ijrelt-v2n1p7-en-1.

pdf

A Critical Overview of Models of Reading

Comprehension with a Focus on Cognitive Aspects
*
Mohammad Taghi Shahnazari
Assistant Professor, University of Isfahan, Iran
[email protected]

Azizollah Dabaghi
Assistant Professor, University of Isfahan, Iran
[email protected]

Abstract
Reading is a cognitive activity involving skills, strategies, attentional
resources, knowledge resources and their integration. The reader’s role is to
decode the written symbols to allow for the recovery of information from
long-term memory to construct a plausible interpretation of the writer’s
message. Various number of reading models have been proposed by
researchers among which some focus on motivational and emotional aspects
of reading. Others highlight the cognitive aspects of reading. In this study,
the models characterizing reading in terms of cognitive aspects are
revieweded, and different viewpoints on the reading process are described.
This may help EFL/ESL teachers to improve their understanding of the
reading process, update their perspectives on teaching reading tasks which
in turn might result in more efficient learning by not putting too much
cognitively demanding reading tasks on EFL/ESL learners.
Keywords: Reading Models; Attentional Resources; EFL Learners;
Cognitive Process.

1. Introduction from visual stimulus to an oral language

Foreign or second language reading has recoding to meaning responses. Since his
been the focus of researchers’ attention over aim was to be illustrative, rather than definitive,
the past twenty years (Macaro, 2003) and a many imprecisely specified stages were left
number of models for this preponderant in his model. Also, Levin and Kaplan
Vol 1-Winter 2014- No. 3

skill have been proposed. During the 1960s (1970), Hockberg (1970), and Mackworth
and early 1970s, a number of researchers (1972) all argued about what a model
proposed more or less formal models of explaining the processes of skilled reading
reading comprehension. For example, Carroll must account for. This work heralded a
(1964) suggested a definition of reading change in conceptions of the reading
along with a simple one-way flow diagram constructs among researchers and practitioners.

* Corresponding Author
 8 / IJRELT

In contrast, Goodman (1965, 1976) described identification than poor readers. Grabe
reading as a psycholinguistic guessing game. (2009) argues that Goodman’s Psycholinguistic
Three distinctive characteristics distinguished Guessing Game model provides a possible
Goodman’s model from other models. First, explanation for an early stage of reading
he believed that the reader relies on existing development. He further argues that Goodman’s
syntactic and semantic knowledge rather Psycholinguistic Guessing Game Model cannot
than graphic information in the process of be a valid alternative to any other models of
reading. Second, he used the term reading, which will be described in the
“decoding” differently from others. While following section (Grabe, 2009).
others used this term to describe what In the following section, seven models of
happens when a reader translates a reading will be discussed in turn: bottom-
graphemic input into a phonemic input, up, top-down, Rumelhurt’s Interactive Model
Goodman used it to illustrate how either a (1977), Stanovich’s (1980) Interactive-
graphemic input or phonemic input gets Compensatory Model, Construction-Integration
translated into a meaning code. He also Model, Verbal Efficiency Theory of Reading,
used the term ‘recoding’ to describe the and Compensatory-Encoding Model. Discussion
process of translating graphemes into of these models follows by a critical
phonemes. Goodman’s and his colleagues’ overview with the focus on the cognitive
efforts were mostly focused on indicating aspects of reading. Other models characterize
the strong procedural preference that reading as a more complex process where
readers of all ages had for depending on the motivational and emotional aspects play an
meaning cues (rather than graphic and important role; however, these aspects are
graphophonemic cues) available in the beyond the scope of this study.
printed message. Third, his model has
arguably had the greatest influence on 1.1. Bottom-Up Model
conceptions about reading pedagogy, to the A bottom-up reading model is a model that
extent that ‘the psycholinguistic approach to focuses on a single-direction, part-to-whole
reading’ or ‘the whole-language approach to processing of a text. More specifically, in
reading’ have become commonly used bottom-up models, the reader is assumed to
terms in the language teaching field be involved in a mechanical process where
(Samuels & Kamil, 1988). he or she decodes the ongoing text letter by
In summary, Goodman (1996) argued letter, word by word, and sentence by
that when an individual reads a text, he or sentence (Grabe, 2009).In these models, the
she makes a set of hypotheses about the reader decodes the text which has been
upcoming text, samples minimally from the previously encoded by the writer. Decoding
text, confirms hypotheses, and then of the text includes a visual focus on the
produces new predictions. However, other identification of the letters, noticing the
researchers (e.g., Grabe, 2000, 2009; Koda, combination of the letters, recognition of
2005; Pressley, 2006) impose some the words, establishing sentences via their
criticisms on this argument. They argue that syntactic structures and finally integrating
there is no persuasive evidence in a fluent sentences into coherent discourse until the
reading that good readers (a) sample from meaning of the text is eventually
Vol 1-Winter 2014 -No. 3

texts and make hypotheses about what determined. The reader’s world knowledge,
words are coming next (b) control their eye contextual information, and other higher-
movements (direct the eye where to go order processing strategies play a minor
during reading to sample from a text). They role, particularly at beginning stages, in
further argue that good readers do not processing information in this model
usually guess upcoming words in a text, (Alderson, 2000; Beach, 1997; Dechant,
and make less use of context for word 1991; Grabe & Stoller, 2002; Koda, 2005).
 A Critical Overview of Models of Reading... / 9

