For me, reading (and comprehension in general) is an event, no part of which is to be discarded. In that event, which is the actualization of meaning, the deep structure plays an important role, but it is not everything; for we comprehend not in terms of the deep structure alone but in terms of a relationship between the unfolding, in time, of the surface structure and a continual checking of it against our projection (always in terms of surface structure) of what the deep structure will reveal itself to be; and when the final discovery has been made and the deep structure is perceived, all the “mistakes” – the positing, on the basis of incomplete evidence, of deep structures that failed to materialize – will not be canceled out. They have been experienced: they have existed in the mental life of the reader; they mean.

Literature, I argue, is the product of a way of reading, of a community agreement about what will count as literature, which leads the members of the community to pay a certain kind of attention and thereby to create literature. Since that way of reading or paying attention is not eternally fixed but will vary with cultures and times, the nature of the literary institution and its relation to other institutions whose configurations are similarly made will be continually changing.

In the procedures I would urge, the reader’s activities are at the center of attention, where they are regarded not as leading to meaning but as having meaning. The meaning they have is a consequence of their not be empty; for they include the making and revising of assumptions, the rendering and regretting of judgements, the coming to and abandoning of conclusions, the giving and withdrawing of approval, the specifying of causes, the asking of questions, the supplying of answers, the solving of puzzles.

Why should two or more readers ever agree, and why should regular, that is, habitual, differences in the career of a single reader ever occur? What is the explanation on the one hand of the stability of interpretation (at least among certain groups at certain times) and on the other of the orderly variety of interpretation if it is not the stability and the variety of texts? The answer to all of these questions is to be found in a notion that has been implicit in my argument, the notion of interpretive communities. Interpretive communities are made up of those who share interpretive strategies not for reading (in the conventional sense) but for writing texts, for constituting their properties and assigning their intentions. In other words, these strategies exist prior to the act of reading and therefore determine the shape of what is read rather than, as is usually assumed, the other way around. […] The assumption in each community will be that the other is not correctly perceiving the “true text,” but the truth will be that each perceives the text (or texts) its interpretive strategies demand and call into being. […] Interpretive communities grow larger and decline, and individuals move from to another; thus, while the alignments are not permanent, there are always there, providing just enough stability for the interpretive battles to go on, and just enough shift and slippage to assure that they will never be settled.

Every description is always an interpretation.

There always is a literal meaning because in any situation there is always a meaning that seems obvious in the sense that it is there indipendently of anything we might do. But that only means that we have already done it, and in another situation, when we have already done something else, there will be another obvious, that is, literal meaning. The stronger sense of literal, in which “single” is inseparable from “once and for all,” would itself make sense only if there were a meaning that was apprehensible apart from any situation whatsoever, a meaning that was not the product of an intepretation but was available indipendently. Every literal meaning comes to us with that claim, but it is a claim that seems to be supportable only because an interpretive act is already in force but is so embedded in the situation (its structure is the structure of the situation) that it doesn’t seem to be an act at all. We are never not in a situation. Because we are never not in a situation, we are never not in the act of interpreting. Because we are never not in the act of interpreting, there is no possibility of reaching a level of meaning beyond or below interpretation. But in every situation some or other meaning will appear to us to be uninterpreted because it is isomorphic with the interpretive structure the situation (and therefore our perception) already has.

The […] proponents of objective interpretation […] are afraid that in the absence of the controls afforded by a normative system of meanings, the self will simply substitute its own meanings for the meanings (usually identified with the intentions of the author) that texts bring with them, the meanings that texts “have”; however, if the self is conceived of not as an indipendent entity but as a social construct whose operations are delimited by the systems of intelligibility that inform it, then the meanings it confers on texts are not its own but have their source in the interpretive community (or communities) of which it is a function.

While there is no core of agreement in the text, there is a core of agreement (although one subject to change) concerning the ways of producing the text. Nowwhere is this set of acceptable ways written down, but it is a part of everyone’s knowledge of what it means to be operating within the literary institution as it is now constituted.

The canon of acceptability [of readings] can change. Moreover, that change is not random but orderly and, to some extent, predictable. A new interpretive strategy always makes its way in some relationship of opposition to the old, which has often marked out a negative space (of things that aren’t done) from which it can emerge into respectability. […] Rhetorically the new position announces itself as a break from the old, but in fact it is radically dependent on the old, because it is only in the context of some differential relationship that it can be perceived as new or, for that matter, perceived at all.

Not only must what one says about a work be related to what has already been said (even if the relation is one of reversal) but as a consequence of saying it the work must be shown to possess in a greater degree than had hitherto been recognized the qualities that properly belong to literary productions […] The new interpretation must not only claim to tell the truth about the the work (in a depedent opposition to the falsehood or partial truths told by its predecessors) but it must claim to make the work better. (The usual phrase is “enhance our appreciation of.”)