Toward a Process-Intervention Model in Literature Teaching
Russell A. Hunt
[as published in College English 44:4 (April 1982), 345-357. Responded to by Martha A. Fisher in "The Cloze: A Comment on "Toward a Process-Intervention Model in Literature Teaching," in turn responded to by me, both in College English 46:5 (April 1984), 507-510.]
Once, without thinking much about it, we assumed that reading was an essentially passive process of decoding the marks on the page to sounds (that is, working from the phone up to larger matters) and then responding to the result as though it were oral language. It is now generally agreed that the process is almost the exact opposite, and that the reader works much more frequently from meaning down than from graphic display up.1 Using information and inferences drawn from knowledge of the world and from whatever portion of the text has already been read at any given point, the reader makes meanings-that is, forms expectations as to what is likely to be encountered, and then checks those expected meanings against the text, the actual marks on the page. These expectations (which we may term hypotheses, with the proviso that they are rarely consciously formed) are based on and tested against a hierarchy of cues in the text, cues that range from grammatical-syntactic, which tend to produce the shortest-term, narrowest "hypotheses," to longer-term, pragmatic matters like organization and tone and purpose.

In any given reading, then, many different combinations of cues and expectations will be operating simultaneously. During the course of this complex "psycholinguistic guessing game" (to use Kenneth Goodman's phrase), there is simply not time for the reader to see or use all-or, in fact, any more than a minuscule portion-of the marks on the page. What seems to happen is that the reader "tests hypotheses" by means of an inconceivably rapid and selective (and, not at all incidentally, often recursive and nonlinear) attention to a very few of those marks, as he or she works toward recognizing and apprehending the "deep structure" (in Ross Winterowd's extension of Chomsky's metaphor2) of the discourse.

There is little doubt that this model, in general if not in detail, reflects the mechanics of the fluent reading process. The extent of its relevance to the process of reading literature, however, is still often considered an open question. Reading at the "literary" or "aesthetic" level is a subject which psycholinguists and reading theorists have tended to avoid.

It seems clear, however, that the processes of reading literature must be to a large extent identical to those of reading nonliterary materialg -- aesthetic" reading is, after all, still reading -- and that the special characteristics we tend to think of as typical of the reading of literary texts are at least very closely analogous to those defined by the model I have briefly described.3 For instance, considerations of genre, plot, character development, form, poetic structure, and patterns of imagery and prosody can easily be seen as sets of cues, operating in the same ways as grammatical and syntactic cues operate in any fluent reading.

While there is much yet to be learned about the act of reading, one thing that is known beyond the possibility of doubt is that it is, in Michael Polanyi's sense, a "skillful performance." Such activities are learned; one can learn to do them better, with more skill and greater elegance. And they can be taught, as surely as swimming or carpentry can be taught. But, as Polanyi's description of the ways in which we learn such skilled activities makes luminously clear,4 they cannot be taught through giving the student information about them.

That such skills are so, rarely -- and, in general, so ineffectually -- taught is due at least in part to our inability to distinguish consistently between the different functions which the English teacher performs. To oversimplify slightly, departments of English have traditionally taught three different subjects: writing, reading, and literary study. They are, of course, closely interrelated, but they are not the same thing and should not be confused; to teach one is not to teach the other two. The most dangerous confusion, it seems to me, is that between teaching the reading of literature itself and teaching literary study.5 It is dangerous because for the student who is not a competent and confident reader to begin with it tends to substitute knowledge about for experience of literature. It tends, in fact, to insulate students from that experience by concentrating on (or allowing and encouraging students to concentrate on) the production of interpretations -- or, even worse, an interpretation.

This is particularly unfortunate because by and large, in introductory literature classes we are not, or should not be, training people to discuss literature (much less write about it). We ought rather to be concerned to teach them to experience and value it. The students who actually make up the overwhelming majority in introductory literature classes are not already committed to reading and to literature. They are the ones who, although they sometimes learn to mouth catch phrases about the profound humane values of literary art, and occasionally learn the lessons taught by books with titles like Writing Themes About Literature, do not generally become readers of literature. After they escape from the toils of Intro Lit, they read, when they read at all, books which neither require nor repay the kind of activity characteristic of a good, active, responsive, synthetic reading.

