Language and the Study of Language/Lecture II

by William Dwight Whitney

LECTURE II.

Nature of the force which produces the changes of language; its modes of action. Language an institution, of historical growth; its study a moral science. Analogies of linguistic sciences with the physical sciences. Its methods historical. Etymology its foundation. Analysis of compound words. Genesis of affixes. Nature of all words as produced by actual composition.

In the preceding lecture, after a very brief survey of the history and objects of linguistic science, we entered upon an inquiry into the means by which we had become possessed of our mother-tongue, an inquiry intended to bring out to our view the mode of transmission and preservation of language in general. And we saw that it is the work of tradition; that each generation passes along to the generation succeeding, with such faithfulness as the nature of the case permits, the store of words, phrases, and constructions which constitute the substance of a spoken tongue. But we also saw that the process of transmission is uniformly an imperfect one; that it never succeeds in keeping any language entirely pure and unaltered: on the contrary, language appeared to us as undergoing, everywhere and always, a slow process of modification, which in course of time effects a considerable change in its constitution, rendering it to all intents and purposes a new tongue. This was illustrated from the history of our English speech, which, by gradual and accumulated alterations made in it, during the past thousand years, by the thirty or forty generations through whose mouths it has passed, has grown from the Anglo-Saxon of King Alfred, through a succession of intermediate phases, into what it is at present. Before, now, we go on to examine in detail the processes of linguistic change, setting forth more fully their causes and modes of action, and exhibiting their results upon a more extended scale, we have to draw from what has been already said one or two important conclusions, touching the nature of the force by which those processes are carried on, and the character, and place among the sciences, of the study which undertakes their investigation.

And, in the first place, we see, I think, from our examination of the manner in which language is learned and taught, in which its life is kept up, what is meant when we speak and write of it as having an independent or objective existence, as being an organism or possessing an organic structure, as having laws of growth, as feeling tendencies, as developing, as adapting itself to our needs, and so on. All these are figurative expressions, the language of trope and metaphor, not of plain fact; they are wholly unobjectionable when consciously employed in their proper character, for the sake of brevity or liveliness of delineation; they are only harmful when we allow them to blind us to the real nature of the truths they represent. Language has, in fact, no existence save in the minds and mouths of those who use it; it is made up of separate articulated signs of thought, each of which is attached by a mental association to the idea it represents, is uttered by voluntary effort, and has its value and currency only by the agreement of speakers and hearers. It is in their power, subject to their will; as it is kept up, so is it modified and altered, so may it be abandoned, by their joint and consenting action, and in no other way whatsoever.

This truth is not only often lost from view by those who think and reason respecting language, but it is also sometimes explicitly denied, and the opposite doctrine is set up, that language has a life and growth independent of its speakers, with which men cannot interfere. A recent popular writer^[1] asserts that, "although there is a continuous change in language, it is not in the power of man either to produce or to prevent it: we might think as well of changing the laws which control the circulation of our blood, or of adding an inch to our height, as of altering the laws of speech, or inventing new words according to our own pleasure." Then, in order to establish the truth of this opinion, he goes on to cite a couple of historical instances, in which two famous emperors, Tiberius of Rome and Sigismund of Germany, committed blunders in their Latin, and were taken to task and corrected by humble grammarians, who informed their imperial majesties that, however great and absolute their power might be, it was not competent to make an alteration in the Latin language. The argument and conclusion we may take to be of this character: If so high and mighty a personage as an emperor could not do so small a thing as alter the gender and termination of a single word—not even, as Sigismund attempted, in a language which was dead, and might therefore be supposed incapable of making resistance to the indignity—much less can any one of inferior consideration hope to accomplish such a change, or any other of the changes, of greater or less account, which make up the history of speech: therefore, language is incapable of alteration by its speakers.

The utter futility of deriving such a doctrine from such a pair of incidents, or from a score, a hundred, or a thousand like them, is almost too obvious to be worth the trouble of pointing out. Against what authority more mighty than their own did these two emperors offend? Simply against the immemorial and well-defined usage of all who wrote and had ever written Latin—nothing more and nothing less. High political station does not confer the right to make and unmake language; a sovereign's grammatical blunders do not become the law of speech to his subjects, any more than do those of the private man. Each individual is, in a way, constantly trying experiments of modification upon his mother-tongue, from the time when, as a child, he drops sounds and syllables which it does not suit his convenience to pronounce, and frames inflections upon mistaken analogies, to that when, as a man, he is guilty of slang, vulgarisms, and bad grammar, or indulges in mannerisms and artificial conceits, or twists words out of their true uses, from ignorance or caprice. But his individual influence is too weak to make head against the consenting usage of the community; his proposals, unless for special reasons, are passed over unnoticed, and he is forced to conform his speech to that of the rest; or, if he insist upon his independence, he is contemned as a blunderer, or laughed at as a humourist.

That an alteration should have been made at the time of Sigismund in any item of Latin grammar, either by the emperor himself, or by all the potentates and learned men of Christendom, was an impossibility. For the language was a dead one; its proprieties of speech were no longer dependent upon the sanction of present usage, but upon the authority of unchanging models. Much that we say is good English, though Shakspeare and Milton knew it not; nothing can be good Latin, unless it be found in Cicero and Virgil, or their compeers. And even under Tiberius, the case was nearly the same: the great authors whose example makes the law of Latin speech had already lived and written; and any deviation from their usage would have been recognized by all coming time as a later corruption. Hence, even had that emperor's blunder been accepted and slavishly imitated by his courtiers, his army, and his subjects at large, their consent could have made it good second-rate Latin only; it might have become the very best usage in the later Italian, French, and Spanish, but it would always have been rejected and avoided by the strict classicists. And all this, not for the reason that man has no power over language, but precisely for the contrary reason, that he has all power over it—that men's usage makes language. He, accordingly, who can direct usage can make or alter language. In this way only can exalted rank confer authority over speech: it can give a more powerful impulse toward that general acceptance and currency which anything must win in order to be language. There are instances on record in which the pun of a monarch has changed for all time the form of a word. Ethnologists well know that the name of the so-called "Tartar" race is properly Tatar, and they are now endeavouring to restore this, its correct orthography. The intrusion of the r is accounted for in the following manner. When, in the reign of St Louis of France, the hordes of this savage race were devastating eastern Europe, the tale of their ravages was brought to the pious king, who exclaimed with horror: "Well may they be called Tartars, for their deeds are those of fiends from Tartarus." The appositeness of the metamorphosed appellation made it take, and from that time French authors—and, after their example, the rest of Europe—have called the Tatar "Tartars." Whether the story is incontestably authentic or not is of small consequence: any one can see that it might be true, and that such causes may have produced such effects times innumerable.

