Previous Next Up Comments
SEHR, volume 4, issue 1: Bridging the Gap
Updated 8 April 1995

inside-out, outside-in

Brian Cantwell Smith

Is it always this way, with master narratives? Or at least with attempts to defend them? Perhaps. Still, it is fascinating to encounter this plain a case.

What one witnesses, in this paper, is how a specific (and not immediately plausible) idea, in the hands of a well-known advocate, is lifted from working hypothesis into general conception through which everything is viewed-so general, in fact, that its specific formulation fades into invisibility, as if requiring no further mention. And yet, as always, it ironically turns out that what is on the one hand totalizing is on the other hand curiously impotent, unable to sustain itself. Crucially, in fact, totalizing myth is held in place by intuitions that run deeper and truer than, but also counter to, its own claims.

All is not lost, however, since what is bad for the account is ultimately good: true intuitions survive to underwrite another day.


Simon starts by embracing a radically internalist theory of meaning. The view is introduced by authority, not argument or rationale: this (so we are told) is how "meaning" is used in cognitive science. Yet no room is left to doubt what the view is taken to signify. When necessary, Simon can be both explicit and strong. Meaning is: "symbols or symbol structures in a reader's memory," "cranial symbol structures," something that "takes place in the collection of neurons called the brain." This concrete, internalist stance runs deep, too, applying equally to "context," "intens(t)ion," and a potpourri of similar terms.

It isn't the explicit bits that are interesting, though; they are straightforward, even a bit flat. The intriguing parts are where Simon is not explicit-passages where words like "meaning," "context," and "emotion" occur unmarked. It is thus that we encounter claims that "emotions are aroused by the meanings. . . .evoked by reading a text." Or that "we can use context to help. . . .narrow. . . .ambiguity." Or that it can be important to understand "the various meanings. . . .attached. . . .by Chinese students, Chinese peasants, and foreigners of various nationalities. . . .to the Statue of Liberty in Tiananmen Square."

Are such claims true? It depends on whose language is being spoken. On an ordinary reading of the words "meaning," "context," "arouse," etc., they are eminently plausible. On an internalist reading, however, at least to my ears, they verge on the preposterous, and are almost surely false. Thus is Simon's strategy betrayed. By writing in unsubscripted language he assumes that what he says explicitly early in the text (e.g., that meaning = neuronal-patterning) will be strong enough to govern the reader's later interpretation. His aim is simple and understandable: he wants to be clear, and assumes that the text's power derives from that clarity. I am not so sure that is really what is going on, however. More seriously, I suspect the text's plausibility surreptitiously relies on the reader's failure to embrace internalist language. To my mind, that is, the text's power, such as it is, depends on an essential and paradoxical unclarity.

To unpack this ambiguity, take "meaning[internal,]" to denote the radically internalist, concrete notion of meaning to which Simon pledges allegiance, and "meaning[actual,]" to denote what I will call meaning in the wild: namely, whatever meaning really is, whatever we get at when we unreflectively ask what someone means, or are troubled by the meaning of a letter. Simon's claim, I take it, is that meaning[actual]=meaningi[nternal ](hence his use of unmarked language); he also presumes to be referring to meaning[internal ]throughout the remainder of the text. Crucially, though, it is meaning[actual] that authors wrestle with, meaning[actual] that undergirds literature's constitutive regularities, and meaning[actual] that critics and literary theorists are interested in-whatever it turns out to be.

Turn then to the argument. Many of Simon's sentences go down ever so easily. Some seem obvious, even banal. Such as that "emotion is usually evoked not by using words like 'sad' or 'happy,' but by creating situations to which we respond with sadness or happiness." But-crucially-they only go down easily when read naturally, in terms of meaning[actual]. That's not how Simon intends them, however-or at least (more on this in a moment) it's not how he thinks he intends them. When he writes "context," he intends to refer to something "that can be made wholly operational," not to the governing situation in 17th-century Germany; when he talks about evoking meaning, that he is "talking about perfectly definite processes that can be executed by mechanisms"; when (as above) he mentions "situations," that he means internal mentalese arrangements, not real-world situations in which the protagonist finds herself, in Paris, say, or Montreal.

It is an amazing ruse. As the paper proceeds, the distinction between readings (actual and internal) gradually fades from view. By middle and late sections-"Criticism as creators of meaning," "Liberal education," "Who owns the text?"-the feat is complete. It would never occur to a naive reader, presented with these sections in isolation, that an internalist interpretation was ever intended.