The proponents of bottom-up models (e.g., Word recognition involves orthographic,

Flesch, 1955; Gough, 1972; LaBerge & phonological and semantic operations.
Samuels, 1974) argue that these models While a word’s meaning is obtained in
work on the premise that the written text is semantic operation, the word’s sound features
hierarchically organized, and the reader’s are achieved in phonological operation.
job is to process the smallest linguistic (i.e. Both of these operations are activated
grapho-phonic) unit first and then combine through orthographic operation and achieved
the smaller unitsto discover and comprehend via an analysis of graphic symbols (Koda,
the higher (e.g., sentence syntax) units 2005; Samuels & Kamil, 1988).
(Alderson, 2000; Dechant, 1991; Field, Orthographic knowledge plays an
2003; Grabe&Stoller, 2002; Koda, 2005; important role in word recognition.
Macaro, 2003; Mitchell, 1982). Research suggested that skilled readers
Word recognition plays an essential role were able to not only analyse and manipulate
in reading comprehension. Koda (2005) word-internal elements such as letters and
defines it as “the processes of extracting letter clusters (e.g., Ehri, 1998; Shankweiler
lexical information from graphic displays of & Liberman, 1972), but also to pronounce
words” (p., 29). Studies on eye movement both individual letters and nonsense letter
indicate that nearly every content word strings (e.g., Siegel & Ryan, 1988; Wanger,
obtains direct visual fixation (Balota, Torgesen, & Rashotte, 1994). This is
Pollasek, &Rayner, 1985; Just & Carpenter, because orthographic knowledge is a
1980, 1987), and the lack of even a single powerful mnemonic device that connects
letter can be disruptive, largely decreasing the written forms of specific words to their
reading efficiency (e.g., Mc Conkie & Zola, pronunciation in memory (Ehri, 1998).
1981; Rayner & Bertera, 1979). Furthermore, Phonological decoding may be the most
based on developmental studies, researchers essential competence for reading acquisition
argued that poor readers could not extract in all languages (Koda, 2005). It is defined
visual information from print, and deficient as the processes involved in accessing,
word recognition is associated with poor storing, and manipulating phonological
comprehension (e.g., Perfetti, 1985; Stanovich, information (Torgesen & Burgess, 1998).
1988). If inefficient word recognition Researchers argued that deficits in
continues, it may have adverse effects, phonological decoding could lead to poor
directly or indirectly, on the acquisition of comprehension in both alphabetic (Abu
reading competence (e.g., Juel, 1988; Juel, Rabia, 1995) and nonalphabetic languages
Griffith, & Gough, 1986). Thus, word such as Japanese and Chinese (Kuhara-
recognition efficiency can result in kojima, Hatano, Saito, & Haebara, 1996;
successful comprehension. Some studies Zhang & Perfetti, 1993).
also indicate that automaticity can be rather The empirical evidence supports the idea
easily achieved in word recognition (e.g, that all of a word’s known meanings are
Adams, 1994; LaBerge & Samuels, 1974; activated by its orthographic input, even
Vol 1-Winter 2014- No. 3

Perfetti & Lesgold, 1979). This may reduce when strong constraints are imposed by the
the processing load in working memory, context (e.g., Seidenberg, Tanenhaus, Leiman,
leaving more capacity for the storage & Bienkowski, 1982). Then contextual
component, and eventually facilitating facilitation helps to determine the
conceptual manipulations of the extracted appropriate meaning of the word in the
information (e.g., Daneman & Carpenter, immediate context at the sentence or
1980; Waters & Caplan, 1996). discourse level. Researchers also argue that
 10 / IJRELT

less skilled readers are more likely to be writer, the reader needs to extract this
dependent on the context to retrieve word meaning and cannot go beyond it
meanings than skilled readers (e.g., (Alderson, 2000; Beach, 1997; Grabe &
Biemiller, 1979; Becker, 1985; Perfetti, Stoller, 2002; Koda, 2005). Therefore, it is
1985; Stanovich, 1988).This supports the not possible to make use of higher-order
idea that poor readers use contextual clues reading skills such as making inferences,
to compensate for their underdeveloped and consequently, background knowledge
visual information sampling skills in order plays virtually no role in deriving and
to decipher a word’s meaning (e.g., Pring & interpreting the meaning of the text in this
Snowling, 1986; Stanovich, 1986). model.
There is an assumed relationship between
contextual effects on word-meaning retrievals 1.2. Top-Down Model
and language proficiency. It is suggested A top-down reading model is a model that
that as L2 proficiency improves, reliance on focuses on what the reader brings to the text
contextual effects to retrieve word meaning to arrive at the meaning. In top-down
diminishes (Becker, 1985; Grabe, 2009; models, it is assumed that the
Pring&Snowling, 1986; Stanovich, 1986). comprehension process is not mechanical,
A large body of studies also indicates that but actively controlled by the reader
efficiency in extracting visual information (Grabe, 2009).The proponents of these
differs among high and low-proficiency models (e.g., Schank, 1978; Smith, 1971)
readers, suggesting that low-proficiency suggested that processing of a text begins in
readers are slower and less accurate in a the mind of the reader with meaning-driven
variety of word recognition tasks (e.g., processes, or an assumption about the
Favreau & Segalowitz, 1982; Haynes & meaning of a text. From this viewpoint,
Carr, 1990; Macnamara, 1970). readers identify letters and words only to
Some other studies suggest that low- confirm their assumptions about the
proficiency readers are more largely meaning of the text (Dechant, 1991). In
involved in word-level than discourse-level these models, the primary purpose of
processing (e.g., Cziko, 1980; Horiba, 1990). reading is deriving meaning from the text
Since low-proficiency readers rely on a rather than mastery of letters, letter-sound
word’s visual information rather than its correspondence, and words (e.g., Alderson,
semantic information (Chamot & El-Dinary, 2000; Grabe & Stoller, 2002; Macaro,
1999; Clarke, 1980), they are less likely to 2003; Smith, 1971). Readers are supposed
engage in conceptual manipulations (such to use meaning and grammatical cues to
as hypothesizing and predicting) than high- identify unfamiliar words, and they are able
proficiency readers (e.g., Anderson, 1991; to comprehend a passage even if they do
Chamot & El-Dinary, 1999). not recognize each word. In this view, the
In bottom-up models, the reader takes a meaning of a text, which is considered an
serial order to process the text, and the important goal to achieve, is accessed by
processing of each component takes place the reader’s activation of prior knowledge
independently of the others (e.g., Alderson, of semantic, pragmatic, syntactic and
Vol 1-Winter 2014 -No. 3

