Lectures on the Science of Language

by F. Max Müller

The first objection which was sure to be raised on the part of such sciences as botany, geology, or physiology is this:—Language is the work of man; it was invented by man as a means of communicating his thoughts, when mere looks and gestures proved inefficient; and it was gradually, by the combined efforts of succeeding generations, brought to that perfection which we admire in the idiom of the Bible, the Vedas, the Koran, and in the poetry of Homer, Virgil, Dante, and Shakespeare. Now it is perfectly true that if language be the work of man, in the same sense in which a statue, or a temple, or a poem, or a law are properly called the works of man, the science of language would have to be classed as an historical science. We should have a history of language as we have a history of art, of poetry, and of jurisprudence, but we could not claim for it a place side by side with the various branches of Natural History. It is true, also, that if you consult the works of the most distinguished modern philosophers you will find that whenever they speak of language, they take it for granted that language is a human invention, that words are artificial signs, and that the varieties of human speech arose from different nations agreeing on different sounds as the most appropriate signs of their different ideas. This view of the origin of language was so powerfully advocated by the leading philosophers of the last century, that it has retained an undisputed currency even among those who, on almost every other point, are strongly opposed to the teaching of that school. A few voices, indeed, have been raised to protest against the theory of language being originally invented by man. But they, in their zeal to vindicate [pg 040] the divine origin of language, seem to have been carried away so far as to run counter to the express statements of the Bible. For in the Bible it is not the Creator who gives names to all things, but Adam. “Out of the ground,” we read, “the Lord God formed every beast of the field, and every fowl of the air; and brought them unto Adam to see what he would call them: and whatsoever Adam called every living creature, that was the name thereof.”[14] But with the exception of this small class of philosophers, more orthodox even than the Bible,[15] the generally received opinion on the origin of language is that which was held by Locke, which was powerfully advocated by Adam Smith in his Essay on the Origin of Language, appended to his Treatise on Moral Sentiments, and which was adopted with slight modifications by Dugald Stewart. According to them, man must have lived for a time in a state of mutism, his only means of communication consisting in gestures of the body, and in the changes of countenance, till at last, when ideas multiplied that could no longer be pointed at with the fingers, “they found it necessary to invent artificial signs of which the meaning was [pg 041] fixed by mutual agreement.” We need not dwell on minor differences of opinion as to the exact process by which this artificial language is supposed to have been formed. Adam Smith would wish us to believe that the first artificial words were verbs. Nouns, he thinks, were of less urgent necessity because things could be pointed at or imitated, whereas mere actions, such as are expressed by verbs, could not. He therefore supposes that when people saw a wolf coming, they pointed at him, and simply cried out, “He comes.” Dugald Stewart, on the contrary, thinks that the first artificial words were nouns, and that the verbs were supplied by gesture; that, therefore, when people saw a wolf coming, they did not cry “He comes,” but “Wolf, Wolf,” leaving the rest to be imagined.[16]

But whether the verb or the noun was the first to be invented is of little importance; nor is it possible for us, at the very beginning of our inquiry into the nature of language, to enter upon a minute examination of a theory which represents language as a work of human art, and as established by mutual agreement as a medium of communication. While fully admitting that if this theory were true, the science of language would not come within the pale of the physical sciences, I must content myself for the present with pointing out that no one has yet explained how, without language, a discussion on the merits of each word, such as must necessarily have preceded a mutual agreement, could have been carried on. But as it is the object of these lectures to prove that language is not a work of human art, in the same sense [pg 042] as painting, or building, or writing, or printing, I must ask to be allowed, in this preliminary stage, simply to enter my protest against a theory, which, though still taught in the schools, is, nevertheless, I believe, without a single fact to support its truth.