So this appears to be the rhetorical situation. By not marking the internal-actual distinction, thus carrying on simultaneously on two different planes, the text cleverly coöpts intuition. The reader's unreflective, tacit, intuitive response agrees with the natural reading. The text, however, tries to coöpt this agreement, by presuming that the reader's unwitting participation bolsters the case for internalism, even if in reality it does no such thing.


As one turns from generalities to details, moreover, the plot thickens, in increasingly ironic ways. Thus consider Simon's intentions, with respect to these two levels.[1] As mentioned, Simon undoubtedly thinks he means meaning[internal] throughout. If challenged, he would (will?) surely say so. But what does that prove? Not necessarily all that much. Just as in politics a perpetrator's intentions (especially explicit intentions) do not exhaust the moral implications of actions, so too, as many critics have argued, authors' intentions (especially explicit intentions) do not necessarily exhaust the meaning of their texts. Ironically, that is, it may not ultimately matter what Simon thinks he means by "meaning." The superficial plausibility of his text, hence its potential power, hence such genuine insights as it is able to convey, hence the intellectual weight borne by the words, hence the responsibility incumbent on Simon for writing them-none is exhausted by Simon's explicit intentions. Those intentions would be exhaustive, of course, if Simon's views on meaning were right. But it won't do to require prior acceptance of Simon's view in order to understand the text that espouses it. That would indeed be vicious circularity.

Note that Simon's own understanding is implicated in this claim. I.e., I am not just saying that readers will partially understand Simon from a natural perspective (take "meaning" to mean "meaning[actual]"), thereby undermining the very conclusion to which they are being enticed. I am also suggesting that the natural perspective may in part be motivating Simon himself, in spite of his conscious intentions. I would be very surprised (to hazard a psychological claim about a psychologist) if the lay power of these claims were not in part responsible for seducing Simon's own allegiance.

How is the ruse maintained for twenty-five pages? If two notions do not line up, after all, as I believe these do not, you might expect their conflation to be sustainable in introductory or summary prose, but to disintegrate under pressure of details.[2] Perhaps it will. Remarkably, however, we are given no details. This is especially curious, given Simon's own protestations. Repeatedly, Simon goes to the mat for precision, claiming for example that "familiar terms like meaning, context, evocation, recognition, image have gained a clarity from the researches of contemporary cognitive science that they did not have in earlier writing and still do not have in literary criticism and its theory." Now precision is not rigour-as AI should have taught us by now.[3] But even assuming we are interested in precision (and setting aside the comment's snide tone), one cannot help wondering whether the lack of details may not be necessary to the maintenance of the ruse-i.e., whether it is not an essential, rather than an accidental, omission in Simon's text.

Space permits consideration of only a single example.[4] Whereas Simon thinks that symbols in the head are meanings, a more plausible (and historically more countenanced) position would hold instead that symbols in the head, if indeed there are such things, mean.[5] To such people (the rest of us?), the question of meaning arises for internal symbols, just as it does for language; it is not answered by them. The internal question seems prior, if you take language to be derivative on thought; subsequent, if you conversely believe that thought is dependent on language. Either way, however, the internal question is an instance of a general issue about how symbols or intentional states can carry or convey meaning-how they signify or represent, or in virtue of what they are intentional. That there are internal symbols, that there are external symbols, and that the two relate is good to know, if it was not already obvious-but that fact alone does not answer the fundamental intentional mysteries.

Moreover, the difference between symbols being meanings (Simon's view) and symbols themselves meaning something (my view) is just the classic distinction between use and mention. Failure to make the distinction-confusion of a name for what is named-is traditionally called a use/mention confusion. This is recursively (and intricately) relevant here because of the self-referential twist that characterises the distinction between meaning[actual] and meaning[internal.] But the bottom line is simple enough: the plausibility of the rhetorical ruse r elies on a semantic unclarity about what is a vehicle of meaning, what is meaning itself.