2000; Grabe, 2009; Koda, 2005; Mitchell, discourse elements. Then he or shewill be
1982). For example, the perception of able to predict and infer the meaning
phonemes is not influenced by the words in underlying propositions and words (e.g.,
which they appear (Carroll, 2008). Since Alderson, 2000; Beach, 1997; Dechant,
there is a single and restricted meaning in 1991; Grabe & Stoller, 2002; Koda, 2005).
the text driven and constructed by the However, this view does not identify what
 A Critical Overview of Models of Reading... / 11

mechanisms the reader draws on to generate made hypothesis is confirmed, strengthening

inferences or how the mental composition connections and built-up layers of
of comprehension works (Grabe, 2009). interpretation by pausing over individual
words and syntactic patterns and their
1. 3. Rumelhart’s (1977) Interactive relationship with other words and phrases
Model of Reading (Macaro, 2003). There fore, the reader is
Since the information in top-down and involved in deriving the meaning of the text
bottom-up models is passed along in one and making inferences through a constant
direction only and the information interaction between the surface structure of
contained in higher stages does not the text and his own knowledge of the topic.
influence the information in lower stages, Since working memory is the workspace
these models could not account for a for the temporary storage and processing of
number of well-known occurrences, such as ongoing information (e.g., Baddeley &
making inferences, which take place while Hitch, 1974, 2000, 2007),it may play a
reading. Thus, to remove this deficiency, significant role in the processes involved in
Rumelhart (1977) proposed an interactive deriving meaning from text(e.g., Alderson,
model of reading. This model, which is a 2000; Beach, 1997; Cain & Oakhill, 2006;
combination of both top-down and bottom- Grabe & Stoller, 2002; Koda, 2005).These
up strategies, is now widely considered a processes consist of maintaining the text
comprehensive explanation of how we information, activating the reader’s world
derive the meaning of a written text. knowledge and retrieving it from long-term
Rumelhart (1977) developed this model memory, integrating the information
based on the fact that meaning does not received from these two sources into
reside in the text alone, but is a co- coherent discourse, and finally deriving the
construction of the writer’s text and the meaning of the text. A substantial body of
reader’s interpretation. So, reading requires L1 (e.g., Daneman & Carpenter, 1980;
an interaction between the reader’s mind Divesta & Dicintio, 1997; Waters & Caplan,
and the writer’s text. This allows the 1996) and L2 research (e.g., Harrington &
information contained in higher stages to Sawyer, 1990; Lesser, 2007) supports the
interact with and influence the information idea that good readers have higher working
in lower stages. memory capacity than poor readers.
In this model, the process starts with the
information picked up by the eyes in the 1.4. Stanovich’s (1980) Interactive-
form of visual features, registered in a Compensatory Model
visual information store, and then sent to Stanovich’s (1980) interactive-compensatory
the central component of the model, the model was a refinement of Rumelhart’s
pattern synthesizer, at the first stage. Then a (1977) interactive model in explaining
wide variety of sources of information skilled and unskilled reading. It is based on
about letter shapes and orthography the principle that a process at any level can
(including what is semantically and compensate for deficiencies at any other
syntactically acceptable in the language, the level. In his words, “… a deficit in any
Vol 1-Winter 2014- No. 3

contextual situation, and information in the knowledge results in a heavier reliance on

mental lexicon) is drawn up from long-term other knowledge sources regardless of their
memory into working memory. Finally, the level in the processing hierarchy” (p. 63).
pattern synthesizer uses this information to So, top-down processing, for a reader weak
work out the more probable interpretation at word recognition, but good at the
of the text. During this process, the already- knowledge of the text topic, may
 12 / IJRELT