Traditionally, we have relied on osmosis and accident to teach those students, and we have measured our success by the achievement of the minority who come to us already reading actively and synthetically. We have merely expected our students to read; how and whether they learned it was their own affair.6 What the majority of them have learned is how to counterfeit the results of reading; they often learn to do so with astonishing skill. If we turn our -- and their -- attention back beyond the results to the process by which readers read, I believe we can avoid this.

Ironically enough it is in the third area for which English teachers have traditionally been responsible -- writing -- where we find suggestive models for making this change. Growing awareness of the importance of the process by which writing is generated has led many composition teachers to try to find ways to intervene in that process while it is underway. Growing understanding of how that process works has begun to produce strategies for intervention which are designed to help the students increase their awareness of, and thus their control over, the process. 7

The implications for the teacher of introductory literature are not difficult to draw. A growing understanding of the reading process puts us in a position to propose a parallel shift for literature teaching. To attend to and intervene in the process by which students read literature offers a way to help them get control over their reading and to make them better readers rather than merely better producers of interpretations. Wolfgang Iser in The Act of Reading (Baltimore, Md.: Johns Hopkins University Press, 1978) has suggested that it should be possible to gain access to the process:

Now it is certainly true that any response to any text is bound to be subjective, but this does not mean that the text disappears into the private world of its individual readers. On the contrary, the subjective processing of a text is generally still accessible to third parties, i.e., available for intersubjective analysis. This, however, is only possible if we pinpoint that which actually happens between text and reader. (p. 49)

The practical difficulty in the classroom is caused by the fact that, as Stanley Fish has acutely observed in "Literature in the Reader" (in Is There a Text in This Class? [Cambridge, Mass.: Harvard University Press, 1980]), "most methods of analysis [of literature] operate at so high a level of abstraction that the basic data of the meaning experience is slighted and/or obscured" (p. 30). In order to get at the ground on which critical analysis is built, we must find mechanisms for clearing away the abstractions. This is particularly difficult because one of the things most students have learned most thoroughly is the art of concealing that ground by creating a jungle of such critical abstractions. Many teachers are even better at such horticulture. (What is most remarkable about this jungle is that it often grows without any ground under it at all.) Classroom discussions in introductory literature classes tend very quickly, as most teachers know, to become discussions of abstractions: that is, interpretations and judgments and evaluations. Our most difficult problem, when everyone involved is vigorously protecting some particular set of abstractions, is to find a path back to concrete interaction with the text.

With many problems, as Yossarian pointed out, the solution lies in avoiding them. So it is with this one. If we concern ourselves not with the results of the reading process, but with the process itself, if we apply in our teaching what we know about that process, we will not face a need to get back to the text, for we will not have left it. What is necessary, then, is what Stanley Fish calls for in criticism: "a method, a machine if you will, which in its operation makes observable, or at least accessible, what goes on below the level of a self-conscious response" ("Literature in the Reader," p. 32). His description in this essay of his own critical method offers some hints as to how this might be done in the classroom.

Essentially what the method does is slow down the reading experience so that "events" one does not notice in normal time, but which do occur, are brought before our analytical attentions. It is as if a slow motion camera with an automatic stop action effect were recording our linguistic experiences and presenting them to us for viewing. (p. 28)

Fish, a practiced and fluent reader, has enough control over his own reading processes that this slowing down poses no problem for him. Our students, however, are at the mercy of the process; one way of characterizing their problems as readers might be to say that they act as though reading were something that happens to them rather than something they do and have control over. The question, then, is what methods we might use to slow down and interrupt the process of reading for our students so that they (and their teachers) can reflect on the transaction between themselves and the text, modify and enrich it, reassemble and re-perform it -- that is, gain some measure of control over it.

Once the question is posed in these terms, it is not difficult to begin to evolve methods: any book or article on the theory of reading, for instance, tends to make one conscious of the process and thus to stimulate the creation of such devices. So too does any activity which tends to slow down and bring to consciousness the process of one's own reading. A book like George Dillon's Language Processing and the Reading of Literature, for instance, will greatly increase any fluent reader's awareness of the processes involved in the unraveling of the complex syntax of Faulkner or Spenser. By analogy it will also increase a teacher's awareness of what students are doing as they process sentences by writers a practiced reader may regard as transcendently lucid.

One obvious way to slow the reading process down for our students is to read aloud. There are many reasons for reinstating the rather unfashionable practice of reading literature aloud in classes.8 One, particularly appropriate to intervention in the process, is that everyone is reading simultaneously, which allows for actually stopping the process so that a discussion may be conducted or a piece of exploratory writing engaged in.