If I am right, moreover, that this conceptual confusion is necessary to sustain the ruse-and hence to underwrite the superficial plausibility of Simon's paper-then one would expect use/mention errors to permeate the paper. And sure enough, they do. Thus we read (i) that "symbol structures. . . .come to awareness," as opposed to being aware of the dinosaurs those structures represent; (ii) that a symbol is "a pattern that denotes. . . .some other pattern" as opposed to denoting something external-or, especially in literature, something fictitious, and thus perhaps not denoting anything at all; (iii) that "symbols can be represented by neuronal patterns" [emphasis added], whereas elsewhere we were led to think that symbols were neuronal patterns; (iv) that a "computer program. . . .is the formal equivalent of the system of differential equations that natural scientists so commonly use to express their theories," a seeming confusion of theory, model, and subject matter; (v) that "computer simulation is the gold coin that stands today behind the promissory notes of neurophysiologists," even though everyone from Searle on down knows that simulation per se (e.g. in the hands of meteorologists) says nothing about the computational nature of the phenomenon being simulated; and (vi) that we must understand "relations among meanings, and corresponding relations among the words and sentences that denote them," whereas normal use of "denote" is not between a sentence and its meaning, but rather between a sentence and lunch.

It is impossible to ignore the sheer peculiarity of the fact that these and other similar passages systematically blur the very semantic distinction that differentiates the view Simon is advocating from elementary observations of commonsense. One simply has to wonder whether this ambiguity does not, as one suspects on first reading, merely reflect a lack of intellectual care, but rather, as one becomes convinced on second reading, may be something that is structurally, even if unintentionally, necessary. Qua text, in sum, Simon's article is not one that could easily withstand being clarified.

Much more remains to be said, but other commentators will undoubtedly fill in those gaps. Are there meanings in the head, to come full circle? I doubt it, at least not in the sense Simon imagines. Does that mean that cognitive science and literary criticism have nothing in common? No. I am as committed as Simon to the idea that, au fond, both communities wrestle with similar problems about symbolism, interpretation, cultural dependency, and the like. I am less convinced, though, that the bond of commonality will stem from either side's yet having pearls to offer the other (or, to speak more modestly, since I know only the cognitive science half, that we have pearls to offer critics). I more suspect that a common sense of purpose will be forged from a joint recognition that we are both stumped, to say nothing of being humbled, by the same problems. Simon's claim that "there is nothing much mysterious about 'meaning'," besides being wholly unsupported by his text, strikes me as spectacularly wrong. I also believe that establishing commonality will be much harder than Simon thinks, ultimately requiring an overhaul of the metaphysical basement, rather than a back-and-forth shuttling of concepts on the third floor. But I attempt no such argument here. In fact my present point is really quite simple: merely an ironic reflection on the fact that both fields apply to the practice of both fields-and thus that a reflection of our practices through each other's lenses may for now prove more illuminating than premature marriage to either's subject matter claims.

Previous Next Up Comments


1. By "intention" in this paragraph I mean what is intended; more on this recalcitrant word below.

2. A similar confusion reigns in object-oriented languages: ambiguity between object-in-the-machine vs. represented-object-in-the-world.

3. Programs are of necessity precise; else they would not run. But as is obvious to anyone who has ever tried to extract defensible intellectual claims out of hacked-up code, precision alone implies neither clarity nor rigour.

4. One more should at least be mentioned: the recalcitrant word "inten(s/t)ion." Clarity, I take it, demands a 3-way distinction among: (i) intension-with-an-s, a term from logic or the philosophy of language, denoting roughly whatever contribution a term or sentence makes in an opaque context to the truth-value of a whole sentence containing it (usually modelled with properties or functions from possible-worlds and perhaps contexts onto extensions); (ii) intention-with-a-t, having essentially to do with anything that is semantical, significant, referential, or "directed," in Brentano's (or at least Chisholm's) original sense, and (iii) something I will call "intention-with-a-d"-namely, the ordinary English word referring to what a person has in mind when they intend to do something. The three senses are famously related: intentional (with-a-t) states are intensional (with-an-s); and intending (to do something, say) is one species of intentional activity, and hence also intensional. But relatedness is different from identity. The literary question, I take it, of whether for example a text means what its author intends, has to do with the third of these (intention-with-a-d). The "broadly philosophical" notion to which Simon explicitly refers can only be the second, though he wavers back and forth between the second and the first, at least as regards spelling (though sometimes, for example in discussing the Constitution, he uses "intension" to refer to what is clearly the third). For the defender of a theory of meaning to be roughshod about so basic a distinction is hardly reassuring.

5. This assumes that there are symbols in the head, of course-itself a contentious view, and one I do not automatically support.