BOOK REVIEWS (2013): Philip Pettit’s (1975) The Concept of Structuralism: A Critical Analysis [Some More]

8 June 2013

Summary (in Once Sentence)

If we want a machine to generate knowledge, then we should what we build generate knowledge and not merely what we want.


Last year in 2012, I set myself the task to read at least ten pages per day, and now I’m not sure if I kept up. I have the same task this year, and I’ve added that I will write a book reaction for each one that I finish (or give up on, if I stop). These reactions will not be Amazon-type reviews, with synopses, background research done on the author or the book itself, unless that strikes me as necessary or if the book inspired me to that when I read it. In general, these amount to assessments of in what ways I found the book helpful somehow.

Consequently, I may provide spoilers, may misunderstand books or get stuff wrong, or get off on a gratuitous tear about the thing in some way, &c. I may say stupid stuff, poorly informed stuff. There are some in the world who expect everyone to be omniscient and can’t be bothered to engage in a human dialogue toward figuring out how to make the world a better place. To the extent that each reaction I offer for a book is a here’s what I found helpful about this, then it is further up to us (you, me, us) to correct, refine, trash and start over, this or whatever it is we see as potentially helpful toward making the world a better place. If you can’t be bothered to take up your end of that bargain, that’s part of the problem to be solved.

This is the third of two (or more) reactions.

Another Reaction To: Philip Pettit’s (1975)[1] The Concept of Structuralism: A Critical Analysis

I aim with the first part of this third response (“Metaphorical Uses of Semiology”) to keep it briefer because what Pettit examines here, the applicability of the structural linguistic model to other domains, involves very high-level or broad-ranging assumptions, so I hope to remain more specific in my reactions. He also at times cursorily, but justly, formulates and refutes claims to veracity or utility by structuralism. The second part of this third response (“Model and Metaphor”) does not aim for brevity.

Metaphorical Uses of Semiology

The basic framing he offers for constructing an analogy between language and other domains—“the narrative, dramatic, and cinematic; non-literary arts such as music, architecture and painting; and the customary arts, as I shall call them—fashion, cuisine, and so on” (42)—involves those “areas where entities are to be found which resemble sentences in the required way: [i.e.,] they have meaning of such a kind that they can exhibit certain states of mind and they produce this meaning by something like a mechanism of articulation” (42).

A high-level objection here would ask: if the only requirement arises that  something has “meaning of such a kind that it can exhibit certain states of mind and produce this meaning by something like a mechanism of articulation,” then understanding language to be a kind of intention engine (or intentional engine), then why should we take it as the prototype?[2] Why not take language as simply one of among a veritable auto lot of intentional engines—the human being itself being first and foremost the most intentional engine of all? Summary of this objection: the bait-and-switch that substitutes “language” for “intentional engine” seem dubious.

In summarizing what he calls straight analysis, Pettit admits that the kinds of quasi-structuralist ridings of texts as if sentences by Barthes or Todorov  do not “identify in narrative texts anything that might count as the ‘language’ of narratives: if the texts are ‘sentences’ there is no ‘language’ to govern their construction” (45). Pettit doesn’t mind this, so long as it does not go too far, i.e., a Tel Quel “fundamentalism [derived] in part from Derrida’s insistence that the play of contrasts in which language produces meaning is an arbitrary play of contrast arbitrarily chosen” (45). He adds:

It is difficult to argue against such fundamentalism. All one can say perhaps is that so long as reading is taken in its ordinary sense—and certainly this is the only sense in which its value is proven—analysis must be subject to the control of something like the principle of [Rawl’s] reflective equilibrium. To reject the control of that principle is to despise the general psychology of reader and drive literary analysis into the sheerest caprice (45).

The mid-level objection here points to Pettit’s naïve insistence that the ‘ordinary sense’ of reading “is the only sense in which its value is proven.” The higher-level objection involves Pettit’s attempt to put a limit on the extent to which structuralism gets taken. The difficulty of his argument with such fundamentalism arises from his having accepted the premises of that fundamentalism in the first place. This seems rather like a torturer becoming astonished that one of his victims died.

If I follow Pettit, he examines a proposal in Barthes’s (1966)[3] Crítique et Vérité, which Pettit reports Barthes never followed up on and subsequently rejected altogether,[4] for a systematic theory of literature. The deal-breaking difficulty noted by Pettit involves that, whereas for sentences one (at least in principle) has a test, the rules of grammar, to distinguish well-formed and poorly formed sentences, we do not have the analogous test for making  a similar determination about a text. Pettit otherwise defends, I believe, other aspects of sentence building as they bear upon text building,[5] but his concern for “literary analysis [as] sheerest caprice” (45) draws him up short. Nonetheless, one might ask why—besides the obviously conventional reason Pettit avers—one needs this analogous test. Poetry in manifold ways—some so obvious we tend not to see them (i.e., the semantic sense of line breaks)—offer nongrammatical examples of language. And this, not merely that poetry reveals previously invisible potentialities, though it does that as well, but may even invent possibilities that did not exist previously.[6] Literary texts may do the same— Bakhtin (1981)[7] certainly seems to think so with the novel—and  the part that addresses the “anything goes” worry of Pettit’s involves whatever rigors exist for those trying to tease new affordances out of language in the first place. Canetti’s (1960)[8] Crowds and Power, claiming status as a work of nonfiction, got justly panned by Phillips (1963)[9] as a bad poem; Here we have just the opposite of what goes on in a good poem: instead of an original and concrete association that puts things in a new light or makes for a new experience, an ordinary observation is given ‘poetic’ overtones, and made to sound more suggestive” (¶3). Certainly in language itself “anything goes” does not prevail; I can write “asdf;lkj qewr;lqew; qe qewlk;rj qewr;lkj” but that doesn’t mean I should or that doing so yields an interesting result.[10]

