
'Some 10,000'
Northton, Isle of Harris (Simpson 1976)
Unpublished
Late Neol.lBronze Age
?
Rossinish, Benbecula (Ballin forthcoming a)
Early Bronze Age
3,532
Jarlshof, Shetland (Hamilton 1956)
Material excavated before 1950
Early-Late Bronze Age
'Hundreds of quartz tools'
Tougs, Shetland (Hedges 1986; Lehane 1986)
Terminological problems*
Early-Late Bronze Age
1,457
Bayanne, Shetland (Ballin, forthcoming c)
Late Bronze Age
2,955
Kebister, Shetland (Clarke 1999)
A very brief report (one page of text + two tables)
B.A./Early Iron Age?
1,155
Quartz assemblage not analysed/published
Table 1. Some important quartz assemblages, their dates and sizes, and problems limiting their inferential use-value.
* 'Terminological problems' generally refers to the use of a typology not including bipolar material. In Hedges (1986) and Lehane (1986) the terminological problems are due to the adoption of Broadbent's (1979) typology.
46
Saville and Ballin
approach was the identification of raw materials
(sub-types of quartz), as the properties of the raw
material might determine the choice of technology
and, to some degree, artefact morphology.
The second element was typological classification and description of all debitage, cores and tools,
and in this connection to discuss relevant typological
concepts. As a third element, the technology was defined and discussed on the basis of raw material and
typological background. And finally, the date and
chronological coherence of the assemblages were
examined. Due to excavation without standard grid
systems it was not possible to test the assemblage
chronology for the selected assemblages via distribution analysis, but this may be a relevant approach in
connection with the main quartz project.
Shieldaig spoil was sieved; at the other sites not) and
different flaking properties (the 'vitreous' varieties of
quartz tend to splinter, the quartz with a 'greasy' lustre does not). Due to differences in date, raw material
availability and technology, the typological compositions of the core and tool groups vary greatly. The
Kilmelfort Cave cores are virtually all bipolar, those
from Lealt Bay and Shieldaig are mostly bipolar but
with a significant presence of platform cores. The
fact that some platform cores are represented in the
Lealt Bay pebble-based assemblage may be due to
one of two possibilities: i) either the pebbles were
larger than in other typical quartz assemblages, or ii)
Mesolithic platform cores are so small that the size of
the pebbles is less of a problem.
The differences in the composition of the tool
group may be explained by differences in site activities, but most of the differences seem to be related to
chronology: i) the Kilmelfort Cave assemblage contains few tool types, and there are no microliths
these facts, combined with the presence of many flint
'button-shaped' scrapers suggest a Bronze Age date
for the quartz assemblage and parts of the flint assemblage; and ii) Lealt Bay and Shieldaig both contain quartz microliths (Lealt Bay: 1; Shieldaig: 18)
and a wide variety of other quartz tool types, suggesting a Mesolithic date for most of the quartz material.
Diagnostic elements in the flint assemblage from
Shieldaig indicate that this assemblage may be contemporary with that from Gleann Mor, Islay (Finlayson and Mithen 1997), which was 14C-dated to
around 7000 BP, though it must be adlnitted that the
chrono-typological sequence for the Scottish Mesolithic is far from resolved (cf. Woodman 1989). Further work towards publication of the whole Shieldaig
assemblage and re-publication of the Kilmelfort cave
assemblage is in progress.
THE THREE ASSEMBLAGES OF THE PILOT
PROJECT
There are some similarities between the three assemblages of the pilot project, but just as many differences. Contrary to 'classic' quartz assemblages from,
for example, Shetland (Calder 1956; Hamilton 1956;
Whittle 1986), all three assemblages form parts of
larger lithic collections contammg substantial
amounts of flint. The quartz at Kilmelfort Cave is
homogeneous milky quartz, the quartz at Lealt Bay is
milky quartz and rock crystal, and there are two types
of quartz in the Shieldaig assemblage: typical 'vitreous' milky quartz and rock crystal, and a variety with
a 'greasy' lustre. These raw material varieties all
have different flaking properties. The quartz at Kilmelfort Cave and Shieldaig is mainly vein quartz,
whereas the quartz at Lealt Bay is pebble quartz.