The speakers of language thus constitute a republic, or rather, a democracy, in which authority is conferred only by general suffrage and for due cause, and is exercised under constant supervision and control. Individuals are abundantly permitted to make additions to the common speech, if there be reason for it, and if, in their work, they respect the sense of the community. When the first schooner ever built, on the coast of Massachusetts, slid from her stocks and floated gracefully upon the water, the chance exclamation of an admiring by-stander, "Oh, how she scoons!" drew from her contriver and builder the answer, "A scooner let her be, then," and made a new English word. The community ratified his act, and accepted the word he proposed, because the new thing wanted a new name, and there was no one else so well entitled as he to name it; if, on the other hand, he had assumed to christen a man-of-war a scooner, no one but his nearest neighbours would ever have heard of the attempt. The discoverer of a new asteroid, again, is allowed to select its title, provided he choose the name of some classical goddess, as is the established precedent for such cases—although, even then, he is liable to have the motives of his choice somewhat sharply looked into. The English astronomer who sought, a few years since, with covert loyalty, to call his planetling "Victoria," was compelled to retract the appellation and offer another. An acute and learned Italian physician, some time in the last century, discovered a new- physical force, and some one called it galvanism, after his name. Many of us well remember how, not long ago, a French savant devised a novel and universally interesting application of certain chemical processes; and here, again, by some person to whose act the community gave its assent, the product was named for its inventor a daguerreotype: and galvanism and daguerreotype, with their derivatives, are now as genuine and well established parts of the English language as are sun and moon, or father and mother. If Galvani had denominated his new principle abracadabra, or if Daguerre had styled his sun-pictures aldiborontiphoscophornios, these names would, indeed, have been not less inherently suitable than the ones actually chosen, in the view of the great majority of those who have since learned to use the latter; for comparatively few have ever heard of the two eminent discoverers, or learned enough of Greek to be able to perceive the etymological aptness of type; yet those who are accustomed to direct public opinion upon such subjects would have revolted, and insisted upon the substitution of other titles, which should seem to them to possess an obvious reason and applicability. The public has looked on quietly, during the last half-century, while the geologists have been bringing into our English speech their flood of new words, nouns, adjectives, and verbs, of various origin and not seldom of uncouth and barbarous aspect, wherewith to signify the new knowledge added by them to the common stock that we all draw from: these gentlemen know best; if they agree among themselves that necessity and propriety require us to say Silurian, palæontological, oölite, post-pleiocene, and the like, we are ready to do so, whether our acquaintance with ancient and modern geography and with the classical tongues be or be not sufficient to enable us to discover or appreciate the reason of each term.

But even in respect to the more intimate and sacred part of language, the words and phrases of universal and every-day use, the community confers some measure of authority upon those who have a just title to it, upon great masters in the art of speech, upon speakers whose eloquence carries captive all hearts, upon writers whose power in wielding the common instrument of thought is felt and acknowledged through all ranks. Such a one may now and then coin a new word, if he follow established analogies; he may revive and bring again into currency one which had fallen into desuetude; he may confer on an old word a new value, not too far differing from that already belonging to it—and the license shall be ratified by general acceptance. A great author may, by his single authority, turn the trembling scale in favour of the admission to good usage of some popular word or phrase, born of an original corruption or blunder, which had hitherto been frowned upon and banned; nay, even his mannerisms and conceits may perhaps become the law of the language. The maxim usus norma loquendi, 'usage is the rule of speech,' is of supreme and uncontrolled validity in every part and parcel of every human tongue, and each individual can make his fellows talk and write as he does just in proportion to the influence which they are disposed to concede to him.

In a language circumstanced like ours, a conscious and detailed discussion sometimes arises on the question of admitting some new word into its recognized vocabulary. We all remember the newspaper controversy, not long ago, as to whether we ought to call a message sent by telegraph a telegraph or a telegram; and many of us, doubtless, are yet waiting to see how the authorities settle it, that we may govern our own usage accordingly. We have a suffix able, which, like a few others that we possess, we use pretty freely in forming new words. Within no very long time past, some writers and speakers have added it to the verb rely, forming the adjective reliable. The same thing must have been done at nearly the same time to other verbs, awakening neither question nor objection; while, nevertheless, reliable is still shut out from the best—or, at least, from the most exclusive—society in English speech. And why? Because, in the first place, say the objectors, the word is unnecessary; we have already trustworthy, which means the same thing: further, it is improperly and falsely formed; as we say "to rely on" anything, our derivative adjective, if we make one, should be relionable, not reliable: finally, it is low-caste; A, B, and C, those prime authorities in English style, are careful never to let it slip from their pens. The other side, however, are obstinate, and do not yield the point. The first objection, they retort, is insufficient; no one can properly oppose the enrichment of the language by a synonym, which may yet be made to distinguish a valuable shade of meaning—which, indeed, already shows signs of doing so, as we tend to say "a trustworthy witness," but "reliable testimony." The second is false: English etymology is by no means so precise in its application of the suffix able as the objectors claim; it admits laughable, meaning 'worthy to be laughed at,' unaccountable, 'not to be accounted for,' indispensable, 'not to be dispensed with,' as well as many other words of the same kind; and even objectionable, 'liable to objection,' marriageable, 'fit for marriage,' and so forth. As for the third objection, whatever A, B, and C may do, it is certain that D, F, and H, with most of the lower part of the alphabet (including nearly all the X's, Y's, and Z's, the unknown quantities), use the new form freely; and it is vain to stand out against the full acceptance of a word which is supported by so much and so respectable authority. How the dispute is likely, or ought, to terminate, need not concern us here; it is only referred to because, while itself carried on in full consciousness, and on paper, it is a typical illustration of a whole class of discussions which go on silently, and even more or less unconsciously, in the minds before which is presented, for acceptance or rejection, any proposed alteration in the subsisting usages of speech. Is it called for? is it accordant with the analogies of the language? is it offered or backed by good authority? these are the considerations by which general consent is won or repelled; and general consent decides every case without appeal.

Downright additions, however, to the vocabulary of a spoken tongue, even those who hold to the doctrine of the organic life of language will probably be willing to ascribe to human agency; since no man in his sober senses, it would seem, could possibly maintain that, when some individual mind has formed a conception or drawn a deduction, or when some individual ingenuity has brought forth a product of any of the modes of activity of which man is capable, language itself spontaneously extrudes a word for its designation! He who sees is likewise he who says; the ingenuity that could find the thing was never at a loss to devise also its appellation.

But the case is not otherwise with those gradual changes which bring about the decay of grammatical structure, or the metamorphosis of phonetic form, in a language. Though they go on in a more covert and unacknowledged way than the augmentations of a vocabulary, they are due to the action of the same forces. If we write knight, and pronounce it nīt, while our ancestors spelled the word cniht, and made its every letter distinctly audible (giving the i our short i-sound, as in pin)—just as the Germans even now both write and speak the same word knecht—we know that it is not because, by any force inherent in the word itself, the fuller form grew into the simpler, but because the combination kn, as initial, was somewhat difficult for men's organs to utter, and therefore began to lose its k, first, in the mouths of careless and easy speakers; and the corruption went on gaining in popularity, until it became the rule of our speech to silence the mute before the nasal in all such words (as in knife, knit, gnat, gnaw, etc.); because, moreover, the sound of the guttural h after a vowel became unpopular, men's organs shrinking from the effort of producing it, and was finally got rid of everywhere (being either left out entirely, as in nigh, ought, or turned into f, as in laugh, cough); while, at the same time, the loss of this consonant led to a prolongation of the vowel i, which was changed into the diphthongal sound we now give it; in company, too, with so many other of the "long i's" of the older language, that our usual name at present for the diphthong is "long i." And so in all the multitude of similar cases. There is no necessity, physiological or other, for the rustic's saying kău for cow; only the former is a lazy drawling utterance, which opens the mouth less widely than the latter. A precisely similar flattening of the simple sound of a, in such words as grăsp, grăft, dănce—which but a brief time since were universally pronounced grâsp, grâft, dânce (â as in far), and are so still in certain localities—is now so common as to have become the accepted mode of utterance; but no one fails to recognize in it a corruption of the previous pronunciation, made current by example and imitation, prompted and recommended by that lazy habit of mouth which has occasioned the dimming of so many of our clear vowels. The pronunciation eīther and neīther seems at the present time to be spreading in our community, and threatening to crowd out of use the better-supported and more analogical^[2] ēither and nēither; but it is only by the deliberate choice of persons who fancy that there is something nicer, more recherché, more "English," in the new sound, and by imitation of these on the part of others. Such phonetic changes, we are accustomed to say, are inevitable, and creep in of themselves; but that is only another way of saying that we know not who in particular is to blame for them. Offences must needs come, but there is always that man by whom they come, could we but trace him out.

It is unnecessary to dwell longer upon this point, or to illustrate it more fully, inasmuch as even those who teach the independent existence and organic growth of language yet allow that phonetic change is the work of men, endeavouring to make things easy to their organs of speech.

A language in the condition in which ours is at present, when thousands of eyes are jealously watching its integrity, and a thousand pens are ready to be drawn, and dyed deep in ink, to challenge and oppose the introduction into it of any corrupt form, of any new and uncalled-for element, can, of course, undergo only the slowest and the least essential alteration. It is when the common speech is in the sole keeping of the uncultivated and careless speakers, who care little for classical and time-honoured usages, to whom the preferences of the moment are of more account than anything in the past or in the future, that mutation has its full course. New dialects are wont to grow up among the common people, while the speech of the educated and lettered class continues to be what it has been. But the nature of the forces in action is the same in the one case as in the other: all change in language is the work of the will of its speakers, which acts under the government of motives, through the organs of speech, and varies their products to suit its necessities and its convenience. Every single item of alteration, of whatever kind, and of whatever degree of importance, goes back to some individual or individuals, who set it in circulation, from whose example it gained a wider and wider currency, until it finally won that general assent which is alone required in order to make anything in language proper and authoritative. Linguistic change must be gradual, and almost insensible while in progress, for the reason that the general assent can be but slowly gained, and can be gained for nothing which is too far removed from former usage, and which therefore seems far-fetched, arbitrary, or unintelligible. The collective influence of all the established analogies of a language is exerted against any daring innovation, as, on the other hand, it aids one which is obvious and naturally suggested. It was, for instance, no difficult matter for popular usage to introduce the new possessive its into English speech, nor to add worked to wrought, as preterit of work, nor to replace the ancient plural kye or kine (Anglo-Saxon cy, from cu, 'cow') by a modern one, cows, formed after the ordinary model: while to reverse either process, to crowd its, worked, and cows out of use by substitution of his, wrought, and kine, would have been found utterly impracticable. The power of resistance to change possessed by a great popular institution, which is bound up with the interests of the whole community, and is a part of every man's thoughts and habitual acts, is not easily to be overestimated. How long has it taken to persuade and force the French people, for instance, into the adoption of the new decimal system of weights and measures! How have they been baffled and shamed who have thought, in these latter days, to amend in a few points, of obvious desirability, our English orthography! But speech is a thing of far nearer and higher importance; it is the most precious of our possessions, the instrument of our thoughts, the organ of our social nature, the means of our culture; its use is not daily or hourly alone, but momently; it is the first thing we learn, the last we forget; it is the most intimate and clinging of our habits, and almost a second nature: and hence its exemption from all sweeping or arbitrary change. The community, to whom it belongs, will suffer no finger to be laid upon it without a reason; only such modifications as commend themselves to the general sense, as are virtually the carrying out of tendencies universally felt, have a chance of winning approval and acceptance, and so of being adopted into use, and made language.

Thus it is indeed true that the individual has no power to change language. But it is not true in any sense which excludes his agency, but only so far as that agency is confessed to be inoperative except as it is ratified by those about him. Speech and the changes of speech are the work of the community; but the community cannot act except through the initiative of its individual members, which it follows or rejects. The work of each individual is done unpremeditatedly, or as it were unconsciously; each is intent only on using the common possession for his own benefit, serving therewith his private ends; but each is thus at the same time an actor in the great work of perpetuating and of shaping the general speech. So each separate polyp on a coral-bank devotes himself simply to the securing of his own food, and excretes calcareous matter only in obedience to the exigencies of his individual life; but, as the joint result of the isolated labours of all, there slowly rises in the water the enormous coral cliff, a barrier for the waves to dash themselves against in vain. To pick out a single man, were he even an emperor, and hold him up to view in his impotence as proof that men cannot make or alter language, is precisely equivalent to selecting one polyp, though the biggest and brightest-coloured of his species, off the growing reef, and exclaiming over him, "See this weak and puny creature! how is it possible that he and his like should build up a reef or an island?" No one ever set himself deliberately at work to invent or improve language—or did so, at least, with any valuable and abiding result; the work is all accomplished by a continual satisfaction of the need of the moment, by ever yielding to an impulse and grasping a possibility which the already acquired treasure of words and forms, and the habit of their use, suggest and put within reach. In this sense is language a growth; it is not consciously fabricated; it increases by a constant and implicit adaptation to the expanding necessities and capacities of men.

This, again, is what is meant by the phrases "organic growth, organic development," as applied to language. A language, like an organic body, is no mere aggregate of similar particles; it is a complex of related and mutually helpful parts. As such a body increases by the accretion of matter having a structure homogeneous with its own, as its already existing organs form the new addition, and form it for a determinate purpose—to aid the general life, to help the performance of the natural functions, of the organized being—so is it also with language: its new stores are formed from, or assimilated to, its previous substance; it enriches itself with the evolutions of its own internal processes, and in order more fully to secure the end of its being, the expression of the thought of those to whom it belongs. Its rise, development, decline, and extinction are like the birth, increase, decay, and death of a living creature.

There is a yet closer parallelism between the life of language and that of the animal kingdom in general. The speech of each person is, as it were, an individual of a species, with its general inherited conformity to the specific type, but also with its individual peculiarities, its tendency to variation and the formation of a new species. The dialects, languages, groups, families, stocks, set up by the linguistic student, correspond with the varieties, species, genera, and so on, of the zoölogist. And the questions which the students of nature are so excitedly discussing at the present day—the nature of specific distinctions, the derivation of species by individual variation and natural selection, the unity of origin of animal life—all are closely akin with those which the linguistic student has constant occasion to treat. We need not here dwell further upon the comparison: it is so naturally suggested, and so fruitful of interesting and instructive analogies, that it has been repeatedly drawn out and employed, by students both of nature and of language.^[3]

Once more, a noteworthy and often-remarked similarity exist between the facts and methods of geology and those of linguistic study. The science of language is, as it were, the geology of the most modern period, the Age of Man, having for its task to construct the history of development of the earth and its inhabitants from the time when the proper geological record remains silent; when man, no longer a mere animal, begins by the aid of language to bear witness respecting his own progress and that of the world about him. The remains of ancient speech are like strata deposited in bygone ages, telling of the forms of life then existing, and of the circumstances which determined or affected them; while words are as rolled pebbles, relics of yet more ancient formations, or as fossils, whose grade indicates the progress of organic life, and whose resemblances and relations show the correspondence or sequence of the different strata; while, everywhere, extensive denudation has marred the completeness of the record, and rendered impossible a detailed exhibition of the whole course of development.

Other analogies, hardly less striking than these, might doubtless be found by a mind curious of such things. Yet they would be, like these, analogies merely, instructive as illustrations, but becoming fruitful of error when, letting our fancy run away with our reason, we allow them to determine our fundamental views respecting the nature of language and the method of its study; when we call language a living and growing organism, or pronounce linguistics a physical science, because zoölogy and geology are such. The point is one of essential consequence in linguistic philosophy. We shall never gain a clear apprehension of the phenomena of linguistic history, either in their individuality or in their totality, if we mistake the nature of the forces which are active in producing them. Language is, in fact, an institution—the word may seem an awkward one, but we can find none better or more truly descriptive—the work of those whose wants it subserves; it is in their sole keeping and control; it has been by them adapted to their circumstances and wants, and is still everywhere undergoing at their hands such adaptation; every separate item of which it is composed is, in its present form—for we are not yet ready for a discussion of the ultimate origin of human speech—the product of a series of changes, effected by the will and consent of men, working themselves out under historical conditions, and conditions of man's nature, and by the impulse of motives, which are, in the main, distinctly traceable, and form a legitimate subject of scientific investigation.

These considerations determine the character of the study of language as a historical or moral science. It is a branch of the history of the human race and of human institutions. It calls for aid upon various other sciences, both moral and physical: upon mental and metaphysical philosophy, for an account of the associations which underlie the developments of signification, and of the laws of thought, the universal principles of relation, which fix the outlines of grammar; upon physiology, for explanation of the structure and mode of operation of the organs of speech, and the physical relations of articulate sounds, which determine the laws of euphony, and prescribe the methods of phonetic change; upon physical geography and meteorology, even, for information respecting material conditions and climatic aspects, which have exerted their influence upon linguistic growth. But the human mind, seeking and choosing expression for human thought, stands as middle term between all determining causes and their results in the development of language. It is only as they affect man himself, in his desires and tendencies or in his capacities, that they can affect speech: the immediate agent is the will of men, working under the joint direction of impelling wants, governing circumstances, and established habits. What makes a physical science is that it deals with material substances, acted on by material forces. In the formation of geological strata, the ultimate cognizable agencies are the laws of matter; the substance affected is tangible matter; the product is inert, insensible matter. In zoölogy, again, as in anatomy and physiology, the investigator has to do with material structures, whose formation is dependent on laws implanted in matter itself, and beyond the reach of voluntary action. In language, on the other hand, the ultimate agencies are intelligent beings, the material is—not articulated sound alone, which might, in a certain sense, be regarded as a physical product, but—sound made significant of thought; and the product is of the same kind, a system of sounds with intelligible content, expressive of the slowly accumulated wealth of the human race in wisdom, experience, comprehension of itself and of the rest of creation. What but an analogical resemblance can there possibly be between the studies of things so essentially dissimilar?

There is a school of modern philosophers who are trying to materialize all science, to eliminate the distinction between the physical and the intellectual and moral, to declare for naught the free action of the human will, and to resolve the whole story of the fates of mankind into a series of purely material effects, produced by assignable physical causes, and explainable in the past, or determinable for the future, by an intimate knowledge of those causes, by a recognition of the action of compulsory motives upon the passively obedient nature of man. With such, language will naturally pass, along with the rest, for a physical product, and its study for a physical science; and, however we may dissent from their general classification, we cannot quarrel with its application in this particular instance. But by those who still hold to the grand distinction of moral and physical sciences, who think the action of intelligent beings, weighing motives and selecting courses of conduct, seeing ends and seeking means to their attainment, to be fundamentally and essentially different from that of atoms moved by gravity, chemical affinity, and the other immutable forces of nature, as we call them—by such, the study of language, whose dependence upon voluntary action is so absolute that not one word ever was or ever will be uttered without the distinct exertion of the human will, cannot but be regarded as a moral science; its real relationship is with those branches of human knowledge among which common opinion is accustomed to rank it—with mental philosophy, with philology, with history.

While, however, we are thus forced to the acknowledgment that everything in human speech is a product of the conscious action of human beings, we should be leaving out of sight a matter of essential consequence in linguistic investigation if we failed to notice that what the linguistic student seeks in language is not what men have voluntarily or intentionally placed there. As we have already seen, each separate item in the production or modification of language is a satisfaction of the need of the moment; it is prompted by the exigencies of the particular case; it is brought forth for the practical end of convenient communication, and with no ulterior aim or object whatsoever; it is accepted by the community only because it supplies a perceived want, and answers an acknowledged purpose in the uses of social intercourse. The language-makers are quite heedless of its position and value as part of a system, or as a record with historical content, nor do they analyze and set before their consciousness the mental tendencies which it gratifies. A language is, in very truth, a grand system, of a highly complicated and symmetrical structure; it is fitly comparable with an organized body; but this is not because any human mind has planned such a structure and skilfully worked it out. Each single part is conscious and intentional; the whole is instinctive and natural. The unity and symmetry of the system is the unconscious product of the efforts of the human mind, grappling with the facts of the world without and the world within itself, and recording each separate result in speech. Herein is a real language fundamentally different from the elaborate and philosophical structures with which ingenious men have sometimes thought to replace them.^[4] These are indeed artful devices, in which the character and bearing of each part is painfully weighed and determined in advance: compared with them, language is a real growth; and human thought will as readily exchange its natural covering for one of them as the growing crustacean will give up its shell for a casing of silver, wrought by the most skilful hands. Their symmetry is that of a mathematical figure, carefully laid out, and drawn to rule and line; in language, the human mind, tethered by its limited capacities in the midst of creation, reaches out as far as it can in every direction and makes its mark, and is surprised at the end to find the result a circle.

In whatever aspect the general facts of language are viewed, they exhibit the same absence of reflection and intention. Phonetic change is the spontaneous working out of tendencies which the individual does not acknowledge to himself; in their effects upon organs of whose structure and workings he is almost or wholly ignorant. Outward circumstances, historical conditions, progress of knowledge and culture, are recorded in speech because its practical uses require that they should be so, not because any one has attempted to depict them. Language shows ethnic descent, not as men have chosen to preserve such evidence of their kindred with other communities and races, but as it cannot be effaced without special effort directed to that end. The operations of the mind, the development of association, the laws of subjective relation, are exhibited there, but only as they are the agencies which govern the phenomena of speech, unrecognized in their working, but inferrible from their effects.

Now it is this absence of reflection and conscious intent which takes away from the facts of language the subjective character that would otherwise belong to them as products of voluntary action. The linguistic student feels that he is not dealing with the artful creations of individuals. So far as concerns the purposes for which he examines them, and the results he would derive from them, they are almost as little the work of man as is the form of his skull, the outlines of his face, the construction of his arm and hand. They are fairly to be regarded as reflections of the facts of human nature and human history, in a mirror imperfect, indeed, but faithful and wholly trustworthy; not as pictures drawn by men's hands for our information. Hence the close analogies which may be drawn between the study of language and some of the physical sciences. Hence, above all, the fundamental and pervading correspondence between its whole method and theirs. Not less than they, it founds itself upon the widest observation and examination of particular facts, and proceeds toward its results by strict induction, comparing, arranging, and classifying, tracing out relations, exhibiting an inherent system, deducing laws of general or universal application, discovering beneath all the variety and diversity of particulars an ever-present unity, in origin and development, in plan and purpose. Beyond all question, it is this coincidence of method which has confused some of the votaries of linguistic science, and blinded their eyes to the true nature of the ultimate facts upon which their study is founded, leading them to deny the agency of man in the production and change of language, and to pronounce it an organic growth, governed by organic forces.

Another motive—a less important one, and in great part, doubtless, unconscious in its action—impelling certain students of language to claim for their favourite branch of investigation a place in the sisterhood of physical sciences, has been, as I cannot but think, an apprehension lest otherwise they should be unable to prove it entitled to the rank of a science at all. There is a growing disposition on the part of the devotees of physical studies—a class greatly and rapidly increasing in importance, and influence—to restrict the honourable title of science to those departments of knowledge which are founded on the unvarying laws of material nature, and to deny the possibility of scientific method and scientific results where the main element of action is the varying and capricious will of man. The considerations adduced above, it is hoped, will remove this apprehension. Nor was it ever otherwise than needless, as the tendency which called it forth is mistaken and unjustifiable. The name "science" admits no such limitation. The vastness of a field of study, the unity in variety of the facts it includes, their connection by such ties that they allow of strict classification and offer fruitful ground for deduction, and the value of the results attained, the truth deduced—these things make a science. And, in all these respects, the study of language need fear a comparison with no one of the physical sciences. Its field is the speech of all mankind, cultivated or savage; the thousands of existing dialects, with all their recorded predecessors; the countless multitudes of details furnished by these, each significant of a fact in human history, external or internal. The wealth of languages is like the wealth of species in the whole animal kingdom. Their tie of connection is the unity of human nature in its wants and capacities, the unity of human knowledge, of existing things and their relations, to be apprehended by the mind and reflected in speech—a bond as infinite in its ramifications among all the varieties of human language, and as powerful in its binding force, as is the unity of plan in vegetable or animal life. The results, finally, for human history, the history of mind, of civilization, of connection of races, for the comprehension of man, in his high endowments and in his use of them, are of surpassing interest. To compare their worth with that of the results derivable from other sciences were to no good purpose: all truth is valuable, and that which pertains to the nature and history of man himself is, to say the least, not inferior in interest to that which concerns his surroundings. Linguistic science, then, has in itself enough of dignity and true scientific character not to need to borrow aught of either from association with other branches of inquiry, which differ from it in subject and scope, while yet they seek by corresponding methods the same ultimate object, the increase of knowledge, and the advancement of man in comprehension of himself and of the universe.

We return, now, from this necessary digression, to follow onward our leading inquiry, "Why we speak as we do?" And we have to push the question a step further than in the last lecture, asking this time, not simply how we ourselves came into possession of the signs of which our mother-tongue is made up, but also how those from whom we learned them came into possession of them before us; how the tradition from whose hands we implicitly accepted them got them in the form in which it passed them on to us; why our words, in short, are what they are, and not otherwise. We have seen that every part and particle of every existing language is a historical product, the final result of a series of changes, working themselves out in time, under the pressure of circumstances, and by the guidance of motives, which are not beyond the reach of our discovery. This fact prescribes the mode in which language is to be fruitfully studied. If we would understand anything which has become what it is, a knowledge of its present constitution is not enough: we must follow it backward from stage to stage, tracing out the phases it has assumed, and the causes which have determined the transition of one into the other. Merely to classify, arrange, and set forth in order the phenomena of a spoken tongue, its significant material, usages and modes of expression, is grammar and lexicography, not linguistic science. The former state and prescribe only; the latter seeks to explain. And when the explanation is historical, the search for it must be of the same character. To construct, then, by historical processes, with the aid of all the historical evidences within his reach, the history of development of language, back to its very beginning, is the main task of the linguistic student; it is the means by which he arrives at a true comprehension of language, in its own nature and in its relations to the human mind and to human history.

Furthermore, it is hardly necessary to point out that the history of language reposes on that of words. Language is made up of signs for thought, which, though in one sense parts of a whole, are in another and more essential sense isolated and independent entities. Each is produced for its own purpose; each is separately exposed to the changes and vicissitudes of linguistic life, is modified, recombined, or dropped, according to its own uses and capacities. Hence etymology, the historical study of individual words, is the foundation and substructure of all investigation of language; the broad principles, the wide-reaching views, the truths of universal application and importance, which constitute the upper fabric of linguistic science, all rest upon word-genealogies. Words are the single witnesses from whom etymology draws out the testimony which they have to give respecting themselves, the tongue to which they belong, and all human speech.

How the study of words is made the means of bringing to light the processes of linguistic growth, and what those processes are, it will, accordingly, be our next duty to examine and set forth by suitable examples. Having only illustration in view, we will avoid all cases of a difficult or doubtful character, noticing only words whose history is well known; choosing, moreover, those which, while they truly exhibit the principles we seek to establish, are at the same time of the simplest kind, and most open to general comprehension.

There is no word or class of words whose history does not exemplify, more or less fully, all the different kinds of linguistic change. It will be more convenient for us, however, to take up these kinds in succession, and to select our instances accordingly. And, as the possibility of etymological analysis depends in no small part on the nature of words as not simple entities, but made of of separate elements, this composite character of the constituents of speech may properly engage our first attention.

That we are in the constant habit of putting together two independent vocables to form a compound word, is an obvious and familiar fact. Instances of such words are fear-inspiring, god-like, break-neck, house-top. They are substitutes for the equivalent phrases inspiring fear, like a god, apt to break one's neck, top of a house. For the sake of more compact and convenient expression, we have given a closer unity to the compound word than belongs to the aggregate which it represents, by omission of connectives, by inversion of the more usual order of arrangement, but most of all by unity of accent: this last is the chief outward means of composition; it converts two entities into one, for the nonce, by subordinating the one of them to the other. Our common talk is strewn with such words, and so gradual is the transition to them from the mere collocations of the phrase, that there are couples, like mother-tongue, well-known, which we hardly know whether to write separately, as collocations only, or with a hyphen, as loose compounds; others, like dial-plate, well-being, usage so far recognizes for compounds that they are always written together, sometimes with the hyphen and sometimes without; others yet, like godlike, herself are so grown together by long contact, by habitual connection, that we hardly think of them as having a dual nature. And even more than this: we have formed some so close combinations that it costs us a little reflection to separate them into their original parts. Of such a character is forehead, still written to accord with its derivation, as a name for the fore part of the head, but so altered in pronunciation that, but for its spelling, its origin would certainly escape the notice of nineteen-twentieths of those who use it. Such, again, is fortnight, altered both in pronunciation and in spelling from the fourteen nights out of which it grew. Such, once more, is our familiar verb breakfast. We gave this name to our morning meal, because it broke, or interrupted, the longest fast of the day, that which includes the night's sleep. 'We said at first breāk fâst—"I broke fast at such an hour this morning;" he, or they, who first ventured to say I breakfasted were guilty of as heinous a violation of grammatical rule as he would be who should now declare I takedinnered, instead of I took dinner; but good usage came over to their side and ratified their blunder, because the community were minded to give a specific name to their earliest meal and to the act of partaking of it, and therefore converted the collocation breākfâst into the real compound brĕakfast.

Yet once more, not only are those words in our language of composite structure, of which at first sight, or on second thought, we thus recognize the constituent elements; not a few, also, which we should not readily conjecture to be other than simple and indivisible entities, and which could not be proved otherwise by any evidence which our present speech contains, do nevertheless, when we trace their history by the aid of other and older languages than ours, admit of analysis into component parts. We will note, as instances, only a familiar word or two, namely such and which. The forms of these words in Anglo-Saxon are swylc and hwylc: with the latter of them the Scottish whilk for which quite closely agrees, and they also find their near correspondents in the German solch and welch. On following up their genealogy, from language to language of our family, we find at last that they are made up of the ancient words for so and who, with the adjective like added to each: such is so-like, 'of that likeness or sort;' which is who-like, 'of what likeness or sort.'

But we turn from compounds like these, in which two originally independent words are fully fused into one, in meaning and form, to another class, of much higher importance in the history of language.

Let us look, first, at our word fearful. This, upon reflection, is a not less evident compound than fear-inspiring: our common adjective full is perfectly recognizable as its final member. Yet, though such be its palpable origin, it is, after all, a compound of a somewhat different character from the other. The subordinate element full, owing to its use in a similar way in a great number of other compounds, such as careful, truthful, plentiful, dutiful, and the frequent and familiar occurrence of the words it forms, has, to our apprehension, in some measure lost the consciousness of its independent character, and sunk to the condition of a mere suffix, forming adjectives from nouns, like the suffix ous in such words as perilous, riotous, plenteous, duteous. It approaches, too, the character of a suffix, in that its compounds are not, like fear-inspiring and house-top, directly translatable back into the elements which form them: plentiful and dutiful do not mean 'full of plenty' and 'full of duty,' but are the precise equivalents of plenteous and duteous. We could with entire propriety form an adjective from a new noun by adding ful to it, without concerning ourselves, as to whether the corresponding phrase, "full of so and so," would or would not make good sense. And when we hear a Scotchman say fearfu', carefu', we both understand him without difficulty, and do not think of inquiring whether he also clips the adjective full to fu'.

The word of opposite meaning, fearless, is not less readily recognizable as a compound, and our first impulse is to see in its final element our common word less, to interpret fearless as meaning 'minus fear,' 'deprived of fear,' and so 'exempt from fear.' A little study of the history of such words, however, as it is to be read in other dialects, shows us that this is a mistake, and that our less has nothing whatever to do with the compound. The Anglo-Saxon form of the ending, leas, is palpably the adjective leas, which is the same with our word loose; and fearless is primarily 'loose from fear,' 'free from fear.' The original subordinate member of the compound has here gone completely through the process of conversion into a suffix, being so divorced from the words which are really akin with it that its derivation is greatly obscured, and a false etymology is suggested to the mind which reflects upon it.

Take, again, such words as godly, homely, brotherly, lovely. Here, as in the other cases, each is composed of two parts; but, while we recognize the one as a noun, having an independent existence in the language, we do not even feel tempted to regard the other as anything but an adjective suffix, destitute of separate significance; it appears in our usage only as an appendage to other words, impressing upon them a certain modification of meaning. What, however, is its history? Upon tracing it up into the older form of our speech, the Anglo-Saxon, we find that our modern usage has mutilated it after the same fashion as the Scottish dialect now mutilates the ful of fearful—by dropping off, namely, an original final consonant: its earlier form was lic. The final guttural letter we find preserved even to the present day in the corresponding suffixes of the other Germanic languages, as in the German lich, Swedish lic, Dutch lijk. These facts lead us naturally to the conjecture that the so-called suffix may be nothing more than a metamorphosis of our common adjective like; and a reference to the oldest Germanic dialect, the Mœso-Gothic, puts the case beyond all question; for there we find the suffix and the independent adjective to be in all respects the same, and the derivatives formed with the suffix to be as evident compounds with the adjective as are our own godlike, childlike, and so on. Words thus composed are common in all the Germanic tongues; but we who speak English have given the same suffix a further modification of meaning, and an extension of application, which belong to it nowhere else. In our usage it is an adverbial suffix, by which any adjective whatever may be converted into an adverb, as in truly, badly, fearfully, fearlessly. In the old Anglo-Saxon, such adverbs were oblique cases of adjectives in lic, and so, of course, were derived only from adjectives formed by this ending; the full adverbial suffix was lice, the e being a case-termination: instances are ânlîce, 'only, singularly,' from ânlîc, 'sole, singular,' literally 'one-like;' leôflîce, 'lovelily,' from leôflîc, 'lovely.' We moderns, now, have suffered the ending to go out of use as one forming adjectives, only retaining the adjectives so formed which we have inherited from the ancient time; but we have taken it up in its adverbial application, and, ignoring both its original character and its former limitation to a single class of adjectives, apply it with unrestricted freedom in making an adverb from any adjective we choose; while, at the same time, we have mutilated its form, casting off as unnecessary the vowel ending, along with the consonant to which it was appended. The history of this adverbial suffix is worthy of special notice, inasmuch as the suffix itself is the latest addition which our grammatical system has gained in the synthetic way, and as its elaboration has taken place during the period when the growth of our language is illustrated by contemporary documents. The successive steps were clearly as follows: the adjective like was first added to a number of nouns, forming a considerable class of adjective compounds, like those now formed by us with full; then, like the latter word, it lost in a measure the consciousness of its origin, and was regarded rather as a suffix, forming derivative adjectives; one of the oblique cases of these adjectives was next often employed in an adverbial sense; and the use of the suffix in its extended form and with its modified application grew in importance and frequency, until finally it threw quite into the shade and supplanted the adjective use—and the independent adjective had become a mere adverbial ending. The mutilation of its form went hand in hand with this obliviousness of its origin and with its transferral to a new office; each helped on the other.

Another Germanic suffix, ship, as in friendship, worship, lordship, is distinctly traceable to its origin in the independent word shape; and its transition of meaning, from 'form' to 'aspect, condition, status, rank,' though perhaps less obvious than those which we have already noted, is evidently a natural and easy one.

A case of somewhat greater difficulty is presented us in such forms as I loved. Here the final d is, as we say, the sign of the preterit tense, added to the root love in order to adapt it to the expression of past time; and, from the evidence presented in our own language, no suspicion of its derivation from an independent word would ever cross our minds. Nor does the Anglo-Saxon, nor any other of the Germanic dialects of the same period, cast any light upon its origin. Since, however, such a sign of past time is one of the distinctive features of the Germanic group of languages, and is found nowhere else in the greater family to which these belong, we cannot help assuming that it has grown up in them since their separation from the rest of the family; just as the adverbial suffix ly, which is peculiar to our own tongue, has grown up in it since its separation from the other Germanic tongues. It is therefore a form of comparatively modern introduction, and we might hope to trace out its genesis. This is, in fact, disclosed to us by the Mœso-Gothic, the most ancient Germanic dialect, which stands toward the rest in somewhat the same relation as the Anglo-Saxon to the English; in its primitive and uncorrupted forms we see clearly that the preterits in question are made by appending to the root of the verb the past tense of another verb, namely did, from to do. We tamed is in Mœso-Gothic tamidêdum, which means not less evidently tame-did-we than the Anglo-Saxon sôthlîce, 'soothly, truly,' means 'in a sooth-like (truth-like) way.' I loved is, then, originally I love did, that is, I did love—as, unconsciously repeating in another way the same old act of composition, we now almost as often say. The history of the suffix has been quite like that of the ly of truly, save that it happened longer ago, and is therefore more difficult to read.

All our illustrations hitherto have been taken from the Germanic part of our language, and they have all been forms which are peculiar to the Germanic dialects, and which we have therefore, as already remarked, every reason to believe of later date than the separation of that group of dialects from the other tongues with which it stands related. Yet, with the exception of the adverbial application of the suffix ly, they are all anterior to the time at which we first make acquaintance with any Germanic tongue in contemporary records. Our confidence in the reality of our etymological analysis, and in the justness of the inferences drawn from it, is not on that account any the less: we feel as sure that the words in question were made by putting together the two parts into which each is still resolvable as if the whole process of composition had gone on under our own observation. If this were not so, if our conclusions respecting the growth of language were to be limited by the possession of strict documentary evidence, our researches in linguistic history would be stopped almost at the outset. Few languages have any considerable portion of their development illustrated by contemporary records; literature is wont, at the best, to cast light upon certain distinct epochs in the history of a dialect, leaving in obscurity the intervening periods, nor do we ever, by such help, reach a point at all nearly approaching that of the actual origin of speech. Hence the necessity resting upon the etymologist of interrogating the material of language itself, of making words yield up, on examination, their own history. He applies the analogy of processes of change and development which are actually going on in language to explain the earlier results of the same or like processes. And, if he work with due caution and logical strictness, his results are no more exposed to question than are those of the geologist, who infers, from the remains of animal and vegetable organisms in deeply-buried rocks, the deposition of those rocks in a period when animal and vegetable life, analogous with that of our own day, was abundant.

If, now, we turn our attention to other portions of our English speech, to those which come to us from the Latin, or which are of an ancient and primitive growth, we note the same condition of things as prevailing there also. The subject admits of the most abundant and varied illustration, but we must limit ourselves to but an instance or two.

In the series of multiplicative numerals, double, triple, quadruple, quintuple, and so on, we have a suffix ple, which is the principal indicator of the grammatical quality of the words. On following them up into the Latin, whence we derive them, we find this brief ending to be a mutilated remnant of the syllable plic, which is a well-known root, meaning 'to bend, to fold.' Double is thus by origin duplic, by abbreviation from duo-plic, and is, in sense, the precise Latin equivalent of our Germanic word two-fold. We still retain the fuller form in duplicate, the learned synonym of double.

Again, one of the oldest words in our familiar speech is am, the first person of the verb to be, nor do we see in it any signs of being otherwise than simple and indivisible. As, however, we trace its history of changes backward, from one to another of the languages with which our own claims kindred, we are enabled to discover that its two sounds are the scanty relics of two separate elements: the first, a, is all that remains of an original syllable as, which expressed the idea of existence; the other, m, represents an ending, mi, which, originally a pronoun, and having the same meaning as the same word, me, still has with us, was employed to limit the predicate of existence to the person speaking; it was, in fact, the suffix universally employed, during the earliest period in the history of our family of languages, to form the first persons singular of verbs. Am, then, really contains a verb and its subject pronoun, and means 'be-I;' that is, 'I exist.' The third person of the same verb, is, possesses virtually a similar character, although linguistic usage, in its caprice, has dealt somewhat differently with it. As am stands for as-mi, 'be-I,' so is stands for as-ti, 'be-that:' we have, indeed, worn off the second element altogether, so that our is is the actual representative only of the radical syllable as; but by far the greater number of the Germanic dialects, and of the other descendants from the primitive tongue in which was first formed the compound asti, have retained at least the initial consonant of the pronominal suffix: witness the German ist, the Slavonian yest, the Latin est, the Greek and Lithuanian esti, the Sanscrit asti, and so on. It is the same t which, in the form of th or s, still does service in the regular scheme of conjugation of our verbs, as ending of the third person singular present: thus, he loveth or loves.

The examples already given may sufficiently answer our purpose as illustrations of the way in which suffixes are produced, and grammatical classes or categories of words created. The adjectives in ful, or the adjectives in less, form together a related group, having a common character, as derivatives from nouns, and derivatives possessing a kindred significance, standing in a certain like relation to their primitives, filling a certain common office in speech, an office of which the sign is the syllable ful, or less, their final member or suffix. With ly, this is still more notably the case: the suffix ly is the usual sign of adverbial meaning; it makes much the largest share of all the adverbs we have. A final m, added to a verbal root, in an early stage of the history of our mother-tongue, and yet more anciently an added syllable mi, made in like manner the first persons singular present of verbs; as an added s, standing for an original syllable ti, does even to the present day make our third persons singular. All these grammatical signs were once independent elements, words of distinct meaning, appended to other words and compounded with them—appended, not in one or two isolated cases only, but so often, and in a sense so generally applicable, that they formed whole classes of compounds. There was nothing about them save this extensibility of their application and frequency of their use to distinguish their compounds from such as house-top, break-neck, forehead, fortnight, and the others of the same class to which we have already referred. Yet this was quite enough to bring about a change of their recognized character, from that of distinct words to that of non-significant appendages to other words. Each passed over into the condition of a formative element; that is to say, an element showing the logical form, the grammatical character, of a derivative, as distinguished from its primitive, the word to which the sign was appended. There was a time when fear-full, fear-loose, fear-free, free-making, fear-struck, love-like, love-rich, love-sick, love-lorn, were all words of the same kind, mere lax combinations; it was only their different degree of availability for answering the ends of speech, for supplying the perceived needs of expression, that caused two or three of them to assume a different character, while the rest remained as they had been.

