Austrian Millionaires, Part Deux (the Third Installment will not be Forthcoming)

For a long time, I have had serious doubts about the Whorf-Sapir hypothesis that the characteristics of particular natural languages such as English or Navaho determine or at least seriously bias or influence the way in which the speakers of these languages understand the world. The vagueness and ambiguity of the hypothesis is part of the problem. Just as an individual determined to achieve atheism is doomed to frustration in the face of interminably mutable theological definitions, you can never be sure whether the Whorf hypothesis you refute is the relevant version or even worth refuting. At best you can raise objections that make a meaningful point against some versions of the hypothesis. Two such considerations:

1. The grammar of particular languages obliges the speaker to provide certain information to the listener. In English, for example, the third person singular pronoun always indicates gender, a linguistic fact that became a real issue in the last 40 years or so as the clumsiness of writing “he or she” over and over again drove many of us to “they.” It is far from clear, however, that it makes much difference whether or not a category like gender or tense is coded in the grammar since the same meaning can just as well be conveyed by extra words. The verb in Homeric Greek has a dual number, but exactly the same meaning can be expressed by words like “pair” in languages such as English that don’t have a dual. It has been claimed that the Arabs have a foggy notion of causality because of the rudimentary systems of tenses in classical Arabic; but the Arabs in fact have a huge historical literature as do the Chinese whose verbs so far from having tenses doesn’t even have conjugations. Meanwhile, the Indians, whose culture is notoriously unhistorical, wrote Sanskrit, a language with a complex and nuanced system of tenses. Arguably, the main effect of putting distinctions into the grammar is to make them less salient and easier to ignore. Why does it make a difference if distinctions are built into the grammar or have to be expressed lexically?

2. The Whorfians suggest that the structure of particular languages makes it hard to say or understand certain things; but this limitation, it seems to me, is an artifact of the methodology of linguistics. A person who studies a language from the outside has no business deciding what a native speaker should say in a particular instance. If there is no recorded instance of a 2nd person plural pluperfect subjunctive middle of the verb “to burp” in the available corpus of late Medieval Utopian, the linguist simply can’t say for sure whether there ever was such an animal or do more than guess what it might have been by analogy to known forms. But native speakers confront this situation all the time. For all I know, no English speaker has ever said “Had you have burped yourself immediately, you wouldn’t have been discommoded by the gas attack;” but the lack of a precedent didn’t slow me down because I have citizen’s rights in my own language and can make things up as I go along. By the same token, if I need of more ways to speak about kinds of snow, I’ll just make them up. It isn’t just people turning their doctoral dissertations into books that possess the magic power of neologism.

The words and expressions in use in a particular natural language doubtlessly betoken something about the culture of the speakers of the language, but it seems more appropriate to say that the culture expresses itself in the language than that the language molds the culture. I was reminded of this issue by an article in last week’s online edition of SCIENCE that reported on an Amazonian tribe that only has three numbers, one, two, and many. These Indians, it turns out, do very poorly on simple arithmetical tasks that involve counting or matching up rows of small objects—story problems were apparently completely out of the question. The finding, which is apparently well attested, was trotted out as a confirmation of Whorf as if the lack of counting numbers accounted for the poor math skills. In fact, what the anthropologists seem to have discovered is an instance of the extremely depauperate culture of a small and isolated group. Everything about these folks is minimalist. The tribe, seem have lost or discarded elements of culture. They don’t talk about much or build very much. In this regard, they are perhaps like the native Tasmanians who forgot how to build boats after they attained their isolated homeland.

This tribe raises the question of just how small a tool kit is enough for a viable human culture, a question analogous to the issue of how small a genome can be. The Indians are certainly economical with numbers. I hadn’t though anybody could improve on the famous Austrian millionaires mentioned in George Gamow’s book One, Two, Three, Infinity. These worthies challenged each other to come up with the biggest number. One millionaire hems and haws, takes a deep breath, grits his teeth, and comes up with “Three!” The other looks stricken, jumps up from his chair, paces around the room, grimaces, and finally despairs, “You win, damn it!” (Incidentally, George Gamow was a great contriver of names, who coined, among other terms, the expressions Big Bang and the Genetic Code, and thus exemplified the super-Worfian creativity of language users. By the way, I don’t know why Gamow settled on the nationality of the idiots in his joke; but I am sure that if he had been an amoeba, he would have called his book One, Two, Four, Infinity.)