PREFACE

Quite a number of years ago Harold Whitehall declared that “criticism in English ought to have a hunger for a sound linguistics” and argued that this was so because “no criticism can go beyond its linguistics.” These claims have not won universal assent. Indeed, twenty years later, there are many who remain unconvinced that linguistics has much to contribute to the understanding of literature or vice-versa. The unsettled status of Whitehall’s claim notwithstanding, I would like to make a comparable one that criticism cannot go beyond its theory of reading, and that most critics operate with an implicit theory that might be characterized as 'watered-down Empson.' This is not a bad theory, by the way, except that it is very inexplicit, and I think criticism would benefit from an attempt to construct a model of reading as explicit and articulated as current knowledge will warrant. The pages that follow are an attempt to suggest the outlines of such an account.

The phrase current knowledge, however, is quite misleading in so far as it suggests a widely accepted, empirically based model of how we comprehend texts. In fact, there are hunches and hypotheses, small portions of which are supported by experimentally obtained data of arguable significance. The theory of language processing is in the very early stages of development, and a discussion of reading based primarily on the reading of difficult literary texts may contribute something to the general understanding of reading that one hopes will emerge from the current ferment.

I will concede at the outset the force of what might be the first axiom of reading research, namely that there are many ways of reading which vary with the type of text and the purposes of the reader—the notion of 'ordinary' or 'plain' reading is probably a chimera, the close relative of 'ordinary language.' What I will describe is an idealization of sorts—what we might do if we tied off every loose end and read not just for the gist but for as exact an account of the text as could be had. This first axiom leaves open the possibility that there may be certain strategies especially appropriate for certain types of texts, say for literary texts, but our procedure will be to assume not, until and unless the texts force us to admit such possibilities.

A few words about the examples: they are cited as proofs or evidence of the operation of certain strategies, but they will not function as such unless the reader struggles to read each of them and does not skip to the analysis. This study does not report the results of experiments; rather, it invites the reader to engage in analysis of his own cognitive processes, and the persuasive force of the book rests on whether the reader experiences the effects described. Read in this fashion, the book becomes quite dense, and so I have kept it as short as possible.

I would like to thank those who fed me examples and puzzled with me over them: my students Judith R. Palmer, Larry Platter, and Helen Twite, and my colleagues Walter Johnston and Philip R. Headings. Another colleague, Jeanette Clausen, gave the completed manuscript a very close reading and offered many useful comments that improved it both in content and in clarity, and the copy editor, Sharon Smith, patiently combed out numerous tangles. Finally I thank my friend and erstwhile collaborator Frederick Kirchhoff, who read and criticized early drafts and late and whose interest and enthusiasm provided me with vital stimulation.