Thursday, December 6, 2012

On the arbitrariness of the signifier

Writing as I recently did about the origins of words and about the instability of meaning reminded me of something.

A while back I read (about half of) the I suppose seminal gang-non-manifesto Deconstruction and Criticism. My feelings on deconstruction, to the extent that I comprehend it well enough to have feelings on it at all, are mixed. I think it asks very real questions that do not have easy or comfortable answers; and like M.H. Abrams I find that it has illuminated aspects of texts that might otherwise not have been so illuminated. On the other hand, I am very sympathetic to the arguments of those (like, again, M.H. Abrams) who point out that a critical method that always knows, every time, what it is going to find in a text before it even starts to read it--and is always right!--is an impoverished critical method. I have various other feelings about it--like that it seems able to deny political and experiential necessity even more than the New Criticism was (though it also seems more open to not denying this); or that the frequency with which sexist conceptions of The Feminine seem to pop up on the way to its aporias is highly problematic; or that I suspect that, despite its applicability to language in general, it just fundamentally makes more sense in French, a language where meaning is far more contingent than it is in English--but that's probably enough for now.

What I want to discuss is something Paul de Man says (and I think it's a common assertion) in his contribution to the book, "Shelley Disfigured." Towards the end of this essay on Percy Bysse's The Triumph of Life, de Man writes:

If, for instance, compelling rhyme schemes such as "billow," "willow," "pillow" or transformations such as "thread" to "tread" or "seed" to "deed" occur at crucial moments in the text, then the question arises whether these particularly meaningful movements or events are not being generated by the random and superficial properties of the signifier rather than by the contraints of meaning. The obliteration of thought by "measure" [a word whose course through the poem de Man has been tracing] would then have to be interpreted as the loss of semantic depth and its replacement by what Mallarmé calls "le hasard infini des conjonctions."
Now, I confess that I am (appropriately?) on very shaky ground here, discussing all this. I have already admitted that I have only the weakest of grasps on deconstructive theory; and here I must admit that I have read neither Shelley nor, except as I shall discuss momentarily, Stéphane Mallarmé. (I hope to investigate both before too long, but have not gotten to either yet--there are a lot of books in the world.) It is entirely possible that de Man here is archly and cattily accurate about Shelley--I wouldn't know. However, in terms of the underlying point he is making about the "superficial properties of the signifier" and what he describes a moment later (in the course, actually, of admitting that there is something more to the text than this) as the "arbitrary element in the alignment between meaning and linguistic articulation," I feel that he has missed something essential.

It is too bad, too, because this is a truly interesting subject, and de Man's discussion of it, interesting so far as it goes, would be far more interesting still if he realized that while there may indeed be an arbitrary element at work here, these similarities in sounds are far from "random"--and farther still from "superficial." To illustrate the point, let us look at the etymologies of those "compelling" rhymes de Man singles out, billow, pillow, and willow. They are a particularly good case study in that they are, if the 1991 edition of Clement Wood and Ronald Bogus's Complete Rhyming Dictionary in collaboration with my own brain can be trusted, the only three common disyllabic words in the English language to share this rhyme.*

*There are polysyllabic rhymes, but even they are as a rule recent and/or highly local importations from other languages that have not as yet changed orthographic form during their presence in English, such as (some pronunciations of) armadillo, cigarillo, and peccadillo; there is also the archaic killow, which does not seem to fit the scheme I am about to describe--and I am hesitantly tempted to suggest that this may be part of the reason the word has fallen out of use.

At the moment I have access only to the Oxford American Dictionary, so I will be limited to the very brief etymologies contained therein--but I think they will be more than suitable to my purpose. They are as follows, each quoted directly and completely:

  • billow: mid 16th cent.: from Old Norse bylgja.
  • pillow: Old English pyle, pylu; related to Dutch peluw and German Pfühl, based on Latin pulvinus 'cushion.'
  • willow: Old English welig, of Germanic origin; related to Dutch wilg.
The first thing to notice about these three word origins is of course their utter diversity: we have one word of Scandinavian origin, one of Italic, and one of Germanic; two words that have been developing with the language since the days of Old English and one which entered the language significantly later (indeed, closer to the time of Shelley than to that of Old English); none which in their "original" forms rhyme with one another or, for that matter, their current forms--and the same is true of the variety of original spellings.

