SOURCE: “Fish's Argument for the Relativity of Interpretative Truth,” in Journal of Aesthetics and Art Criticism, Vol. 48, No. 3, Summer, 1990, pp. 223-30.

[In the following essay, Stecker examines Fish's theoretical claims about the contextual modes of literary meaning and interpretation, as presented in Is There a Text in This Class; Stecker concludes that Fish's effort to assert the validity of interpretative assumptions as an alternative to relativism or foundationalism ultimately results in its own form of relativism.]

There are four interrelated philosophical problems about the interpretation of literature. While I speak here of literature, these problems can be extended to any interpretive procedures concerned with human action or the products of human agency: the interpretation of all art, of all texts, of individual behavior, of history, etc. One problem concerns the correctness of interpretations. Are interpretations correct or incorrect (true or false), or neither correct nor incorrect? A second problem concerns the standard of acceptability. If there are correct interpretations, is this correctness what we are chiefly concerned with in deciding whether an interpretation is acceptable? A third problem concerns the number of acceptable interpretations. Are there many acceptable interpretations of a given work or just one? A final problem concerns the relative importance of the writer’s and the interpreter’s point of view. Do acceptable interpretations of a work recover the writer’s intention? Does the standard of acceptability vary with interpreters point of view?

In a series of essays written between 1970 and 1985, Stanley Fish developed a comprehensive theory of interpretation which answers all these questions. According to Fish, interpretations are correct or incorrect but only relative to a set of interpretive assumptions. Fish is mainly concerned with the question: What makes an interpretation correct? This suggests he regards this as the most important standard of acceptability. However, nothing Fish says strictly commits him to this. Fish certainly believes that there are many acceptable interpretations of a text since he believes that there are many sets of assumptions from which it can be interpreted. Finally, in theory at least, the interpreter’s point of view is very important. Not only is an interpretation true only relative to an interpreter’s assumptions, but it is those assumptions which, according to Fish, create the object of, and evidence for, an interpretation.

Fish’s work has attracted much criticism but this has not produced any resolution of the issues separating Fish and his critics. There are two reasons for this. First, criticism has been direct. It has attempted to directly refute Fish’s key theses. This is very difficult to do conclusively. Second, Fish and his critics tend to disagree about the meaning and consequences of his theses.

My own objectives are more modest, therefore more promising. First, I approach Fish’s theory with a bit more sympathy than most of his critics. I will claim that his theory is based on a genuine insight. Second, I will argue that all of Fish’s controversial theses have one sense in which they are relatively innocuous and another in which they are quite radical. It is Fish’s equivocation between these senses that explains the confusion over the meaning and consequences of these theses. Third, instead of attacking the genuinely radical theses directly, I argue that they do not follow from their premises. It is much easier to establish that A does not follow from B than that A is false. However given the highly controversial nature of these theses, it is almost as damaging to establish the former.

I. THE DENIAL OF SENTENCE MEANING

Fish’s literary theory rests on an insight which is important to understand and, at least sometimes, to act on. One of the clearest statements is expressed in a sentence half quoted from J. L. Austin: “‘What we have to study is not the sentence’ in its pure unattached form but the issuing of an utterance in a situation by a human being.”¹

If a sentence in its “pure unattached form” is regarded not as a concatenation of marks but as a set of marks associated by linguistic rules with a finite set of distinct, “objective” meanings, then Fish believes there is no such things as the sentence in its pure unattached form.

This claim, if true, could provide the foundation for Fish’s insight. The claim that there is no such thing as the sentence in its pure unattached form implies that we cannot study the sentence in its pure unattached form. This makes plausible the claim that what we have to study are utterances (though it’s possible to think of other alternatives).

However, Fish does not start with a denial of the existence of sentences (or sentence meanings). Typically, the argument starts with the fact that utterance meaning varies from situation to situation and concludes that there is no such thing as “objective,” context independent sentence meaning.