But there are other objections besides this which would seem to bar the admission of the science of language to the circle of the physical sciences. Whatever the origin of language may have been, it has been remarked with a strong appearance of truth, that language has a history of its own, like art, like law, like religion; and that, therefore, the science of language belongs to the circle of the historical, or, as they used to be called, the moral, in contradistinction to the physical sciences. It is a well-known fact, which recent researches have not shaken, that nature is incapable of progress or improvement. The flower which the botanist observes to-day was as perfect from the beginning. Animals, which are endowed with what is called an artistic instinct, have never brought that instinct to a higher degree of perfection. The hexagonal cells of the bee are not more regular in the nineteenth century than at any earlier period, and the gift of song has never, as far as we know, been brought to a higher perfection by our nightingale than by the Philomelo of the Greeks. “Natural History,” to quote Dr. Whewell's words,[17] “when systematically treated, excludes all that is historical, for it classes objects by their permanent and universal properties, and has nothing to do with the narration of particular or casual facts.” Now, if we consider the large number of tongues spoken in different parts of [pg 043] the world with all their dialectic and provincial varieties, if we observe the great changes which each of these tongues has undergone in the course of centuries, how Latin was changed into Italian, Spanish, Portuguese, Provençal, French, Wallachian, and Roumansch; how Latin again, together with Greek, and the Celtic, the Teutonic, and Slavonic languages, together likewise with the ancient dialects of India and Persia, must have sprung from an earlier language, the mother of the whole Indo-European or Aryan family of speech; if we see how Hebrew, Arabic, and Syriac, with several minor dialects, are but different impressions of one and the same common type, and must all have flowed from the same source, the original language of the Semitic race; and if we add to these two, the Aryan and Semitic, at least one more well-established class of languages, the Turanian, comprising the dialects of the nomad races scattered over Central and Northern Asia, the Tungusic, Mongolic, Turkic,[18] Samoyedic, and Finnic, all radii from one common centre of speech:—if we watch this stream of language rolling on through centuries in these three mighty arms, which, before they disappear from our sight in the far distance, clearly show a convergence towards one common source: it would seem, indeed, as if there were an historical life inherent in language, and as if both the will of man and the power of time could tell, if not on its substance, at least on its form. And even if the mere local varieties of speech were not considered sufficient ground for excluding language from the domain of natural science, there would still remain the greater [pg 044] difficulty of reconciling with the recognized principles of physical science the historical changes affecting every one of these varieties. Every part of nature, whether mineral, plant, or animal, is the same in kind from the beginning to the end of its existence, whereas few languages could be recognized as the same after the lapse of but a thousand years. The language of Alfred is so different from the English of the present day that we have to study it in the same manner as we study Greek and Latin. We can read Milton and Bacon, Shakespeare and Hooker; we can make out Wycliffe and Chaucer; but, when we come to the English of the thirteenth century, we can but guess its meaning, and we fail even in this with works previous to the Ormulum and Layamon. The historical changes of language may be more or less rapid, but they take place at all times and in all countries. They have reduced the rich and powerful idiom of the poets of the Veda to the meagre and impure jargon of the modern Sepoy. They have transformed the language of the Zend-Avesta and of the mountain records of Behistún into that of Firdusi and the modern Persians; the language of Virgil into that of Dante, the language of Ulfilas into that of Charlemagne, the language of Charlemagne into that of Goethe. We have reason to believe that the same changes take place with even greater violence and rapidity in the dialects of savage tribes, although, in the absence of a written literature, it is extremely difficult to obtain trustworthy information. But in the few instances where careful observations have been made on this interesting subject, it has been found that among the wild and illiterate tribes of Siberia, Africa, and Siam, two or three generations are [pg 045] sufficient to change the whole aspect of their dialects. The languages of highly civilized nations, on the contrary, become more and more stationary, and seem sometimes almost to lose their power of change. Where there is a classical literature, and where its language is spread to every town and village, it seems almost impossible that any further changes should take place. Nevertheless, the language of Rome, for so many centuries the queen of the whole civilized world, was deposed by the modern Romance dialects, and the ancient Greek was supplanted in the end by the modern Romaic. And though the art of printing and the wide diffusion of Bibles, and Prayer-books, and newspapers have acted as still more powerful barriers to arrest the constant flow of human speech, we may see that the language of the authorized version of the Bible, though perfectly intelligible, is no longer the spoken language of England. In Booker's Scripture and Prayer-book Glossary[19] the number of words or senses of words which have become obsolete since 1611, amount to 388, or nearly one fifteenth part of the whole number of words used in the Bible. Smaller changes, changes of accent and meaning, the reception of new, and the dropping of old words, we may watch as taking place under our own eyes. Rogers[20] said that “cóntemplate is bad enough, but bálcony makes me sick,” whereas at present no one is startled by cóntemplate instead of contémplate, and bálcony has become more usual than balcóny. Thus Roome and chaney, layloc and goold, have but lately been driven from the stage by Rome, china, [pg 046] lilac, and gold, and some courteous gentlemen of the old school still continue to be obleeged instead of being obliged. Force,[21] in the sense of a waterfall, and gill, in the sense of a rocky ravine, were not used in classical English before Wordsworth. Handbook,[22] though an old Anglo-Saxon word, has but lately taken the place of manual, and a number of words such as cab for cabriolet, buss for omnibus, and even a verb such as to shunt tremble still on the boundary line between the vulgar and the literary idioms. Though the grammatical changes that have taken place since the publication of the authorized version are yet fewer in number, still we may point out some. The termination of the third person singular in th is now entirely replaced by s. No one now says he liveth, but only he lives. Several of the irregular imperfects and participles have assumed a new form. No one now uses he spake, and he drave, instead of he spoke, and he drove; holpen is replaced by helped; holden by held; shapen by shaped. The distinction between ye and you, the former being reserved for the nominative, the latter for all the other cases, is given up in modern English; and what is apparently a new grammatical form, the possessive pronoun its, has sprung into life since the beginning of the seventeenth century. It never occurs in the Bible; and though it is used three or four times by Shakespeare, Ben Jonson does not recognize it as yet in his English Grammar.[23]

It is argued, therefore, that as language, differing thereby from all other productions of nature, is liable to historical alterations, it is not fit to be treated in the [pg 047] same manner as the subject-matter of all the other physical sciences.

There is something very plausible in this objection, but if we examine it more carefully, we shall find that it rests entirely on a confusion of terms. We must distinguish between historical change and natural growth. Art, science, philosophy, and religion all have a history; language, or any other production of nature, admits only of growth.