And yet somehow, over different time scales, these three words converged to the point where the only difference in both sound and spelling is one letter (or phoneme). Possibly the process by which this occurred is in each case understood--indeed, once I finish writing this and take it to a computer with internet access to post it online, I no doubt could research each case individually. That I am not doing this is just one more instance of the shaky ground from which I write, I suppose. However interesting each of these case studies likely is, though, I suspect that they would illuminate little of direct relevance to my purpose here.

What will illuminate that purpose, as is so often the case, will be recourse to Samuel R. Delany, who writes at section 54 of his extraordinary discursive (in both senses of the word!) essay (in the current and older senses of the word!), "Shadows":

Think of grammar solely as the phonic redundancies that serve to transform a heard utterance from the interpretive field, through the range of assocations in the hearer/speaker's memory that includes "his [sic] language," into the hearer/speaker's generative field as an utterance.

In the qui, quae, quo of Latin, for instance, I'm sure the Roman brain (if not the Roman grammarian) considered the redundancy of the intial "qu" sound as grammatically significant (in my sense of "grammar"), as it considered significant, say, the phonic redundancy between the "ae" at the end of "quae" and the "ae" at the end of "pullae." (We must rid ourselves of the notion of grammar as something that applies only to the ends of the words!) In English, the initial sound of the, this, that, these, those, and there are all grammatically redundant in a similar way. (The "th" sound indicates, as it were, "indication"; the initial "qu" sound, in Latin, indicates "relation," just as the terminal "ae" sound indicates, in that language, "more than one female.") What one can finally say of this "grammar" is: When a phonic redundancy does relate to the way that a sound is employed in conjunction with the other sounds/meanings, then that phonic element of the grammar is regular. When a phonic redundancy does not relate, that element is irregular. (The terminal "s" sound on "these" and "those" is redundant with the terminal "s" of loaves, horses, sleighs--it indicates plurality, and is therefore regular with those words. The terminal "s" on "this" is irregular with them. The terminal "s" at the end of "is," "wants," "has," and "loves" all imply singularity. Should the terminal "s" on "this" be considered regular with these others? I suspect in many people's version of English it is.) For all we know, in the ordinary English hearer/speaker's brain, "cream," "loam," "foam," and "spume" are all associated, by that final "m" sound, with the concept of "matter difficult to individuate"--in other words, the "m" is a grammatically regular structure of that particular word group. Such associations with this particular terminal "m" may explain why most people seldom use "ham" in the plural--though nothing empirically or traditionally grammatical prevents it. They may also explain why "cream," when pluralized, in most people's minds immediately assumes a different viscosity (i.e., referentially, becomes a different word; what the dictionary indicates by a "second meaning"). I suspect that, in a very real sense, poets are most in touch with the true "deep grammar" of the language. Etymology explains some of the sound-redundancy/meaning-associations that are historical. Others that are accidental, however, may be no less meaningful.

This, I think, is very interesting! And if we apply this thought process to our three -illow rhymes, it is easy to see a way that they are "regular" with one another in this way, though I am finding it a bit difficult to put it into other words: something to do with a bulging, but not awkward, shape or action, perhaps, would be the best way to put it.

A digression: after immediately making this connection as I read de Man, further reflection led to my being flummoxed for a time by willow--or, rather, the adjective willowy, which in its meaning of long, tall, and slender seems highly "irregular" both with the related adjectives pillowy and billowy and with the mental image I, at least, conjure up on encountering the word willow, which is without fail one of a weeping willow--hence bulging, etc. In conversation with my mother the plant expert, she pointed out that willows tend to grow in very fast, straight shoots before curving over again, and that perhaps this was what was meant; and indeed the both of us as well as my father agreed that to us, despite the absence of this in the dictionary definition,* willowy always carried a connotation of bending, waving; a person so tall and slender that they and their limbs seem to curve, bend, blow in the wind--and it strikes me as at least possible that this aspect of our internal definition might come at least as much from association with billow and pillow as from association with the trees themselves. Digression over.

*OAD: "(of a person) tall, slim, and lithe."

This is all, to be sure, on some level very silly, and if we attempt to take it in any deterministic or predictive direction it all falls apart instantly: for there are many hundreds of words more closely linked through "bulging" than these three that share no element of sound.* However, if we view the matter through the non-teleological lens of actual language use through time, I don't think it at all far-fetched to think that three words of different derivations but with moderately similar phonemic qualities could, over time, tend to converge phonetically and orthographically under the influence of one another and their hazily shared qualities.