It was possible to define some technological differences between the three assemblages, mostly associated with the use of different raw materials.
Quartz from a pebble source usually results in a
technology heavily dominated by bipolar technique,
with vein quartz resulting in a technology based on
platform technique, but with bipolar technique applied to otherwise exhausted platform cores.
The numbers of the individual debitage categories differ from assemblage to assemblage, mainly as
a result of differences in recovery policies (at
QUARTZ-BASED RESEARCH PROSPECTS
AND PROBLEMS
Due to its general properties, quartz is undoubtedly
more demanding to examine than, say, flint or chert,
but if quartz assemblages are excavated and examined as meticulously as assemblages in other raw
47
_-._1
Quartz Technology in Scottish Prehistory
materials, they offer similar potential for information
on prehistoric societies.
nologies, in particular combined with the chafnes
operatoire approach (Eriksen 2000).
Raw Materials
When dealing with quartz, it is important to realize
that this resource encompasses a group of closely
related, but more or less distinctive, materials. In the
pilot project milky quartz, rock crystal and 'greasy'
quartz are three varieties with different appearances
and different flaking properties, and they would
probably have been perceived by prehistoric people
as three different materials with different functional
and perhaps symbolic values.
Distribution analysis
Analysis of quartz assemblages may include distribution analyses to either: a) shed light on intra-site settlement organization; or b) test the chronology of the
site and/or the quartz assemblage. The main precondition for carrying out a distribution analysis is the
existence of a standard grid system with square grids
of maximum 1m2 (or preferably 0.25m2 or less). In
Scottish archaeology, quartz has rarely (if ever) been
used in distribution analyses. This is not due to
quartz being less suited to this kind of investigation;
it is more a combination of traditional low expectancies and recovery/documentation policies. Most
known quartz assemblages are from excavations, but
unfortunately many of these were documented in
ways inappropriate for distribution analysis (as, for
example, in the case of the three assemblages studied
in the pilot project). Today, it is common practice in
Scotland to excavate Stone and Bronze Age sites
with accurate recording of finds to either standard
grids or well-defined contexts, and many assemblages excavated in the last 15-20 years are wellsuited for analysis of settlement organization.
Typology
It is well known that it can be difficult to identify
retouch, and thereby tools, in quartz. Another problem is that, due to the different flaking properties,
many quartz blanks and tools have a different appearance to that of artefacts in flint-like silica. For
this reason some scholars have attempted to develop
a typology solely for quartz artefacts (Broadbent
1979; Callahan 1987; Lehane 1986). Unfortunately,
the introduction of a separate quartz typology prevents comparison of quartz tools and those in other
raw materials, and we believe this practice should be
discouraged. As demonstrated in the lithic analyses
already undertaken for this proj ect (Ballin 2001), it is
quite practicable to apply the same typology to all
raw materials. A general difficulty with Scottish
lithic research in particular has been the lack of consensus on one typologyi, which is a problem in need
of serious attention.
Dating
As demonstrated in the pilot project report, quartz is
just as useful for dating purposes as other lithic materials, as long as diagnostic artefacts or attributes are
present. At a general level, dating quartz assemblages
by the use of diagnostic artefacts or technological
attributes is hampered by the lack in Scotland of an
adequate typo-chronological framework for the
Mesolithic and Neolithic periods. It must be an important aim in Scottish Stone Age research to improve the chronological framework by identifying
new diagnostic types or diagnostic technological attributes.
Technology
The common use of bipolar technique in connection
with reduction of quartz material, combined with the
fact that the bipolar technique was not generally
recognized as such until the mid-1980s, means that
most older presentations of quartz assemblages are
heavily flawed (Table 1). Bipolar material was
classified as, inter alia, opposed-platform cores,
wedges and scrapers, which firstly creates a bias in
inferences about the assemblages and activities
associated with them, and secondly makes
comparison based on, or including, older literature
highly problematic. With bipolar material classified
correctly, it is possible to achieve a detailed picture
of prehistoric lithic technologies, in particular
PROPOSALS FOR THE MAIN PROJECT
To satisfy the aims set up by this project, a substantial number of representative quartz assemblages
must be examined and compared.
