The Sapir-Whorf Hypothesis
Geoffrey Sampson
The subject of this chapter is not so much a geographically or chronologically distinguishable school of thinkers, as an idea which has held a perennial fascination for linguists of diverse schools, and indeed for very many people who have never been students of language in any formal sense. This idea—that a man’s language moulds his perception of reality, or that the world a man inhabits is a linguistic construct—although in one form or another a very old one, has become associated with the names of the Americans Edward Sapir (1884-1939) and Benjamin Lee Whorf (1897-1941), and more particularly with the latter.
The work of these writers might well have been treated in the last chapter, since it fell squarely within the tradition initiated by Boas. I have chosen to discuss Sapir and Whorf in a separate chapter, because the aspect of their work which we shall examine represents a rather special development within the Descriptivist school, and one which conflicted fairly sharply with the thought of many other members of that school. Sapir and Whorf fully shared the relativism of Boas and his Descriptivist successors, with its emphasis on the alienness of exotic languages, while never being influenced by the behaviorism (in either “good” or “bad” senses) of Bloomfield. (Behaviorism was an element which Bloomfield imported into the Descriptivist tradition rather than finding it already there—Boas, and indeed Bloomfield himself in his early writing, were happy to discuss meanings and spent little time worrying about the logical status of linguists’ data. But Bloomfield succeeded in taking most of his colleagues with him in his conversion to behaviorism, which is why I say that there was a conflict between the ideas summarized as the “Sapir-Whorf hypothesis” and the ideas of other Descriptivists.)
Sapir studied languages of the Pacific coast of North America, and began his career in charge of anthropological research at the Canadian National Museum; in 1925 he moved to the University of Chicago, and in 1931 to Yale. Much of his work was quite comparable to that of other Descriptivist linguists, though he differed from the behaviorists in stressing that patterns revealed by linguistic analysis were patterns in speaker’s minds (it is significant that the collection of his papers published in 1949 bears the title Selected Writings in Language, Culture and Personality), and Sapir took it for granted that if one wants to know how a language is structured for its speakers it is appropriate to ask them1. Sapir’s independence of his American colleagues’ assumptions is particularly obvious in his notion of linguistic “drift”, behind the more-or-less random fluctuations which make up the detailed history of any language, Sapir thought, there was a long-term tendency for that language to modify itself in some particular direction, as the coming and going of waves on a beach marks a steady long-term tidal movement (Sapir 1921, ch. 7). This idea comes very close to implying that a language has life of its own in some more than metaphorical sense, and it would clearly have been anathema to a methodological individualist such as Bloomfield.
On the issue with which this chapter is concerned, Sapir was by no means single-minded. The
2
occurrence of his name in the term “Sapir-Whorf hypothesis” is perhaps due more to the fact that Whorf took his general approach to linguistics from Sapir than to Sapir’s being one of the most active proponents of that hypothesis. (The term was introduced by J. B. Carroll (Whorf 1956, p. 27).) In his popular book Language, indeed, Sapir suggests that differences between languages are merely differences in modes of expressing a common range of experiences, rather than corresponding to differences in the experiences themselves (Sapir 1921, p. 218). Later, though, Sapir changed his mind. Consider, for example, the following passages:
Human beings do not live in the objective world alone, nor alone in the world of social activity as ordinarily understood, but are very much at the mercy of the particular language which has become the medium of expression for their society. It is quite an illusion to imagine that one adjusts to reality essentially without the use of language and that language is merely an incidental means of solving specific problems of communication or reflection. The fact of the matter is that the “real world” is to a large extent unconsciously built up on the language habits of the group. No two languages are ever sufficiently similar to be considered as representing the same social reality. The worlds in which different societies live are distinct worlds, not merely the same world with different labels attached. [1929, p. 209; my italics]
Language … not only refers to experience largely acquired without its help but actually defines experience for us by reason of its formal completeness and because of our unconscious projection of its implicit expectations into the field of experience…. Such categories as number, gender, case, tense, … are not so much discovered in experience as imposed upon it because of the tyrannical hold that linguistic form has upon our orientation in the world. [1931; my italics]
These remarks might be interpreted as mere truisms, but if taken literally they are strong statements. The special contribution of Whorf was, by means of detailed analysis of certain American Indian languages, to make as convincing a case as has ever been made for believing that we must acknowledge the view expressed by Sapir as true in a quite radical, untrivial sense.