Let us consider, first, 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. As man is the lord of nature only if he knows her laws and submits to them, the poet and the philosopher become the lords of language only if they know its laws and obey them.

When the Emperor Tiberius had made a mistake, and was reproved for it by Marcellus, another grammarian of the name of Capito, who happened to be present, remarked that what the emperor said was good Latin, or, if it were not, it would soon be so. Marcellus, more of a grammarian than a courtier, replied, “Capito is a liar; for, Cæsar, thou canst give the Roman citizenship to men, but not to words.” A similar anecdote is told of the German Emperor Sigismund. When presiding at the Council of Costnitz, he addressed the assembly in a Latin speech, exhorting them to eradicate the schism of the Hussites. “Videte Patres,” he said, “ut eradicetis schismam Hussitarum.” He was very unceremoniously called [pg 048] to order by a monk, who called out, “Serenissime Rex, schisma est generis neutri.”[24] The emperor, however, without losing his presence of mind, asked the impertinent monk, “How do you know it?” The old Bohemian school-master replied, “Alexander Gallus says so.” “And who is Alexander Gallus?” the emperor rejoined. The monk replied, “He was a monk.” “Well,” said the emperor, “and I am Emperor of Rome; and my word, I trust, will be as good as the word of any monk.” No doubt the laughers were with the emperor; but for all that, schisma remained a neuter, and not even an emperor could change its gender or termination.

The idea that language can be changed and improved by man is by no means a new one. We know that Protagoras, an ancient Greek philosopher, after laying down some laws on gender, actually began to find fault with the text of Homer, because it did not agree with his rules. But here, as in every other instance, the attempt proved unavailing. Try to alter the smallest rule of English, and you will find that it is physically impossible. There is apparently a very small difference between much and very, but you can hardly ever put one in the place of the other. You can say, “I am very happy,” but not “I am much happy,” though you may say “I am most happy.” On the contrary, you can say “I am much misunderstood,” but not “I am very misunderstood.” Thus the western Romance dialects, Spanish and Portuguese, together [pg 049] with Wallachian, can only employ the Latin word magis for forming comparatives:—Sp. mas dulce; Port. mais doce; Wall, mai dulce; while French, Provençal, and Italian only allow of plus for the same purpose: Ital. più dolce; Prov. plus dous; Fr. plus doux. It is by no means impossible, however, that this distinction between very, which is now used with adjectives only, and much, which precedes participles, should disappear in time. In fact, “very pleased” and “very delighted” are Americanisms which may be heard even in this country. But if that change take place, it will not be by the will of any individual, nor by the mutual agreement of any large number of men, but rather in spite of the exertions of grammarians and academies. And here you perceive the first difference between history and growth. An emperor may change the laws of society, the forms of religion, the rules of art: it is in the power of one generation, or even of one individual, to raise an art to the highest pitch of perfection, while the next may allow it to lapse, till a new genius takes it up again with renewed ardor. In all this we have to deal with the conscious acts of individuals, and we therefore move on historical ground. If we compare the creations of Michael Angelo or Raphael with the statues and frescoes of ancient Rome, we can speak of a history of art. We can connect two periods separated by thousands of years through the works of those who handed on the traditions of art from century to century; but we shall never meet with that continuous and unconscious growth which connects the language of Plautus with that of Dante. The process through which language is settled and unsettled combines in one the two opposite [pg 050] elements of necessity and free will. Though the individual seems to be the prime agent in producing new words and new grammatical forms, he is so only after his individuality has been merged in the common action of the family, tribe, or nation to which he belongs. He can do nothing by himself, and the first impulse to a new formation in language, though given by an individual, is mostly, if not always, given without premeditation, nay, unconsciously. The individual, as such, is powerless, and the results apparently produced by him depend on laws beyond his control, and on the co-operation of all those who form together with him one class, one body, or one organic whole.

But, though it is easy to show, as we have just done, that language cannot be changed or moulded by the taste, the fancy, or genius of man, it is very difficult to explain what causes the growth of language. Ever since Horace it has been usual to compare the growth of languages with the growth of trees. But comparisons are treacherous things. What do we know of the real causes of the growth of a tree, and what can we gain by comparing things which we do not quite understand with things which we understand even less? Many people speak, for instance, of the terminations of the verb, as if they sprouted out from the root as from their parent stock.[25] But what ideas can they connect with such expressions? If we must compare language with a tree, there is one point which may be illustrated by this comparison, and this is that neither language nor the tree can exist or grow by itself. Without the soil, without air and light, the tree could not live; it could not even be conceived to live. It is the same [pg 051] with language. Language cannot exist by itself; it requires a soil on which to grow, and that soil is the human soul. To speak of language as a thing by itself, as living a life of its own, as growing to maturity, producing offspring, and dying away, is sheer mythology; and though we cannot help using metaphorical expressions, we should always be on our guard, when engaged in inquiries like the present, against being carried away by the very words which we are using.

Now, what we call the growth of language comprises two processes which should be carefully distinguished, though they may be at work simultaneously. These two processes I call,