This article was downloaded by: [Tufts University]
On: 16 March 2012, At: 10:17
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,
UK
Scientific Studies of Reading
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/hssr20
Reading Fluency and Its
Intervention
Maryanne Wolf & Tami Katzir-Cohen
Available online: 19 Nov 2009
To cite this article: Maryanne Wolf & Tami Katzir-Cohen (2001): Reading Fluency and
Its Intervention, Scientific Studies of Reading, 5:3, 211-239
To link to this article: http://dx.doi.org/10.1207/S1532799XSSR0503_2
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/termsand-conditions
This article may be used for research, teaching, and private study purposes.
Any substantial or systematic reproduction, redistribution, reselling, loan,
sub-licensing, systematic supply, or distribution in any form to anyone is
expressly forbidden.
The publisher does not give any warranty express or implied or make any
representation that the contents will be complete or accurate or up to
date. The accuracy of any instructions, formulae, and drug doses should be
independently verified with primary sources. The publisher shall not be liable
for any loss, actions, claims, proceedings, demand, or costs or damages
whatsoever or howsoever caused arising directly or indirectly in connection
with or arising out of the use of this material.
SCIENTIFIC STUDIES OF READING, 5(3), 211–239
Copyright © 2001, Lawrence Erlbaum Associates, Inc.
Reading Fluency and Its Intervention
Downloaded by [Tufts University] at 10:17 16 March 2012
Maryanne Wolf and Tami Katzir-Cohen
Center for Reading and Language Research
Tufts University
This 3-part article represents an effort to confront 3 large lacunae in the research on
reading fluency: definition, component structure, and theory-based intervention. The
1st section describes several historical approaches to fluency and the components of
fluent reading that are implicit in these approaches. We then present our own developmental- and component-based definition of reading fluency. In the 2nd section we
discuss how different types of current fluency interventions correspond to particular
components in fluency’s structure and to particular phases of its development. The
last section presents an overview of an experimental fluency program that attempts to
address multiple components in the development of fluent reading. Finally, we argue
that increased exploration of the issues surrounding fluency and comprehension will
contribute to our understanding of both reading development and dyslexia subtypes.
The history of fluency research in the field of reading might best be characterized as
intellectually spasmodic: There are periods of great effort and creativity, followed
by fallow periods of relative disinterest. In 1983 fluency was described as the “most
neglected” reading skill (Allington, 1983). In 1997 Lyon and Moats summarized
the state of intervention research with a call to refocus attention on fluency:
Improvements in decoding and word-reading accuracy have been far easier to obtain
than improvements in reading fluency and automaticity. This persistent finding indicates there is much we have to learn about the development of componential reading
skills and how such skills mediate reading rate and reading comprehension. (p. 579)
The recent attention to fluency and fluency intervention is the result, we believe, of a convergence of three major factors. First, the systematic research on the
role of phonological processes in reading failure and intervention has proven both
Requests for reprints should be sent to Maryanne Wolf, Center for Reading and Language Research,
Miller Hall, Tufts University, Medford, MA 02155. E-mail: [email protected]
Downloaded by [Tufts University] at 10:17 16 March 2012
212
WOLF AND KATZIR-COHEN
highly successful and insufficient in dealing with the heterogeneity of reading disabilities and the complexity of reading breakdown—particularly in the area of fluency. (For a recent comprehensive review, see Meyer & Felton, 1999; also see
Breznitz & Share, 1992; Rashotte & Torgesen, 1985; Torgesen et al., 1999;
Torgesen, Rashotte, & Wagner, 1997; Young & Bowers, 1995.) The wish to address the needs of children who do not completely respond to phonological-based
treatment is a motivating impulse in turning to additional explanatory principles
like fluency (Torgesen, Rashotte, & Alexander, in press).
The second factor is an increased awareness of the multiple underlying sources
that can contribute to or impede fluency development. The concerted effort to understand the predictive ability of naming speed in reading failure is an example of
this focus (see reviews in Wolf & Bowers, 1999, 2000), as are the extensive studies
that explore a range of time-related deficits in children with reading disabilities in
various perceptual and motor areas (see reviews in Farmer & Klein, 1995;
Nicolson & Fawcett, 1994; Stein, in press; Waber, in press; Wolf, Bowers, &
Biddle, 2000; Wolff, 2000).
The third factor is related to naming-speed research and involves cumulative evidence from subtype research in developmental dyslexia. A growing body of work
demonstrates that there are discrete groups of children with reading disabilities who
can be characterized by single deficits in either naming speed or phonological processes or combined deficits in both areas (Badian, 1996; Compton, Chayna,
DeFries, Gayan, & Olson, in press; Levy, in press; Lovett, 1987; Lovett, Steinbach,
& Frijters, 2000; Manis, Doi, & Bhadha, 2000; Wolf, Bowers, & Biddle, 2000). This
conceptualization, known as the Double-Deficit Hypothesis, has several fluency-related implications (Wolf & Bowers, 1999, 2000). First, children with single naming-speed deficits are frequently difficult to diagnose in early primary years, but go
on to develop fluency and comprehension problems by the end of Grade 3. Second,
children with both phonological and naming-speed deficits are consistently found to
possess the most severe problems in reading and reading fluency. Third, and most
important, this work provides a theoretical rationale and foundation for intervention
that specifically addresses issues of speed of processing and fluency. Until very recently, intervention was largely directed to treatment for phonologically based decoding problems. Children with either single processing-speed deficits or combined
deficits would be only partially served by such a focus, thus augmenting the ranks of
children who do not respond to treatment.
Represented by this special issue of Scientific Studies of Reading, the new,
evolving work on reading fluency and fluency-based intervention has much to
learn from the 2 decades of work that precede it in theoretically based, phonologically focused treatment. Specifically, there must be a greater concentration of effort (a) in defining what we mean by fluency, (b) in charting the development of its
component structure, (c) in understanding the breadth and nature of processing-speed and fluency deficits in reading subtypes, and (d) in applying this knowl-
Downloaded by [Tufts University] at 10:17 16 March 2012
READING FLUENCY
213
edge to the development of intervention programs. In the last years we have
directed considerable attention to understanding the third area of concern, the processes underlying naming speed and other rate-related deficits in developmental
dyslexia. We view this article as a working paper on the other three of these areas.
The first section describes several historical approaches to fluency and the components of fluent reading implicit in these approaches. We then provide our own
evolving definition. In the second section we discuss how different types of current fluency interventions correspond to particular components in fluency’s structure and particular phases of its development. Finally, the last section presents an
overview of our own first efforts to construct a theory-based, comprehensive fluency program that attempts to address multiple components in fluent reading.
TOWARD A DEFINITION OF FLUENCY AND AN
UNDERSTANDING OF ITS COMPONENT STRUCTURE
We view this article as a working paper because there are still no consensual definitions of what is meant by fluency and what its relation might be to the subset of
time-related terms most frequently related to it (e.g., automaticity, speed of processing, reading rate/speed, and word recognition rate/proficiency). In this section,
our goals are foundational for the rest of the article and for our future work. We review several historical and current definitions and approaches to fluency as the basis for our own definition and as the context for examining the underlying components of fluency. These and other efforts toward gaining greater clarity about
fluency will push forward not only our understanding of the component structure of
reading fluency, but also the development of better assessment tools and more comprehensive interventions.
BACKGROUND: APPROACHES AND DEFINITIONS
Early Research
Perhaps predictably, any historical review of perspectives on fluency begins with
the work of William MacKeen Cattell and Sir Edmond Huey. Cattell (1886), a
19th-century experimental psychologist, found that letters and words are named
faster than other symbolic categories, such as colors, and other more concrete, semantic categories, such as pictured objects. Cattell was the first researcher to highlight the automatic-like rates of recognition achieved in letter naming and word
reading, with words read as fast as letters and reading speed increased when semantic and syntactic information are provided (as in sentences). This set of findings was
completely replicated almost a century later by Doehring (1976).
214
WOLF AND KATZIR-COHEN
Downloaded by [Tufts University] at 10:17 16 March 2012
Also a century ahead of his time, Huey (1908/1968) presciently described what
most cognitive scientists refer to as automaticity in reading (LaBerge & Samuels,
1974; Logan, 1988). More specifically, according to Huey, the development of
fluent reading involves the steady accumulation and synthesis of “more and more
complex constituent acts as these are progressively welded together by practice”
(p. 105). An integral aspect of this synthesis to Huey was the development of a rate
of processing, which through “repetition progressively frees the mind from attention to details, makes facile the total act, shortens the time, and reduces the extent
to which consciousness must concern itself with the process” (p. 65).
