SOURCE: “Fish's Consequences,” in British Journal of Aesthetics, Vol. 29, No. 1, Winter, 1989, pp. 57-64.

[In the following essay, Kivy examines various contradictions and logical flaws of Fish's theoretical perspective, particularly those involving distinctions between demonstration and persuasion models of criticism and their respective implications for literary interpretation.]

I want to examine, in this essay, the consequences for the practice of criticism there would necessarily be if what Stanley Fish says about that practice were indeed the truth. I am not interested, here, in truth, but in consequences. Fish believes that the implications of his view for the practice of criticism would be business as usual. I will argue that this cannot be correct, and that Fish’s position, in this regard, is deeply confused.

To put the matter baldly, there are, in Fish’s view, two opposing models of the literary critic’s activity: what Fish calls the demonstration model, and his own, which he calls the model of persuasion. As he describes them:

In a demonstration model our task is to be adequate to the description of objects that exist independently of our activities; we may fail or we may succeed, but whatever we do the objects of our attention will retain their ontological separateness and still be what they were before we approached them. In a model of persuasion, however, our activities are directly constitutive of these objects, and the terms in which they can be described, and of the standards by which they can be evaluated.¹

Now the persuasion model has what appear to be deeply troubling, even paradoxical consequences, as various critics of Fish have pointed out, and as Fish himself quite candidly admits. ‘Not only is such a view disturbing but it seems counterintuitive given the very real sense we all have, both as critics and teachers, of advancing toward a clearer sight of our object.’ For if we believe that we, as interpreters, construct the text rather than find it out, ‘if we really believe that a text has no determinate meaning, then how can we presume to judge our students’ approximations of it, and, for that matter, how can we presume to teach them anything at all?’²

But it is Fish’s view that these troubles are apparent only. Embracing his conception of criticism, ‘We have everything that we always had … texts, standards, norms, criteria of judgment, critical histories, and so on. We can convince others that they are wrong, argue that one interpretation is better than another, cite evidence in support of the interpretations we prefer’.³ How can this be?

The answer is that, on Fish’s view (although he does not say it in so many words), the evidence that is constituted by how critics behave, and by how we might, in various gedanken experiments, imagine them to go on behaving, is just as consistent with the persuasion model of criticism as it is with the demonstration model. This requires some spelling out.

On the demonstration model, texts possess meanings and critics try to discover them. If a critic thinks he knows what the meaning of a text is, he will, quite naturally, try to convince others. He will do so by adducing arguments, as he made his discovery, or, perhaps, verified it, by marshalling evidence. If he has thought about methodological matters at all, and is not just ‘doin’ what comes naturally’, then he believes that the procedures he follows are those suited to gaining his end, which is discovery of meaning; and he will insist that others abide by them as well, if they share the enterprise. Doubtless he will also believe, if he has thought about it, that these procedures are inviolable, although he may or may not have opinions about what their epistemic grounds are.

Now Fish does not believe any of this. On his view, the critic is not discovering meaning in the text but constructing meaning on the text, or, in effect, constructing the text. But he does so under the institutional constraints of the accepted procedures. ‘No longer is the critic the humble servant of texts whose glories exist independently of anything he might do; it is what he does, within the constraints embedded in the literary institution, that brings texts into being and makes them available for analysis and appreciation.’⁴ But these procedural constraints are not, as the demonstration model would have it, inviolable rules of reason: ‘all arguments are made within assumptions and presuppositions that are themselves subject to challenge and change’.⁵ Nevertheless, because the critic works within this system, he or she behaves, critically, just as if the demonstration model were true. That is to say, what reveals itself to the meta-critic, from above, to be a system of institutional constraints on text construction, to a degree ephemeral and subject to change, is seen from below to be the inescapable, inviolable method for finding out the truth of meaning in the text; and, by consequence, what reveals itself from above to be construction presents itself from below as discovery.

