The claim is so bold and so extraordinary as to invite a parodic response: presumably, any objection depends on mistaking the meaning of the argument, since the authors intend it to be irrefutable. (See Goodman's proof that p.) So, criticism is a treacherous business.
How do Knapp and Michaels argue for the trivial truth of what would seem to be a controversial claim? They appeal to a now-famous example: the "wave poem", in which the random lapping of water on a beach inscribes, by astonishing coincidence, the characters of Wordsworth's lyric, "A slumber did my spirit steal". As they point out, the absence of intention seems to make the marks meaningless: "what you thought to be poetry turns out not to be poetry at all [...] because it isn't language."
Apart from its obvious defect of logic – it could show, at most, that the author's intention is necessary, not sufficient, for meaning – the use of this example is puzzling. No claim about the relation between linguistic meaning and intention could settle the dispute about "intentionalism", since it arises just as well for non-linguistic art. Is the meaning of painting a matter of the painter's intention? Or the meaning of a symphony just what the composer intended it to be? In any case, the meaning of the sentences of a novel, however it relates to the intentions of the author, underdetermines what we call "the meaning of the text": it leaves the most interesting thematic and psychological questions about the fiction wide open.
Critics of Knapp and Michaels sometimes stop at this point. (Philosophers have been especially harsh.) But we can see how absurd it is to interpret them as making a claim about the meaning of words by repeating a question pressed, albeit with some embarrassment, by W. J. T. Mitchell:
What happens [...] if we (whether theorists or plain practitioners) intend these terms ["meaning" and "intention"] to mean something different?Taken as a theory about linguistic meaning, the Knapp-Michaels doctrine is latently paradoxical. They must be up to something else.
The problem is that distinguishing linguistic and artistic meaning doesn't seem to make their claim more plausible. Intention is not sufficient to generate the meaning of a sculpture, or a sonnet, or a song. Art is not magic: one can fail, and failure is precisely a discrepancy between the actual meaning of your creation, and what you wanted it to be. This shows, in turn, that art can mean something unintended: intention is not necessary, either. ("I meant to write a biting satire of materialism, but ended up with an affectionate farce." "I meant to depict a glorious hero, but created a comic fool.") There is such a thing a biographical speculation about an artist, but it is not what we principally mean when we talk about "interpretation". The "platitude" is not a platitude, after all.
When I reflect on what I am trying to do when I engage in interpretation – and here I make no claim to generality – it seems to be the opposite of "intentionalism": casting a veil of ignorance over the artist's actual intentions, but not her literary and historical conditions, I ask what hypothetical intentions would make most sense of the work. "Making most sense", here, goes beyond the ordinary principle of charity: the target is not the most plausible intentional story, but the one that attributes the most intricate intentionality. The more significance it finds, the better the interpretation; and the best gives meaning to everything: it leaves nothing inert.
Put this way, it seems a pretty queer activity. Why abstract from intention but nothing else? Why give in to this hunger for meaning, which flatly ignores the limits of the artist? It has its satisfactions, but so do many other games one can play with a text, a composition, or an installation. It doesn't seem to correspond to anything. (Thus the pull of critical pluralism.)
Here we find the real force of the wave poem example. Just as no string of marks has linguistic meaning unless it was the product of intention, so, I believe, nothing counts as a work of art unless it was intended to be one. But if interpretation is what I have said it is (at least for me), why should it be attentive to this one intention, among others? Why not play the same game with the wave poem, with a cauliflower that happens to be shaped like a human head, or with nature itself?