compensate for this deficit. On the other phase. In the construction phase, a reader
hand, a reader good at word recognition, develops propositions from the incoming
but lacking knowledge of the topic may rely text information in order to generate a
on bottom-up processes for this mental model of the text. This model is
compensation (Samuels & Kamil, 1988). provisional and incoherent since it includes
The research also supports the idea that both relevant and irrelevant information
prior knowledge of the topic can be used by which have been activated; when an
the learner as a strategy to reduce the individual reads a word, all the meanings of
cognitive load when syntactic complexity the word as well as the semantic associates
makes access to meaning difficult (Barry & of that word are automatically activated in
Lazarte, 1998). From a theoretical his or her long-term memory (Graesser,
perspective, Stanovich (1988) made a Millis & Zwaan, 1997). In the integration
unique contribution to reading models by phase, the reader evaluates the propositions
providingan explanation of compensation he or she has developed within a global
strategies, which account for why poor context with the goal of making a stable
readers show greater sensitivity to contextual activation pattern or a coherent mental
constraints under some circumstancesthan network. In doing so, the propositions
good readers (e.g., Alderson, 2000; Beach, which are compatible within the context are
1997; Grabe & Stoller, 2002; Samuels & connected to form the network, and those
Kamil, 1988). which are incompatible are disregarded. At
Unlike the models described above, the this phase, the integration of text
following two models (Construction- information with the reader’s background
Integration and Verbal-Efficiency models) knowledge yields a coherent mental model
are experimental/behaviour models of which captures the global and local
reading where the researchers draw on a relations and consequently results in
range of experimental evidence to develop comprehension. All these processes in
and support their assumptions (Grabe, construction and integration phases are
2009). Moreover, they envision an manipulated by working memory. More
important role for working memory and specifically, working memory is involved
automatic bottom-up processing in reading in the processes of making propositions,
process as it will be described in the suppressing irrelevant information, and
following section. developing a coherent mental network
which result in reading comprehension.
1.5. Construction-Integration Model This suggests that working memory plays a
of Reading strong role in reading comprehension.
Construction-Integration Model was proposed The assumptions of cognitive capacity
by Kintsch and his colleague (Kintsch, limitations in comprehension processes
1988a, 1998b; Kintsch& van Dijk, 1978; followed by the integration processes
van Dijk & Kintsch, 1983). Based on this (summarizing processes due to being
model, automatic lower-level reading overlapping associations among propositions)
processes are combined with higher level distinguish the Construction-Integration
Vol 1-Winter 2014 -No. 3

reading processes to create a coherent Model from the models reviewed before.
discourse representation of a text, and these The automatic lower-level processes and
processes are supported by a limited limited pool of attentional resources in
capacity pool of attentional resources. working memory are also considered as
There are two phases in this model; a important assumptions in the Verbal-
construction phase and an integration Efficiency Model, proposed by Perfetti
 A Critical Overview of Models of Reading... / 13

(1985, 2007). However, the emphasis is on meaning for the proposition in the context
automatic word-recognition skills which is selected. Third, initial propositions are
result in reserving more attentional created from the propositional encoding of
resources for higher level processes, and each word and maintained in working
consequently better reading performance. memory. Finally, new propositions are
integrated with previous propositions held
1.6. Verbal Efficiency Theory of Reading in working memory to give a representation
This model was proposed by Perfetti (1985, of the text. This representation remains
1999, 2007). It is an example of an active in working memory to be further
interactive model which is very constrained processed by text-modelling processes
by the bottom-up view of reading (Hudson, (higher-level processes) (Perfetti, 1985).
2007). Efficient word-recognition skills Text-modelling processes are used to
play a very important role in good reading combine the representation of the text with
performance in this model. It is argued that a reader’s background knowledge to fill the
problems with higher-level comprehension gaps in the propositional base and make
skills originate from inefficient word- him or her create inferences. It is at this
recognition skills which, in turn, stem from stage that comprehension (text-modelling)
low-quality lexical representations (Perfetti, takes place and causes the propositions to
2007). Perfetti and his colleague (Perfetti, make sense as a whole. A continual
2007; Perfetti& Hart, 2001, 2002) argue updating process occurs during the reading
that there are three constituent information by reconciling incoming text processing
sources for word recognition including with background knowledge. For this
phonological, orthographic and semantic process to be efficient, the processes for
information. These constituents work word-recognition components (phonological,
together and share information until a word orthographic, and semantic) must be
is recognized. automatic. As there are limited attentional
Based on Verbal Efficiency Theory of resources in working memory, automatic
reading, skilled readers have automatic processes reduce the amount of attentional
lower-level processes (e.g., efficient word resources for processing letter and word
recognition skills), and this allows them to identification, and consequently leave
draw on their limited attentional resources further attentional resources for processing
in working memory for higher level higher level comprehension skills. This
comprehension skills. More specifically, suggests that efficient working memory
there are two sets of processes in this processes play an important role in reading
model, local text processes and text- comprehension (Hudson, 2007), particularly
modelling processes, which have interactions for low-proficiency readers who have not
in reading process. The central principle of obtained automaticity in their local
this model is that the comprehension of a processes. If this is the case, working
text is partially constrained by the efficient memory is expected to explain individual
operation of the local processes. The local differences in reading comprehension.
processes involve the processes that the However, it is not clear yet whether or not
Vol 1-Winter 2014- No. 3

reader uses to encode contextually the role of working memory may change as
appropriate meanings and propositions. a result of language proficiency development.
When a text is read, first, the possible Overall, this model is compatible with
meanings associated with each word in the the Construction-Integration Model where
text are activated in working memory. working memory with a limited capacity
Second, the most appropriate semantic pool of resources is central to manipulating
 14 / IJRELT

reading processes. It appears that Verbal- Stanovich’s model, include higher-level