48
Saville and Ballin
Chronology
It is most likely that the appearance of quartz artefacts will vary over time. To detect this variation it
will be necessary to examine assemblages from all
quartz-using periods and phases. To test whether the
observed variation is in fact chronological (that is,
not due to differences in territoriality, raw material
availability, or site economy/activities), it will be
vital to include in the project as many different assemblages as possible from each period and phase.
Activities
As the specific subsistence economy and activities at
a particular site may influence the composition of its
assemblage, the quartz project should include assemblages from different types of site. Relevant site
types are: i) open-air sites (i.e. sites not associated
with more substantial dwelling structures); ii) house
sites; iii) ritual sites; and iv) burial sites. One could,
for example, envisage a marked difference between
assemblages from sites of type i/ii and iii/iv. The
composition of assemblages from the former group
of sites might be dominated by utilitarian choices,
and they might include a certain amount of production refuse. The composition of assemblages from the
latter group of sites might be dominated by choices
involving the symbolic value of artefact raw materials and style, and might contain less refuse, fewer
blanks, and more tools.
Territoriality
To test whether territoriality is a causal factor in the
morphology and technology of quartz artefacts (stylistic variation), it is imperative that all regions of the
Scottish quartz province are represented by suitable
assemblages. Stylistic variation (Wobst 1977; Wiessner 1983; Gebauer 1988) is usually associated with
the distinction of social territories (Clark 1975, 12),
and, for example, Scandinavian research (Bruen 01sen & Alsaker 1984; Andersen 1983; 1995a; 1995b)
suggests social territories to be identical to specific
sets of biotopes ('economical zones') delimited by
significant topographical features (fjords, rivers,
mountain ranges or water divides). For this reason,
the quartz proj ect should attempt to cover as many
economic zones as possible.
CONCLUSION
The demands on the project's selection process are
extensive, and possibly the best answer, in terms of a
project structure, is a general survey of the 'Scottish
quartz province'. This 'geographical concept corresponds roughly to N and NW Scotland. The proj ect
should attempt to cover:
Resources
The effect of chronology and territoriality on the
variation within and between quartz assemblages is
as yet unproven, but it is fairly certain that some of
the observed variation is due to raw material differences. The following points have been established: i)
throughout the Scottish quartz province different
types of quartz were used; ii) within the individual
quartz types, some varieties are better suited for the
manufacture of flaked tools than others; and iii) in
some assemblages quartz was supplemented by other
materials, or it was itself a supplement. In the attempt
to analyse the influence of raw material variability on
assemblage variability, quartz assemblages will be
selected from all geological zones of the Scottish
quartz province (mainly N and NW Scotland). Assemblages from central and southern Scotland rarely
yield more than a few individual pieces of quartz.
•
assemblages from all Stone and Bronze Age
phases (as well as the Early Iron Age?)
•
assemblages from as many biotopes/economic
zones as possible
•
assemblages in different types and sub-types of
quartz, and assemblages in which quartz is supplemented by other raw materials
•
•
assemblages from all main geological zones
assemblages from different site types
Assemblages should be examined from museums
within the 'quartz zone' and from the major collections in Edinburgh and Glasgow. The project should
be structured in three stages:
•
49
an initial planning stage, during which it is attempted, via the archaeological literature and
communication with the relevant museums, to
list and locate relevant assemblages. At this
Quartz Technology in Scottish Prehistory
stage a system is established for the on-site classification and cataloguing of assemblages. This
includes preparation of database schemas for direct storage of lithic information.
•
a 'travel stage', during which all relevant assemblages are examined, combined with some
sample collection of natural quartz from relevant
locations.