Thus, in “Is There a Text in This Class,” Fish argues from the fact that utterances of the title question have different meanings in different situations to the conclusion that the question has no literal meaning other than the obvious meaning an utterance of the question has in a given situation. In the essay, “Normal Circumstances … and Other Special Cases,” Fish goes through the same argument for a series of expression: “private members only,” “this stuff is light enough to carry,” “Let’s go to the movies tonight,” and “I have to eat popcorn, tonight.”

The argument is invalid. Even if the premise is granted, the conclusion can be consistently denied. There are philosophers of language who would agree with everything Fish says about utterance meaning but who would claim that there are sentences or sentence meanings in the same sense that Fish denies there are. They can do this because they believe that utterance meaning is a function of sentence meaning plus various pragmatic considerations.

Can the distinction between semantic (sentence meaning) and pragmatic considerations be maintained? It is far from certain that it can. However, there are reasons for thinking the distinction plausible. Thus it seems possible to distinguish cases where utterance meaning varies for semantic reasons from cases where it varies for pragmatic reasons. Consider the sentence “The air is crisp.” E. D. Hirsch put forward this sentence as possessing a clear “valid meaning” accessible to all speakers of English: a “rough meteorological description predicting a certain quality in the local atmosphere” (p. 309). However, this is not the only meaning this sentence can have. Contrary to Fish, one does not have to be in, or imagine, a new context to realize this. One can predict it simply by looking up dictionary entries for “air” and “crisp,” which will tell you that “air” can refer to a piece of music which can also be crisp or crisply played. The dictionary entries suggest that the sentence itself has different meanings (is ambiguous), not merely that different utterances of the sentence will have different meanings.

On the other hand, the fact that an utterance of “I have to eat popcorn tonight” can “mean” “Sorry, I have to work” cannot be predicted from nor explained by the semantics of the sentence alone. No knowledge of the meanings of the words of the sentence and its syntax will enable one to predict this meaning (unlike the musical meaning of “The air is crisp”). Nor will the fact that an utterance of this sentence means “Sorry, I have to work,” imply that the sentence is ambiguous. The explanation and prediction of the meaning of the utterance can only be accomplished by introducing the pragmatics of the situation in which the sentence is uttered: the fact that the speaker is a taster in a popcorn factory and someone has just asked him if he wants to go to the movies.

Fish never acknowledges the existence of such considerations. But he does make remarks that suggest how he might reply to them. Fish claims it makes no sense to talk about ambiguity (as I did with “The air is crisp”) unless there is a “kind of language to which ambiguous language could be opposed” (p. 281). Fish denies that there is. However, it is not true that there has to be unambiguous sentences for it to make sense to speak of ambiguous ones. To say a sentence is ambiguous is to say it has more than one meaning. It is certainly conceivable that all sentences have this property. Also, Fish has never demonstrated that all sentences of English are ambiguous. By his own principles, he could not do so since, by his principles, he could only show that different utterances of a sentence have different meanings. That claim is compatible with the existence of unambiguous sentences, and a distinction between the semantic and the pragmatic.

Secondly, Fish might say that the semantic/pragmatic distinction leads to a distortion of the facts. On my account, it is plausible to say that one arrives at the utterance meaning of “I have to eat popcorn tonight” by inference. From the semantics of the sentence combined with facts or beliefs about the speaker one infers the utterance meaning. But Fish would say the meaning is immediately seen. No inference is involved. However, Fish is making this claim simply on the basis of phenomenology of the recognition of meaning. The hearer may not be aware of making an inference, but there is reason to think that there are many inferences we make that we are not aware of. Phenomenological considerations are bad reasons for denying inferences.

Fish’s insight is compatible with the existence of sentence meaning. Though Fish emphasizes the conventional nature of interpretation, he ignores the existence of linguistic conventions that cut across special situations and institutions (e.g., that “popcorn” can and commonly does refer to popcorn, that “text” can and commonly does refer to pieces of writing).² Until Fish acknowledges the considerations in favor of such conventions, he will be unable to give a very strong argument against them.

II. RELATIVISM AND INTERPRETIVE ASSUMPTIONS

Fish writes: “The shared basis of agreement sought by Abrams and others is never not already found although it is not always the same one” (p. 318). This passage can be read in at least two ways one of which suggests a “debilitating relativism,” the other does not.

There is nothing debilitating about the view that utterances of the same sentences in different situations have different meanings. In a way this is a relativistic view being the view that utterance meaning depends on various pragmatic considerations as well as sentence meaning (if there is such a thing). But this relativism is not the type anyone fears. It is compatible, for example, with there being a uniquely correct interpretation for each utterance. It is compatible with denying that there are many, sometimes incompatible, equally legitimate interpretation of an utterance that result from different, equally legitimate interpretive assumptions about the situation. It is compatible with claiming that the situation determines a single correct set of interpretive assumptions.

The other interpretation of Fish’s remark is that, for utterances, there is no single, stable basis of agreement that determines their meaning. Their meaning is relative to interpretive assumptions, the correctness of which cannot be adjudicated. This is the sort of relativism that Abrams and others fear.

In many places, Fish argues for a position that corresponds to the first, more innocuous interpretation of his remark. The student who utters “is there a text in this class” was asking a quite definite question. The teacher she addressed it to at first misunderstood its meaning, and then understood it. When Mr. Newlin raises his hand in Fish’s classroom, his gesture means that he wants to ask a question, not that he wants to go to the bathroom or that he is pointing to the sky, though in other situations the same gesture might have those meanings. Fish acknowledges, indeed exaggerates the determinateness of utterance meaning. “Listeners always know what speech act is being performed … because in any set of circumstances the illocutionary force a sentence may have will already have been determined” (p. 292. See also pp. 277 and 307).

It is surprising, therefore, that he adopts a position that corresponds to the second, less innocuous interpretation of his remark. He adopts the relativism that Abrams and others fear. Fish tries to defuse this relativism by arguing that it is a theoretical position that has no practical consequences and that it is severely constrained by the authority of interpretive communities. I return to these arguments later. For the time being, I am concerned with how Fish moves from his insight to theoretical relativism.³

In the essay “Is There a Text in This Class,” the move is made not as the result of an argument but by a subtle shift in the reference of the word “situation.” Consider the following: “the possibility of a context, or institution specific norm surely rules out the possibility of a norm whose validity is recognized by everyone no matter what his situation” (p. 319). Fish originally supported such a non-context specific norm (though the mention of context, institution, or situation in the statement of the norm may seem paradoxical), namely, that the meaning of an utterance is a function of the situation (context, institution) in which it is made. Anyone, whatever their situation, can try to follow this norm. What Fish is now suggesting is not that one interprets an utterance within its situation rather than in its pure unattached form, but that whenever one interprets one is doing so from one’s own situation. True or false, this is a completely new thought.

Fish makes a similar shift when he talks about interpretive assumptions. Sometimes Fish talks as if the interpretive assumptions are imposed by the institution or situation within which the utterance is delivered as if it were an objective feature of the overall context of utterance. It is not simply an assumption we impose on Mr. Newlin’s hand raising that it signifies the intention to ask a question. It is a convention of classroom conduct in institutions of higher learning (and that is how Fish describes it).

On the other hand, whenever Fish talks about interpretive assumptions in connection with literature, he shifts from claiming that the assumptions come from the situation of utterance (the writer’s situation) to claiming that the assumptions come from the critic’s situation.

Has Fish simply made a mistake, conflating two quite different views without being aware of doing so? The evidence of “Is There a Text in This Class” suggests that he is prone to such a conflation. However, since the shift is most commonly made in connection with literary interpretation, it is worth asking whether there is a difference between Mr. Newlin’s classmates interpreting his gesture and a critic interpreting a literary work.

One difference is that Mr. Newlin and his classmates are part of the same situation so it might be thought that the interpretive situation and the situation of utterance are identical. On the other hand, the situation of poets writing their poems is typically not the same as the critic interpreting the poem. When the situation of utterance and the situation of reception split apart, perhaps interpreters have no choice but to use interpretive assumptions from their situation. The literary critic is always such an interpreter. However, there is no single set of assumptions that constrain all critics at a given time. Furthermore, sets of assumptions change with time. Hence, the truth of relativism.

Fish never explicitly distinguishes between interpretive situations in this way. At best, such a distinction is implicit in his practice (as a theorist rather than a critic). One explanation of the difference might be a view about the accessibility of intention. When Mr. Newlin raises his hand to seek permission to speak, his intention to do this is a product of the same institutional conventions that produce the interpretive assumptions of his classmates. It is easy to imagine a situation where this ceases to be the case. Mr. Newlin might be an extraordinarily precocious eleven year old university freshman who raised his hand to seek permission to go the bathroom. The interpretation of this situation would differ from the original one in two ways. First, Mr. Newlin’s intention would no longer be transparent. His classmates would probably mistake his intention. This, however, does not mean that his intention is not accessible. Mr. Newlin could clear up matters quite easily. Second, even after Mr. Newlin’s intention was known, there could be a controversy about what Mr. Newlin’s hand raising meant. One side would maintain that the hand raising still had the conventional meaning of requesting permission to speak. This position would distinguish utterer’s meaning from utterance meaning and would claim that Mr. Newlin was requesting permission to speak even though he did not intend to. The other side would claim that Mr. Newlin’s hand raising meant that he wanted to go to the bathroom. This position would claim that Mr. Newlin’s intention determines the meaning of his gesture. Hence it determines what he does.

When we turn to the interpretation of such things as literature or law, not only are there subcommunities endorsing different interpretive assumptions (such as the ones just mentioned for interpreting Mr. Newlin’s gesture), but Fish appears to believe that intentions themselves are not accessible. Thus, about the legal reasoning in Riggs vs. Palmer, Fish comments:

This would seem to suggest that one need only recover the maker’s intention to arrive at the correct literal reading; but the documents … that would give us that intention are no more available to a literal reading (are no more uninterpreted) than the literal reading it would yield. However one specifies what is in the statute—whether by some theory of strict constructionism or by some construction of the original intention—that specification will have the same status as the specification of what is in Samson Agonistes or of what PRIVATE MEMBERS ONLY means. It can always be made, but as the situation and the purposes which inform them change, it will have to be made again.

(p. 281)

The passage seems to give an argument for the inaccessibility of intention. However it conflates two quite different points without arguing for either. The first point has to do with the accessibility of intention. Fish points out that evidence for the intention of the maker of a law (or of a literary work) will come in the form of further pieces of writing which will also need interpretation. Beginning with “however,” the passage shifts to the second point. The principles by which one interprets a law (or a literary work) will vary with the situation and purpose of the interpreter.

Fish is right about part of the second point. People can interpret with different purposes. One person might be interested in the conventional meaning of the eleven year old Mr. Newlin’s gesture; another with Mr. Newlin’s meaning (intention). But in referring to the interpreter’s situation, Fish makes the shift from the utterer’s situation that we have been trying to justify. The justification might be thought to be found in the first point. But the first point does not show that the intentions of legislators or writers are inaccessible. That the evidence is also in need of interpretation does not show that the evidence never clearly supports a particular hypothesis. The correct interpretation may be obvious. When the eleven year old Mr. Newlin explains, “in raising my hand, I intended to request permission to go to the bathroom, not permission to speak,” his words need interpretation just as much as the original gesture. Furthermore, they could be interpreted in many ways. Nevertheless, they could also make utterly obvious what Mr. Newlin meant by his gesture. So could evidence about legislators and writers.

III. INTERPRETATIONS CREATE THEIR OWN EVIDENCE

What ultimately justifies the relativistic move from utterers situation to interpreters situation is a view Fish holds about evidence. According to Fish, evidence “is always a function of what it is evidence for and is never independently available” (p. 272). Fish believes this shows that interpretation-independent facts are never available. Hence, for Fish, the shift from the utterer’s situation (never accessible in itself) to the interpreter’s situation is necessary.

There is a relatively innocuous sense in which there are no interpretation-independent facts, namely, that the discovery of any fact requires interpretation of data. But Fish means to claim something more radical. He claims that facts are created by interpretations. Similarly, the claim that evidence is a function of interpretation has a more and a less radical reading. The less radical reading is that the interpretation dictates the kind of thing that will count as evidence for it. For example, an interpretation that claims to recover an author’s intention requires evidence of intention. Fish sometimes speaks as if this is his view: “interpretation determines what will count as evidence for it” (p. 272, emphasis added) But Fish’s argument relies on the more radical reading, namely, that interpretations create their own evidence. This is brought out by Fish’s claim that evidence for an interpretation is not available independently of an interpretation.

I think I know why Fish is tempted to make this more radical claim. Consider the occurrence of the word “forest” in Blake’s “The Tyger.” Rival incompatible interpretations cite the occurrence of this word as evidence for each interpretation. How can the same word, the same bit of evidence, be evidence for incompatible readings? Fish’s answer is that the evidence only appears to be the same. The way we perceive the word, the content we assign to it, is a function of the interpretation. Thus, Kathleen Raine interprets the poem as inspired by cabalistic writing. Interpreted that way, “forest” takes on a symbolic significance of a fallen world. E. D. Hirsch interprets the poem as celebrating the “holiness of tygerness.” Now “forest” will be perceived in an entirely different way. To Hirsch, it suggests “tall; straight forms.” Similarly according to Fish, if one hypothesizes a typological reading of Milton’s Samson Agonistes, the absence of explicit reference to Christ is evidence of implicit, typological reference. “Once [a typological] characterization of Milton’s intention has been specified, the text will immediately assume the shape Madsen proceeds to describe” (p. 233).

It is possible that Fish’s account of evidence describes some attempts by critics to justify their interpretations. Such attempts are unsound. No doubt, once one is aware of an interpretation of a text, it is often easy to read it in terms of that interpretation. But that is not evidence in favor of the interpretation. In the case of Samson Agonistes, we can read the work according to Madsen’s interpretation because the absence of explicit reference to Christ is compatible with a typological interpretation. However, this absence of reference to Christ is not positive evidence for the interpretation either before or after the interpretation has been given. Evidence for Madsen’s interpretation would give us positive reason to believe that Madsen’s characterization of Milton’s intention is the true one. Such evidence is not a function of what it is evidence for, and, if it is available at all, it is available independently of the interpretation.

The same is true of the debate between Raine and Hirsch. The fact that one can find a significance in individual words that accords with an interpretation shows at best that the occurrence of the word is compatible with the interpretation. Since both Raine and Hirsch aim at recovering Blake’s intention, evidence for these interpretations would have to show that the significance an interpretation gives to the words of the poem is the significance Blake intended. The fragment of Hirsch’s argument that Fish reports fails to do this. Raines makes two evidential claims which, if true, do support her interpretation: that “The Tyger” is inspired by certain cabalistic writings; that Blake always uses “forest” to refer to a fallen world. These are the right sort of claims to support her interpretation. If true, they would go a long way toward establishing Blake’s intention. They can be debated independently of any commitment to, or even knowledge of, Raines’s interpretation of “The Tyger.”

In the face of similar criticism by John Reichert, Fish replies that “standards of right and wrong, [of what counts as evidence], do not exist apart from assumptions but follow from them” (p. 296). It is important to see that this is a retreat from Fish’s more to Fish’s less radical claim about evidence. This is not the claim that interpretations create their own evidence. It is not the claim that evidence is unavailable independently of an interpretation. It is simply the claim, which I would accept without reservation, that looking for evidence for an interpretation is risky. Not only might the evidence not turn up, or go against the interpretation, but, in searching for evidence, one will make assumptions about when one has it or what constitutes it. Some of these assumptions may be questionable or even false. However one’s assumptions may be true! This is a possibility that Fish does nothing to rule out though he does not take it seriously because he is overly impressed with the fact that one never knows whether all one’s assumptions are true. But this lack of knowledge is simply part of the riskiness of the enterprise.⁴

IV. DEFUSING RELATIVISM

I will finally consider Fish’s attempt to take the sting out of relativism. This attempt is based on two claims: A. that it is a theoretical position that has no practical consequences; B. that relativism is severely constrained by the authority of interpretive communities.