Efficiency Model is more prominent in skills and metacognitive strategies (e.g., goal
explaining efficient word-recognition skills checking, comprehension monitoring). When
which result in automaticity in reading process the process of reading proceeds under time
which leads to leaving more attentional pressure, the compensatory processes do
resources in working memory for higher- not play a role and instead lower-level
level processes such as making inferences. processes become influential and play a
The last model reviewed here is the predictive role in reading performance
Compensatory Encoding Model which also (Grabe, 2009).
conceives of working memory as having a
key role in reading process. However, a 2. Discussion and Conclusion
compensatory process is assumed in this Overall,each model can contribute to our
model which distinguishes it from the understanding of the reading process. With
Verbal Efficiency Model. Moreover, unlike each new model building on previous work,
the Verbal Efficiency model, it is a a developing understanding of the reading
descriptive model where a synthesis of the process has emerged from this rich research
most important evidence is used to explain history. The increasing specification of the
how a cognitive process like reading works role of cognitive processing in reading is of
(Grabe, 2009). particular relevance to this study, and
makes it possible to more clearly
1.7. The Compensatory-Encoding Model understand the role of cognitive resources
This model of reading was proposed based in the reading process. Except for bottom-
on verbal efficiency model and adopted its up and top-down models, there are some
basic assumptions including automatized commonalities among the models reviewed
lower-level processing, well-developed above. They all conceive of the reading
lexical representations, and efficient working process as involving both lower-level (e.g.,
memory processes (Walczyk, 1995, 2000; word-recognition skills, syntactic parsing)
Walczyk, Marsiglia, Bryan, & Naquin, and higher-level (e.g., making inferences)
2001). This model assumes an additional processes. The proponents of these models
process that is a compensatory process. The suggest that word-recognition skills play a
compensatory process in this model differs very important role in reading comprehension.
from that of Stanovich’s interactive As Perfetti (2007) suggests,one explanation
compensatory model in that it is used could be that word recognition involves the
continually to counter inefficiencies and interaction of orthographic, phonological,
weaknesses in reading skills. In Stanovich’s semantic and syntactic processes which are
model, higher-level skills and strategies are cognitively demanding. Thus, those readers
used only when needed. Based on the who are good at word recognition (due to
compensatory encoding model, the possessing well-represented lexical information)
compensatory processes play an influential leave much of their attentional resources for
and a predictive role in reading performance higher-level reading processes which in
when there is no time constraint on a turn result in better comprehension.
Vol 1-Winter 2014 -No. 3

reading task. This could work well Moreover, they argue that basic
particularly for readers with lower working grammatical information can be extracted
memory capacity as they may employ these to support clause-level meaning and
strategies to compensate for their inefficient proposition formation. However, these
working memory processes. These models differ in explaining the nature and
compensatory processes, similar to those in role of these processes.
 A Critical Overview of Models of Reading... / 15

Reading is not a mechanical process, as processing. However, as familiarity with

assumed by bottom-up models, nor is it second language increases, reading
carried out in a serial order as it is becomes more automatized and greater
envisioned by the bottom-up and top-down attention can be given to higher-level skills
models, but it is an interactive cognitive such as making inferences. In Construction-
process involving simultaneous lower-level integration model, comprehension
and higher level processes which are processes are carried out within the
manipulated in working memory. Of the attentional capacity limitations of working
reading models described above, Verbal- memory. This model may explain how
Efficiency (Perfetti, 1999, 2007) and limited-attentional resources in working
Construction-Integration models (Kintsch, memory could be drawn on to develop a
1998) are better in explaining the reading local representation of a text (a set of main
process. They specify the role of cognitive idea and supporting details), and then to
processes in reading comprehension more integrate this representation with
precisely than other models of reading. background knowledge to make an
They provide a reasonably complete interpretation of the text in a global context.
explanation of reading abilities in terms of There are still some limitations among
cognitive processes with empirical evidence, all the models of reading described above.
indicating how reading performance may None of these models explain how
vary under different conditions (Grabe, executive control processes in working
2009). They specify how reading memorywork in fluent readingand how
performance may vary due to individual reading strategies are used when reading
differences in reading abilities. These more difficult texts or learning from texts.
individual differences could stem from In both phases of Construction-Integration
either attentional resources or reading skills model, the abilities of monitoring
(e.g., word-recognition skills). For example, comprehension, using strategies, and
in Verbal-Efficiency model, skilled readers reassessing and re-establishing goals are
are distinguished from poor readers in used to repair comprehension problems.
terms of possessing automatized lower- However, it is not completely clear how the
level processes (e.g, more efficient word- operation of monitoring, as an attentional
recognition skills). Thus, the central demanding process and an aspect of
assumption of Verbal Efficiency model executive control processing in working
could be used to explain the automaticity memory, is manipulated cognitively.
aspect of the reading process. Based on this Moreover, these models do not explain how
assumption, automatized reading processes working memory handles the cognitive
are not very cognitively demanding of processes in comprehending longer and
attentional resources, so more attentional more complex texts.
resources are left in working memory for Finally, the current study can have two
higher-level reading processes, which, in implications for EFL/ESL teachers in
turn, enhance the reading performance. This teaching L2 reading. The first implication is
assumption could be used to explain what concerned with the type of reading tasks
Vol 1-Winter 2014- No. 3