There is something which should be remembered about reading aloud, however. It distorts the process of reading; in normal, fluent silent reading we do not (in spite of Stanley Fish's assertion that an "informed reader" should and does) attend to our "developing responses to the words as they succeed one another on the page.9 We respond not to individual words, but to ideas and speech acts and structure; and we assemble them not strictly in chronological order, detonating each word as we come to it as though a sentence were a string of firecrackers, but rather in large chunks, as our eyes and our attention synthesize meaning out of the graphic display on the page. Oral reading plods along in the wake of this dance; the meaning must be synthesized before the sentence can be delivered aloud. The better the reading, the more sophisticated an understanding it requires beforehand -- and the more the reading will embody an aesthetic interpretation as well as a literal comprehension of the language. Thus it is very easy for someone else's reading to supplant the student's own experience of the text. A good oral reading can also, of course, enhance and modify the reader's own experience, and as such it is a most valuable teaching technique. But it remains one whose tendency is as much upward toward abstraction and interpretation as downward toward concrete interaction with the text. (This tendency, of course, is not an argument against oral reading, but rather a warning that it can become another crutch by means of which students can allow someone else's interpretation to supplant their own firsthand experience of reading the text.)

Another strategy, which also allows the reading process to be stopped for examination but which ensures that the process which is stopped is that of the student's own reading, is the use of a prepared text, where (for instance) the display of the text ends and continues only after a page has been turned or a further section displayed on an overhead projector. This arresting of the process can also, of course, be done simply by indicating stopping points in the text.10 Any such strategy, however, will only work when the student's attention is focused on the experience of the text rather than on the pressure to guess "correctly" what will happen next, to "get the right answer." It normally takes some time and tact to lessen that fear of being "wrong" which, as John Holt pointed out fifteen years ago in How Children Fail (New York: Pitman, 1964: see especially chapter three, "Fear and Failure"), is so common a product of the educational situation.

The choice of a stopping point and the kinds of issues to which attention can be directed are obviously interrelated. Many questions -- of plot, and tone, and characterization of speaker, for instance -- will be posed almost automatically by the simple act of arresting the reading, as writers of serial fiction have known for two hundred years at least.11 Similarly, a lack of overt connection between two juxtaposed elements of a story or poem -- or essay -- can create an intolerable itch for resolution and explanation. In either case, attention can be drawn to the cause of the impetus to continue reading -- and, later, to the satisfaction involved in finding the event or the connection.12

Because the most dynamic and important property of the reading act is anticipation, one of the most fruitful sources of questions and discussion will be the reader's expectations -- and one of the most important things a student can learn is that wrong expectations are often exactly the right response. Reading Shirley Jackson's "The Lottery" a few paragraphs at a time, for instance, will dramatize a consistent pattern of wrong expectations -- hich are precisely, of course, what the story needs in order to work. 13

More subtle, and perhaps harder to attend to, are matters involving frustrated or gratified grammatical and syntactic expectations. The first line of Shakespeare's Sonnet 33, for instance,14 creates an expectation that one phrase is finished and a new one will start with the next line: "Full many a glorious morning have I seen." In fact, of course, Shakespeare is actually going to continue the phrase, and thus enjamb the lines very tightly.

Full many a glorious morning have I seen
Flatter the mountain-tops with sovereign eye.
In context, of course, the momentary hesitation about phrase boundaries is so quickly resolved that it may not be noticed -- though, in Stanley Fish's phrase, it is "an event that does occur" in many readers, and one that has consequences.

Larger questions of what may be expected (and why) can, of course, be raised; attention can be focused on diction and tone as a way of creating expectations in the reader. Stopping after the first sentences of a text like Orwell's "A Hanging," for instance ("It was in Burma, a sodden morning of the rains. A sickly light, like yellow tinfoil, was slanting over the high walls into the jail yard."), offers an opportunity to attend to the texture of the passage and how it affects the anticipations of a reader.

Even broader are the sorts of tacit hypotheses we form regarding long-term matters like point of view. In Joyce's "Ivy Day in the Committee Room," for instance, the constraint of the purely visual, camera-eye point of view is part of the information we can use to assemble a pattern of expectations of what is to come.