Information-Processing Models
Similar to Huey’s notions, more recent perspectives on fluency in the last quarter of
the 20th century emphasize automatic, effortless rates of processing that allow the
reallocation of attention. For example, LaBerge and Samuels (1974), whose work
ushered in an era of renewed attention to fluency, proposed a model of automaticity
with principles derived from information-processing theory (e.g., Norman, 1968;
Posner, Lewis, & Conrad, 1972). According to their model, reading becomes increasingly fluent as the result of the development of automaticity of the subskills:
“When one describes a skill at the macro level as being automatic, it follows that the
subskills at the micro level and their interrelations must also be automatic” (p. 295).
The first stage of these subskills in their model involves the visual code and the unitization of visual stimuli. These stimuli may include letters, spelling patterns,
words, and highly frequent word groups (e.g., high school). With exposure and
practice, the visual features in stimuli like letters become unitized and then perceived as a single unit. As these units accumulate and letter perception becomes increasingly automatic, attention to early visual coding processes decreases. This decrease allows attentional resources to be reallocated to other areas, such as the
semantic (or meaning) code.
The earlier physiological work of Donald Hebb (1949) undergirds the concept
of unitization used here. An example from the visual system is helpful. When an
unknown visual stimuli is first seen by the retina, there is an activation in the visual cortex of multiple individual cells. These cells correspond to each aspect of
what the retina sees and are responsible for very specific types of information
(e.g., curved lines, diagonals, etc.). After multiple exposures to the same stimulus, the individual cells in the visual areas become a working unit, or cell assembly. These unified groups of neurons learn to work together in synchrony so that
recognition of frequently viewed stimuli such as letters becomes so efficient, it is
virtually automatic. Hebb argued that the ultimate result of these cell assemblies
in the visual area is a reservoir of mental representations of practiced, frequently
viewed visual stimuli.
Downloaded by [Tufts University] at 10:17 16 March 2012
READING FLUENCY
215
LaBerge and Samuels (1974) argued that a similar principle is found at more
complex visual levels (and also in other modalities, such as phonological representations). For example, the orthography of every language permits particular, frequently seen spelling patterns or letter combinations (e.g., th in English, schr in
German). With sufficient exposures and practice, these orthographic patterns also
become quickly recognized. A similar principle, they believed, may apply to many
frequently read words, although LaBerge and Samuels raised questions about how
these larger units are perceived and processed by different readers (see, e.g.,
Gough, 1972).
The primary contributions of the LaBerge and Samuels model for work in fluency (McCormick & Samuels, 1979) include several principles about how we can
process written information at highly rapid rates: (a) the concept of unitization; (b)
the key notion that, with the increased speed of lower level subskills, attention can
be reallocated elsewhere; and (c) the more reading-specific concept that with
automaticity, attention can be shifted from lower level decoding to higher level
comprehension skills. A fundamental premise of our work is based on their conceptualization that macrolevel fluency is based on the automaticity of microlevel
subskills and their connections.
Consciously building on the work of Cattell, Hebb, and LaBerge and Samuels,
Doehring (1976) provided one of the most systematic studies of the development
of fluency in reading subskills in children from kindergarten until the end of high
school. In his seminal monograph “Acquisition of Rapid Reading Responses,”
Doehring studied the acquisition of rates of processing in symbols (colors and objects), letters, letter combinations, words, random word sequences, and sentences.
In so doing, he sought to map
the relative course of acquisition of skills for processing the graphological features of
letters, the orthographic regularities of letter combinations, the semantic features of
words, and the semantic-syntactic constraints of word sequences. These are the skills
that must be mastered beyond the level of simple accuracy to the point where accurate
processing becomes rapid enough to be classified as fluent reading. (p. 2)
The data in this study went well beyond a replication of Cattell’s first findings about
the “time it takes” human beings to acquire automatic-like rates in various processes; to our minds, Doehring described the first data in a developmental, component-process approach to reading fluency.
Perfetti’s (1977, 1985) verbal efficiency theory represents another highly influential perspective in the history of fluency-related research. Like LaBerge and
Samuels (1974) and Doehring (1976), Perfetti worked within an information-processing approach. Perfetti did not refer to the word fluency, and he did not equate
verbal efficiency with fluency. Rather, Perfetti discussed fluency-related concepts
in terms of an overall account of reading ability in which verbal efficiency theory
Downloaded by [Tufts University] at 10:17 16 March 2012
216
WOLF AND KATZIR-COHEN
played a major role. Within this context, fluency is closest to what Perfetti and
what Jackson and McClelland (1979) called “effective reading speed,” construed
as the outcome of comprehension accuracy and reading speed (words per minute).
Perfetti’s (1985) account of verbal efficiency was a theoretical effort to explain
how “individual differences in reading comprehension are produced by individual
differences in the efficient operation of local processes. The local processes are
those by which temporary representations of text are established” (p. 100; here he
included orthographic, phonological, and semantic processes). Some of the critical
components in Perfetti’s model that contributed to the efficiency of local processes
were (a) general symbol activation and retrieval, (b) recognition processes, (c) lexical access and retrieval, and (d) working memory.
Learning and practice also were thought to play a pivotal role in the acquisition
of efficiency in the local processes. This efficiency was considered necessary to
construct high-quality mental representations, a central concept in the theory. The
quality of representations was hypothesized to affect both the further development
of efficiency (by enhancing the ease of lexical retrieval) and working memory:
“To the extent that these codes are retrieved rapidly and are high in quality, the system is efficient” (Perfetti, 1985, p. 118). When the underlying systems were efficient, the individual was considered able to free cognitive resources to focus on
higher level demands in reading, which was important for comprehension. The
converse was also predicted here. For example, Perfetti suggested that an inefficient system resulting in a slow rate of word recognition could obstruct the individual’s ability to hold large units of text in working memory, which, in turn would
affect comprehension and recall.
Important to work today, Perfetti (1985) stressed that the quality of representations and of each of the local processes, components, and their various forms of integration (depending on the type of reading task) were potentially rate-limiting
factors in reading acquisition. Furthermore, as Doehring (1976) suggested earlier,
different emphases would be placed on processing systems, depending on development. We emphasize LaBerge and Samuels’ (1974), Doehring’s (1976), and
Perfetti’s (1985) focus on the development of efficiency in lower levels of processing because this aspect of fluency development is frequently neglected in present
fluency interventions.
There were immediate and long-term effects of this model on reading and reading disabilities research, particularly concerning short-term memory and lexical
retrieval. Shankweiler and Crain’s (1986) influential article on reading disability
incorporated aspects of Perfetti’s theory by hypothesizing that the demands of orthographic decoding and limited working memory capacity contribute to comprehension difficulties of poor readers. Swanson (as cited in Johnston & Anderson,
1998) proposed that many poor readers have difficulty in reading because they fail
to establish effective, efficient visual–verbal connections, which may lead to difficulty in accessing or laying down long-term memory traces.
READING FLUENCY
217
Perfetti’s description of the role of an efficient lexical retrieval process in reading development has influenced much of our own work on word-retrieval and its
relation to reading disability (Wolf, 1982, 1991; Wolf & Goodglass, 1986; Wolf &
Obregon, 1992). Perfetti used knowledge about retrieval and all components in
verbal efficiency theory as ways to explicate individual differences in reading
comprehension. In doing so, Perfetti’s work provides the figure–ground perspective from which to view fluency today—as a means to reading comprehension.
Downloaded by [Tufts University] at 10:17 16 March 2012
Rauding Theory
Carver (1991, 1997), who also focused on the links between fluency and comprehension, introduced an emphasis on the different purposes of reading and their respective rates. According to Carver, most reading is done in the rauding mode,
which involves the fastest rate at which an individual can successfully understand
complete thoughts in each sentence.
Carver’s work includes a model that depicts four levels of factors that affect the
achievement of optimal reading rate and accuracy. Some of these factors are age,
teaching variables, “aptitude” factors (i.e., verbal knowledge, decoding processes,
and cognitive speed), decoding speed, and naming speed (see Spring & Capps,
1974). Carver reintroduced the complex concept of cognitive speed to reading fluency theory, a line of research (Kail, 1991, 1992) in reading and in other areas of
cognitive development (see also Breznitz, in press; Marcus, 1997).
Connectionist Models
The last body of research with implications for fluency in this review involves the
evolution of connectionist models of reading. These models are profoundly different from earlier information-processing models, particularly in the role of retrieval
and unitization. Connectionist models emphasize the continuous, distributed interaction of phonological, orthographic, syntactic, and semantic-processing codes
during word recognition (Adams, 1990; Caramazza, 1997; Foorman, 1994;
Seidenberg & McClelland, 1989). There are no retrieval mechanisms, as such;
rather, all codes are computed for every word. In these powerful models, the
amount and level of activation in each system and the speed of processing within
and among these systems are products of internal factors (e.g., learning) and external factors (e.g., the consistency of correspondence between a given letter and a
sound pattern). These models add to any discussion of fluency the importance of
accounting not only for intra- and intersystem factors in the learner (the emphasis in
earlier information-processing models), but also for such linguistic features as frequency, regularity, and the amount of processing-code connections among individ-
218
WOLF AND KATZIR-COHEN
ual words. The latter factor, we believe, has special importance for learning and, by
extension, for the design of fluency interventions.