•
a final writing-up stage - the assemblages are
compared and a publication is prepared, including the commissioning of extensive illustration.
Ballin, T.B. forthcoming b. The lithic assemblage. In:
P. Ashmore: (ed.), Callanais, Lewis, Western
Isles.
Ballin, T.B. forthcoming c. The quartz assemblage.
In: H. Moore and G. Wilson (ed.), The Bayanne
Project, Shetland.
Broadbent, N. 1979. Coastal Resources and Settlement Stability. A Critical Study of a Mesolithic
Site Complex in Northern Sweden. Uppsala:
AUN3.
Bruen Olsen, A., & Alsaker, S. 1984. Greenstone and
diabase utilization in the stone age of western
Norway: technological and socio-cultural aspects
of axe and adze production and distribution. Norwegian Archaeological Review 17: 71-103.
Calder, C.S.T. 1956. Report on the discovery of numerous stone age house-sites in Shetland. Proceedings of the Society of Antiquaries of Scotland 89: 340-397.
Callahan, E. 1987. An evaluation of the lithic technology in Middle Sweden during the Mesolithic
and Neolithic. Uppsala: AUN 8.
Clark, J.G.D. 1975. The Earlier Stone Age Settlement
of Scandinavia. Cambridge: Cambridge Archaeological Press
Clarke, A. 1999. The flaked quartz. In: O. Owen and
C. Lowe (ed.), Kebister: The-Four-ThousandYear-Old Story of One Shetland Township. Edinburgh: Society of Antiquaries of Scotland,
Monograph 14. pp 164-166
Coles, J.M. 1983. Excavations at Kilmelfort Cave,
Argyll. Proceedings of the Society of Antiquaries
of Scotland 113: 11-21.
Eriksen, B.V. 2000. 'Chaine operatoire' - den operative proces og kunsten at tcenke som en flinthugger. In: B.V. Eriksen (ed.), Flintstudier-en handbog i systematiske analyser af flintinventarer.
Aarhus. pp 101-126.
Finlayson, B. 1992. Carding Mill Bay: lithic report.
Glasgow Archaeological Journal 17: microfiche
M52-M58.
Finlayson, B. 1996. Worked flint and quartz. In:
R.P.J. McCullogh and R. Tipping (ed.), The
Lairg Project 1988-1996. The Evolution of an
Archaeological Landscape in Northern Scotland.
Edinburgh: AOC Monograph 3. pp 132-139.
Finlayson, B. forthcoming. The flint and quartz artefacts. In: 1. Armit forthcoming: Excavations at
ACKNOWLEDGEMENT
The authors are grateful to Historic Scotland for being the main grant-aider of the pilot project, which
was managed through the National Museums of
Scotland. In particular the authors wish to thank Mr
Patrick Ashmore at Historic Scotland for his interest
and support.
BIBLIOGRAPHY
Andersen, S.H. 1983. Stenalderfolk fra Tybrind Vig,
Vestfyn. Fynske Minder: 7-28.
Andersen, S.H. 1995a. Coastal adaptation and marine
exploitation in Late Mesolithic Denmark - with
special emphasis on the Limfjord Region. In: A.
Fischer (ed.), Man and Sea in the Mesolithic.
Coastal Settlement Above and Below Present Sea
Level. Oxford: Oxbow Monographs 53. pp 41-66.
Andersen, S.H. 1995b. Erteb011e trappers and wild
boar hunters in eastern Jutland.
A survey.
Journal ofDanish Archaeology 12: 13-59.
Ballin, T.B. 2000. Classification and description of
lithic artefacts. A discussion of the basic lithic
terminology. Lithics 21: 9-14
Ballin, T.B. 2001. Quartz Technology in Scottish
Prehistory: Pilot Project. Edinburgh: unpublished report commissioned by Historic Scotland
and the National Museums of Scotland.
Ballin, TB. forthcoming a. The lithic assemblage.
In: 1. Shepherd & A. Shepherd (ed.), Rosinish,
Benbecula, Outer Hebrides.