Old Jack raked the cinders together with a piece of cardboard and spread them judiciously over the whitening dome of coals. When the dome was thinly covered his face lapsed into darkness but as he set himself to fan the fire again, his crouching shadow ascended the opposite wall and his face slowly re-emerged into light. It was an old man's face, very bony and hairy.
Simply to stop at that point, without allowing access to the continuation of the story, has almost automatically the effect of raising questions about authorial choices and patterns regarding, for instance, limitations on the physical placement of the eye and the consequent selection of detail.

It is, then, clearly not difficult to find stopping places which will generate comment and direct attention to the process of reading. Fish's notion that "everything counts" -- that there are no long stretches in literature where "there is no response worth talking about because nothing much is happening" (p. 63) -- reminds us that virtually anything can serve as the springboard for a useful discussion or writing assignment, or simply a moment of contemplation. The difficulty lies in finding places which will stimulate reflection and generate comment likely to help students move on another step in their awareness of reading as a process. I know of no easy rule here. Education is not an efficient process. But a teacher who understands the process will, I am convinced, have no difficulty finding such stopping places.

There are, of course, difficulties with interrupted reading, and surely no one would argue in favor of using it exclusively. Like reading aloud, it depends in part on conceiving of reading as a chronological word-by-word and phrase-by-phrase and sentence-by-sentence affair, when of course it is no such thing. A device which allows more scope for a reader's power of synthesis utilizes a prepared text from which selected words (or sentences, or paragraphs) have been eliminated. Something very like this procedure is used in reading research. The so-called "cloze procedure" was originally proposed as a way of measuring the readability of texts and depends on eliminating words arbitrarily.15 What I am describing, however, involves a very careful selection of items to eliminate. Even more important, I am not suggesting it as a device for measuring either reading skill or the readability of texts, but rather for directing attention to the ongoing interaction between reader and text.

Through this device much can be done with the tension between what one might expect and what one actually gets. In a poem like Robert Frost's "Design," for instance, specific words can be eliminated: "I saw a __________ spider, fat and white." Common student guesses as to what might appropriately fill that blank are "repulsive" and "disgusting." (At some point, one might discuss the pressure of scansion on a reader's expectation, and what a reader who is subject to that pressure is bringing to the poem.) Less common but more fruitful is a word like "bloated." The information that "dimpled" was Frost's actual choice can lead to an illuminating consideration of the relations between connotation and context, of various reactions to spiders (and to "dimpled"), and of the role of closeness and accuracy of observation in making writing effective. It can only do these things, however, if there has been an opportunity for the students to consider the question in the absence of the "right" answer, and if it is clear that they are not being judged on their ability to guess correctly. Under such circumstances the confrontation with the blank space in a line like Herbert's "Sweet rose, whose hue and brave" can engender genuine admiration for the brilliance of "angry" in the context of this poem ("Virtue").

More important, perhaps, in helping students become conscious of their synthetic role as readers is what can be done with respect to structural response by this means. The careful elimination of words and phrases from a poem, for instance, can draw a student's attention to the way in which it is the reader who is responsible for synthesizing the structure of a poem like "Virtue":

Sweet day, so cool, so calm, so bright,
The bridal of the earth and sky:
The dew shall weep thy fall tonight,
For thou must die.

Sweet rose, whose hue __________ and brave
Bids the rash gazer wipe his eye:
Thy root is ever in its grave,
For thou must die.

Sweet spring, full of sweet ____________ and _____________
A box where sweets compacted lie:
My music shows ye have your closes,
And _________ must die.

Only a sweet and virtuous soul,
Like seasoned timber, never gives:
But though the whole world turns to coal,
Then chiefly lives.

The inference (from the rhyme scheme) that line nine probably ends in "roses" can lead to the perception of the pattern that if stanzas one and two begin with "days" and "roses" then line nine might include them both. Consequently, a reader may guess that the substitution in line twelve of "all" for "thou" (which is the virtually universal first guess) reinforces the cumulative way the structure of the poem leads toward the generalization of the fourth stanza.