Downloaded by [Tufts University] at 10:17 16 March 2012
Current Research and Definitions
Throughout the period between LaBerge and Samuels (1974) and Carver (1997),
the general consensus was that fluent reading could be defined as “that level of
reading competence at which textual material can be effortlessly, smoothly, and automatically understood” (Schreiber, 1980, p. 177). Current research is largely consistent with this view. Indeed Meyer and Felton (1999), in a recent review of the literature, defined fluency in a similar manner as “the ability to read connected text
rapidly, smoothly, effortlessly, and automatically with little conscious attention to
the mechanics of reading such as decoding” (p. 284). Along similar lines, Hudson,
Mercer, and Lane (2000) defined fluency as “accurate reading at a minimal rate
with appropriate prosodic features (expression) and deep understanding” (p. 16).
Although this approach to fluency effectively captures the end goal of fluency—that is, effortless reading with good comprehension—it does not yet permit
ease in validation or in conceptualizing the components underlying fluency. For
example, Torgesen et al. (in press) argued that most previous fluency definitions
do not allow empirical validation; they preferred, therefore, the minimalist definition of “rate and accuracy in oral reading” used in curriculum-based assessment research (Shinn, Good, Knutson, Tilly, & Collins, 1992). The National Reading
Panel’s (2000) definition of fluency as “the immediate result of word recognition
proficiency” (pp. 3–5) allows the simple procedure of testing for a particular level
of proficiency in word recognition, just as Torgesen et al.’s view can be assessed
by performance on an oral reading measure that incorporates rate and accuracy
(e.g., the Gray Oral Reading Test; Wiederholt & Bryant, 1992).
Although we concur in principle with Torgesen et al.’s (in press) general admonition to use definitions that can be validated, their emphasis on oral reading and
the National Reading Panel’s emphasis on word recognition ignore the multiple
other dimensions of fluency, particularly lower level subskills, that were emphasized in earlier accounts. Few current approaches attempt to define fluency in
terms of either its component parts or its various levels of reading subskills—that
is, letter, letter pattern, word, sentence, and passage.
More to the point, almost all of the definitions until this time posit fluency as the
outcome of learned skills. Kame’enui, Simmons, Good, and Harn (in press) suggested a figure–ground shift for the field. In a broad-ranging article, they conceptualized fluency in a more developmentally encompassing manner both as the
development of proficiency in underlying lower level and component skills of
reading (e.g., phoneme awareness) and as the outcome of proficiency in higher
level processes and component skills (e.g., accuracy in comprehension; see also
Downloaded by [Tufts University] at 10:17 16 March 2012
READING FLUENCY
219
Logan, 1997). Such a perspective, we believe, has profoundly different implications for prevention, intervention, and assessment than the traditional view. For,
within a developmental perspective, efforts to address fluency must begin at the
beginning of the reading process—that is, during acquisition—not after reading is
already acquired. The importance of working proactively to prevent the development of difficult fluency problems is a major theme in the recent studies by
Torgesen et al. (in press) and by Kame’enui et al. (in press).
Berninger, Abbott, Billingsley, and Nagy (in press) took an equally multidimensional view with a systems approach to fluency. In their conceptualization,
fluency is influenced by (a) the characteristics of stimulus input (e.g., rate and persistence of a visual signal or speech signal), (b) the efficiency and automaticity of
internal processes (e.g., the development of phonological, orthographic, and morphological systems), and (c) the coordination of responses by the executive functions system. As we discuss later, Berninger is one of the few researchers to place
special importance on the role of morphological knowledge about words in facilitating the development of orthographic rate and overall fluency.
We regard Kame’enui’s (in press) developmental and Berninger’s (in press)
systems-analysis approaches as the context for our own working definition:
In its beginnings, reading fluency is the product of the initial development of
accuracy and the subsequent development of automaticity in underlying
sublexical processes, lexical processes, and their integration in single-word
reading and connected text. These include perceptual, phonological, orthographic, and morphological processes at the letter, letter-pattern, and word
levels, as well as semantic and syntactic processes at the word level and connected-text level. After it is fully developed, reading fluency refers to a level
of accuracy and rate where decoding is relatively effortless; where oral reading is smooth and accurate with correct prosody; and where attention can be
allocated to comprehension.
THE COMPONENT STRUCTURE OF FLUENCY
This developmental and component-based definition of fluency poses a difficult
challenge for empirical validation; at the same time it provides a foundation for the
specification of processes, skills, and factors that underlie or influence reading fluency and that can guide assessment and validation efforts. In this subsection we unpack the previously discussed definitions and approaches in terms of the components that are implied as necessary in the development of fluency.
Most researchers describe three major types of processes as prominent in fluency’s development: orthographic, phonological, and semantic. Berninger et al.
(in press) and Adams (1990) added emphases on morphological and syntactic
Downloaded by [Tufts University] at 10:17 16 March 2012
220
WOLF AND KATZIR-COHEN
knowledge systems. For example, Adams proposed that both orthographic and semantic information are necessary for learning morphological knowledge about the
patterns and roots that make up many words (e.g., derivational and compound
words; Latin- and Greek-derived roots). The direct learning of these morpheme
patterns enhances vocabulary acquisition, fluency, and reading comprehension.
First, such knowledge provides rich semantic associations that can be applied to
acquiring new vocabulary words. Second, morphological knowledge makes orthographic chunks more visible, more familiar, and more quickly retrieved. Third, the
combination of rapid recognition of orthographic units and faster retrieval of
meaning facilitates comprehension. The working notion here (and in the intervention described later) is that one retrieves faster what one knows better, thus continuously emphasizing the connections that link orthographic, semantic,
phonological, and morphological systems. This conclusion is reinforced by recent
research showing that morphological awareness is significantly related to reading
fluency in third and fourth grades (Berninger et al., in press).
Multiple Components
The implications of a developmental, component-based conceptualization of fluency are far from simple. At a minimum such an approach implies three related premises: first, that multiple processes contribute to fluency development; second, that
multiple sources of dysfluency are possible; and third, that assessment and intervention should be based on knowledge about the development and breakdown of
these components. With regard to the first premise, the fluency-related processes
and components described in the research summarized here include lower level attention and visual perception, orthographic (letter-pattern) representation and identification, auditory perception, phonological representation and phoneme awareness, short-term and long-term memory, lexical access and retrieval, semantic
representation, decoding and word identification, morphosyntactic and prosodic
knowledge, and connected-text knowledge and comprehension. Berninger et al. (in
press) went further and argued that a systems approach to fluency would include the
executive function’s coordination of all these discussed internal-processing systems so that they perform smoothly and in synchrony. Wood, Flowers, and
Grigorenko (in press) added to the set of executive functions an emphasis on the
role of anticipatory facilitation for tasks that become fluent.
In other words, the unsettling conclusion is that reading fluency involves every
process and subskill involved in reading. We do not shy away from this conclusion;
rather, we wish to underscore it. Unlike reading accuracy, which can be executed
without utilizing some important reading components like semantic processes, we
argue that fluency is influenced by the development of rapid rates of processing in all
the components of reading. Kame’enui et al. (in press) would emphasize that reading
READING FLUENCY
221
fluency involves the development of accuracy and proficiency in every underlying
component. Researchers within connectionist approaches (Adams, 1990; Foorman,
1994; Seidenberg & McClelland, 1989) would stress the explicit linkages or connections among the orthographic, semantic, and phonological processes. Berninger et
al. (in press) and Adams (1990) would add the connections between
morphosyntactic knowledge and these other processes.
Downloaded by [Tufts University] at 10:17 16 March 2012
Multiple Sources of Dysfluency
The second premise is no less complex and is a logical outgrowth of the first premise and of time-allocation principles. Impairment in any one or more of the underlying processes could increase the processing time both within that process and in
reading outcome behaviors (Wolf, 1991; Wolf, Bowers, & Biddle, 2000). Further,
it appears that inefficiency not only can arise from accuracy or timing problems
within a processing system but also may be based on the coordination and integration of information across processes. An example of the latter is found in recent research by Breznitz (in press), who proposed that a temporal discrepancy (i.e.,
asynchrony) between the speed of processing of visual information and the speed of
processing of auditory information prevents the efficient cross-modal integration
necessary for visual–verbal processing in reading.