*Interestingly, bulge does not seem to be etymologically related, as one might suppose it would be, to billow's root bylgja, coming instead from Latin bulga, leather bag.

The ground we are on is certainly aptly described by Mallarmé's hasard infini. But it is important to remember--and this is what I think de Man, in his eagerness, has forgotten--that "infinite chance" is not the same thing as "randomness." Right now (for since I said I did not have internet access I have relocated and am now writing at work--don't tell anyone) I can go to my favorite website,, and generate a string of however many truly random numbers from whatever range I choose--say, 5 integers between 1 and 100. When I do it at this moment, I get 61, 80, 33, 32, 82. I can come up with any number of ways in which I find these numbers to be linked (the adjacency of 33 and 32, the shared final digit of 32 and 82, the shared initial digit of 80 and 82, etc.), but this is just a game; in truth, there is no relationship, no reason whatsoever that 32 followed 33 rather than any other number; and if you click on my link and generate your own list, it will be different, and there will be no relationship between your numbers and mine--even if the two lists share items--even if they happen to be entirely identical lists, which, though unlikely, could happen--and even in its unlikelihood would mean nothing at all.

This is randomness; and perhaps one might be able in some constrained, to me unacceptably constrained, fashion to think of it as "infinite chance." But when Mallarmé speaks of le hasard infini des conjonctions I suspect that he has something rather different in mind.

As I mentioned before, I have not read Mallarmé. But in thinking about the issues I am discussing in this post, I decided that I should attempt to verify to whatever extent was possible my suspicions about this quote and de Man's use of it, so (on an earlier occasion at work) I googled for "Igitur," the work from which the quote comes, and found Mary Ann Caws's translation. My reading of "Igitur," I will freely admit, has been rapid, distracted, and superficial,* and I would not pretend to be more familiar with Mallarmé than Paul de Man is! However, I have noticed some things that seem to complicate the straightforward implication de Man seems to be reading out of these words.

*And though I have no reason to distrust Caws's translation, I also have no reason to trust it--I have no idea.

First of all, as I had suspected, it does seem that when Mallarmé says "conjonctions," he really means it: that is to say, he is talking about the infinite chance of conjunctions, not of words in general--and in fact though the grammatical sense of conjunction is very much on his mind ("igitur" is itself a conjunction, after all, and a particularly tendentious one at that), it is not the only sense on his mind. Now, speaking for the moment just linguistically, I am not so naïve that I think an assault on conjunctions could remain limited to just this one grammatical class. Calling one aspect of language into question, as (I hope) we saw with Elizabeth Willis's treatment of "to be," inevitably calls all of language into question. But it is still important to remember, as de Man seems not to here, the kernel: in Willis's case, it is statements of existence or equivalence that become so problematic as to unhinge all speech; in the current case (that of Mallarmé), it is specifically the infinite chance or randomness of conjunctions that troubles.

This, on a personal note, is something I can very much relate to: as (so I flatter myself) a writer conjunctions are for me, along with the verb to be, the possessive pronouns, and some others, among the most fraught of grammatical categories. William Empson's admission (in Seven Types of Ambiguity) that he has "usually said 'either...or' when meaning 'both...and'" is noble as far as it goes, but to me does not go nearly far enough; for while "either...or" is untenably restrictive in its way, so too can be "both...and." Both, all, conjunctions are inadequate; in the right mood I might go so far as to say perverse. How can it be anything other than perverse that I am forced to say "but" in that sentence of a moment ago ("noble as far as it goes, but to me...") when in many (but not all!) ways what I want to say is "and"? How can it be anything but perverse when I can't even be sure what the difference, finally, is?

And so when we use a conjunction, it is in many ways arbitrary; two things, two concepts are linked together, somehow, and we must try to express this conjunction--but (and!) the grammatical conjunctions we have available to us are set up in such a way that they must either exclude (such as the "either...or" in which I am currently engaged), or by their inclusiveness exclude the notion of exclusion. It is dizzying; it is a problem.

