This page is located at:

There's Reading . . . And Then There's Reading

There's Reading...and Then There's Reading

Process Models and Instruction

by Victoria Purcell-Gates
What does it mean when adults come to us for help with reading? What is it that they want help doing? What do they mean by 'reading'? What do we, as teachers of adults, think they mean? And what do we mean when we say we teach 'reading'?

These are not meant to be metaphysical questions. They are questions that are immediately relevant to our daily instruction and to our ongoing decisions about the needs of learners. As Harste (1984) put it, our assumptions do matter. If we are to guide and direct our students, we need to know where we are going, which paths are the most likely to get us there, and which paths are most likely to be dead ends. This means that, as teachers of reading, we must be cognizant of our underlying beliefs or theories of literacy development: how one begins to learn to read and how one develops from that point into an increasingly effective reader with a broadening range of texts. As teachers of adults, we must know -- in the sense of holding beliefs that are grounded in experience and information -- how this literacy development is affected by the knowledge, experiences, and cognitive stage of adults.

In this article, I will discuss some current, broad theories of the reading process and their implications for adult literacy instruction. I am defining ‘theory' as an explanation of a phenomenon like the reading process; a 'model' thus serves as a metaphor for that phenomenon, or in this case, reading process (Ruddell, Ruddell, & Singer, 1994). So while some of the theories I will explore have accompanying models, not all of them do -- and may never, because of the difficulty of capturing in a two-dimensional model the dynamism and socially-contextualized nature of the reading process described by the theories.

The pie of reading process theories can be cut several ways. I am going to categorize these approaches as they reflect different emphases on pieces of the process rather than attempt to draw totally competing theories like ‘top-down' versus ‘bottom-up.' Most reading theorists, such as Marilyn Adams, P. David Pearson, and Michael Pressley, have abandoned such all or nothing approaches and embrace some form of interactive theory of the reading process, while prioritizing different parts of it. First I will describe some theories that emphasize the role of letter and word recognition, then discuss theories that stress the role of comprehension over letter and word recognition, and finally, explore newer theories that highlight the role of social and cultural context in the process of reading. As you read these descriptions, you will probably recognize yourself in more than one camp. This is perfectly reasonable, given that all theories are attempts to characterize the same process.

Letter and Word Recognition is Key

"You can't read unless you look at the print and recognize the letters and then recognize the words."  Who can argue with this?

Several prominent theories detail the letter and word recognition parts of the reading process, with slightly different foci. Gough's (1972) model took an information processing lens and tried to describe the flow of information during the reading process. According to Gough's model, the reading process is linear, with letters being recognized first feature-by-feature by a visual system and then transferred to a sound (phonemic) system for recognition and held until the next letter is processed in the same way. When words are recognized (processed) they are held in working memory until they are processed for underlying meaning and finally understood as sentences and ultimately texts. In summary, Gough's description of the reading process is letter-by-letter and word-by-word in an additive fashion.

LaBerge and Samuels' model (1974) of the reading process brought us the concept of 'automaticity' in word recognition. Like Gough, word recognition is primary and needed for the later work of comprehension. Comprehension is made possible, according to LaBerge and Samuels, when readers no longer have to expend all of their cognitive attention on the recognition of letters and words. This is because it is not possible to consciously attend to, and thus process information from, more than one event or stimulus at a time. A familiar example of this is the ability of an experienced driver to navigate a route by car while engaging in a spirited conversation with a passenger. The driver must cease talking when an unexpected event occurs in the surrounding traffic that requires the transfer of attention. What makes driving in familiar circumstances while engaging in conversation possible is the automaticity of the driving processes. When a process becomes so familiar and 'expert,' usually through repetition and practice, it becomes ‘automatic' and does not require selective, or conscious, attention. This frees selective attention for other processes that do require it.

Decoding takes selective attention when it is unfamiliar or not fully mastered. And comprehension also requires attention. So, it is necessary to practice decoding skills to the point of mastery before the process of comprehension -- the real goal of reading -- is possible. The faster one becomes an automatic decoder -- recognizes words without having to break them down and ‘figure them out' -- the sooner one can attend to comprehending text.