Describing two kinds of paradigmatic approaches to narrative, these descriptive theories focus on form and content. And what Pettit concludes for one “what is possible is not always interesting” (50), he concludes for both, with the same caveat in each case for “considerable non-theoretical interest. [Such theories] may offer a set of terms for approaching a certain sort of comparative analysis of texts. Indeed, on reflection, this is precisely what Propp’s[11] theory provides: its goal, after all, is the comparative analysis of Russian folk-takes, not an abstract theory in any strict sense of the term” (52). He neglects to mention how such comparative analysis, if done cursorily and ad hoc,, may provide grist for literary writing. I recall reading somewhere that the twentieth century more than any previous era displays a reciprocal and mutually informing relationship between literature, literary critics, and literary criticism.

Systematic analysis [exemplified by Barthes’s (1970)[12] S/Z], like straight analysis, requires nothing more by way of assumption than what the central structural argument provides … that when a text is given an interpretation by a reader, it is given a structure that puts limits on what may replace any event in the text. What systematic analysis does is to show for each event the significance which it has, by contrast with possible replacements, in terms of setting, story, and other dimension (53).

The second premise of the argument is that the meaning of any part in a non-literary work of art is determined by the background contrast with what might have replaced it without making nonsense of the whole (56).[13]

An important trace that Pettit keeps repeating in the environs of the word ‘systematic’ shows up, for example, in the phrase “if a part changes in its meaning the meaning of the whole changes too” (56).[14] Hence, “the non-literary work of art, like a sentence, produces its meaning by the way in which it articulates its parts … that the ‘audience’ of such a work projects a structure on it which defines the possible replacements for any element” (58).[15],[16] However, also this: “for any … [art, or experience] to be given meaning it must be assigned a structure which puts limits on what can replace any element in the piece” (63) or else interpretation devolves to “anything goes.” We might sidestep this fussiness on Pettit’s part, which desires to avoid letting the intellectual life of the academy turn into an empty game,[17] and ask instead whether or not we do this in everyday life—or, more precisely, why in the structures we assign to everything we encounter in daily life does “anything goes” not already prevail? I will return to this below.

Against a gestalt of information theory vision of art that overemphasizes “the power of a work to challenge an surprise” (61), Pettit notes:

What is of equal importance … is satisfaction. I meant he word etymologically: the repetition in a work, usually through variations, which ensures that the aesthetic subject is given ‘enough’, that he is satisfied and not just tantalised. It is a point of the most common experience that the human being enjoys having his expectations fulfilled—as rhyme fulfills expectations in poetry for instance. A work of art is satisfying only if this desire is indulged in some measure (61).

Reprising his earlier argument in a new setting, Pettit questions the value of generative or descriptive theory given that, unlike the ‘test’ of grammar that permits the determination of well-formed and poorly formed sentences, we find no similar sort of test for well-formed or poorly formed customary arts (in fashion, cuisine, &c); “with no piece of customary art is our intuitive sense of structure firm enough to bear representation in a grammar of strict recursive rules” (63). He sees structure more as a gestalt in this case.  However, in the most general way, one might assign ‘culture’ as the test for customary arts analogous to ‘grammar’ as the test for linguistic performance.

I do not want to pursue this problematic proposal at the moment, except to suggest that to the extent that grammar itself already proposes a reflection of culture, then ‘culture’—as the set of constraints on human behavior in a given society, modifiable by that society—may provide the more adequately descriptive ground for a “master metaphor” than language. Specifically, rather than a ‘language’ of dance or film or architecture, for instance, we might speak of a ‘culture’ of dance, architecture, film, &c, each with their own way of speaking (their own discourse). By this sense of, for example, the ‘culture’ of film, I specifically point not to any culture (as a milieu) of the film industry and whatnot, but in an expressly metaphorical sense standing in exactly the position where one normally finds ‘language of film’ (and the like). What this might entail I leave aside for now. I will point out, though, that when Saussure elected to bracket out parole—as the customary way of speaking in a given culture—he thereby “invented” language as a reification of that customary way of speaking and thus falsified it, or, rather, erased it. So also Lévi-Strauss, who responds to myths as “messages of a most disheartening complexity, and which previously appeared to defeat all attempts to decipher them” (qtd in Pettit, p. 73, emphasis added)[18] by his objective, his desire rather, to analyze that structure of myth that “he takes to remain constant from version to version” (73):[19] “the supposed multiplicity of features proves to be largely illusory” (1958, qtd in Pettit, p. 76).[20]

Model and Metaphor

Pettit ends his book with a discussion of models beginning by a characterization of metaphor. On this view, he describes models as a systematic metaphor. We might, taking as an example Donne’s poetry in a literary tradition, call such a systematic metaphor a conceit. A conceit, precisely (and in Donne’s case, in poetic form) offers not simply a metaphor—as the famous likening of a pair of lovers to a mathematical compass—but an extended one, whose use at first may seem unwarranted but, by the end and in its fuller articulation, seems surprisingly apt after all.

