CONCLUSION

Toward a Specification of Response

LET the last question of the Introduction be the first of the conclusion: what sorts of principles guide, or should guide, readers and critics in reading literary texts? The first four chapters sketched a rather narrow view of reading—as narrow, explicit, and coherent, in fact, as I could make it—and later two chapter considered ways that it might be modified to deal with certain texts. In its narrowest and simplest, this model of reading holds that readers process texts into the best propositional structure which renders a texv grammatical (according to the reader’s knowledge). The reader’s experiences of pursuing garden paths, making premature closures, and losing the thread constitute evidence that he is reading for sentence and propositional structure according to the strategies we have outlined. Assuming the operation of these strategies, we have been able to predict experiences of difficulty and misperception with some success, though, as we have seen, semantic, referential, and structural strategies work together in complex ways. This model presumably characterizes the way readers approach most texts and the way they might approach literary texts as well.

Consider now the reader who opens a text in Early Modern English. The spellings are unfamiliar, impeding word recognition; the punctuation is different and, therefore, sometimes misleading as a clue to sentence structure; the words themselves may have different grammatical and semantic properties, so that one is never sure whether the word one thinks one recognizes will function just as it does in Modern English; finally, the rules of syntax may be different, so that structural clues he customarily depends on may be absent. When he begins reading the text, he does not know the extent to which his knowledge and assumptions—and the strategies based on them—will serve him, and he must read with a certain tentativeness, making do with a conditional certainty. In effect, he analogizes: “if these words are words I know, and this text works according to rules I know, then this means'...'.” One will have to settle sometimes for a near match, as in readings we have considered correct in previous chapters ("to seek if she might attain her strayed champion"; “when they had hunted through (the labyrinth) by tract(?)”; “her swollen heart seemed to bereave her speech”). There is a strain involved for the reader in all of this—Spenser and Milton (and Shakespeare) are hard to read, probably impossible to read with our accustomed certainty that we have got the sentence right. At this point, the reader must either begin to adapt his customary strategies to the text or try to make out some sort of meaning without much regard for sentence structure. One can hack up a complex structure into simple chunks which are not wildly off base, as, for example, when one interprets Her swollen hart her speach seemd to bereaue as “she was sad, so she stopped talking, or, she choked up” (this was offered me as a paraphrase by a graduate student). But it is not on target either, since the seemed gets lost and turns out to be crucial (she is a hypocrite). To be sure, perhaps Spenser wants the reader to mistake the show of feeling for true feeling just as his somewhat dim hero does, but he is also trying to train the reader to see through hypocrisy.

One value of the extended analysis we have conducted is that it becomes possible to specify some of the adjustments the reader needs to make to get the sentence structure of these particular writers. When reading Spenser, one will have to process a lot of which when constructions, Subjectless subordinate clauses, and Object + to + verb sequences—with all of these, I have suggested, he may extend strategies he already has for other constructions by analogy, as it were. He will also find it useful to rely fairly heavily on line ends as indicators of what can be grouped with what, and it will serve him well to reduce reliance on serial-ordering strategies for identifying Subjects and Objects in favor of semantic (and thematic) strategies. The last adjustment is also useful for Milton, but line ends must be virtually disregarded as a clue to syntactic grouping. Some strategy for putting split coordinate elements back together will be useful—the one suggested is based on checking for shared semantic features. He will also have to process a goodly number of detached and strayed participials and relative clauses, and it will help to devalue his usual assumption that the noun phrase to the immediate left is the head to which the participial must be attached or which the relative clause modifies. When reading James, one will have to cope with parenthetical material spreading the main sentence elements. With Faulkner, one had better reduce his reliance on the canonical order strategy, since the S V A O order is common, and be wary of apparently transitive verbs. With Stevens, one must struggle to interpret each appositive and not let a string of them dissolve into a blur. These are practical suggestions: even when students do not find my suggested adjustment the best one for them, they do find that description of the sources of difficulty and some practice in coping with the constructions as types do make their reading easier, better, and more satisfying. One cannot, after all, experience most of the effects described in the previous chapter if he has given up trying to specify reference, attachment, and syntactic structure generally.

The model of reading sketched here does seem to furnish the beginning of a predictive and normative specification of response of the type Fish desires. In fact, it is a restriction of the more general notion of response that he presents: “The category of response includes any and all of the activities provoked by a string of words: the projection of syntactical and/or lexical probabilities; their subsequent occurrence or non-occurrence; attitudes toward persons, or things, or ideas referred to; the reversal or questioning of those attitudes; and much more" (Self-Consuming Artifacts, p. 388). It is the any that is the problem rather than the all. I have tried to distinguish between questions the reader must ask from those he may, and we can also distinguish assumptions he is likely to make from those that he may. It is a plain fact that not all readers find all of Fish’s 'difficult' passages as difficult as he says they are. Ralph Rader picks out some vulnerable examples which, he says, the mind’s “meaning-oriented direction-finder” guides one through,¹ and one might put the matter the other way as well: if a putative difficulty does not follow from an axiom of processing theory, one should be cautious about claiming that the reader will experience it. Fish claims that a that-clause at the beginning of a sentence will generally be taken to be factive (i.e., as if the fact had been ellipsed before that), but this does not follow from any principle of processing or fact of grammar: many verbs and adjectival predicates take non-factive that-clause Subjects, and an initial that-clause need not even be a Subject. The reasonable procedure is to look ahead to the governing predicate, at which point assignment of the clause as the Subject will settle the matter of its factivity. Obviously this methodological point will gain in force as the theory of language processing is developed.

The point of Rader’s “meaning-oriented direction-finder” is in part that Fish’s 'method' is too responsive to any response: it does not sort out responses, nor give any primacy to the processes of reconstructing propositional content. Rader is trying to defend the objectivity of structure in the text: we are not free to impose any grouping of words and elements that we find interesting nor do we attempt to exercise such freedom when we read. We are 'oriented' or direct ourselves when we read toward the construction of propositional content, and this orientation shapes and constrains our apprehension of the text. It may clarify the issues to distinguish between the goals or questions to be answered and the means of answering them. Some of the questions we try to answer as we read are those having to do with propositional content and reference. The very terms of these questions involve some apprehension of phrases, and hence reading cannot avoid positing and identifying some syntactic units. As we have seen, however, there are many ways of getting answers to these questions. The answers may not involve syntactic solutions and may be the incidental byproduct of some process of interpretation or inference. Interpretation has also its units and patterns of analysis (e.g., 'image7 clusters and oppositions) which may or may not coincide with syntactic units and patterns. We need to know more about comprehension and interpretation and how they can interact with perception before we can decide the question of the primacy of perception of propositional structure. But from where we are now, we can at least see a little more clearly the outlines of a model of reading.