Learning from Language: Symmetry, Asymmetry, and Literary Humanism

2. Two Famous Asymmetrists

18 2 Two Famous Asymmetrists When Plato took up the question of symmetry directly, in his dialogue Cratylus, he knew that he was stepping into an ongoing, high-stakes discussion. He knew that it was Heraclitus, the celebrated pre-Socratic philosopher of change, who first proposed an educative relationship between words and the things they named. Heraclitus, the philosopher who famously proclaimed that a person cannot step into the same river twice, was impressed with what he regarded as the relative stability of language. In his view, external features of the world existed in such a rapid state of flux that words about them, because they changed more slowly than other things, were capable of retaining important truths now lost to direct observation (Poster 15–17). The Stoic philosophers, however, wrote more extensively on the subject (though much of this writing is now lost), and to Plato they presented a more powerful and disturbing argument. The Stoics maintained that the forms and categories of language, in spite of all kinds of surface evidence to the contrary, existed continuously in basic symmetry with the forms of the world. They must in fact do so, the Stoics argued, or we would not be able to know anything about the world or do anything useful in it. tWo famouS aSymmetriStS 19 In ancient philosophy, the symmetry question was also linked to another question specifically about language, the so-called analogy-anomaly question. Here, we deal with the fact that the surface of every language, like the surface of the world itself, is full of irregularities. Why do we have “bring,” “brought,” “has brought,” rather than “bring,” “bringed,” “has bringed,” or “bring,” “brang,” and “has brung”? To some thinkers, such facts were, in themselves, evidence enough that language was an irregular (and therefore probably unreliable) instrument. This is the anomaly position. To the analogists, however, this view seemed as superficial as the evidence on which it was based. Apparent anomalies were either the surface manifestations of a deeper system of order (roughly akin to the deep structures of modern linguistics), or they were corrupted versions of a previously existing order. The analogy-anomaly question runs parallel to the symmetry question , and ancient treatments of it tended to provide further support for symmetry. Nobody expects the existing, outward form of the word duck to be like a duck or expects an active sentence to always signify action. Language’s reference is not to things, in the sense of objects or events, but to the (truer, more stable) ideas of things. If you look deep enough, or go back far enough, you will likely find the connection. “Likely,” however, turns out to be a big word in that sentence. The explanation seems very reasonable, but it can be exceedingly difficult to demonstrate consistently. And this is why in all ages the case for symmetry has come forward more as an article of conviction and belief than as a statement of fact. In the ancient discussion, the need to keep the faith might have seemed especially strong, because the issue of language connected to two important philosophical issues. The first, as already noted, was the epistemology question: Are there stable, knowable realities out there? How can we know them? And how can we communicate reliably what we know? If something is being claimed in language, is there anything in language itself that authenticates or discredits the claim? The second, related issue is more specifically about human institutions —government, religion, the law—and that is the ancient debate of nature versus convention: Allowing for all sorts of surface differences (everybody knows the constitution of Sparta is different from that of Athens), do such institutions derive from a natural order of things, and are they therefore subject to universal laws? Or do they exist by convention—by agreements among humans at different times and places, under different circumstances. If the latter is true, then they are not accountable to uni- 20 tWo famouS aSymmetriStS versal principles, since none are better or truer than another, except as a matter of practical working. How do questions about the regularity and symmetry of language relate to all of this? First of all, human institutions are language-saturated. They are conducted in language, and in an important sense (a literal one, in fact, in the case of constitutions and laws) constructed by language. If language has no special hold on the...