I emphasize this in part because Pettit has already expressed a concern against “anything goes” and, in the advocacy of the sense of systematic metaphor for model, he remains at some pains to not let just any old metaphorical linking occur. While anticipating Lakoff and Johnson (2003)[21] in some respects and citing other previous critics of the metaphor as a device, Pettit teases out semantic aspects of various metaphors that help to ground their plausibility—that (I would say) help to suade the reader to them as conceits.  In particular, Pettit notes—I leave out his examples to avoid bogging down in them—that when we propose a metaphor, there exists simultaneously some sense of analogy and disanalogy between the two things metaphorically linked, and that the more necessary culture finds it to assert and maintain the disanalogical portions of the comparison, the less likely the case that the metaphor will die as a metaphor and transform into a nonmetaphorical piece of language, perhaps even an adage or proverb, e.g., “misery loves company”.

At one point, he utilizes earlier commentary to add something significant to the notion of metaphors.

In the “subcategorial” aspect of linguistic description, what Pettit calls “selectional subcategorisations” help to structure the listener’s expectation of what type of word should or likely will follow; thus, when we see “John plays _____,” the selectional rule sets up the expectation that something like “music” or “golf” will follow, e.g., “John plays golf.” Pettit identifies violations of these selectional expectations as expressly metaphorical, e.g., “Golf plays John.” A first point to note about this, Pettit observes, involves that the sentence does not seem wholly nonsense—one could say that John has become a golf addict and in that sense golf plays him; Foucault’s “we do not speak discourse, discourse speaks us” proposes a more dire but similar example.

However, I would not overestimate this capacity to paraphrase. Even without providing a “realistic” sense-paraphrase for “Golf plays John,” it nonetheless does generate a sense of meaning—however much a nebulous, vague, or difficult to express meaning (precisely because, as the sentence makes clear, the only adequate way to express the idea comes about exactly in the form “golf plays John” and so we have no ready-made conventional meaning for this “new” idea). This differs notably from the other famous example (from Chomsky) that Pettit borrows: “colorless green ideas sleep furiously”. I do not intend to seriously analyze the merits of different examples of metaphors, but simply want to point out that the construction of this second example presupposes multiple contradictions—“colorless” and “green”, color as a property of an abstract idea, the ascription of an action ‘sleep” to an abstract noun, and the contradiction of sleeping furiously, &c. For me, because these several elements get heaped up together the meaning-making capacity of the sentence gets clabbered up for me or, slightly more precisely, it doesn’t have the same stark numinousness that “Golf plays John” has (for me).

It may seem a digression, but I suspect it would bear fruit to examine this more closely. If we imagine various subphrases of “colorless green ideas sleep furiously” in a poem—because in a poem, we will tolerate almost anything in language—then we might consider “a colorless green” and “ideas sleep” and “sleep furiously” keeping in the back of our mind what Pettit says about aptness in metaphor as it relates to the analogy and disanalogy each proposes.

Of the three phrases, “a colorless green” “ideas sleep” an “sleep furiously” I sense the strongest metaphorical pull or suggestive strangeness in “ideas sleep”. For me, “sleeps furiously” seems already nonmetaphorical—anyone who has had a rough night of trying to sleep next to someone tossing and turning has witnessed someone else “sleeping furiously”. One may say that while “sleeps furiously” on the face of it seems selectionally barred, i.e., it proposes a word one shouldn’t expect after “sleeps,” nonetheless, once we see it, it seems in retrospect not barred on “realistic” or conventional grounds after all. In the same way, “runs slowly” seems equally already nonmetaphorical after all.

“A colorless green,”—I think the phrase requires the addition of an ‘a’ to more generate a sense of meaning in the first place—reads more like a straight contradiction at first, but ‘a colorless green” after all could mean a green that has had all the hue removed from it. (One might ‘abuse” the example, and take “green” to indicate the green of a golf course, which as a noun may be given an adjective like colorless, however badly that fares for the greens keeper of the course.) Thus, “a colorless green” shifts my attention from green in some absolute sense and suggestively implies that its greenness instead has gotten skewed, distorted, or modified in some way so that it now looks like a “colorless green”. In exactly the same way I might imagine a “light darkness” or a “dark lightness,” both phrases having a tendency to point me—without justification—toward a sort of visual imagining of “grey”. This has no justification because a “light darkness” would not, strictly speaking, amount to grey, but the suggestiveness of the phrase, as a metaphor, has precisely some kind of evocative power in similar to “Golf plays John.”

However, there remain gradations of suggestiveness or strangeness going on in these examples, and Todorov (1974)[22] description of the literary genre of the fantastic sheds some light here. Briefly, this genre operates by affecting a hesitancy in the reader (and often in the protagonist) of a story. Classically, in ghost stories, this hesitancy arises from uncertainty, for instance, whether the ghost actually exists or represents some sort of a hallucination in the narrator (or protagonist’s) mind. However, this bated-breath sense of hesitancy lasts only so long as we do not resolve this question, and most stories will come down on one side or the other of it finally. Thus, the genre of the fantastic teeters perilously between two adjacent genres, the uncanny and the marvelous. In the former, the ghost—as in all episodes of Scooby-Doo—turns out to be a hoax, so that a rational and human explanation occurs, while in the case of the marvelous, the supernatural prevails despite any evidence to the contrary: the climax of The Exorcist goes this way. Sometimes, however, the author does not resolve the matter and/or sometimes the reader does not resolve the matter either, and rather than lapsing into one of the “side genres”—of the supernatural debunked or the supernatural affirmed—we remain hovering in a (pleasingly tantalizing or perhaps aggravating) uncertainty and hesitancy.