If the two premises of a developmental, component-based conceptualization of
reading fluency are true, a long and complex set of processes and variables could
be implicated when fluency is not attained. What is the evidence that impaired
readers break down across this continuum of processes? A full discussion of all the
evidence that might be brought to bear on this large question is well outside the
scope of this article. Meyer and Felton (1999), however, provided a first summary
of the existing research on explanations of dysfluency. They divided this research
into three major areas:
1. A breakdown in the lower level processes. At this level, dysfluent reading is
based on deficits in phonological processing and/or orthographic processing systems that affect the timing and coordination of these systems. More specifically,
deficits may arise from phonological, visio-spatial, and/or working memory processes. Breznitz (in press) used extensive evoked potential and behavioral data to
indicate that the largest contribution to slowed word-reading rate among dyslexic
readers is the slowed speed of processing in perceptual stages (see also Farmer &
Klein, 1995; Stein, in press; Tallal, Miller, & Fitch, 1993).
2. A failure to make higher order semantic and phonological connections between words, meaning, and ideas (Adams, 1990). In this level, dysfluent reading is
based on deficits that occur after perceptual identification has been completed.
One manifestation may involve the slowed retrieval of names, meaning, or both
222
WOLF AND KATZIR-COHEN
Downloaded by [Tufts University] at 10:17 16 March 2012
(see work on confrontation naming and receptive vocabulary by German, 1992;
Haynes, 1994; Segal & Wolf, 1993; for naming speed, see Wolf, Bowers, &
Biddle, 2000).
3. A breakdown in syntactic processing (Schreiber, 1980), with deficits exhibited in a lack of prosody and rhythm in oral reading and a lack of sensitivity to
prosodic cues. These deficits may become apparent only at the level of connected
text for reading but may be able to be assessed much earlier in speech development
(e.g., through measures of prosody and syntax).
Assessment and Intervention
Meyer and Felton’s (1999) organization of possible areas of deficit implies that
breakdown in fluent reading could originate at the sublexical, lexical, sentence, and
higher-conceptual integration levels. In future efforts we will examine evidence for
possible sources of dysfluency at the component level as well. There are significant
implications for diagnosis, as just sketched, that can utilize the application of many
existing measures. The most important implication, however, of Meyer and Felton’s
review—and of our working definition of fluency—concerns a new, more comprehensive approach to intervention. In the second section of this article we briefly examine how current fluency interventions correspond to the different phases of fluency development and to the range of component processes (e.g., orthographic) and
levels (e.g., sublexical, lexical) incorporated in this view of reading fluency.
THE CORRESPONDENCE BETWEEN CURRENT
FLUENCY INTERVENTION AND THE DEVELOPMENTAL,
COMPONENT-BASED VIEW OF READING FLUENCY
As discussed in work by Stahl and his colleagues (Stahl, Heubach, & Cramond,
1997), the oldest and most commonly used method for facilitating fluency is the repeated reading technique. Here readers are simply asked to read a passage of connected text (at a level appropriate to the learner) several times until a particular
reading rate (words per minute) is attained. At each juncture the reader is given further passages at that level until the optimal rate is reached. Dowhower (1994)
showed that practicing a series of passages helps build a large repertoire of quickly
identified words, a principle also emphasized in the development of fluency by
Torgesen et al. (in press) and by Stahl et al. (1997). Repeated reading can be assisted and thus necessitate oral reading (see Young, Bowers, & MacKinnon, 1996;
also see their use of prosodic assistance methods) or unassisted and involve silent
reading. First described in the literature by Dahl (1974) and named by Samuels
(1985), repeated reading methods were designed to increase reading rate for the
READING FLUENCY
223
Downloaded by [Tufts University] at 10:17 16 March 2012
given material and other similar materials and to improve comprehension. (As discussed earlier, this is based on the principle that comprehension can be allocated
more time when the rate of lower level decoding skills is increased.)
First, does repeated reading achieve its own specified goals? Second, how does
it fall on a developmental continuum of fluency’s growth, and what processes does
it address? In their careful review of the most well conducted repeated reading
studies, Meyer and Felton (1999) underscored the
complexity involved in assessing comprehensive gains from fluency instruction, and
suggest factors that need to be considered such as student age, reading level, instructional method type (types of repeated reading or rapid decoding of single words), and
cueing. Furthermore, the question of whether fluency and comprehension reciprocally influence one another is unanswered. (p. 294)
There is by now a considerable history of studies documenting the relation between fluency and comprehension (Fuchs, Fuchs, & Maxwell, 1988; Jenkins,
Fuchs, Espin, van den Broek, & Deno, 2000; Levy, in press; Levy, Abello, &
Lysynchuk, 1997; Torgesen et al., in press). As Meyer and Felton (1999) suggested, however, whether the reciprocal direction also exists between comprehension and fluency remains unresolved. Equally unresolved to our minds is whether
repeated reading or any existing fluency instruction significantly changes accuracy, rate, and comprehension of trained materials with transfer to other materials.
AN EXAMINATION OF FLUENCY STUDIES
To address this latter question, as well as issues about developmental stages and
components that different interventions emphasize, we have taken a second look at
many of the studies included in Meyer and Felton’s (1999) review and added several new fluency intervention studies. Depicted in Table 1, our framework examines the following variables: (a) the intensity of the intervention and the nature of
the gains (i.e., rate, accuracy, comprehension, transfer), (b) the developmental level
of reading subskills addressed (e.g., letter, letter-pattern, word, or passage) and the
grade of the individuals, and (c) the components of fluency focused on in the intervention. In the remainder of this section we summarize what we found about each
of the variables.
Intensity of Interventions
The length of interventions in the existing literature is extremely limited (most
range between 1 and 15 days). Most of the existing studies were intended to be
Downloaded by [Tufts University] at 10:17 16 March 2012
TABLE 1
Fluency Studies in Chronological Order: Gains Across Rate, Accuracy, Comprehension, and Transfer to New Materials
Length of
Intervention
Nature of
Intervention
Level and
Component
Emphasized
No. of
Rereadings
in
Intervention
Reading Rate
Gains
Accuracy
Gains
Comprehension
Gains
Transfer to
Untrained
Passages
Study
Population
N
O’Shea,
Sindelar,
and
O’Shea
(1985)
Average to
aboveaverage
readers
30 third
graders
read at
70 to
119
wpm
3 sessions
Half cued for
comprehension, half for
rate and
accuracy
Passage
Phonological
decoding/
rate
1, 3, and 7
Rate group
1–3: 24.7 wpm
3–7: 13.6 wpm
Comp group
1–3: 22.6 wpm
3–7: 10.3 wpm
+ For rate
group
Comprehension
gains only
from 1 to 3
rereading.
Comprehension
NA
Rashotte
and
Torgesen
(1985)
Nonfluent
readers
with
reading
disabilities
12 2–5
grades
read at
65 wpm
or less
on
Grade 2
level
7 days, 21
sessions in
small-group
instruction
settings
The three
conditions:
1. Repeated
reading within
session of
unrelated
stories.
2. Repeated
reading of
stories with
shared words.
3. Four
different
stories a day.
Passage
Word
recognition
Exposure
to words
4
1. 35.3 wpm
2. 33.0 wpm
3. 5.2 wpm
(all significant)
1. 2.52
error
decrease
2. 2.16
error
decrease
3. .82
error
decrease
(only 1&2
significant)
Comprehension
was high to
begin with,
therefore may
be ceiling effect
1. More
transfer when
degree of word
commonality
among stories
is high.
2. Word
commonality
had less effect
on accuracy or
comprehension.
Downloaded by [Tufts University] at 10:17 16 March 2012
Levy,
Nicholas,
and Kohen
(1993)
Poor
readers
Good
readers
24 poor
readers
24 good
readers
Grade
3–5
One or two
sessions
Read silently
stories, cross
out misspelled
words
Passage
Orthographic
patterns
4
Across grades
both groups
improved,
poor readers
showed more
improvement
+ For
both
groups
Stoddard,
Valcante,
Sindelar,
O’Shea,
and
Algozzine
(1993)
Reading
disabled
30
grades
4–5
(read at
least 70
wpm)
15
Half received
training in
intonation.
Half received
training in
sentence
segmentation.
Passage
Prosody/
Syntax
7
Reading rate
improved
significantly
for both
groups from 1
to 3 to 7
repetitions.
Intonation: 30
wpm;
Segmentation:
33 wpm
NA
Lemoine,
Levy, and
Hutchinson
(1993)
Study 1
Poor
readers
Grade 4
32
4 sessions
Blocked
presentation
of words
families or
scrambled
presentation
of word
families
Sublexical
Lexical
Morphology
34
repetitions
Blocked-150
msec;
Scrambled350 msec
Word
recognition
+ For
both
groups
Gains in both
groups
Only on word
recognition,
not on
comprehension
Improved from
1 to 3 readings,
but not to 7;
did not differ
among groups
NA
Number or
errors decreased
equally for the
two training
methods
No
generalization
(continued)
Downloaded by [Tufts University] at 10:17 16 March 2012
TABLE 1 (Continued)
Nature of
Intervention
Level and
Component
Emphasized
No. of
Rereadings
in
Intervention
Reading Rate
Gains
Accuracy
Gains
Comprehension
Gains
Transfer to
Untrained
Passages
Study
Population
N
Length of
Intervention
Lemoine et
al. (1993)
Study 2
Good
readers
Poor
readers
Grade 3
40
5 sessions
Readers were
trained on 50
regular words
or 50 irregular
words
Lexical
Word
recognition
25
Good readers:
Named both
equally quickly, decrease of
100 msec over
the first 4 trials. Poor readers: The difference between regular
and irregular
decreased over
trials
+ For
both
groups
+ For both
groups
Number of
repetitions in
training affected retention
for both
groups on both
sets of words.
Poor readers
benefited more
from repetitions, both in
rate gains as in
accuracy gains
Young,
Bowers,
and
MacKinnon
(1996)
Poor
readers
Grade 5
40
1 session
Modeling
Text practice
Both
Neither
Passage
3
Practice of
text: 36.9
wpm. Both: 29
wpm. Only
modeling: 17.7
wpm. Neither:
25.5 wpm.
+ In all
groups
approx.
2.5
Practice of text:
10, Modeling:
4, Both: 7,
Neither: 7
+ Only for the
repeated
reading group,
in accuracy
gains
Downloaded by [Tufts University] at 10:17 16 March 2012
Levy,
Abello, and
Lysynchuk
(1997)
Poor
readers
20 (FN)
20 (SN)
Grade 4
4 sessions
Repeated oral
reading of
isolated
content words
Lexical
Passage
Word
recognition
5 of isolated
words 4 of
stories
Fast RAN:
25.3, Slow
RAN: 33.5 (on
stories), Fast
RAN: 80 sec,
Slow RAN: 60
sec (on single
words)
Trained
stories
were read
more accurately,
no gains
from
rereading
in both
groups
+ For both
groups
Breznitz
(1997)
Study 1
Average
readers,
age
matched
(age 6.9).
Poordyslexic
readers
(age 9.1),
mostly
boys
52
2 sessions
Reading at
two rate
conditions:
Self-paced or
fast-paced +
auditory
masking
Passage
Rate
3 of parallel
texts:
Self
Fast
Self
Average
achievers: 9.13
sec; Poor
readers: 8.43
sec (both sig.
increase of
20%)
Both
groups
made the
least errors in the
masked,
fast-pace
condition.
Sig. difference
only for
poor
reader.
In fast pace both
groups gained
on comp. for
poor readers
more marked.
NA
Breznitz
(1997)
Study 2
Dyslexic
readers
(age 9.3)
23:
18 boys
5 girls
1 session
Reading at
two rates:
selfpaced, fast
paced
Passage
Rate
3 parallel
sets of
materials
0.1 letters per
second (sig.)
Decrease
in 7.9 errors per
passage
(sig.)
Recall increased
in 8% (sig.)
NA
Note. Most of these studies are from experimental studies; intervention was done in a 1:1 setting, except Rashotte and Torgesen (1985).
+ Word
recognition
+ Rate
+Comprehension
228
WOLF AND KATZIR-COHEN
brief, experimental explorations of what different versions of repeating reading
methods could do to facilitate accuracy and rate gains in oral reading. Several of the
studies measured comprehension, but in ways that varied widely from study to
study. It is not our intention to criticize the goals of these studies, but to indicate the
need for more intensive studies that allow a more meaningful evaluation of comprehension and transfer.
Downloaded by [Tufts University] at 10:17 16 March 2012
Developmental Level of Reading Subskills
Addressed by Interventions
The second general need that can be inferred from Table 1 is that the majority of the
experimental studies focus on passage-level reading with no attention to lexical-level, much less sublexical-level, instruction to increase rate. This means that
fluency was being addressed late in its development at a time when only a few underlying processes could be addressed (e.g., prosody). The few fluency studies that
did emphasize increased single-word rate showed mixed results for comprehension
(e.g., Tan & Nicholson, 1997, found gains). Levy is one of the few researchers to
conduct systematic, experimental rate studies on sublexical levels, although her
major interest in these studies was in increasing accuracy and rate of word recognition, not comprehension (Levy, in press; Levy, Bourassa, & Horn, 1999). Interestingly, in a recent study Levy et al. (1999) found little difference between single-word and connected-text repeated reading practice on the rate of reading.
Torgesen et al. (in press) concluded from studies like these that “the primary locus
of the repeated reading effect is on individual word reading efficiency.” Torgesen
et al. (in press) argued that this sight vocabulary is key to closing the large gap between the words read by average or good readers—estimated by Nagy and Anderson (1984) to be 1 million words a year—and those read by less skilled readers, estimated at 100,000 words.
Several questions emerge from such a conclusion: first, whether an emphasis
solely on word-level efficiency is sufficient to change comprehension; second,
whether increases in word-reading efficiency are enough to build a sufficiently
large sight vocabulary. As seen in Table 1, there are real gains that are possible in
oral reading accuracy and rate from various repeated reading methods at both the
connected-text and word levels. It appears reasonable to us that such gains in
word-reading rate—if accumulated over time within more intensive intervention
programs—could result in an expanded working sight vocabulary for children.
Whether there are concomitant gains in comprehension remains as unresolved for
this summary as for Meyer and Felton’s (1999).
These kinds of questions and the enormity of the existing vocabulary discrepancy between good and impaired readers have led some researchers, including
Torgesen et al. (in press), to conclude that intervention efforts are best invested in
the prevention of such a gap. Many, including us, are in agreement concerning the
Downloaded by [Tufts University] at 10:17 16 March 2012
READING FLUENCY
229
need for developmental prevention models that address the growth of accuracy
and fluency before fluency problems ever have an opportunity to develop. As
Torgesen et al. (in press) and Kame’enui et al. (in press) have begun to demonstrate, there are impressive ways we can prevent an enormous number of at-risk
children from developing later reading and reading fluency problems.
Despite all our best prevention efforts, however, some children—for various
well-discussed reasons (see Wolf & Bowers, 1999, 2000)—will continue to develop serious reading fluency and comprehension problems. The fluency needs of
most of these children, we believe, will be insufficiently addressed by most of the
experimental interventions described in Table 1. These studies address only the final levels of fluency development and few of the components underlying it.
Levels of Subskills and Component Processes
of Reading Fluency Instruction
The most difficult question in this article concerns what components that fluency
intervention should emphasize. The most difficult implication of our view of reading fluency is that each component and each level of subskills in reading should be
addressed in fluency intervention. Following connectionist principles (see Adams,
1990; Foorman, 1994; Seidenberg & McClelland, 1989), we believe that explicit
instruction is necessary to link phonological, orthographic, semantic, and morphological processes to sublexical and word-level subskills (see Berninger et al., in
press). As seen in Table 1, only a very small number of intervention programs include any systematic attention to the explicit linking of phonological, orthographic,
and semantic processes in the acquisition of reading (see discussion of Decoding
Pilot Program and RAVE-O program in Meyer & Felton, 1999). Furthermore, we
believe that systematic instruction should be directed to accuracy and then to rate at
each developmental level of the acquisition of reading subskills—that is, at the
level of the phoneme, grapheme, letter, letter pattern (orthographic chunk), word,
phrase and sentence, and passage.
On the surface, this is a great deal to ask. In actuality it is asking that each level in
the teaching of reading incorporate both an accuracy and a rate of processing component. For example, following Ehri’s (1998) framework, we believe that instruction
for fluency development should begin with an emphasis on the accuracy of underlying representations (i.e., phonological, orthographic, semantic, and morphological)
for each level up to the word level. In the second phase the focus should be on facilitating the rate of processing in these same lower level processes and subskills, until
they become automatic and can be orally read within the context of connected text.
In what we refer to as the fluency outcome stage (the equivalent of Ehri’s speed
phase), instruction would emphasize attainment of an efficient reading speed with
good comprehension of passage-level connected text (see Biemiller, 1977; Chall,
1983; Doehring, 1976, for average reading rates at different ages).
230
WOLF AND KATZIR-COHEN
Downloaded by [Tufts University] at 10:17 16 March 2012
In the last section of this article we discuss an experimental fluency program
based on the developmental, componential view of reading fluency articulated in
this article. This program represents our first efforts to squarely confront the complexity and developmental changes in fluency development. We do not conceptualize this program as the only solution to the questions raised in this article. Rather,
we see it as an evolving approach that will be empirically tested and refined as our
collective understanding of fluency increases.
TOWARD A COMPREHENSIVE APPROACH TO
FLUENCY INTERVENTION: THE RAVE-O
PROGRAM PRINCIPLES
Following concepts outlined in the last section, the RAVE-O program (Retrieval,
Automaticity, Vocabulary, Engagement, and Orthography; see complete description in Wolf, Miller, & Donnelly, 2000) was designed as a comprehensive, fluency-based intervention with three major goals: (a) accuracy and automaticity in
the components and subskills underlying reading; (b) fluency in word identification, word attack, connected text, and comprehension; and (c) a transformed attitude in children with reading disabilities toward written language. As such, the program is one example of a developmental, component-based approach to fluency
development. The program emerged out of theoretical-based knowledge about
dyslexic children whose particular deficits go beyond phonological-based deficits.
