Chapter VI: Part 6

Instructive forms may be cited from Nootka. We shall return to our "fire in the house."[107] The Nootka word _inikw-ihl_ "fire in the house" is not as definitely formalized a word as its translation, suggests. The radical element _inikw-_ "fire" is really as much of a verbal as of a nominal term; it may be rendered now by "fire," now by "burn," according to the syntactic exigencies of the sentence. The derivational element _-ihl_ "in the house" does not mitigate this vagueness or generality; _inikw-ihl_ is still "fire in the house" or "burn in the house." It may be definitely nominalized or verbalized by the affixing of elements that are exclusively nominal or verbal in force. For example, _inikw-ihl-'i_, with its suffixed article, is a clear-cut nominal form: "the burning in the house, the fire in the house"; _inikw-ihl-ma_, with its indicative suffix, is just as clearly verbal: "it burns in the house." How weak must be the degree of fusion between "fire in the house" and the nominalizing or verbalizing suffix is apparent from the fact that the formally indifferent _inikwihl_ is not an abstraction gained by analysis but a full-fledged word, ready for use in the sentence. The nominalizing _-'i_ and the indicative _-ma_ are not fused form-affixes, they are simply additions of formal import. But we can continue to hold the verbal or nominal nature of _inikwihl_ in abeyance long before we reach the _-'i_ or _-ma_. We can pluralize it: _inikw-ihl-'minih_; it is still either "fires in the house" or "burn plurally in the house." We can diminutivize this plural: _inikw-ihl-'minih-'is_, "little fires in the house" or "burn plurally and slightly in the house." What if we add the preterit tense suffix _-it_? Is not _inikw-ihl-'minih-'is-it_ necessarily a verb: "several small fires were burning in the house"? It is not. It may still be nominalized; _inikwihl'minih'isit-'i_ means "the former small fires in the house, the little fires that were once burning in the house." It is not an unambiguous verb until it is given a form that excludes every other possibility, as in the indicative _inikwihl-minih'isit-a_ "several small fires were burning in the house." We recognize at once that the elements _-ihl_, _-'minih_, _-'is_, and _-it_, quite aside from the relatively concrete or abstract nature of their content and aside, further, from the degree of their outer (phonetic) cohesion with the elements that precede them, have a psychological independence that our own affixes never have. They are typically agglutinated elements, though they have no greater external independence, are no more capable of living apart from the radical element to which they are suffixed, than the _-ness_ and _goodness_ or the _-s_ of _books_. It does not follow that an agglutinative language may not make use of the principle of fusion, both external and psychological, or even of symbolism to a considerable extent. It is a question of tendency. Is the formative slant clearly towards the agglutinative method? Then the language is "agglutinative." As such, it may be prefixing or suffixing, analytic, synthetic, or polysynthetic.

[Footnote 107: See page 110.]

[Transcriber's note: Footnote 107 refers to the paragraph beginning on line 3331.]

To return to inflection. An inflective language like Latin or Greek uses the method of fusion, and this fusion has an inner psychological as well as an outer phonetic meaning. But it is not enough that the fusion operate merely in the sphere of derivational concepts (group II),[108] it must involve the syntactic relations, which may either be expressed in unalloyed form (group IV) or, as in Latin and Greek, as "concrete relational concepts" (group III).[109] As far as Latin and Greek are concerned, their inflection consists essentially of the fusing of elements that express logically impure relational concepts with radical elements and with elements expressing derivational concepts. Both fusion as a general method and the expression of relational concepts in the word are necessary to the notion of "inflection."

[Footnote 108: See Chapter V.]

[Footnote 109: If we deny the application of the term "inflective" to fusing languages that express the syntactic relations in pure form, that is, without the admixture of such concepts as number, gender, and tense, merely because such admixture is familiar to us in Latin and Greek, we make of "inflection" an even more arbitrary concept than it need be. At the same time it is true that the method of fusion itself tends to break down the wall between our conceptual groups II and IV, to create group III. Yet the possibility of such "inflective" languages should not be denied. In modern Tibetan, for instance, in which concepts of group II are but weakly expressed, if at all, and in which the relational concepts (e.g., the genitive, the agentive or instrumental) are expressed without alloy of the material, we get many interesting examples of fusion, even of symbolism. _Mi di_, e.g., "man this, the man" is an absolutive form which may be used as the subject of an intransitive verb. When the verb is transitive (really passive), the (logical) subject has to take the agentive form. _Mi di_ then becomes _mi di_ "by the man," the vowel of the demonstrative pronoun (or article) being merely lengthened. (There is probably also a change in the tone of the syllable.) This, of course, is of the very essence of inflection. It is an amusing commentary on the insufficiency of our current linguistic classification, which considers "inflective" and "isolating" as worlds asunder, that modern Tibetan may be not inaptly described as an isolating language, aside from such examples of fusion and symbolism as the foregoing.]

