Interpretation

«In other words interpretative communities are no more stable than texts because interpretative strategies are not natural or universal, beu learned. This does not mean that there is a point at which an individual has not yet learned any. The ability to interpret is not acquired; it is constitutive of being human. What is acquired are the ways of interpreting and those same ways can also be forgotten or supplanted, or complicated or dropped from favor ("no one reads that way anymore"). When any of these things happens, there is a corresponding change in texts, not because they are being read differently, but because they are being written differently.
The only stability, then, inheres in the fact (at least in my model) that interpretative strategies are always beign deployed, and this means that communication is a much more chancy affair than we are accustomed to think it. For it there are no fixed texts, but only interpretative strategies making them, and if interpretative strategies are not natural, but learned (and are therefore unavailable to a finite description), what is it utterers are in the business of handing over ready-made or prefabricated meanings. Those meanings are said to be enconded, and the code is assumed to be in the world independently of the individuals who are obliged to attach themselves to it (if they do not they run the danger of being declared deviant). In my model, however, meanings are not extracted but made and made not by encoded forms but by interpretative strategies that call forms into being. It follows then that what utterers do is give hearers and readers the opportunity to make meanings (and texts) by inviting them to put into execution a set of strategies. It is presumed that the invitation will be recognized, and that presumption rests on a projection on the part of a speaker or author of the moves he would make if confronted by the sounds or marks he is uttering or setting down.
It would seem at first that this account of things simply reintroduces the old objection; for isn't this an admission that there is after all a formal encoding, not perhaps of meanings, but of the directions for making them, for executing interpretative strategies? The answer is that they will only be  directions to those who already have the interpretative strategies in the first place. Rather than producing interpretative acts, they are the products of one. An author hazards his projection, not because of something "in" the marks, but because of something he assumes to be in his reader. The very existence of the "marks" is a function of an interpretative community, for they will be recognized (that is, made) only by its members. Those outside that community will be deploying a different set of interpretative strategies (interpretation cannot be withheld) and will therefore be making different marks.»

Stanley Fish, «Interpretating the Variorum», in Is there a text in this class?