role working memory plays in reading given to EFL/ESL learners. As mentioned

process and whether this role changes as before, attentional resources could be one
proficiency increases. For example, lower of the sources of individual differences in
proficiency learners may struggle with reading comprehension. This is because
word-recognition processes, and attentional these resources are limited and if these
resources are directed towards lower-level learners are given cognitively demanding
 16 / IJRELT

tasks beyond their proficiency level, this Alderson, J.C. (2000). Assessing reading.
may disturb their reading process. This Cambridge: Cambridge University Press.
suggests thatEFL/ESL teachers should be Anderson, N.J. (1991). Individual differences in
cautious not to place a burden on language strategy use in second language reading and
learners beyond their capabilities. One way testing. Modern Language Journal, 75, 460-472.
to reduce this burden is giving language Baddeley, A.D., & Hitch, G. (1974). Working
learners the reading tasks which match with memory revised. American Psychologist, 19,
their proficiency level. 851-864.
The second implication of the present Balota, D., Pollasek, A., & Rayner, K. (1985).
study is closely associated with enhancing The interaction of contextual constraints and
EFL/ESL learners' word-recognition skills. Parafoveal visual information in reading.
As research shows, word recognition can be Cognitive Psychology,17, 364-390.
very challenging and effortful, particularly Barry, S.,&Lazarte, A.A. (1998). Evidence for
for lower proficiency readers, because it mental models: how do prior knowledge,
involves processing orthographic, phonological syntactic complexity, and reading topic
and semantic information (e.g., Perfetti, affect inference generation in a recall task
1985; Segalowitz, Poulsen & Komoda, for nonnative readers of Spanish? Modern
Language Journal, 82(2), 176-193.
1991; Stanovich, 1988). Thus, EFL/ ESL
teachers should provide these learners with Becker, C.A. (1985). What do we really know
sufficient practice, for example, on sound about semantic context during reading? In
D. Besner, T. Walker & G. Mackinnon
discrimination, detecting individual phones,
(Eds.), Reading research: Advances in
distinctive sound unites (phonemes), and
theory and practice, 5 (pp. 125-166).
phonological sensitivity to make sure they NewYork: Academic Press.
are competent enough in these abilities and
Beech, J.R. (1997). Assessment of memory and
then they are able to identify individual
reading. In Beech, J.R. and Singleton, C.
words easily. This may also help these (Eds.), The psychological assessment of
language learners to gain automaticity on reading. London: Routledge.
word identification which resultsin drawing
Biemiller, A. (1979). Changes in the use of
on less attentional resources. As a result, graphic and contextual information as
further resources are leftfor executing other functions of passage difficulty and reading
lower and higher level reading processes achievement level. Journal of Reading
such assyntactic parsing, semantic Behavior, 11, 307-319.
proposition formation, inference making, Cain, K. & Oakhill, J. (2006). Assessment
and comprehension monitoring which in matters: Issues in the measurement of
turn leads to better reading performance. reading comprehension. British Journal of
Educational Psychology,76, 697-708.
Carroll, D.W. (2008). Psychology of language.
References
Thomson Wadsworth: Canada.
Abu Rabia, S. (2003). The influence of working
Carroll J.B. (1964). Language and thought.
memory on reading and creative writing
Englewood Cliffs, N.J. Prentice Hall.
processes in a second language. Educational
Vol 1-Winter 2014 -No. 3

Psychology, 23(2) 209-222. Chamot, A.U., & El-Dinary, P.B. (1999).

Children’s reading strategies in language
Adams, M.J. (1994). Modelling the connections
between word recognition and reading. In immersion classrooms. Modern Language
R.B. Ruddell, M.R. Ruddell, & H. Singer Journal, 83, 319-338.
(Eds.), Theoretical models and processes of Clarke, M.A. (1980). The short circuit
reading (pp. 830-863). Newark, DE: hypothesis of ESL reading: Or when
International Reading Association. language competence interferes with reading
 A Critical Overview of Models of Reading... / 17

performance. Modern Language Journal, Grabe, W. (2009). Reading in a second

64, 203-209. language: Moving from theory to practice.
Cziko, G. (1980). Language competence and New York: Cambridge University Press.
reading strategies: A comparison of first and Grabe, W. & Stoller, F. L. (2002). Teaching
second language oral reading errors. and researching reading. London: Longman.
Language Learning,30, 101-116. Grabe, W. & Stoller, F.L. (2011). Teaching and
Daneman, M. & Carpenter, P.A. (1980). researching reading (2nd Ed.). London:
Individual Differences in Working Memory Longman.
and Reading.Journal of Verbal Learning Graesser, A., Millis, K., &Zwaan, R. (1997).
and Verbal Behaviour,19, 450-466. Discourse comprehension. Annual Review of
Dechant, E. (1991). Understanding and Psychology,48, 163-189.
teaching reading: An interactive model. Harrington, M., & Sawyer, M. (1992). L2
Hillsdale, NJ: Lawrence Erlbaum. working memory capacity and L2 reading
Divesta, F. J. & Dicintio, M. J. (1997). skill. Studies in Second Language Acquisition,
Interactive effects of working memory span 14, 25-38.
and text context on reading comprehension Haynes, M., & Carr, T.H. (1990). Writing
and retrieval. Learning and Individual system background and second language
Differences,9(3), 215-231. reading: A component skills analysis of
Ehri, L.C. (1998). Grapheme-phoneme English reading by native speaking-readers
knowledge is essential for learning to read of Chinese. In T.H. Carr & B.A. Levy
words in English. In J.L. Metsala& L.C. (Eds.), Reading and its development:
Ehri (Eds.), Word recognition in beginning Component skills approaches (pp. 375-421),
literacy (pp. 3-40). Mahwah, NJ: Lawrence San Diego: Academic Press.
Erlbaum Associates. Horiba, Y. (1990). Narrative comprehension
Favreau, M., &Segalowitz, N. (1982). Second processes: A study of native and non-native
language reading in fluent bilinguals. readers of Japanese. Modern Language
Applied Psycholinguistics, 3, 329-341. Journal,74, 188-202.
Field, J. (2003). Psycholinguistics: A resource Hockberg, J. (1970). Components of literacy
book for students.Routledge: London. speculations and exploratory research. In H.
Flesch, R. (1955). Why Johnny can’t read? Levin & J.P. Williams (Eds.), Basic studies
New York: Harper and Row. on reading. New York: Basic Books.
Goodman, K.S. (1965). A linguistic study of Hudson, T. (2007). Teaching second language
cues and miscues in reading. Elementary reading. New York, NY: Oxford University
English 42, 639-643. Press.
Goodman, K.S. (1976), Reading: A Juel, C. (1988). Learning to read and write: A
Psycholinguistic guessing game. Journal of longitudinal study of fifty-four children from
the Reading Specialist,6, 126-135. first through fourth grade. Journal of
Educational Psychology, 80, 437-447.
Goodman, K. (1996). On reading. Portsmouth,
NH: Heinemnn. Juel, C., Griffith, P.L., & Gough, P.B. (1986).
Acquisition of literacy: A longitudinal study
Gough, P.B. (1972). One second of reading. In
of children in first and second grade.
J.F. Kavanagh & I.G. Mattingly (Eds.), In
Journal of Educational Psychology, 78, 243-
Language by ear and by eye (pp. 331-358).
Vol 1-Winter 2014- No. 3

255.
Cambridge, Mass: MIT Press.
Just, M.A., & Carpenter, P.A. (1980). A theory
Grabe, W. (2000). Reading research and its
of reading: From eye fixation to comprehension,
implications for reading assessment. In A.
Psychological Review,87, 329-354.
Kunnan (Ed.), Fairness and validation in
language assessment (Studies in Language Just, M.A., & Carpenter, P.A. (1987). The
Testing 9, pp. 226-62). Cambridge: Cambridge psychology of reading and language
University Press. comprehension. Boston: Allyn& Bacon.
 18 / IJRELT

Kintsch, W. (1988a). The role of knowledge in A cognitive analysis of fluent reading

discourse comprehension: A construction- andlearning to read. John Wiley & Sons
integration model. Psychological Review,95, Ltd.
163-182. Perfetti, C.A. (1985). Reading ability. Oxford:
Kintsch, W. (1998b). Comprehension: A frame Oxford University Press.
work for cognition. NewYork: Cambridge Perfetti, C. (1999). Comprehending written
University Press. language: A blueprint for the reader. In C.
Kintsch, W., & van Dijk, T. (1978). Toward a Brown & P. Hagoort (Eds.), Neurocognition
model of text comprehension and production. of language (pp. 167-208). Oxford: Oxford
Psychological Review,85, 363-394. UniversityPress.
Koda, K. (2005). Insights into second language Perfetti, C. (2007). Reading ability: Lexical
reading: A cross-linguistic approach. quality to comprehension. Scientific Studies
Cambridge University Press. of Reading, 8, 293-304.
Koda, K. (2007). Reading and language Perfetti, C., & Hart, L. (2001). The lexical basis
learning. Blackwell Publishing: U.S.A. of comprehension skill. In D. Gorfien (Ed.),
Kuhara-Kojima, K., Hatano, G., Saito, H., On the consequences of meaning selection
&Haebara, T. (1996). Vocalization latencies (pp. 67-86). Washington, DC: American
of skilled and less skilled comprehenders for Psychological Association.
words written in Hiragana and Kanji. Perfetti, C., & Hart, L. (2002). The lexical
Reading Research Quarterly, 31, 158- 171. quality hypothesis. In L. Verhoeven, C.
LaBerge, D., & Samuels, S.J. (1974). Toward a Elbro, & P. Reitsma (Eds.). Precursors of
theory of automatic information processing functional literacy (pp. 189-213).
in reading. Cognitive Psychology,6, 293- Amsterdam: John Benjamins.
323. Perfetti, C.A., &Lesgold, A.M. (1979). Coding
Levin, H. & Kaplan, E.L. (1970). Grammatical and comprehension in skilled reading and
structure and reading. In H. Levin & J.P. implications for reading instruction. In L.B.
Williams (Eds.), Basic studies on reading Resnick& P. Weaver (Eds.), Theory
(pp. 62-75). New York: Basic Books. andpractice of early reading (pp. 57-84).
Hillsdale, NJ: Erlbaum.
Lesser, J.M. (2007). Learner-based factors in
L2 reading comprehension and processing Pressley, M. (2006). Reading instruction that
grammatical form: Topic familiarity and works (3rded.). New York: Guilford Press.
working memory, Language Learning, 57, Pring, L. &Snowling, M. (1986). Developmental
2, 229-270. changes in word recognition: An
Macaro, E. (2003). Teaching and learning a information-processing account. Quarterly
second language: A guide to recent research Journal of Experimental Psychology, 38,
and its applications. Continuum: London. 395-418.
Mackworth, J.F. (1972). Some models of the Rayner, K., Bertera, J.H. (1979). Reading
reading process: Learners and skill readers. without a fovea.Science, 206, 468-469.
Reading research Quarterly, 7, 701-733. Rumelhart, D.E. (1977). Toward an interactive
Macnamara, J. (1970). Comparative studies of model of reading. In S. Dornic (Eds.),
reading and problem solving in two Attention and performance (pp. 573-603).
Vol 1-Winter 2014 -No. 3