A. According to Fish’s theory critical activity constitutes its objects (p. 355). (This claim, which can only be based on his claim that we have no interpretation-independent access to the objects of interpretation, does not, in fact, follow from it. But this is one move of Fish’s that I will not dwell on.) A consequence of this is that an interpretation cannot be more or less adequate to an independently existing work. There is no such thing as progress in criticism if progress is defined as the production of increasingly accurate interpretations of a given work (p. 366).

An interpretation can be more or less adequate to the assumptions or rules on which it is based. Furthermore, these assumptions can be criticized in the light of other assumptions (p. 360). However, since such criticisms have their own set of underlying assumptions, such critiques do not bring about progress, only change (p. 361).

This is Fish’s theory, his relativism. Why does Fish think this theory has no consequences for critical practice? In particular, why doesn’t this theory have the consequence that we ought to become skeptical of the truth of our interpretations because the best interpretations arising from one set of assumptions are no better justified than incompatible interpretations arising from other assumptions?

In Is There a Text in This Class, Fish gave a very simple answer to this question: since we must believe something, it is impossible to sufficiently distance ourselves from our beliefs to become skeptical of them (pp. 360–361). Since we often draw conclusions from beliefs, propositions held to be true, our drawing of conclusions will feel like progress.

In a later essay, “Consequences,” Fish gives a more complicated answer.⁵ It isn’t just that we forget theory whenever we engage in practice (though Fish continues to maintain this).⁶ Fish now claims that his antifoundational theory does not have any skeptical consequences. It is simply a theory about how we acquire beliefs. We acquire them against a background of beliefs and assumptions. The claim that beliefs so acquired are unjustified, or arbitrary, or are not knowledge is not a part of antifoundationalism.

Neither of these answers are satisfactory. The first is unsatisfactory because it does not imply the conclusion Fish needs. It may be true that we always must believe something but we can still be skeptical about our interpretations. Anyone who believes Fish’s theory, and unlike Fish, believes that it has skeptical consequences, can stop believing his interpretations and still believe something (Fish’s theory).

Fish assumes that when one engages in a “practical” activity like criticism, it is one’s theoretical beliefs that will be given up in favor of beliefs one acquires in the course of immersing oneself in actual practice. However, this is an unjustified assumption. Some will do this; some won’t.⁷

Fish’s more complicated response is also unsatisfactory. It is unsatisfactory because it is disingenuous. Certainly, Fish says something about how we acquire beliefs. However, as Fish admits at the end of “Consequences,” what he says severely limits the sort of justification of which interpretations are capable. The question is: should these limitations, if we accept them, make us skeptical of our interpretations? Yes, they should. Suppose I believe that Satan is the hero of Paradise Lost as a result of correctly applying the assumptions of my critical community. Suppose another critic (whether or not she is my contemporary does not matter) believes that Satan is not the hero of Paradise Lost as a result of correctly applying the assumptions of her critical community. On Fish’s view, each belief is as well justified as an interpretive belief can be. Now suppose I add this theoretical belief to my stock of beliefs. I now believe: 1. Satan is the hero of Paradise Lost; 2. My belief that Satan is the hero of Paradise Lost is a consequence of a set of assumptions that are ultimately no better justified than another set of assumptions that imply that Satan is not the hero of Paradise Lost. If I believe 2, then, I am bound to admit that I am no better justified in believing 1 than someone else is in believing that Satan is not the hero of Paradise Lost on the basis of no more, but no less, arbitrary assumptions. Furthermore, I cannot become more justified in believing 1 than I already am. If this isn’t a good reason to suspend my belief in 1, it is hard to know what is.⁸

Even if I am wrong that Fish’s theory implies skepticism, it still has consequences. After denying this for twenty pages in “Consequences,” Fish finally admits this. Someone who accepts his theory, Fish says, would likely stop looking for certain kinds of justification. This is precisely the sort of guidance of practice by theory that Fish tried to deny.