Other possible uses of this "cloze procedure" will not be hard to imagine. A more complex device involves juxtaposing a doctored text with the real one (usually without revealing which is the "authorized" version) and initiating a discussion of the differences. Students who might never have noticed it otherwise may discover the leveling effect, and the consequent characterization of the narrator, of the obsessive use of simple coordination in a story like Hemingway's “After the Storm," which (in the original) begins:

It wasn't about anything, something about making punch, and then we started fighting and I slipped and he had me down kneeling on my chest and choking me with both hands like he was trying to kill me and all the time I was trying to get the knife out of my pocket to cut him loose. Everybody was too drunk to pull him off me. He was choking me and hammering my head on the floor and I got the knife out and opened it up; and I cut the muscle right across his arm and he let go of me. He couldn't have held on if he wanted to.
Applying a few elementary sentence combining techniques to improve Hemingway's syntactic maturity yields an entirely different impression:
It wasn't about anything; it was just something about making punch, but suddenly we started fighting. I slipped and in a moment he had me down kneeling on my chest, choking me with both hands like he was trying to kill me. All the time, I was trying to get the knife out of my pocket to cut him loose. Everybody was too drunk to pull him off me. He was choking me and hammering my head on the floor, but I got the knife out and opened it up. I cut the muscle right across his arm, and so of course he let go of me; he couldn't have held on if he'd wanted to.
A discussion or an impromptu exploratory writing assignment concerning the students' reactions to each of these passages, and to the differences between them, will focus attention on concrete aspects of the texts and on the subtleties of the relation between the speaker and the reader in a way that no abstract question about Hemingway's use of style as a characterizing device could ever do. (The most useful part of the discussion is often the argument over whether the second alternative is not better than the first because it better follows the kinds of rules for writing most of the students have been taught at some point in their educational careers.)

An important benefit of this device is that it helps students apprehend the distinction which the Russian formalist critics make between fabula and sjuzhet, or what Seymour Chatman terms "story" and "discourse." Response to this distinction is an important characteristic of "literary" as opposed to less complex and multileveled forms of reading, and it is commonly a very difficult step for students to take.

Another way of approaching the same distinction involves rewriting a story from a different point of view and presenting the students with both alternatives. (Were "After the Storm" told by an omniscient narrator. for instance, it might begin like this: "It wasn't about anything, just something about making punch, but they started fighting anyway. Jeff slipped so that his opponent got him down . . . .") The discussion here might well include even a point like the fact that in order to tell the story from this point of view the character must be named, whereas Hemingway's original is not.

This exercise can lead to one even more profound: asking students themselves to rewrite part of a story from a different point of view. Here they will learn more about the constraints and advantages of choosing a certain narrator or set of conventions than any abstract discussion could teach them, and such a task is in itself a startlingly efficacious way to slow the reading process, to help students attend to what they are doing as they read.

What all these devices have in common is that they have at their center the very act of reading itself. They focus attention away from interpretation and evaluation, back toward the process which takes place before a formal interpretation can be settled on or an evaluation arrived at.16 All of them are ways to slow or interrupt the reading process so that what is happening can be thought about, discussed, perhaps modified or enriched. Although short pieces of exploratory writing are appropriate, none of these devices is pointed toward the production of essays about literature. All are concerned to implement ways of asking effectively the question suggested by Stanley Fish: "What does that (word, phrase, sentence, paragraph) DO?" as opposed to the more traditional, "What does it mean?"17 All are attempts to help students, in Fish's phrase again, "to see the value of considering effects and begin to be able to think of language as an experience rather than a repository of extractable meaning" (p. 65). All, finally, are intended to be ways of helping students improve the way they read (as opposed to improving the quality of the products of that reading). As teachers of literature, after all, our most central concern should be with the way our students approach the next text they read, not with the results of their approaches to the last text.

Responding to cues and making choices about hypotheses, assuming an appropriate stance, assembling wholes out of parts -- all this is automatic for skilled readers, so automatic that they are almost never conscious of what they are doing and very often assume that it is in the text itself that all these things are done. (Studies of narrative cohesion, for instance, almost always assume that cohesion is to be found in the narratives themselves, rather than in the connecting that the reader is invited or enabled to do.) As we learn more about the reading process, however, we become more aware of the extent to which the choices and connections are made not in the text, but rather in a transaction between text and reader -- and that it is not uncommon for readers to be unsuccessful in their end of the transaction. Making readers more conscious of their activity, of their part in the transaction, may be a way of helping them to succeed more regularly. Self-consciousness is not, however, an end in itself. Ideally, the consciousness of the process should fade-as with driving a car, or playing a guitar, or any other skillful performance -- and the readers should ultimately do it automatically (in Michael Polanyi's useful terms, the awareness of all these matters must be subsidiary or tacit for the activity to be effectively and efficiently performed).