But to have thus defined inflection is to doubt the value of the term as descriptive of a major class. Why emphasize both a technique and a particular content at one and the same time? Surely we should be clear in our minds as to whether we set more store by one or the other. "Fusional" and "symbolic" contrast with "agglutinative," which is not on a par with "inflective" at all. What are we to do with the fusional and symbolic languages that do not express relational concepts in the word but leave them to the sentence? And are we not to distinguish between agglutinative languages that express these same concepts in the word--in so far inflective-like--and those that do not? We dismissed the scale: analytic, synthetic, polysynthetic, as too merely quantitative for our purpose. Isolating, affixing, symbolic--this also seemed insufficient for the reason that it laid too much stress on technical externals. Isolating, agglutinative, fusional, and symbolic is a preferable scheme, but still skirts the external. We shall do best, it seems to me, to hold to "inflective" as a valuable suggestion for a broader and more consistently developed scheme, as a hint for a classification based on the nature of the concepts expressed by the language. The other two classifications, the first based on degree of synthesis, the second on degree of fusion, may be retained as intercrossing schemes that give us the opportunity to subdivide our main conceptual types.

It is well to recall that all languages must needs express radical concepts (group I) and relational ideas (group IV). Of the two other large groups of concepts--derivational (group II) and mixed relational (group III)--both may be absent, both present, or only one present. This gives us at once a simple, incisive, and absolutely inclusive method of classifying all known languages. They are:

A. Such as express only concepts of groups I and IV; in other words, languages that keep the syntactic relations pure and that do not possess the power to modify the significance of their radical elements by means of affixes or internal changes.[110] We may call these _Pure-relational non-deriving languages_ or, more tersely, _Simple Pure-relational languages_. These are the languages that cut most to the bone of linguistic expression.

B. Such as express concepts of groups I, II, and IV; in other words, languages that keep the syntactic relations pure and that also possess the power to modify the significance of their radical elements by means of affixes or internal changes. These are the _Pure-relational deriving languages_ or _Complex Pure-relational languages_.

C. Such as express concepts of groups I and III;[111] in other words, languages in which the syntactic relations are expressed in necessary connection with concepts that are not utterly devoid of concrete significance but that do not, apart from such mixture, possess the power to modify the significance of their radical elements by means of affixes or internal changes.[112] These are the _Mixed-relational non-deriving languages_ or _Simple Mixed-relational languages_.

D. Such as express concepts of groups I, II, and III; in other words, languages in which the syntactic relations are expressed in mixed form, as in C, and that also possess the power to modify the significance of their radical elements by means of affixes or internal changes. These are the _Mixed-relational deriving languages_ or _Complex Mixed-relational languages_. Here belong the "inflective" languages that we are most familiar with as well as a great many "agglutinative" languages, some "polysynthetic," others merely synthetic.

[Footnote 110: I am eliminating entirely the possibility of compounding two or more radical elements into single words or word-like phrases (see pages 67-70). To expressly consider compounding in the present survey of types would be to complicate our problem unduly. Most languages that possess no derivational affixes of any sort may nevertheless freely compound radical elements (independent words). Such compounds often have a fixity that simulates the unity of single words.]

[Transcriber's note: Footnote 110 refers to the three paragraphs beginning on line 2066.]

[Footnote 111: We may assume that in these languages and in those of type D all or most of the relational concepts are expressed in "mixed" form, that such a concept as that of subjectivity, for instance, cannot be expressed without simultaneously involving number or gender or that an active verb form must be possessed of a definite tense. Hence group III will be understood to include, or rather absorb, group IV. Theoretically, of course, certain relational concepts may be expressed pure, others mixed, but in practice it will not be found easy to make the distinction.]

[Footnote 112: The line between types C and D cannot be very sharply drawn. It is a matter largely of degree. A language of markedly mixed-relational type, but of little power of derivation pure and simple, such as Bantu or French, may be conveniently put into type C, even though it is not devoid of a number of derivational affixes. Roughly speaking, languages of type C may be considered as highly analytic ("purified") forms of type D.]

This conceptual classification of languages, I must repeat, does not attempt to take account of the technical externals of language. It answers, in effect, two fundamental questions concerning the translation of concepts into linguistic symbols. Does the language, in the first place, keep its radical concepts pure or does it build up its concrete ideas by an aggregation of inseparable elements (types A and C _versus_ types B and D)? And, in the second place, does it keep the basic relational concepts, such as are absolutely unavoidable in the ordering of a proposition, free of an admixture of the concrete or not (types A and B _versus_ types C and D)? The second question, it seems to me, is the more fundamental of the two. We can therefore simplify our classification and present it in the following form: _ I. Pure-relational _/ A. Simple Languages \_ B. Complex _ II. Mixed-relational _/ C. Simple Languages \_ D. Complex

The classification is too sweeping and too broad for an easy, descriptive survey of the many varieties of human speech. It needs to be amplified. Each of the types A, B, C, D may be subdivided into an agglutinative, a fusional, and a symbolic sub-type, according to the prevailing method of modification of the radical element. In type A we distinguish in addition an isolating sub-type, characterized by the absence of all affixes and modifications of the radical element. In the isolating languages the syntactic relations are expressed by the position of the words in the sentence. This is also true of many languages of type B, the terms "agglutinative," "fusional," and "symbolic" applying in their case merely to the treatment of the derivational, not the relational, concepts. Such languages could be termed "agglutinative-isolating," "fusional-isolating" and "symbolic-isolating."

