The importance of having words by which concepts may be distinguished and isolated from one another will become clearer by a brief reminder of the nature of reflection. Thinking is in large part (as will be discussed in detail in chapter XIII) concerned with the breaking-up of an experience into its significant elements. But experience begins with objects, and so far as perceptual experience is concerned, ends there. We perceive objects, not qualities, actions, or ideas apart from objects. And the elements into which thinking analyzes an experience are never present, save in connection with, as parts of, a sensibly perceived object. Thus we never perceive whiteness save in white objects; warmth save in warm objects; red save in red objects. We never, for that matter, perceive so abstract a thing as an "object." We experience red houses or red flags; white flowers, white shoes, white paper; warm stoves, warm soup, and warm plates. Even houses and stoves and shoes are, in a sense, abstractions. No two of these are ever alike. But it is of the highest importance for us to have some means of identifying and preserving in memory the significant resemblances between our experiences. Else we should be, as it were, utterly astounded every time we saw a chair or a table or a fork. Though they may, in each case in which we experience them, differ in detail, chairs, tables, forks have certain common features which we can "abstract" from the gross total experience, and by a word or "term," define, record, communicate, and recall. The advantage of a precise technical vocabulary over a loose "popular" one is that we can by means of the former more accurately single out the specific and important elements of an experience and distinguish them from one another. The common nouns, or "general names" in a language indicate to what extent and in what manner that language, through some or other of its users, classifies its experiences. Highly developed languages make it possible to classify similarities not easily detected in crude experience. They make it possible to identify other things than merely directly sensed objects.
In primitive languages experience is described and classified only in so far as it is perceptual. In other words, primitive languages have names for objects only, not for ideas, qualities, or relations. Thus it is impossible in some Indian languages to express the concept of a "brother" by the same word, unless the "brother" is in every case in the same identical circumstances. One cannot use the same word for "man" in different relations: "man-eating," "man-sleeping," "man-standing-here," and "man-running-there" would all be separate compound words. Among the Fuegians there is one word which means "to look at one another, hoping that each will offer to do something which both parties desire but are unwilling to do."[1] Marett writes in this connection:
[Footnote 1: Marett: Anthropology, p. 140.]
Take the inhabitants of that cheerless spot, Tierra del Fuego, whose culture is as rude as that of any people on earth. A scholar who tried to put together a dictionary of their language found that he had got to reckon with more than thirty thousand words, even after suppressing a large number of forms of lesser importance. And no wonder that the tally mounted up. For the Fuegians had more than twenty words, some containing four syllables, to express what for us would be either "he" or "she"; then they had two names for the sun, two for the moon, and two more for the full moon, each of the last named containing four syllables and having no elements in common.[2]
[Footnote 2: Ibid., pp. 138-39.]
It is easy to see how very little refinement or abstraction from experience could be made with such a cumbersome and inflexible vocabulary. The thirty thousand word vocabulary expressed a poverty of linguistic technique rather than a richness of ideas.
At the other extreme stands a language like English, which is, to an extraordinary degree, an "analytic" language. It has comparatively no inflections. This means that words can be used and moved about freely in different situations and relations. Thus the dominant elements of an experience can be freely isolated. A noun standing for a certain object or relation is not chained to a particular set of accompanying circumstances. "Man" stands as a definite concept, whether it be used with reference to an ancient Greek, a wounded man, a brave, a wretched, a competent, or a tall man. We can give the accompanying circumstances by additional adjectives, which are again freely movable verbally and intellectually. Thus we can speak of a brave child and a tall tower as well as a brave man and a tall man. In Marett's words:
The evolution of language then, on this view, may be regarded as a movement away from the holophrastic [compound] in the direction of the analytic. When every piece in your playbox of verbal bricks can be dealt with separately, because it is not joined on in all sorts of ways to the other pieces, then only can you compose new constructions to your liking. Order and emphasis, as is shown by English, and still more conspicuously by Chinese, suffice for sentence-building. Ideally, words should be individual and atomic. Every modification they suffer by internal change of sound, or by having prefixes or suffixes tacked on to them, involves a curtailment of their free use and a sacrifice of distinctness. It is quite easy, of course, to think confusedly, even whilst employing the clearest type of language.... On the other hand, it is not feasible to attain a high degree of clear thinking, when the only method of speech available is one that tends toward wordlessness—that is to say, one that is relatively deficient in verbal forms that preserve their identity in all contexts.[1]
[Footnote 1: Marett: loc. cit., pp. 141-42.]
Languages differ not only in being more or less analytic, but in their general modes of classification. That is, not only do they have more or less adequate vocabularies, but in their syntax, their sentence structure, their word forms, they variously organize experience. It is important to note that in these divergent classifications no one of them is more final than another. We are tempted, despite this fact, to think that the grammar, spelling, and phonetics of our own language constitute the last word in the rational conveyance of thought.