But (and!) not only a problem. Just a moment before he speaks of this infinite chance of conjunctions, Mallarmé suggests that "(t)he infinite emerges from chance, which you have denied." In some ways this could, and (or!) should, be taken as addressing the conjunctions themselves: by pretending to be absolute, finite, comprehensive and comprehendible, they claim to deny chance, and thus collapse what should, or could, be the infinity of our perspective to the finite realm of language, rationality, singularity. But there are other ways to take this, too: as an admonishment, perhaps, against those who would deny that there is an infinite chance to conjunctions; perhaps there is a suggestion here that we should throw ourselves into this infinite chance, embrace it, use these arbitrary conjunctions in the gleeful knowledge that that is what they are--and allow the infinite to emerge from this chance.

The problem with trying to assign, or glean, specific, one-sided meaning from Mallarmé's utterances in "Igitur" is of course the same as the problem with trying to place one specific conjunction between two other grammatical bodies: the nature of these utterances is such that their relation, their conjunction, is infinite, multiplex. In the two bits we have quoted here, "the infinite chance of conjunctions" and "the infinite emerges from chance," we see only a small portion of the work's constant rearrangement of and play with these words and concepts: chance, infinity. They appear in every conceivable conjunction, and more perhaps inconceivable; and this infinite chance in itself defeats the efforts of those who, as de Man seems to me to be doing, would paradoxically attempt to read some kind of absolute statement of unabsoluteness out of, or into,* them.

*Prepositions, too, are problematic.

I am probably being simultaneously a poor and an uncharitable reader of de Man. I'm not sure what to do about the former beyond to continue reading. As for the latter, I should point out that, though I've been discussing his use of Mallarmé's conjunctions solely in the grammatical sense of the word, de Man probably didn't intend it to be taken this way; after all, one conjunction he's referring to is the supposedly arbitrary one of the pillow/willow/billow rhyme scheme in Shelley. This is of course a direct conjunction of objects in the world (despite that the objects are in this case words printed on a page) rather than a solely grammatical function--though to the extent that I understand deconstruction, its adherents might object to this distinction, I don't know.

But to return to de Man's comments on rhyme specifically, I should begin noting, which I think it is worth doing, that they are not applicable, or at least not equally applicable, to every language as they are to English: in French, for example, many more words rhyme; and in Italian almost literally everything does, or can be made to--there, rhyme is more a matter of matching parts of speech than anything else (and even that is not often necessary), which is why it was much easier for Dante to write however many million lines of terza rima without being obnoxious than it is for his English translators, many of whom don't even try, or for that matter than it was for Shelley in the very poem de Man is discussing. This may seem irrelevant to the point--which after all arises in an essay written in English on a poem written in English--but I think it is important, when discussing the contingency, arbitrariness, and putative randomness of language, to remember that these things are different, and apply to differing degrees, in different languages--in other words, linguistics knows no universals, and though to a certain extent we can speak of language itself, in most cases we're more properly speaking of a language in particular. (This also is where my earlier comment on deconstruction making more sense in French comes in.)

And when we're talking about a language, we're talking about something that has a history, a particular context for its past and present use. And it seems to me that it is this history, like all histories, that contains the dizzying arbitrariness and randomness which de Man attributes to rhyme. For who could have predicted, a thousand years ago, that bylgja, pyle, and welig would converge to the point where they could serenely punctuate a Romantic's revival of a scheme of classical artistic order? ...And yet today we can look back on this history, see what happened (to whatever extent it is possible to do so), formulate all kinds of theories about it, and allow the resonances from these theories and the feelings these words have gathered over all this time being used in ways similar and different from one another to color our reading of the poetry they so punctuate.

A favorite example of this for me is the bizarre word guerdon, particularly as it is used by Hart Crane in the "Proem: To Brooklyn Bridge" which opens The Bridge.

And obscure as that heaven of the Jews,
Thy guerdon . . . Accolade thou dost bestow
Of anonymity time cannot raise:
Vibrant reprieve and pardon thou dost show.
(The italics and ellipses are Crane's.)

Guerdon, my dictionary (again the OAD) tells me, is "chiefly archaic" and means "a reward or recompense." This definition allows us to make some sense of the stanza, to be sure; but I am equally sure that in order to feel what Crane is saying we must also be aware, consciously or not, that guerdon rhymes with, suggests, and to a certain extent shares in the meanings of, burden, a word strongly present in its absence, both through this phonic resemblance and through the surrounding Jews, reprieves, pardons, obscurations, and even bestowings. It is difficult to resolve the question of whether one is meant to superimpose the meanings of the two words, resulting in something like "a reward or payment that is in itself a burden,"* or whether the "vibrant reprieve and pardon" of the nevertheless "obscure" guerdon simply lessens or removes (reprieves and pardons) a burden; and this difficulty itself is I think a central factor in the not insignificant impact of the stanza. It is a difficulty that would not exist had Crane used the word burden directly, thus collapsing its range of potential meanings into one specific one; and it is a difficulty that could not exist did not guerdon and burden rhyme.