This brings up the important general consideration that the method of handling one group of concepts need not in the least be identical with that used for another. Compound terms could be used to indicate this difference, if desired, the first element of the compound referring to the treatment of the concepts of group II, the second to that of the concepts of groups III and IV. An "agglutinative" language would normally be taken to mean one that agglutinates all of its affixed elements or that does so to a preponderating extent. In an "agglutinative-fusional" language the derivational elements are agglutinated, perhaps in the form of prefixes, while the relational elements (pure or mixed) are fused with the radical element, possibly as another set of prefixes following the first set or in the form of suffixes or as part prefixes and part suffixes. By a "fusional-agglutinative" language we would understand one that fuses its derivational elements but allows a greater independence to those that indicate relations. All these and similar distinctions are not merely theoretical possibilities, they can be abundantly illustrated from the descriptive facts of linguistic morphology. Further, should it prove desirable to insist on the degree of elaboration of the word, the terms "analytic," "synthetic," and "polysynthetic" can be added as descriptive terms. It goes without saying that languages of type A are necessarily analytic and that languages of type C also are prevailingly analytic and are not likely to develop beyond the synthetic stage.

But we must not make too much of terminology. Much depends on the relative emphasis laid on this or that feature or point of view. The method of classifying languages here developed has this great advantage, that it can be refined or simplified according to the needs of a particular discussion. The degree of synthesis may be entirely ignored; "fusion" and "symbolism" may often be combined with advantage under the head of "fusion"; even the difference between agglutination and fusion may, if desired, be set aside as either too difficult to draw or as irrelevant to the issue. Languages, after all, are exceedingly complex historical structures. It is of less importance to put each language in a neat pigeon-hole than to have evolved a flexible method which enables us to place it, from two or three independent standpoints, relatively to another language. All this is not to deny that certain linguistic types are more stable and frequently represented than others that are just as possible from a theoretical standpoint. But we are too ill-informed as yet of the structural spirit of great numbers of languages to have the right to frame a classification that is other than flexible and experimental.

The reader will gain a somewhat livelier idea of the possibilities of linguistic morphology by glancing down the subjoined analytical table of selected types. The columns II, III, IV refer to the groups of concepts so numbered in the preceding chapter. The letters _a_, _b_, _c_, _d_ refer respectively to the processes of isolation (position in the sentence), agglutination, fusion, and symbolism. Where more than one technique is employed, they are put in the order of their importance.[113]

[Footnote 113: In defining the type to which a language belongs one must be careful not to be misled by structural features which are mere survivals of an older stage, which have no productive life and do not enter into the unconscious patterning of the language. All languages are littered with such petrified bodies. The English _-ster_ of _spinster_ and _Webster_ is an old agentive suffix, but, as far as the feeling of the present English-speaking generation is concerned, it cannot be said to really exist at all; _spinster_ and _Webster_ have been completely disconnected from the etymological group of _spin_ and of _weave (web)_. Similarly, there are hosts of related words in Chinese which differ in the initial consonant, the vowel, the tone, or in the presence or absence of a final consonant. Even where the Chinaman feels the etymological relationship, as in certain cases he can hardly help doing, he can assign no particular function to the phonetic variation as such. Hence it forms no live feature of the language-mechanism and must be ignored in defining the general form of the language. The caution is all the more necessary, as it is precisely the foreigner, who approaches a new language with a certain prying inquisitiveness, that is most apt to see life in vestigial features which the native is either completely unaware of or feels merely as dead form.]

Note.--Parentheses indicate a weak development of the process in question.

+----------------+---+----+---+--------------+----------+--------------+ |Fundamental Type"II |III |IV |Technique "Synthesis "Examples | +----------------+---+----+---+--------------+----------+--------------+ | A " | | | " " | |(Simple Pure- "-- |-- |a |Isolating "Analytic "Chinese; | | relational) " | | | " "Annamite | | " | | | " " | | "(d)|-- |a,b|Isolating "Analytic "Ewe | | " | | |(weakly " "(Guinea Coast)| | " | | |agglutinative)" " | | " | | | " " | | "(b)|-- |a, |Agglutinative "Analytic "Modern Tibetan| | " | |b,c|(mildly " " | | " | | |agglutinative-" " | | " | | |fusional) " " | | " | | | " " | | B " | | | " " | |(Complex Pure- "b, |-- |a |Agglutinative-"Analytic "Polynesian | | relational) "(d)| | |isolating " " | | " | | | " " | | "b |-- |a, |Agglutinative-"Polysyn- "Haida | | " | |(b)|isolating "thetic " | | " | | | " " | | "c |-- |a |Fusional- "Analytic "Cambodgian | | " | | |isolating " " | | " | | | " " | | "b |-- |b |Agglutinative "Synthetic "Turkish | | " | | | " " | | "b,d|(b) |b |Agglutinative "Polysyn- "Yana (N. | | " | | |(symbolic "thetic "California) | | " | | |tinge) " " | | " | | | " " | | "c, |-- |a,b|Fusional- "Synthetic "Classical | | "d, | | |agglutinative "(mildly) "Tibetan | | "(b)| | |(symbolic " " | | " | | |tinge) " " | | " | | | " " | | "b |-- |c |Agglutinative-"Synthetic "Sioux | | " | | |fusional "(mildly " | | " | | | "polysyn- " | | " | | | "thetic) " | | " | | | " " | | "c |-- |c |Fusional "Synthetic "Salinan (S.W. | | " | | | " "California) | | " | | | " " | | "d,c|(d) |d, |Symbolic "Analytic "Shilluk | | " | |c,a| " "(Upper Nile) | | " | | | " " | | C " | | | " " | |(Simple Mixed- "(b)|b |-- |Agglutinative "Synthetic "Bantu | | relational) " | | | " " | | "(c)|c, |a |Fusional "Analytic "French[114] | | " |(d) | | "(mildly " | | " | | | "synthetic)" | | " | | | " " | | D " | | | " " | |(Complex Mixed- "b, |b |b |Agglutinative "Polysyn- "Nootka | | relational) "c,d| | | "thetic "(Vancouver | | " | | | "(symbolic "Island)[115] | | " | | | "tinge) " | | " | | | " " | | "c, |b |-- |Fusional- "Polysyn- "Chinook (lower| | "(d)| | |agglutinative "thetic "Columbia R.) | | " | | | "(mildly) " | | " | | | " " | | "c, |c, |-- |Fusional "Polysyn- "Algonkin | | "(d)|(d),| | "thetic " | | " |(b) | | " " | | " | | | " " | | "c |c,d |a |Fusional "Analytic "English | | " | | | " " | | "c,d|c,d |-- |Fusional "Synthetic "Latin, Greek, | | " | | |(symbolic " "Sanskrit | | " | | |tinge) " " | | " | | | " " | | "c, |c,d |(a)|Fusional "Synthetic "Takelma | | "b,d| | |(strongly " "(S.W. Oregon) | | " | | |symbolic) " " | | " | | | " " | | "d,c|c,d |(a)|Symbolic- "Synthetic "Semitic | | " | | |fusional " "(Arabic, | | " | | | " "Hebrew) | +----------------+---+----+---+--------------+----------+--------------+

