Popular Science Monthly/Volume 29/May 1886/The Evolution of Language
EVEN if the study of words, as it is carried on by the method of the natural sciences, did not furnish evidence that all language is traceable back to primordial monosyllabic elements, observation of the language-processes in children would lead to that conclusion. Gestures and physiognomical motions preceded language proper, or articulate language; and on this point it is of interest to compare man with the monkeys, which are able to express a considerable variety of feelings by the play of the muscles of the forehead and the eyebrows, the lips, nose, and jaws. If asked on what vocalization depends, we should answer that it depends solely on a particular sensation being stronger than others. With the infant, voice is provoked at first by some uneasiness or suffering; and it is not till a later period that it responds to a feeling of comfort and satisfaction. But in either case the first emissions have nothing intentional about them, and there is no link of volition between the feeling and the vocal manifestation of it. The time comes at last when the child, beginning to perceive what is going on around him, remarks that they always come to his help when he has committed the act of utterance; and he has from that time learned by experiment the use of his vocal power. He employs it at first in a very general and vague way; but, as he is taught by experience, he learns to exercise it more precisely, more in accordance with his volition, and to adapt the vocal emission to the results he wishes to bring about. He also perceives the greater facility of expression it gives him, and so goes on developing his precious faculty as he continues to exercise it. Tylor has clearly brought out the fact that savages have in a high degree the power of expressing their ideas directly by emotional tones. These tones, or interjections, are the first elements of grammatical language. The same author has also remarked another fact, that children not more than three or four years old, for example, are wont to observe the play of features, attitude, and gestures of the person who is speaking to them, in order to get the exact sense of the words which they hear.
We mention here, without dwelling upon it, that the faculty of language stands in close relation with a certain one of the frontal convolutions of the brain, which the inferior monkeys do not possess and which is found in a rudimentary state in the anthropoids, but the full acquisition and most complete development of which have made man, what he is, the master of articulate speech.
We thus perceive that the study of language belongs to the domain of the natural sciences. The objections that have been made to this view have little force with us. The first of them is that language is not transmitted with the blood. This confounds the transmission of the art of speech with that of the faculty of language. The faculty is hereditarily transmitted; it is intimately related to the cerebral development, and goes down with the structure, nature^ and qualities of the brain. As to the way in which the transmitted organ shall perform its functions, the parents of the child are there to stimulate and direct it, and to teach their offspring how to use the faculty it has inherited from them. We must not confound the faculty with the use that may be made of it. That use is an art, which the child acquires by tradition. But, we repeat, in the period of formation of a language, sonorous expression is only the more intense formulation of an emotion, usually associated with mimicry, the general attitude and face-play, a formulation which has the advantage of being more striking to strangers. In any case, it originally required to be complemented by gesture; and peoples little advanced in civilization may still be cited, among whom conversation is difficult in the dark, where mimicry can not be brought in to aid it. Bon wick relates that the Tasmanians had to recur to gestures and signs to establish the exact sense of their words; and Spix and Martins say the same of some of the savages of South America, and Cranz of the Greenlanders. These observations are far from being the only ones that have been made to the same effect.
A sound reason for including the study of language among the natural sciences lies in the fact that no man or group of men is competent arbitrarily to change the structure of its language. Fashion may sometimes admit particular words or banish others, but that has nothing to do with the structure. The morphological evolution of language defies all convention, all encroachment; it goes on by virtue of its own force, more or less slowly or speedily, but without the fancy or the pleasure of men having any power to divert it from its course. In short, we must avoid confounding changes in the vocabulary with linguistic, or, as we might call them, morphological changes. Among some Polynesian peoples, words are sometimes abolished; they will cease, for example, to employ in conversation the syllables that occur in the name of a chief; some people of the Bantu race will not pronounce any syllable that is found in the name of a near male relative. But these are special usages, temporary fashions, and have nothing whatever to do with the structure of the language. Then, again, we witness the creation of new words every day, but these words are always formed according to analogies with already existing words. They may be happy inventions or awkward attempts, but they are never pure creations or wholly fanciful.