Thus far the representation seems plausible: but only so long as Fish refrains from making reference to his own critical behaviour, or that of a critic who has been convinced of Fish’s view. It is easy, perhaps, to imagine that there would be no difference in critical behaviour between a possible world in which the demonstration model was correct, and critics believed it, and a possible world in which the persuasion model was correct, but critics believed the demonstration model was. What is difficult to imagine is how behaviour would be invariant with the belief that the persuasion model is correct as opposed to the belief that the demonstration model is. ‘How’, Fish himself asks, ‘can someone who believes that the force and persuasiveness of an interpretation depends on institutional circumstances (rather than any normative standard of correctness), and that those circumstances are continually changing, argue with conviction for the interpretation he happens to hold at the present time?’ Here is his reply:

The answer is that the general or metacritical belief (to which I am trying to persuade you in these lectures) does not in any way affect the belief or set of beliefs (about the nature of literature, the proper mode of critical inquiry, the forms of literary evidence, and so on) which yields the interpretation that now seems to you (or me) to be inescapable and obvious. I may, in some sense, know that my present reading of Paradise Lost follows from assumptions that I did not always hold and may not hold in a year or so, but that ‘knowledge’ does not prevent me from knowing that my present reading of Paradise Lost is the correct one. This is because the reservation with which I might offer my reading amounts to no more than saying ‘of course I may someday change my mind’, but the fact that my mind may someday be other than it now is does not alter the fact that it is what it now is; no more than the qualifying ‘as far as I know’ with which someone might preface an assertion means that he doesn’t know what he knows—he may someday know something different, and when he does, that something will then be as far as he knows and he will know it no less firmly than what he knows today.⁶

The passage bristles with epistemological difficulties. Of this Fish cannot be totally unaware; for both ‘know’ and ‘knowledge’ are introduced with some diffidence. The critic is said to know only ‘in some sense’. (What sense?) And ‘knowledge’ is introduced in quotation marks to indicate, I presume, a nonstandard sense of the term. (Again: What sense?) Perhaps in response to these difficulties, Fish moves, a bit later on, from knowledge to belief, in what I take to be another attempt to express the same thought: ‘If one believes what one believes, then one believes what one believes is true, and conversely, one believes that what one doesn’t believe is not true, even if that is something one believed a moment ago’.⁷

I shall take belief rather than knowledge, therefore, to be the operative concept in both of these passages. And I shall introduce, for purposes of discussion, two critics whom I will call, not very imaginatively, Critic (1) and Critic (2), the former a believer in the demonstration model, the latter a believer in the persuasion one; by which I mean, Critic (1) believes that the demonstration model correctly characterizes the critical process, while Critic (2) believes that the persuasion model does. In other words, Critic (1) believes that meanings are discovered in texts, true or correct interpretations being such discoveries; and Critic (2) believes that an interpretation is true or correct in that it conforms to one of the currently accepted ‘ways of producing the text’, which are neither ‘monolithic’ nor ‘stable’, but always subject to change.⁸

Now on Fish’s view, critical behaviour is invariant under the belief that the persuasion model is true or that the demonstration model is. Two questions seem in order. The first is: would critical behaviour be invariant? The second: should it be? That is to say, we want to know not only whether people would go on doing what they are now doing if they became convinced that the demonstration model is incorrect, the persuasion model correct. We want also to know whether they would be rationally justified in their behaviour. I shall argue that it is highly unlikely a critic would behave in the same way if he or she were persuaded to Fish’s view, as under the assumption of the demonstration model; and, further, that if one were to behave in a way appropriate to the demonstration model while believing the persuasion to be the correct one, he or she would either be behaving irrationally, or in bad faith.