B. Fish likes to point out that his theory does not have the consequence of extreme subjectivism. It is not the individual critic who invents standards that are no better or worse than the next critic’s. There are shared standards, common strategies of reading. These are provided by sets of institutional assumptions (p. 367).

There are two reasons why I would not find the existence of institutional assumptions comforting, were I to accept Fish’s theory.

First, I worry that my assumptions are arbitrary: not justified or justifiable, nor, better justified than the assumptions of critics I disagree with. I find no comfort in the fact that my arbitrary assumptions come from an institution rather than from myself.

Second, while Fish never doubts that we (critics of like mind) do share the same institutional assumptions, his theory should make him doubt it. Even if we grant that there is an institution issuing assumptions about how to interpret texts, we have no more direct access to these assumptions than to poems or utterance meanings. Individuals approach this institution with their own idiosyncrasies and views, with prior sets of assumptions transmitted by other institutions and practices. To grasp the assumptions of the institution of criticism, they interpret them in light of the beliefs, assumptions, and special interests one already has. It is quite likely that these are different for each person. Each person’s interpretation of the assumptions of criticism are likely to be different.

Fish has not escaped extreme subjectivism after all.

V. CONCLUDING REMARKS

To the charge that his theory implies a skeptical view about the justification of interpretation, Fish sometimes replies: the charge presupposes foundationalism, i.e., the view that justification requires appeal to a common foundation against which all beliefs can be tested. It is only foundationalists who think the theory leads to skepticism.

I have tried to avoid making foundationalist assumptions. Instead, I draw skeptical conclusions from Fish’s theory on the basis of the argument of section 4.

So far, what Fish, has given us are attacks on foundationalism (essentialism, realism) combined with antifoundational theories that leave me wondering what is so cognitively respectable about justification relative to a set of assumptions (conceived the way Fish conceives it). If that is all we can get, why shouldn’t we be skeptics? Perhaps an explanation why we shouldn’t be can be given, but it has yet to be provided.

Notes

1. Stanley Fish, Is There a Text in This Class? (Harvard University Press, 1980), p. 231. References within the text are to this book.

2. This is James Carney’s chief criticism of Fish in “Literary Relativism,” Journal of Aesthetic Education 21 (1987), p. 13.

3. Gerald Graff notes Fish’s tendency to shift between claiming that meaning is determinate in context and claiming that it is relative to interpretive assumptions in “Interpretation of Tlon: a Response to Stanley Fish,” New Literary History 17 (1985), pp. 113–114.

4. For further criticism of Fish’s views about evidence, see Annette Barnes, On Interpretation (Oxford: Basil Blackwell, 1988), pp. 86–95.

5. Stanley Fish, “Consequences,” Critical Inquiry 11 (1985): 433–458.

6. Ibid., p. 450.

7. In “Consequences,” Fish sharply distinguishes between beliefs and theories. Beliefs have consequences, theories don’t. But beliefs can be theoretical—being about assumptions underlying certain practices, about the relation of evidence to interpretation—about, in short, the stuff of Fish’s theory. If Fish likes, we can talk in the following way: we can say that beliefs, including theoretical beliefs, have consequences, but theories, when unincorporated into systems of belief, don’t. That reduces the issue to less than a quibble.

8. Peter Kivy makes a different though related criticism of Fish’s claim that his theory has no practical consequences in “Fish’s Consequences,” British Journal of Aesthetics 29 (1989): 57–64. A difference between Kivy and me is that he regards Fish’s theory as giving a noncognitive model of interpretation, while I try to preserve what I take to be Fish’s intention to give a cognitive model. However, I agree with Kivy that there is much reason to think that Fish cannot give such a model consistently.