So, we might array the final disposition (of the story or the reader) for these three genres in terms of “ah, that all makes sense after all” (the uncanny), “uh, that makes neither sense nor nonsense” (the fantastic), and “oh, that makes no sense” (the marvelous), where the starting point (for the story or reader) stood initially on the tightrope of the fantastic.

These three possibilities characterize the subtle variations of strangeness in the paired words of “colorless green ideas sleep furiously”. With respect to the kind of strangeness apparent in “sleeps furiously” the selectional violation in it (as the generative mechanism for the sense of fantastic or strangeness in the first place) falls easily down on the side of “ah, that makes sense after all”. Conversely, “colorless green” primarily because it propose a semantically literal contradiction would most likely fall down on the side of “oh, that makes no sense” An important point about this: to say “it makes no sense” has a misleading quality; we might say “that’s non-sense” rather, the marker “non” sense being precisely what the phrase generated. The phrase does not generate nothing; it generates non sense. This remains true in the genre of the marvelous as well—the conclusion of the story that affirms the existence of the ghost asks us to accept a reality of ghosts, which some people incline to already, but for them a ghost story falls in the category of literary realism then. The whole drift of The exorcist, its narrative wiliness, lies precisely in allowing the mother’s unceasing skepticism—her refusal to believe that a demon has actually possessed her daughter –to get worn down along with the reader’s, so that the supernatural affirmation at the end of the book offers as much terror in the prospect of girls like Regen behaving horribly but also the “demonstration” that demons (pun not intended) exist. We may say, “oh, that’s nonsense” as a way to deal with this (marvelous) conclusion, but we do so from a standpoint that the author has said something, has asserted a meaning rather than something meaningless. Similarly, we may dismiss “colorless green” as “oh, that’s nonsense,’ but not because nothing got meant.

This leaves “ideas sleep” to suggest the fantastic. Certainly, with this phrase I sense a “strangeness” most akin to what I encounter in “Golf plays John”. Here, we encounter metaphor in the sense that Pettit especially underscores where his point about analogy and disanalogy specifically apply. As an initial reaction, whatever ideas “do,” sleep seems not one of those things, nonetheless, in trying to sort out the “paraphrase” for what this means, we may refer to Todorov again. He particularly ascribes the composition of the fantastic, logically enough, so that in at least one sense we can blame the authors for whether the story ends ultimately—and thus comes to reveal itself as in—the genre of the uncanny, the fantastic, or the marvelous. However, readers do not always follow an author’s intentions, so that (let us imagine an example of the fantastic) the reader may decide she has read an example of the uncanny or the marvelous instead. One way this readily happens arises when the reader discerns or elects to read the story as an allegory (hence the uncanny) or poetry (hence the marvelous). Thus, one might dismiss the meaning of “colorless green” as a piece of poetry; one might read the insistence on the supernatural in The Exorcist as an allegory. Readers’ readings varying, I propose no definitive pronouncements about how these variations might play out. One could take “sleeps furiously” as a piece of poetry (analogous to the marvelous); for me, the phrase seems ultimately commonplace even if it initially and briefly surprises me. &c.

The purpose of this—as well as Todorov’s excursus—does not propose to explain all literature genres or reactions to them, however. It particularly applies to the moment of hesitancy prior to making an either/or determination (sense, or non-sense) or suspending that determination (neither sense nor nonsense). So in particularly when confronted by the phrase “ideas sleep,” I feel a longer sense of pause confronting it than with “colorless green” or “sleeps furiously”. All three phrases have a strangeness—fantastic, marvelous, an uncanny, respectively—but we may distinguish them in their further qualities of that strangeness.

Of these three selectional violations, “ideas sleep”[23] resonates most strongly (for me) as a metaphor in Pettit’s sense[24] even as the disanalogy between them, i.e., the set of qualities one might attribute to sleep not typically shared with “idea”, overshadows the sense or meaning that the metaphor generates (for me). I find myself quickly turning the phrase first into a piece of “mere” poetry and then secondly personifying “ideas” and imagining them carrying on with their day in some sense that amounts to an allegory. I feel the “gravity” of this either/or strongly, i.e., I find myself spending less time on whatever “ideas sleep” itself and more considering it in its poetic or allegorical senses. Other people will react differently of course.

By contrast, “golf plays John” leaves me lingering much longer in the strange world where something like golf in some sense plays John. I have available to me taking this in an allegorical sense—just as Pettit does when he proposes that John’s golf addiction has got the better of him—or I can take this in the “poetic” way one might read the transformation of Gregor Samsa into “an insect”—more properly, some kind of “vermin”—at the beginning of Kafka’s Metamorphosis. But in both the poetic and allegorical case, these “explanations” do not have the same degree of satisfaction that a poetic (or marvelous) and allegorical (or uncanny) reading of “colorless green” and “sleeps furiously” have (for me). Pettit especially emphasizes that along with the pleasure of surprise in literature, we find also the pleasure of having our expectations affirmed. He specifically uses the word satisfaction to describe this, pointing to the etymological sense of the word as “enough”.