Benjamin Lee Whorf, a descendant of seventeenth-century English emigrants to Massachusetts, was in his scholarly work an outstanding example of the brilliant amateur. After taking a degree in chemical engineering he began a successful career as a fire-prevention inspector with an insurance company in Hartford, Connecticut, and despite several offers of academic posts he continued to work for the same company until his death at the age of 44. (Whorf learned lessons from his professional work which encouraged his belief that world-view is moulded by language. In analysing a large number of reports of how fires had started, Whorf tells us (1941a, p. 135), he began by assuming that only physical factors would be relevant but came to realize that language often played an important role: for instance, people behaved cautiously near what they categorized as “full petrol drums” but carelessly near “empty petrol drums”, although the “empty” drums contained explosive petrol vapour and were thus even more dangerous than the full ones.) Whorf’s linguistic interests were originally rather diverse; when in 1931 Sapir moved to Yale University, only thirty-odd miles from Hartford, Whorf became a regular collaborator of his and began to focus his attention mainly on Hopi, a language of Arizona. Much of Whorf’s writing discusses the special, very un-European world-view which he believed to be implied by various features of Hopi grammar.
Whorf makes the point that only certain grammatical categories in any language are marked 3
overtly, as, for example, the distinction between present and past tense is marked in every finite verb which occurs in English. There exist also numerous “covert” categories, or “cryptotypes” as Whorf sometimes calls them. For instance, in English the names of countries and towns form a “cryptotype” because, although they outwardly resemble other nouns, they cannot be reduced to pronouns after the prepositions in, at, to, from (Whorf 1945, p. 92). Thus one can say I live in it when “it” refers back to a phrase such as that house or that basement, but not when it refers to Kendal or Bulgaria— even though I live in Kendal, I live in Bulgaria are perfectly correct. Whorf felt that such covert categories were more telling than the overt categories of a language in establishing the world-view of its speakers, on the ground that the use of overt markers may be merely learned by rote but “cryptotype” can be manipulated consistently only if the categorization which they imply is real for the speaker. (If all country-names and town-names ended in some special suffix, say -ia, then an Englishman could simply remember “nouns ending in -ia may not pronominalize after a preposition” , but since they in fact have no special form we must think of them as a semantic class.) In Hopi rain-prayers, it seems that clouds are spoken of as if they were alive. Whorf points out that from this alone one cannot know whether the usage “is some metaphor or special religious or ceremonial figure of speech”, or whether the Hopi actually believe that clouds are living beings. However, the distinction between animate and inanimate exists as a covert category in Hopi. Any noun used to refer to a living being is pluralized in a special way (even when the noun is not basically animate, so that, for example, the Rolling Stones in Hopi would take the animate plural of “stone”); and the word for “cloud” is invariably pluralized in the animate way, which demonstrates that the Hopi do indeed believe clouds to be alive (Whorf 1956, p. 79).2
Although this neatly illustrates Whorf’s point about the importance of covert categories, it is not a particularly good example of the differences Whorf claims to exist between Hopi and European world-views: in this case the categories animate/inanimate are perfectly normal for a European, and the only question concerns the status of clouds with respect to these categories. (We shall consider a better example of Whorf’s thesis about linguistic diversity shortly.) Nevertheless, even here it is possible to take a sceptical stance. Thus, suppose that we encounter another tribe in which sex is a “covert category”, so that, say, all nouns referring to females evoke special suffixes in words modifying them; and suppose further that many words for inanimate objects, such as “stone”, “water”, “moon”, belong to the female cryptotype, while others, such as “iron”, “fire”, “sun” behave like the words for males. Clearly Whorf would have to conclude that this tribe holds some sort of animistic view of Nature, according to which everything that exists is alive and has a sex. But there is such a tribe: they live just across the Channel form Dover, and if there is one thing the French are not it is surely animists. Whorf did not in fact apply his notions to differences between the familiar European languages; he felt that these all presupposed the same world-view because of the long period in which Europe had shared a common culture, and he referred to them collectively as “Standard Average European”. It is perhaps appropriate to be cautious, at least, in accepting a theory which says that certain communities see the world in ways startlingly different from ours, but which is illustrated almost wholly by reference to primitive tribes about whose beliefs we have little independent evidence. The non-European language with which the present writer is best acquainted is Chinese; although traditional Chinese ideas about the world differ greatly from the European ideas, the two intellectual systems do not seem to possess quite the same quality of mutual incommensurability that Whorf alleges to occur with Hopi vis-à-vis 4
“Standard Average European”. One cannot help wondering whether this may be because Chinese civilization, although, like that of the Hopi, quite independent of Europe, has been articulate enough to refute the flights of fancy in which a Whorf might be inclined to indulge on the basis of formal characteristics of Chinese grammar.
In fact, the various contrasts in world-view for which Whorf argues differ greatly in the extent to which they are surprising or controversial. Boas had already made the point that, for instance, where English has on word snow Eskimo has separate basic roots for snow falling, snow on the ground, drifting snow, and so forth; at this relatively concrete level disparities between the conceptual schemes of different languages are fairly familiar, and there is no doubt that they influence perception—it can be shown that people’s conceptions of their surroundings are modified by the conceptual categories their language happens to provide (Lenneberg and Roberts 1956, p. 31; cf. Herman et al. 1957; Hanson 1958). Whorf discusses cases of this kind, but they are not what he is primarily interested in. “What surprises most”, Whorf rightly says, “is to find that various grand generalizations of the Western world, such as time, velocity and matter, are not essential to the construction of a consistent picture of the universe” (1940, p. 216). Hopi, in particular, “may be called a timeless language”: the language does not recognize time as a linear dimension which can be measured and divided into units like spatial dimensions, so that for instance Hopi never borrow spatial terms to refer to temporal phenomena in the way so common in European languages (before the door ~ before noon, between London and Brighton ~ between 9 and 10 a.m., in the box ~ in the morning), nor does Hopi permit phrases such as five days since daytime is not a thing like an apple of which one can have one or several. Furthermore, Hopi verbs do not have tenses comparable to those of European languages. And since there is no concept of time, there can be no concept of speed, which is the ratio of distance to time: Hopi has no word for “fast”, and their nearest equivalent or “He runs fast” would translate more literally as sometimes like He very runs. If the Hopi rather than Europeans had developed sophisticated scientific theories, Whorf suggests, modern physics would be very different from what it is, though it might be equally self-consistent and satisfactory.
Notes:
1 Sapir has been cited by members of the modern, Chomskyan school as a forerunner of their own movement. I find this judgment somewhat forced; Sapir did not construct explicit arguments against the behaviorist principle, as Chomsky has done, he merely remained uninfluenced by the arguments for behaviorism ( Sapir was interested in questions of substance rather than of methodology).
2 The following English parallel may be worth mentioning. In conservative dialects of English, including my own, animateness is a covert category in that only animate nouns may take the “Germanic Genitive” (N’s as opposed to of N; innovating dialects permit expressions such as the car’s wheels, the theory’s influence, but these phrases are ill-formed for me). I have been mildly disturbed to notice that the one noun in my own speech which consistently takes the Germanic genitive in prima facie violation of this rule is the noun computer.
(From Geoffrey Sampson, 1980. Chapter 4, Schools of Linguistics: Competition and Evolution. London: Hutchinson.)