50
Saville and Ballin
Woodman, P.C. 1989. A reView of the Scottish
Mesolithic: a plea for normality! Proceedings of
the Society ofAntiquaries of Scotland 119: 1-32.
B.C. Oxford: British Archaeological Reports 33. pp
221-231.
Walker, M. 1973: Archaeological excavation of a
microlithic assemblage at Shieldaig, Wester
Ross, Scotland, 24/iii/73-6/iv/73: preliminary report. Unpublished interim report (copy in the
NMRS).
Whittle, A. 1986. Scord of Brouster. An Early Agricultural Settlement on Shetland. Oxford: Oxford
University Committee for Archaeology, Monograph 9.
Wickham-Jones, C. 1990. Rhum. Mesolithic and
Later Sites at Kinloch. Excavations 1984-86. Edinburgh: Society of Antiquaries of Scotland,
Monograph 7.
Wiessner, P. 1983. Style and Social Information in
Kalahari San Projectile Points. American Antiquity
48: 253-276.
W obst, H.M. 1977. Stylistic Behavior and Information
Exchange. Museum ofAnthropology, University of
Michigan. Anthropological Papers 61: 317-342.
the Neolithic site of Eilean Domhnuill, Loch
Olabhat, North Uist.
Finlayson, B., & Mithen, S. 1997. The microwear
and morphology of microliths from Gleann Mor.
In: H. Knecht (ed.), Projectile Technology. New
York: Academic Press. pp 107-129.
Gebauer, A.B. 1987. Stylistic analysis. a critical review of concepts, models, and interpretations.
Journal ofDanish Archaeology 6: 223-229.
Haggarty, A. 1991. Machrie Moor, Arran: recent excavations at two stone circles. Proceedings of the
Society ofAntiquaries of Scotland 121: 51-94.
Hamilton, J.R.C.1956. Excavations at Jarlshof, Shetland. Edinburgh: Ministry of Works, Archaeological Reports, I.
Hedges, J.W. 1986. Bronze Age structures at Tougs,
Burra Isle, Shetland. Glasgow Archaeological
Journal 13: 1-43.
Lehane, D. 1986. Attribute analysis [Tougs, Burra
Isle, Shetland]. Glasgow Archaeological Journal
13: microfiche M3-M8.
Lindgren, C. 1998. Shapes of quartz and shapes of
minds. In: L. Holm & K. Knutsson (eds), Third
Flint Alternatives Conference at Uppsala. Uppsala: Uppsala University, Department of Archaeology and Ancient History, Occasional Papers in
Archaeology 16. pp 95-103.
Mercer, J. 1968. Stone tools from a washing-limit
deposit of the highest Postglacial transgression,
Lealt Bay, Isle of Jura. Proceedings of the Society ofAntiquaries of Scotland 100: 1-46.
Mercer, J. 1971. A regression-time stoneworkers'
camp, 33 ft. OD, Lussa River, Isle of Jura. Proceedings of the Society of Antiquaries of Scotland 103: 1-32.
Mercer, J. 1972. Microlithic and Bronze Age camps,
75-26ft OD, N Carn, Isle of Jura. Proceedings of
the Society ofAntiquaries of Scotland 104: 1-22.
Saville, A. 1994. Exploitation of lithic resources for
stone tools in earlier prehistoric Scotland. In:
N.Ashton & A. David (ed.), Stories in Stone.
London: Lithic Studies Society Occasional Paper
4. pp 57-70.
Simpson, D.D.A. 1976. The Later Neolithic & Beaker
Settlement Site at Northton, Isle of Harris. In: C.
Burgess & R. Miket (eds.), Settlement and Economy in the Third and Second Millenia
ENDNOTES
In the Scottish archaeological literature it is possible
to find up to four different definitions of the same
lithic type (e.g. 'chips'), which hinders comparison
of assemblages (Ballin 2000).
1
Saville
Archaeology Department
National Museums of Scotland
Edinburgh
EH1 IJF
Ballin
Lithic Research
Banknock Cottage
Denny
Stirlingshire
FK65NA
51