Like talking -- or the basic skill of literacy -- the reading of literature is an activity that ought not to have to be taught at all. The more we learn about such activities, the clearer it becomes that they employ skills which human beings find easy to acquire, and which are acquired most easily and most thoroughly -- one is tempted to say "most naturally" -- when they are engaged with in a real context, as wholes, rather than broken down into chunks and taught. We learn such activities most effectively, Polanyi argues, in an apprenticeship context -- i.e., in the presence of the activity being carried on in the "real world."18

On the face of it, such an observation would suggest that the best way to teach literature is the way it has traditionally been taught: the teacher performs the act of reading and interpreting for and with the students, who learn it as a whole. Unfortunately, it is perfectly clear that such a method simply does not work for the vast majority of the students most college English teachers face in courses in introductory literature. What's wrong?

Here is a hypothesis. Students who are going to learn "naturally" will have done so before they enter university. The others, the majority, have not, for whatever reasons, learned and now it is too late for the process to occur "naturally." Like children taught to read by some straitjacket method such as phonics or look-say (many university-level nonreaders were almost certainly taught to read in such ways), they now require measures which perhaps might be called remedial.

Basically, there are two ways of going at the teaching of any skillful activity. One, the more natural, is to increase attention to the focus of the activity: e.g., a batting coach urging a child, "Just meet the ball. Watch it till it hits the bat." The other -- less natural, perhaps -- involves focusing on activities which should be subsidiary and attempting to modify them. Here the batting coach may pull the batter's focus from the ball meeting the bat to the grip or the stance. "Hold the bat this way. Bring it back this far before you swing. As the bat begins to move into the ball, step forward. Keep your swing level."

Clearly, a batter who continues to concentrate on grip or on which foot is bearing weight is not very likley to send a screaming liner past the shortstop. On the other hand, if that batter can change grip or stance and then make the new action subsidiary, tacit, part of the unconscious activity focused on meeting the ball, a screaming liner may well be exactly what happens. The important thing for the coach to remember is not to make something that ought to be subsidiary into the main point.

Perhaps the central reason that attending to the process of reading is a promising method of teaching literature has to do with confidence. In part, it is an attempt to produce students who approach a text with some measure of confidence -- confidence that the text will make sense, that its elements do cohere and that they will be able to bring their synthetic skill to bear on the text and make the elements combine into a whole. The most general and universal difference between good readers and poor ones is that the good ones have this confidence.19 The best way to acquire it is to have read effectively and successfully. No critical method or set of abstractions -- much less anyone else's reading of a text -- can move a student toward its acquisition. Faced with students who have not read effectively and successfully in childhood, or who have been convinced by years of miseducation that they cannot do so, it seems clear that what a teacher ought to do is attempt to help them through that process, to increase their faith in their own ability and in the fact that texts do repay attention with coherence and satisfaction.

In the 60s and 70s literature teachers often attempted to give their students such confidence by pretending that any interpretation was as good as any other, that each student's view of literature was equivalent to any other. They didn't believe us. We didn't believe ourselves. It didn't work. But it may just be possible, if we can escape the trap set by our obsessive attention to the product -- interpretations, evaluations, discourse about texts -- and attend to the process, that we can find a way to help our students attain (in many cases, reattain) that confidence.


1. Publications which I have found particularly important to the development of this view of reading include: Frank Smith, Understanding Reading-A Psycholinguistic Analysis of Reading and Learning to Read (New York: Holt, Rinehart and Winston, 1971); Kenneth S. Goodman, "Behind the Eye: What Happens in Reading," in Reading: Process and Program, by Kenneth S. Goodman and Olive S. Niles (Urbana, Ill.: NCTE, 1970), and Goodman's "Reading: A Psycholinguistic Guessing Game," in Theoretical Models and Processes of Reading, ed. H. Singer and R. B. Ruddell (Newark, Del.: International Reading Association, 1970). George L. Dillon's Language Processing and the Reading of Literature: Toward a Theory of Comprehension (Bloomington: Indiana University Press, 1978) is extremely useful in suggesting ways in which understanding of basic language processing activities can be applied to more "literary" kinds of reading.

