Only the text was both indisputably there and stable

«To the degree that this argument was influential (and it was enormously so) it constrained in advance the form any couterargument might take. In order to dislodge the effective fallacy, for example, one would have to show first that the text was not the self-sufficient repository of meaning and, second, that something else was, at the very least, contributory. This was exactly my strategy in the first of the articles presented in this book. I challenged the self-sufficiency of the text by pointing out that its (apparently) spatial form belied the temporal dimension in which its meanings were actualized, and I argued that it was the developing shape of that actualization, rather than the static shape of the printed page, that should be the object of critical description. In short, I substituted the structure of the reader's experience for the formal structures of the text on the grouds that while the latter were the more visible, they acquired significance only in the context of the former. This general position had many consequences. First of all, the activities of the reader were given a prominence and importance they did not have before: if meaning is embedded in the text, the reader's responsabilites are limited to the job of getting it out; but if meaning develops, and if it develops in a dynamic relationship with the reader's expectations, projections, conclusions, judgements, and assumptions, these activities (the things the reader does) are not merely instrumental, or mechanical, but essential, and the act of description must both begin and end with them. In practice, this resulted in the replacing of one question - what does this mean? - by another - what does it do? - with "do" equivocating between a reference to the action of the text on a reader and the actions performed by a reader as he negotiates (and, in some sense, actualizes) the text. This equivocation allowed me to retain the text as a stable entity at the same time that I was dislodging it as the privileged container of meaning. The reade was now given joint responsibility for the production of a meaning that was itself redefined as an event rather than an entity. That is, one could not point to this meaning as one could if it were the property of the text; rather, one could observe or follow its gradual emergence in the interaction between the text, conceived of as a succession of words, and the developing response of the reader.»

Stanley Fish, »Introduction, or How I Stopped Worrying and Learned To Love Interpretation», in Is There a Text in this Class?, the Authority of Interpretive Communities