Phonemic Awareness

Phonemic awareness is the final theory I will discuss that focuses on the primacy of letter and word recognition. This theory was recently eloquently explicated by Adams (1990), who, within a theory of the need for the attainment of automaticity, synthesizes a broad array of research to highlight the importance of letter-to-sound decoding in what we call ‘reading.' Adam's review of research is formulated to respond to theorists -- such as Frank Smith (1978) and Kenneth Goodman (1994) -- who claim that skilled readers do not decode every single letter from right to left during actual reading. Rather, these theorists claim that readers process words as wholes and may even skip words, parts of words, or whole sections of text without losing meaning. Adams, however, claims that skilled readers do process words letter-by-letter, and that these letters must be translated into sound units to be pronounced and thus understood. Developing readers must learn to do this to progress beyond the limits that are imposed by relying solely on visual recognition and memory of the thousands of words (Juel, 1991) one would need to recognize at sight for competent reading. However, perceiving language at the phoneme level (the smallest, isolatable unit of sound) -- hearing the individual sounds in cat (/k/ /ae/ /t/), for example -- is not easy and, for some readers, impossible without skilled intervention. Adams' model includes mental processors for the visual recognition of letters, context, and meaning as well as for the phonological processing I have briefly described. That makes her model similar to the interactive models discussed below. I put her in this letter-word recognition category because of the emphasis by her and those who cite her on the crucial role of accurate and automatic phonological processing.

Comprehension is Primary

"Reading is comprehending from print. Decoding is not reading. Only comprehension is reading."  Who can argue with that?

Standing opposed to those theorists discussed in the preceding section are those who do not want to view decoding as preceding comprehension either in on-line processing of print or in emphasis or sequence in reading instruction. These theorists see the process turned around, so to speak. Readers, they say, bring all of their experiences and background knowledge to the reading task. They expect meaning from print and they coordinate various language cue systems (pragmatic, syntactic, semantic, and graphophonic) to get at that meaning. Kenneth Goodman's model (1994) of the reading process probably typifies this theoretical stance best for most professionals. He originally referred to his model as a psycholinguistic one to reflect its language-processing essence; he now terms it a ‘Transactional Socio-Psycholinguistic Model of Reading,' capturing the recent theories of trans-action and the influence on language use of social settings and pragmatics: rules for language use that differ according to different social contexts.

According to this model -- and, as above, I am simplifying it without, I hope, changing its essence -- readers begin the act of reading by recognizing that they are reading and sampling and selecting from the visual array (letters and words) based on their predictions of what they expect to find. These predictions result from informed inferences about the meaning as it is supported and conveyed by the syntax and the graphophonics (letter/sound units). As readers read in this fashion, they confirm or disconfirm their predictions by using their knowledge of the different language cue systems to detect when meaning breaks down. When a breakdown in meaning occurs, and is detected, readers go back and correct, again employing the different language cue systems as they are appropriate. Goodman calls these actions by the reader ‘cognitive strategies.'

Reading, within this theory, is not linear but cyclical. Visual, perceptual, syntactic, and semantic cycles are constantly in play, each dependent on and enabling the others. In addition, reading, according to this theory, is goal oriented, with the goal being meaning. Using inference and prediction, the reader can continue to move toward meaning without completing the optical, perceptual, and syntactic cycles. Thus, as opposed to the theories described in the preceding section, comprehension is not dependent on linear, accurate, automatic decoding and letter-word recognition.

While all the reading theories discussed in this article include comprehension as an important aspect of reading, Goodman's theory -- and, by implication, all others categorized with it -- makes it the center of the process, beginning, during, and after. He specifies comprehension as a transaction between reader and text. Within a transactional theory of the reading process (Rosenblatt, 1994), the meaning does not reside solely in the text. Nor does it reside only in the reader. Rather, when reader and text interact -- i.e., when a reader reads a text -- the meaning that the reader constructs from the text is a third entity, not exactly what the author created nor what the reader could have created without a text. Therefore, each reading -- whether by the same reader or by a different one -- involves a different transaction and, thus, a somewhat different meaning.

Under this theory, beginning readers learn to read much the way they learned to talk: by being immersed in many purposeful and functional instances of language (print) use, they will feel compelled to join in, taking from the input the necessary cues to make sense and grow increasingly competent and conventional as readers and writers. As they participate in actual reading and writing, they will deduce and generalize from the linguistic input the underlying rules and patterns of the language in use.

Balanced Theories

"Learners need both to focus on meaning with real, authentic texts and to work on skills." Sounds reasonable, even the best of both worlds, for increasing numbers of reading teachers.