The RAVE-O program was specifically designed to address the needs for increased
rate of processing and reading fluency among two well-represented subtypes of
children with reading disabilities (Wolf & Bowers, 1999).
RAVE-O is aimed at increasing processing speed both in underlying components (i.e., visual and auditory recognition, orthographic pattern recognition, lexical retrieval, and semantic activation processes) and in three reading outcome
behaviors (i.e., word identification, word attack, and comprehension). Drawing on
research by Doehring (1976) and connectionist models of the reading process (Adams, 1990; Caramazza, 1997; Foorman, 1994), RAVE-O simultaneously addresses the need for automaticity in phonological, orthographic, and semantic
systems and the importance of teaching explicit connections among these three
systems.
PROGRAM IN PRACTICE
Goals 1 and 2
The program is to be taught only in combination with a program that teaches systematic, phonological analysis and blending. Children are taught a group of core
Downloaded by [Tufts University] at 10:17 16 March 2012
READING FLUENCY
231
words each week that exemplify critical phonological, orthographic, and semantic
principles. Each core word is chosen on the basis of (a) shared phonemes with the
phonological-treatment program, (b) sequenced orthographic patterns, and (c) semantic richness (e.g., each core word has at least three different meanings). First,
the multiple meanings of core words are introduced within their varied possible semantic contexts. Second, children are directly taught to connect the phoneme analysis skills (taught in the phonological program) with the trained orthographic patterns in RAVE-O. For example, children are taught individual phonemes in the
phonological program (like a, t, and m) and orthographic chunks with the same
phonemes in RAVE-O (e.g., at and am, along with their word families).
A central program emphasis is on rapid recognition and practice of the most frequent orthographic letter patterns in English. A special set of computerized games
(see Speed Wizards; Wolf & Goodman, 1996) was designed both to allow for
maximal practice and to increase the speed of orthographic pattern recognition
(i.e., onset and rime) in a fun fashion. In addition, the format of the games was designed to enhance the speed and accuracy of the multiple underlying components
like visual scanning, auditory discrimination at the phoneme and phoneme cluster
levels, and letter and word identification (see more complete description in Wolf,
Miller, & Donnelly, 2000).
Another key emphasis in the program involves the direct integration of vocabulary development and increased lexical retrieval skills. The dual emphasis on vocabulary and retrieval is based on earlier work in vocabulary development that
suggested that one retrieves fastest what one knows best (see Beck, Perfetti, &
McKeown, 1982; German, 1992; Kame’enui, Dixon, & Carnine, 1987; Wolf &
Segal, 1999). Vocabulary growth is conceptualized as essential both to rapid retrieval (in oral and written language) and to improved comprehension, an ultimate
goal in the program.
The combination of the vocabulary and retrieval areas also helps address several clinical and theoretical issues in research on children with reading disabilities.
First, in the clinical domain, although many dyslexic children begin with adequate
to superb vocabularies, their vocabulary falls behind the level of average readers
who are reading increasingly sophisticated texts with words unavailable in oral
discourse. This is a key part of the sight word vocabulary discrepancy discussed by
Torgesen et al. (in press) and Nagy and Anderson (1984). Second, although a child
with reading disabilities may well know a given word, our work has demonstrated
that many dyslexic readers often cannot retrieve known words (Wolf &
Goodglass, 1986; Wolf & Obregon, 1992). Third, whether it is an antecedent or a
consequence of reading and fluency problems, or a combination of the two, some
children with reading disabilities appear to have a less flexible set toward determining the often multiple meanings of words. They simply may not be able to allocate more time to processing more than one meaning to a known word, thus
affecting later comprehension development.
Downloaded by [Tufts University] at 10:17 16 March 2012
232
WOLF AND KATZIR-COHEN
RAVE-O’s daily structure combats flagging vocabulary development,
word-retrieval problems, and inflexibility in word usage by teaching an imaginative set toward language from the start. Every core word in the program with its
varied meanings and syntactic uses is practiced every day in different ways. For
example, “Word-webs” are created for many of the words as a teaching tool to illustrate the principle “If you know one word, you know a hundred!”
To enhance the speed and accuracy of lexical retrieval, four metacognitive
strategies for word retrieval—called Sam Spade Strategies—are taught. Each
strategy incorporates some of the phonological, orthographic, and semantic principles of the program (see earlier RAVE program in Wolf & Segal, 1999). For example, children are taught during “Tip of the Tongue” occurrences to try to remember
an onset, rime, or semantic associate of the missed word.
Sam Spade also appears as a character in the series of timed and untimed comprehension stories (e.g., Minute Mysteries). These stories accompany each week
of RAVE-O and directly address fluency in comprehension in several ways. The
controlled vocabulary in the stories incorporates the week’s particular orthographic patterns and emphasizes the multiple meanings and semantically related
words of the week’s core words. In addition, the stories provide a good vehicle for
repeated reading practice, which facilitates fluency in connected text. Thus, the
Minute Mysteries facilitate fluency in phonological, orthographic, and semantic
systems at the same time that they build comprehension skills.
The Third Goal: Engagement With Language
An important, implicit goal of the RAVE-O program concerns a whole new cognitive–affective set toward language use and is specifically geared toward children
who have begun to feel disenfranchised from their own language. As Chukovsky
(1963) once stated in another context, many of our children have begun to feel “like
strangers in the realm of their own language” (p. 9). RAVE-O is built around the notion that a combination of systematic goals, incremental successes, and whimsical
imaginative activities go hand in hand with building a new attitude to language,
reading, and learning in general. From colorfully animated computer games to
webs of words, children learn they can master this task of reading. The motivational
component of RAVE-O is one of its real strengths for both teacher and learner, particularly in overcrowded urban schools.
PRELIMINARY ANALYSES
Preliminary data on the RAVE-O program are now available for 200 second- and
third-grade children with severe reading impairments who participated in the large
Downloaded by [Tufts University] at 10:17 16 March 2012
READING FLUENCY
233
National Institute for Child Health and Human Development (NICHD) intervention project by Morris, Lovett, and Wolf (1995). In this project (described in Wolf,
Miller, & Donnelly, 2000), children were selected if they met either low achievement or ability–achievement (regressions corrected) discrepancy criteria for reading disability on the Woodcock Reading Mastery Tests (Woodcock, 1987). All
children in the program were taught in small-group (4:1), pullout settings for 70
sessions, with 30 min of RAVE-O and 30 min of the phonological program (see
Lovett et al., 2000, for description of Phonological Analysis and Blending).
The preliminary data indicate significant gains in word attack, word identification, oral reading rate and accuracy, and passage comprehension. Although these
data will be presented in full report in a subsequent study, the preliminary analyses
offer our first hint that a developmental–componential approach to fluency intervention can change the reading rate of severely impaired readers on standardized
oral reading measures and, most important, increase their comprehension on standardized measures. The findings, therefore, provide one promising piece of evidence that severe processing speed deficits in individuals with reading disabilities
may be amenable to treatment. That said, these data do not yet represent an empirical validation of the developmental–componential view to reading fluency that we
offer in this article; rather, they offer a starting place for further efforts.
SUMMARY
There are many unresolved theoretical and applied questions in research on reading
fluency and fluency-based deficits in individuals with reading disabilities. Prominent among the theoretical questions are issues of definition and clarification of
time-related terms. There is no consensus concerning how we use such basic terms
as rate, automaticity, speed of processing, temporal processing, dynamic processing, or precise timing, much less fluency. Yet, as illustrated here, our definitions
have critical implications for how and who we diagnose and for how we construct
and evaluate intervention. We argue strongly for a definition of fluency that is developmental- and component-based, where rate and speed are the characteristics of
the components and subskills of reading, and where accuracy and automaticity are
assessable outcome stages of reading and reading fluency. Within this admittedly
complex view, there are large lacunae in our understanding of how and how much
individual component processes (e.g., orthographic, semantic, morphosyntactic)
contribute to fluency and how fluency and comprehension influence each other.
There are also frank omissions even in this broad view of fluency concerning the
contributions of basic processes like general cognitive speed (see Breznitz, in
press; Carver, 1991; Kail, 1992). There remains much to understand.
At such a moment it is good to reflect on our own recent history as a field.
Thirty years ago, powerful questions about the role of phonological processes in
234
WOLF AND KATZIR-COHEN
reading led to a systematic body of research on phonological-core deficits, diagnostic measures, and phonological-based treatments. In a similar vein, we believe
that the exploration of the basic structure and development of reading fluency will
lead to a heightened understanding of fluency-based reading problems in children,
their assessment, and their amelioration. The articles in this special issue are a part
of that exploration.