Often, as every one knows, there is an accumulation of formative elements in the same word. In truthfully, for example, we have the adverbial suffix ly added to the primitive truthful; in which, again, the adjective suffix ful has performed the same office toward the remoter primitive truth. By the use of a formative element of another kind, a prefix, we might have made the yet more intricate compound untruthfully. Nay, further, truth itself contains a suffix, and is a derivative from the adjective true, as appears from its analogy with wealth from well, width from wide, strength from strong, and many other like words; and even true, did we trace its history to the beginning, we should find ending in a formative element, and deriving its origin from a verbal root meaning 'to be firm, strong, reliable.' The Latin part of our language, which includes most of our many-syllabled words, offers abundant instances of a similar complicated structure. Thus, the term inapplicabilities contains two prefixes, the negative in and the preposition ad which means 'to,' and three suffixes, able, forming adjectives, ty, forming abstract nouns from adjectives, and s, the plural ending, all clustered about the verbal root plic, which we have already seen itself forming a suffix, in double, triple, and so forth, and which conveys the idea of 'bending' or 'folding.' By successive extensions and modifications of meaning, by transferral from one category to another through means of their appropriate signs, we have developed this simple idea into a form which can only be represented by the long paraphrase 'numerous conditions of being not able to bend (or fit) to something.'

With but few exceptions—which, moreover, are only apparent ones—all the words of our language admit of such analysis as this, which discovers in them at least two elements, whereof the one conveys the central or fundamental idea, and the other indicates a restriction, application, or relation of that idea. Even those brief vocables which appear to us of simplest character can be proved either to exhibit still, like am for as-mi, the relic of a mutilated formative element, or, like is for as-ti, to have lost one which they formerly possessed. This, then, in our language (as in the whole family of languages to which ours is related), is the normal constitution of a word: it invariably contains a radical and a formal portion; it is made up of a root combined with a suffix, or with a suffix and prefix, or with more than one of each. In more technical phrase, no word is unformed; no one has been a mere significant entity, without designation of its relation, without a sign putting it in some class or category.

It is plain, therefore, that a chief portion of linguistic analysis must consist, not in the mere dismembering of such words as we usually style compounded, but in the distinction from one another of radical and formal elements; in the isolation of the central nucleus, or root, from the affixes which have become attached to it, and the separate recognition of each affix, in its individual form and office. But our illustrations have, as I think, made it not less plain that there is no essential and ultimate difference in the two cases: in the one, as in the other, our process of analysis is the retracing of a previous synthesis, whereby two independent elements were combined and integrated. That this is so to a certain extent is a truth so palpable as to admit of neither denial nor doubt. Had there been in the Germanic languages no such adjective as full, no such derivative adjectives as fearful and truthful would have grown up in them; if they had possessed no adjective like, they would never have gained such adjectives as godly and lovely, nor such adverbs as fearfully and truly. So also with friendship, with loved, with am and is, and the rest. No inconsiderable number of the formative elements of our tongue, in every department of grammar and of word-formation, can be thus traced back to independent words, with which they were at first identical, out of which they have grown. It is true, at the same time, that a still larger number do not allow their origin to be discovered. But we have not, on that account, the right to conclude that their history is not of the same character. In grammar, as everywhere else, like effects presuppose like causes. We have seen how the formative elements are liable to become corrupted and altered, so that the signs of their origin are obscured, and may even be obliterated. The full in truthful is easy enough to recognize, but a little historical research is necessary in order to show us the like which is contained in truly. Hateful is, for aught we know, as old a compound as lovely, but linguistic usage has chanced to be more merciful to the evidence of descent in the former case than in the latter. A yet more penetrating investigation is required ere we discover our pronoun me in the word am, or our imperfect did in I loved; and, but for the happy chance that preserved to us the one or two fragmentary manuscripts in which are contained our only records of Mœso-Gothic speech, the genesis of the latter form would always have remained an unsolved problem, a subject for ingenious conjecture, but beyond the reach of demonstration. The loss of each intermediate stage, coming between any given dialect and its remotest ancestor, wipes out a portion of the evidence which would explain the origin of its forms. If English stood all alone among the other languages of the earth, but an insignificant part of its word-history could be read; its kindred dialects, contemporary and older, help us to the discovery of a much larger portion; and the preservation of authentic records of every period of its life would, as we cannot hesitate to believe, make clear the rest. There is no break in the chain of analogical reasoning which compels the linguistic student to the conviction that his analyses are everywhere real, and distinguish those elements by the actual combination of which words were originally made up. On this conviction rests, for him, the value of his analytical processes: if they are to be regarded as in part historical and real, in part only theoretical and illusory, his researches into the history of language are baffled; he is in pursuit of a phantom, and not of truth.

Wherever, then, our study of words brings us to the recognition of an element having a distinct meaning and office, employed in combination with other elements for the uses of expression, there we must recognize an originally independent entity. The parts of our words were once themselves words.

Some of the remoter consequences involved in this principle will engage our attention at a more advanced stage of our inquiries into the history of human speech: our present purpose only requires us to notice that, since all known words have been constructed by putting together previously existing items of speech, the combination of old materials into new forms, the making of compounds, with frequent accompanying reduction of one of their members to a merely formal significance, is a very prominent part of the mechanism of language, one of the most fundamental and important of the processes by which are carried on its perpetual growth and change, its organic development. What other processes are the concomitants and auxiliaries of this one we shall go on to inquire in the next lecture.

Notes

1. Professor Max Müller, in his Lectures on the Science of Language, first series, second lecture.
2. The only English word in which ei has the "long i" sound is height, and even there it is nothing but an old orthographical blunder; there was no reason for divorcing the derivative noun in spelling from its theme, high.
3. For instance, by Lyell (Antiquity of Man, chapter xxiii.), who has founded upon it a lucid and able analogical argument bearing on the Darwinian theory of the mutation of species. Professor August Schleicher (Die Darwinsche Theorie und die Sprachwissenschaft, Weimar, 1863) attempts absolutely to prove by its aid the truth of the Darwinian theory, overlooking the fact that the relation between the two classes of phenomena is one of analogy only, not of essential agreement.
4. For an account of some of these attempts at an artificial language, of theoretically perfect structure, and designed for universal use, see Professor Max Müller's Lectures on Language, second series, second lecture.