A second objection to the classification of linguistic among scientific studies rests upon the fact that whole peoples, and even races, are capable of abandoning their own language and adopting another. The fact is undeniable; but it is also undeniable that language is independent of history; and, to take one example among many, we have seen Latin go on in its evolution in Gaul, Spain, and Roumania, after having been adopted by the barbarians.
It is proper to say something here about so-called mixed languages, which are, however, not at all hybrid in their structure, but have simply admitted foreign words into their vocabularies. With all the Persian and Arabic words it contains, the Turkish language is evidently and only Altaic. The Araucanian language, although it has received a host of Spanish words, is a purely American idiom. English is Germanic, although its vocabulary is loaded with words of Latin origin. The French language was introduced into England by the Norman conquest in the eleventh century. From the two languages which were then found in the presence of one another, the Anglo-Saxon and the French, it has been usually said that a mixed language was formed the—English. This assertion is very inexact, from the morphological point of view. French, after the conquest, became the language of the court and of justice, while it entered into the popular language, the Anglo-Saxon, only as to its vocabulary; but there it made a deep impression. Of 43,000 words in the English language, as they occur in the dictionary, more than 29,000 are of Roman origin, while only 13,000 or 14,000 are of Germanic origin, or Anglo-Saxon; yet the English language is wholly Germanic in its structure. The remains of the declensions of nouns and of the conjugations of verbs are Germanic, with no Latin about them. Another example of the kind is found in the Basque language, three quarters of the vocabulary of which is to-day Romanic; yet the fact does not prevent the language from having a peculiarly individual structure and form wholly free from Romanic elements in its grammar.
In short, the processes of linguistic study—which have nothing in common with those of the study of philology—demonstrate that the linguist studies the anatomy of forms just as the botanist and zoölogist do.
Another objection to the scientific view of linguistics is more specious, but not more solid. It is that, since articulate language can not be produced without vocal organs, it can not be regarded as an independent organism; besides, the sounds or vocal emissions do not become a language till they acquire significance by means of an operation that escapes us. It is easy to answer to this that, while language is in relation with a mental operation, it nevertheless constitutes a fact which is perceived by a sense—the sense of hearing. Of course, it is only in the abstract sense that we can regard language as an organism, but there is no doubt that in reality it behaves like an organism, and that it is in a constant state of evolution. And it is to this condition of evolution that I invite attention.
The phases of this evolution, as we understand it, are those of formation, growth, maturity, and decay. The variation is continual. Languages arise, are developed, pass on to decadence, and perish, like other organized beings. That their historical development is modified in the course of ages, according to certain conditions, is incontestable; but the observer of these modifications never sees in them anything other than phenomena of natural evolution. The evident proof of this fact is that the evolution is, as a whole, the same in linguistic families essentially different from one another.
Abel Rémusat has, in his "Recherches sur les langues tartares," indicated the general nature of the evolution of idioms: "In studying them attentively," he says, "we are tempted to believe that they are as constant in their march as the physical constitution that gives origin to them. . . . Possibly there prevails in languages less of the arbitrary than we have been accustomed to suppose; and, if we bring to their study the necessary care, we may be able to find in them signs as sure, as pronounced, and as characteristic as 'those which we can deduce from physiognomy, the color of the skin, or any other physical and external peculiarity." This "necessary care has been carried into the study of languages, and we shall see to what conclusions it has led us.
We are not acquainted with any language in its embryonic condition, if such a term is admissible. All of the languages submitted to our direct observation, even those of the most primitive stage, have passed the period of formation, which was prehistoric, and are now in the historical period, and generally in their decay. But by methodically separating and comparing their formative elements we can put ourselves, as it were, into the period of their formation.
The result of such comparative researches has confirmed the theory proposed in 1818 by William Schlegel, that languages first passed through a monosyllabic period; that some of them rose to the stage of development called the agglutinative; and that a small number of these last reach a later stage of flexion. The structure of the languages of the first class is simple, that of the second class is complex, and that of the third class is still more complex.