Imagine, now, the critic returning to the cave, after conversion from the demonstration to the persuasion model. If his new critical beliefs have any real content at all, and are not just a re-description of the demonstration model, it seems to me to be very unlikely that his critical activities should fail to at least begin to conform with the new belief system. It would certainly not be my expectation that someone who previously believed that criticism was discovery, and now believes that it is creation, should display the same pattern of critical behaviour as before. How might it be expected to change? My contention is that it would be bound to become more liberal, more changeful, less stable, because, of course, although one would not cease to have critical beliefs—i.e., that work W means M—the content of those beliefs would be radically different under the assumption of the persuasion model, as opposed to the demonstration model, and, hence, the process of one’s belief formation markedly different as well. What would be rational behaviour under one assumption would be irrational under the other; and vice versa; and although there is never expected to be a complete conformity, rationally speaking, between belief and action, there is no reason to believe there would be no conformity here at all. And even if there were none, I take it that it is Fish’s wish, and we have every right to demand it of him, to convince us not only that critical behaviour will remain the same on the assumption of the persuasion model, but that it would be rational for the critic to go on as before; and if the former seems to me to be unlikely, the latter seems patently false.

Imagine now that Critic (2) believes the following: ‘Mills of Satan’ is an illusion to the churches of Blake’s day. What is the content of his belief? It is that (among other things), as the rules and procedures of the institution of literary criticism are now constituted, this is a possible construction one can make of Blake’s phrase: there is a critical mechanism in place for producing it, and no rule forbidding it. He also believes that these rules and procedures change; and that in he future there may not be a critical mechanism in place for generating this construction, or there may be a rule making it inappropriate. And he believes, quite literally, that these rules and procedures are for the purpose of constructing texts, not discovering meanings in them. They are, in other words, not analogues to the rules and procedures that the analytic chemist, for example, adheres to for the purpose of finding out what is in an unknown substance. Rather, they are analogous to the rules of sixteenth-century counterpoint that forbid certain ways of constructing musical pieces, and countenance others.

This analogy suggests a rather important disanalogy between what would constitute rational (or irrational) behaviour for Critic (2), and what for Critic (1). Suppose that Critic (1) believes, with Critic (2), that ‘Mills of Satan’ alludes to churches. Imagine now it is proposed to both that the phrase refers to the factories of the Industrial Revolution. And imagine, further, that both prefer that interpretation over the one they now hold; that is to say, they would rather it be true than the other (although it is somewhat questionable whether Critic (2) is entitled to the concept of ‘truth’ here). Never mind why they prefer it: perhaps because it is more interesting, makes for a better poem, or makes their poetic hero more prophetic. The point I want to make, and develop, is that preference—‘mere’ preference—can never be a relevant consideration for Critic (1): indeed, his model of criticism, the demonstration model, has built into it a positive proscription against it. (‘Don’t let your desires cloud your disinterested quest for the truth, or you will follow the path of Lysenko, and others of his ilk.’) Whereas to suggest that preferring an interpretation cannot figure, perhaps even decisively, in Critic (2)’s interpretative decision would be like suggesting that preferring a certain musical structure over another cannot figure in a composer’s choice of that structure for his latest composition. And the force of saying to the critic, ‘This interpretation breaks a rule’, or ‘There is no approved procedure for generating that interpretation’ would be about as great, if the persuasion model were assumed, as the force of saying to Beethoven that he had violated the accepted rules of counterpoint in the Credo of the Missa Solemnis, or of saying to Schoenberg that there were no harmonic procedures, currently accepted, for generating the harmonic structure of the Kammersymphonie. For once one sees, as the persuasionist has, that the ‘rules’ and ‘procedures’ of criticism are not the universal rules and procedures for finding out the truth of some matter or other, but the changeable rules and procedures for making things—and imaginative, non-practical things at that, without such constraints as the physical world places on the makers of bombs and bridges—one will hardly take the rules and procedures as presently constituted, as overriding constraints on what one can or cannot make of a text, anymore than a Beethoven or a Schoenberg would the rules and procedures of musical composition. The critic, in being persuaded to the persuasion model, has joined the ranks of the creators, and is to the rules, if he wishes, as Walther to Beckmesser.