‘Balanced reading instruction' is a relatively new term and the concept probably rests on a theory of reading that is close to Rumelhart's (1994) Interactive Model of Reading, although with a sociocultural aspect that is absent in Rumelhart's model. Rumelhart acknowledges the reciprocal influence of different levels of knowledge held by a reader -- from letter featural knowledge of the features of letters to semantic knowledge -- and models these levels of knowledge as operating in interaction with each other. In other words, readers read by focusing on comprehension and on letter features roughly at the same time, granting that the reading event begins with graphic input. Rumelhart differs from Goodman (1994) in that he sees the reader as processing all of the different letters and words but acknowledges that meaning and syntactic context influence perception and recognition of letters and words.

A sociocultural view of literacy questions the view that literacy is a generic process that is the same for everyone in all instances (Santa Barbara Discourse Group, 1994). Rather, this theory holds that literacy -- and thus reading and writing -- looks different and is practiced differently by different social groups, such as students, churchgoers, business people, clerks, and so on. Concerned much more with social setting, structures, and influences on literacy practice, socioculturalists do not espouse any internal model of the reading process. As I understand this perspective, though, if such a model were to be conceived, it would reflect these sociocultural influences on the different levels of processing modeled by Rumelhart and Goodman.

Implications for Instruction

Skills-Driven Models

Teachers who operate under beliefs about reading process and development similar to those discussed in the section ‘Letter and Word Recognition is Key' will want to focus heavily on the skills and abilities needed to accurately and automatically recognize letters and words. They will see these skills as not only crucial but primary, in that they must be mastered before comprehension can take place. Therefore, students will be taught how to discriminate and recognize individual letters first (visual) and then to discriminate and assign individual letter, or letter combinations, to individual sounds. While there are many approaches to teaching decoding via sounding out, all such instruction that reflects the letter-word primacy of reading emphasizes decoding to automatic levels first and then increased focus on comprehension. While comprehension difficulties, in and of themselves, are possible, students who ‘cannot read' or ‘cannot read well enough' are viewed as potential decoding problems first until this is ruled out through diagnostic tests or diagnostic teaching.

Drill and practice in decoding skills, as well as phonemic awareness training, are recommended and practiced by proponents of this view. There is no concern about teaching ‘isolated skills' since, within this reading theory, to isolate pieces of the decoding process is to make them easier to grasp and practice. In addition, though, the authors of these theories suggest that their theories implicate the following instructional activities as well: repeated readings of text to increase accurate and automatic word recognition; providing texts that are easier to read if the student's problem is decoding; and the reading of interesting and emotionally satisfying texts as long as one is able to accurately decode the words.

Comprehension-Driven Models

Teachers who hold beliefs about reading similar to the Transactional Sociopsycholinguistic model of reading engage in radically different instructional decisions and activities. Because meaning, or comprehension, is the driving force of the reading process, and because this process involves cyclical cognitive strategy use and synergistic relations between the language cuing systems, learners must always be dealing with whole texts which are read for authentic purposes. Breaking the process down into isolated pieces for purposes of practice is counterproductive because the reader does not have all of the cue systems to use nor can the reader engage in full cyclical strategy use in pursuit of meaning. Thus, isolated skill teaching should never be done. Rather, teachers help their students get control of different parts of the process while they are reading -- and writing -- whole texts.

Teachers who operate within this theoretical paradigm will have their students reading texts that they want to read, for self-chosen purposes. They will teach their students that, when they read, they should make sense of the texts at all times. These teachers will deemphasize accurate reading, often actively dis-couraging it in favor of ‘translating' the text to the degree that it makes personal sense to the individual reader. While students in classes that focus on the primacy of letter-word recognition may often be asked to read aloud so that the teacher can check for accurate word reading, students in Goodman-like classes may never be asked to read aloud for accuracy checks because teachers do not want to over focus the students on accuracy at the expense of compre-hension. If students do read aloud, it will be to share favorite parts or for the teacher to gain insight into the ways in which a student is coordinating appropriate strategies in pursuit of meaning.

Since comprehension, from this theoretical stance, is a personal transaction between reader and text, teachers will eschew the traditional comprehension questions for activities like response journals, book shares, or book discussions, all of which involve individual responses to text that are not judged correct or incorrect. Teachers would never want their students to

read simplified or altered texts that are created for the sole purpose of teaching someone to read because the act of simplifying would corrupt the natural language of the text and thus prevent the reader from employing innate language knowledge and strategies.