In the first phase of language, the root and the word are one, and each word-root or root-word is monosyllabic. The phrase is therefore a pure and simple succession of isolated roots. It is evident that the first process of elocution was of this character. Expression was found in uttering, one after another, monosyllables which were sometimes undoubtedly onomatopoetic—imitations of noises, sounds, and cries.
Existing monosyllabic languages have singularly improved upon this primitive process, while they have still remained monosyllabic. They have not created grammar, there being no structure in their words, but they have created a syntax. This syntax consists in the position in the phrase given to the different root-words. The place which any monosyllable occupies in the phrase determines the meaning of that monosyllable. The same process of syntactical arrangement comes back into use in the existing analytical languages that are most advanced in decadence. When, for example, we say in English, "Peter likes John," we are obliged to put the word Peter at the beginning of the phrase, and John at the end; for both words have lost every morphological distinction that could show which of them is the subject and which the object. It is not so in the synthetical languages in which the subject and the object are distinguished by the form of the word, and position in the phrase is of little importance. Thus, to say in Latin that the Helvetians sent legates, we say indifferently, Helvetii legatos miserunt, or Legatos miserunt Helvetii; the form in which the two nouns are put defining their respective functions.
In Chinese, the root which is to be the subject, or nominative in a phrase, takes its place before the root that has the significance of a verb. By thus assigning to the subject-word a fixed place in the phrase, the want of the grammatical elements which in Greek and Latin characterize the nominative case is obviated. In a monosyllabic language, in short, there is no grammar; there are no substantive forms, no verbal forms, or declensions, or conjugations, or gender, moods, or tenses, nothing but syntax, or "putting together." This, moreover, is what we shall more easily grasp in studying the transition from monosyllablism to agglutination, or the passage from the first to the second linguistic phase.
This transition or evolution takes place in a very simple way. Some word-roots abdicate a part of their meaning and become simple elements of relation, while others retain their full and independent signification. In Chinese and in other existing monosyllabic languages, we find this division of words into "full" words (which we may translate into English by a noun or a verb) and "vacant" words, the primary sense of which has gradually become obscured, and which have come to define more exactly or limit the broad sense of the "full" words. It is an interesting fact that a similar process has been employed at a much later stage in languages which have reached a high degree of development. Thus, in Latin, besides the word circus, a circle, we find circum, around, a kind of vacant word, denoting only relation; and examples of the kind might be multiplied in that language. So in Chinese, a monosyllabic language, the word for with, the sign of the instrumental case ("with the arm," "with a stick") is simply the root which when a "full" word signifies to make use of.
In the monosyllabic languages, the full words and the vacant words follow one another without ever amalgamating; that is, the roots are always isolated from each other, and there is never a word of several syllables. It is true that we can form something like compounds by bringing two words together, but without uniting them. Thus, in Chinese, the words fú, father, and mù, mother, brought together under the form of fú-mù, signify parents; and in the same way the words for "far" and "near" are made to signify distance. But there is nothing of derivation in this. Neither of the two words serves as an element of relation to the other, but each keeps all of its personality.
A step further is taken at a certain moment of linguistic development. The word indicating relation, the vacant word, is joined to the full word, and a polysyllabic form arises. A new word is formed, consisting of something else than a simple root, by the agglomeration of different elements, and we are in a secondary or agglutinative stage. We have no longer two full words juxtaposed to form a composite word; but an annexation to the principal word of a word playing the part of a secondary derivative and defining the relations of the root to which it is joined. When this derivative element is placed after the radical form, it is called a suffix; when it is placed at the beginning of the word, it is a prefix. Sometimes it is intercalated in the body of the word, and is then called an infix; but that method of derivation is rare.