Nor can it be claimed that I force upon Fish an analogy completely alien to his views; for the analogy fairly jumps from his text. Here is what Fish says; I quoted it at greater length at the outset. In the persuasion model, ‘our activities are directly constitutive of these objects, and the terms in which they can be described, and of the standards by which they can be evaluated’.⁹ In even stronger language, ‘No longer is the critic the humble servant of texts whose glories exist independently of anything he might do…’; for he, in fact, ‘brings texts into being…’. And although Fish attempts to soften this blow somewhat by adding that the critic, in creating texts, works ‘within the constraints embedded in the literary institution’,¹⁰ it is clear that that cannot be wholly true, since he has also insisted, as we have just seen, that the activities of critics are directly constitutive of the terms in which texts can be described and the standards by which they can be evaluated—which I take it can only mean that the activities of critics are, in other words, directly constitutive of the critical constraints, just mentioned, that are embedded in the literary institution. But if that is the case, then it is difficult to see how it could be true that the critics always work, or would feel obliged always to work ‘within the constraints embedded in the literary institution’. Since it is their activity that makes those constraints, if they always worked within them, the constraints would never change, which is exactly contrary to what Fish is saying. It is by obeying the constraints, one presumes, that critics keep them in place; and it is by disobeying them, one must also presume, that they are dislodged and replaced by new ones, just as by disobeying the rules of counterpoint and harmony that are given them, composers constitute new ones by their compositional practice. Once the critic comes to believe that the rules and procedures of critics are constituted, both in the breach, as well as in the observance, by his or her critical behaviour, then I can see no reason why mere preference for an interpretation cannot be decisive in motivating the critic to break accepted rules and circumvent accepted practices in order to construct it. It sometimes was enough for Beethoven and Schoenberg.

Now Fish’s stratagem, as I take it, for palliating what must strike some readers as this excessive critical freedom, fully countenanced by his theory, is to explain it away as nothing more than the freedom that is allowed to anyone who states that he or she ‘merely believes’ (as opposed to ‘knows’ or ‘is certain’) that something is the case. Two things about ‘merely believing’ are relevant to what Fish is claiming in this regard. First, when one states one’s belief, one is implicitly allowing that one might be mistaken; that one might, in light of (unknown) future evidence, alter one’s belief. And, second, even though one makes that proviso by stating one ‘merely believes’ something, this does not alter the fact that one does now believe it, and, more important still for present purposes, one cannot cease to believe it as a freely willed mental action. What one believes, one believes, and cannot do otherwise, even though one knows that one might, in the future, not believe that thing. I am certain, now, that some of my present beliefs are false; but I cannot tell you which ones because, of course, if I knew, they would not be my beliefs. If I am convinced by whatever evidence I have, I cannot will not to believe what I am convinced of (as Spinoza long ago argued against Descartes).¹¹

It is this dual concept of belief that Fish wishes to apply to the persuasion model. He wants to argue that the critic who is convinced of the persuasion model and who acknowledges, therefore, that his interpretation of a given work may change with the change in the rules and procedures of the literary institution, is saying no more nor less than the person who says: ‘This is what I believe; but, of course, it is merely my belief, and I may be wrong: my belief may change, in the light of new evidence, or in the discrediting of old’.

But the analogy will not do; for in taking up the position that he has, the follower of the persuasion model has given up any right to the non-volitional character of his critical ‘beliefs’; for they are not beliefs about what is in the world at all: indeed, they are not ‘beliefs’ at all but ‘creations’; and what one creates is, unlike what one believes, subject to the will (within, of course, the limits of human capability). The ‘freedom of belief’ explanation (if I may so call it) would only function successfully in cases in which the critic is convinced of the demonstration model. Convinced of the persuasion model, however, the critic no longer believes that he is discovering something in the world—i.e., in the text—but believes that he is constructing something in the world—i.e., the text. His critical ‘statements’ cease to be, for him, statements of belief, and become creations of texts or meanings. And whereas beliefs about texts are not subject to the will, the construction of texts is.

Now, of course, how Critic (2) constructs texts is itself governed by beliefs; and these beliefs he cannot wilfully change. Critic (2) holds beliefs about (among other things) what the rules and procedures embedded in the literary institution are. These beliefs may be true or false, and are, like any other beliefs, subject to change, but not volitionally. However, one very strange consequence of the persuasion model is that even if Critic (2)’s beliefs are false (and why might they not be?), it would not matter a jot or a tittle, since if he believes that x, y, and z are the reigning rules and procedures, whether they are or not, they can enable him to generate his construction. And, as we have seen, even if there are no rules in place for generating the construction he wants, or even if there are, and he doesn’t believe it, he can generate the construction he wants, in any case, since, like the artist or composer, he is a creator, and like any creator, may rebel against what rules and constraints he does not acquiesce in. Indeed, that is at least part of the mechanism, on Fish’s account, by which changes in the rules and procedures of literary criticism are effected.

What must be abundantly clear, at this point, is that ‘business as usual’ cannot possibly be the result of the critical community’s giving up the demonstration model for the persuasion one. The only way to reach the conclusion that it would be, is by a kind of double-think, whereby one fudges the distinction between the critic who does, and the critic who does not believe that the persuasion model is true, and, in so doing, makes illicit use of epistemic characteristics of concepts such as ‘belief’, ‘knowledge’, and ‘truth’, that, as a convert to persuasion, one is no longer entitled to. For the picture each model—demonstration or persuasion—paints of the critical world is vastly different; and a person who went from the former to the latter view would have beliefs so altered as to alter his or her behaviour radically, or, failing that, be open to the charge of bad faith or irrationality.

It is, of course, an empirical question as to whether the whole critical community, or a significant part of it might, in a mass epidemic of bad faith or irrationality, go on behaving as if meanings were discovered in texts, all the while believing they were constructed on them, with all that implies (as I have tried to show) for critical practice. My own empirical intuitions, for what they are worth, are against the hypothesis that critical behaviour would remain unchanged in a wholesale conversion of the critical community from the demonstration model to the persuasion one. It seems to me utterly incredible. Indeed, just as there is evidence that the increase in belief in psychological determinism changes our attitudes and behaviour towards criminal and immoral acts, I perceive some evidence that critical practice has altered in ways to be expected in light of the growing influence of Fish’s view, and views like it, in the critical community.

But it now remains merely to end, as I began, with the caveat that I am interested in consequences, here, and not in truth. That is to say, none of the conclusions I have drawn in this essay is meant as counting against Fish’s persuasion model of critical discourse. If what I have claimed to be the implications of his view do indeed follow from it, and are found to be unpalatable implications, that no more constitutes a counter-argument than would the well-known claim that atheism leads to immorality be an argument against it and for the existence of God. If, however, those implications suggest not only the unpalatable but the unintelligible—if, that is to say, critical practice, thought, and discourse turned out, were Fish’s view correct, to be impossible for us to square with our ‘way of life’ in some deep sense of ‘impossible’—then, indeed, the persuasion model would be in trouble. But that must be the subject for another occasion.¹²

References

1. Stanley Fish, ‘Demonstration vs. Persuasion: Two Models of Critical Activity’, Is There a Text in This Class?: The Authority of Interpretive Communities (Cambridge, Mass., and London: Harvard U.P., 1980), p. 367.

2. Ibid., pp. 358–359.

3. Ibid., p. 367.

4. Ibid., p. 368.

5. Ibid.

6. Ibid., pp. 359–360.

7. Ibid., p. 361.

8. ‘What Makes an Interpretation Acceptable?’, Is There a Text in This Class?, pp. 342–343.

9. See Reference 1 above.

10. See Reference 4 above.

11. And see Fish, ‘Demonstration vs. Persuasion’ (p. 363): ‘she can’t will a belief in the Aspects model any more than she can will a disbelief in the arguments that persuaded her that it was unworkable’.

12. I am grateful to my colleague, Laurent Stern, for reading an earlier version of this paper, and for saving me from a number of errors.