*This is an important concept, and it is interesting and equally important that English does not have a word for it. Some other languages do--there is a Korean word, which I unfortunately can't seem to conjure up out of the internet at the moment, that can be used to describe someone, or the actions of someone, who has burdened you with their kindness. It is, again, interesting and important to note that poetry is capable of making up for a language's lacunae in this way.

And how did they come to rhyme? Burden has a very straightforward etymology, arising from Old English byrþen, with the same meaning, which requires only the function of very well-known and familiar principles of linguistic drift to transform into the form familiar today. But guerdon is an entirely different story. The OAD gives the etymology as follows:

late Middle English: from Old French, from medieval Latin widerdonum, alteration (by association with Latin donum 'gift') of a West Germanic compound represented by Old High German widarlōn 'repayment.'
This has to be one of the most peculiar etymologies I've ever come across. There is the odd zig-zagging of the word's journey, out of Germanic languages into Romance and back into (Germanic) English, rather than passing as most words do directly from Germanic or Romance roots to English usage; there is the religious implication of its having passed through medieval Latin (as well as the peculiar feeling one gets from learning a word comes from Latin, though it was never actually used by Romans); there is the merging of distinct words similar in elements of sound and meaning, reminiscent of what we have speculated above with the -illow words; and then there is the uncommented-on transformation of widerdonum into guerdon, which I can, just barely, explain to my own satisfaction through a handful of processes familiar to me, from simple elision (-donum to -don, sure) to the interchangeability, between Romance and Germanic languages, of gu and w (as in guerre/war, Guillaume/William; or, within English, such pairs as guile/wile and guarantee/warranty).

To read this brief etymology, presented so unassumingly in the dictionary, is, to me at least, to be overtaken by a kind of vertigo, though not an unpleasant one. It is to see a word, or rather a few words, jumbling their way through history, through languages, in and out of notability; bumping into one another and sticking; curated, perhaps a bit sloppily, by monks; released back out into the wild, carried from French to English through some mechanism other than the usual one of the Norman conquerors, whereupon it presumably flails around a bit before dying--only breathed back into life, fitfully, by the occasional oddball poet.

Le hasard infini...

Infinite chance, yes. But the infinite arises, perhaps, from chance, which we are ill-advised to deny. Beyond all this that I've been talking about, there is indeed as de Man points out a large degree of arbitrariness in the assigning of sound to meaning, signifier to signified--there is no reason, say, why the -illow sound should represent this bulging shape or action common to willows, pillows, and billows, rather than any other sound. But once this representation occurs, it becomes a sort of center of gravity, drawing other representations into it, allowing (or even forcing) others to circle around it. The history of what falls in, what circles, and what remains aloof is peculiar, wonderful in the fullest sense of the word, and perhaps to a degree random. But in the words as they are used, in their relationships of signifier to signifier, there is a kind of logic, an algebra even, representing sometimes obvious, sometimes unexpected similarities in the signified, and this, far from superficial, is what rhymed (or, sometimes, unrhymed) verse ideally taps into--perhaps giving us in the process, if not the infinite itself, then at least hints towards it.


Or, doggone it, you could do like William Empson did--which I saw too late to work into this essay--and sum up just about everything I'm trying to say in one footnote. From the third edition of Seven Types of Ambiguity:

What you normally get from a likeness of sound is an added force to the Paget effect in cases where there is a clear group of words with similar sound and meaning (e.g. skate, skid, skee, scrape). But this makes you feel the meaning of the one word more vividly, not confuse it with the meanings of the others. On the other hand, it might be argued that a controlled partial confusion of this kind is the only real point of using alliteration and rhyme.
A "controlled partial confusion" in combination with "feel[ing] the meaning of the one word more vividly"--among other reasons, by reinforcement from the similar meanings of the other phonically related words--is very much what I'm arguing goes on in cases like this.

No comments: