Language and the Study of Language/Lecture VII
LECTURE VII.
The last two lectures have given us a view of the Indo-European family of languages. We have glanced at the principal dialects, ancient and modern, of which it is composed, noticing their exceeding variety and the high antiquity of some among them—the unequalled sweep, of time and of historic development together, which they include and cover. The family has been shown to be of preëminent importance and interest to the linguistic student, because the peoples to whom it belongs have taken during the past two thousand years or more a leading or even the foremost part in the world's history, because it includes the noblest and most perfect instruments of human thought and expression, and because upon its study is mainly founded the present science of language. We examined, in a general way, the method pursued in its investigation—namely, a genetic analysis, effected chiefly by the aid of a widely extended comparison of the kindred forms of related dialects (whence the science gets its familiar name of "comparative philology")—and noted briefly some of the misapprehensions and misapplications to which this was liable. At present, before going on to survey the other great families of language, and to consider the relation in which they stand to the Indo-European, we have to pause long enough to look at the main facts in the history of growth of the latter—of our own form of speech, using the word "our" in the widest sense to which we have as yet extended it. This we do, partly on account of the intrinsic interest of the subject, and partly because the results thus won will be found valuable, and even almost indispensable, in the course of our farther inquiries.
The history of Indo-European language has been more carefully read, and is now more thoroughly understood, than that of any other of the grand divisions of human speech. Not that our knowledge of it is by any means complete, or is not marked even by great and numerous deficiencies and obscurities: owing in no small part to the obliteration of needed evidence, and hence irreparable; but in part also to incomplete comparison and analysis of the material yet preserved, and therefore still admitting and sure ere long to receive amendment. Such deficiencies, however, are more concerned with matters of minor detail, and less with facts and principles of fundamental consequence, here than elsewhere. Hence the mode of development of language in general, even from its first commencement, can in no other way be so well exemplified as by tracing its special history in this single family.
Our first inquiry concerns the primitive stage of Indo-European language, its historical beginnings.
The general processes of linguistic growth and change, as they have for long ages past been going on in all the dialects of our kindred, were set forth and illustrated with some detail in the early part of our discussions respecting language (in the second and third lectures). We there saw that, in order to provide new thought and knowledge with its appropriate signs, and to repair the waste occasioned by the loss of words from use and memory, and the constant wearing out of forms, new combinations were made out of old materials, words of independent significance reduced to the position and value of modifying appendages to other words, and meanings variously altered and transferred. These processes may, for aught we can see, work on during an indefinite period in the future, with never-ending evolution out of each given form of speech of another slightly differing from it; even until every now existing dialect shall have divided into numerous descendants, and each of these shall have varied so far from its ancestor that their kindred shall be scarcely, or not at all, discoverable. Have we, now, any good reason to believe that they have not worked on thus indefinitely in the past also, with a kaleidoscopic resolution of old forms and combination of new, changing the aspect of language without altering its character as a structure? Or, are we able to find distinct traces of a condition of speech which may be called primitive in comparison with that in which it at present exists?
This question admits an affirmative answer. The present structure of language has its beginnings, from which we are not yet so far removed that they may not be clearly seen. Our historical analysis does not end at last in mere obscurity; it brings us to the recognition of elements which we must regard as, if not the actual first utterances of men, at least the germs out of which their later speech has been developed. It sets before our view a stage of expression essentially different from any of those we now behold among the branches of our family, and serving as their common foundation.
It must be premised that this belief rests entirely upon our faith in the actuality of our analytical processes, as being merely a retracing of the steps of a previous synthesis—in the universal truth of the doctrine that the elements into which we separate words are those by the putting together of which those words were at first made up. The grounds upon which such a faith reposes were pretty distinctly set forth in the second lecture (p. 66), but the importance of the subject will justify us in a recapitulation of the argument there presented.
No one can possibly suppose that we should ever have come to call our morning meal breakfast, if there had not already existed in our language the two independent words break and fast; any more than that we should say telegraph-wire, hickory-pole, campaign-document, gun-boat, without previous possession of the simple words of which are formed these modern compounds. Fearful and fearless, in like manner, imply the existence beforehand of the noun fear, and of the adjectives full and loose, or their older equivalents, which have assumed, with reference to that noun, the quality of suffixes. Nor should we have any adverbial suffix ly, if we had not earlier had the adjective like, nor any preterits in d (as I love-d), but for the fact that our Germanic ancestors owned an imperfect corresponding to our did, which they added to their new verbs to express past action. Any one, I think, will allow that elements distinguishable by word-analysis which can thus be identified with independent words are thereby proved to have been themselves once in possession of an independent status in the language, and to have been actually reduced by combination to the form and office with which our analysis finds them endowed. But farther, few or none will be found to question that all those formative elements which belong to the Germanic languages alone, of which no traces are to be discovered in any other of the branches of the Indo-European family, which constitute the peculiar patrimony of some or all of the dialects of our branch, must have been gained by the latter since their separation from the common stock, and in the same way with the rest, even though we can no longer demonstrate the origin of each affix. With the disguising and effacing effects of the processes of linguistic change fully present to our apprehensions, we shall not venture to conclude that those cases in which our historical researches fail to give us the genesis of both the elements of a compound form are fundamentally different from those in which it fully succeeds in doing so. The difference lies, not in the cases themselves, but in our attitude toward them; in our accidental possession of information as to the history of the one, and our lack of it as to that of the other. This reasoning, however, obviously applies not to Germanic speech alone; it is equally legitimate and cogent in reference to all Indo-European language. We cannot refuse to believe that the whole history of this family of languages has been, in its grand essential features, the same; that their structure is homogeneous throughout. There is no reason whatever for our assuming that the later composite forms are made up, and not the earlier; that the later suffixes are elaborated out of independent elements, and not the earlier. So far back as we can trace the history of language, the forces which have been efficient in producing its changes, and the general outlines of their modes of operation, have been the same; and we are justified in concluding, we are even compelled to infer, that they have been the same from the outset. There is no way of investigating the first hidden steps of any continuous historical process, except by carefully studying the later recorded steps, and cautiously applying the analogies thence deduced. So the geologist studies the forces which are now altering by slow degrees the form and aspect of the earth's crust, wearing down the rocks here, depositing beds of sand and pebbles there, pouring out floods of lava over certain regions, raising or lowering the line of coast along certain seas; and he applies the results of his observations with confidence to the explanation of phenomena dating from a time to which men's imaginations, even, can hardly reach. The legitimacy of the analogical reasoning is not less undeniable in the one case than in the other. You may as well try to persuade the student of the earth's structure that the coal-bearing rocks lie in parallel layers, of alternating materials, simply because it pleased God to make them so when he created the earth; or that the impressions of leaves, the stems and trunks of trees, the casts of animal remains, shells and bones, which they contain, the ripple and rain-marks which are seen upon them, are to be regarded as the sports of nature, mere arbitrary characteristics of the formation, uninterpretable as signs of its history—as to persuade the student of language that the indications of composition and growth which he discovers in the very oldest recorded speech, not less than in the latest, are only illusory, and that his comprehension of linguistic development must therefore be limited to the strictly historical period of the life of language. It is no prepossession, then, nor à priori theory, but a true logical necessity, a sound induction from observed facts, which brings us to the conclusion that all linguistic elements possessing distinct meaning and office, variously combined and employed for the uses of expression, are originally independent entities, having a separate existence before they entered into mutual combination.
In the light of these considerations let us examine a single word in our language, the word irrevocability. It comes to us from the Latin, where it had the form irrevocabilitas (genitive -tatis). It is clearly made by the addition of ty (tas, tatis) to a previously existing irrevocable (irrevocabili-s), just as we now form a new abstract noun from any given adjective by adding ness: for example, doughfacedness. Again, revocable (revocabilis) preceded irrevocable, as dutiful preceded undutiful. Further, if there had been no verb to revoke (revocare), there would have been no adjective revocable, any more than lovable without the verb to love. Yet once more: although we in English have the syllable voke only in composition with prefixes, as revoke, evoke, invoke, provoke, yet in Latin, as the verb vocare, 'to call,' it is, of course, older than any of these its derivatives, as stand is older than understand and withstand. Thus far our way is perfectly clear. But while, in our language, voke appears as a simple syllable, uncombined with suffixes, this is only by the comparatively recent effect of the wearing-out processes, formerly illustrated (in the third lecture); in the more original Latin, it is invariably associated with formative elements, which compose with its forms like vocare, 'to call,' vocat, 'he calls,' vocabar, 'I was called;' or, in substantive uses, vocs (vox), 'a calling, a voice,' vocum, 'of voices;' and so on. There is nothing, so far as concerns the formative elements themselves, to distinguish this last class of cases from the others, before analyzed; each suffix has its distinct meaning and office, and is applied in a whole class of analogous words; and some of them, at least, are traceable back to the independent words out of which they grew. The only difference is that here, if we cut off the formative elements, we have left, not a word, actually employed as such in any ancient language of our family, but a significant syllable, expressing the general and indeterminate idea of 'calling,' and found to occur in connected speech only when limited and defined by the suffixes which are attached to it. This is not, however, a peculiarity which can exempt the words so formed from a like treatment, leading to like conclusions, with the rest; we must still trust in the reality of our analysis; and especially, when we consider such forms as the Sanskrit vak-mi, vak-shi, vak-ti, where the mi, shi, and ti are recognizable pronouns, making compounds which mean clearly 'call-I,' 'call-thou,' 'call-he,' we cannot doubt that the element voc (vak) had also once an independent status, that it was a word, a part of spoken speech, and that the various forms which contain it were really produced by the addition of other elements to it, and their fusion together into a single word, in the same manner in which we have fused truth and full into truthful, truth and loose into truthless, true and like into truly.
The same conclusion may be stated in more general terms, as follows. The whole body of suffixes, of formative endings, is divided into two principal classes: first, primary, or such as form derivatives directly from roots; second, secondary, or such as form derivatives from other derivatives, from themes containing already a formative element. But the difference between these two classes is in their use and application, not in their character and origin. No insignificant portion of each is traceable back to independent words, and the presumption alike for each is that in all its parts it was produced in the same manner. If, then, we believe that the themes to which the secondary endings are appended were historical entities, words employed in actual speech before their further composition, we must believe the same respecting the roots to which are added the primary endings: these are not less historical than the others.
The conclusion is one of no small consequence. Elements like voc, each composing a single syllable, and containing no traceable sign of a formative element, resisting all our attempts at reduction to a simpler form, are what we arrive at as the final results of our analysis of the Indo-European vocabulary; every word of which this is made up—save those whose history is obscure, and cannot be read far back toward its beginning—is found to contain a monosyllabic root as its central significant portion, along with certain other accessory portions, syllables or remnants of syllables, whose office it is to define and direct the radical idea. The roots are never found in practical use in their naked form; they are (or, as has been repeatedly explained, have once been) always clothed with suffixes, or with suffixes and prefixes; yet they are no mere abstractions, dissected out by the grammarian's knife from the midst of organisms of which they were ultimate and integral portions; they are rather the nuclei of gradual accretions, parts about which other parts gathered to compose orderly and membered wholes; germs, we may call them, out of which has developed the intricate structure of later speech. And the recognition of them in this character is an acknowledgment that Indo-European language, with all its fulness and inflective suppleness, is descended from an original monosyllabic tongue; that our ancestors talked with one another in single syllables, indicative of the ideas of prime importance, but wanting all designation of their relations; and that out of these, by processes not differing in their nature from those which are still in operation in our own tongue, was elaborated the marvellous and varied structure of all the Indo-European dialects.
Such is, in fact, the belief which the students of language have reached, and now hold with full confidence. New and strange but a few years ago, it commands at present the assent of nearly all comparative philologists, and is fast becoming a matter of universal opinion. Since, however, it is still doubted and opposed by a few even among linguistic scholars, and is doubtless more or less unfamiliar and startling to a considerable part of any educated community, it will be proper that we combine with our examination of it some notice and refutation of the arguments by which it is assailed.
It is surely unnecessary, in the first place, to protest against any one's taking umbrage at this theory of a primitive monosyllabic stage of Indo-European language out of regard for the honour and dignity of our remote ancestors. The linguist is making a historical inquiry into the conditions of that branch of the human family to which we belong, and should no more be shocked at finding them talking in single syllables than dwelling in caves and huts of branches, or clad in leaves and skins. To require, indeed, for man's credit that he should have been sent upon the earth with a fully developed language miraculously placed in his mouth, with lists of nouns, verbs, and adverbs stored away in his memory, to be drawn upon at will, is not more reasonable than to require that the first human beings should have been born in full suits of clothes, and with neat cottages, not destitute of well-stocked larders, ready built over their heads. It surely is most of all to the honour of human nature that man should have been able, on so humble a foundation, to build up this wondrous fabric of speech; and also, as we may already say, that he should have been allowed to do so is more in accordance with the general plan of the Creator, who has endowed him with high capacities, and left him to work them out to their natural and intended results.
Nor, again, will any one venture to object that it would have been impossible to make so imperfect and rudimentary a language answer any tolerable purpose as a means of expression and communication—any one, at least, who knows aught of the present condition of language among the other races of the globe. One tongue, the Chinese—as we shall see more particularly farther on (in the ninth lecture)—has never advanced out of its primitive monosyllabic stage; its words remain even to the present day simple radical syllables, closely resembling the Indo-European roots, formless, not in themselves parts of speech, but made such only by their combination into sentences, where the connection and the evident requirements of the sense show in what signification and relation each is used. Yet this scanty and crippled language has served all the needs of a highly cultivated and literary people for thousands of years.
After these few words of reply to one or two of the difficulties which sometimes suggest themselves at first blush to those before whom is brought the view we are defending, we will next proceed to examine in more detail the original monosyllabism of Indo-European language, and see of what character it was.
The roots of our family of languages are divided into two distinct classes: those ultimately indicative of position merely, and those significant of action or quality. The former class are called demonstrative or pronominal roots; the latter class are styled predicative or verbal roots.
The pronominal roots are subjective in their character; they have nothing to do with the inherent qualities of objects, but mark them simply in their relation to the speaker, and primarily their local relation; they give the distinction between the this and the that, the nearer and the remoter object of attention, myself here, you there, and the third person or thing yonder, present or absent. By their nature, they are not severally and permanently attachable to certain objects or classes of objects, nor are they limited in their application; each of them may designate any and every thing, according to the varying relation sustained by the latter to the person or thing with reference to which it is contemplated. Only one thing can be called the sun; only certain objects are white; but there is nothing which may not be I, and you, and it, alternately, as the point from which it is viewed changes. In this universality of their application, as dependent upon relative situation merely, and in the consequent capacity of each of them to designate any object which has its own specific name besides, and so, in a manner, to stand for and represent that other name, lies the essential character of the pronouns.[1] From the pronominal roots come most directly the demonstrative pronouns, of which the personal are individualized forms, and the interrogatives; from these are developed secondarily the possessives and relatives, and the various other subordinate classes. They also generate adverbs of position and of direction. To examine in detail the forms they take, and the variations of the fundamental distinction between this and that which they are applied to express, would lead us too far. So much as this may be pointed out: those beginning with m are especially employed to denote the subject, the ego, 'me myself;' those with t and n are used more demonstratively, and those with k interrogatively. They are few in number, hardly counting a dozen all together, including some which are probably variants of the same original. They are of the simplest phonetic structure, consisting either of a pure vowel, like a or i, or of a vowel combined with a single preceding consonant, forming an open syllable, which is the easiest that the organs of articulation can be called upon to utter; instances are ma, na, ta, tu, ka.
The roots of the other class, those of action or quality, are very much more numerous, being reckoned by hundreds; and, they are of more complicated structure, illustrating every variety of the syllable, from the pure single vowel to the vowel preceded or followed, or both, by one consonant, or even by more than one. They are of objective import, designating the properties and activities inherent in natural objects—and prevailingly those that are of a sensible phenomenal character, such as modes of motion and physical exertion, of sound, and so forth. Let us notice a few instances of roots which are shown to have belonged to the original language of our family by being still met with in all or nearly all of its branches. Such are i and ga, denoting simple motion; ak, swift motion; stā, standing; ās and sad, sitting; kī, lying; pad, walking; vas, staying; sak, following; vart, turning; sarp, creeping; pat, flying; plu, flowing; ad, eating; pā, drinking; an, blowing; vid, seeing; klu, hearing; vak, speaking; dhā, putting; dā, giving; labh, taking; garbh, holding; dik, pointing out; bhar, bearing; kar, making; tan, stretching; skid and dal, dividing; bandh, binding; star, strewing; par, filling; mar, rubbing; bhā, shining; bhū, growing, etc., etc.
In endeavouring to apprehend the significance of these roots, we must divest their ideas of the definite forms of conception which we are accustomed to attach to them: each represents its own meaning in nakedness, in an indeterminate condition from which it is equally ready to take on the semblance of verb or of noun. We may rudely illustrate their quality by comparing them with such a word in our own language as love, which, by the wearing off of the formative elements with which it was once clothed, has reverted to the condition of a bare root, and which must therefore now be placed in such connection, or so pregnantly and significantly uttered, as to indicate to the intelligent and sympathizing listener in what sense it is meant and is to be understood—whether as verb, in "I love," or as substantive, in "my love," or as virtual adjective, in "love-letter."
The inquiry, which might naturally enough be raised at this point, how the radical syllables of which we are treating were themselves originated, and whether there be any natural and necessary connection between them, or any of them, and the ideas which they represent, such as either necessitated or at least recommended the allotment of the particular sign to the particular conception, we must pass by for the present, having now to do only with that for which direct evidence is to be found in language itself, with the historically traceable beginnings of Indo-European speech; this question, with its various dependent questions of a more theoretical and recondite nature, is reserved for consideration at a later time (in the eleventh lecture).
It deserves to be renewedly urged that, in this account of the primitive stage of Indo-European language, there is nothing which is not the result of strict and careful induction from the facts recorded in the dialects of the different members of the family. No one's theory as to what the beginnings of language must have been, or might naturally have been expected to be, has had anything to do with shaping it. It has been a matter of much controversy among linguistic theorizers what parts of speech language began with; whether nouns or verbs were the first words; but I am not aware that any acute thinker ever devised, upon à priori grounds, a theory at all closely agreeing with the account of the matter of which comparative philology soon arrived through her historical researches. That the first traceable linguistic entities are not names of concrete objects, but designate actions, motions, phenomenal conditions, is a truth resting on authority that overrides all preconceived theories and subjective opinions. How far and why it is accordant with what a sound theory, founded on our general knowledge of human nature and human speech, would teach, and is therefore entitled to be accepted as a satisfactory explanation of the way in which men began to talk, we shall inquire in the lecture devoted to such subjects.
Thus is it, also, as regards the division of the roots into two classes, pronominal and verbal: this division is so clearly read in the facts of language that its acceptance cannot be resisted. Some are loth to admit it, and strive to find a higher unity in which it shall disappear, the two classes falling together into one; or to show how the pronominal may be relics of verbal roots, worn down by linguistic usage to such brief form and unsubstantial significance; but their efforts must at least be accounted altogether unsuccessful hitherto, and it is very questionable whether they are called for, or likely ever to meet with success. As regards the purposes of our present inquiry, the double classification is certainly primitive and absolute; back to the very earliest period of which linguistic analysis gives us any knowledge, roots verbal and roots pronominal are to be recognized as of wholly independent substance, character, and office.
But, it may very properly be asked, how do we know that the roots which we have set up, and the others like them, are really ultimate and original? why may they not be the results of yet more ancient processes of linguistic change—like love and lie, and so many others, which have been repeatedly cited, and shown to have taken in our language the place of earlier complicated forms, such as lagamasi and laganti? how should they be proved different from our word count, for example, which we treat like an original root, expanding it by means of suffixes into various forms—as he counts, they counted, counting, counter, countable—while yet it is only a modern derivative from a Latin compound verb containing a preposition, namely computare, 'to think together, combine in thought,' got through the medium of the French compter (where the p is still written, though not pronounced)—in fact, the same word as the evidently made-up compute? Of apparent monosyllabic verbal roots like this, which are readily proved by a little historical study to be of polysyllabic origin, or to contain the relics of formative processes, our language contains no small number: other instances are preach from pre-dicare, vend from venum-dare, blame from Greek blas-phēmein; don and doff from do on and do off; learn, of which the n is a passive ending, added to lere, 'teach,' whence comes lore, 'doctrine;' to throng, a denominative from the noun throng, which is derived from thring (Anglo-Saxon thringan), 'press,' lost in our modern use (as if we were to lose sing, and substitute for it to song, from the derived noun song); to blast, a like denominative from blast, a derivative from blæsan, 'to blow, blare;' and so on. Such are to be found also abundantly in other languages, modern and ancient; why not as well among the alleged Indo-European roots? Now there can be no question whatever that such additions to the stock of verbal expression have been produced at every period of the growth of language, not only throughout its recorded career, but also in times beyond the reach of historic analysis. There is not a known dialect of our family which does not exhibit a greater or less number of seeming roots peculiar to itself; and of these the chief part may be proved, or are to be assumed, to be of secondary origin, and not at all entitled to lay claim to the character of relics from the original stock, lost by the sister dialects. Even the Sanskrit, upon which we have mainly to rely for our restoration of Indo-European roots, possesses not a few which are such only in seeming, which are of special Aryan or Indian growth, and valueless for the construction of general Indo-European etymologies. And, yet farther, among those very radical syllables whose presence in the tongues of all the branches proves them a possession of the original community before its dispersion, there are some which show the clearest signs of secondary formation. As a single example, let us take the root man, 'think' (in Latin me-min-i, moneo, mens; Greek men-os, man-tis; Lithuanian men-û; Mœso-Gothic man, German mein-en, our I mean): distinct analogies lead us to see in it a development—probably through a derivative noun, of which it is the denominative—of the older root mā, meaning either 'to make' or 'to measure;' a designation for the mental process having been won by figuratively regarding it as a mental manufacture or production, or else as an ideal mensuration of the object of thought, a passing from point to point of it, in estimation of its dimension and quality. Some linguistic scholars go much farther than others in their attempts at analyzing the Indo-European roots, and referring them to more primitive elements; all the methods of secondary origin which we have illustrated above have been sought for and thought to be recognized among them; and there are those who are unwilling to believe that any absolutely original root can have ended otherwise than in a vowel, or begun with more than a single consonant, and who therefore regard all radical syllables not conforming with their norm as the product of composition or fusion with formative elements. We need not here enter into the question as to the justice of these extreme views, or a criticism of the work of the root-analysts; we are compelled at any rate to concede that the results of growth are to be seen among even the earliest traceable historical roots; that we must be cautious how we claim ultimateness for any given radical syllable, unless we can succeed in establishing an ultimate and necessary tie between it and the idea it represents; and that the search after the absolutely original in human speech is a task of the most obscure and recondite character.
But these concessions do not impair our claim that the inflective structure of Indo-European speech is built up upon a historical foundation of monosyllabic roots. If the particular roots to which our analysis brings us are not in all cases the products of our ancestors' first attempts at articulation, they are at any rate of the same kind with these, and represent to us the incipient stage of speech. If in every dissyllable whose history we can trace we recognize a compound structure, if in every nominal and verbal form we find a formative element which gives it character as noun or verb, then we must believe that the germs out of which our language grew were not more complicated than single syllables, and that they possessed no distinct character as nouns or verbs, but were equally convertible into both. Our researches are only pointed a step farther back, without a change of method or result. That in these roots we approach very near to, if we do not quite touch, the actual beginnings of speech, is proved by other considerations. In order to bring into any language new apparent roots, and give them mobility by clothing them with inflections, a system of inflections must have been already elaborated by use with other roots in other forms. We cannot apply our d as sign of the imperfect tense to form such words as I electrified, I telegraphed, until we have worked down our preterit did, in substance and meaning, to such a mere formative element. And when we have traced the suffix back until we find it identical with the independent word out of which it grew, we know that we are close upon the beginning of its use, and have before us virtually that condition of the language in which its combinations were first made. So also with the adverbial suffix ly, when we have followed it up to lîce, a case of the adjective lîc, 'like.' Now, in connection with the roots of which examples have been given above, we see in actual process of elaboration the general system of Indo-European inflection, the most ancient, fundamental, and indispensable part of our grammatical apparatus; and we infer that these roots and their like are the foundation of our speech, the primitive material out of which its high and complicated fabric has been reared. It is not possible to regard them as the worn-down relics of a previous career of inflective development. The English, it is true, has been long tending, through the excessive prevalence of the wearing-out processes, toward a state of flectionless monosyllabism; but such a monosyllabism, where the grammatical categories are fully distinguished, where relational words and connectives abound, where every vocable inherits the character which the former possession of inflection has given it, where groups of related terms are applied to related uses, is a very different thing from a primitive monosyllabism like that to which the linguistic analyst is conducted by his researches among the earliest representatives of Indo-European language; and he finds no more difficulty in distinguishing the one from the other, and recognizing the true character of each, than does the geologist in distinguishing a primitive crystalline formation from a conglomerate, composed of well-worn pebbles, of diverse origin and composition, and containing fragments of earlier and later fossils. If the English were strictly reduced to its words of one syllable, it would still contain an abundant repertory of developed parts of speech, expressing every variety of idea, and illustrating a rich phonetic system. The Indo-European roots are not parts of speech, but of indeterminate character, ready to be shaped into nouns and verbs by the aid of affixes; they are limited in signification to a single class of ideas, the physical or sensual, the phenomenal, out of which the intellectual and moral develop themselves by still traceable processes; and in them is represented a system of articulated sounds of great simplicity. It will be not uninstructive to set down here, for comparison with the spoken alphabet of our modern English, already given (see p. 91), that scanty scheme of articulations, containing but three vowels and twelve consonants, which alone is discoverable in the earliest Indo-European language; it is as follows:
| a | Vowels. | ||
| i | u | ||
| l, r | Semivowel. | ||
| n | m | Nasals. | |
| h[2] | Aspiration. | ||
| s | Sibilant. | ||
| g | d | b | Mutes. |
| k | t | p |
The first beginning of polysyllabism seems to have been made by compounding together roots of the two classes already described, pronominal and verbal. Thus were produced true forms, in which the indeterminate radical idea received a definite significance and application. The addition, for example, to the verbal root vak, 'speaking,' of pronominal elements mi, si, ti (these are the earliest historically traceable forms of the endings: they were probably yet earlier ma, sa, ta), in which ideas of the nearer and remoter relation, of the first, second, and third persons, were already distinguished, produced combinations vakmi, vaksi, vakti, to which usage assigned the meaning 'I here speak,' 'thou there speakest,' 'he yonder speaks,' laying in them the idea of predication or assertion, the essential characteristic which makes a verb instead of a noun, just as we put the same into the ambiguous element, love, when we say I love. Other pronominal elements, mainly of compound form, indicating plurality of subject, made in like manner the three persons of the plural; they were masi (ma-si, 'I-thou', i.e. 'we'), tasi (ta-si, 'he-thou,' i.e. 'ye'), and anti (of more doubtful genesis). A dual number of the same three persons was likewise added; but the earliest form and derivation of its endings cannot be satisfactorily made out. Thus was produced the first verbal tense, the simplest and most immediate of all derivative forms from roots. The various shapes which its endings have assumed in the later languages of the family have already more than once been referred to, in the way of illustration of the processes of linguistic growth: our th or s, in he goeth or goes, still distinctly represents the ti of the third person singular; and in am we have a solitary relic of the mi of the first. Doubtless the tense was employed at the outset as general predicative form, being neither past, present, nor future, but all of them combined, and doing duty as either, according as circumstances required, and as sense and connection explained; destitute, in short, of any temporal or modal character; but other verbal forms by degrees grew out of it, or allied themselves with it, assuming the designation of other modifications of predicative meaning, and leaving to it the office of an indicative present. The prefixion of a pronominal adverb, a or ā, the so-called "augment," pointing to a 'there' or 'then' as one of the conditions of the action signified, produced a distinctively past or preterit tense. Although only very scanty and somewhat dubious traces of such an augment-preterit (aorist or imperfect) are found in any languages of the family beside the Aryan and the Greek, it is looked upon as an original formation, once shared by them all. Again, the repetition of the root, either complete, or by "reduplication," as we term it, the repetition of its initial part, was made to indicate symbolically the completion of the action signified by the root, and furnished another past tense, a perfect: for example, from the root dā, 'give,' Sanskrit dadāu, Greek dedōka, Latin dedi; from dhā, 'put, make,' Greek tetheika, Old High-German tēta, Anglo-Saxon dide, our did. This reduplicated perfect, as is well known, is a regular part of the scheme of Greek conjugation; in the Latin, not a few of the oldest verbs show the same, in full, or in more or less distinct traces; the Mœso-Gothic has preserved it in a considerable number of verbs (for example, in haihald, 'held,' from haldan, 'hold;' saislep, 'slept,' from slepan, 'sleep'); in the other Germanic dialects it is nearly confined to the single word did, already quoted. Moods were added by degrees: a conjunctive, having for its sign a union-vowel, a, interposed between root and endings, and bearing perhaps a symbolical meaning; and an optative, of which the sign is i or ia in the same position, best explained as a verbal root, meaning 'wish, desire.' From this optative descends the "subjunctive" of all the Germanic dialects. The earliest future appears to have been made by compounding with the root the already developed optative of the verb 'to be,' as-yâ-mi; for 'I shall call,' then, the language literally said 'I may be calling' (vak-s-yâ-mi). Of primitive growth, too, was a reflexive or "middle" voice, characterized by an extension of the personal endings, which is most plausibly explained as a repetition of them, once as subject and once as object: thus, vak-mai, for vak-ma-mi, 'call-I-me,' i.e. 'I call myself:' it was also soon employed in a passive sense, 'I am called'—as reflexives, of various age and form, have repeatedly been so employed, or have been converted into distinct passives, in the history of Indo-European language.[3] Other secondary forms of the verb, as intensives, desideratives, causatives, were created by various modifications of the root, or compositions with other roots; yet such verbal derivatives have played only a subordinate part in the development of the languages of our family, and need not be dwelt upon here. Of more consequence is the frequent formation of a special theme for the present tense, to which was then added a corresponding imperfect, made by means of the augment. This was accomplished in various ways: either by vowel-increment (as in Greek leipō, from lip, 'leave'), by reduplication (as in Greek dadāmi, from dā: the repetition of the root doubtless indicated repetition or continuity of the action), or by the addition or even insertion of formative elements (as in Greek deiknumi from dik, 'point out,' Sanskrit yunajmi from yuj, 'join;' Greek gignōscō, Latin gnosco, from gnā, 'know'); these last are, at least in part, noun-suffixes, and the forms they make are by origin denominatives.
Of this system of primitive verbal forms, produced before the separation of the family into branches, almost every branch has abandoned some part, while each has also new forms of its own to show, originated partly for supplying the place of that which was lost, partly in order to fill up the scheme to greater richness, and capacity of nicer and more varied expression. The Greek verb is, among them all, the most copious in its wealth, the most subtle and expressive in its distinctions: it has lost hardly anything that was original, and has created a host of new forms, some of which greatly tax the ingenuity of the comparative philologist who would explain their genesis. The Latin follows not very far behind, having made up its considerable losses, and supplied some new uses, by combinations of secondary growth: such are its imperfect in bam, its future in bo, and its derivative perfects in ui and si, in all of which are seen the results of composition with the roots of the substantive verb. Both these are greatly superior to the Sanskrit, in copiousness of forms, and in preciseness of their application. The Germanic verb was reduced at one period almost to the extreme of poverty, having saved only the ancient present, which was used also in the sense of a future, and a preterit, the modern representative of the original reduplicated perfect; each of the two tenses having also its subjunctive mood. The existing dialects of the branch have supplied a host of new expressions for tense and mood by the extensive employment of auxiliaries, which, in their way, afford an admirable analytic substitute for the old synthetic forms. To trace out and describe in full the history of the Indo-European verb, in these and in the other branches of the family, showing the contractions and expansions which it has undergone, down even to such recent additions as the future of the Romanic tongues, and our own preterit in d (the reason and method of whose creation have been explained above, in the third lecture), would be a most interesting and instructive task; but it is one which we may not venture here to undertake.
To follow back to its very beginnings the genesis of nouns, and of the forms of nouns, is much more difficult than to explain the origin of verbal forms. Some nouns—of which the Latin vox (voc-s), 'a calling, a voice,' and rex (reg-s), 'one ruling, a king,' are as familiar examples as any within our reach—are produced directly from the roots, by the addition of a different system of inflectional endings; the idea of substantiation or impersonation of the action expressed by the root being arbitrarily laid in them by usage, as was the idea of predication in the forms of the verb. The two words we have instanced may be taken as typical examples of the two classes of derivatives coming most immediately and naturally from the root: the one indicating the action itself, the other, either adjectively or substantively, the actor; the one being of the nature of an infinitive, or abstract verbal noun, the other of a participle, or verbal adjective, easily convertible into an appellative. Even such derivatives, however, as implying a greater modification of the radical idea than is exhibited by the simplest verbal forms, appear to have been from the first mainly made by means of formative elements, suffixes of derivation, comparable with those which belong to the moods and tenses, and the secondary conjugations of the verb. Precisely what these suffixes were, in their origin and primitive substance, and what were the steps of the process by which they lost their independence, and acquired their peculiar value as modifying elements, it is not in most cases feasible to tell. But they were obviously in great part of pronominal origin, and in the acts of linguistic usage which stamped upon them their distinctive value there is much which would seem abrupt, arbitrary, or even perhaps inconceivable, to one who has not been taught by extensive studies among various tongues how violent and seemingly far-fetched are the mutations and transfers to which the material of linguistic structure is often submitted—on how remote an analogy, how obscure a suggestion, a needed name or form is sometimes founded. Verbal roots, as well as pronominal, were certainly also pressed early into the same service: composition of root with root, of derived form with form, the formation of derivative from derivative, went on actively, producing in sufficient variety the means of limitation and individualization of the indeterminate radical idea, of its reduction to appellative condition, so as to be made capable of designating by suitable names the various beings, substances, acts, states, and qualities, observed both in the world of matter and in that of mind.
This class of derivatives from roots was provided with another, a movable, set of suffixes, which we call case-endings, terminations of declension. Where, as in the case of our two examples vox and rex, the theme of declension was coincident with the verbal root, the declensional endings themselves were sufficient to mark the distinction of noun from verb, without the aid of a suffix of derivation. They formed a large and complicated system, and were charged with the designation of various relations. In the first place, they indicated case, or the kind of relation sustained by the noun to`which they were appended to the principal action of the sentence in which it was used, whether as subject, as direct object, or as indirect object with implication of meanings which we express by means of prepositions, such as with, from, in, of. Of cases thus distinguished there were seven. Three of them distinctly indicated local relations: the ablative (of which the earliest traceable form has t or d for its ending: thus, Sanskrit açvāt, Old Latin equod, 'from a horse') denoted the relation expressed by from; the locative (with the ending i), that expressed by in; the instrumental (with the ending ā), that expressed by with, or by—the idea of adjacency or accompaniment passing naturally into that of means, instrument, or cause. Two cases, the dative and genitive, designated relations of a less physical character: the former (with the ending ai) we should render by for before the noun; the latter (its ending is asya or as) expressed general pertinence or possession. Then the accusative (with the sign m) assumed the office of indicating the directest dependent relation, that which even with us is expressed without the aid of a preposition—the objective—as well as that most immediate relation of motion which we signify by to. The nominative, finally, has also its ending, s, in the presence of which is strikingly exhibited the tendency of the earliest Indo-European language to make every vocable a true form, to give to every theme, in every relation, a sign of its mode of application, a formative element. Besides these seven proper cases, the vocative or interjectional case, the form of address, also makes a part of the scheme of declension; it has no distinctive ending, but is identical with the theme or the nominative case, or is only phonetically altered from them.
The declensional endings which we have instanced are those of the singular number. To explain their origin in any such way as shows us their precise value as independent elements, and the nature of the act of transfer by which they were made signs of case-relations, is not practicable. Pronominal elements are distinctly traceable in most of them, and may have assumed something of a prepositional force before their combination. The genitive affix is very likely to have been at the first, like many genitive affixes of later date in the history of the Indo-European languages, one properly forming a derivative adjective: and it is not impossible that the dative ending was of the same nature.
There are many existing tongues which have for the plurals of their nouns precisely the same case-endings as for the singular, only adding them along with a special pluralizing suffix. The attempt has been made[4] to find such a plural-suffix also among the plural endings of our earliest nouns, but with only faint and doubtful success; if these are actually of composite derivation, the marks of their composition are hidden almost beyond hope of discovery. We must be content to say for the present, at least, that the suffixes of declension indicate by their differences the distinctions of number as well as of case. And, among the nouns as well as the verbs of the primitive language, not only a plural, but also a dual, was distinguished from the singular by its appropriate endings, which are of not less problematical derivation, and, in the earliest condition of speech that we can trace, much fewer in number, being limited to three.
One other distinction, that of gender, was partially dependent for its designation upon the case-endings. We have already (in the third lecture) had occasion to refer to the universal classification of objects named, by the earliest language-makers of our family, according to gender, as masculine, feminine, or neuter—a classification only partially depending upon the actual possession of sexual qualities, and exhibiting, in the modern dialects which have retained it, an aspect of almost utter and hopeless arbitrariness. Nor, as was before remarked, is it possible even in the oldest Indo-European tongues to trace and point out otherwise than most dimly and imperfectly the analogies, apparent or fanciful, which have determined the grammatical gender of the different words and classes of words: such is the difficulty and obscurity of the subject that we must avoid here entering into any details respecting it. It appears that, in the first place, from the masculine, as the fundamental form, certain words were distinguished as possessed of feminine qualities, and marked by a difference of derivative ending, often consisting in a prolongation of the final vowel of the ending; while to all the derivatives formed by certain endings like qualities were attributed. The distinction was doubtless made in the beginning by the endings of derivation alone, those of case having no share in it; but it passed over to some extent into those of case also, the feminine here again showing a tendency to broader and fuller forms. The separation of neuter from masculine was both later in origin and less substantially marked, having little to do with suffixes of derivation, and extending through only a small part of the declensional endings (it is mainly limited to the nominative and accusative).
This system of Indo-European declension has suffered not less change in the history of the various branches of the family than has that of conjugational inflection. The dual number was long ago given up, as of insignificant practical value, by most of the branches: the oldest Aryan dialects exhibit it most fully; it also makes some figure in ancient Greek; but even the most antique Germanic tongues have a dual only in the personal pronouns of the first and second persons; and the Latin shows but the faintest traces of it (in the peculiar nominative and accusative endings of duo, 'two,' and ambo, 'both'). As regards, again, the cases, the complete scheme only appears in the Indian and Persian; and even there the process of its reduction has begun, by the fusion, in one or another number, and in one or another class of words, of two cases into one—that is to say, the loss of the one as a distinct form, and the transference of its functions to another. In the oldest known condition of the classic tongues, this process has gone yet farther; in Latin, the locative and instrumental are thus fused with the dative and ablative; and in Greek, the genitive and ablative have been also compressed into one. The oldest Germanic dialects have nominative, accusative, genitive, and dative; with traces of the instrumental, which the later tongues have lost. But the modern development of the prepositions, and their rise to importance as independent indicators of the relations formerly expressed by the case-endings, has brought with it a yet more sweeping abandonment of the latter. We, in English, have saved a single oblique case, the ancient genitive, so restricting its use at the same time as to make a simple "possessive" of it—and further, among the pronouns, an accusative or "objective" (me, us, etc., and whom); in the Romanic languages, the noun has become wholly stripped of case-inflection. In what manner we have rid ourselves of the distinctions of grammatical gender has been shown in a previous lecture (the third); we still keep up a linguistic distinction of natural gender by the use of our generic pronouns of the third person, he, she, and it; the modern Persian has abandoned even that, and the consideration of sex no longer enters into it in any way, save in the vocabulary, in the use of such words as son and daughter, bull and cow. Of the other modern tongues of the family, some, like these two, have eliminated from their grammatical systems the distinctions of gender; some, like the French, have reduced the three genders to two, by effacing the differences of masculine and neuter; but the larger part, like the German, still faithfully adhere to the inherited distinction of masculine, feminine, and neuter, so long ago established.
The ancient Indo-European language made no difference, as regarded declension, between its two classes of nouns, nouns substantive and nouns adjective. In their genesis, the two are but one; the same suffixes, to no small extent, form both; each passes by the most easy and natural transfer into the other; whether a given word indicating the possession of quality should be used attributively or predicatively, or as an appellative, was a question of subordinate consequence. The pronouns, also, both substantive and adjective, were inflected by a declension mainly corresponding, although marked by some peculiarities, and tending earlier to irregular forms.
With conjugation and declension, the subject of grammatical structure is, in fact, as good as exhausted: everything in language is originally either verb or noun. To the other parts of speech, then, which have been developed out of these, we shall need to give but a brief consideration.
Adverbs, the most ancient and necessary class of indeclinable words, or particles, are by origin, in the earliest stage of language as in the latest, forms of declension, cases of substantives, or adjectives, or pronouns. We have seen already how our adverbs in ly were elaborated out of former oblique cases (instrumentals) of adjectives in lîc ('like'); so also the usual adverbial ending ment of the Romanic languages is the Latin ablative mente, 'with mind' (thus, French bonnement, 'kindly,' is bonâ mente, 'with kind intent'); the ōs which forms Greek adverbs (for example, kakōs, 'ill,' from kakos, 'bad') is the original ablative case-ending: and we are doubtless to infer that both the general classes of adverbs, made by means of apparent adverbial suffixes, and the more irregular and obscure single words, of kindred meaning and office, which we trace in the earliest vocabulary of the family, are of like derivation. Those parts of speech which we call prepositions were originally such, not in our present understanding of the term, but according to its etymological signification; they were adverbial prefixes to the verb, serving to point out more clearly the direction of the verbal action; it was only later, and by degrees, that they detached themselves from the verb, and came to belong to the noun, furthering the disappearance of its case-endings, and assuming their office. The earliest of them, as was to be expected from their designation of direction, trace their origin chiefly to pronominal roots; but in part, also, they come from verbal. Conjunctions, connectives of sentences, are almost altogether of comparatively late growth; the earliest style was too simple to call for their use: we have seen examples already (in the third lecture) of the mode in which they were arrived at, by attenuation of the meaning of words possessing by origin a more full and definite significance. Other products of a like attenuation, made generally at a decidedly modern date, are the articles: the definite article always growing out of a demonstrative pronoun; the indefinite, out of the numeral one.
The interjections, finally, however expressive and pregnant with meaning they may be, are not in a proper sense parts of speech; they do not connect themselves with other words, and enter into the construction of sentences; they are either the direct outbursts of feeling, like oh! ah! or else, like st! sh! mere "vocal gestures," immediate intimations of will—in both cases alike, substitutes for more elaborate and distinct expression. They require, however, to be referred to here, not merely for the sake of completeness, but also because many words come to be employed only interjectionally which were once full parts of speech; even a whole phrase being, as it were, reduced to a single pregnantly uttered exclamation: examples are alas! that is, O me lasso, 'oh weary me!' zounds! 'I swear by God's wounds,' dear me! that is, dio mio, 'my God!' and many others.
Such are, compendiously and briefly stated, the steps by which Indo-European language was developed out of monosyllabic weakness into the wealth and fertility of inflective speech. At what rate they went on, how rapid was the growth after its first inception, we know not, and we can hardly hope ever to know. The conditions of that primitive period, and the degree in which they might have been able to quicken the now sluggish processes of word-combination and formation, are so much beyond our ken that even our conjectures respecting them have—at least as yet—too little value to be worth recording. What may have been the numbers of the community which laid the foundation of all the Indo-European tongues, and what its relation to other then existing communities, are also points hitherto involved in the deepest obscurity. But we know that, before the separation, whether simultaneous or successive, of this community into the parts which afterward became founders of the different tongues of Europe and south-western Asia, the principal part of the linguistic development had already taken place—enough for its traces to remain ineffaceable, even to the present day, in the speech of all the modern representatives of the family; the inflective character of Indo-European language, the main distinctions of its parts of speech, its methods of word-formation and inflection, were elaborated and definitely established.
But, though we cannot pretend to fix the length of time required for this process of growth, in terms of centuries or of thousands of years, we can at least see clearly that it must have gone on in a slow and gradual manner, and occupied no brief period. Such is the nature of the forces by which all change in language has been shown to be effected, that anything like a linguistic revolution, a rapid and sweeping modification of linguistic structure, is wholly impossible—and most especially, a revolution of a constructive character, building up a fabric of words and forms. Every item of the difference by which a given dialect is distinguished from its ancestor, or from another dialect having the same ancestry, is the work of a gradual change of usage made by the members of a community in the speech which they were every day employing as their means of mutual communication, and which, if too rapidly altered, would not answer the purposes of communication. It takes time for even that easiest of changes, a phonetic corruption or abbreviation, to win the assent of a community, and become established as the law of their speech: it takes decades, and even generations, or centuries, for an independent word to run through the series of modifications in form and meaning which are necessary to its conversion into a formative element. That the case was otherwise at the very beginning, we have not the least reason for believing, The opinion of those who hold that the whole structure of a language was produced "at a single stroke" is absolutely opposed to all the known facts of linguistic history; it has no inductive basis whatever; it rests upon arbitrary assumption, and is supported by à priori reasoning. There must have been a period of some duration—and, for aught we know, it may have been of very long duration—when the first speakers of our language talked together in their scanty dialect of formless monosyllables. The first forms, developed words containing a formal as well as a radical element, cannot have come into existence otherwise than by slow degrees, worked out by the unconscious exercise of that ingenuity in the adaptation of means to ends, of that sense for symmetry, for finished, even artistic, production, which have ever been qualities especially characterizing our division of the human race. Every form thus elaborated led the way to others: it helped to determine a tendency, to establish an analogy, which facilitated their further production. A protracted career of formal development was run during that primitive period of Indo-European history which preceded the dispersion of the branches: words and forms were multiplied until even a maximum of synthetic complexity, of fullness of inflective wealth, had been reached, from which there has been in later times, upon the whole, a gradual descent and impoverishment.
Here we must pause a little, to consider an objection urged by some linguistic scholars of rank and reputation against the truth of the views we have been defending, as to the primitive monosyllabism of Indo-European language, and its gradual emergence out of that condition—an objection which has more apparent legitimacy and force than any of those hitherto noticed. It is this. In ascending the current of historical development of the languages of our family, say the objectors, instead of approaching a monosyllabic condition, we seem to recede farther and farther from it. The older dialects are more polysyllabic than the later: where our ancestors used long and complicated forms, we are content with brief ones, or we have replaced them with phrases composed of independent words. Thus, to recur once more to a former example, for an earlier lagamasi we say we lie; thus, again, for the Latin fuisset, the French says simply fût, while we express its meaning by four distinct words, he might have been. Modern languages are full of verbal forms of this latter class, which substitute syntactical for substantial combinations. The relations of case, too, formerly signified only by means of declensional endings, have lost by degrees this mode of expression, and have come to be indicated by prepositions, independent words. This is what is well known as the "analytical" tendency in linguistic growth. Our own English tongue exhibits its effects in the highest known degree, having reduced near half the vocabulary it possesses to a monosyllabic form, and got rid of almost all its inflections, so that it expresses grammatical relations chiefly by relational words, auxiliaries and connectives: but it is only an extreme example of the results of a movement generally perceptible in modern speech. If, then, during the period when we can watch their growth step by step, languages have become less synthetic, words less polysyllabic, must we not suppose that it was always so; that human speech began with highly complicated forms, which from the very first have been undergoing reduction to simpler and briefer shape?
This is, as we have confessed, a plausible argument, but it is at the same time a thoroughly unsound and superficial one. It skims the surface of linguistic phenomona, without penetrating to the causes which produce them. It might pass muster, and be allowed to determine our opinions, if the analytical tendency alone had been active since our knowledge of language began; if we had seen old forms worn out, but no new forms made; if we had seen words put side by side to furnish analytic combinations, but no elements fused together into synthetic union. But we know by actual experience how both synthetic and analytic forms are produced, and what are the influences and circumstances which favour the production of the one rather than of the other. The constructive as well as the destructive forces in language admit of illustration, and have been by us illustrated, with modern as well as with ancient examples. Both have been active together, during all the ages through which we can follow linguistic growth. There have never been forms which were not undergoing continual modification and mutilation, under the influence of the already recognized tendencies to forget the genesis of a word in its later application, and then to reduce it to a shape adapted to more convenient utterance; there was also never a time when reparation was not making for this waste in part by the fresh development of true forms out of old materials. Nor has the tendency been everywhere and in all respects downward, toward poverty of synthetic forms, throughout the historic period. If the Greek and Latin system of declension is scantier than that of the original language of the family, their system of conjugation, especially the Greek, is decidedly richer, filled up with synthetic forms of secondary growth; the modern Romanic tongues have lost something of this wealth, but they have also added something to it, and their verb, leaving out of view its compound tenses, will bear favourable comparison with that which was the common inheritance of the branches. Some of the modern dialects of India, on the other hand, having once lost, in the ordinary course of phonetic corruption, the ancient case-terminations of the Sanskrit, have replaced them by a new scheme, not less full and complete than its predecessor. The Russian of the present day possesses in some respects a capacity of synthetic development hardly, if at all, excelled by that of any ancient tongue. For example, it takes the two independent words bez Boga, 'without God,' and fuses them into a theme from which it draws a whole list of derivatives. Thus, first, by adding an adjective suffix, it gets the adjective bezbozhnüǐ, 'godless;' a new suffix appended to this makes a noun, bezbozhnik, 'a godless person, an atheist;' the noun gives birth to a denominative verb, bezbozhnikhat, 'to be an atheist;' from this verb, again, come a number of derivatives, giving to the verbal idea the form of adjective, agent, act, and so on: the abstract is bezbozhnichestvo, 'the condition of being an atheist;' while, once more, a new verb is made from this abstract, namely bezbozhnichestvovat, literally 'to be in the condition of being a godless person.' A more intricate synthetic form than this could not easily be found in Greek, Latin, or Sanskrit; but it is no rare or exceptional case in the language from which we have extracted it; it rather represents, by a striking instance, the general character of Russian word-formation and derivation.
It is obviously futile, then, to talk of an uninterrupted and universal reduction of the resources of synthetic expression among the languages of the Indo-European family, or to allow ourselves to be forced by an alleged pervading tendency toward analytic forms into accepting synthesis, inflective richness, as the ultimate condition of the primitive tongue from which they are descended. If certain among them have replaced one or another part of their synthetic structure by analytic forms, if some—as the Germanic family in general, and, above all, the English—have taken on a prevailingly analytic character, these are facts which we are to seek to explain by a careful study of the circumstances and tendencies which have governed their respective development. If, moreover, as has been conceded, the general bent has for a long time been toward a diminution of synthesis and a predominance of analytic expressions, another question, of wider scope, is presented us for solution; but the form in which it offers itself is this: why should the forces which produce synthetic combinations have reached their height of activity during the ante-historic period of growth, and have been gradually gained upon later, at varying rates in different communities, by those of another order? We do not in the least feel impelled to doubt the historic reality of the earliest combinations, their parallelism, in character and origin, with those which we see springing up in modern times. That we now say analytically I did love, or deal, or lead is no ground for questioning that our ancestors said compositely I love-did, deal-did, lead-did, and then worked them down into the true synthetic forms I loved, dealt, led. The cause which produced the different nature of the two equivalent expressions I loved and I did love, composed, as they are, of identical elements, was a difference in habit of the language at the periods when they were respectively generated. Any language can do what it is in the habit of doing. We can turn almost any substantive in our vocabulary into a quasi adjective—saying a gold watch, a grass slope, a church mouse, and so on—because, through the intermediate step of loose compounds like goldsmith, grasshopper, churchman, we have acquired the habit of looking upon our substantives as convertible to adjective uses without alteration and without ceremony. Neither the Frenchman nor the German can do the same thing, simply because his speech presents no analogies for such a procedure. We, on the other hand, like the French, have lost the power to form compounds with anything like the facility possessed by the ancient tongue from which ours is descended and by some of its modern representatives, as the German; not because they would not be intelligible if we formed them, but because, under the operation of traceable circumstances in our linguistic history, we have grown out of the habit of so combining our words, and into the habit of merely collocating them, with or without connectives. Now we have only to apply this principle upon a wider scale, and under other conditions of language, in order to find, as I think, a sufficient answer to the question which is engaging our attention. When once, after we know not how long a period of expectation and tentative effort, the formation of words by synthesis had begun in the primitive Indo-European language, and had been found so fruitful of the means of varied and distinct expression, it became the habit of the language. The more numerous the new forms thus produced, the greater was the facility of producing more, because the material of speech was present to the minds of its speakers as endowed with that capacity of combination and fusion of which the results in every part of its structure were so apparent. But the edifice after a time became, as it were, complete; a sufficient working-apparatus of declensional, conjugational, and derivative endings was elaborated to answer the purposes of an inflective tongue; fewer and rarer additions were called for, as occasional supplements of the scheme, or substitutes for lost forms. Thus began a period in which the formative processes were more and more exclusively an inheritance from the past, less and less of recent acquisition; and as the origin of forms was lost sight of, obscured by the altering processes of phonetic corruption, it became more and more difficult to originate new ones, because fewer analogies of such forms were present to the apprehension of the language-makers, as incentives and guides to their action. On the other hand, the expansion of the whole vocabulary to wealth of resources, to the possession of varied and precise phraseology, furnished a notably increased facility of indicating ideas and relations by descriptive phrases, by groups of independent words. This mode of expression, then, always more or less used along with the other, began to gain ground upon it, and, of course, helped to deaden the vitality of the latter, and to render it yet more incapable of extended action. That tendency to the conscious and reflective use of speech which comes in with the growth of culture especially, and which has already been repeatedly pointed out as one of the main checks upon all the processes of linguistic change, cast its influence in the same direction; since the ability to change the meaning and application of words, even to the degree of reducing them to the expression of formal relations, is a much more fundamental and indefeasible property of speech than the ability to combine and fuse them bodily together. Then, when peculiar circumstances in the history of a language have arisen, to cause the rapid and general decay and effacement of ancient forms, as in our language and the Romanic, the process of formative composition; though never wholly extinct, has been found too inactive to repair the losses; they have been made up by syntactical collocation, and the language has taken on a prevailingly analytic character.
These considerations and such as these, I am persuaded, furnish a satisfactory explanation of the preponderating tendency to the use of analytic forms exhibited by modern languages; as they also account for the greatly varying degree in which the tendency exhibits itself. But even should they be found insufficient, this would only throw open for a renewed investigation the question respecting the ground of the tendency; the general facts in the history of earliest development of our languages would still remain sure, beyond the reach of cavil, since they are established by evidence which cannot be gainsaid, contained in the structure of the most ancient forms. We are compelled to believe that the formative processes which we see going on, in decreasing abundance, in the historically recorded ages of linguistic life, are continuations and repetitions of the same constructive acts by which has been built up the whole homogeneous structure of inflective speech.
One more theoretic objection to the doctrine of a primitive Indo-European monosyllabism we may take the time to notice, more on account of the respectability of its source than for any cogency which it in itself possesses. M. Renan, namely,[5] asserts that this doctrine is the product of a mistaken habit of mind, taught us by the artificial scholastic methods of philosophizing, and leading us to regard simplicity as, in the order of time, anterior to complexity; while, in fact, the human mind does not begin with analysis; its first acts being, on the contrary, complex, obscure, synthetic, containing all the parts, indistinctly heaped together. To this claim respecting the character of the mental act we may safely yield a hearty assent; but, instead of inferring from it that "the idea expressed itself at the beginning with its whole array of determinatives and in a perfect unity," and that hence, "in the history of languages, synthesis is primitive, and analysis, far from being the natural form of the human mind, is only the low result of its development," we shall be conducted to a precisely contrary conclusion. The synthetic forms which we are asked to regard as original have not the character of something indistinctly heaped together; they contain the clear and express designation of the radical idea and of its important relations; they represent by a linguistic synthesis the results of a mental analysis. The idea is, indeed, conceived in unity, involving all its aspects and relations; but these cannot be separately expressed until the mind has separated them, until practice in the use of language has enabled it to distinguish them, and to mark each by an appropriate sign. In amabor, the (Latin) word cited as an example of synthesis, are contained precisely the same designations as in the equivalent English analytic phrase, 'I shall be loved:' ama expresses 'loving;' bo unites future-sign and ending designating the first person; and the r is the sign of passivity. Who can possibly maintain that a system of such forms, gathered about a root, exhibits the results of experience, of developed acuteness, in thought and speech, any less clearly than the analytic forms of our English conjugation? The two are only different methods of expressing the same "array of determinatives." The first synthetic mental act, on the contrary, is truly represented by the bare root: there all is, indeed, confused and indiscrete. The earliest radical words, when first uttered, stood for entire sentences, expressed judgments, as undeniably as the fully elaborated phrases which we now employ, giving every necessary relation its proper designation. It is thus that, even at present, children begin to talk; a radical word or two means in their mouths a whole sentence: up signifies 'take me up into your lap;' go walk, 'I want to go out to walk,' or 'I went to walk,' or various other things, which the circumstances sufficiently explain; but forms, inflections, connectives, signs of tense and mode and condition, they do not learn to use until later, when their minds have acquired power to separate the indistinct cognition into its parts. M. Renan, in short, has made a very strange confusion of analytic style of expression with mental analysis: all expression of relations, whether by means that we call synthetic or analytic, is the result and evidence of analysis; and his own thesis respecting the complexity in obscurity of unpractised and uninstructed thought, brings us directly to a recognition of the radical stage of Indo-European language as the necessary historical basis of its inflective development.
This development, it may be remarked in conclusion, has been gradual and steadily progressive, being governed in both its synthetic and analytic phases by the same causes which universally regulate linguistic growth, and which have been here repeatedly set forth or referred to: namely, on the one hand, the traditional influence of the stores of expression already worked out and handed down, consisting in the education given by them to thought, and the constraining force exerted by their analogies; and, on the other hand, the changing character and capacity, the varying circumstances and needs, of the community of speakers, during the different periods of their history. It has experienced no grand revolution, no sudden shift of direction, no pervading change of tendency. There is no cleft, as is sometimes assumed, parting ancient tongues from modern, justifying the recognition of different forces, the admission of different possibilities, in the one and in the other. Nor are we to regard the energies of a community as absorbed in the work of language-making more at one period than at another. Language-making is always done unconsciously and by the way, as it were: it is one of the incidents of social life, an accompaniment and result of intellectual activity, not an end toward which effort is directed, nor a task in whose performance is expended force which might have been otherwise employed. The doctrine that a race first constructs its language, and then, and not till then, is ready to commence its historic career, is as purely fanciful as anything in the whole great chapter of à priori theorizings about speech. No living language ever ceases to be constructed, or is less rapidly built upon in ages of historic activity: only the style of the fabric is, even more than the rate, determined by external circumstances. It is because the very earliest epoch of recorded history are still far distant from the beginnings of Indo-European language, as of human language generally, that we find its peculiar structure completely developed when it is first discovered by our researches. We have fully acknowledged the powerful influence exerted by culture over the growth of language: but neither the accident of position and accessibility to other nations that at a certain time brings a race forward into the light of record, and makes it begin to be an actor or a factor in the historic drama, nor its more gradual and independent advance to conspicuousness in virtue of acquired civilization and political power, can have any direct effect whatever upon its speech. The more thorough we are in our study of the living and recent form of human language, the more rigorous in applying the deductions thence drawn to the forms current in ante-historic periods, the more cautious about admitting forces and effect in unknown ages whereof the known afford us no example or criterion, so much the more sound and trustworthy will be the conclusions at which we shall arrive. It is but a shallow philology, as it is a shallow geology, which explains past changes by catastrophes and cataclysms.
We have now long enough given our almost exclusive attention to the language of the Indo-European race, and, in the next two lectures, shall proceed to define the boundaries and sketch the characters, as well as we may, of the other grand divisions of human speech.
Notes[edit]
- ↑ Their Hindu title, sarvanâman, 'name for everything, universal designation,' is therefore more directly and fundamentally characteristic than the one we give them, pronoun, 'standing for a name.'
- ↑ The aspiration is not found as a separate letter, but only in close combination with the mutes, forming the aspirated mutes gh, dh, bh, and (probably by later development) kh, th, ph. These aspirates, though historically they are independent and important members of the system of spoken sounds, I have not given separately in the scheme, because phonetically they are compound, containing the aspiration as a distinctly audible element following the mute.
- ↑ The Latin passive, for instance, is of reflexive origin, as is that of the Scandinavian Germanic dialects. Among modern European tongues, the Italian is especially noticeable for its familiar use of reflexive phrases in a passive sense: thus, si dice, 'it says itself,' for 'it is said.'
- ↑ By Professor Schleicher, in his Compendium of Indo-European Comparative Grammar.
- ↑ In his work on the Origin of Language, seventh chapter.