2. See W. Ross Winterowd, Contemporary Rhetoric: A Conceptual Background with Readings (New York: Harcourt Brace Jovanovich, 1975), pp. 164-165. Winterowd suggests that perhaps an "analogical bridge" can be built "between the grammar of the sentence and the grammar of form." See also Seymour Chatman's Story and Discourse (Ithaca, N.Y.: Cornell University Press, 1978) for a characterization of what Chatman calls "reading out" and defines as "decoding from surface to deep narrative structures" (pp. 37-42).

3. Louise Rosenblatt's now well-known distinction between "efferent" and "aesthetic" reading is clearly relevant here. Her initial insistence in The Reader, the Text, the Poem (Carbondale: Southern Illinois University Press, 1978), that "the reader performs very different activities" during these processes (see especially pp. 22-30) has more recently been softened. In "On the Aesthetic as the Basic Model of the Reading Process," in Bucknell Review: Theories of Reading, Looking and Listening, ed. Harry R. Garvin (Lewisburg, Pa.: Bucknell University Press, 1980), she suggests that it is not unreasonable to consider aesthetic reading as "the source of the basic model for all reading processes." Whatever the state of the empirical question, it seems to me useful and productive to think of the two as analogous.

4. Personal Knowledge: Toward a Post-Critical Philosophy (Chicago: University of Chicago Press, 1958). See especially chapter four. "Skills," pp. 49-65.

5. Hugo Verdaasdonk and C. J. van Rees make this distinction usefully in "Reading a Text vs. Analyzing a Text:” Poetics, 6 (1977), 55-76. although their primary point is that "analysis" is not more objective or accurate than "reading." Roland Barthes has suggested that all critical reading is and must be re-reading (S/Z: Essais [Paris: Editions du Seuil. 1970]. pp. 22-23) and is thus different in kind from the sort of transaction that noncritical readers have with a text. Keith Fort extends this point to suggest reasons why the two may be mutually inimical, in "Form, Authority, and the Critical Essay," College English, 32 (1971).629-639. The article is reprinted in Winterowd, Contemporary Rhetoric, pp. 172-183. In introducing it Winterowd comments: "As far as literature is concerned, the critical expository essay has been disastrous for English classes. Responses to art are many and varied, and only the professional responds with a formal critical essay. Needless to say, the object of the English class is to produce sensitive, eager readers, not professional responders" (p. 172). I am not at all sure that it is needless to say it.

6. This point has been made with regard to the writing process, I think first by Francis Christensen. See his Notes Toward a New Rhetoric: Nine Essays for Teachers, second edition, ed. Bonnie-jean Christensen (New York: Harper and Row, 1978). In "A Generative Rhetoric of the Sentence" (pp. 23-44) he says. "In composition courses we do not really teach our captive charges to write better-we merely expect them to. And we do not teach them how to write better because we do not know how to teach them to write better. And so we merely go through the motions" (p. 25).

7. A paper delivered at the Wyoming Conference on Freshman and Sophomore English by James A. Reither, and currently being prepared for publication, outlines a number of strategies devised by one composition teacher for intervening in the writing process (as opposed to teaching students about the process). Frank Smith's brilliant new book, Writing and the Writer (New York: Holt, Rinehart and Winston, 1982), attentive to language processes throughout, ends with two chapters centrally concerned with intervention in the writing process.

8. Roger Shattuck, in "How to Rescue Literature," New York Review of Books, April 17, 1980. pp. 29-35, offers an eloquent statement of many other reasons than the one I offer for reinstituting this practice.

9. "Literature in the Reader," pp. 27, 46, 59. It is interesting that Fish varies the phrase; twice he refers to the words succeeding one another "in time." rather than "on the page." but he appears to make no distinction between the two, as though the reader were as much at the mercy of the order of the text as a listener would be to the order of spoken words.

10. Similar devices are sometimes suggested in teaching reading at an elementary level, as aids to comprehension and to "reading for meaning." See Russell Stauffer's The Language-Experience Approach to the Teaching of Reading, second edition (New York: Harper and Row, 1980), especially his description of "Directed Reading-Thinking Activities," pp. 184-213; and David Doake, "Comprehension and Teaching Strategies," in New Horizons in Reading, ed. J. E. Merritt (Newark, Del.: International Reading Association, 1976), pp. 125-140. Doake's suggestion that we insert "adjunct questions" into reading activities (pp. 136-137) is recommended by John Downing in Reading and Reasoning (New York: Springer-Verlag, 1979): see pp. 53-54. An objection that colleagues have made to this procedure is that it offers an insult to the reading process and may traumatize it or even send it into shock. I know of no evidence to support this view; the reading process is a tough and resilient one, as any reader of novels (which are virtually always interrupted and resumed) knows. Italo Calvino's recent novel, If on a Winter's Night a Traveller, makes this resilience almost a structural principle.