[Footnote 114: Might nearly as well have come under D.]

[Footnote 115: Very nearly complex pure-relational.]

I need hardly point out that these examples are far from exhausting the possibilities of linguistic structure. Nor that the fact that two languages are similarly classified does not necessarily mean that they present a great similarity on the surface. We are here concerned with the most fundamental and generalized features of the spirit, the technique, and the degree of elaboration of a given language. Nevertheless, in numerous instances we may observe this highly suggestive and remarkable fact, that languages that fall into the same class have a way of paralleling each other in many details or in structural features not envisaged by the scheme of classification. Thus, a most interesting parallel could be drawn on structural lines between Takelma and Greek,[116] languages that are as geographically remote from each other and as unconnected in a historical sense as two languages selected at random can well be. Their similarity goes beyond the generalized facts registered in the table. It would almost seem that linguistic features that are easily thinkable apart from each other, that seem to have no necessary connection in theory, have nevertheless a tendency to cluster or to follow together in the wake of some deep, controlling impulse to form that dominates their drift. If, therefore, we can only be sure of the intuitive similarity of two given languages, of their possession of the same submerged form-feeling, we need not be too much surprised to find that they seek and avoid certain linguistic developments in common. We are at present very far from able to define just what these fundamental form intuitions are. We can only feel them rather vaguely at best and must content ourselves for the most part with noting their symptoms. These symptoms are being garnered in our descriptive and historical grammars of diverse languages. Some day, it may be, we shall be able to read from them the great underlying ground-plans.

[Footnote 116: Not Greek specifically, of course, but as a typical representative of Indo-European.]

Such a purely technical classification of languages as the current one into "isolating," "agglutinative," and "inflective" (read "fusional") cannot claim to have great value as an entering wedge into the discovery of the intuitional forms of language. I do not know whether the suggested classification into four conceptual groups is likely to drive deeper or not. My own feeling is that it does, but classifications, neat constructions of the speculative mind, are slippery things. They have to be tested at every possible opportunity before they have the right to cry for acceptance. Meanwhile we may take some encouragement from the application of a rather curious, yet simple, historical test. Languages are in constant process of change, but it is only reasonable to suppose that they tend to preserve longest what is most fundamental in their structure. Now if we take great groups of genetically related languages,[117] we find that as we pass from one to another or trace the course of their development we frequently encounter a gradual change of morphological type. This is not surprising, for there is no reason why a language should remain permanently true to its original form. It is interesting, however, to note that of the three intercrossing classifications represented in our table (conceptual type, technique, and degree of synthesis), it is the degree of synthesis that seems to change most readily, that the technique is modifiable but far less readily so, and that the conceptual type tends to persist the longest of all.

[Footnote 117: Such, in other words, as can be shown by documentary or comparative evidence to have been derived from a common source. See Chapter VII.]

The illustrative material gathered in the table is far too scanty to serve as a real basis of proof, but it is highly suggestive as far as it goes. The only changes of conceptual type within groups of related languages that are to be gleaned from the table are of B to A (Shilluk as contrasted with Ewe;[118] Classical Tibetan as contrasted with Modern Tibetan and Chinese) and of D to C (French as contrasted with Latin[119]). But types A : B and C : D are respectively related to each other as a simple and a complex form of a still more fundamental type (pure-relational, mixed-relational). Of a passage from a pure-relational to a mixed-relational type or _vice versa_ I can give no convincing examples.

[Footnote 118: These are far-eastern and far-western representatives of the "Soudan" group recently proposed by D. Westermann. The genetic relationship between Ewe and Shilluk is exceedingly remote at best.]