languages. TESOL Quarterly 4, 107-116. New York: Academic Press.

McConkie, G.W., & Zola, D. (1981). Language Samuels, S.J. & Kamil, M.L. (1988). Models of
constraints and the functional stimulus in the reading process. In Carrell, P., Devine, J.
reading. In A.M. Lesgold& C.A. Perfetti &Eskey, D. (Eds.), Interactive approaches
(Eds.), Interactive process in reading (pp. to second language reading (pp. 22-36).
155-175). Hillsdale, NJ: Elrbaum. Cambridge: Cambridge University Press.
Mitchell, D.C., (1982). The process of reading: Schank, R.C. (1978). Predictive understanding.
 A Critical Overview of Models of Reading... / 19

In R.N. Campbell & P.T. Smith (Eds.), Torgesen, J.K., & Burgess, S.R. (1998).
Recent advances in the psychology of Consistency of reading related phonological
language–formal and experimental processes throughout early childhood:
Approaches (pp. 91-101).New York, NJ: Evidence from longitudinal, correlational
Plenum Press. and instructional studies. In J.L, Metsala&
Schankweiler, D., & Liberman, I. Y. (1972). L.C. Ehri (Eds.), Word recognition in
Misreading. A research for causes. In J.F. beginning reading(p. 161-188). Hillsdale,
Kavanaugh & I. G. Mattingly (Eds.), NJ: Erlbaum.
Language by eye and by ear (pp. 293-317). Van Dijk, T., & Kintsch, W. (1983). Strategies
Cambridge: MIT Press. of discourse comprehension. New York:
Segalowitz, N., Poulson, C. and Komoda, M. Academia Press.
(1991). Lower level of components reading Walczyk, J. (1995). Testing a compensatory-
skill in higher level bilinguals: Implications encoding model. Reading Research
for reading instruction. Reading in Two Quarterly, 30, 396-408.
Languages, ALLA Review, 8, 15-30.
Walczyk, J. (2000). The interplay between
Seidenberg, M.S., Tanehaus, M. K., Leiman, automatic and control processes in reading.
J.M., &Bienkowski, M. (1982). Automatic Reading Research Quarterly, 35, 554-566.
access of the meanings of ambiguous words
Walczyk, J., Marsiglia, C., Bryan, K., &
in context: Some limitations of knowledge-
Naquin, P. (2001). Overcoming inefficient
based processing. Cognitive Psychology, 14,
reading skills. Journal of Educational
489-537.
Psychology, 93, 750-757.
Stanovich, K.E. (1980). Toward an interactive-
Wagner, R.K., Torgesen, J.K. and Rashott, C.
compensatory model of individual
(1994). Development on reading related
differences in the development of reading
phonological processing abilities: New
fluency. Reading Research Quarterly,
evidence of bidirectional causality from a
Stanovich, K.E. (1986). Matthew effects in latent variable longitudinal study.
reading: Some consequences of individual Developmental Psychology, 30, 73-87.
differences in the acquisition of literacy.
Waters, G.S., &Caplan, D. (1996). The
Reading Research Quarterly, 21, 360-406.
measurement of verbal working memory
Stanovich, K.E. (1988). The language code: capacity and its relation to reading
Issues in word recognition. In S.R. Yussen comprehension. The Quarterly Journal of
& M.C. Smith (Eds.), Reading across the Experimental Psychology,49A, 51-79.
life span, New York: Springer-Verlag.
Zhang, S., &Perfetti, C.A. (1993). The tongue-
Smith, F. (1971). Understanding reading. twister effect in reading Chinese. Journal of
NewYork, MY: Holt, Rinehart and Winston. Experimental Psychology: Learning,
Stahl, S.A. (2003). Vocabulary and readability: Memory and Cognition, 19, 1-12.
How knowing word meanings affects
comprehension. Topics in Language
Disorders, 23, 241-274.
Vol 1-Winter 2014- No. 3