What constitutes “enough” for one person differs from another, so again no one may utter definitive laws about how this plays out in a given text. For me, I find it ‘enough’ to take “ideas sleep” as oscillating alternatively between the uncanny and the marvelous, the poetic and the allegorical, while dwelling for a much less appreciable length of time in the fantastic. I find the opposite true for “golf plays John”; the specific weirdness and charm of this proposal, but also my inability to find “enough” in the poetic and allegorical standpoints to account for “what the phrase means” marks out something I deem more like the fantastic as Todorov describes it and the apt (systematic) metaphor (or conceit) as Pettit describes it.

One might compare “misery loves company” to “golf plays John” to see how strikingly two otherwise extremely similar constructions may resonate: in this case, because misery loves company has become a platitude. The fact that usage moves this phrase into common parlance means also we may turn it metaphorical by reversing it: company loves misery. So we see that whatever analogies or disanalogies we might assert between “company’ and “misery,” the force of cliché has pressed all of that juice out, as from an olive with its pit removed.

At the same time, I hardly know where to begin in trying to pretend what analogies or disanologies must inhere between “golf” and “John”. The phrase seems witty at the very least, quite apart from whether it “means” (or points to) John’s golf addiction playing him, i.e., leading him about by the nose, to marshal in another metaphorical platitude. One could pretend that despite being a proper name “John” simply offers a specific but fungible term for “person” but even on that assumption, what then do we label as the analogies between (the game) “golf” and “persons” (who play the game). Borrowing Hjelmslev’s commutation test, we may swap terms in the phrase and examine the results: music plays John; golf plays Joan;[25] dead plays John; death plays John; golf slays John.

The term “plays” seem obligatory in this example, so that rather than trying to find analogies between “golf” and “John,” we might suspect that at least something significant lurks in the verb. A conventionalized kind of metaphorical construction will often offer something of the form “X is Y”, i.e., war is hell. In this case, “is” makes the identity proposed by the metaphor explicit and so, perhaps for that reason, on the one hand makes the project of determining the analogies and disanologies of the metaphor more visibly necessary but on the other masks that necessity to the extent that such an analogy, like misery loves company, no longer seems metaphorical. Nonetheless, this kind of “is” metaphor changes starkly when reversed: war is hell and hell is war, or love is war and war is love.

For “Golf plays John,” it presupposes, one might say, the reversed version. We know very well what it means that john plays golf, but in light of that, what might “golf plays John” mean then? Whatever charm or wit one discerns in “company loves misery” would seem to rely on a similar contrast.

Pettit, citing numerous stands of linguistics, constantly invokes the notion that language “means”—explicitly in its paradigmatic sense—by contrasts with background words or a sense (by a reader or listener) of what words were chosen relative to those that were not. The play of language I resorted to above belies my autobiography, not facts about language, though I do assume that my processing of language would bear (does bear) similarities to whatever process you have experience reading these examples. Golf plays John” has the numinous, metaphorical charge and charm I read from it because I already had the “corrected” (reversed) version “John plays golf”. But even without this in advance, one might stumble across it simply by reversing the terms—otherwise “Golf plays John” might seem a perfectly bland thing to say, no more or less metaphorical than “John plays golf”.[26]

I emphasize all of this because typically the verb of an analogy gets overlooked, perhaps because we construct the most obvious ones in the form “X is Y”.  One may see (a structurally similar) example to “golf plays John” in “the boy may frighten sincerity”. For me, this tends sooner to a poetic or allegorical sense on the face of it, tempered somewhat when I contextualize it by its reversal: sincerity may frighten the boy.

And all I point to by these examples illustrate (my) possible readings of phrases and the kinds of reactions they generate (in me) as meanings. The technique of swapping terms—given a technical formality in Hjelmslev’s commutation test—does not presuppose a necessity but only points to a way of engaging an otherwise “unsatisfactory” phrase to (attempt to) make it satisfactory, to make it provide “enough”. In the case of “sleeps furiously” I have “enough” almost instantaneously, and even a sense that the phrase seems belabored. Perhaps as an adverb “furiously” already seems an afterthought, and that adds that much more to its sense as already conventional language; perhaps “furiously” itself signals the belabored sense. With “colorless green,” the obviously deliberate contradiction also all but instantaneously gives me “enough” (there at the beginning of the phrase) an I complacently file it under a “poetic usage”. With “ideas sleep” (or, at this point more precisely, “colorless green ideas”), the predisposition to “poetry” in the opening two words may set up the frame to take “ideas sleep” in a poetic sense as well, whether furiously or not. And so I can parse the whole thing as if straight out of cummings or Ogden Nash.

I recognize of course that one needn’t parse the sentence in two word phrases; to do so would adjust the meanings meant—one might read “colorless green ideas” and “ideas sleep furiously” and so forth—but I don’t intend to claim what this phrase means, unless without recognizing also the plurality of the word “means”.