[Footnote 119: This case is doubtful at that. I have put French in C rather than in D with considerable misgivings. Everything depends on how one evaluates elements like _-al_ in _national_, _-té_ in _bonté_, or _re-_ in _retourner_. They are common enough, but are they as alive, as little petrified or bookish, as our English _-ness_ and _-ful_ and _un-_?]

The table shows clearly enough how little relative permanence there is in the technical features of language. That highly synthetic languages (Latin; Sanskrit) have frequently broken down into analytic forms (French; Bengali) or that agglutinative languages (Finnish) have in many instances gradually taken on "inflective" features are well-known facts, but the natural inference does not seem to have been often drawn that possibly the contrast between synthetic and analytic or agglutinative and "inflective" (fusional) is not so fundamental after all. Turning to the Indo-Chinese languages, we find that Chinese is as near to being a perfectly isolating language as any example we are likely to find, while Classical Tibetan has not only fusional but strong symbolic features (e.g., _g-tong-ba_ "to give," past _b-tang_, future _gtang_, imperative _thong_); but both are pure-relational languages. Ewe is either isolating or only barely agglutinative, while Shilluk, though soberly analytic, is one of the most definitely symbolic languages I know; both of these Soudanese languages are pure-relational. The relationship between Polynesian and Cambodgian is remote, though practically certain; while the latter has more markedly fusional features than the former,[120] both conform to the complex pure-relational type. Yana and Salinan are superficially very dissimilar languages. Yana is highly polysynthetic and quite typically agglutinative, Salinan is no more synthetic than and as irregularly and compactly fusional ("inflective") as Latin; both are pure-relational, Chinook and Takelma, remotely related languages of Oregon, have diverged very far from each other, not only as regards technique and synthesis in general but in almost all the details of their structure; both are complex mixed-relational languages, though in very different ways. Facts such as these seem to lend color to the suspicion that in the contrast of pure-relational and mixed-relational (or concrete-relational) we are confronted by something deeper, more far-reaching, than the contrast of isolating, agglutinative, and fusional.[121]

[Footnote 120: In spite of its more isolating cast.]

[Footnote 121: In a book of this sort it is naturally impossible to give an adequate idea of linguistic structure in its varying forms. Only a few schematic indications are possible. A separate volume would be needed to breathe life into the scheme. Such a volume would point out the salient structural characteristics of a number of languages, so selected as to give the reader an insight into the formal economy of strikingly divergent types.]

VII

LANGUAGE AS A HISTORICAL PRODUCT: DRIFT

Every one knows that language is variable. Two individuals of the same generation and locality, speaking precisely the same dialect and moving in the same social circles, are never absolutely at one in their speech habits. A minute investigation of the speech of each individual would reveal countless differences of detail--in choice of words, in sentence structure, in the relative frequency with which particular forms or combinations of words are used, in the pronunciation of particular vowels and consonants and of combinations of vowels and consonants, in all those features, such as speed, stress, and tone, that give life to spoken language. In a sense they speak slightly divergent dialects of the same language rather than identically the same language.

There is an important difference, however, between individual and dialectic variations. If we take two closely related dialects, say English as spoken by the "middle classes" of London and English as spoken by the average New Yorker, we observe that, however much the individual speakers in each city differ from each other, the body of Londoners forms a compact, relatively unified group in contrast to the body of New Yorkers. The individual variations are swamped in or absorbed by certain major agreements--say of pronunciation and vocabulary--which stand out very strongly when the language of the group as a whole is contrasted with that of the other group. This means that there is something like an ideal linguistic entity dominating the speech habits of the members of each group, that the sense of almost unlimited freedom which each individual feels in the use of his language is held in leash by a tacitly directing norm. One individual plays on the norm in a way peculiar to himself, the next individual is nearer the dead average in that particular respect in which the first speaker most characteristically departs from it but in turn diverges from the average in a way peculiar to himself, and so on. What keeps the individual's variations from rising to dialectic importance is not merely the fact that they are in any event of small moment--there are well-marked dialectic variations that are of no greater magnitude than individual variations within a dialect--it is chiefly that they are silently "corrected" or canceled by the consensus of usage. If all the speakers of a given dialect were arranged in order in accordance with the degree of their conformity to average usage, there is little doubt that they would constitute a very finely intergrading series clustered about a well-defined center or norm. The differences between any two neighboring speakers of the series[122] would be negligible for any but the most microscopic linguistic research. The differences between the outer-most members of the series are sure to be considerable, in all likelihood considerable enough to measure up to a true dialectic variation. What prevents us from saying that these untypical individuals speak distinct dialects is that their peculiarities, as a unified whole, are not referable to another norm than the norm of their own series.

[Footnote 122: In so far as they do not fall out of the normal speech group by reason of a marked speech defect or because they are isolated foreigners that have acquired the language late in life.]