11. Wolfgang Iser has discussed what he calls "cutting techniques" in The Act of Reading (pp. 191-192). He points out, for instance, that "one common means of intensifying the reader's imaginative activity is suddenly to cut to new characters or even to different plotlines, so that the reader is forced to try to find connections between the hitherto familiar story and the new, unforeseeable situations." See also Iser's "Indeterminacy and the Reader's Response in Prose Fiction," in the 1971 English Institute Essays, Aspects of Narrative, ed. J. Hillis Miller (New York: Columbia University Press, 1971), pp. 14 ff.

12. In one sense, attending to this pressure to continue reading is attending to the form of the work itself. In Counter-Statement (Los Altos, Cal.: Hermes Publishing, 1953), Kenneth Burke defines form in literature as "an arousing and fulfillment of desires" (p. 124; the essay, "The Nature of Form," is reprinted in Winterowd, Contemporary Rhetoric, pp. 183-198).

13. One of the characteristics of a more "literary" or sophisticated reading, as opposed to what Rosenblatt terms an "efferent" reading, is that such responses do not evaporate on second reading, but are experienced again, tempered by a growing awareness of the patterns underlying them. It is true, however, that many students do not imaginatively reconstruct a "wrong" expectation on their second reading. The inability or unwillingness to do this may be one of the defining characteristics of readers who have not learned to deal with literary texts on their own terms.

14. This example is drawn from George Dillon, Language Processing and the Reading of Literature, p. 31, who in turn draws it from Stephen Booth, An Essay on Shakespeare's Sonnets (New Haven. Ct.: Yale University Press, 1969), pp. 55-56.

15. The earliest mention of "cloze procedure" in this connection is W. L. Taylor, "Cloze Procedure: A New Tool for Measuring Readability," Journalism Quarterly, 30 (1953), 415 ff. I have not been able to discover the origin of the word "cloze," but it is apparently related to the concept of "closure" as developed in gestalt psychology.

16. Susan Sontag, of course, stressed two decades ago the destructive effects of focusing on the production of interpretations of literature (see her widely reprinted essay, "Against Interpretation"). Fish extends the point more directly to the classroom by noting that "the act of interpretation is often so removed from the act of reading that the latter (in time the former) is hardly remembered" ("Literature in the Reader," p. 52).

17. The phrase, which runs through "Literature in the Reader," was I think first used by Susan Sontag: "None of us can ever retrieve that innocence before an theory when art knew no need to justify itself, when one did not ask of a work of art what it said because one knew (or thought one knew) what it did" (Against Interpretation and Other Essays [New York: Dell Publishing, 1969]. p. 14).

18. This idea is important throughout Personal Knowledge, but perhaps the statement most directly relevant to language learning occurs in Part Two, "The Tacit Component": "All arts are learned by intelligently imitating the way they are practiced by other persons in whom the learner places his confidence. To know a language is an art, carried on by tacit judgments and the practice of unspecifiable skills. . . . Spoken communication is the successful application by two persons of the linguistic knowledge and skill acquired by such apprenticeship, one person wishing to transmit, the other to receive, information" (p. 206). More specifically relevant to the teaching situation, perhaps, is the statement in the section on "Skills": "To learn by example is to submit to authority. You follow your master because you trust his manner of doing things even when you cannot analyze and account in detail for its effectiveness. By watching the master and emulating his efforts in the presence of his example, the apprentice unconsciously picks up the rules of the art, including those which are not explicitly known to the master himself” (p. 53).

19. This "confidence" is clearly related to the issue, often raised in discussions of reading at the primary level, of metalinguistic awareness. John Downing, for instance, argues at some length that children must learn about reading in order to acquire sufficient motivation to practice it. (See Reading and Reasoning, especially chapters two and three.) That such knowledge might be (perhaps ought to be) tacit, in Polanyi's sense, Downing does not seem to consider. In both Understanding Reading and in Writing and the Writer, Frank Smith makes a strong case that it is often destructive to attend to such matters consciously.