Because speaking involves an iterated sequence, then sequence matters—an precisely that it matters means as well we may find times when the priority of sequence does not completely dominate (a sentence). Thus, if someone wanted to say, if with some difficulty, that “Sleep furiously colorless green ideas” still means in essence or more or less the same thing as “colorless green ideas sleep furiously,” we can’t say the same for “John plays golf” and “golf plays John”. Similarly, to say “green colorless ideas sleep furiously” seems (to me) to stray too far from the original to still call it the same— “green colorless ideas” being in the final analysis (to me) not identical enough to “colorless green ideas”—and this, I would say, because (for whatever reason) here sequence matters more than in the case of “sleep furiously colorless green ideas”.[27] Even at the granularity of “ideas sleep,” a poet might trick us by letting us assume that “ideas” presents a plural noun but eventually discloses “ideas” to be a third-person conjugation of the verb to idea (whatever that means) with sleep as the subject that ideas:

as night’s eyes
sleep’s wide black

so wide
pries sleep
night’s black eye

the eye’s knight
while black wards

and eyes
as black
sleep’s wide-backed

All of this to lay some further lines of suggestion for making the metaphor of metaphor into a model for models. Perhaps the main thing to note concerns reversibility, not only the meaning that emerges when we reverse  metaphor but also the intransigence against doing so. We mays similarly remain sensitive to the notion that a model (as a systematic metaphor, as a conceit) in all likelihood hovers more in the zone of the fantastic than collapsing into one of the side genes—much less starting in one. Pettit speaks at length of the necessary integration of philosophy and empiricism, especially for all of the humanistic disciplines and the social studies that mistakenly call themselves science. The center of these projects in particular stands as the phenomenological subject, precisely in a delicate (fantastic) position on a tightrope between the threat of lapsing into positivism—as Pettit makes clear that Lévi-Strauss did—or the solipsism of “anything goes”.

I might suggest further attention paid to the verbs of metaphors in Pettit’s sense; what actions, what processes, constitute the relation between the parts of the metaphor. The conventional assumption that “is” links the parts of a metaphor lays the groundwork for a misleading claim to transform (in a mathematical sense) the one into the other, but the unbounded license Lévi-Strauss permitted himself to argue one myth represented a transform of another, the mathematical pretenses of Jakobson and the Prague school of linguistics, and the “original” bracketing out of all human users of language by Saussure and its reinvention in its syntagmatic and paradigmatic transform all point to the manifold problems, deliberate or accidental, that come from metaphors with “is”. If a metaphor emerges as a violation of what Pettit call selectional rules—without requiring we subscribe to the whole theory that comes with—then in what sense do models, as systematic metaphors, represent something of such a selectional violation?

In a remarkable section of his book, Peter Taylor (2005)[28] empirically demonstrates that models never do not work; to the extent that they generate knowledge at all, they do not cease to do so simply when “wrong”. Moreover, simply because a model generates apparent contradictions or impossibilities (in some of its elements) does not mean it fails as a model—he demonstrates this with a model conventionally dismissed as erroneous (because it generates results populations smaller than zero) but which nonetheless demonstrates accurate predictive power.  To summarize it too breezily, these anomalous results may arise “simply” from the analytical terms of the model, but the clear point Taylor makes asserts that no amount of refinement in the number of kinds of those terms will entirely erase all anomalousness—at best, it may reach a point where such anomalies remain left out of focus. Taylor’s further point advocates developing a thick skin about impossibilities and not necessarily building assumptions into a model to exclude those possibilities, e.g., ruling out the possibility in advance of populations less than zero.

I bring this up to say two things. First, in ramifying our models as conceits (as systematic or extended metaphors), this greater articulation proposes a reduction in anomalies rather than an approach to any “truth,” which we can never reach. Second, it points to a right orientation for understanding the inadequacy of overly simplified models—such as structuralism’s unregulated desire to cast everything into binary operations. The problem of such an approach doesn’t arise from too much simplicity—Taylor makes clear that all models if they generate knowledge at all will generate knowledge—and not necessarily even in the likely legion numbers of anomalies an analytical inadequate model would generate. The objection, rather, arises where the modeler begins making exceptional, i.e., making exceptions, about those anomalies, and especially offering appeals outside the framework of the model itself. Fundamentally, this issue involves vanity, institutional continuity, and issues themselves that stand outside of the model, but at root this involves moving the goal-posts; formalized, we call this data massage statistics.


[1] Pettit, P. (1975). The concept of structuralism: a critical analysis. Berkeley, CA: University of California Press, pp. i–v, 1–117.

[2] In the phrase “can exhibit certain states of mind,” Pettit underscores an emphasis on intentionality, without which one might say that inert matter “exhibits” as well. Thus, this framing offered for constructing an analogy between language and other domains characterizes language as an intentional engine apart from some brainless or mindless process.

[3] Barthes, R (1966). Crítique et vérité. Paris: Seuill.

[4] Barthes, R (1971). A conversation, in S. Heath, C McCabe, and C. Prendergast (eds). Signs of the times, pp. 44–5. Cambridge: Granta.

[5] In effect, he seems to give voice to my sarcastic question, which I ask when someone speaks of a “language of film” and the like, “what, then, comprise the nouns of this text? What grammatical case or cases does this text move through?” A non-systematic approach to this characterizes, it seems, what Pettit calls straight analysis. It also reflects, though he only dimly acknowledges this, something of the approach in Russian formalism.