If the speech of any member of the series could actually be made to fit into another dialect series,[123] we should have no true barriers between dialects (and languages) at all. We should merely have a continuous series of individual variations extending over the whole range of a historically unified linguistic area, and the cutting up of this large area (in some cases embracing parts of several continents) into distinct dialects and languages would be an essentially arbitrary proceeding with no warrant save that of practical convenience. But such a conception of the nature of dialectic variation does not correspond to the facts as we know them. Isolated individuals may be found who speak a compromise between two dialects of a language, and if their number and importance increases they may even end by creating a new dialectic norm of their own, a dialect in which the extreme peculiarities of the parent dialects are ironed out. In course of time the compromise dialect may absorb the parents, though more frequently these will tend to linger indefinitely as marginal forms of the enlarged dialect area. But such phenomena--and they are common enough in the history of language--are evidently quite secondary. They are closely linked with such social developments as the rise of nationality, the formation of literatures that aim to have more than a local appeal, the movement of rural populations into the cities, and all those other tendencies that break up the intense localism that unsophisticated man has always found natural.

[Footnote 123: Observe that we are speaking of an individual's speech as a whole. It is not a question of isolating some particular peculiarity of pronunciation or usage and noting its resemblance to or identity with a feature in another dialect.]

The explanation of primary dialectic differences is still to seek. It is evidently not enough to say that if a dialect or language is spoken in two distinct localities or by two distinct social strata it naturally takes on distinctive forms, which in time come to be divergent enough to deserve the name of dialects. This is certainly true as far as it goes. Dialects do belong, in the first instance, to very definitely circumscribed social groups, homogeneous enough to secure the common feeling and purpose needed to create a norm. But the embarrassing question immediately arises, If all the individual variations within a dialect are being constantly leveled out to the dialectic norm, if there is no appreciable tendency for the individual's peculiarities to initiate a dialectic schism, why should we have dialectic variations at all? Ought not the norm, wherever and whenever threatened, automatically to reassert itself? Ought not the individual variations of each locality, even in the absence of intercourse between them, to cancel out to the same accepted speech average?

If individual variations "on a flat" were the only kind of variability in language, I believe we should be at a loss to explain why and how dialects arise, why it is that a linguistic prototype gradually breaks up into a number of mutually unintelligible languages. But language is not merely something that is spread out in space, as it were--a series of reflections in individual minds of one and the same timeless picture. Language moves down time in a current of its own making. It has a drift. If there were no breaking up of a language into dialects, if each language continued as a firm, self-contained unity, it would still be constantly moving away from any assignable norm, developing new features unceasingly and gradually transforming itself into a language so different from its starting point as to be in effect a new language. Now dialects arise not because of the mere fact of individual variation but because two or more groups of individuals have become sufficiently disconnected to drift apart, or independently, instead of together. So long as they keep strictly together, no amount of individual variation would lead to the formation of dialects. In practice, of course, no language can be spread over a vast territory or even over a considerable area without showing dialectic variations, for it is impossible to keep a large population from segregating itself into local groups, the language of each of which tends to drift independently. Under cultural conditions such as apparently prevail to-day, conditions that fight localism at every turn, the tendency to dialectic cleavage is being constantly counteracted and in part "corrected" by the uniformizing factors already referred to. Yet even in so young a country as America the dialectic differences are not inconsiderable.

Under primitive conditions the political groups are small, the tendency to localism exceedingly strong. It is natural, therefore, that the languages of primitive folk or of non-urban populations in general are differentiated into a great number of dialects. There are parts of the globe where almost every village has its own dialect. The life of the geographically limited community is narrow and intense; its speech is correspondingly peculiar to itself. It is exceedingly doubtful if a language will ever be spoken over a wide area without multiplying itself dialectically. No sooner are the old dialects ironed out by compromises or ousted by the spread and influence of the one dialect which is culturally predominant when a new crop of dialects arises to undo the leveling work of the past. This is precisely what happened in Greece, for instance. In classical antiquity there were spoken a large number of local dialects, several of which are represented in the literature. As the cultural supremacy of Athens grew, its dialect, the Attic, spread at the expense of the rest, until, in the so-called Hellenistic period following the Macedonian conquest, the Attic dialect, in the vulgarized form known as the "Koine," became the standard speech of all Greece. But this linguistic uniformity[124] did not long continue. During the two millennia that separate the Greek of to-day from its classical prototype the Koine gradually split up into a number of dialects. Now Greece is as richly diversified in speech as in the time of Homer, though the present local dialects, aside from those of Attica itself, are not the lineal descendants of the old dialects of pre-Alexandrian days.[125] The experience of Greece is not exceptional. Old dialects are being continually wiped out only to make room for new ones. Languages can change at so many points of phonetics, morphology, and vocabulary that it is not surprising that once the linguistic community is broken it should slip off in different directions. It would be too much to expect a locally diversified language to develop along strictly parallel lines. If once the speech of a locality has begun to drift on its own account, it is practically certain to move further and further away from its linguistic fellows. Failing the retarding effect of dialectic interinfluences, which I have already touched upon, a group of dialects is bound to diverge on the whole, each from all of the others.

[Footnote 124: It is doubtful if we have the right to speak of linguistic uniformity even during the predominance of the Koine. It is hardly conceivable that when the various groups of non-Attic Greeks took on the Koine they did not at once tinge it with dialectic peculiarities induced by their previous speech habits.]

[Footnote 125: The Zaconic dialect of Lacedaemon is the sole exception. It is not derived from the Koine, but stems directly from the Doric dialect of Sparta.]