Downloaded by [Tufts University] at 10:17 16 March 2012
ACKNOWLEDGMENTS
We acknowledge the support of National Institute for Child Health and Human Development Grant OD30970–01A1 and the Haan Foundation for Children during
ongoing intervention research. We thank present and past members of the Center
for Reading and Language Research: Heidi Bally, Kathleen Biddle, Theresa
Deeney, Katharine Donnelly, Wendy Galante, Calvin Gidney, Julie Jeffery, Terry
Joffe, Cynthia Krug, Lynne Miller, Mateo Obregon, Beth O’Brien, Alyssa
O’Rourke, and Maya Rom. The work cited in this article could never have been
done without their efforts. We also are indebted to Pat Bowers, Zvia Breznitz,
Robin Morris, and Maureen Lovett for their intellectual collaborations on this topic
over many years, and to our colleagues whose contributions to the National Dyslexia Research Foundation’s conference—Time, Fluency, and Dyslexia—are cited
throughout this article.
REFERENCES
Adams, M. J. (1990). Beginning to read: Thinking and learning about print. Cambridge, MA: MIT
Press.
Allington, R. L. (1983). Fluency: The neglected reading goal in reading instruction. The Reading
Teacher, 36, 556–561.
Badian, N. (1996, November). Dyslexia: Does it exist? Dyslexia, garden-variety poor reading, and the
Double-Deficit Hypothesis. Paper presented at the Orton Dyslexia Society, Boston.
Beck, I. L., Perfetti, C. A., & McKeown, M. G. (1982). Effects of long-term vocabulary instruction on
lexical access and reading comprehension. Journal of Educational Psychology, 74, 506–521.
Berninger, V. W. (in press). Specific reading and writing disabilities in young children: Assessment,
prevention, and intervention (2nd ed.). San Diego, CA: Academic.
Berninger, V. W., Abbott, R. D., Billingsley, F., & Nagy, W. (in press). Processes underlying timing and
fluency of reading: Efficiency, automaticity, coordination, and morphological awareness. In M.
Wolf (Ed.), Dyslexia, fluency, and the brain. York Press.
Biemiller, A. (1977). Relationship between oral reading rates for letters, words, and simple text in the
development of reading achievement. Reading Research Quarterly, 13, 223–253.
Breznitz, Z. (1997). Effects of accelerated reading rate on memory for text among dyslexic readers.
Journal of Educational Psychology, 89, 289–297.
Downloaded by [Tufts University] at 10:17 16 March 2012
READING FLUENCY
235
Breznitz, Z. (in press). The role of inter-modality temporal features of speed of information processing
in asynchrony between visual-orthographic and auditory-phonological processing. In M. Wolf
(Ed.), Dyslexia, fluency, and the brain. York Press.
Breznitz, Z., & Share, D. L. (1992). Effects of accelerated reading rate on memory for text among dyslexic readers. Journal of Educational Psychology, 89, 289–297.
Caramazza, A. (1997). How many levels of processing are there in lexical access? Cognitive
Neuropsychology, 14, 177–209.
Carver, R. P. (1991). Using letter-naming speed to diagnose reading disability. Remedial and Special
Education, 12(5), 33–43.
Carver, R. P. (1997). Reading for one second, one minute, or one year from the perspective of rauding
theory. Scientific Studies of Reading, 1, 3–43.
Cattell, M. (1886). The time it takes to see and name objects. Mind, 2, 63–85.
Chall, J. S. (1983). Stages of reading development. New York: McGraw-Hill.
Chukovsky, K. (1963). From two to five. Berkeley, CA: University of California Press.
Compton, D. L., Chayna, J. D., DeFries, D. J., Gayan, J., & Olson, R. K. (in press). Genetic and environmental influences on reading and RAN: An overview of results from the Colorado twin study. In M.
Wolf (Ed.), Dyslexia, fluency, and the brain. York Press.
Dahl, P. (1974). An experimental program for teaching high speed word recognition and comprehension
skills (Final Rep. Project No. 3–1154). Washington, DC: National Institute of Education.
Doehring, D. G. (1976). Acquisition of rapid reading responses. Monograph of the Society for Research
in Child Development, 165(2).
Dowhower, S. L. (1994). Repeated reading revisited: Research into practice. Reading and Writing
Quarterly: Overcoming Learning Difficulties, 10, 343–358.
Ehri, L. (1998). Grapheme–phoneme knowledge is essential for learning to read words in English. In J.
Metsala & L. Ehri (Eds.), Word recognition in beginning reading (pp. 3–40). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Farmer, M. E., & Klein, R. M. (1995). The evidence for a temporal processing deficit linked to dyslexia:
A review. Psychonomic Bulletin & Review, 2, 460–493.
Foorman, B. R. (1994). The varieties of orthographic knowledge I: Theoretical and developmental issues. Dordrecht, The Netherlands: Kluwer.
Fuchs, L. S., Fuchs, D., & Maxwell, L. (1988). The validity of informal reading comprehension measures. Remedial and Special Education, 9(2), 20–29.
German, D. J. (1992). Word finding intervention for children and adolescents. Topics in Learning Disorders, 13(1), 33–50.
Gough, P. B. (1972). One second of reading. In J. F. Kavanaugh & I. G. Mattingly (Eds.), The relationship between speech and hearing (pp. 331–358). Cambridge, MA: MIT Press.
Haynes, C. (1994). Differences between name recognition and name retrieval abilities in relationship to
reading performance. Unpublished doctoral dissertation, Harvard University, Cambridge, MA.
Hebb, D. O. (1949). The organization of behaviour. New York: Wiley.
Hudson, R., Mercer, C. D., & Lane, H. (2000). Exploring reading fluency: A paradigmatic overview.
Unpublished manuscript, University of Florida, Gainesville.
Huey, S. E. (1968). The psychology and pedagogy of reading. Cambridge, MA: MIT Press. (Original
work published 1908)
Jackson, M. D., & McClelland, J. L. (1979). Processing determinants of reading speed. Journal of Experimental Psychology: General, 108, 151–181.
Jenkins, J. R., Fuchs, L. S., Espin, C., van den Broek, P., & Deno, S. L. (2000, February). Effects of task
format and performance dimension on word reading measures: Criterion validity, sensitivity to impairment, and context facilitation. Paper presented at Pacific Coast Research Conference, San
Diego, CA.
Downloaded by [Tufts University] at 10:17 16 March 2012
236
WOLF AND KATZIR-COHEN
Johnston, R. S., & Anderson, M. (1998). Memory span, naming speed, and memory strategies in poor
and normal readers. Memory, 6, 143–163.
Kail, R. (1991). Developmental change in speed of processing during childhood and adolescence. Psychological Bulletin, 109, 490–501.
Kail, R. (1992). Processing speed, speech rate, and memory. Developmental Psychology, 28, 899–904.
Kame’enui, E. J., Dixon, R. C., & Carnine, D. W. (1987). Issues in the design of vocabulary instruction.
In M. G. McKeown & M. E. Curtis (Eds.), The nature of vocabulary acquisition (pp. 129–145).
Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Kame’enui, E. J., Simmons, D. C., Good, R. H., & Harn, B. A. (in press). The use of fluency-based measures in early identification and evaluation of intervention efficacy in schools. In M. Wolf (Ed.),
Dyslexia, fluency, and the brain. York Press.
LaBerge, D., & Samuels, S. J. (1974). Toward a theory of automatic information processing in reading.
Cognitive Psychology, 6, 293–323.
Lemoine, H. E., Levy, B. A., & Hutchinson, A. (1993). Increasing the naming speed of poor readers:
Representations formed across repetitions. Journal of Experimental Child Psychology, 55,
297–328.
Levy, B. A. (in press). Moving the bottom: Improving reading fluency. In M. Wolf (Ed.), Dyslexia, fluency, and the brain. York Press.
Levy, B. A., Abello, B., & Lysynchuk, L. (1997). Transfer from word training to reading in context:
Gains in reading fluency and comprehension. Learning Disability Quarterly, 20, 173–188.
Levy, B. A., Bourassa, D. C., & Horn, C. (1999). Fast and slow namers: Benefits of segmentation and
whole word training. Journal of Experimental Child Psychology, 73, 115–138.
Levy, B. A., Nicholas, A., & Kohen, D. (1993). Repeated readings: Process benefits for good and poor
readers. Journal of Experimental Child Psychology, 56, 303–327.
Logan, G. D. (1988). Toward an instant theory of automatization. Psychological Review, 95, 492–527.
Logan, G. D. (1997). Automaticity and reading: Perspectives from the instance theory of automation.
Reading and Writing Quarterly, 13, 146.
Lovett, M. W. (1987). A developmental approach to reading disability: Accuracy and speed criteria of
normal and deficient reading skill. Child Development, 58, 234–260.
Lovett, M. W., Steinbach, K. A., & Frijters, J. C. (2000). Remediating the core deficits of developmental
reading disability: A double-deficit perspective. Journal of Learning Disabilities, 33, 334–358.
Lyon, G. R., & Moats, L. C. (1997). Critical conceptual and methodological considerations in reading
intervention research. Journal of Learning Disabilities, 30, 578–588.
Manis, F. R., Doi, L. M., & Bhadha, B. (2000). Naming speed, phonological awareness, and orthographic knowledge in second graders. Journal of Learning Disabilities, 33, 325–333.
Marcus, D. (1997). An investigation of the relationship of naming speed, processing speed, and reading
in young, reading-impaired children. Unpublished master’s thesis, Tufts University, Boston.
McCormick, C., & Samuels, S. J. (1979). Word recognition by second graders: The unit of perception
and interrelationships among accuracy, latency, and comprehension. Journal of Reading Behaviour,
11, 107–118.
Meyer, M. S., & Felton, R. H. (1999). Repeated reading to enhance fluency: Old approaches and new directions. Annals of Dyslexia, 49, 283–306.
Morris, R., Lovett, M., & Wolf, M. (1995). Treatment of developmental reading disabilities. National
Institute for Child Health and Human Development Grant proposal #1R55HD/OD30970–01A1.
Nagy, W., & Anderson, R. C. (1984). How many words are there in printed school English? Reading Research Quarterly, 19, 304–330.
National Reading Panel. (2000). Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction. Washington, DC:
National Institute of Child Health and Human Development.
Downloaded by [Tufts University] at 10:17 16 March 2012
READING FLUENCY
237
Nicholson, R., & Fawcett, A. J. (1994). Reaction times and dyslexia. Quarterly Journal of Experimental
Psychology, 47, 29–48.
Norman, D. A. (1968). Toward a theory of memory and attention. Psychological Review, 75,
522–536.
O’Shea, L. J., Sindelar, P. T., & O’Shea, D. J. (1985). The effects of repeated readings and attentional
cues on reading fluency and comprehension. Journal of Reading Behavior, 17, 129–142.
Perfetti, C. (1977). Language comprehension and fast decoding: Some psycholinguistic prerequisites
for skilled reading comprehension. In J. T. Guthrie (Ed.), Cognition, curriculum, and comprehension (pp. 20–41). Newark, DE: International Reading Association.
Perfetti, C. A. (1985). Reading ability. New York: Oxford Press.
Posner, M. I., Lewis, J. L., & Conrad, C. (1972). Component processes in reading: A performance analysis. In J. F. Kavanaugh & I. G. Mattingly (Eds.), Language by ear and by eye (pp. 159–192). Cambridge, MA: MIT Press.
Rashotte, C., & Torgesen, J. (1985). Repeated reading and reading fluency in learning disabled children.
Reading Research Quarterly, 20, 180–188.
Samuels, S. J. (1985). Automaticity and repeated reading. Lexington, MA: Lexington.
Schreiber, P. A. (1980). On the acquisition of reading fluency. Journal of Reading Behavior, 12,
177–186.
Segal, D., & Wolf, M. (1993). Cognitive, linguistic, and developmental perspectives on learning disorders. Boston: Little, Brown.
Seidenberg, M., & McClelland, J. (1989). A distributed developmental model of word recognition and
naming. Psychological Review, 96, 523–568.
Shankweiler, D., & Crain, S. (1986). Language mechanisms and reading disorder: A modular approach.
Cognition, 24, 139–168.
Shinn, M. R., Good, R. H., Knutson, N., Tilly, W. D., & Collins, V. L. (1992). Curriculum based measurement of oral reading fluency: A confirmatory analysis of its relation to reading. School Psychology Review, 21, 459–479.
Spring, C., & Capps, C. (1974). Encoding speed, rehearsal, and probed recall of dyslexic boys. Journal
of Educational Psychology, 66, 780–786.
Stahl, S. A., Heubach, K., & Cramond, B. (1997). Fluency-oriented reading instruction (Reading Research 79). Ahtens, GA: National Reading Research Center.
Stein, J. (in press). The neurobiology of reading difficulties. In M. Wolf (Ed.), Dyslexia, fluency, and the
brain. York Press.
Stoddard, K., Valcante, G., Sindelar, P., O’Shea, L., & Algozzine, B. (1993). Increasing reading rate and
comprehension: The effects of repeated readings, sentence segmentation, and intonation training.
Reading Research and Instruction, 32, 53–65.
Tallal, P., Miller, S., & Fitch, R. H. (1993). Neurobiological basis of speech: A case for the preeminence
of temporal processing. In P. Tallal, A. Galaburda, & C. Von Euler (Eds.), Temporal processing in
the nervous system. Annals of the New York Academy of Sciences (Vol. 682, pp. 7–47). New York:
New York Academy of Sciences.
Tan, A., & Nicholson, T. (1997). Flash cards revisited: Training poor readers to read words faster improves their comprehension of text. Journal of Educational Psychology, 89, 276–288.
Torgesen, J., Rashotte, C., & Alexander, A. (in press). The prevention and remediation of reading fluency problems. In M. Wolf (Ed.), Dyslexia, fluency, and the brain. York Press.
Torgesen, J., Rashotte, C., & Wagner, R. (1997, November). Research on instructional interventions for
children with reading disabilities. Paper presented at the International Dyslexia Association Conference, Chicago.
Torgesen, J., Wagner, R., Rashotte, C., Rose, E., Lindamood, P., Conway, T., & Garven, C. (1999). Preventing reading failure in young children with phonological processing disabilities: Group and individual responses to instruction. Journal of Educational Psychology, 91, 1–15.
Downloaded by [Tufts University] at 10:17 16 March 2012
238
WOLF AND KATZIR-COHEN
Waber, D. P. (in press). Aberrations of timing in children with impaired reading: Cause or effect? In M.
Wolf (Ed.), Dyslexia, fluency, and the brain. York Press.
Wiederholt, J. L., & Bryant, B. (1992). Gray Oral Reading Test (GORT–3). Austin, TX: PRO-ED.
Wolf, M. (1982). The word-retrieval process and reading in children and aphasics. In K. Nelson (Ed.),
Children’s language (pp. 437–486). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Wolf, M. (1991). Naming speed and reading: The contribution of the cognitive neurosciences. Reading
Research Quarterly, 26, 123–141.
Wolf, M., & Bowers, P. (1999). The “Double-Deficit Hypothesis” for the developmental dyslexias.
Journal of Educational Psychology, 91(3), 1–24.
Wolf, M., & Bowers, P. (2000). The question of naming-speed deficits in developmental reading disabilities: An introduction to the Double-Deficit Hypothesis. Journal of Learning Disabilities, 33,
322–324.
Wolf, M., Bowers, P., & Biddle, K. (2000). Naming-speed processes, timing, and reading: A conceptual
review. Journal of Learning Disabilities, 33, 387–407.
Wolf, M., & Goodglass, H. (1986). Dyslexia, dysnomia, and lexical retrieval. Brain and Language, 28,
154–168.
Wolf, M., & Goodman, G. (1996). Speed Wizards [Computerized reading program]. Tufts University,
Boston, and Rochester Institute of Technology, Rochester, NY.
Wolf, M., Miller, L., & Donnelly, K. (2000). The Retrieval, Automaticity, Vocabulary Elaboration, Orthography (RAVE-O): A comprehensive fluency-based reading intervention program. Journal of
Learning Disabilities, 33, 375–386.
Wolf, M., & Obregon, M. (1992). Early naming deficits, developmental dyslexia, and a specific deficit
hypothesis. Brain and Language, 42, 219–247.
Wolf, M., & Segal, D. (1999). Retrieval-rate, accuracy and vocabulary elaboration (RAVE) in reading-impaired children: A pilot intervention program. Dyslexia, 5, 1–27.
Wolff, P. H. (2000). Impaired temporal resolution in developmental dyslexia. Annals of the New York
Academy of Sciences, 682, 87–103.
Wood, F., Flowers, L., & Grigorenko, E. (in press). On the functional neuroanatomy of fluency, or why
walking is just as important to reading as talking is. In M. Wolf (Ed.), Dyslexia, fluency, and the
brain. York Press.
Woodcock, R. W. (1987). Woodcock Reading Mastery Tests. Circle Pines, MN: American Guidance
Service.
Young, A., & Bowers, P. G. (1995). Individual differences and text difficulty determinants of reading
fluency and expressiveness. Journal of Experimental Child Psychology, 60, 428–454.
Young, A. R., Bowers, P. G., & MacKinnon, G. E. (1996). Effects of prosodic modeling and repeated
reading on poor readers’ fluency and comprehension. Applied Psycholinguistics, 17, 59–84.
Manuscript received November 9, 2000
Final revision received December 4, 2000
Accepted December 4, 2000