[6] This involves a tricky claim. A chair affords use as a murder weapon, but shall we call this affordance for murder an originally present potential or a newly discovered possibility? Saying this, we might more cogently note how these affordances (either to sit or to kill) arise from the interface of the chair itself with a human user. Saying this proposes an alternative way of talking about objects like chairs and subjects like humans that re-enmeshes them together as an interactive unit otherwise stereotypically (and incorrectly) taken as separate elements. By “incorrectly” I mean that a description of objects and subjects apart from the affordances arising from their interface involves and offers a relatively unfit way of describing the human experience involved in such interfaces and affordances; I do not mean that the description “is wrong”. The relative unfitness of this description (that separates objects and subjects) leads to debates less well described as fruitless and more accurately as iteratively perpetuating a status quo. For example, by separating the human subject from the object of a gun, both sides of the gun debate may blame the other side, and the social consequence of that blaming substitutes debating about the issue for taking action on the issue (one way or another). When summarizing a debate between the merits of a patriotic versus a cosmopolitan outlook, Nussbaum (2002) asks and answers:

May I give my daughter an expensive college education, while children all over the world re starving and effective relief agencies exist? May Americans enjoy their currently high standard of living, when there are reasons to think the globe as a whole could not sustain that level of consumption? These are hard questions, and there will and should be much debate about the proper answers (137).

In the face of wholly unjust human suffering around the world, Nussbaum concludes that “these are hard questions, and there will and should be much debate about the proper answers” (137)—a remarkably inhuman conclusion.

[7] Bakhtin, MM (1981). The dialogic imagination: four essays (ed. and trans. M. Holquist and C. Emerson). Austin: University of Texas Press.

[8] Canetti, E. (1981). Crowds and Power (trans. Carol Stewart), 6th printing. New York: NY: Noonday Press.

[9] Phillips, W. (1963, 1 February). History on the couch. [online] The New York Review of Books from here.

[10] Unless someone in a position of some variety of institutional authority (whether academic or merely wealthy) declares it as such and markets the phrase as significant (not just interesting).

[11] Propp, V. (1968). Morphology of the folktale, 2nd ed., Austin, TX: University of Texas Press.

[12] Barthes, R (1970. S/Z, Paris: Seuill.

[13] Regarding this point, Pettit alludes to Hjelmslev’s (1953) ‘commutation test,’ which “consists in replacing a word in a sentence an watching the effect on the whole of different replacements” (57). Moles (1966)** takes this further: “One of the most general heuristic procedures of aesthetics, based on the materiality of the work of art, consists of progressively destroying the work by known, perceptible quantities, and in following the variations in esthetic sensation, value, and knowledge as a function of this destruction” (201).

* Hjelmslev, L (1953). Prolegomena to a theory of language. Bloomington, IN: Indiana University Pres.

** Moles, A (1966). Information theory and aesthetic perception. Urbana, IL: University of Illinois Press.

[14] Or later, again, almost as if without acknowledging that he has written this earlier:

There are two premises to the argument. The first is that the meaning of the whole depends on the meaning of its parts so that it changes if the meaning of any part changes. The second is that the meaning of a part is determine by its background contrasts with what might have replaced it without making nonsense of the whole. These premises … imply that any … piece of art produces its meaning by having its parts set up against appropriate contrasts (62).

[15] Pettit links systematic analysis of nonliterary arts with the literary arts (stylistics and narratology): “

Systematic analysis corresponds to stylistics when the latter describes the significance of the words of a poem in semantic, metrical and symbolic terms, and to narrative analysis when it describes the significance of events in a text in terms of narrator’s presentation, from a definitive perspective n with a definite purpose, of setting, character, or story (58).

[16] In an aside, Pettit contrasts descriptive systematics in music in two forms, using the terms technical and expressive to characterize them; one recognizes in these the analogs of form and content in descriptive systematics in literature—the former aspiring to describe how the parts achieve their effects, the latter to describe the effects of a piece’s parts. The faults of these two approaches in music my apply to the literary as well: the former “leave much out of account—most notoriously, rhythmic pattern—and can be accused of giving only the appearance of rigour … [for the latter] the worst fault is that the terms it uses are necessarily vague and offer no hope of characterizing artistic effects with any precision” (59). The disconnect here consists, in the incapacity in the former (technical/formal) case to satisfactorily explain how various motives, (musical) gestures, even chords and key changes, for instance, actually affect the critic’s claimed effects, while in the latter (expressive/content)  case the incapacity arises in not satisfactorily establishing that the affects noted by the critic constitute the effects at all. Thus, in the former, we may often have more an autobiography of the critic’s capacity to sniff out patterns and in the latter an autobiography of the critic’s sensibilities in their works, both of them claiming (sometimes not entirely without some justification) to write history, not autobiography. Or, in a nonacademic setting (such as music critic for a major newspaper), they my unabashedly embrace autobiography while still arrogating to themselves the claim (again, with some justification since in some case such critics have made and broke musical personalities) to make history. In this, we may discern that history and autobiography (in the fields of music and literature) stand as analogs to form and content in music or literature, or technical and expressive descriptive systematics in criticism of music or literature.