In course of time each dialect itself splits up into sub-dialects, which gradually take on the dignity of dialects proper while the primary dialects develop into mutually unintelligible languages. And so the budding process continues, until the divergences become so great that none but a linguistic student, armed with his documentary evidence and with his comparative or reconstructive method, would infer that the languages in question were genealogically related, represented independent lines of development, in other words, from a remote and common starting point. Yet it is as certain as any historical fact can be that languages so little resembling each other as Modern Irish, English, Italian, Greek, Russian, Armenian, Persian, and Bengali are but end-points in the present of drifts that converge to a meeting-point in the dim past. There is naturally no reason to believe that this earliest "Indo-European" (or "Aryan") prototype which we can in part reconstruct, in part but dimly guess at, is itself other than a single "dialect" of a group that has either become largely extinct or is now further represented by languages too divergent for us, with our limited means, to recognize as clear kin.[126]

[Footnote 126: Though indications are not lacking of what these remoter kin of the Indo-European languages may be. This is disputed ground, however, and hardly fit subject for a purely general study of speech.]

All languages that are known to be genetically related, i.e., to be divergent forms of a single prototype, may be considered as constituting a "linguistic stock." There is nothing final about a linguistic stock. When we set it up, we merely say, in effect, that thus far we can go and no farther. At any point in the progress of our researches an unexpected ray of light may reveal the "stock" as but a "dialect" of a larger group. The terms dialect, language, branch, stock--it goes without saying--are purely relative terms. They are convertible as our perspective widens or contracts.[127] It would be vain to speculate as to whether or not we shall ever be able to demonstrate that all languages stem from a common source. Of late years linguists have been able to make larger historical syntheses than were at one time deemed feasible, just as students of culture have been able to show historical connections between culture areas or institutions that were at one time believed to be totally isolated from each other. The human world is contracting not only prospectively but to the backward-probing eye of culture-history. Nevertheless we are as yet far from able to reduce the riot of spoken languages to a small number of "stocks." We must still operate with a quite considerable number of these stocks. Some of them, like Indo-European or Indo-Chinese, are spoken over tremendous reaches; others, like Basque,[128] have a curiously restricted range and are in all likelihood but dwindling remnants of groups that were at one time more widely distributed. As for the single or multiple origin of speech, it is likely enough that language as a human institution (or, if one prefers, as a human "faculty") developed but once in the history of the race, that all the complex history of language is a unique cultural event. Such a theory constructed "on general principles" is of no real interest, however, to linguistic science. What lies beyond the demonstrable must be left to the philosopher or the romancer.

[Footnote 127: "Dialect" in contrast to an accepted literary norm is a use of the term that we are not considering.]

[Footnote 128: Spoken in France and Spain in the region of the Pyrenees.]

We must return to the conception of "drift" in language. If the historical changes that take place in a language, if the vast accumulation of minute modifications which in time results in the complete remodeling of the language, are not in essence identical with the individual variations that we note on every hand about us, if these variations are born only to die without a trace, while the equally minute, or even minuter, changes that make up the drift are forever imprinted on the history of the language, are we not imputing to this history a certain mystical quality? Are we not giving language a power to change of its own accord over and above the involuntary tendency of individuals to vary the norm? And if this drift of language is not merely the familiar set of individual variations seen in vertical perspective, that is historically, instead of horizontally, that is in daily experience, what is it? Language exists only in so far as it is actually used--spoken and heard, written and read. What significant changes take place in it must exist, to begin with, as individual variations. This is perfectly true, and yet it by no means follows that the general drift of language can be understood[129] from an exhaustive descriptive study of these variations alone. They themselves are random phenomena,[130] like the waves of the sea, moving backward and forward in purposeless flux. The linguistic drift has direction. In other words, only those individual variations embody it or carry it which move in a certain direction, just as only certain wave movements in the bay outline the tide. The drift of a language is constituted by the unconscious selection on the part of its speakers of those individual variations that are cumulative in some special direction. This direction may be inferred, in the main, from the past history of the language. In the long run any new feature of the drift becomes part and parcel of the common, accepted speech, but for a long time it may exist as a mere tendency in the speech of a few, perhaps of a despised few. As we look about us and observe current usage, it is not likely to occur to us that our language has a "slope," that the changes of the next few centuries are in a sense prefigured in certain obscure tendencies of the present and that these changes, when consummated, will be seen to be but continuations of changes that have been already effected. We feel rather that our language is practically a fixed system and that what slight changes are destined to take place in it are as likely to move in one direction as another. The feeling is fallacious. Our very uncertainty as to the impending details of change makes the eventual consistency of their direction all the more impressive.

[Footnote 129: Or rather apprehended, for we do not, in sober fact, entirely understand it as yet.]

[Footnote 130: Not ultimately random, of course, only relatively so.]

Sometimes we can feel where the drift is taking us even while we struggle against it. Probably the majority of those who read these words feel that it is quite "incorrect" to say "Who did you see?" We readers of many books are still very careful to say "Whom did you see?" but we feel a little uncomfortable (uncomfortably proud, it may be) in the process. We are likely to avoid the locution altogether and to say "Who was it you saw?" conserving literary tradition (the "whom") with the dignity of silence.[131] The folk makes no apology. "Whom did you see?" might do for an epitaph, but "Who did you see?" is the natural form for an eager inquiry. It is of course the uncontrolled speech of the folk to which we must look for advance information as to the general linguistic movement. It is safe to prophesy that within a couple of hundred years from to-day not even the most learned jurist will be saying "Whom did you see?" By that time the "whom" will be as delightfully archaic as the Elizabethan "his" for "its."[132] No logical or historical argument will avail to save this hapless "whom." The demonstration "I: me = he: him = who: whom" will be convincing in theory and will go unheeded in practice.