Beginning readers must engage with whole texts, under this theory, so teachers often employ highly predictable or memorized texts for these readers. Patterned language, such as jingles, songs, and poetry, allows beginning readers to employ their predicting, sampling, and selecting strategies and to successfully engage in confirming and disconfirming, with correction. Language experience texts, obtained by writing down texts the learners dictate, are also often used. Since these dictated texts involve the readers' meaning-making processes as well as language, they are highly predictable and thus appropriate for this level of reader.

Integrated Models

Teachers who espouse a balanced view of reading instruction (McIntyre & Pressley, 1996) appear all along the reading process continuum, from the ‘skills' end to the ‘holistic.' They all view comprehension as the only purpose for reading and stress meaning-making in their instructional activities. The more skills-oriented will not hesitate to teach isolated skills as well as involving their students in reading and writing of au-thentic and compelling texts. The more holistic-oriented view skills teaching as best done in the context of authentic and compelling reading and writing.

Sometimes referred to as ‘whole-part-whole' instruction (Purcell-Gates, 1995, 1996), this teaching first involves students in purposeful (to the student) reading and writing, then pulls out some skills -- ranging from decoding to text structure and comprehension -- for focused work. The worked-upon skills are then ‘plugged' into the literacy activity for consolidation and practice in the actual process of reading and writing.

Acknowledging the role of social context in language use, these teachers strive to teach, in a direct and straight-forward manner, the different con-ventions and ‘ways of doing' associated with reading and writing texts in different genres and for different purposes and audiences, such as science writing and reading, newspaper writing and reading, document writing and reading, science fiction writing and reading, etcetera.

In Conclusion

In conclusion, as teachers it is critical that we identify our assumptions and beliefs -- and many of them may be quite implicit -- about what it is we are trying to help our students do: what process we assume is the process these students are trying to master. With our own process models made explicit, we can better choose materials, activities, instructional procedures, and assessments that serve to foster this mastery.


Adams, M.J. (1990). Beginning to read: Thinking and learning about print. Cambridge, MA: MIT Press.

Goodman, K.S. (1994). "Reading, writing, and written texts: A transactional sociopsycho-linguistic view." In R.B. Ruddell, M.R. Ruddell, & H. Singer (eds.), Theoretical models and processes of reading, 4th edition. Newark, DE: International Reading Association.

Gough, P.B. (1972). "One second of reading." In J.F. Kavenaugh & I.G. Mattingly (eds.), Language by ear and by eye. Cambridge, MA: MIT Press.

Harste, J. (1984). "Examining our assump-tions: A transactional view of literacy and learning." Research in the Teaching of English, 18, 84-10.

Juel, C. (1991). "Beginning reading." In R. Barr, M.L. Kamil, P. Mosenthal, & P.D. Pearson (eds.), Handbook of reading research: Volume 2 . New York: Erlbaum.

LaBerge, D. & Samuels, S.J. (1974). "Toward a theory of automatic information processing in reading." Cognitive Psychology, 6, 293-323.

McIntyre, E. & Pressley, M. (eds.) (1996), Balanced instruction: Strategies and skills in whole language. Norwood, MA: Christopher-Gordon.

Purcell-Gates, V. (1995). Other people's words: The cycle of low literacy. Cambridge, MA: Harvard University Press.

Purcell-Gates, V. (1996). "Process teaching with direct instruction and feedback in a university-based clinic." In E. McIntyre & M. Pressley (eds.), Balanced instruction: Strategies and skills in whole language, Norwood, MA: Christopher-Gordon.

Rosenblatt, L. (1994), "The transactional theory of reading and writing." In R.B. Ruddell, M.R. Ruddell, & H. Singer (eds.), Theoretical models and processes of reading, 4th edition. Newark, DE: International Reading Association.

Ruddell, R.B., Ruddell, M.R., & Singer, H., (eds.) (1994). Theoretical models and processes of reading, 4th edition. Newark, DE: International Reading Association.

Santa Barbara Discourse Group, (1994). "Constructing literacy in classrooms: literate action as social accomplishment." In R.B. Ruddell, M.R. Ruddell, & H. Singer (eds.), Theoretical models and processes of reading, 4th edition. Newark, DE: International Reading Association.

Smith, F. (1978). Understanding reading, 2nd ed. New York: Holt, Rinehart, and Winston.