It may be added that there are no limitations to derivation. The derived word may be the beginning of a second compound, and this of a third, and so on. Thus, in Magyar, the derivative zárat means he causes to shut, and zárhat he can shut; then, by a secondary derivation, we form zárathat, he can cause to shut. In like manner, záratgat, he causes to shut often, is a secondary, and záratgathat, he can cause to shut often, is a tertiary derivative. The languages of the third period of evolution, Latin for example, present a considerable number of these secondary and tertiary derivatives. The Latin word pater, father, is a primary derivative, of which the full or radical element is pa and the limiting element is ter. Paternus, whence our paternal, is a secondary, and paternitas, corresponding with our paternity, is a tertiary derivative. But our languages have not the extraordinary facility in derivation possessed by some simply agglutinative idioms. Thus, in the Turkish language, a single word may be made to introduce an indefinite number of ideas: as, sèvmèk, to love; sèvmèmèk, not to love; sèvilmèk, to be loved; sèvilmèmèk, not to be loved; sèvdirmèk, to make to love; sèvdirmèmèk, not to make to love; sèvinmèk, to love one's self; and so on, in which the derivative elements indicate, in the various forms, negation, causation, the reflexive quality, and other ideas, which in our language have to be expressed by separate words.
The larger number of languages are in the secondary or agglutinative stage. Among them are the negro, Malay, Polynesian, Dravidian, Altaic, Basque, and American languages or families of languages. But community of structure is no sign of relationship; it only indicates that two or more languages are in the same stage of evolution.
Some languages have made but little progress in agglutination, while others have advanced a great way in it. Some of the Western African negro languages still use, with agglutinative forms, processes that appertain to the monosyllabic structure. These are not cases of return to ancient forms, but are survivals of ancient forms in the midst of more complex formations. Some idioms, also, perpetually betray the evidences of the passage from monosyllablism to agglutination. Such languages have no literary value, and are not at all prominent; but they are like those obscure vegetable or animal species which are frequently richer in facts for the botanist or zoologist than other species that are usually esteemed much more useful or beautiful.
It is not quite so easy to explain the phenomena of the evolution from agglutination to flexion. The principle by which the evolution takes place is that of a phonic modification of the root. In the Indo-European languages, among which are included the Sanskrit, Persian, Greek, Latin, etc., evolution took place, according to M. Victor Henry, not only in this way, but by an agglutination of infixes also. But this point is not yet cleared up.
If we consider the ancient languages of the Indo-European family—Sanskrit, Greek, and Latin—we shall find that they are in different degrees synthetic; while, if we examine the characters of the modern branches of the family, we shall discover that they are analytical. This effect is the work of linguistic decadence, which has been less rapid in the Slavic languages than in the Germanic, in the Germanic than in the Romanic languages.
This decadence, which constitutes a new phase of evolution, is not brought about by chance. Regarding it phonetically, we see in it the results of the least effort. Diphthongs are condensed, as when in Latin veicos and deivos become vicus and deus. Assimilation takes place among the consonants, as when noctem, night, becomes notte, or septem, seven, sette, or when the earlier s-sound is softened into a simple aspirate. A considerable number of phonetic variations, which baffle persons not familiar with linguistic studies, are justified by comparison with other words.
Grammatical decadence also corresponds with a simplification. The ancient Indo-European language, of which the comparison of the Sanskrit, Greek, Latin, Slavic, and Germanic languages has enabled us to restore the important forms, possessed a rich system of declensions. Latin lost a part of its cases, and had of others only vestiges. Old French went a step further, and only kept two cases, the subject and the object cases; and even this greatly simplified declension disappeared in the fourteenth century, and the French language became wholly analytical, yet not without preserving traces of the two cases of the middle ages in the double forms of some of its words.
The simplification of declension appears in all modern languages. In Persian there is, properly, no declension. The dative and accusative are expressed by adding prepositions to the noun, the genitive by syntactical arrangement. Modern Greek has lost the forms of the dual number and of the dative case. Among the Semitic languages, current or spoken Arabic has dropped the terminations by which the cases are distinguished in literary Arabic. In vulgar Arabic the cases are distinguished by the position of the words or the use of prepositions. The same analytical phenomena may be observed in the conjugations. In the original Indo-European system, the perfect was formed by the reduplication of the root. Latin formed its perfects by compositions of words in which the auxiliaries were partly disguised as terminations, and in modern languages the analytical process has been further carried out. The same process is going on in the future tenses, which in English have reached the ultimate stage of it. Decadence sometimes proceeds by the primary value of a form or a word being forgotten. French affords some curious examples of this. Take the words luette, uvula, and lierre, ivy, which are from the Latin uveta, hedera. In old French they were written uette, hierre. When the article was prefixed they appeared as l'uette, l'hierre. Then the meaning of the article was forgotten or misconceived, and it was written as a part of the words. It then had to be supplied again, and so the French say now la luette, la lierre. This deformation took place naturally and without intention.
I come now to speak of the struggle for existence which is constantly going on between languages geographically near to one another and between different dialects of the same language. Unless one of the idioms is specially favored in the struggle by political circumstances, it is evident that the one which is most advanced in evolution will gain upon those which are less advanced: this fact can be established by many examples. Thus, in the territory which is now France, Latin, introduced into Gaul by a relatively small number of persons, shortly surpassed the Celtic dialects. The French language is wholly Latin, having retained from the Celtic only a few recollections in its vocabulary; but, when the Germans established themselves in a large part of Gaul, instead of giving their language to the conquered population, they abandoned it in the end and adopted the neo-Latin, which afterward became French; and the French language is no more Germanic than it is Celtic. Natural selection has caused the disappearance of a considerable number of idioms. Languages which come into conflict are like groups of animals that have to struggle with one another for existence. They must gain upon their competitors, or resign themselves to disappear before them. Just as, in the contest for life and development, the best-armed races finally prevail over those which are less favored, so languages which are best served by their own aptitudes and by external circumstances prevail over those whose evolutive force is less considerable, and over those which historical conditions have less well prepared for the combat. In France, the French, the ancient langue d'oil gradually supplanted the langue d'oc, the Corsican, the Breton, the Flemish, and the Basque. In the British Islands, English eclipsed the Celtic languages, Irish, Scotch, Manx, and Gaelic, and will shortly have supplanted the Cornish. German has overcome a number of Slavic idioms.
Another kind of selection is going on within the language itself with reference to the use of particular forms and words. In reference to this, the study of dialects is of great interest. Dialects should not be regarded as degenerate conditions of literary languages. These languages are simply fortunate dialects, whose rival dialects have been less favored. We are constantly meeting in dialects forms and words which their sister literary languages have not preserved; and this fact gives dialects an important place in the study of the natural history of language.
The fact that some idioms have been lost has been disadvantageous to linguistic studies because intermediate forms have thereby disappeared, the existence of which would have explained many living forms. In this, again, we have presented in language something comparable to what has taken place among animals and plants. Moreover, a linguistic species, once extinct, can never be brought back to life. It has been only a little while since the Tasmanians disappeared, and their language with them. Those people who were the product of a long ethnic evolution can never be brought back; no more can a language like theirs, which was also the product of a long development, be revived. So in the world of animals and plants, the disappearance of a species is always definitive; to bring it back to a new life would require the impossible return of the conditions of every kind which had brought it up to the stage which it had reached at the moment of its extinction.
I should be satisfied if I could believe that this review, perhaps too rapid, has made evident the interesting fact of the life and evolution of languages. To say life of language does not seem sufficient, for that word only gives the idea of a simple state of activity. The word evolution is more rigorously exact. We find ourselves, in fact, in the presence of successive developments of an entirely natural order. The organic perfectionment of the brain gives to the highest of the primates the faculty of articulate speech; that faculty, brought into play, gives rise to an extremely rudimentary system of expression, the source of which, as Lucretius has observed with much force and justice, lies in an imperious need. This need is, in fact, the creator of words. Gradually the monosyllabic words become differentiated into principal words and words of secondary signification. A new phase begins with the closer association of words, and the different processes of derivation develop themselves more and more. The third phase is characterized at first by a remarkable synthetic process, which soon, however, undergoes simplification, and yields under the influence of a more rapid civilization to a more and more accentuated analytical precision. The ultimate form has evidently not yet been reached by the English and French languages; but since language was born with man, and is his single characteristic, though laboriously and slowly developed as all his powers have expanded, it is destined to be transformed into more and more perfect forms of expression as man himself continues to ascend in the scale of superiority.—Translated for the Popular Science Monthly from the Revue Scientifique.