[Footnote 131: In relative clauses too we tend to avoid the objective form of "who." Instead of "The man whom I saw" we are likely to say "The man that I saw" or "The man I saw."]

[Footnote 132: "Its" was at one time as impertinent a departure as the "who" of "Who did you see?" It forced itself into English because the old cleavage between masculine, feminine, and neuter was being slowly and powerfully supplemented by a new one between thing-class and animate-class. The latter classification proved too vital to allow usage to couple males and things ("his") as against females ("her"). The form "its" had to be created on the analogy of words like "man's," to satisfy the growing form feeling. The drift was strong enough to sanction a grammatical blunder.]

Even now we may go so far as to say that the majority of us are secretly wishing they could say "Who did you see?" It would be a weight off their unconscious minds if some divine authority, overruling the lifted finger of the pedagogue, gave them _carte blanche_. But we cannot too frankly anticipate the drift and maintain caste. We must affect ignorance of whither we are going and rest content with our mental conflict--uncomfortable conscious acceptance of the "whom," unconscious desire for the "who."[133] Meanwhile we indulge our sneaking desire for the forbidden locution by the use of the "who" in certain twilight cases in which we can cover up our fault by a bit of unconscious special pleading. Imagine that some one drops the remark when you are not listening attentively, "John Smith is coming to-night." You have not caught the name and ask, not "Whom did you say?" but "Who did you say?" There is likely to be a little hesitation in the choice of the form, but the precedent of usages like "Whom did you see?" will probably not seem quite strong enough to induce a "Whom did you say?" Not quite relevant enough, the grammarian may remark, for a sentence like "Who did you say?" is not strictly analogous to "Whom did you see?" or "Whom did you mean?" It is rather an abbreviated form of some such sentence as "Who, did you say, is coming to-night?" This is the special pleading that I have referred to, and it has a certain logic on its side. Yet the case is more hollow than the grammarian thinks it to be, for in reply to such a query as "You're a good hand at bridge, John, aren't you?" John, a little taken aback, might mutter "Did you say me?" hardly "Did you say I?" Yet the logic for the latter ("Did you say I was a good hand at bridge?") is evident. The real point is that there is not enough vitality in the "whom" to carry it over such little difficulties as a "me" can compass without a thought. The proportion "I : me = he : him = who : whom" is logically and historically sound, but psychologically shaky. "Whom did you see?" is correct, but there is something false about its correctness.

[Footnote 133: Psychoanalysts will recognize the mechanism. The mechanisms of "repression of impulse" and of its symptomatic symbolization can be illustrated in the most unexpected corners of individual and group psychology. A more general psychology than Freud's will eventually prove them to be as applicable to the groping for abstract form, the logical or esthetic ordering of experience, as to the life of the fundamental instincts.]

It is worth looking into the reason for our curious reluctance to use locutions involving the word "whom" particularly in its interrogative sense. The only distinctively objective forms which we still possess in English are _me_, _him_, _her_ (a little blurred because of its identity with the possessive _her_), _us_, _them_, and _whom_. In all other cases the objective has come to be identical with the subjective--that is, in outer form, for we are not now taking account of position in the sentence. We observe immediately in looking through the list of objective forms that _whom_ is psychologically isolated. _Me_, _him_, _her_, _us_, and _them_ form a solid, well-integrated group of objective personal pronouns parallel to the subjective series _I_, _he_, _she_, _we_, _they_. The forms _who_ and _whom_ are technically "pronouns" but they are not felt to be in the same box as the personal pronouns. _Whom_ has clearly a weak position, an exposed flank, for words of a feather tend to flock together, and if one strays behind, it is likely to incur danger of life. Now the other interrogative and relative pronouns (_which_, _what_, _that_), with which _whom_ should properly flock, do not distinguish the subjective and objective forms. It is psychologically unsound to draw the line of form cleavage between _whom_ and the personal pronouns on the one side, the remaining interrogative and relative pronouns on the other. The form groups should be symmetrically related to, if not identical with, the function groups. Had _which_, _what_, and _that_ objective forms parallel to _whom_, the position of this last would be more secure. As it is, there is something unesthetic about the word. It suggests a form pattern which is not filled out by its fellows. The only way to remedy the irregularity of form distribution is to abandon the _whom_ altogether for we have lost the power to create new objective forms and cannot remodel our _which_-_what_-_that_ group so as to make it parallel with the smaller group _who-whom_. Once this is done, _who_ joins its flock and our unconscious desire for form symmetry is satisfied. We do not secretly chafe at "Whom did you see?" without reason.[134]

[Footnote 134: Note that it is different with _whose_. This has not the support of analogous possessive forms in its own functional group, but the analogical power of the great body of possessives of nouns (_man's_, _boy's_) as well as of certain personal pronouns (_his_, _its_; as predicated possessive also _hers_, _yours_, _theirs_) is sufficient to give it vitality.]