[17] It must be said that so-called postmodernism, as a reflection of late-order (one hopes) capitalism, to turn the intellectual life of the academy into an empty game provides a crucial element. Much else needs adding to this to flesh out the point, but what we see missing from Pettit’s analysis, who after all wrote this in 1975 before the most glaring aspects of academia’s empty game had come to the fore as they since have, involves his failure to analyze who benefits from this empty game, whose interests get served by it. Thus the liens of power in late-order capitalism co-opt the artist and the intellectual—in a sense the analogous figures of content and form, expression and technique, autobiography and history—and politically neutralize their meaning-generating function (see my comments on Suttner, 2005 here) for culture.

[18] Lévi-Strauss, C. (1964). The raw and the cooked. London: Cape.

[19] Pettit provides a weird moment in Lévi-Strauss, when Lévi-Strauss (1963)* avers, “behind every sense there is a non-sense and not vice-versa. For me meaning is always phenomenal” (qtd in Pettit, p. 73). As an anti-phenomenologist, this functions rather like ‘this sentence is false,” and so points to a (perhaps in this case deliberate) obfuscation of terms. After all, if Lévi-Strauss in some sense purports to describe the savage mind, then what savagery shall we read out of the merely phenomenal meaning of his text? Another gesture of colonialism? He seems to acknowledge this: ‘because we are prisoners of subjectivity we cannot understand things simultaneously from within and without” (ibid, qtd in Pettit, p. 72). His phrase “prisoners of subjectivity” belies a (his?) desire for freedom, but if phenomenology provides, re Foucault, Althuser, Lacan, &c, an illusion of freedom and escape, then Lévi-Strauss delivers his lectures from his prison cell while imagining he stands on Sinai. This reflects no mere semantic detail; beginning with Jakobson’s phonology, which he believed had “reached beyond the superficial conscious and historical phenomena to attain fundamental and objective realities consisting of systems of relations which are the products of unconscious thought processes” (1958, p. 58, qtd in Pettit, p. 76; see note 20 below), he concludes that “all forms of social life ‘consist of systems of behavior that represent the projection, on the level of conscious and socialized thought, of universal laws which regulate the unconscious activities of the mind’ [1958, p. 59]” (Pettit, 1975, 76). These propose freedoms as well—or rather, solace from the prison of subjectivity at least in offering an alternative. Similarly, when Lévi-Strauss insists that “the time of myth … is reversible” (c.f., 1958, p. 209–12) and that “the purpose of myth is to provide a logical model capable of overcoming a contradiction (n impossible achievement if, as it happens, the contradiction is real” (ibid, p. 229, qtd in Pettit, p. 81), these too betoken an escape from what Eliade (1954)** describes as the terror of history that the linear advance of time proposes and an escape from the (for some experientially) unbearable contradiction of the Self in Jung’s sense as simultaneously ego-consciousness and the unconscious. One may sympathize over the desire to avoid the terror of history or the contradiction of human existence, but to respond in opposition to an existential or phenomenological understanding of human beings by describing human beings as an “intolerable spoiled child who for too long has held the philosophical scene and prevented any serious work, drawing exclusive attention to itself’ (1971, p. 614–5,*** qtd in Pettit, p. 77) sounds an awful lot like a petulant temper tantrum thrown by an intolerably spoiled child.

*Lévi-Strauss, C. (1963). Auteur de la Pensée sauvage: réponses à quelques questions, Esprit (November 1963). Available online in French here.

** Eliade, M. (1954). The myth of the eternal return. [New York]: Pantheon Books.

*** Lévi-Strauss, C. (1971). L’homme nu, Paris: Plon.

[20] Lévi-Strauss, C. (1958). Structural anthropology. Harmandsworth: Penguin.

[21] Lakoff, G., & Johnson, M. (2003). Metaphors we live by. [New ed.]. Chicago: University of Chicago Press.

[22] Todorov, T. (1993). The fantastic: a structural approach to a literary genre (trans. Richard Howard). Ithaca: NY: Cornell University Press.

[23] In a pleasant irony, Pettit’s chapter critiquing Lévi-Strauss’ structuralism quotes the Frenchman that if structuralism may at present sleep, it will one day wake up in the natural sciences. Pettit responds, “on the day when semiological analysis awakes it will not be in the natural sciences” (78).

[24] Pettit insists, broadly, that a selectional violation generates something at least metaphorical if not a metaphor itself. So the objection that “sleeps furiously” and “colorless green” do not propose metaphors but simply some unusual combination of words specifically betray as a commitment—justified or not—to a particular sense of metaphor.

[25] To the extent that “golf plays John” means differently than “golf plays Joan” shows the patriarchal biases of language.

[26] Thus, a degree of linguistic acculturation changes the meaning of “golf plays John”—a detail that gets marshaled in by assumption in the form of linguistic competence or cultural competence, &c. But we needn’t assume this; rather, we might stop pretending that language stands separable from users of it.  Sentence in a foreign language I do not speak “means” precisely non-sense to me. Here, I encounter no hesitancy a la the fantastic whatsoever, but simply the non-sense (and non-sensical premise, the supernatural premise, that an Other might speak and encode the world differently than I do).

[27] One may rightly object that I have changed the verb tense to permit the rearrangement. You might experiment with rearrangements that do not change the verb tense to read off your own hangs of meaning in the sentence so deranged.

[28] Taylor, PJ (2005). Unruly complexity: ecology, interpretation, engagement. Chicago: University of Chicago Press

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: