[1]

[2]

LANGUAGE
ITS NATURE
DEVELOPMENT
AND ORIGIN

[3]

BY
OTTO JESPERSEN
PROFESSOR IN THE UNIVERSITY OF COPENHAGEN

LONDON: GEORGE ALLEN & UNWIN LTD.
RUSKIN HOUSE, 40 MUSEUM STREET, W.C. 1

[4]

First published in 1922

(All rights reserved)

[5]

TO
VILHELM THOMSEN

[6]

[7]

The distinctive feature of the science of language as conceived nowadays is its historical character: a language or a word is no longer taken as something given once for all, but as a result of previous development and at the same time as the starting-point for subsequent development. This manner of viewing languages constitutes a decisive improvement on the way in which languages were dealt with in previous centuries, and it suffices to mention such words as ‘evolution’ and ‘Darwinism’ to show that linguistic research has in this respect been in full accordance with tendencies observed in many other branches of scientific work during the last hundred years. Still, it cannot be said that students of language have always and to the fullest extent made it clear to themselves what is the real essence of a language. Too often expressions are used which are nothing but metaphors—in many cases perfectly harmless metaphors, but in other cases metaphors that obscure the real facts of the matter. Language is frequently spoken of as a ‘living organism’; we hear of the ‘life’ of languages, of the ‘birth’ of new languages and of the ‘death’ of old languages, and the implication, though not always realized, is that a language is a living thing, something analogous to an animal or a plant. Yet a language evidently has no separate existence in the same way as a dog or a beech has, but is nothing but a function of certain living human beings. Language is activity, purposeful activity, and we should never lose sight of the speaking individuals and of their purpose in acting in this particular way. When people speak of the life of words—as in celebrated books with such titles as La vie des mots, or Biographies of Words—they do not always keep in view that a word has no ‘life’ of its own: it exists only in so far as it is pronounced or heard or remembered by somebody, and this kind of existence cannot properly be compared with ‘life’ in the original and proper sense of that word. The only unimpeachable definition of a word is that it is a human habit, an habitual act on the part of one human individual which has, or may have, the effect of evoking some idea in the mind[8] of another individual. A word thus may be rightly compared with such an habitual act as taking off one’s hat or raising one’s fingers to one’s cap: in both cases we have a certain set of muscular activities which, when seen or heard by somebody else, shows him what is passing in the mind of the original agent or what he desires to bring to the consciousness of the other man (or men). The act is individual, but the interpretation presupposes that the individual forms part of a community with analogous habits, and a language thus is seen to be one particular set of human customs of a well-defined social character.

It is indeed possible to speak of ‘life’ in connexion with language even from this point of view, but it will be in a different sense from that in which the word was taken by the older school of linguistic science. I shall try to give a biological or biographical science of language, but it will be through sketching the linguistic biology or biography of the speaking individual. I shall give, therefore, a large part to the way in which a child learns his mother-tongue (Book II): my conclusions there are chiefly based on the rich material I have collected during many years from direct observation of many Danish children, and particularly of my own boy, Frans (see my book Nutidssprog hos börn og voxne, Copenhagen, 1916). Unfortunately, I have not been able to make first-hand observations with regard to the speech of English children; the English examples I quote are taken second-hand either from notes, for which I am obliged to English and American friends, or from books, chiefly by psychologists. I should be particularly happy if my remarks could induce some English or American linguist to take up a systematic study of the speech of children, or of one child. This study seems to me very fascinating indeed, and a linguist is sure to notice many things that would be passed by as uninteresting even by the closest observer among psychologists, but which may have some bearing on the life and development of language.

Another part of linguistic biology deals with the influence of the foreigner, and still another with the changes which the individual is apt independently to introduce into his speech even after he has fully acquired his mother-tongue. This naturally leads up to the question whether all these changes introduced by various individuals do, or do not, follow the same line of direction, and whether mankind has on the whole moved forward or not in linguistic matters. The conviction reached through a study of historically accessible periods of well-known languages is finally shown to throw some light on the disputed problem of the ultimate origin of human language.

Parts of my theory of sound-change, and especially my objections[9] to the dogma of blind sound-laws, date back to my very first linguistic paper (1886); most of the chapters on Decay or Progress and parts of some of the following chapters, as well as the theory of the origin of speech, may be considered a new and revised edition of the general chapters of my Progress in Language (1894). Many of the ideas contained in this book thus are not new with me; but even if a reader of my previous works may recognize things which he has seen before, I hope he will admit that they have been here worked up with much new material into something like a system, which forms a fairly comprehensive theory of linguistic development.

Still, I have not been able to compress into this volume the whole of my philosophy of speech. Considerations of space have obliged me to exclude the chapters I had first intended to write on the practical consequences of the ‘energetic’ view of language which I have throughout maintained; the estimation of linguistic phenomena implied in that view has bearings on such questions as these: What is to be considered ‘correct’ or ‘standard’ in matters of pronunciation, spelling, grammar and idiom? Can (or should) individuals exert themselves to improve their mother-tongue by enriching it with new terms and by making it purer, more precise, more fit to express subtle shades of thought, more easy to handle in speech or in writing, etc.? (A few hints on such questions may be found in my paper “Energetik der Sprache” in Scientia, 1914.) Is it possible to construct an artificial language on scientific principles for international use? (On this question I may here briefly state my conviction that it is extremely important for the whole of mankind to have such a language, and that Ido is scientifically and practically very much superior to all previous attempts, Volapük, Esperanto, Idiom Neutral, Latin sine flexione, etc. But I have written more at length on that question elsewhere.) With regard to the system of grammar, the relation of grammar to logic, and grammatical categories and their definition, I must refer the reader to Sprogets Logik (Copenhagen, 1913), and to the first chapter of the second volume of my Modern English Grammar (Heidelberg, 1914), but I shall hope to deal with these questions more in detail in a future work, to be called, probably, The Logic of Grammar, of which some chapters have been ready in my drawers for some years and others are in active preparation.

I have prefixed to the theoretical chapters of this work a short survey of the history of the science of language in order to show how my problems have been previously treated. In this part (Book I) I have, as a matter of course, used the excellent works on the subject by Benfey, Raumer, Delbrück (Einleitung in das Sprachstudium, 1st ed., 1880; I did not see the 5th ed., 1908, till[10] my own chapters on the history of linguistics were finished), Thomsen, Oertel and Pedersen. But I have in nearly every case gone to the sources themselves, and have, I think, found interesting things in some of the early books on linguistics that have been generally overlooked; I have even pointed out some writers who had passed into undeserved oblivion. My intention has been on the whole to throw into relief the great lines of development rather than to give many details; in judging the first part of my book it should also be borne in mind that its object primarily is to serve as an introduction to the problems dealt with in the rest of the book. Throughout I have tried to look at things with my own eyes, and accordingly my views on a great many points are different from those generally accepted; it is my hope that an impartial observer will find that I have here and there succeeded in distributing light and shade more justly than my predecessors.

Wherever it has been necessary I have transcribed words phonetically according to the system of the Association Phonétique Internationale, though without going into too minute distinction of sounds, the object being, not to teach the exact pronunciation of various languages, but rather to bring out clearly the insufficiency of the ordinary spelling. The latter is given throughout in italics, while phonetic symbols have been inserted in brackets [ ]. I must ask the reader to forgive inconsistency in such matters as Greek accents, Old English marks of vowel-length, etc., which I have often omitted as of no importance for the purpose of this volume.

I must express here my gratitude to the directors of the Carlsbergfond for kind support of my work. I want to thank also Professor G. C. Moore Smith, of the University of Sheffield: not only has he sent me the manuscript of a translation of most of my Nutidssprog, which he had undertaken of his own accord and which served as the basis of Book II, but he has kindly gone through the whole of this volume, improving and correcting my English style in many passages. His friendship and the untiring interest he has always taken in my work have been extremely valuable to me for a great many years.

OTTO JESPERSEN.

University of Copenhagen,
June 1921.

[11]

[13]

[16]

' stands before the stressed syllable.

· indicates length of the preceding sound.

[a·] as in alms.
[ai] as in ice.
[au] as in house.
[æ] as in hat.
[ei] as in hate.
[ɛ] as in care; Fr. tel.
[ə] indistinct vowels.
[i] as in fill; Fr. qui.
[i·] as in feel; Fr. fille.
[o] as in Fr. seau.
[ou] as in so.
[ɔ] open o-sounds.
[u] as in full; Fr. fou.
[u·] as in foorl; Fr. épouse.
[y] as in Fr. vu.
[ʌ] as in cut.
[ø] as in Fr. feu.
[œ] as in Fr. sœur.
[~] French nasalization.
[c] as in G. ich.
[x] as in G., Sc. loch.
[ð] as in this.
[j] as in you.
[þ] as in thick.
[ʃ] as in she.
[ʒ] as in measure.
[’] in Russian palatalization, in Danish glottal stop.

The science of language began, tentatively and approximately, when the minds of men first turned to problems like these: How is it that people do not speak everywhere the same language? How were words first created? What is the relation between a name and the thing it stands for? Why is such and such a person, or such and such a thing, called this and not that? The first answers to these questions, like primitive answers to other riddles of the universe, were largely theological: God, or one particular god, had created language, or God led all animals to the first man in order that he might give them names. Thus in the Old Testament the diversity of languages is explained as a punishment from God for man’s crimes and presumption. These were great and general problems, but the minds of the early Jews were also occupied with smaller and more particular problems of language, as when etymological interpretations were given of such personal names as were not immediately self-explanatory.

The same predilection for etymology, and a similar primitive kind of etymology, based entirely on a more or less accidental similarity of sound and easily satisfied with any fanciful connexion in sense, is found abundantly in Greek writers and in their Latin imitators. But to the speculative minds of Greek thinkers the problem that proved most attractive was the general and abstract one, Are words natural and necessary expressions of the notions underlying them, or are they merely arbitrary and conventional signs for notions that might have been equally well expressed by any other sounds? Endless discussions were carried on about this question, as we see particularly from Plato’s Kratylos, and no very definite result was arrived at, nor could any be expected so long as one language only formed the basis of the discussion—even in our own days, after a century of comparative philology, the question still remains an open one. In Greece, the two catchwords phúsei (by nature) and thései (by convention) for centuries[20] divided philosophers and grammarians into two camps, while some, like Sokrates in Plato’s dialogue, though admitting that in language as actually existing there was no natural connexion between word and thing, still wished that an ideal language might be created in which words and things would be tied together in a perfectly rational way—thus paving the way for Bishop Wilkins and other modern constructors of philosophical languages.

Such abstract and a priori speculations, however stimulating and clever, hardly deserve the name of science, as this term is understood nowadays. Science presupposes careful observation and systematic classification of facts, and of that in the old Greek writers on language we find very little. The earliest masters in linguistic observation and classification were the old Indian grammarians. The language of the old sacred hymns had become in many points obsolete, but religion required that not one iota of these revered texts should be altered, and a scrupulous oral tradition kept them unchanged from generation to generation in every minute particular. This led to a wonderfully exact analysis of speech sounds, in which every detail of articulation was carefully described, and to a no less admirable analysis of grammatical forms, which were arranged systematically and described in a concise and highly ingenious, though artificial, terminology. The whole manner of treatment was entirely different from the methods of Western grammarians, and when the works of Panini and other Sanskrit grammarians were first made known to Europeans in the nineteenth century, they profoundly influenced our own linguistic science, as witnessed, among other things, by the fact that some of the Indian technical terms are still extensively used, for instance those describing various kinds of compound nouns.

In Europe grammatical science was slowly and laboriously developed in Greece and later in Rome. Aristotle laid the foundation of the division of words into “parts of speech” and introduced the notion of case (ptôsis). His work in this connexion was continued by the Stoics, many of whose grammatical distinctions and terms are still in use, the latter in their Latin dress, which embodies some curious mistakes, as when genikḗ, “the case of kind or species,” was rendered genitivus, as if it meant “the case of origin,” or, worse still, when aitiatikḗ, “the case of object,” was rendered accusativus, as if from aitiáomai, ‘I accuse.’ In later times the philological school of Alexandria was particularly important, the object of research being the interpretation of the old poets, whose language was no longer instantly intelligible. Details of flexion and of the meaning of words were described and referred to the two categories of analogy or regularity and anomaly or irregularity, but real insight into the nature of language[21] made very little progress either with the Alexandrians or with their Roman inheritors, and etymology still remained in the childlike stage.

Nor did linguistic science advance in the Middle Ages. The chief thing then was learning Latin as the common language of the Church and of what little there was of civilization generally; but Latin was not studied in a scientific spirit, and the various vernacular languages, which one by one blossomed out into languages of literature, even less so.

The Renaissance in so far brought about a change in this, as it widened the horizon, especially by introducing the study of Greek. It also favoured grammatical studies through the stress it laid on correct Latin as represented in the best period of classical literature: it now became the ambition of humanists in all countries to write Latin like Cicero. In the following centuries we witness a constantly deepening interest in the various living languages of Europe, owing to the growing importance of native literatures and to increasing facilities of international traffic and communication in general. The most important factor here was, of course, the invention of printing, which rendered it incomparably more easy than formerly to obtain the means of studying foreign languages. It should be noted also that in those times the prevalent theological interest made it a much more common thing than nowadays for ordinary scholars to have some knowledge of Hebrew as the original language of the Old Testament. The acquaintance with a language so different in type from those spoken in Europe in many ways stimulated the interest in linguistic studies, though on the other hand it proved a fruitful source of error, because the position of the Semitic family of languages was not yet understood, and because Hebrew was thought to be the language spoken in Paradise, and therefore imagined to be the language from which all other languages were descended. All kinds of fanciful similarities between Hebrew and European languages were taken as proofs of the origin of the latter; every imaginable permutation of sounds (or rather of letters) was looked upon as possible so long as there was a slight connexion in the sense of the two words compared, and however incredible it may seem nowadays, the fact that Hebrew was written from right to left, while we in our writing proceed from left to right, was considered justification enough for the most violent transposition of letters in etymological explanations. And yet all these flighty and whimsical comparisons served perhaps in some measure to[22] pave the way for a more systematic treatment of etymology through collecting vast stores of words from which sober and critical minds might select those instances of indubitable connexion on which a sound science of etymology could eventually be constructed.

The discovery and publication of texts in the old Gothonic (Germanic) languages, especially Wulfila’s Gothic translation of the Bible, compared with which Old English (Anglo-Saxon), Old German and Old Icelandic texts were of less, though by no means of despicable, account, paved the way for historical treatment of this important group of languages in the seventeenth and eighteenth centuries. But on the whole, the interest in the history of languages in those days was small, and linguistic thinkers thought it more urgent to establish vast treasuries of languages as actually spoken than to follow the development of any one language from century to century. Thus we see that the great philosopher Leibniz, who took much interest in linguistic pursuits and to whom we owe many judicious utterances on the possibility of a universal language, instigated Peter the Great to have vocabularies and specimens collected of all the various languages of his vast empire. To this initiative taken by Leibniz, and to the great personal interest that the Empress Catherine II took in these studies, we owe, directly or indirectly, the great repertories of all languages then known, first Pallas’s Linguarum totius orbis vocabularia comparativa (1786-87), then Hervas’s Catálogo de las lenguas de las naziones conocidas (1800-5), and finally Adelung’s Mithridates oder allgemeine Sprachenkunde (1806-17). In spite of their inevitable shortcomings, their uncritical and unequal treatment of many languages, the preponderance of lexical over grammatical information, and the use of biblical texts as their sole connected illustrations, these great works exercised a mighty influence on the linguistic thought and research of the time, and contributed very much to the birth of the linguistic science of the nineteenth century. It should not be forgotten, moreover, that Hervas was one of the first to recognize the superior importance of grammar to vocabulary for deciding questions of relationship between languages.

It will be well here to consider the manner in which languages and the teaching of languages were generally viewed during the centuries preceding the rise of Comparative Linguistics. The chief language taught was Latin; the first and in many cases the only grammar with which scholars came into contact was Latin grammar. No wonder therefore that grammar and Latin grammar came in the minds of most people to be synonyms. Latin grammar played an enormous rôle in the schools, to the exclusion of many subjects (the pupil’s own native language, science, history, etc.)[23] which we are now beginning to think more essential for the education of the young. The traditional term for ‘secondary school’ was in England ‘grammar school’ and in Denmark ‘latinskole,’ and the reason for both expressions was obviously the same. Here, however, we are concerned with this privileged position of Latin grammar only in so far as it influenced the treatment of languages in general. It did so in more ways than one.

Latin was a language with a wealth of flexional forms, and in describing other languages the same categories as were found in Latin were applied as a matter of course, even where there was nothing in these other languages which really corresponded to what was found in Latin. In English and Danish grammars paradigms of noun declension were given with such cases as accusative, dative and ablative, in spite of the fact that no separate forms for these cases had existed for centuries. All languages were indiscriminately saddled with the elaborate Latin system of tenses and moods in the verbs, and by means of such Procrustean methods the actual facts of many languages were distorted and misrepresented. Discriminations which had no foundation in reality were nevertheless insisted on, while discriminations which happened to be non-existent in Latin were apt to be overlooked. The mischief consequent on this unfortunate method of measuring all grammar after the pattern of Latin grammar has not even yet completely disappeared, and it is even now difficult to find a single grammar of any language that is not here and there influenced by the Latin bias.

Latin was chiefly taught as a written language (witness the totally different manner in which Latin was pronounced in the different countries, the consequence being that as early as the sixteenth century French and English scholars were unable to understand each other’s spoken Latin). This led to the almost exclusive occupation with letters instead of sounds. The fact that all language is primarily spoken and only secondarily written down, that the real life of language is in the mouth and ear and not in the pen and eye, was overlooked, to the detriment of a real understanding of the essence of language and linguistic development; and very often where the spoken form of a language was accessible scholars contented themselves with a reading knowledge. In spite of many efforts, some of which go back to the sixteenth century, but which did not become really powerful till the rise of modern phonetics in the nineteenth century, the fundamental significance of spoken as opposed to written language has not yet been fully appreciated by all linguists. There are still too many writers on philological questions who have evidently never tried to think in sounds instead of thinking in letters and symbols,[24] and who would probably be sorely puzzled if they were to pronounce all the forms that come so glibly to their pens. What Sweet wrote in 1877 in the preface to his Handbook of Phonetics is perhaps less true now than it was then, but it still contains some elements of truth. “Many instances,” he said, “might be quoted of the way in which important philological facts and laws have been passed over or misrepresented through the observer’s want of phonetic training. Schleicher’s failing to observe the Lithuanian accents, or even to comprehend them when pointed out by Kurschat, is a striking instance.” But there can be no doubt that the way in which Latin has been for centuries made the basis of all linguistic instruction is largely responsible for the preponderance of eye-philology to ear-philology in the history of our science.

We next come to a point which to my mind is very important, because it concerns something which has had, and has justly had, enduring effects on the manner in which language, and especially grammar, is viewed and taught to this day. What was the object of teaching Latin in the Middle Ages and later? Certainly not the purely scientific one of imparting knowledge for knowledge’s own sake, apart from any practical use or advantage, simply in order to widen the spiritual horizon and to obtain the joy of pure intellectual understanding. For such a purpose some people with scientific leanings may here and there take up the study of some out-of-the-way African or American idiom. But the reasons for teaching and learning Latin were not so idealistic. Latin was not even taught and learnt solely with the purpose of opening the doors to the old classical or to the more recent religious literature in that language, but chiefly, and in the first instance, because Latin was a practical and highly important means of communication between educated people. One had to learn not only to read Latin, but also to write Latin, if one wanted to maintain no matter how humble a position in the republic of learning or in the hierarchy of the Church. Consequently, grammar was not (even primarily) the science of how words were inflected and how forms were used by the old Romans, but chiefly and essentially the art of inflecting words and of using the forms yourself, if you wanted to write correct Latin. This you must say, and these faults you must avoid—such were the lessons imparted in the schools. Grammar was not a set of facts observed but of rules to be observed, and of paradigms, i.e. of patterns, to be followed. Sometimes this character of grammatical instruction is expressly indicated in the form of the precepts given, as in such memorial verses as this: “Tolle -me, -mi, -mu, -mis, Si declinare domus vis!” In other words, grammar was prescriptive rather than descriptive.

[25]

The current definition of grammar, therefore, was “ars bene dicendi et bene scribendi,” “l’art de bien dire et de bien écrire,” the art of speaking and writing correctly. J. C. Scaliger said, “Grammatici unus finis est recte loqui.” To attain to correct diction (‘good grammar’) and to avoid faulty diction (‘bad grammar’), such were the two objects of grammatical teaching. Now, the same point of view, in which the two elements of ‘art’ and of ‘correctness’ entered so largely, was applied not only to Latin, but to other languages as well, when the various vernaculars came to be treated grammatically.

The vocabulary, too, was treated from the same point of view. This is especially evident in the case of the dictionaries issued by the French and Italian Academies. They differ from dictionaries as now usually compiled in being not collections of all and any words their authors could get hold of within the limits of the language concerned, but in being selections of words deserving the recommendations of the best arbiters of taste and therefore fit to be used in the highest literature by even the most elegant or fastidious writers. Dictionaries thus understood were less descriptions of actual usage than prescriptions for the best usage of words.

The normative way of viewing language is fraught with some great dangers which can only be avoided through a comprehensive knowledge of the historic development of languages and of the general conditions of linguistic psychology. Otherwise, the tendency everywhere is to draw too narrow limits for what is allowable or correct. In many cases one form, or one construction, only is recognized, even where two or more are found in actual speech; the question which is to be selected as the only good form comes to be decided too often by individual fancy or predilection, where no scientific tests can yet be applied, and thus a form may often be proscribed which from a less narrow point of view might have appeared just as good as, or even better than, the one preferred in the official grammar or dictionary. In other instances, where two forms were recognized, the grammarian wanted to give rules for their discrimination, and sometimes on the basis of a totally inadequate induction he would establish nice distinctions not really warranted by actual usage—distinctions which subsequent generations had to learn at school with the sweat of their brows and which were often considered most important in spite of their intrinsic insignificance. Such unreal or half-real subtle distinctions are the besetting sin of French grammarians from the ‘grand siècle’ onwards, while they have played a much less considerable part in England, where people have been on the whole more inclined to let things slide as best they may on the[26] ‘laissez faire’ principle, and where no Academy was ever established to regulate language. But even in English rules are not unfrequently given in schools and in newspaper offices which are based on narrow views and hasty generalizations. Because a preposition at the end of a sentence may in some instances be clumsy or unwieldy, this is no reason why a final preposition should always and under all circumstances be considered a grave error. But it is of course easier for the schoolmaster to give an absolute and inviolable rule once and for all than to study carefully all the various considerations that might render a qualification desirable. If the ordinary books on Common Faults in Writing and Speaking English and similar works in other languages have not even now assimilated the teachings of Comparative and Historic Linguistics, it is no wonder that the grammarians of the seventeenth and eighteenth centuries, with whom we are here concerned, should be in many ways guided by narrow and insufficient views on what ought to determine correctness of speech.

Here also the importance given to the study of Latin was sometimes harmful; too much was settled by a reference to Latin rules, even where the modern languages really followed rules of their own that were opposed to those of Latin. The learning of Latin grammar was supposed to be, and to some extent really was, a schooling in logic, as the strict observance of the rules of any foreign language is bound to be; but the consequence of this was that when questions of grammatical correctness were to be settled, too much importance was often given to purely logical considerations, and scholars were sometimes apt to determine what was to be called ‘logical’ in language according to whether it was or was not in conformity with Latin usage. This disposition, joined with the unavoidable conservatism of mankind, and more particularly of teachers, would in many ways prove a hindrance to natural developments in a living speech. But we must again take up the thread of the history of linguistic theory.

The problem of a natural origin of language exercised some of the best-known thinkers of the eighteenth century. Rousseau imagined the first men setting themselves more or less deliberately to frame a language by an agreement similar to (or forming part of) the contrat social which according to him was the basis of all social order. There is here the obvious difficulty of imagining how primitive men who had been previously without any speech came to feel the want of language, and how they could agree on what sound was to represent what idea without having already[27] some means of communication. Rousseau’s whole manner of putting and of viewing the problem is evidently too crude to be of any real importance in the history of linguistic science.

Condillac is much more sensible when he tries to imagine how a speechless man and a speechless woman might be led quite naturally to acquire something like language, starting with instinctive cries and violent gestures called forth by strong emotions. Such cries would come to be associated with elementary feelings, and new sounds might come to indicate various objects if produced repeatedly in connexion with gestures showing what objects the speaker wanted to call attention to. If these two first speaking beings had as yet very little power to vary their sounds, their child would have a more flexible tongue, and would therefore be able to, and be impelled to, produce some new sounds, the meaning of which his parents would guess at, and which they in their turn would imitate; thus gradually a greater and greater number of words would come into existence, generation after generation working painfully to enrich and develop what had been already acquired, until it finally became a real language.

The profoundest thinker on these problems in the eighteenth century was Johann Gottfried Herder, who, though he did little or nothing in the way of scientific research, yet prepared the rise of linguistic science. In his prize essay on the Origin of Language (1772) Herder first vigorously and successfully attacks the orthodox view of his age—a view which had been recently upheld very emphatically by one Süssmilch—that language could not have been invented by man, but was a direct gift from God. One of Herder’s strongest arguments is that if language had been framed by God and by Him instilled into the mind of man, we should expect it to be much more logical, much more imbued with pure reason than it is as an actual matter of fact. Much in all existing languages is so chaotic and ill-arranged that it could not be God’s work, but must come from the hand of man. On the other hand, Herder does not think that language was really ‘invented’ by man—although this was the word used by the Berlin Academy when opening the competition in which Herder’s essay gained the prize. Language was not deliberately framed by man, but sprang of necessity from his innermost nature; the genesis of language according to him is due to an impulse similar to that of the mature embryo pressing to be born. Man, in the same way as all animals, gives vent to his feelings in tones, but this is not enough; it is impossible to trace the origin of human language to these emotional cries alone. However much they may be refined and fixed, without understanding they can never become human, conscious language. Man differs from brute animals not in degree or in the addition of[28] new powers, but in a totally different direction and development of all powers. Man’s inferiority to animals in strength and sureness of instinct is compensated by his wider sphere of attention; the whole disposition of his mind as an unanalysable entity constitutes the impassable barrier between him and the lower animals. Man, then, shows conscious reflexion when among the ocean of sensations that rush into his soul through all the senses he singles out one wave and arrests it, as when, seeing a lamb, he looks for a distinguishing mark and finds it in the bleating, so that next time when he recognizes the same animal he imitates the sound of bleating, and thereby creates a name for that animal. Thus the lamb to him is ‘the bleater,’ and nouns are created from verbs, whereas, according to Herder, if language had been the creation of God it would inversely have begun with nouns, as that would have been the logically ideal order of procedure. Another characteristic trait of primitive languages is the crossing of various shades of feeling and the necessity of expressing thoughts through strong, bold metaphors, presenting the most motley picture. “The genetic cause lies in the poverty of the human mind and in the flowing together of the emotions of a primitive human being.” Another consequence is the wealth of synonyms in primitive language; “alongside of real poverty it has the most unnecessary superfluity.”

When Herder here speaks of primitive or ‘original’ languages, he is thinking of Oriental languages, and especially of Hebrew. “We should never forget,” says Edward Sapir,[1] “that Herder’s time-perspective was necessarily very different from ours. While we unconcernedly take tens or even hundreds of thousands of years in which to allow the products of human civilization to develop, Herder was still compelled to operate with the less than six thousand years that orthodoxy stingily doled out. To us the two or three thousand years that separate our language from the Old Testament Hebrew seems a negligible quantity, when speculating on the origin of language in general; to Herder, however, the Hebrew and the Greek of Homer seemed to be appreciably nearer the oldest conditions than our vernaculars—hence his exaggeration of their ursprünglichkeit.”

Herder’s chief influence on the science of speech, to my mind, is not derived directly from the ideas contained in his essay on the actual origin of speech, but rather indirectly through the whole of his life’s work. He had a very strong sense of the value of everything that had grown naturally (das naturwüchsige); he prepared the minds of his countrymen for the manysided recep[29]tiveness of the Romanticists, who translated and admired the popular poetry of a great many countries, which had hitherto been terræ incognitæ; and he was one of the first to draw attention to the great national value of his own country’s medieval literature and its folklore, and thus was one of the spiritual ancestors of Grimm. He sees the close connexion that exists between language and primitive poetry, or that kind of spontaneous singing that characterizes the childhood or youth of mankind, and which is totally distinct from the artificial poetry of later ages. But to him each language is not only the instrument of literature, but itself literature and poetry. A nation speaks its soul in the words it uses. Herder admires his own mother-tongue, which to him is perhaps inferior to Greek, but superior to its neighbours. The combinations of consonants give it a certain measured pace; it does not rush forward, but walks with the firm carriage of a German. The nice gradation of vowels mitigates the force of the consonants, and the numerous spirants make the German speech pleasant and endearing. Its syllables are rich and firm, its phrases are stately, and its idiomatic expressions are emphatic and serious. Still in some ways the present German language is degenerate if compared with that of Luther, and still more with that of the Suabian Emperors, and much therefore remains to be done in the way of disinterring and revivifying the powerful expressions now lost. Through ideas like these Herder not only exercised a strong influence on Goethe and the Romanticists, but also gave impulses to the linguistic studies of the following generation, and caused many younger men to turn from the well-worn classics to fields of research previously neglected.

Where questions of correct language or of the best usage are dealt with, or where different languages are compared with regard to their efficiency or beauty, as is done very often, though more often in dilettante conversation or in casual remarks in literary works than in scientific linguistic disquisitions, it is no far cry to the question, What would an ideal language be like? But such is the matter-of-factness of modern scientific thought, that probably no scientific Academy in our own days would think of doing what the Berlin Academy did in 1794 when it offered a prize for the best essay on the ideal of a perfect language and a comparison of the best-known languages of Europe as tested by the standard of such an ideal. A Berlin pastor, D. Jenisch, won the prize, and in 1796 brought out his book under the title Philosophisch-kritische vergleichung und würdigung von vierzehn ältern und neuern sprachen[30] Europens—a book which is even now well worth reading, the more so because its subject has been all but completely neglected in the hundred and twenty years that have since intervened. In the Introduction the author has the following passage, which might be taken as the motto of Wilhelm v. Humboldt, Steinthal, Finck and Byrne, who do not, however, seem to have been inspired by Jenisch: “In language the whole intellectual and moral essence of a man is to some extent revealed. ‘Speak, and you are’ is rightly said by the Oriental. The language of the natural man is savage and rude, that of the cultured man is elegant and polished. As the Greek was subtle in thought and sensuously refined in feeling—as the Roman was serious and practical rather than speculative—as the Frenchman is popular and sociable—as the Briton is profound and the German philosophic—so are also the languages of each of these nations.”

Jenisch then goes on to say that language as the organ for communicating our ideas and feelings accomplishes its end if it represents idea and feeling according to the actual want or need of the mind at the given moment. We have to examine in each case the following essential qualities of the languages compared, (1) richness, (2) energy or emphasis, (3) clearness, and (4) euphony. Under the head of richness we are concerned not only with the number of words, first for material objects, then for spiritual and abstract notions, but also with the ease with which new words can be formed (lexikalische bildsamkeit). The energy of a language is shown in its lexicon and in its grammar (simplicity of grammatical structure, absence of articles, etc.), but also in “the characteristic energy of the nation and its original writers.” Clearness and definiteness in the same way are shown in vocabulary and grammar, especially in a regular and natural syntax. Euphony, finally, depends not only on the selection of consonants and vowels utilized in the language, but on their harmonious combination, the general impression of the language being more important than any details capable of being analysed.

These, then, are the criteria by which Greek and Latin and a number of living languages are compared and judged. The author displays great learning and a sound practical knowledge of many languages, and his remarks on the advantages and shortcomings of these are on the whole judicious, though often perhaps too much stress is laid on the literary merits of great writers, which have really no intrinsic connexion with the value of a language as such. It depends to a great extent on accidental circumstances whether a language has been or has not been used in elevated literature, and its merits should be estimated, so far as this is possible, independently of the perfection of its literature. Jenisch’s prejudice[31] in that respect is shown, for instance, when he says (p. 36) that the endeavours of Hickes are entirely futile, when he tries to make out regular declensions and conjugations in the barbarous language of Wulfila’s translation of the Bible. But otherwise Jenisch is singularly free from prejudices, as shown by a great number of passages in which other languages are praised at the expense of his own. Thus, on p. 396, he declares German to be the most repellent contrast to that most supple modern language, French, on account of its unnatural word-order, its eternally trailing article, its want of participial constructions, and its interminable auxiliaries (as in ‘ich werde geliebt werden, ich würde geliebt worden sein,’ etc.), with the frequent separation of these auxiliaries from the main verb through extraneous intermediate words, all of which gives to German something incredibly awkward, which to the reader appears as lengthy and diffuse and to the writer as inconvenient and intractable. It is not often that we find an author appraising his own language with such severe impartiality, and I have given the passage also to show what kind of problems confront the man who wishes to compare the relative value of languages as wholes. Jenisch’s view here forms a striking contrast to Herder’s appreciation of their common mother-tongue.

Jenisch’s book does not seem to have been widely read by nineteenth-century scholars, who took up totally different problems. Those few who read it were perhaps inclined to say with S. Lefmann (see his book on Franz Bopp, Nachtrag, 1897, p. xi) that it is difficult to decide which was the greater fool, the one who put this problem or the one who tried to answer it. This attitude, however, towards problems of valuation in the matter of languages is neither just nor wise, though it is perhaps easy to see how students of comparative grammar were by the very nature of their study led to look down upon those who compared languages from the point of view of æsthetic or literary merits. Anyhow, it seems to me no small merit to have been the first to treat such problems as these, which are generally answered in an off-hand way according to a loose general judgement, so as to put them on a scientific footing by examining in detail what it is that makes us more or less instinctively prefer one language, or one turn or expression in a language, and thus lay the foundation of that inductive æsthetic theory of language which has still to be developed in a truly scientific spirit.

The nineteenth century witnessed an enormous growth and development of the science of language, which in some respects came to present features totally unknown to previous centuries. The horizon was widened; more and more languages were described, studied and examined, many of them for their own sake, as they had no important literature. Everywhere a deeper insight was gained into the structures even of such languages as had been for centuries objects of study; a more comprehensive and more incisive classification of languages was obtained with a deeper understanding of their mutual relationships, and at the same time linguistic forms were not only described and analysed, but also explained, their genesis being traced as far back as historical evidence allowed, if not sometimes further. Instead of contenting itself with stating when and where a form existed and how it looked and was employed, linguistic science now also began to ask why it had taken that definite shape, and thus passed from a purely descriptive to an explanatory science.

The chief innovation of the beginning of the nineteenth century was the historical point of view. On the whole, it must be said that it was reserved for that century to apply the notion of history to other things than wars and the vicissitudes of dynasties, and thus to discover the idea of development or evolution as pervading the whole universe. This brought about a vast change in the science of language, as in other sciences. Instead of looking at such a language as Latin as one fixed point, and instead of aiming at fixing another language, such as French, in one classical form, the new science viewed both as being in constant flux, as growing, as moving, as continually changing. It cried aloud like Heraclitus[33] “Pánta reî,” and like Galileo “Eppur si muove.” And lo! the better this historical point of view was applied, the more secrets languages seemed to unveil, and the more light seemed also to be thrown on objects outside the proper sphere of language, such as ethnology and the early history of mankind at large and of particular countries.

It is often said that it was the discovery of Sanskrit that was the real turning-point in the history of linguistics, and there is some truth in this assertion, though we shall see on the one hand that Sanskrit was not in itself enough to give to those who studied it the true insight into the essence of language and linguistic science, and on the other hand that real genius enabled at least one man to grasp essential truths about the relationships and development of languages even without a knowledge of Sanskrit. Still, it must be said that the first acquaintance with this language gave a mighty impulse to linguistic studies and exerted a lasting influence on the way in which most European languages were viewed by scholars, and it will therefore be necessary here briefly to sketch the history of these studies. India was very little known in Europe till the mighty struggle between the French and the English for the mastery of its wealth excited a wide interest also in its ancient culture. It was but natural that on this intellectual domain, too, the French and the English should at first be rivals and that we should find both nations represented in the pioneers of Sanskrit scholarship. The French Jesuit missionary Cœurdoux as early as 1767 sent to the French Institut a memoir in which he called attention to the similarity of many Sanskrit words with Latin, and even compared the flexion of the present indicative and subjunctive of Sanskrit asmi, ‘I am,’ with the corresponding forms of Latin grammar. Unfortunately, however, his work was not printed till forty years later, when the same discovery had been announced independently by others. The next scholar to be mentioned in this connexion is Sir William Jones, who in 1796 uttered the following memorable words, which have often been quoted in books on the history of linguistics: “The Sanscrit language, whatever be its antiquity, is of a wonderful structure; more perfect than the Greek, more copious than the Latin and more exquisitely refined than either; yet bearing to both of them a stronger affinity, both in the roots of verbs and in the forms of grammar, than could possibly have been produced by accident; so strong, indeed, that no philologer could examine them all three without believing them to have sprung from some common source, which, perhaps, no longer exists. There is a similar reason, though not quite so forcible, for supposing that both the Gothic and the Celtic ... had the same origin with the Sanscrit; and the old Persian might be added to[34] the same family.” Sir W. Jones, however, did nothing to carry out in detail the comparison thus inaugurated, and it was reserved for younger men to follow up the clue he had given.

One of the books that exercised a great influence on the development of linguistic science in the beginning of the nineteenth century was Friedrich von Schlegel’s Ueber die sprache und weisheit der Indier (1808). Schlegel had studied Sanskrit for some years in Paris, and in his romantic enthusiasm he hoped that the study of the old Indian books would bring about a revolution in European thought similar to that produced in the Renaissance through the revival of the study of Greek. We are here concerned exclusively with his linguistic theories, but to his mind they were inseparable from Indian religion and philosophy, or rather religious and philosophic poetry. He is struck by the similarity between Sanskrit and the best-known European languages, and gives quite a number of words from Sanskrit found with scarcely any change in German, Greek and Latin. He repudiates the idea that these similarities might be accidental or due to borrowings on the side of the Indians, saying expressly that the proof of original relationship between these languages, as well as of the greater age of Sanskrit, lies in the far-reaching correspondences in the whole grammatical structure of these as opposed to many other languages. In this connexion it is noticeable that he is the first to speak of ‘comparative grammar’ (p. 28), but, like Moses, he only looks into this promised land without entering it. Indeed, his method of comparison precludes him from being the founder of the new science, for he says himself (p. 6) that he will refrain from stating any rules for change or substitution of letters (sounds), and require complete identity of the words used as proofs of the descent of languages. He adds that in other cases, “where intermediate stages are historically demonstrable, we may derive giorno from dies, and when Spanish so often has h for Latin f, or Latin p very often becomes f in the German form of the same word, and c not rarely becomes h [by the way, an interesting foreshadowing of one part of the discovery of the Germanic sound-shifting], then this may be the foundation of analogical conclusions with regard to other less evident instances.” If he had followed up this idea by establishing similar ‘sound-laws,’ as we now say, between Sanskrit and other languages, he would have been many years ahead of his time; as it is, his comparisons are those of a dilettante, and he sometimes falls into the pitfalls of accidental similarities while overlooking the real correspondences. He is also led astray by the idea of a[35] particularly close relationship between Persian and German, an idea which at that time was widely spread[2]—we find it in Jenisch and even in Bopp’s first book.

Schlegel is not afraid of surveying the whole world of human languages; he divides them into two classes, one comprising Sanskrit and its congeners, and the second all other languages. In the former he finds organic growth of the roots as shown by their capability of inner change or, as he terms it, ‘flexion,’ while in the latter class everything is effected by the addition of affixes (prefixes and suffixes). In Greek he admits that it would be possible to believe in the possibility of the grammatical endings (bildungssylben) having arisen from particles and auxiliary words amalgamated into the word itself, but in Sanskrit even the last semblance of this possibility disappears, and it becomes necessary to confess that the structure of the language is formed in a thoroughly organic way through flexion, i.e. inner changes and modifications of the radical sound, and not composed merely mechanically by the addition of words and particles. He admits, however, that affixes in some other languages have brought about something that resembles real flexion. On the whole he finds that the movement of grammatical art and perfection (der gang der bloss grammatischen kunst und ausbildung, p. 56) goes in opposite directions in the two species of languages. In the organic languages, which represent the highest state, the beauty and art of their structure is apt to be lost through indolence; and German as well as Romanic and modern Indian languages show this degeneracy when compared with the earlier forms of the same languages. In the affix languages, on the other hand, we see that the beginnings are completely artless, but the ‘art’ in them grows more and more perfect the more the affixes are fused with the main word.

As to the question of the ultimate origin of language, Schlegel thinks that the diversity of linguistic structure points to different beginnings. While some languages, such as Manchu, are so interwoven with onomatopœia that imitation of natural sounds must have played the greatest rôle in their formation, this is by no means the case in other languages, and the perfection of the oldest organic or flexional languages, such as Sanskrit, shows that they cannot be derived from merely animal sounds; indeed, they form an additional proof, if any such were needed, that men did not everywhere start from a brutish state, but that the clearest and intensest reason existed from the very first beginning. On all these points Schlegel’s ideas foreshadow views that are found in later works; and it is probable that his fame as a writer outside the philological field gave to his linguistic speculations a notoriety which his often[36] loose and superficial reasonings would not otherwise have acquired for them.

Schlegel’s bipartition of the languages of the world carries in it the germ of a tripartition. On the lowest stage of his second class he places Chinese, in which, as he acknowledges, the particles denoting secondary sense modifications consist in monosyllables that are completely independent of the actual word. It is clear that from Schlegel’s own point of view we cannot here properly speak of ‘affixes,’ and thus Chinese really, though Schlegel himself does not say so, falls outside his affix languages and forms a class by itself. On the other hand, his arguments for reckoning Semitic languages among affix languages are very weak, and he seems also somewhat inclined to say that much in their structure resembles real flexion. If we introduce these two changes into his system, we arrive at the threefold division found in slightly different shapes in most subsequent works on general linguistics, the first to give it being perhaps Schlegel’s brother, A. W. Schlegel, who speaks of (1) les langues sans aucune structure grammaticale—under which misleading term he understands Chinese with its unchangeable monosyllabic words; (2) les langues qui emploient des affixes; (3) les langues à inflexions.

Like his brother, A. W. Schlegel places the flexional languages highest and thinks them alone ‘organic.’ On the other hand, he subdivides flexional languages into two classes, synthetic and analytic, the latter using personal pronouns and auxiliaries in the conjugation of verbs, prepositions to supply the want of cases, and adverbs to express the degrees of comparison. While the origin of the synthetic languages loses itself in the darkness of ages, the analytic languages have been created in modern times; all those that we know are due to the decomposition of synthetic languages. These remarks on the division of languages are found in the Introduction to the book Observations sur la langue et la littérature provençale (1818) and are thus primarily meant to account for the contrast between synthetic Latin and analytic Romanic.

We now come to the three greatest names among the initiators of linguistic science in the beginning of the nineteenth century. If we give them in their alphabetical order, Bopp, Grimm and Rask, we also give them in the order of merit in which most subsequent historians have placed them. The works that constitute their first claims to the title of founder of the new science came in close succession, Bopp’s Conjugationssystem in 1816, Rask’s Undersøgelse in 1818, and the first volume of Grimm’s Grammatik in[37] 1819. While Bopp is entirely independent of the two others, we shall see that Grimm was deeply influenced by Rask, and as the latter’s contributions to our science began some years before his chief work just mentioned (which had also been finished in manuscript in 1814, thus two years before Bopp’s Conjugationssystem), the best order in which to deal with the three men will perhaps be to take Rask first, then to mention Grimm, who in some ways was his pupil, and finally to treat of Bopp: in this way we shall also be enabled to see Bopp in close relation with the subsequent development of Comparative Grammar, on which he, and not Rask, exerted the strongest influence.

Born in a peasant’s hut in the heart of Denmark in 1787, Rasmus Rask was a grammarian from his boyhood. When a copy of the Heimskringla was given him as a school prize, he at once, without any grammar or dictionary, set about establishing paradigms, and so, before he left school, acquired proficiency in Icelandic, as well as in many other languages. At the University of Copenhagen he continued in the same course, constantly widened his linguistic horizon and penetrated into the grammatical structure of the most diverse languages. Icelandic (Old Norse), however, remained his favourite study, and it filled him with enthusiasm and national pride that “our ancestors had such an excellent language,” the excellency being measured chiefly by the full flexional system which Icelandic shared with the classical tongues, partly also by the pure, unmixed state of the Icelandic vocabulary. His first book (1811) was an Icelandic grammar, an admirable production when we consider the meagre work done previously in this field. With great lucidity he reduces the intricate forms of the language into a consistent system, and his penetrating insight into the essence of language is seen when he explains the vowel changes, which we now comprise under the name of mutation or umlaut, as due to the approximation of the vowel of the stem to that of the ending, at that time a totally new point of view. This we gather from Grimm’s review, in which Rask’s explanation is said to be “more astute than true” (“mehr scharfsinnig als wahr,” Kleinere schriften, 7. 515). Rask even sees the reason of the change in the plural blöð as against the singular blað in the former having once ended in -u, which has since disappeared. This is, so far as I know, the first inference ever drawn to a prehistoric state of language.

In 1814, during a prolonged stay in Iceland, Rask sent down to Copenhagen his most important work, the prize essay on the origin of the Old Norse language (Undersøgelse om det gamle nordiske eller islandske sprogs oprindelse) which for various reasons was not printed till 1818. If it had been published when it was finished, and especially if it had been printed in a language[38] better known than Danish, Rask might well have been styled the founder of the modern science of language, for his work contains the best exposition of the true method of linguistic research written in the first half of the nineteenth century and applies this method to the solution of a long series of important questions. Only one part of it was ever translated into another language, and this was unfortunately buried in an appendix to Vater’s Vergleichungstafeln, 1822. Yet Rask’s work even now repays careful perusal, and I shall therefore give a brief résumé of its principal contents.

Language according to Rask is our principal means of finding out anything about the history of nations before the existence of written documents, for though everything may change in religion, customs, laws and institutions, language generally remains, if not unchanged, yet recognizable even after thousands of years. But in order to find out anything about the relationship of a language we must proceed methodically and examine its whole structure instead of comparing mere details; what is here of prime importance is the grammatical system, because words are very often taken over from one language to another, but very rarely grammatical forms. The capital error in most of what has been written on this subject is that this important point has been overlooked. That language which has the most complicated grammar is nearest to the source; however mixed a language may be, it belongs to the same family as another if it has the most essential, most material and indispensable words in common with it; pronouns and numerals are in this respect most decisive. If in such words there are so many points of agreement between two languages that it is possible to frame rules for the transitions of letters (in other passages Rask more correctly says sounds) from the one language to the other, there is a fundamental kinship between the two languages, more particularly if there are corresponding similarities in their structure and constitution. This is a most important thesis, and Rask supplements it by saying that transitions of sounds are naturally dependent on their organ and manner of production.

Next Rask proceeds to apply these principles to his task of finding out the origin of the Old Icelandic language. He describes its position in the ‘Gothic’ (Gothonic, Germanic) group and then looks round to find congeners elsewhere. He rapidly discards Greenlandic and Basque as being too remote in grammar and vocabulary; with regard to Keltic languages he hesitates, but finally decides in favour of denying relationship. (He was soon to see his error in this; see below.) Next he deals at some length with Finnic and Lapp, and comes to the conclusion that the simi[39]larities are due to loans rather than to original kinship. But when he comes to the Slavonic languages his utterances have a different ring, for he is here able to disclose so many similarities in fundamentals that he ranges these languages within the same great family as Icelandic. The same is true with regard to Lithuanian and Lettic, which are here for the first time correctly placed as an independent sub-family, though closely akin to Slavonic. The comparisons with Latin, and especially with Greek, are even more detailed; and Rask in these chapters really presents us with a succinct, but on the whole marvellously correct, comparative grammar of Gothonic, Slavonic, Lithuanian, Latin and Greek, besides examining numerous lexical correspondences. He does not yet know any of the related Asiatic languages, but throws out the hint that Persian and Indian may be the remote source of Icelandic through Greek. Greek he considers to be the ‘source’ or ‘root’ of the Gothonic languages, though he expresses himself with a degree of uncertainty which forestalls the correct notion that these languages have all of them sprung from the same extinct and unknown language. This view is very clearly expressed in a letter he wrote from St. Petersburg in the same year in which his Undersøgelse was published; he here says: “I divide our family of languages in this way: the Indian (Dekanic, Hindostanic), Iranic (Persian, Armenian, Ossetic), Thracian (Greek and Latin), Sarmatian (Lettic and Slavonic), Gothic (Germanic and Skandinavian) and Keltic (Britannic and Gaelic) tribes” (SA 2. 281, dated June 11, 1818).

This is the fullest and clearest account of the relationships of our family of languages found for many years, and Rask showed true genius in the way in which he saw what languages belonged together and how they were related. About the same time he gave a classification of the Finno-Ugrian family of languages which is pronounced by such living authorities on these languages as Vilhelm Thomsen and Emil Setälä to be superior to most later attempts. When travelling in India he recognized the true position of Zend, about which previous scholars had held the most erroneous views, and his survey of the languages of India and Persia was thought valuable enough in 1863 to be printed from his manuscript, forty years after it was written. He was also the first to see that the Dravidian (by him called Malabaric) languages were totally different from Sanskrit. In his short essay on Zend (1826) he also incidentally gave the correct value of two letters in the first cuneiform writing, and thus made an important contribution towards the final deciphering of these inscriptions.

His long tour (1816-23) through Sweden, Finland, Russia, the Caucasus, Persia and India was spent in the most intense study[40] of a great variety of languages, but unfortunately brought on the illness and disappointments which, together with economic anxieties, marred the rest of his short life.

When Rask died in 1832 he had written a great number of grammars of single languages, all of them remarkable for their accuracy in details and clear systematic treatment, more particularly of morphology, and some of them breaking new ground; besides his Icelandic grammar already mentioned, his Anglo-Saxon, Frisian and Lapp grammars should be specially named. Historical grammar in the strict sense is perhaps not his forte, though in a remarkable essay of the year 1815 he explains historically a great many features of Danish grammar, and in his Spanish and Italian grammars he in some respects forestalls Diez’s historical explanations. But in some points he stuck to erroneous views, a notable instance being his system of old Gothonic ‘long vowels,’ which was reared on the assumption that modern Icelandic pronunciation reflects the pronunciation of primitive times, while it is really a recent development, as Grimm saw from a comparison of all the old languages. With regard to consonants, however, Rask was the clearer-sighted of the two, and throughout he had this immense advantage over most of the comparative linguists of his age, that he had studied a great many languages at first hand with native speakers, while the others knew languages chiefly or exclusively through the medium of books and manuscripts. In no work of that period, or even of a much later time, are found so many first-hand observations of living speech as in Rask’s Retskrivningslære. Handicapped though he was in many ways, by poverty and illness and by the fact that he wrote in a language so little known as Danish, Rasmus Rask, through his wide outlook, his critical sagacity and aversion to all fanciful theorizing, stands out as one of the foremost leaders of linguistic science.[3]

Jacob Grimm’s career was totally different from Rask’s. Born in 1785 as the son of a lawyer, he himself studied law and came under the influence of Savigny, whose view of legal institutions as the outcome of gradual development in intimate connexion with popular tradition and the whole intellectual and moral life of the[41] people appealed strongly to the young man’s imagination. But he was drawn even more to that study of old German popular poetry which then began to be the fashion, thanks to Tieck and other Romanticists; and when he was in Paris to assist Savigny with his historico-legal research, the old German manuscripts in the Bibliothèque nationale nourished his enthusiasm for the poetical treasures of the Middle Ages. He became a librarian and brought out his first book, Ueber den altdeutschen meistergesang (1811). At the same time, with his brother Wilhelm as constant companion and fellow-worker, he began collecting popular traditions, of which he published a first instalment in his famous Kinder- und hausmärchen (1812 ff.), a work whose learned notes and comparisons may be said to have laid the foundation of the science of folklore. Language at first had only a subordinate interest to him, and when he tried his hand at etymology, he indulged in the wildest guesses, according to the method (or want of method) of previous centuries. A. W. Schlegel’s criticism of his early attempts in this field, and still more Rask’s example, opened Grimm’s eyes to the necessity of a stricter method, and he soon threw himself with great energy into a painstaking and exact study of the oldest stages of the German language and its congeners. In his review (1812) of Rask’s Icelandic grammar he writes: “Each individuality, even in the world of languages, should be respected as sacred; it is desirable that even the smallest and most despised dialect should be left only to itself and to its own nature and in nowise subjected to violence, because it is sure to have some secret advantages over the greatest and most highly valued language.” Here we meet with that valuation of the hitherto overlooked popular dialects which sprang from the Romanticists’ interest in the ‘people’ and everything it had produced. Much valuable linguistic work was directly inspired by this feeling and by conscious opposition to the old philology, that occupied itself exclusively with the two classical languages and the upper-class literature embodied in them. As Scherer expresses it (Jacob Grimm, 2te ausg., Berlin, 1885, p. 152): “The brothers Grimm applied to the old national literature and to popular traditions the old philological virtue of exactitude, which had up to then been bestowed solely on Greek and Roman classics and on the Bible. They extended the field of strict philology, as they extended the field of recognized poetry. They discarded the aristocratic narrowmindedness with which philologists looked down on unwritten tradition, on popular ballads, legends, fairy tales, superstition, nursery rimes.... In the hands of the two Grimms philology became national and popular; and at the same time a pattern was created for the scientific study of all the peoples of the earth and[42] for a comparative investigation of the entire mental life of mankind, of which written literature is nothing but a small epitome.”

But though Grimm thus broke loose from the traditions of classical philology, he still carried with him one relic of it, namely the standard by which the merits of different languages were measured. “In reading carefully the old Gothonic (altdeutschen) sources, I was every day discovering forms and perfections which we generally envy the Greeks and Romans when we consider the present condition of our language.”... “Six hundred years ago every rustic knew, that is to say practised daily, perfections and niceties in the German language of which the best grammarians nowadays do not even dream; in the poetry of Wolfram von Eschenbach and of Hartmann von Aue, who had never heard of declension and conjugation, nay who perhaps did not even know how to read and write, many differences in the flexion and use of nouns and verbs are still nicely and unerringly observed, which we have gradually to rediscover in learned guise, but dare not reintroduce, for language ever follows its inalterable course.”

Grimm then sets about writing his great historical and comparative Deutsche Grammatik, taking the term ‘deutsch’ in its widest and hardly justifiable sense of what is now ordinarily called Germanic and which is in this work called Gothonic. The first volume appeared in 1819, and in the preface we see that he was quite clear that he was breaking new ground and introducing a new method of looking at grammar. He speaks of previous German grammars and says expressly that he does not want his to be ranged with them. He charges them with unspeakable pedantry; they wanted to dogmatize magisterially, while to Grimm language, like everything natural and moral, is an unconscious and unnoticed secret which is implanted in us in youth. Every German therefore who speaks his language naturally, i.e. untaught, may call himself his own living grammar and leave all schoolmasters’ rules alone. Grimm accordingly has no wish to prescribe anything, but to observe what has grown naturally, and very appropriately he dedicates his work to Savigny, who has taught him how institutions grow in the life of a nation. In the new preface to the second edition there are also some noteworthy indications of the changed attitude. “I am hostile to general logical notions in grammar; they conduce apparently to strictness and solidity of definition, but hamper observation, which I take to be the soul of linguistic science.... As my starting-point was to trace the never-resting (unstillstehende) element of our language which changes with time and place, it became necessary for me to admit one dialect after the other, and I could not even[43] forbear to glance at those foreign languages that are ultimately related with ours.”

Here we have the first clear programme of that historical school which has since then been the dominating one in linguistics. But as language according to this new point of view was constantly changing and developing, so also, during these years, were Grimm’s own ideas. And the man who then exercised the greatest influence on him was Rasmus Rask. When Grimm wrote the first edition of his Grammatik (1819), he knew nothing of Rask but the Icelandic grammar, but just before finishing his own volume Rask’s prize essay reached him, and in the preface he at once speaks of it in the highest terms of praise, as he does also in several letters of this period; he is equally enthusiastic about Rask’s Anglo-Saxon grammar and the Swedish edition of his Icelandic grammar, neither of which reached him till after his own first volume had been printed off. The consequence was that instead of going on to the second volume, Grimm entirely recast the first volume and brought it out in a new shape in 1822. The chief innovation was the phonology or, as he calls it, “Erstes buch. Von den buchstaben,” which was entirely absent in 1819, but now ran to 595 pages.

This first book in the 1822 volume contains much, perhaps most, of what constitutes Grimm’s fame as a grammarian, notably his exposition of the ‘sound shift’ (lautverschiebung), which it has been customary in England since Max Müller to term ‘Grimm’s Law.’ If any one man is to give his name to this law, a better name would be ‘Rask’s Law,’ for all these transitions, Lat. Gr. p = f, t = þ (th), k = h, etc., are enumerated in Rask’s Undersøgelse, p. 168, which Grimm knew before he wrote a single word about the sound shift.

Now, it is interesting to compare the two scholars’ treatment of these transitions. The sober-minded, matter-of-fact Rask contents himself with a bare statement of the facts, with just enough well-chosen examples to establish the correspondence; the way in which he arranges the sounds shows that he saw their parallelism clearly enough, though he did not attempt to bring everything under one single formula, any more than he tried to explain why these sounds had changed.[4] Grimm multiplies the examples and[44] then systematizes the whole process in one formula so as to comprise also the ‘second shift’ found in High German alone—a shift well known to Rask, though treated by him in a different place (p. 68 f.). Grimm’s formula looks thus:

which may be expressed generally thus, that tenuis (T) becomes aspirate (A) and then media (M), etc., or, tabulated:

For this Grimm would of course have deserved great credit, because a comprehensive formula is more scientific than a rough statement of facts—if the formula had been correct; but unfortunately it is not so. In the first place, it breaks down in the very first instance, for there is no media in High German corresponding to Gr. p and Gothic f (cf. poûs, fotus, fuss, etc.); secondly, High German has h just as Gothic has, corresponding to Greek k (cf. kardía, hairto, herz, etc.), and where it has g, Gothic has also g in accordance with rules unknown to Grimm and not explained till long afterwards (by Verner). But the worst thing is that the whole specious generalization produces the impression of regularity and uniformity only through the highly unscientific use of the word ‘aspirate,’ which is made to cover such phonetically disparate things as (1) combination of stop with following h, (2) combination of stop with following fricative, pf, ts written z, (3) voiceless fricative, f, s in G. das, (4) voiced fricative, v, ð written th, and (5) h. Grimm rejoiced in his formula, giving as it does three chronological stages in each of the three subdivisions (tenuis, media, aspirate) of each of the three classes of consonants (labial, dental, ‘guttural’). This evidently took hold of his fancy through the mystic power of the number three, which he elsewhere (Gesch 1. 191, cf. 241) finds pervading language generally: three original vowels, a, i, u, three genders, three numbers (singular, dual, plural), three persons, three ‘voices’ (genera: active, middle, passive), three tenses (present, preterit, future), three declensions through a, i, u. As there is here an element of mysticism, so is there also in Grimm’s highflown[45] explanation of the whole process from pretended popular psychology, which is full of the cloudiest romanticism. “When once the language had made the first step and had rid itself of the organic basis of its sounds, it was hardly possible for it to escape the second step and not to arrive at the third stage,[5] through which this development was perfected.... It is impossible not to admire the instinct by which the linguistic spirit (sprachgeist) carried this out to the end. A great many sounds got out of joint, but they always knew how to arrange themselves in a different place and to find the new application of the old law. I am not saying that the shift happened without any detriment, nay from one point of view the sound shift appears to me as a barbarous aberration, from which other more quiet nations abstained, but which is connected with the violent progress and craving for freedom which was found in Germany in the beginning of the Middle Ages and which initiated the transformation of Europe. The Germans pressed forward even in the matter of the innermost sounds of their language,” etc., with remarks on intellectual progress and on victorious and ruling races. Grimm further says that “die dritte stufe des verschobnen lauts den kreislauf abschliesse und nach ihr ein neuer ansatz zur abweichung wieder von vorn anheben müsse. Doch eben weil der sprachgeist seinen lauf vollbracht hat, scheint er nicht wieder neu beginnen zu wollen” (GDS 1. 292 f., 299). It would be difficult to attach any clear ideas to these words.

Grimm’s idea of a ‘kreislauf’ is caused by the notion that the two shifts, separated by several centuries, represent one continued movement, while the High German shift of the eighth century has really no more to do with the primitive Gothonic shift, which took place probably some time before Christ, than has, for instance, the Danish shift in words like gribe, bide, bage, from gripæ, bitæ, bakæ (about 1400), or the still more recent transition in Danish through which stressed t in tid, tyve, etc., sounds nearly like [ts], as in HG. zeit. There cannot possibly be any causal nexus between such transitions, separated chronologically by long periods, with just as little change in the pronunciation of these consonants as there has been in English.[6]

[46]

Grimm was anything but a phonetician, and sometimes says things which nowadays cannot but produce a smile, as when he says (Gr 1. 3) “in our word schrift, for instance, we express eight sounds through seven signs, for f stands for ph”; thus he earnestly believes that sch contains three sounds, s and the ‘aspirate’ ch = c + h! Yet through the irony of fate it was on the history of sounds that Grimm exercised the strongest influence. As in other parts of his grammar, so also in the “theory of letters” he gave fuller word lists than people had been accustomed to, and this opened the eyes of scholars to the great regularity reigning in this department of linguistic development. Though in his own etymological practice he was far from the strict idea of ‘phonetic law’ that played such a prominent rôle in later times, he thus paved the way for it. He speaks of law at any rate in connexion with the consonant shift, and there recognizes that it serves to curb wild etymologies and becomes a test for them (Gesch 291). The consonant shift thus became the law in linguistics, and because it affected a great many words known to everybody, and in a new and surprising way associated well-known Latin or Greek words with words of one’s own mother-tongue, it became popularly the keystone of a new wonderful science.

Grimm coined several of the terms now generally used in linguistics; thus umlaut and ablaut, ‘strong’ and ‘weak’ declensions and conjugations. As to the first, we have seen that it was Rask who first understood and who taught Grimm the cause of this phenomenon, which in English has often been designated by the German term, while Sweet calls it ‘mutation’ and others better ‘infection.’ With regard to ‘ablaut’ (Sweet: gradation, best perhaps in English apophony), Rask termed it ‘omlyd,’ a word which he never applied to Grimm’s ‘umlaut,’ thus keeping the two kinds of vowel change as strictly apart as Grimm does. Apophony was first discovered in that class of verbs which Grimm called ‘strong’; he was fascinated by the commutation of the vowels in springe, sprang, gesprungen, and sees in it, as in bimbambum, something mystic and admirable, characteristic of the old German spirit. He was thus blind to the correspondences found in other languages, and his theory led him astray in the second volume, in which he constructed imaginary verbal roots to explain apophony wherever it was found outside the verbs.

[47]

Though Grimm, as we have seen, was by his principles and whole tendency averse to prescribing laws for a language, he is sometimes carried away by his love for mediæval German, as when he gives as the correct nominative form der boge, though everybody for centuries had said der bogen. In the same way many of his followers would apply the historical method to questions of correctness of speech, and would discard the forms evolved in later times in favour of previously existing forms which were looked upon as more ‘organic.’

It will not be necessary here to speak of the imposing work done by Grimm in the rest of his long life, chiefly spent as a professor in Berlin. But in contrast to the ordinary view I must say that what appears to me as most likely to endure is his work on syntax, contained in the fourth volume of his grammar and in monographs. Here his enormous learning, his close power of observation, and his historical method stand him in good stead, and there is much good sense and freedom from that kind of metaphysical systematism which was triumphant in contemporaneous work on classical syntax. His services in this field are the more interesting because he did not himself seem to set much store by these studies and even said that syntax was half outside the scope of grammar. This utterance belongs to a later period than that of the birth of historical and comparative linguistics, and we shall have to revert to it after sketching the work of the third great founder of this science, to whom we shall now turn.

The third, by some accounted the greatest, among the founders of modern linguistic science was Franz Bopp. His life was uneventful. At the age of twenty-one (he was born in 1791) he went to Paris to study Oriental languages, and soon concentrated his attention on Sanskrit. His first book, from which it is customary in Germany to date the birth of Comparative Philology, appeared in 1816, while he was still in Paris, under the title Ueber des conjugationssystem der sanskritsprache in vergleichung mit jenem der griechischen, lateinischen, persischen und germanischen sprache, but the latter part of the small volume was taken up with translations from Sanskrit, and for a long time he was just as much a Sanskrit scholar, editing and translating Sanskrit texts, as a comparative grammarian. He showed himself in the latter character in several papers read before the Berlin Academy, after he had been made a professor there in 1822, and especially in his famous Vergleichende grammatik des sanskrit, ṣend, armenischen, griechischen, lateinischen, litauischen, altslawischen, gotischen und deutschen, the first edition of which was[48] published between 1833 and 1849, the second in 1857, and the third in 1868. Bopp died in 1867.

Of Bopp’s Conjugationssystem a revised, rearranged and greatly improved English translation came out in 1820 under the title Analytical Comparison of the Sanskrit, Greek, Latin and Teutonic Languages. This was reprinted with a good introduction by F. Techmer in his Internationale zeitschrift für allgem. sprachwissenschaft IV (1888), and in the following remarks I shall quote this (abbreviated AC) instead of, or alongside of, the German original (abbreviated C).

Bopp’s chief aim (and in this he was characteristically different from Rask) was to find out the ultimate origin of grammatical forms. He follows his quest by the aid of Sanskrit forms, though he does not consider these as the ultimate forms themselves: “I do not believe that the Greek, Latin, and other European languages are to be considered as derived from the Sanskrit in the state in which we find it in Indian books; I feel rather inclined to consider them altogether as subsequent variations of one original tongue, which, however, the Sanskrit has preserved more perfect than its kindred dialects. But whilst therefore the language of the Brahmans more frequently enables us to conjecture the primitive form of the Greek and Latin languages than what we discover in the oldest authors and monuments, the latter on their side also may not unfrequently elucidate the Sanskrit grammar” (AC 3). Herein subsequent research has certainly borne out Bopp’s view.

After finding out by a comparison of the grammatical forms of Sanskrit, Greek, etc., which of these forms were identical and what were their oldest shapes, he tries to investigate the ultimate origin of these forms. This he takes to be a comparatively easy consequence of the first task, but he was here too much under the influence of the philosophical grammar then in vogue. Gottfried Hermann (De emendanda ratione Græcæ grammaticæ, 1801), on purely logical grounds, distinguishes three things as necessary elements of each sentence, the subject, the predicate, and the copula joining the first two elements together; as the power of the verb is to attribute the predicate to the subject, there is really only one verb, namely the verb to be. Bopp’s teacher in Paris, Silvestre de Sacy, says the same thing, and Bopp repeats: “A verb, in the most restricted meaning of the term, is that part of speech by which a subject is connected with its attribute. According to this definition it would appear that there can exist only one verb, namely, the substantive verb, in Latin esse; in English, to be.... Languages of a structure similar to that of the Greek, Latin, etc., can express by one verb of this kind a whole logical proposition, in which, however, that part of speech which expresses the connexion[49] of the subject with its attribute, which is the characteristic function of the verb, is generally entirely omitted or understood. The Latin verb dat expresses the proposition ‘he gives,’ or ‘he is giving’: the letter t, indicating the third person, is the subject, da expresses the attribute of giving, and the grammatical copula is understood. In the verb potest, the latter is expressed, and potest unites in itself the three essential parts of speech, t being the subject, es the copula, and pot the attribute.”

Starting from this logical conception of grammar, Bopp is inclined to find everywhere the ‘substantive verb’ to be in its two Sanskrit forms as and bhu as an integral part of verbal forms. He is not the first to think that terminations, which are now inseparable parts of a verb, were originally independent words; thus Horne Tooke (in Epea pteroenta, 1786, ii. 429) expressly says that “All those common terminations in any language ... are themselves separate words with distinct meanings,” and explains, for instance, Latin ibo from i, ‘go’ + b, ‘will,’ from Greek boúl(omai) + o ‘I,’ from ego. Bopp’s explanations are similar to this, though they do not imply such violent shortenings as that of boúl(omai) to b. He finds the root Sanskrit as, ‘to be,’ in Latin perfects like scrip-s-i, in Greek aorists like e-tup-s-a and in futures like tup-s-o. That the same addition thus indicates different tenses does not trouble Bopp greatly; he explains Lat. fueram from fu + es + am, etc., and says that the root fu “contains, properly, nothing to indicate past time, but the usage of language having supplied the want of an adequate inflexion, fui received the sense of a perfect, and fu-eram, which would be nothing more than an imperfect, that of a pluperfect, and after the same manner fu-ero signifies ‘I shall have been,’ instead of ‘I shall be’” (AC 57). All Latin verbal endings containing r are thus explained as being ultimately formed with the substantive verb (ama-rem, etc.); thus among others the infinitives fac-ere, ed-ere, as well as esse, posse: “E is properly, in Latin, the termination of a simple infinitive active; and the root Es produced anciently ese, by adding e; the s having afterwards been doubled, we have esse. This termination e answers to the Greek infinitive in ai, eînai ...” (AC 58).

If Bopp found a master-key to many of the verbal endings in the Sanskrit root es, he found a key to many others in the other root of the verb ‘to be,’ Sanskrit bhu. He finds it in the Latin imperfect da-bam, as well as in the future da-bo, the relation between which is the same as that between er-am and er-o. “Bo, bis, bit has a striking similarity with the Anglo-Saxon beo, bys, byth, the future tense of the verb substantive, a similarity which cannot be considered as merely accidental.” [Here neither the form nor the function of the Anglo-Saxon is stated quite correctly.] But[50] the ending in Latin ama-vi is also referred to the same root; for the change of the b into v we are referred to Italian amava, from Lat. amabam; thus also fui is for fuvi and potui is for pot-vi: “languages manifest a constant effort to combine heterogeneous materials in such a manner as to offer to the ear or eye one perfect whole, like a statue executed by a skilful artist, that wears the appearance of a figure hewn out of one piece of marble” (AC 60).

The following may be taken as a fair specimen of the method followed in these first attempts to account for the origin of flexional forms: “The Latin passive forms amat-ur, amant-ur, would, in some measure, conform to this mode of joining the verb substantive, if the r was also the result of a permutation of an original s; and this appears not quite incredible, if we compare the second person ama-ris with the third amat-ur. Either in one or the other there must be a transposition of letters, to which the Latin language is particularly addicted. If ama-ris, which might have been produced from ama-sis, has preserved the original order of letters, then ama-tur must be the transposition of ama-rut or ama-sut, and ama-ntur that of ama-runt or ama-sunt. If this be the case, the origin of the Latin passive can be accounted for, and although differing from that of the Sanskrit, Greek, and Gothic languages, it is not produced by the invention of a new grammatical form. It becomes clear, also, why many verbs, with a passive form, have an active signification; because there is no reason why the addition of the verb substantive should necessarily produce a passive sense. There is another way of explaining ama-ris, if it really stands for ama-sis; the s may be the radical consonant of the reflex pronoun se. The introduction of this pronoun would be particularly adapted to form the middle voice, which expresses the reflexion of the action upon the actor; but the Greek language exemplifies the facility with which the peculiar signification of the middle voice passes into that of the passive.” The reasoning in the beginning of this passage (the only one contained in C) carries us back to a pre-scientific atmosphere, of which there are few or no traces in Rask’s writings; the latter explanation (added in AC) was preferred by Bopp himself in later works, and was for many years accepted as the correct one, until scholars found a passive in r in Keltic, where the transition from s to r is not found as it is in Latin; and as the closely corresponding forms in Keltic and Italic must obviously be explained in the same way, the hypothesis of a composition with se was generally abandoned. Bopp’s partiality for the abstract verb is seen clearly when he explains the Icelandic passive in -st from s = es (C 132); here Rask and Grimm saw the correct and obvious explanation.

[51]

Among the other explanations given first by Bopp must be mentioned the Latin second person of the passive voice -mini, as in ama-mini, which he takes to be the nominative masculine plural of a participle corresponding to Greek -menos and found in a different form in Lat. alumnus (AC 51). This explanation is still widely accepted, though not by everybody.

With regard to the preterit of what Grimm was later to term the ‘weak’ verbs, Bopp vacillates between different explanations. In C 118 he thinks the t or d is identical with the ending of the participle, in which the case endings were omitted and supplanted by personal endings; the syllable ed after d [in Gothic sok-id-edum; ‘Greek,’ p. 119, must be a misprint for Gothic] is nothing but an accidental addition. But on p. 151 he sees in sokidedun, sokidedi, a connexion of sok with the preterit of the verb Tun, as if the Germans were to say suchetaten, suchetäte; he compares the English use of did (did seek), and thinks the verb used is G. tun, Goth. tanjan. The theory of composition is here restricted to those forms that contain two d’s, i.e. the plural indicative and the subjunctive. In the English edition this twofold explanation is repeated with some additions: d or t as in Gothic sok-i-da and oh-ta originates from a participle found in Sanskr. tyak-ta, likh-i-ta, Lat. -tus, Gr. -tós; this suffix generally has a passive sense, but in neuter verbs an active sense, and therefore would naturally serve to form a preterit tense with an active signification. He finds a proof of the connexion between this preterit and the participle in the fact that only such verbs as have this ending in the participle form their preterit by means of a dental, while the others (the ‘strong’ verbs, as Grimm afterwards termed them) have a participle in an and reduplication or a change of vowel in the preterit; and Bopp compares the Greek aorist passive etúphth-ēn, edóth-ēn, which he conceives may proceed from the participle tuphth-eís, doth-eís (AC 37 ff.). This suggestion seems to have been commonly overlooked or abandoned, while the other explanation, from dedi as in English did seek, which Bopp gives p. 49 for the subjunctive and the indicative plural, was accepted by Grimm as the explanation of all the forms, even of those containing only one dental; in later works Bopp agreed with Grimm and thus gave up the first part of his original explanation. The did explanation had been given already by D. von Stade (d. 1718, see Collitz, Das schwache präteritum, p. 1); Rask (P 270, not mentioned by Collitz) says: “Whence this d or t has come is not easy to tell, as it is not found in Latin and Greek, but as it is evident from the Icelandic grammar that it is closely connected with the past participle and is also found in the preterit subjunctive, it seems clear that it must have been an old characteristic of the past tense in every mood, but was lost[52] in Greek when the above-mentioned participles in tos disappeared from the verbs” (cf. Ch. XIX § 12).

With regard to the vowels, Bopp in AC has the interesting theory that it is only through a defect in the alphabet that Sanskrit appears to have a in so many places; he believes that the spoken language had often “the short Italian e and o,” where a was written. “If this was the case, we can give a reason why, in words common to the Sanskrit and Greek, the Indian akāra [that is, short a] so often corresponds to ε and ο, as, for instance, asti, he is, ἐστί; patis, husband, πόσις; ambaras, sky, ὄμβρος, rain, etc.” Later, unfortunately, Bopp came under the influence of Grimm, who, as we saw, on speculative grounds admitted in the primitive language only the three vowels a, i, u, and Bopp and his followers went on believing that the Sanskrit a represented the original state of language, until the discovery of the ‘palatal law’ (about 1880) showed (what Bopp’s occasional remark might otherwise easily have led up to, if he had not himself discarded it) that the Greek tripartition into a, e, o represented really a more original state of things.

In a chapter on the roots in AC (not found in C), Bopp contrasts the structure of Semitic roots and of our own; in Semitic languages roots must consist of three letters, neither more nor less, and thus generally contain two syllables, while in Sanskrit, Greek, etc., the character of the root “is not to be determined by the number of letters, but by that of the syllables, of which they contain only one”; thus a root like i, ‘to go,’ would be unthinkable in Arabic. The consequence of this structure of the roots is that the inner changes which play such a large part in expressing grammatical modifications in Semitic languages must be much more restricted in our family of languages. These changes were what F. Schlegel termed flexions and what Bopp himself, two years before (C 7), had named “the truly organic way” of expressing relation and mentioned as a wonderful flexibility found in an extraordinary degree in Sanskrit, by the side of which composition with the verb ‘to be’ is found only occasionally. Now, however, in 1820, Bopp repudiates Schlegel’s and his own previous assumption that ‘flexion’ was characteristic of Sanskrit in contradistinction to other languages in which grammatical modifications were expressed by the addition of suffixes. On the contrary, while holding that both methods are employed in all languages, Chinese perhaps alone excepted, he now thinks that it is the suffix method which is prevalent in Sanskrit, and that “the only real inflexions ... possible[53] in a language, whose elements are monosyllables, are the change of their vowels and the repetition of their radical consonants, otherwise called reduplication.” It will be seen that Bopp here avoids both the onesidedness found in Schlegel’s division of languages and the other onesidedness which we shall encounter in later theories, according to which all grammatical elements are originally independent subordinate roots added to the main root.

In his Vocalismus (1827, reprinted 1836) Bopp opposes Grimm’s theory that the changes for which Grimm had introduced the term ablaut were due to psychological causes; in other words, possessed an inner meaning from the very outset. Bopp inclined to a mechanical explanation[7] and thought them dependent on the weight of the endings, as shown by the contrast between Sanskr. vēda, Goth. vait, Gr. oîda and the plural, respectively vidima, vitum, ídmen. In this instance Bopp is in closer agreement than Grimm with the majority of younger scholars, who see in apophony (ablaut) an originally non-significant change brought about mechanically by phonetic conditions, though they do not find these in the ‘weight’ of the ending, but in the primeval accent: the accentuation of Sanskrit was not known to Bopp when he wrote his essay.

The personal endings of the verbs had already been identified with the corresponding pronouns by Scheidius (1790) and Rask (P 258); Bopp adopts the same view, only reproaching Scheidius for thinking exclusively of the nominative forms of the pronouns.

It thus appears that in his early work Bopp deals with a great many general problems, but his treatment is suggestive rather than exhaustive or decisive, for there are too many errors in details and his whole method is open to serious criticism. A modern reader is astonished to see the facility with which violent changes of sounds, omissions and transpositions of consonants, etc., are gratuitously accepted. Bopp never reflected as deeply as Rask did on what constitutes linguistic kinship, hence in C he accepts the common belief that Persian was related more closely to German than to Sanskrit, and in later life he tried to establish a relationship between the Malayo-Polynesian and the Indo-European languages. But in spite of all this it must be recognized that in his long laborious life he accomplished an enormous amount of highly meritorious work, not only in Sanskrit philology, but also in comparative grammar, in which he gradually freed himself of his worst methodical errors. He was constantly widening his range of vision, taking into consideration more and more cognate languages. The ingenious way in which he explained the curious Keltic shiftings in initial[54] consonants (which had so puzzled Rask as to make him doubt of a connexion of these languages with our family, but which Bopp showed to be dependent on a lost final sound of the preceding word) definitely and irrefutably established the position of those languages. Among other things that might be credited to his genius, I shall select his explanation of the various declensional classes as determined by the final sound of the stem. But it is not part of my plan to go into many details; suffice it to say that Bopp’s great Vergleichende grammatik served for long years as the best, or really the only, exposition of the new science, and vastly contributed not only to elucidate obscure points, but also to make comparative grammar as popular as it is possible for such a necessarily abstruse science to be.

In Bopp’s Vergleichende grammatik (1. § 108) he gives his classification of languages in general. He rejects Fr. Schlegel’s bipartition, but his growing tendency to explain everything in Aryan grammar, even the inner changes of Sanskrit roots, by mechanical causes makes him modify A. W. Schlegel’s tripartition and place our family of languages with the second instead of the third class. His three classes are therefore as follows: I. Languages without roots proper and without the power of composition, and thus without organism or grammar; to this class belongs Chinese, in which most grammatical relations are only to be recognized by the position of the words. II. Languages with monosyllabic roots, capable of composition and acquiring their organism, their grammar, nearly exclusively in this way; the main principle of word formation is the connexion of verbal and pronominal roots. To this class belong the Indo-European languages, but also all languages not comprised under the first or the third class. III. Languages with disyllabic roots and three necessary consonants as sole bearers of the signification of the word. This class includes only the Semitic languages. Grammatical forms are here created not only by means of composition, as in the second class, but also by inner modification of the roots.

It will be seen that Bopp here expressly avoids both expressions ‘agglutination’ and ‘flexion,’ the former because it had been used of languages contrasted with Aryan, while Bopp wanted to show the essential identity of the two classes; the latter because it had been invested with much obscurity on account of Fr. Schlegel’s use of it to signify inner modification only. According to Schlegel, only such instances as English drink / drank / drunk are pure flexion, while German trink-e / trank / ge-trunk-en, and still more Greek leip-ō / e-lip-on / le-loip-a, besides an element of ‘flexion’ contain also affixed elements. It is clear that no language can use ‘flexion’ (in Schlegel’s sense) exclusively, and consequently this[55] cannot be made a principle on which to erect a classification of languages generally. Schlegel’s use of the term ‘flexion’ seems to have been dropped by all subsequent writers, who use it so as to include what is actually found in the grammar of such languages as Sanskrit and Greek, comprising under it inner and outer modifications, but of course not requiring both in the same form.

In view of the later development of our science, it is worthy of notice that neither in the brothers Schlegel nor in Bopp do we yet meet with the idea that the classes set up are not only a distribution of the languages found side by side in the world at this time, but also represent so many stages in historical development; indeed, Bopp’s definitions are framed so as positively to exclude any development from his Class II to Class III, as the character of the underlying roots is quite heterogeneous. On the other hand, Bopp’s tendency to explain Aryan endings from originally independent roots paved the way for the theory of isolation, agglutination and flexion as three successive stages of the same language.

In his first work (C 56) Bopp had already hinted that in the earliest period known to us languages had already outlived their most perfect state and were in a process of decay; and in his review of Grimm (1827) he repeats this: “We perceive them in a condition in which they may indeed be progressive syntactically, but have, as far as grammar is concerned, lost more or less of what belonged to the perfect structure, in which the separate members stand in exact relation to each other and in which everything derived has still a visible and unimpaired connexion with its source” (Voc. 2). We shall see kindred ideas in Humboldt and Schleicher.

To sum up: Bopp set about discovering the ultimate origin of flexional elements, but instead of that he discovered Comparative Grammar—“à peu près comme Christophe Colomb a découvert l’Amérique en cherchant la route des Indes,” as A. Meillet puts it (LI 413). A countryman of Rask may be forgiven for pushing the French scholar’s brilliant comparison still further: in the same way as Norsemen from Iceland had discovered America before Columbus, without imagining that they were finding the way to India, just so Rasmus Rask through his Icelandic studies had discovered Comparative Grammar before Bopp, without needing to take the circuitous route through Sanskrit.

This will be the proper place to mention one of the profoundest thinkers in the domain of linguistics, Wilhelm von Humboldt (1767-1835), who, while playing an important part in the political[56] world, found time to study a great many languages and to think deeply on many problems connected with philology and ethnography.[8]

In numerous works, the most important of which, Ueber die Kawisprache auf der Insel Jawa, with the famous introduction “Ueber die Verschiedenheit des menschlichen Sprachbaues und ihren Einfluss auf die geistige Entwickelung des Menschengeschlechts,” was published posthumously in 1836-40, Humboldt developed his linguistic philosophy, of which it is not easy to give a succinct idea, as it is largely couched in a most abstruse style; it is not surprising that his admirer and follower, Heymann Steinthal, in a series of books, gave as many different interpretations of Humboldt’s thoughts, each purporting to be more correct than its predecessors. Still, I believe the following may be found to be a tolerably fair rendering of some of Humboldt’s ideas.

He rightly insists on the importance of seeing in language a continued activity. Language is not a substance or a finished work, but action (Sie selbst ist kein werk, ergon, sondern eine tätigkeit, energeia). Language therefore cannot be defined except genetically. It is the ever-repeated labour of the mind to utilize articulated sounds to express thoughts. Strictly speaking, this is a definition of each separate act of speech; but truly and essentially a language must be looked upon as the totality of such acts.[57] For the words and rules, which according to our ordinary notions make up a language, exist really only in the act of connected speech. The breaking up of language into words and rules is nothing but a dead product of our bungling scientific analysis (Versch 41). Nothing in language is static, everything is dynamic. Language has nowhere any abiding place, not even in writing; its dead part must continually be re-created in the mind; in order to exist it must be spoken or understood, and so pass in its entirety into the subject (ib. 63).

Humboldt speaks continually of languages as more perfect or less perfect. Yet “no language should be condemned or depreciated, not even that of the most savage tribe, for each language is a picture of the original aptitude for language” (Versch 304). In another place he speaks about special excellencies even of languages that cannot in themselves be recognized as superlatively good instruments of thought. Undoubtedly Chinese of the old style carries with it an impressive dignity through the immediate succession of nothing but momentous notions; it acquires a simple greatness because it throws away all unnecessary accessory elements and thus, as it were, takes flight to pure thinking. Malay is rightly praised for its ease and the great simplicity of its constructions. The Semitic languages retain an admirable art in the nice discrimination of sense assigned to many shades of vowels. Basque possesses a particular vigour, dependent on the briefness and boldness of expression imparted by the structure of its words and by their combination. Delaware and other American languages express in one word a number of ideas for which we should require many words. The human mind is always capable of producing something admirable, however one-sided it may be; such special points decide nothing with regard to the rank of languages (Versch 189 f.). We have here, as indeed continually in Humboldt, a valuation of languages with many brilliant remarks, but on the whole we miss the concrete details abounding in Jenisch’s work. Humboldt, as it were, lifts us to a higher plane, where the air may be purer, but where it is also thinner and not seldom cloudier as well.

According to Humboldt, each separate language, even the most despised dialect, should be looked upon as an organic whole, different from all the rest and expressing the individuality of the people speaking it; it is characteristic of one nation’s psyche, and indicates the peculiar way in which that nation attempts to realize the ideal of speech. As a language is thus symbolic of the national character of those who speak it, very much in each language had its origin in a symbolic representation of the notion it stands for; there is a natural nexus between certain sounds and certain general ideas, and consequently we often find similar sounds used for the[58] same, or nearly the same, idea in languages not otherwise related to one another.

Humboldt is opposed to the idea of ‘general’ or ‘universal’ grammar as understood in his time; instead of this purely deductive grammar he would found an inductive general grammar, based upon the comparison of the different ways in which the same grammatical notion was actually expressed in a variety of languages. He set the example in his paper on the Dual. His own studies covered a variety of languages; but his works do not give us many actual concrete facts from the languages he had studied; he was more interested in abstract reasonings on language in general than in details.

In an important paper, Ueber das Entstehen der grammatischen Formen und ihren Einfluss auf die Ideenentwickelung (1822), he says that language at first denotes only objects, leaving it to the hearer to understand or guess at (hinzudenken) their connexion. By and by the word-order becomes fixed, and some words lose their independent use and sound, so that in the second stage we see grammatical relations denoted through word-order and through words vacillating between material and formal significations. Gradually these become affixes, but the connexion is not yet firm, the joints are still visible, the result being an aggregate, not yet a unit. Thus in the third stage we have something analogous to form, but not real form. This is achieved in the fourth stage, where the word is one, only modified in its grammatical relations through the flexional sound; each word belongs to one definite part of speech, and form-words have no longer any disturbing material signification, but are pure expressions of relation. Such words as Lat. amavit and Greek epoíēsas are truly grammatical forms in contradistinction to such combinations of words and syllables as are found in cruder languages, because we have here a fusion into one whole, which causes the signification of the parts to be forgotten and joins them firmly under one accent. Though Humboldt thus thinks flexion developed out of agglutination, he distinctly repudiates the idea of a gradual development and rather inclines to something like a sudden crystallization (see especially Steinthal’s ed., p. 585).

Humboldt’s position with regard to the classification of languages is interesting. In his works we continually meet with the terms agglutination[9] and flexion by the side of a new term, ‘incorporation.’ This he finds in full bloom in many American languages, such as Mexican, where the object may be inserted into the verbal form between the element indicating person and the[59] root. Now, Humboldt says that besides Chinese, which has no grammatical form, there are three possible forms of languages, the flexional, the agglutinative and the incorporating, but he adds that all languages contain one or more of these forms (Versch 301). He tends to deny the existence of any exclusively agglutinative or exclusively flexional language, as the two principles are generally commingled (132). Flexion is the only method that gives to the word the true inner firmness and at the same time distributes the parts of the sentence according to the necessary interlacing of thoughts, and thus undoubtedly represents the pure principle of linguistic structure. Now, the question is, what language carries out this method in the most consistent way? True perfection may not be found in any one language: in the Semitic languages we find flexion in its most genuine shape, united with the most refined symbolism, only it is not pursued consistently in all parts of the language, but restricted by more or less accidental laws. On the other hand, in the Sanskritic languages the compact unity of every word saves flexion from any suspicion of agglutination; it pervades all parts of the language and rules it in the highest freedom (Versch 188). Compared with incorporation and with the method of loose juxtaposition without any real word-unity, flexion appears as an intuitive principle born of true linguistic genius (ib.). Between Sanskrit and Chinese, as the two opposed poles of linguistic structure, each of them perfect in the consistent following one principle, we may place all the remaining languages (ib. 326). But the languages called agglutinative have nothing in common except just the negative trait that they are neither isolating nor flexional. The structural diversities of human languages are so great that they make one despair of a fully comprehensive classification (ib. 330).

According to Humboldt, language is in continued development under the influence of the changing mental power of its speakers. In this development there are naturally two definite periods, one in which the creative instinct of speech is still growing and active, and another in which a seeming stagnation begins and then an appreciable decline of that creative instinct. Still, the period of decline may initiate new principles of life and new successful changes in a language (Versch 184). In the form-creating period nations are occupied more with the language than with its purpose, i.e. with what it is meant to signify. They struggle to express thought, and this craving in connexion with the inspiring feeling of success produces and sustains the creative power of language (ib. 191). In the second period we witness a wearing-off of the flexional forms. This is found less in languages reputed crude or rough than in refined ones. Language is exposed to the most[60] violent changes when the human mind is most active, for then it considers too careful an observation of the modifications of sound as superfluous. To this may be added a want of perception of the poetic charm inherent in the sound. Thus it is the transition from a more sensuous to a more intellectual mood that works changes in a language. In other cases less noble causes are at work. Rougher organs and less sensitive ears are productive of indifference to the principle of harmony, and finally a prevalent practical trend may bring about abbreviations and omissions of all kinds in its contempt for everything that is not strictly necessary for the purpose of being understood. While in the first period the elements still recall their origin to man’s consciousness, there is an æsthetic pleasure in developing the instrument of mental activity; but in the second period language serves only the practical needs of life. In this way such a language as English may reduce its forms so as to resemble the structure of Chinese; but there will always remain traces of the old flexions; and English is no more incapable of high excellences than German (Versch 282-6). What these are Humboldt, however, does not tell us.

Humboldt here foreshadowed and probably influenced ideas to which Jacob Grimm gave expression in two essays written in his old age and which it will be necessary here to touch upon. In the essay on the pedantry of the German language (Ueber das pedantische in der deutschen sprache, 1847), Grimm says that he has so often praised his mother-tongue that he has acquired the right once in a while to blame it. If pedantry had not existed already, Germans would have invented it; it is the shadowy side of one of their virtues, painstaking accuracy and loyalty. Grimm’s essay is an attempt at estimating a language, but on the whole it is less comprehensive and less deep than that of Jenisch. Grimm finds fault with such things as the ceremoniousness with which princes are spoken to and spoken of (Durchlauchtigster, allerhöchstderselbe), and the use of the pronoun Sie in the third person plural in addressing a single person; he speaks of the clumsiness of the auxiliaries for the passive, the past and the future, and of the word-order which makes the Frenchman cry impatiently “J’attends le verbe.” He blames the use of capitals for substantives and other peculiarities of German spelling, but gives no general statement of the principles on which the comparative valuation of different languages should be based, though in many passages we see that he places the old stages of the language very much higher than the language of his own day.

[61]

The essay on the origin of language (1851) is much more important, and may be said to contain the mature expression of all Grimm’s thoughts on the philosophy of language. Unfortunately, much of it is couched in that high-flown poetical style which may be partly a consequence of Grimm’s having approached the exact study of language through the less exact studies of popular poetry and folklore; this style is not conducive to clear ideas, and therefore renders the task of the reporter very difficult indeed. Grimm at some length argues against the possibility of language having been either created by God when he created man or having been revealed by God to man after his creation. The very imperfections and changeability of language speak against its divine origin. Language as gradually developed must be the work of man himself, and therein is different from the immutable cries and songs of the lower creation. Nature and natural instinct have no history, but mankind has. Man and woman were created as grown-up and marriageable beings, and there must have been created at once more than one couple, for if there had been only one couple, there would have been the possibility that the one mother had borne only sons or only daughters, further procreation being thus rendered impossible (!), not to mention the moral objections to marriages between brother and sister. How these once created beings, human in every respect except in language, were able to begin talking and to find themselves understood, Grimm does not really tell us; he uses such expressions as ‘inventors’ of words, but apart from the symbolical value of some sounds, such as l and r, he thinks that the connexion of word and sense was quite arbitrary. On the other hand, he can tell us a great deal about the first stage of human speech: it contained only the three vowels a, i, u, and only few consonant groups; every word was a monosyllable, and abstract notions were at first absent. The existence in all (?) old languages of masculine and feminine flexions must be due to the influence of women on the formation of language. Through the distinction of genders Grimm says that regularity and clearness were suddenly brought about in everything concerning the noun as by a most happy stroke of fortune. Endings to indicate person, number, tense and mood originated in added pronouns and auxiliary words, which at first were loosely joined to the root, but later coalesced with it. Besides, reduplication was used to indicate the past; and after the absorption of the reduplicational syllable the same effect was obtained in German through apophony. All nouns presuppose verbs, whose material sense was applied to the designation of things, as when G. hahn (‘cock’) was thus called from an extinct verb hanan, corresponding to Lat. canere, ‘to sing.’

[62]

In what Grimm says about the development of language it is easy to trace the influence of Humboldt’s ideas, though they are worked out with great originality. He discerns three stages, the last two alone being accessible to us through historical documents. In the first period we have the creation and growing of roots and words, in the second the flourishing of a perfect flexion, and in the third a tendency to thoughts, which leads to the giving up of flexion as not yet (?) satisfactory. They may be compared to leaf, blossom and fruit, “the beauty of human speech did not bloom in its beginning, but in its middle period; its ripest fruits will not be gathered till some time in the future.” He thus sums up his theory of the three stages: “Language in its earliest form was melodious, but diffuse and straggling; in its middle form it was full of intense poetical vigour; in our own days it seeks to remedy the diminution of beauty by the harmony of the whole, and is more effective though it has inferior means.” In most places Grimm still speaks of the downward course of linguistic development; all the oldest languages of our family “show a rich, pleasant and admirable perfection of form, in which all material and spiritual elements have vividly interpenetrated each other,” while in the later developments of the same languages the inner power and subtlety of flexion has generally been given up and destroyed, though partly replaced by external means and auxiliary words. On the whole, then, the history of language discloses a descent from a period of perfection to a less perfect condition. This is the point of view that we meet with in nearly all linguists; but there is a new note when Grimm begins vaguely and dimly to see that the loss of flexional forms is sometimes compensated by other things that may be equally valuable or even more valuable; and he even, without elaborate arguments, contradicts his own main contention when he says that “human language is retrogressive only apparently and in particular points, but looked upon as a whole it is progressive, and its intrinsic force is continually increasing.” He instances the English language, which by sheer making havoc of all old phonetic laws and by the loss of all flexions has acquired a great force and power, such as is found perhaps in no other human language. Its wonderfully happy structure resulted from the marriage of the two noblest languages of Europe; therefore it was a fit vehicle for the greatest poet of modern times, and may justly claim the right to be called a world’s language; like the English people, it seems destined to reign in future even more than now in all parts of the earth. This enthusiastic panegyric forms a striking contrast to what the next great German scholar with whom we have to deal, Schleicher, says about the same language, which to him shows only “how rapidly the language of a nation important both in history and literature can decline” (II. 231).

Bopp and Grimm exercised an enormous influence on linguistic thought and linguistic research in Germany and other countries. Long even before their death we see a host of successors following in the main the lines laid down in their work, and thus directly and indirectly they determined the development of this science for a long time. Through their efforts so much new light had been shed on a number of linguistic phenomena that these took a quite different aspect from that which they had presented to the previous generation; most of what had been written about etymology and kindred subjects in the eighteenth century seemed to the new school utterly antiquated, mere fanciful vagaries of incompetent blunderers, whereas now scholars had found firm ground on which to raise a magnificent structure of solid science. This feeling was especially due to the undoubted recognition of one great family of languages to which the vast majority of European languages, as well as some of the most important Asiatic languages, belonged: here we had one firmly established fact of the greatest magnitude, which at once put an end to all the earlier whimsical attempts to connect Latin and Greek words with Hebrew roots. As for the name of that family of languages, Rask hesitated between different names, ‘European,’ ‘Sarmatic’ and finally ‘Japhetic’ (as a counterpart of the Semitic and the Hamitic languages); Bopp at first had no comprehensive name, and on the title-page of his Vergl. grammatik contents himself with enumerating the chief languages described, but in the work itself he says that he prefers the name ‘Indo-European,’ which has also found wide acceptance, though more in France, England and Skandinavia than in Germany. Humboldt for a long while said ‘Sanskritic,’ but later he adopted ‘Indo-Germanic,’ and this has been the generally recognized name used in Germany, in spite of Bopp’s protest who said that ‘Indo-klassisch’ would be more to the point; ‘Indo-[64]Keltic’ has also been proposed as designating the family through its two extreme members to the East and West. But all these compound names are clumsy without being completely pertinent, and it seems therefore much better to use the short and convenient term ‘the Aryan languages’: Aryan being the oldest name by which any members of the family designated themselves (in India and Persia).[10]

Thanks to the labours of Bopp and Grimm and their co-workers and followers, we see also a change in the status of the study of languages. Formerly this was chiefly a handmaiden to philology—but as this word is often in English used in a sense unknown to other languages and really objectionable, namely as a synonym of (comparative) study of languages, it will be necessary first to say a few words about the terminology of our science. In this book I shall use the word ‘philology’ in its continental sense, which is often rendered in English by the vague word ‘scholarship,’ meaning thereby the study of the specific culture of one nation; thus we speak of Latin philology, Greek philology, Icelandic philology, etc. The word ‘linguist,’ on the other hand, is not infrequently used in the sense of one who has merely a practical knowledge of some foreign language; but I think I am in accordance with a growing number of scholars in England and America if I call such a man a ‘practical linguist’ and apply the word ‘linguist’ by itself to the scientific student of language (or of languages); ‘linguistics’ then becomes a shorter and more convenient name for what is also called the science of language (or of languages).

Now that the reader understands the sense in which I take these two terms, I may go on to say that the beginning of the nineteenth century witnessed a growing differentiation between philology and linguistics in consequence of the new method introduced by comparative and by historical grammar; it was nothing less than a completely new way of looking at the facts of language and trying to trace their origin. While to the philologist the Greek or Latin language, etc., was only a means to an end, to the linguist it was an end in itself. The former saw in it a valuable, and in fact an indispensable, means of gaining a first-hand knowledge of the literature which was his chief concern, but the linguist cared not for the literature as such, but studied languages for their own sake, and might even turn to languages destitute of literature because they were able to throw some light on the life of language in general or on forms in related languages. The philologist as such would not think of studying the Gothic of Wulfila, as a know[65]ledge of that language gives access only to a translation of parts of the Bible, the ideas of which can be studied much better elsewhere; but to the linguist Gothic was extremely valuable. The differentiation, of course, is not an absolute one; besides being linguists in the new sense, Rask was an Icelandic philologist, Bopp a Sanskrit philologist, and Grimm a German philologist; but the tendency towards the emancipation of linguistics was very strong in them, and some of their pupils were pure linguists and did no work in philology.

In breaking away from philology and claiming for linguistics the rank of a new and independent science, the partisans of the new doctrine were apt to think that not only had they discovered a new method, but that the object of their study was different from that of the philologists, even when they were both concerned with language. While the philologist looked upon language as part of the culture of some nation, the linguist looked upon it as a natural object; and when in the beginning of the nineteenth century philosophers began to divide all sciences into the two sharply separated classes of mental and natural sciences (geistes- und naturwissenschaften), linguists would often reckon their science among the latter. There was in this a certain amount of pride or boastfulness, for on account of the rapid rise and splendid achievements of the natural sciences at that time, it began to be a matter of common belief that they were superior to, and were possessed of a more scientific method than, the other class—the same view that finds an expression in the ordinary English usage, according to which ‘science’ means natural science and the other domains of human knowledge are termed the ‘arts’ or the ‘humanities.’

We see the new point of view in occasional utterances of the pioneers of linguistic science. Rask expressly says that “Language is a natural object and its study resembles natural history” (SA 2. 502); but when he repeats the same sentence (in Retskrivningslære, 8) it appears that he is thinking of language as opposed to the more artificial writing, and the contrast is not between mental and natural science, but between art and nature, between what can and what cannot be consciously modified by man—it is really a different question.

Bopp, in his review of Grimm (1827, reprinted Vocalismus, 1836, p. 1), says: “Languages are to be considered organic natural bodies, which are formed according to fixed laws, develop as possessing an inner principle of life, and gradually die out because they do not understand themselves any longer [!], and therefore cast off or mutilate their members or forms, which were at first significant, but gradually have become more of an extrinsic mass....[66] It is not possible to determine how long languages may preserve their full vigour of life and of procreation,” etc. This is highly figurative language which should not be taken at its face value; but expressions like these, and the constant use of such words as ‘organic’ and ‘inorganic’ in speaking of formations in languages, and ‘organism’ of the whole language, would tend to widen the gulf between the philological and the linguistic point of view. Bopp himself never consistently followed the naturalistic way of looking at language, but in § 4 of this chapter we shall see that Schleicher was not afraid of going to extremes and building up a consistent natural science of language.

The cleavage between philology and linguistics did not take place without arousing warm feeling. Classical scholars disliked the intrusion of Sanskrit everywhere; they did not know that language and did not see the use of it. They resented the way in which the new science wanted to reconstruct Latin and Greek grammar and to substitute new explanations for those which had always been accepted. Those Sanskritists chatted of guna and vrddhi and other barbaric terms, and even ventured to talk of a locative case in Latin, as if the number of cases had not been settled once for all long ago![11]

Classicists were no doubt perfectly right when they reproached comparativists for their neglect of syntax, which to them was the most important part of grammar; they were also in some measure right when they maintained that linguists to a great extent contented themselves with a superficial knowledge of the languages compared, which they studied more in grammars and glossaries than in living texts, and sometimes they would even exult when they found proof of this in solecisms in Bopp’s Latin translations from Sanskrit, and even on the title-page of Glossarium Sanscritum a Franzisco Bopp. Classical scholars also looked askance at the growing interest in the changes of sounds, or, as it was then usual to say, of letters. But when they were apt here to quote the scriptural phrase about the letter that killeth, while the spirit giveth life, they overlooked the fact that Nature has rendered it impossible for anyone to penetrate to the mind of anyone else except through its outer manifestations, and that it is consequently impossible to get at the spirit of a language except through its sounds: phonology must therefore form the necessary basis and prerequisite of the scientific study of any group of languages. Still, it cannot be denied that sometimes comparative phonology was treated in such a mechanical way as partly to dehumanize the study of language.

[67]

When we look back at this period in the history of linguistics, there are certain tendencies and characteristics that cannot fail to catch our attention. First we must mention the prominence given to Sanskrit, which was thought to be the unavoidable requirement of every comparative linguist. In explaining anything in any of the cognate languages the etymologist always turned first to Sanskrit words and Sanskrit forms. This standpoint is found even much later, for instance in Max Müller’s Inaugural Address (1868, Ch. 19): “Sanskrit certainly forms the only sound foundation of Comparative Philology, and it will always remain the only safe guide through all its intricacies. A comparative philologist without a knowledge of Sanskrit is like an astronomer without a knowledge of mathematics.” A linguist of a later generation may be excused for agreeing rather with Ellis, who says (Transact. Philol. Soc., 1873-4, 21): “Almost in our own days came the discovery of Sanskrit, and philology proper began—but, alas! at the wrong end. Now, here I run great danger of being misunderstood. Although for a scientific sifting of the nature of language I presume to think that beginning at Sanskrit was unfortunate, yet I freely admit that, had that language not been brought into Europe ... our knowledge of language would have been in a poor condition indeed.... We are under the greatest obligations to those distinguished men who have undertaken to unravel its secrets and to show its connexion with the languages of Europe. Yet I must repeat that for the pure science of language, to begin with Sanskrit was as much beginning at the wrong end as it would have been to commence zoology with palæontology—the relations of life with the bones of the dead.”

Next, Bopp and his nearest successors were chiefly occupied with finding likenesses between the languages treated and discovering things that united them. This was quite natural in the first stage of the new science, but sometimes led to one-sidedness, the characteristic individuality of each language being lost sight of, while forms from many countries and many times were mixed up in a hotch-potch. Rask, on account of his whole mental equipment, was less liable to this danger than most of his contemporaries; but Pott was evidently right when he warned his fellow-students that their comparative linguistics should be supplemented by separative linguistics (Zählmethode, 229), as it has been to a great extent in recent years.

Still another feature of the linguistic science of those days is the almost exclusive occupation of the student with dead languages. It was quite natural that the earliest comparativists should first give their attention to the oldest stages of the languages compared, since these alone enabled them to prove the essential[68] kinship between the different members of the great Aryan family. In Grimm’s grammar nearly all the space is taken up with Gothic, Old High German, Old Norse, etc., and comparatively little is said about recent developments of the same languages. In Bopp’s comparative grammar classical Greek and Latin are, of course, treated carefully, but Modern Greek and the Romanic languages are not mentioned (thus also in Schleicher’s Compendium and in Brugmann’s Grammar), such later developments being left to specialists who were more or less considered to be outside the sphere of Comparative Linguistics and even of the science of language in general, though it would have been a much more correct view to include them in both, and though much more could really be learnt of the life of language from these studies than from comparisons made in the spirit of Bopp.

The earlier stages of different languages, which were compared by linguists, were, of course, accessible only through the medium of writing; we have seen that the early linguists spoke constantly of letters and not of sounds. But this vitiated their whole outlook on languages. These were scarcely ever studied at first-hand, and neither in Bopp nor in Grimm nor in Pott or Benfey do we find such first-hand observations of living spoken languages as play a great rôle in the writings of Rask and impart an atmosphere of soundness to his whole manner of looking at languages. If languages were called natural objects, they were not yet studied as such or by truly naturalistic methods.

When living dialects were studied, the interest constantly centred round the archaic traits in them; every survival of an old form, every trace of old sounds that had been dropped in the standard speech, was greeted with enthusiasm, and the significance of these old characteristics greatly exaggerated, the general impression being that popular dialects were always much more conservative than the speech of educated people. It was reserved for a much later time to prove that this view is completely erroneous, and that popular dialects, in spite of many archaic details, are on the whole further developed than the various standard languages with their stronger tradition and literary reminiscences.

It was from this archæological point of view only that Grimm encouraged the study of dialects, but he expressly advised students not to carry the research too far in the direction of discriminating minutiæ of sounds, because these had little bearing on the history of language as he understood it. In this connexion we may[69] mention an episode in the history of early linguistics that is symptomatic. K. M. Rapp brought out his Versuch einer Physiologie der Sprache nebst historischer Entwickelung der abendländischen Idiome nach physiologischen Grundsätzen in four volumes (1836, 1839, 1840, 1841). A physiological examination into the nature and classification of speech sounds was to serve only as the basis of the historical part, the grandiose plan of which was to find out how Greek, Latin and Gothic sounded, and then to pursue the destinies of these sound systems through the Middle Ages (Byzantine Greek, Old Provençal, Old French, Old Norse, Anglo-Saxon, Old High German) to the present time (Modern Greek, Italian, Spanish, etc., down to Low and High German, with different dialects). To carry out this plan Rapp was equipped with no small knowledge of the earlier stages of these languages and a not contemptible first-hand observation of living languages. He relates how from his childhood he had a “morbidly sharpened ear for all acoustic impressions”; he had early observed the difference between dialectal and educated speech and taken an interest in foreign languages, such as French, Italian and English. He visited Denmark, and there made the acquaintance of and became the pupil of Rask; he often speaks of him and his works in terms of the greatest admiration. After his return he took up the study of Jacob Grimm; but though he speaks always very warmly about the other parts of Grimm’s work, Grimm’s phonology disappointed him. “Grimm’s theory of letters I devoured with a ravenous appetite for all the new things I had to learn from it, but also with heartburning on account of the equally numerous things that warred against the whole of my previous research with regard to the nature of speech sounds; fascinated though I was by what I read, it thus made me incredibly miserable.” He set to his great task with enthusiasm, led by the conviction that “the historical material gives here only one side of the truth, and that the living language in all its branches that have never been committed to writing forms the other and equally important side which is still far from being satisfactorily investigated.” It is easy to understand that Rapp came into conflict with Grimm’s Buchstabenlehre, that had been based exclusively on written forms, and Rapp was not afraid of expressing his unorthodox views in what he himself terms “a violent and arrogating tone.” No wonder, therefore, that his book fell into disgrace with the leaders of linguistics in Germany, who noticed its errors and mistakes, which were indeed numerous and conspicuous, rather than the new and sane ideas it contained. Rapp’s work is extraordinarily little known; in Raumer’s Geschichte der germanischen Philologie and similar works it is not even mentioned, and when I disinterred it[70] from undeserved oblivion in my Fonetik (1897, p. 35; cf. Die neueren Sprachen, vol. xiii, 1904) it was utterly unknown to the German phoneticians of my acquaintance. Yet not only are its phonetic observations[12] deserving of praise, but still more its whole plan, based as it is on a thorough comprehension of the mutual relations of sounds and writing, which led Rapp to use phonetic transcription throughout, even in connected specimens both of living and dead languages; that this is really the only way in which it is possible to obtain a comprehensive and living understanding of the sound-system of any language (as well as to get a clear perception of the extent of one’s own ignorance of it!) has not yet been generally recognized. The science of language would have made swifter and steadier progress if Grimm and his successors had been able to assimilate the main thoughts of Rapp.

Another (and still earlier) work that was overlooked at the time was the little pamphlet Om Aarsagerne til Sprogenes Forandringer (1821) by the Dane J. H. Bredsdorff. Bopp and Grimm never really asked themselves the fundamental question, How is it that language changes: what are the driving forces that lead in course of time to such far-reaching differences as those we find between Sanskrit and Latin, or between Latin and French? Now, this is exactly the question that Bredsdorff treats in his masterly pamphlet. Like Rapp, he was a very good phonetician; but in the pamphlet that concerns us here he speaks not only of phonetic but of other linguistic changes as well. These he refers to the following causes, which he illustrates with well-chosen examples: (1) Mishearing and misunderstanding; (2) misrecollection; (3) imperfection of organs; (4) indolence: to this he inclines to refer nine-tenths of all those changes in the pronunciation of a language that are not due to foreign influences; (5) tendency towards analogy: here he gives instances from the speech of children and explains by analogy such phenomena as the extension of s to all genitives, etc.; (6) the desire to be distinct; (7) the need of expressing new ideas. He recognizes that there are changes that cannot be brought under any of these explanations, e.g. the Gothonic sound shift (cf. above, p. 43 note), and he emphasizes the many ways in which foreign nations or foreign languages may influence a language. Bredsdorff’s explanations may not always be correct;[71] but what constitutes the deep originality of his little book is the way in which linguistic changes are always regarded in terms of human activity, chiefly of a psychological character. Here he was head and shoulders above his contemporaries; in fact, most of Bredsdorff’s ideas, such as the power of analogy, were the same that sixty years later had to fight so hard to be recognized by the leading linguists of that time.[13]

In Rapp, and even more in Bredsdorff, we get a whiff of the scientific atmosphere of a much later time; but most of the linguists of the twenties and following decades (among whom A. F. Pott deserves to be specially named) moved in essentially the same grooves as Bopp and Grimm, and it will not be necessary here to deal in detail with their work.

August Schleicher (1821-68) in many ways marks the culmination of the first period of Comparative Linguistics, as well as the transition to a new period with different aims and, partially at any rate, a new method. His intimate knowledge of many languages, his great power of combination, his clear-cut and always lucid exposition—all this made him a natural leader, and made his books for many years the standard handbooks of linguistic science. Unlike Bopp and Grimm, he was exclusively a linguist, or, as he called it himself, ‘glottiker,’ and never tired of claiming for the science of linguistics (‘glottik’), as opposed to philology, the rank of a separate natural science. Schleicher specialized in Slavonic and Lithuanian; he studied the latter language in its own home and took down a great many songs and tales from the mouths of the peasants; he was for some years a professor in the University of Prague, and there acquired a conversational knowledge of Czech; he spoke Russian, too, and thus in contradistinction to Bopp and Grimm had a first-hand knowledge of more than one foreign language; his interest in living speech is also manifested in his specimens of the dialect of his native town, Volkstümliches aus Sonneberg. When he was a child his father very severely insisted on the constant and correct use of the educated language at home; but the boy, perhaps all the more on account of the paternal prohibition, was deeply attracted to the[72] popular dialect he heard from his playfellows and to the fascinating folklore of the old townspeople, which he was later to take down and put into print. In the preface he says that the acquisition of foreign tongues is rendered considerably easier through the habit of speaking two dialects from childhood.

What makes Schleicher particularly important for the purposes of this volume is the fact that in a long series of publications he put forth not only details of his science, but original and comprehensive views on the fundamental questions of linguistic theory, and that these had great influence on the linguistic philosophy of the following decades. He was, perhaps, the most consistent as well as one of the clearest of linguistic thinkers, and his views therefore deserve to be examined in detail and with the greatest care.

Apart from languages, Schleicher was deeply interested both in philosophy and in natural science, especially botany. From these he fetched many of the weapons of his armoury, and they coloured the whole of his theory of language. In his student days at Tübingen he became an enthusiastic adherent of the philosophy of Hegel, and not even the Darwinian sympathies and views of which he became a champion towards the end of his career made him abandon the doctrines of his youth. As for science, he says that naturalists make us understand that in science nothing is of value except facts established through strictly objective observation and the conclusions based on such facts—this is a lesson that he thinks many of his colleagues would do well to take to heart. There can be no doubt that Schleicher in his practice followed a much more rigorous and sober method than his predecessors, and that his Compendium in that respect stands far above Bopp’s Grammar. In his general reasonings on the nature of language, on the other hand, Schleicher did not always follow the strict principles of sober criticism, being, as we shall now see, too dependent on Hegelian philosophy, and also on certain dogmatic views that he had inherited from previous German linguists, from Schlegel downwards.

The Introductions to Schleicher’s two first volumes are entirely Hegelian, though with a characteristic difference, for in the first he says that the changes to be seen in the realm of languages are decidedly historical and in no way resemble the changes that we may observe in nature, for “however manifold these may be, they never show anything but a circular course that repeats itself continually” (Hegel), while in language, as in everything mental, we may see new things that have never existed before. One generation of animals or plants is like another; the skill of animals has no history, as human art has; language is specifically human and mental: its development is therefore analogous to history, for in[73] both we see a continual progress to new phases. In Schleicher’s second volume, however, this view is expressly rejected in its main part, because Schleicher now wants to emphasize the natural character of language: it is true, he now says, that language shows a ‘werden’ which may be termed history in the wider sense of this word, but which is found in its purest form in nature; for instance, in the growing of a plant. Language belongs to the natural sphere, not to the sphere of free mental activity, and this must be our starting-point if we would discover the method of linguistic science (ii. 21).

It would, of course, be possible to say that the method of linguistic science is that of natural science, and yet to maintain that the object of linguistics is different from that of natural science, but Schleicher more and more tends to identify the two, and when he was attacked for saying, in his pamphlet on the Darwinian theory, that languages were material things, real natural objects, he wrote in defence Ueber die bedeutung der sprache für die naturgeschichte des menschen, which is highly characteristic as the culminating point of the materialistic way of looking at languages. The activity, he says, of any organ, e.g. one of the organs of digestion, or the brain or muscles, is dependent on the constitution of that organ. The different ways in which different species, nay even different individuals, walk are evidently conditioned by the structure of the limbs; the activity or function of the organ is, as it were, nothing but an aspect of the organ itself, even if it is not always possible by means of the knife or microscope of the scientist to demonstrate the material cause of the phenomenon. What is true of the manner of walking is true of language as well; for language is nothing but the result, perceptible through the ear, of the action of a complex of material substances in the structure of the brain and of the organs of speech, with their nerves, bones, muscles, etc. Anatomists, however, have not yet been able to demonstrate differences in the structures of these organs corresponding to differences of nationality—to discriminate, that is, the organs of a Frenchman (quâ Frenchman) from those of a German (quâ German). Accordingly, as the chemist can only arrive at the elements which compose the sun by examining the light which it emits, while the source of that light remains inaccessible to him, so must we be content to study the nature of languages, not in their material antecedents but in their audible manifestations. It makes no great difference, however, for “the two things stand to each other as cause and effect, as substance and phenomenon: a philosopher [i.e. a Hegelian] would say that they are identical.”

Now I, for one, fail to understand how this can be what Schleicher believes it to be, “a refutation of the objection that language is[74] nothing but a consequence of the activity of these organs.” The sun exists independently of the human observer; but there could be no such thing as language if there was not besides the speaker a listener who might become a speaker in his turn. Schleicher speaks continually in his pamphlet as if structural differences in the brain and organs of speech were the real language, and as if it were only for want of an adequate method of examining this hidden structure that we had to content ourselves with studying language in its outward manifestation as audible speech. But this is certainly on the face of it preposterous, and scarcely needs any serious refutation. If the proof of the pudding is in the eating, the proof of a language must be in the hearing and understanding; but in order to be heard words must first be spoken, and in these two activities (that of producing and that of perceiving sounds) the real essence of language must consist, and these two activities are the primary (or why not the exclusive?) object of the science of language.

Schleicher goes on to meet another objection that may be made to his view of the ‘substantiality of language,’ namely, that drawn from the power of learning other languages. Schleicher doubts the possibility of learning another language to perfection; he would admit this only in the case of a man who exchanged his mother-tongue for another in his earliest youth; “but then he becomes by that very fact a different being from what he was: brain and organs of speech develop in another direction.” If Mr. So-and-So is said to speak and write German, English and French equally well, Schleicher first inclines to doubt the fact; and then, granting that the same individual may “be at the same time a German, a Frenchman and an Englishman,” he asks us to remember that all these three languages belong to the same family and may, from a broader point of view, be termed species of the same language; but he denies the possibility of anyone’s being equally at home in Chinese and German, or in Arabic and Hottentot, etc., because these languages are totally different in their innermost essence. (But what of bilingual children in Finland, speaking Swedish and Finnish, or in Greenland, speaking Danish and Eskimo, or in Java, speaking Dutch and Malay?) Schleicher has to admit that our organs are to some extent flexible and capable of acquiring activities that they had not at first; but one definite function is and remains nevertheless the only natural one, and thus “the possibility of a man’s acquiring foreign languages more or less perfectly is no objection to our seeing the material basis of language in the structure of the brain and organs of speech.”

Even if we admit that Schleicher is so far right that in nearly all (or all?) cases of bilingualism one language comes more naturally[75] than the other, he certainly exaggerates the difference, which is always one of degree; and at any rate his final conclusion is wrong, for we might with the same amount of justice say that a man who has first learned to play the piano has acquired the structure of brain and fingers peculiar to a pianist, and that it is then unnatural for him also to learn to play the violin, because that would imply a different structure of these organs. In all these cases we have to do with a definite proficiency or skill, which can only be obtained by constant practice, though of course one man may be better predisposed by nature for it than another; but then it is also the fact that people who speak no foreign language attain to very different degrees of proficiency in the use of their mother-tongue. It cannot be said too emphatically that we have here a fundamental question, and that Schleicher’s view can never lead to a true conception of what language is, or to a real insight into its changes and historical development.

Schleicher goes on to say that the classification of mankind into races should not be based on the formation of the skull or on the character of the hair, or any such external criteria, as they are by no means constant, but rather on language, because this is a thoroughly constant criterion. This alone would give a perfectly natural system, one, for instance, in which all Turks would be classed together, while otherwise the Osmanli Turk belongs to the ‘Caucasian’ race and the so-called Tataric Turks to the ‘Mongolian’ race; on the other hand, the Magyar and the Basque are not physically to be distinguished from the Indo-European, though their languages are widely dissimilar. According to Schleicher, therefore, the natural system of languages is also the natural system of mankind, for language is closely connected with the whole higher life of men, which is therefore taken into consideration in and with their language. In this book I am not concerned with the ethnographical division of mankind into races, and I therefore must content myself with saying that the very examples adduced by Schleicher seem to me to militate against his theory that a division of mankind based on language is the natural one: are we to reckon the Basque’s son, who speaks nothing but French (or Spanish) as belonging to a different race from his father? And does not Schleicher contradict himself when on p. 16 he writes that language is “ein völlig constantes merkmal,” and p. 20 that it is “in fortwährender veränderung begriffen”? So far as I see, Schleicher never expressly says that he thinks that the physical structure conditioning the structure of a man’s language is hereditary, though some of his expressions point that way, and that may be what he means by the expression ‘constant.’ In other places (Darw. 25, Bed. 24) he allows external conditions[76] of life to exercise some influence on the character of a language, as when languages of neighbouring peoples are similar (Aryans and Semites, for example, are the only nations possessing flexional languages). On such points, however, he gives only a few hints and suggestions.

In the question of the classification of languages Schleicher introduces a deductive element from his strong preoccupation with Hegelian ideas. Hegel everywhere moves in trilogies; Schleicher therefore must have three classes, and consequently has to tack together two of Pott’s four classes (agglutinating and incorporating); then he is able philosophically to deduce the tripartition. For language consists in meaning (bedeutung; matter, contents, root) and relation (beziehung; form), tertium non datur. As it would be a sheer impossibility for a language to express form only, we obtain three classes:

I. Here meaning is the only thing indicated by sound; relation is merely suggested by word-position: isolating languages.

II. Both meaning and relation are expressed by sound, but the formal elements are visibly tacked on to the root, which is itself invariable: agglutinating languages.

III. The elements of meaning and of relation are fused together or absorbed into a higher unity, the root being susceptible of inward modification as well as of affixes to denote form: flexional languages.

Schleicher employs quasi-mathematical formulas to illustrate these three classes: if we denote a root by R, a prefix by p and a suffix by s, and finally use a raised x to denote an inner modification, we see that in the isolated languages we have nothing but R (a sentence may be represented by R R R R ...), a word in the second class has the formula R s or p R or p R s, but in the third class we may have p R^x s (or R^x s).

Now, according to Schleicher the three classes of languages are not only found simultaneously in the tongues of our own day, but they represent three stages of linguistic development; “to the nebeneinander of the system corresponds the nacheinander of history.” Beyond the flexional stage no language can attain; the symbolic denotation of relation by flexion is the highest accomplishment of language; speech has here effectually realized its object, which is to give a faithful phonetic image of thought. But before a language can become flexional it must have passed through an isolating and an agglutinating period. Is this theory borne out by historical facts? Can we trace back[77] any of the existing flexional languages to agglutination and isolation? Schleicher himself answers this question in the negative: the earliest Latin was of as good a flexional type as are the modern Romanic languages. This would seem a sort of contradiction in terms; but the orthodox Hegelian is ready with an answer to any objection; he has the word of his master that History cannot begin till the human spirit becomes “conscious of its own freedom,” and this consciousness is only possible after the complete development of language. The formation of Language and History are accordingly successive stages of human activity. Moreover, as history and historiography, i.e. literature, come into existence simultaneously, Schleicher is enabled to express the same idea in a way that “is only seemingly paradoxical,” namely, that the development of language is brought to a conclusion as soon as literature makes its appearance; this is a crisis after which language remains fixed; language has now become a means, instead of being the aim, of intellectual activity. We never meet with any language that is developing or that has become more perfect; in historical times all languages move only downhill; linguistic history means decay of languages as such, subjugated as they are through the gradual evolution of the mind to greater freedom.

The reader of the above survey of previous classifications will easily see that in the matter itself Schleicher adds very little of his own. Even the expressions, which are here given throughout in Schleicher’s own words, are in some cases recognizable as identical with, or closely similar to, those of earlier scholars.

He made one coherent system out of ideas of classification and development already found in others. What is new is the philosophical substructure of Hegelian origin, and there can be no doubt that Schleicher imagined that by this addition he contributed very much towards giving stability and durability to the whole system. And yet this proved to be the least stable and durable part of the structure, and as a matter of fact the Hegelian reasoning is not repeated by a single one of those who give their adherence to the classification. Nor can it be said to carry conviction, and undoubtedly it has seemed to most linguists at the same time too rigid and too unreal to have any importance.

But apart from the philosophical argument the classification proved very successful in the particular shape it had found in Schleicher. Its adoption into two such widely read works as Max Müller’s and Whitney’s Lectures on the Science of Language contributed very much to the popularity of the system, though the former’s attempt at ascribing to the tripartition a sociological[78] importance by saying that juxtaposition (isolation) is characteristic of the ‘family stage,’ agglutination of ‘the nomadic stage’ and amalgamation (flexion) of the ‘political stage’ of human society was hardly taken seriously by anybody.

The chief reasons for the popularity of this classification are not far to seek. It is easy of handling and appeals to the natural fondness for clear-cut formulas through its specious appearance of regularity and rationality. Besides, it flatters widespread prejudices in so far as it places the two groups of languages highest that are spoken by those nations which have culturally and religiously exercised the deepest influence on the civilization of the world, Aryans and Semites. Therefore also Pott’s view, according to which the incorporating or ‘polysynthetic’ American languages possess the same characteristics that distinguish flexion as against agglutination, only in a still higher degree, is generally tacitly discarded, for obviously it would not do to place some languages of American Indians higher than Sanskrit or Greek. But when these are looked upon as the very flower of linguistic development it is quite natural to regard the modern languages of Western Europe as degenerate corruptions of the ancient more highly flexional languages; this is in perfect keeping with the prevalent admiration for classical antiquity and with the belief in a far past golden age. Arguments such as these may not have been consciously in the minds of the framers of the ordinary classification, but there can be no doubt that they have been unconsciously working in favour of the system, though very little thought seems to be required to show the fallacy of the assumption that high civilization has any intrinsic and necessary connexion with the grammatical construction of the language spoken by the race or nation concerned. No language of modern Europe presents the flexional type in a purer shape than Lithuanian, where we find preserved nearly the same grammatical system as in old Sanskrit, yet no one would assert that the culture of Lithuanian peasants is higher than that of Shakespeare, whose language has lost an enormous amount of the old flexions. Culture and language must be appraised separately, each on its own merits and independently of the other.

From a purely linguistic point of view there are many objections to the usual classification, and it will be well here to bring them together, though this will mean an interruption of the historical survey which is the main object of these chapters.

First let us look upon the tripartition as purporting a comprehensive classification of languages as existing side by side without any regard to historic development (the nebeneinander[79] of Schleicher). Here it does not seem to be an ideal manner of classifying a great many objects to establish three classes of such different dimensions that the first comprises only Chinese and some other related languages of the Far East, and the third only two families of languages, while the second includes hundreds of unrelated languages of the most heterogeneous character. It seems certain that the languages of Class I represent one definite type of linguistic structure, and it may be that Aryan and Semitic should be classed together on account of the similarity of their structure, though this is by no means quite certain and has been denied (by Bopp, and in recent times by Porzezinski); but what is indubitable is that the ‘agglutinating’ class is made to comprehend languages of the most diverse type, even if we follow Pott and exclude from this class all incorporating languages. Finnish is always mentioned as a typically agglutinative language, yet there we meet with such declensional forms as nominative vesi ‘water,’ toinen ‘second,’ partitive vettä, toista, genitive veden, toisen, and such verbal forms as sido-n ‘I bind,’ sido-t ‘thou bindest,’ sito-o ‘he binds,’ and the three corresponding persons in the plural, sido-mme, sido-tte, sito-vat. Here we are far from having one unchangeable root to which endings have been glued, for the root itself undergoes changes before the endings. In Kiyombe (Congo) the perfect of verbs is in many cases formed by means of a vowel change that is a complete parallel to the apophony in English drink, drank, thus vanga ‘do,’ perfect venge, twala ‘bring,’ perfect twele or twede, etc. (Anthropos, ii. p. 761). Examples like these show that flexion, in whatever way we may define this term, is not the prerogative of the Aryans and Semites, but may be found in other nations as well. ‘Agglutination’ is either too vague a term to be used in classification, or else, if it is taken strictly according to the usual definition, it is too definite to comprise many of the languages which are ordinarily reckoned to belong to the second class.

It will be seen, also, that those writers who aim at giving descriptions of a variety of human tongues, or of them all, do not content themselves with the usual three classes, but have a greater number. This began with Steinthal, who in various works tried to classify languages partly from geographical, partly from structural points of view, without, however, arriving at any definite or consistent system. Friedrich Müller, in his great Grundriss der Sprachwissenschaft, really gives up the psychological or structural division of languages, distributing the more than hundred different languages that he describes among twelve races of mankind, characterized chiefly by external criteria that have nothing to do with language. Misteli establishes six main types: I. Incorporating. II. Root-[80]isolating. III. Stem-isolating. IV. Affixing (Anreihende). V. Agglutinating. VI. Flexional. These he also distributes so as to form four classes: (1) languages with sentence-words: I; (2) languages with no words: II, III and IV; (3) languages with apparent words: V; and (4) languages with real words: VI. But the latter division had better be left alone; it turns on the intricate question “What constitutes a word?” and ultimately depends on the usual depreciation of ‘inferior races’ and corresponding exaltation of our own race, which is alone reputed capable of possessing ‘real words.’ I do not see why we should not recognize that the vocables of Greenlandic, Malay, Kafir or Finnish are just as ‘real’ words as any in Hebrew or Latin.

Our final result, then, is that the tripartition is insufficient and inadequate to serve as a comprehensive classification of languages actually existing. Nor shall we wonder at this if we see the way in which the theory began historically in an obiter dictum of Fr. v. Schlegel at a time when the inner structure of only a few languages had been properly studied, and if we consider the lack of clearness and definiteness inherent in such notions as agglutination and flexion, which are nevertheless made the corner-stones of the whole system. We therefore must go back to the wise saying of Humboldt quoted on p. 59, that the structural diversities of languages are too great for us to classify them comprehensively.

In a subsequent part of this work I shall deal with the tripartition as representing three successive stages in the development of such languages as our own (the nacheinander of Schleicher), and try to show that Schleicher’s view is not borne out by the facts of linguistic history, which give us a totally different picture of development.

From both points of view, then, I think that the classification here considered deserves to be shelved among the hasty generalizations in which the history of every branch of science is unfortunately so rich.

Probably Schleicher’s most original and important contribution to linguistics was his reconstruction of the Proto-Aryan language, die indogermanische ursprache. The possibility of inferentially constructing this parent language, which to Sanskrit, Greek, Latin, Gothic, etc., was what Latin was to Italian, Spanish, French, etc., was early in his thoughts (see quotations illustrating the gradual growth of the idea in Oertel, p. 39 f.), but it was not till the first edition of his Compendium that he carried it out in[81] detail, giving there for each separate chapter (vowels, consonants, roots, stem-formation, declension, conjugation) first the Proto-Aryan forms and then those actually found in the different languages, from which the former were inferred. This arrangement has the advantage that the reader everywhere sees the historical evolution in the natural order, beginning with the oldest and then proceeding to the later stages, just as the Romanic scholar begins with Latin and then takes in successive stages Old French, Modern French, etc. But in the case of Proto-Aryan this procedure is apt to deceive the student and make him take these primitive forms as something certain, whose existence reposes on just as good evidence as the forms found in Sanskrit literature or in German or English as spoken in our own days. When he finds some forms given first and used to explain some others, there is some danger of his forgetting that the forms given first have a quite different status to the others, and that their only raison d’être is the desire of a modern linguist to explain existing forms in related languages which present certain similarities as originating from a common original form, which he does not find in his texts and has, therefore, to reconstruct. But apart from this there can be no doubt that the reconstruction of older forms (and the ingenious device, due to Schleicher, of denoting such forms by means of a preposed asterisk to distinguish them from forms actually found) has been in many ways beneficial to historical grammar. Only it may be questioned whether Schleicher did not go too far when he wished to base the whole grammar of all the Aryan languages on such reconstructions, instead of using them now and then to explain single facts.

Schleicher even ventured (and in this he seems to have had no follower) to construct an entire little fable in primitive Aryan: see “Eine fabel in indogermanischer ursprache,” Beiträge zur vergl. sprachforschung, 5. 206 (1868). In the introductory remarks he complains of the difficulty of such attempts, chiefly because of the almost complete lack of particles capable of being inferred from the existing languages, but he seems to have entertained no doubt about the phonetic and grammatical forms of the words he employed. As the fable is not now commonly known, I give it here, with Schleicher’s translation, as a document of this period of comparative linguistics.

The question here naturally arises: Is it possible in the way initiated by Schleicher to reconstruct extinct linguistic stages, and what degree of probability can be attached to the forms thus created by linguists? The answer certainly must be that in some instances the reconstruction may have a very strong degree of probability, namely, if the data on which it is based are unambiguous and the form to be reconstructed is not far removed from that or those actually found; but that otherwise any reconstruction becomes doubtful, and naturally the more so according to the extent of the reconstruction (as when a whole text is constructed) and to the distance in time that intervenes between the known and the unknown stage. If we look at the genitives of Lat. genus and Gr. génos, which are found as generis and génous, it is easy to see that both presuppose a form with s between two vowels, as we see a great many intervocalic s’s becoming r in Latin and disappearing in Greek; but when Schleicher gives as the prototype of both (and of corresponding forms in the other languages) Aryan ganasas, he oversteps the limits of the permissible in so far as he ascribes to the vowels definite sounds not really warranted by the known forms. If we knew the modern Scandinavian languages and English only, we should not hesitate to give to the Proto-Gothonic genitive of the word for ‘mother’ the ending -s, cf. Dan. moders, E. mother’s; but G. der mutter suffices to show that the conclusion is not safe, and as a matter of fact, both in Old Norse and in Old English the genitive of this[83] word is without an s. An analogous case is presented when Schleicher reconstructs the nom. of the word for ‘father’ as patars, because he presupposes -s as the invariable sign of every nom. sg. masc., although in this particular word not a single one of the old languages has -s in the nominative. All Schleicher’s reconstructions are based on the assumption that Primitive Aryan had a very simple structure, only few consonant and fewer vowel sounds, and great regularity in morphology; but, as we shall see, this assumption is completely gratuitous and was exploded only a few years after his death. Gabelentz (Spr 182), therefore, was right when he said, with a certain irony, that the Aryan ursprache had changed beyond recognition in the short time between Schleicher and Brugmann. The moral to be drawn from all this seems to be that hypothetical and starred forms should be used sparingly and with the extremest caution.

With regard to inferential forms denoted by a star, the following note may not be out of place here. Their purely theoretical character is not always realized. An example will illustrate what I mean. If etymological dictionaries give as the origin of F. ménage (OF. maisnage) a Latin form *mansionaticum, the etymology may be correct although such a Latin word may never at any time have been uttered. The word was framed at some date, no one knows exactly when, from the word which at various times had the forms (acc.) mansionem, *masione, maison, by means of the ending which at first had the form -aticum (as in viaticum), and finally (through several intermediate stages) became -age; but at what stage of each the two elements met to make the word which eventually became ménage, no one can tell, so that the only thing really asserted is that if the word had been formed at a very early date (which is far from probable) it would have been mansionaticum. It would, therefore, perhaps be more correct to say that the word is from mansione + -aticum.

Second only to Schleicher among the linguists of those days was Georg Curtius (1820-85), at one time his colleague in the University of Prague. Curtius’s special study was Greek, and his books on the Greek verb and on Greek etymology cleared up a great many doubtful points; he also contributed very much to bridge the gulf between classical philology and Aryan linguistics. His views on general questions were embodied in the book Zur Chronologie der indogermanischen Sprachforschung (1873). While Schleicher died when his fame was at its highest and his theories were seemingly victorious in all the leading circles, Curtius had[84] the misfortune to see a generation of younger men, including some of his own best disciples, such as Brugmann, advance theories that seemed to him to be in conflict with the most essential principles of his cherished science; and though he himself, like Schleicher, had always been in favour of a stricter observance of sound-laws than his predecessors, his last book was a polemic against those younger scholars who carried the same point to the excess of admitting no exceptions at all, who believed in innumerable analogical formations even in the old languages, and whose reconstructions of primitive forms appeared to the old man as deprived of that classical beauty of the ursprache which was represented in his own and Schleicher’s works (Zur Kritik der neuesten Sprachforschung, 1885). But this is anticipating.

If Curtius was a comparativist with a sound knowledge of classical philology, Johan Nikolai Madvig was pre-eminently a classical philologist who took a great interest in general linguistics and brought his critical acumen and sober common sense to bear on many of the problems that exercised the minds of his contemporaries. He was opposed to everything of a vague and mystical nature in the current theories of language and disliked the tendency of some scholars to find deep-lying mysterious powers at the root of linguistic phenomena. But he probably went too far in his rationalism, for example, when he entirely denied the existence of the sound-symbolism on which Humboldt had expatiated. He laid much stress on the identity of the linguistic faculty in all ages: the first speakers had no more intention than people to-day of creating anything systematic or that would be good for all times and all occasions—they could have no other object in view than that of making themselves understood at the moment; hence the want of system which we find everywhere in languages: a different number of cases in singular and plural, different endings, etc. Madvig did not escape some inconsistencies, as when he himself would explain the use of the soft vowel a to denote the feminine gender by a kind of sound-symbolism, or when he thought it possible to determine in what order the different grammatical ideas presented themselves to primitive man (tense relation first in the verb, number before case in the noun). He attached too little value to phonological and etymological research, but on the whole his views were sounder than many which were set forth on the same subjects at the time; his papers, however, were very little known, partly because they were written in Danish, partly because his style was extremely heavy and difficult, and when he finally brought out his Kleine philologische schriften in German (1875), he expressed his regret in the preface at finding that many of the theories he had put forward years before in Danish[85] had in the meantime been independently arrived at by Whitney, who had had the advantage of expressing them in a world-language.

One of the most important features of the period with which we are here dealing is the development of a number of special branches of historical linguistics on a comparative basis. Curtius’s work on Greek might be cited as one example; in the same way there were specialists in Sanskrit (Westergaard and Benfey among others), in Slavonic (Miklosich and Schleicher), in Keltic (Zeuss), etc. Grimm had numerous followers in the Gothonic or Germanic field, while in Romanic philology there was an active and flourishing school, headed by Friedrich Diez, whose Grammatik der romanischen Sprachen and Etymologisches Wörterbuch der romanischen Sprachen were perhaps the best introduction to the methodical study of linguistics that anyone could desire; the writer of these lines looks back with the greatest gratitude to that period of his youth when he had the good fortune to make the acquaintance of these truly classical works. Everything was so well arranged, so carefully thought out and so lucidly explained, that one had everywhere the pleasant feeling that one was treading on firm ground, the more so as the basis of the whole was not an artificially constructed nebulous ursprache, but the familiar forms and words of an historical language. Here one witnessed the gradual differentiation of Latin into seven or eight distinct languages, whose development it was possible to follow century by century in well-authenticated texts. The picture thus displayed before one’s eyes of actual linguistic growth in all domains—sounds, forms, word-formation, syntax—and (a very important corollary) of the interdependence of these domains, could not but leave a very strong impression—not merely enthusiasm for what had been achieved here, but also a salutary skepticism of theories in other fields which had not a similarly solid basis.

Working, as we have seen, in many fields, linguists had now brought to light a shoal of interesting facts affecting a great many languages and had put forth valuable theories to explain these facts; but most of their work remained difficult of access except to the specialist, and very little was done by the experts to impart to educated people in general those results of the new science which might be enjoyed without deeper study. But in 1861 Max Müller gave the first series of those Lectures on the Science of Language which, in numerous editions, did more than anything else to popularize linguistics and served to initiate a great many students into our science. In many ways these lectures were[86] excellently adapted for this purpose, for the author had a certain knack of selecting interesting illustrations and of presenting his subject in a way that tended to create the same enthusiasm for it that he felt himself. But his arguments do not bear a close inspection. Too often, after stating a problem, he is found to fly off at a tangent and to forget what he has set out to prove for the sake of an interesting etymology or a clever paradox. He gives an uncritical acceptance to many of Schleicher’s leading ideas; thus, the science of linguistics is to him a physical science and has nothing to do with philology, which is an historical science. If, however, we look at the book itself, we shall find that everything that he counts on to secure the interest of his reader, everything that made his lectures so popular, is really non-naturalistic: all those brilliant exposés of word-history are really like historical anecdotes in a book on social evolution; they may have some bearing on the fundamental problems, but these are rarely or never treated as real problems of natural science. Nor does he, when taken to task, maintain his view very seriously, but partly retracts it and half-heartedly ensconces himself behind the dictum that everything depends on the definition you give of “physical science” (see especially Ch 234, 442, 497)—thus calling forth Whitney’s retort that “the implication here is that our author has a right at his own good pleasure to lay down such a definition of a physical science as should make the name properly applicable to the study of this particular one among the products of human capacities.... So he may prove that a whale is a fish, if you only allow him to define what a fish is” (M 23 f.).

Though Schleicher and Max Müller in their own day had few followers in defining linguistics as a natural or physical science—the opposite view was taken, for instance, by Curtius (K 154), Madvig and Whitney—there can be no doubt that the naturalistic point of view practically, though perhaps chiefly unconsciously, had wide-reaching effects on the history of linguistic science. It was intimately connected with the problems chiefly investigated and with the way in which they were treated. From Grimm through Pott to Schleicher and his contemporaries we see a growing interest in phonological comparisons; more and more “sound-laws” were discovered, and those found were more and more rigorously applied, with the result that etymological investigation was attended with a degree of exactness of which former generations had no idea. But as these phonological studies were not, as a rule, based on a real, penetrating insight into the nature of speech-sounds, the work of the etymologist tended more and more to be purely mechanical, and the science of language was to a great extent deprived of those elements which are more[87] intimately connected with the human ‘soul.’ Isolated vowels and consonants were compared, isolated flexional forms and isolated words were treated more and more in detail and explained by other isolated forms and words in other languages, all of them being like dead leaves shaken off a tree rather than parts of a living and moving whole. The speaking individual and the speaking community were too much lost sight of. Too often comparativists gained a considerable acquaintance with the sound-laws and the grammatical forms of various languages without knowing much about those languages themselves, or at any rate without possessing any degree of familiarity with them. Schleicher was not blind to the danger of this. A short time before his death he brought out an Indogermanische Chrestomathie (Weimar, 1869), and in the preface he justifies his book by saying that “it is of great value, besides learning the grammar, to be acquainted, however slightly, with the languages themselves. For a comparative grammar of related languages lays stress on what is common to a language and its sisters; consequently, the languages may appear more alike than they are in reality, and their idiosyncrasies may be thrown into the shade. Linguistic specimens form, therefore, an indispensable supplement to comparative grammar.” Other and even more weighty reasons might have been adduced, for grammar is after all only one side of a language, and it is certainly the best plan, if one wants to understand and appreciate the position of any language, to start with some connected texts of tolerable length, and only afterwards to see how its forms are related to and may be explained by those of other languages.

Though the mechanical school of linguists, with whom historical and comparative phonology was more and more an end in itself, prevailed to a great extent, the trend of a few linguists was different. Among these one must especially mention Heymann Steinthal, who drew his inspiration from Humboldt and devoted numerous works to the psychology of language. Unfortunately, Steinthal was greatly inferior to Schleicher in clearness and consistency of thought: “When I read a work of Steinthal’s, and even many parts of Humboldt, I feel as if walking through shifting clouds,” Max Müller remarks, with good reason, in a letter (Life, i. 256). This obscurity, in connexion with the remoteness of Steinthal’s studies, which ranged from Chinese to the language of the Mande negroes, but paid little regard to European languages, prevented him from exerting any powerful influence on the linguistic thought of his generation, except perhaps through his emphatic assertion of the truth that language can only be understood and explained by means of psychology: his explanation of syntactic attraction paved the way for much in Paul’s Prinzipien.

[88]

The leading exponent of general linguistics after the death of Schleicher was the American William Dwight Whitney, whose books, Language and the Study of Language (first ed. 1867) and its replica, The Life and Growth of Language (1875), were translated into several languages and were hardly less popular than those of his antagonist, Max Müller. Whitney’s style is less brilliant than Max Müller’s, and he scorns the cheap triumphs which the latter gains by the multiplication of interesting illustrations; he never wearies of running down Müller’s paradoxes and inconsistencies,[14] from which he himself was spared by his greater general solidity and sobriety of thought. The chief point of divergence between them was, as already indicated, that Whitney looked upon language as a human institution that has grown slowly out of the necessity for mutual understanding; he was opposed to all kinds of mysticism, and words to him were conventional signs—not, of course, that he held that there ever was a gathering of people that settled the meaning of each word, but in the sense of “resting on a mutual understanding or a community of habit,” no matter how brought about. But in spite of all differences between the two they are in many respects alike, when viewed from the coign of vantage of the twentieth century: both give expression to the best that had been attained by fifty or sixty years of painstaking activity to elucidate the mysteries of speech, and especially of Aryan words and forms, and neither of them was deeply original enough to see through many of the fallacies of the young science. Consequently, their views on the structure of Proto-Aryan, on roots and their rôle, on the building-up and decay of the form-system, are essentially the same as those of their contemporaries, and many of their theories have now crumbled away, including much of what they probably thought firmly rooted for all time.

In works of this period one frequently meets with expressions of pride and joy in the wonderful results that had been achieved in comparative linguistics in the course of a few decades. Thus Max Müller writes: “All this becomes clear and intelligible by the light of Comparative Grammar; anomalies vanish, exceptions prove the rule, and we perceive more plainly every day how in language, as elsewhere, the conflict between the freedom claimed by each individual and the resistance offered by the community at large establishes in the end a reign of law most wonderful, yet perfectly rational and intelligible”; and again: “There is nothing accidental, nothing irregular, nothing without a purpose and meaning in any part of Greek or Latin grammar. No one who has once discovered this hidden life of language, no one who has once found out that what seemed to be merely anomalous and whimsical in language is but, as it were, a petrification of thought, of deep, curious, poetical, philosophical thought, will ever rest again till he has descended as far as he can descend into the ancient shafts of human speech,” etc. (Ch 41 f.). Whitney says: “The difference between the old haphazard style of etymologizing and the modern scientific method lies in this: that the latter, while allowing everything to be theoretically possible, accepts nothing as actual which is not proved by sufficient evidence; it brings to bear upon each individual case a wide circle of related facts; it imposes upon the student the necessity of extended comparison and cautious deduction; it makes him careful to inform himself as thoroughly as circumstances allow respecting the history of every word he deals with” (L 386). And Benfey, in his Geschichte der Sprachwissenschaft (1869, see pp. 562 f. and 596), arrives at the conclusion that the investigation of Aryan languages has already attained a very great degree of certainty, and that the reconstruction of Primitive Aryan, both in grammar and[90] vocabulary, must be considered as in the main settled in such a way that only some details are still doubtful; thus, it is certain that the first person singular ended in -mi, and that this is a phonetic reduction of the pronoun ma, and that the word for ‘horse’ was akva. This feeling of pride is certainly in a great measure justified if we compare the achievements of linguistic science at that date with the etymologies of the eighteenth century; it must also be acknowledged that 90 per cent. of the etymologies in the best-known Aryan languages which must be recognized as established beyond any reasonable doubt had already been discovered before 1870, while later investigations have only added a small number that may be considered firmly established, together with a great many more or less doubtful collocations. But, on the other hand, in the light of later research, we can now see that much of what was then considered firm as a rock did not deserve the implicit trust then placed in it.

This is true in the first place with regard to the phonetic structure ascribed to Proto-Aryan. A series of brilliant discoveries made about the year 1880 profoundly modified the views of scholars about the consonantal and still more about the vocalic system of our family of languages. This is particularly true of the so-called palatal law.[15] So long as it was taken for granted that Sanskrit had in all essential points preserved the ancient sound system, while Greek and the other languages represented younger stages, no one could explain why Sanskrit in some cases had the palatals c and j (sounds approximately like the initial sounds of E. chicken and joy) where the other languages have the velar sounds k and g. It was now recognized that so far from the distribution of the two classes of sounds in Sanskrit being arbitrary, it followed strict rules,[91] though these were not to be seen from Sanskrit itself. Where Sanskrit a following the consonant corresponded to Greek or Latin o, Sanskrit had velar k or g; where, on the other hand, it corresponded to Greek or Latin e, Sanskrit had palatal c or j. Thus we have, for instance, c in Sansk. ca, ‘and’ = Greek te, Lat. que, but k in kakša = Lat. coxa; the difference between the two consonants in a perfect like cakara, ‘have done,’ is dependent on the same vowel alternation as that of Greek léloipa; c in the verb pacati, ‘cooks,’ as against k in the substantive pakas, ‘cooking,’ corresponds to the vowels in Greek légei as against lógos, etc. All this shows that Sanskrit itself must once have had the vowels e and o instead of a; before the front vowel e the consonant has then been fronted or palatalized, as ch in E. chicken is due to the following front vowel, while k has been preserved before o in cock. Sanskrit is thus shown to be in some important respects less conservative than Greek, a truth which was destined profoundly to modify many theories concerning the whole family of languages. As Curtius said, with some resentment of the change in view then taking place, “Sanskrit, once the oracle of the rising science and trusted blindly, is now put on one side; instead of the traditional ex oriente lux the saying is now in oriente tenebræ” (K 97).

The new views held in regard to Aryan vowels also resulted in a thorough revision of the theory of apophony (ablaut). The great mass of Aryan vowel alternations were shown to form a vast and singularly consistent system, the main features of which may be gathered from the following tabulation of a few select Greek examples, arranged into three columns, each representing one ‘grade’:

It is outside our scope to show how this scheme gives us a natural clue to the vowels in such verbs as E. I ride, II rode, III ridden (2), G. I werde, II ward, III geworden (4), or I binde, II band, III gebunden (5). It will be seen from the Greek examples that grade I is throughout characterized by the vowel e and grade II by the vowel o; as for grade III, the vowel of I and II has entirely disappeared in (1), where there is no vowel between the[92] two consonants, and in (2) and (3), where the element found after e and o and forming a diphthong with these has now become a full (syllabic) vowel i and u by itself. In (4) Sanskrit has in grade III a syllabic r (adrçam = Gr. édrakon), while Greek has ra, or in some instances ar, and Gothonic has ur or or according to the vowel of the following syllable. It was this fact that suggested to Brugmann his theory that in (5) Greek a, Lat. in, Goth. un in the third grade originated in syllabic ṇ, and that tatós thus stood for *tṇtós; he similarly explained Gr. déka, Lat. decem, Gothic taihun, E. ten from *dekṃ with syllabic m. I do not believe that his theory is entirely correct; but so much is certain, that in all instances grade III is characterized by a reduction of the vowel that appears in the two other grades as e and o, and there can be no doubt that this reduction is due to want of stress. This being so, it becomes impossible to consider lip the original root-form, which in leip and loip has been extended, and the new theory of apophony thus disposes of the old theory, based on the Indian grammarians’ view that the shortest form was the root-form, which was then raised through ‘guna’ and ‘vrddhi.’ This now is reversed, and the fuller form is shown to be the oldest, which in some cases was shortened according to a process paralleled in many living languages. Bopp was right in his rejection of Grimm’s theory of an inner, significatory reason for apophony, as apophony is now shown to have been due to a mechanical cause, though a different one from that suggested by Bopp (see above, p. 53); and Grimm was also wrong in another respect, because apophony is found from the first in noun-formations as well as in verbs, where Grimm believed it to have been instituted to indicate tense differences, with which it had originally nothing to do. Apophony even appears in other syllables than the root syllable; the new view thus quite naturally paved the way for skepticism with regard to the old doctrine that Aryan roots were necessarily monosyllabic; and scholars soon began to admit dissyllabic ‘bases’ in place of the old roots; instead of lip, the earliest accessible form thus came to be something like leipo or leipe. In this way the new vowel system had far-reaching consequences and made linguists look upon many problems in a new light. It should be noted, however, that the mechanical explanation of apophony from difference in accent applies only to grade III, in contradistinction to grades I and II; the reason of the alternation between the e of I and the o of II is by no means clear.

The investigations leading to the discovery of the palatal law and the new theory of apophony were only a part of the immense labour of a number of able linguists in the ’seventies[93] and ’eighties, which cleared up many obscure points in Aryan phonology and morphology. One of the most famous discoveries was that of the Dane Karl Verner, that a whole series of consonant alternations in the old Gothonic languages was dependent on accent, and (more remarkable still) on the primeval accent, preserved in its oldest form in Sanskrit only, and differing from that of modern Gothonic languages in resting in some instances on the ending and in others on the root. When it was realized that the fact that German has t in vater, but d in bruder, was due to a different accentuation of the two words three or four thousand years ago, or that the difference between s and r in E. was and were was connected with the fact that perfect singulars in Sanskrit are stressed on the root, but plurals on the ending, this served not only to heighten respect for the linguistic science that was able to demonstrate such truths, but also to increase the feeling that the world of sounds was subject to strict laws comparable to those of natural science.

The ‘blind’ operation of phonetic laws became the chief tenet of a new school of ‘young-grammarians’ or ‘junggrammatiker’ (Brugmann, Delbrück, Osthoff, Paul and others), who somewhat noisily flourished their advance upon earlier linguists and justly roused the anger not only of their own teachers, including Curtius, but also of fellow-students like Johannes Schmidt and Collitz. For some years a fierce discussion took place on the principles of linguistic science, in which young-grammarians tried to prove deductively the truth of their favourite thesis that “Sound-laws admit of no exceptions” (first, it seems, enounced by Leskien). Osthoff wrongly maintained that sound changes belonged to physiology and analogical change to psychology; but though that distribution of the two kinds of change to two different domains was untenable, the distinction in itself was important and proved a valuable, though perhaps sometimes too easy instrument in the hands of the historical grammarian. It was quite natural that those who insisted on undeviating phonetic laws should turn their attention to those cases in which forms appeared that did not conform to these laws, and try to explain them; and thus they inevitably were led to recognize the immense importance of analogical formations in the economy of all languages. Such formations had long been known, but little attention had been paid to them, and they were generally termed ‘false analogies’ and looked upon as corruptions or inorganic formations found only[94] or chiefly in a degenerate age, in which the true meaning and composition of the old forms was no longer understood. Men like Curtius were scandalized at the younger school explaining so many even of the noble forms of ancient Greek as due to this upstart force of analogy. His opponents contended that the name of ‘false analogy’ was wrong and misleading: the analogy in itself was perfect and was handled with unerring instinct in each case. They likewise pointed out that analogical formations, so far from being perversions of a late age, really represented one of the vital principles of language, without which it could never have come into existence.

One of the first to take the new point of view and to explain it clearly was Hermann Paul. I quote from an early article (as translated by Sweet, CP 112) the following passages, which really struck a new note in linguistic theory:

“There is one simple fact which should never be left out of sight, namely, that even in the parent Indogermanic language, long before its split-up, there were no longer any roots, stems, and suffixes, but only ready-made words, which were employed without the slightest thought of their composite nature. And it is only of such ready-made words that the store is composed from which everyone draws when he speaks. He has no stock of stems and terminations at his disposal from which he could construct the form required for each separate occasion. Not that he must necessarily have heard and learnt by heart every form he uses. This would, in fact, be impossible. He is, on the contrary, able of himself to form cases of nouns, tenses of verbs, etc., which he has either never heard or else not noticed specially; but, as there is no combining of stem and suffix, this can only be done on the pattern of the other ready-made combinations which he has learnt from his fellows. These latter are first learnt one by one, and then gradually associated into groups which correspond to the grammatical categories, but are never clearly conceived as such without special training. This grouping not only greatly aids the memory, but also makes it possible to produce other combinations. And this is what we call analogy.”

“It is, therefore, clear that, while speaking, everyone is incessantly producing analogical forms. Reproduction by memory and new-formation by means of association are its two indispensable factors. It is a mistake to assume a language as given in grammar and dictionary, that is, the whole body of possible words and forms, as something concrete, and to forget that it is nothing but an abstraction devoid of reality, and that the actual language exists only in the individual, from whom it cannot be separated even in scientific investigation, if we will understand[95] its nature and development. To comprehend the existence of each separate spoken form, we must not ask ‘Is it current in the language?’ or ‘Is it conformable to the laws of the language as deduced by the grammarians?’ but ‘Has he who has just employed it previously had it in his memory, or has he formed it himself for the first time, and, if so, according to what analogy?’ When, for instance, anyone employs the plural milben in German, it may be that he has learnt it from others, or else that he has only heard the singular milbe, but knows that such words as lerche, schwalbe, etc., form their plural lerchen, etc., so that the association milbe-milben is unconsciously suggested to him. He may also have heard the plural milben, but remembers it so imperfectly that he would forget it entirely were it not associated in his mind with a series of similar forms which help him to recall it. It is, therefore, often difficult to determine the share memory and creative fancy have had in each separate case.”

Linguists thus set about it seriously to think of language in terms of speaking individuals, who have learnt their mother-tongue in the ordinary way, and who now employ it in their daily intercourse with other men and women, without in each separate case knowing what they owe to others and what they have to create on the spur of the moment. Just as Sokrates fetched philosophy down from the skies, so also now linguists fetched words and forms down from vocabularies and grammars and placed them where their natural home is, in the minds and on the lips of ordinary men who are neither lexicographers nor grammarians, but who nevertheless master their language with sufficient ease and correctness for all ordinary purposes. Linguists now were confronted with some general problems which had not greatly troubled their predecessors (with the solitary exception of Bredsdorff, whose work was entirely overlooked), namely, What are the causes of changes in language? How are they brought about, and how should they be classified? Many articles on these questions appeared in linguistic periodicals about the year 1880, but the profoundest and fullest treatment was found in a masterly book by H. Paul, Prinzipien der Sprachgeschichte, the first edition of which (1880) exercised a very considerable influence on linguistic thought, while the subsequent editions were constantly enlarged and improved so as to contain a wealth of carefully sifted material to illustrate the various processes of linguistic change. It should also be noted that Paul paid more and more attention to syntax, and that this part of grammar, which had been neglected by Bopp and Schleicher and their contemporaries, was about this time taken up by some[96] of the leading linguists, who showed that the comparative and historical method was capable of throwing a flood of light on syntax no less than on morphology (Delbrück, Ziemer).

While linguists in the ’eighties were taking up, as we have seen, a great many questions of vast general importance that had not been treated by the older generation, on the other hand they were losing interest in some of the problems that had occupied their predecessors. This was the case with the question of the ultimate origin of grammatical endings. So late as 1869 Benfey included among Bopp’s ‘brilliant discoveries’ his theory that the s of the aorist and of the future was derived from the verb as, ‘to be,’ and that the endings of the Latin imperfect -bam and future -bo were from the synonymous verb fu = Sanskrit bhu (Gesch 377), and the next year Raumer reckons the same theories among Bopp’s ‘most important discoveries.’ But soon after this we see that speculations of this kind somehow go out of fashion. One of the last books to indulge in them to any extent is Scherer’s once famous Zur Geschichte der deutschen Sprache (2nd ed., 1878), in the eighth chapter of which the writer disports himself among primitive roots, endings, prepositions and pronouns, which he identifies and differentiates with such extreme boldness and confidence in his own wild fancies that a sober-minded man of the twentieth century cannot but feel dazed and giddy. The ablest linguists of the new school simply left these theories aside: no new explanations of the same description were advanced, and the old ones were not substantiated by the ascertained phenomena of living languages. So much was found in these of the most absorbing interest that scholars ceased to care for what might lie behind Proto-Aryan; some even went so far as to deprecate in strong expressions any attempts at what they termed ‘glottogonic’ theories. To these matter-of-fact linguists all speculations as to the ultimate origin of language were futile and nebulous, a verdict which might be in no small degree justified by much of what had been written on the subject by quasi-philosophers and quasi-linguists. The aversion to these questions was shown as early as 1866, when La Société de Linguistique was founded in Paris. Section 2 of the statutes of the Society expressly states that “La Société n’admet aucune communication concernant, soit l’origine du langage, soit la création d’une langue universelle”—both of them questions which, as they can be treated in a scientific spirit, should not be left exclusively to dilettanti.

[97]

The last forty years have witnessed an extraordinary activity on the part of scholars in investigating all domains of the Aryan languages in the light of the new general views and by the aid of the methods that have now become common property. Phonological investigations have no doubt had the lion’s share and have to a great extent been signalized by that real insight into physiological phonetics which had been wanting in earlier linguists; but very much excellent work has also been done in morphology, syntax and semantics; and in all these domains much has been gained by considering words not as mere isolated units, but as parts of sentences, or, better, of connected speech. In phonetics more and more attention has been paid to sentence phonetics and ‘sandhi phenomena’; the heightened interest in everything concerning ‘accent’ (stress and pitch) has also led to investigations of sentence-stress and sentence-melody; the intimate connexion between forms and their use or function in the sentence, in other words their syntax, has been more and more recognized; and finally, if semantics (the study of the significations of words) has become a real science instead of being a curiosity shop of isolated specimens, this has only been rendered possible through seeing words as connected with other words to form complete utterances. But this change of attitude could not have been brought about unless linguists had studied texts in the different languages to a far greater extent than had been done in previous periods; thus, naturally, the antagonism formerly often felt between the linguistic and the purely philological study of the same language has tended to disappear, and many scholars have produced work both in their particular branch of linguistics and in the corresponding philology. There can be no doubt that this development has been profitable to both domains of scientific activity.

Another beneficial change is the new attitude taken with regard to the study of living speech. The science of linguistics had long stood in the sign of Cancer and had been constantly looking backwards—to its own great loss. Now, with the greater stress laid on phonetics and on the psychology of language, the necessity of observing the phenomena of actual everyday speech was more clearly perceived. Among pioneers in this respect I must specially mention Henry Sweet; now there is a steadily growing interest in living speech as the necessary foundation of all general theorizing. And with interest comes knowledge.

It is outside the purpose of this volume to give the history of linguistic study during the last forty years in the same way as I have attempted to give it for the period before 1880, and I must therefore content myself with a few brief remarks on[98] general tendencies. I even withstand the temptation to try and characterize the two greatest works on general linguistics that have appeared during this period, those by Georg v. d. Gabelentz and Wilhelm Wundt: important and in many ways excellent as they are, they have not exercised the same influence on contemporary linguistic research as some of their predecessors. Personally I owe incomparably much more to the former than to the latter, who is much less of a linguist than of a psychologist and whose pages seem to me often richer in words than in fertilizing ideas. As for the rest, I can give only a bare alphabetical list of some of the writers who during this period have dealt with the more general problems of linguistic change or linguistic theory, and must not attempt any appreciation of their works: Bally, Baudouin de Courtenay, Bloomfield, Bréal, Delbrück, van Ginneken, Hale, Henry, Hirt, Axel Kock, Meillet, Meringer, Noreen, Oertel, Pedersen, Sandfeld (Jensen), de Saussure, Schuchardt, Sechehaye, Streitberg, Sturtevant, Sütterlin, Sweet, Uhlenbeck, Vossler, Wechssler. In the following parts of my work there will be many opportunities of mentioning their views, especially when I disagree with them, for I am afraid it will be impossible always to indicate what I owe to their suggestions.

In the history of linguistic science we have seen in one period a tendency to certain large syntheses (the classification of languages into isolating, agglutinative and flexional, and the corresponding theory of three periods with its corollary touching the origin of flexional endings), and we have seen how these syntheses were later discredited, though never actually disproved, linguists contenting themselves with detailed comparisons and explanations of single words, forms or sounds without troubling about their ultimate origin or about the evolutionary tendencies of the whole system or structure of language. The question may therefore be raised, were Bopp and Schleicher wrong in attempting these large syntheses? It would appear from the expressions of some modern linguists that they thought that any such comprehensive generalization or any glottogonic theory were in itself of evil. But this can never be admitted. Science, of its very nature, aims at larger and larger generalizations, more and more comprehensive formulas, so as finally to bring about that “unification of knowledge” of which Herbert Spencer speaks. It was therefore quite right of the early linguists to propound those great questions; and their failure to solve them in a way that could satisfy the stricter demands of a later generation should not be charged too heavily against them. It was also quite right of the moderns to reject their premature solutions (though this was often done without any adequate examination), but[99] it was decidedly wrong to put the questions out of court altogether.[16] These great questions have to be put over and over again, till a complete solution is found; and the refusal to face these difficulties has produced a certain barrenness in modern linguistics, which must strike any impartial observer, however much he admits the fertility of the science in detailed investigations. Breadth of vision is not conspicuous in modern linguistics, and to my mind this lack is chiefly due to the fact that linguists have neglected all problems connected with a valuation of language. What is the criterion by which one word or one form should be preferred to another? (most linguists refuse to deal with such questions of preference or of correctness of speech). Are the changes that we see gradually taking place in languages to be considered as on the whole beneficial or the opposite? (most linguists pooh-pooh such questions). Would it be possible to construct an international language by which persons in different countries could easily communicate with one another? (most linguists down to the present day have looked upon all who favour such ideas as visionaries and Utopians). It is my firm conviction that such questions as these admit of really scientific treatment and should be submitted to serious discussion. But before tackling those of them which fall within the plan of this work, it will be well to deal with some fundamental facts of what is popularly called the ‘life’ of language, and first of all with the manner in which a child acquires its mother-tongue. For as language exists only in individuals and means some specific activities of human beings which are not inborn, but have to be learnt by each of them separately from his fellow-beings, it is important to examine somewhat in detail how this interaction of the individual and of the surrounding society is brought about. This, then, will occupy us in Book II.

[103]

A Danish philosopher has said: “In his whole life man achieves nothing so great and so wonderful as what he achieved when he learnt to talk.” When Darwin was asked in which three years of his life a man learnt most, he said: “The first three.”

A child’s linguistic development covers three periods—the screaming time, the crowing or babbling time, and the talking time. But the last is a long one, and must again be divided into two periods—that of the “little language,” the child’s own language, and that of the common language or language of the community. In the former the child is linguistically an individualist, in the latter he is more and more socialized.

Of the screaming time little need be said. A child’s scream is not uttered primarily as a means of conveying anything to others, and so far is not properly to be called speech. But if from the child’s side a scream is not a way of telling anything, its elders may still read something in it and hurry to relieve the trouble. And if the child comes to remark—as it soon will—that whenever it cries someone comes and brings it something pleasant, if only company, it will not be long till it makes use of this instrument whenever it is uneasy or wants something. The scream, which was at first a reflex action, is now a voluntary action. And many parents have discovered that the child has learnt to use its power of screaming to exercise a tyrannical power over them—so that they have had to walk up and down all night with a screaming child that prefers this way of spending the night to lying quietly in its cradle. The only course is brutally to let the baby scream till it is tired, and persist in never letting it get its desire because it screams for it, but only because what it desires is good for it. The child learns its lesson, and a scream is once more what it was at first, an involuntary, irresistible result of the fact that something is wrong.

[104]

Screaming has, however, another side. It is of physiological value as an exercise of all the muscles and appliances which are afterwards to be called into play for speech and song. Nurses say—and there may be something in it—that the child who screams loudest as a baby becomes the best singer later.

Babbling time produces pleasanter sounds which are more adapted for the purposes of speech. Cooing, crowing, babbling—i.e. uttering meaningless sounds and series of sounds—is a delightful exercise like sprawling with outstretched arms and legs or trying to move the tiny fingers. It has been well said that for a long time a child’s dearest toy is its tongue—that is, of course, not the tongue only, but the other organs of speech as well, especially the lips and vocal chords. At first the movements of these organs are as uncontrolled as those of the arms, but gradually they become more systematic, and the boy knows what sound he wishes to utter and is in a position to produce it exactly.

First, then, come single vowels or vowels with a single consonant preceding them, as la, ra, lö, etc., though a baby’s sounds cannot be identified with any of ours or written down with our letters. For, though the head and consequently the mouth capacity is disproportionally great in an infant and grows more rapidly than its limbs, there is still a great difference between its mouth capacity and that required to utter normal speech-sounds. I have elsewhere (PhG, p. 81 ff.) given the results of a series of measurings of the jaw in children and adults and discussed the importance of these figures for phonetic theory: while there is no growth of any importance during the talking period (for a child of five may have the same jaw-length as a man of thirty-seven), the growth is enormous during the first months of a child’s life: in the case of my own child, from 45 mm. a few days after birth to 60 mm. at three months old and 75 mm. at eleven months, while the average of grown-up men is 99 mm. and of women 93 mm. The consequence is that the sounds of the baby are different from ours, and that even when they resemble ours the mechanism of production may be different from the normal one; when my son during the first weeks said something like la, I was able to see distinctly that the tip of the tongue was not at all in the position required for our l. This want of congruence between the acoustic manners of operation in the infant and the adult no doubt gives us the key to many of the difficulties that have puzzled previous observers of small children.

Babbling or crowing begins not earlier than the third week; it may be, not till the seventh or eighth week. The first sound exercises are to be regarded as muscular exercises pure and simple, as is clear from the fact that deaf-mutes amuse themselves with[105] them, although they cannot themselves hear them. But the moment comes when the hearing child finds a pleasure in hearing its own sounds, and a most important step is taken when the little one begins to hear a resemblance between the sounds uttered by its mother or nurse and its own. The mother will naturally answer the baby’s syllables by repeating the same, and when the baby recognizes the likeness, it secures an inexhaustible source of pleasure, and after some time reaches the next stage, when it tries itself to imitate what is said to it (generally towards the close of the first year). The value of this exercise cannot be over-estimated: the more that parents understand how to play this game with the baby—of saying something and letting the baby say it after, however meaningless the syllable-sequences that they make—the better will be the foundation for the child’s later acquisition and command of language.

It is generally said that the order in which the child learns to utter the different sounds depends on their difficulty: the easiest sounds are produced first. That is no doubt true in the main; but when we go into details we find that different writers bring forward lists of sounds in different order. All are agreed, however, that among the consonants the labials, p, b and m, are early sounds, if not the earliest. The explanation has been given that the child can see the working of his mother’s lips in these sounds and therefore imitates her movements. This implies far too much conscious thought on the part of the baby, who utters his ‘ma’ or ‘mo’ before he begins to imitate anything said to him by his surroundings. Moreover, it has been pointed out that the child’s attention is hardly ever given to its mother’s mouth, but is steadily fixed on her eyes. The real reason is probably that the labial muscles used to produce b or m are the same that the baby has exercised in sucking the breast or the bottle. It would be interesting to learn if blind children also produce the labial sounds first.

Along with the labial sounds the baby produces many other sounds—vowel and consonant—and in these cases one is certain that it has not been able to see how these sounds are produced by its mother. Even in the case of the labials we know that what distinguishes m from b, the lowering of the soft palate, and b from p, the vibrations of the vocal chords, is invisible. Some of the sounds produced by means of the tongue may be too hard to pronounce till the muscles of the tongue have been exercised in consequence of the child having begun to eat more solid things than milk.

[106]

By the end of the first year the number of sounds which the little babbler has mastered is already considerable, and he loves to combine long series of the same syllables, dadadada ..., nenenene ..., bygnbygnbygn ..., etc. That is a game which need not even cease when the child is able to talk actual language. It is strange that among an infant’s sounds one can often detect sounds—for instance k, g, h, and uvular r—which the child will find difficulty in producing afterwards when they occur in real words, or which may be unknown to the language which it will some day speak. The explanation lies probably in the difference between doing a thing in play or without a plan—when it is immaterial which movement (sound) is made—and doing the same thing of fixed intention when this sound, and this sound only, is required, at a definite point in the syllable, and with this or that particular sound before and after. Accordingly, great difficulties come to be encountered when the child begins more consciously and systematically to imitate his elders. Some sounds come without effort and may be used incessantly, to the detriment of others which the child may have been able previously to produce in play; and a time even comes when the stock of sounds actually diminishes, while particular sounds acquire greater precision. Dancing masters, singing masters and gymnastic teachers have similar experiences. After some lessons the child may seem more awkward than it was before the lessons began.

The ‘little language’ which the child makes for itself by imperfect imitation of the sounds of its elders seems so arbitrary that it may well be compared to the child’s first rude drawings of men and animals. A Danish boy named Gustav (1.6)[17] called himself [dodado] and turned the name Karoline into [nnn]. Other Danish children made skammel into [gramn] or [gap], elefant into [vat], Karen into [gaja], etc. A few examples from English children: Hilary M. (1.6) called Ireland (her sister) [a·ni], Gordon M. (1.10) called Millicent (his sister) [dadu·]. Tony E. (1.11) called his playmate Sheila [dubabud].

As the child gets away from the peculiarities of his individual ‘little language,’ his speech becomes more regular, and a linguist can in many cases see reasons for his distortions of normal words. When he replaces one sound by another there is always some common element in the formation of the two sounds, which causes[107] a kindred impression on the ear, though we may have difficulty in detecting it because we are so accustomed to noticing the difference. There is generally a certain system in the sound substitutions of children, and in many instances we are justified in speaking of ‘strictly observed sound-laws.’ Let us now look at some of these.

Children in all countries tend to substitute [t] for [k]: both sounds are produced by a complete stoppage of the breath for the moment by the tongue, the only difference being that it is the back of the tongue which acts in one case, and the tip of the tongue in the other. A child who substitutes t for k will also substitute d for g; if he says ‘tat’ for ‘cat’ he will say ‘do’ for ‘go.’

R is a difficult sound. Hilary M. (2.0) has no r’s in her speech. Initially they become w, as in [wʌn] for ‘run,’ medially between vowels they become l, as in [veli, beli] for ‘very, berry,’ in consonantal combinations they are lost, as in [kai, bʌʃ] for ‘cry, brush.’ Tony E. (1.10 to 3.0) for medial r between vowels first substituted d, as in [vedi] for ‘very,’ and later g [vegi]; similarly in [mu·gi] for ‘Muriel,’ [tægi] for ‘carry’; he often dropped initial r, e.g. oom for ‘room.’ It is not unusual for children who use w for r in most combinations to say [tʃ] for tr and [dʒ] for dr, as in ‘chee,’ ‘jawer’ for ‘tree,’ ‘drawer.’ This illustrates the fact that what to us is one sound, and therefore represented in writing by one letter, appears to the child’s ear as different sounds—and generally the phonetician will agree with the child that there are really differences in the articulation of the sound according to position in the syllable and to surroundings, only the child exaggerates the dissimilarities, just as we in writing one and the same letter exaggerate the similarity.

The two th sounds offer some difficulties and are often imitated as f and v respectively, as in ‘frow’ and ‘muvver’ for ‘throw’ and ‘mother’; others say ‘ze’ or ‘de’ for ‘the.’ Hilary M. (2.0) has great difficulty with th and s; th usually becomes [ʃ], [beʃ, ti·ʃ, ʃri·] for ‘Beth,’ ‘teeth,’ ‘three’; s becomes [ʃ], e.g. [franʃiʃ, ʃti·m] for ‘Francis,’ ‘steam’; in the same way z becomes [ʒ] as in [lʌbʒ, bouʒ] for ‘loves,’ ‘Bowes’; sw becomes [fw] as in [fwiŋ, fwi·t] for ‘swing,’ ‘sweet.’ She drops l in consonantal combinations, e.g. [ki·n, kaim, kɔk, ʃi·p] for ‘clean,’ ‘climb,’ ‘clock,’ ‘sleep.’

Sometimes it requires a phonetician’s knowledge to understand the individual sound-laws of a child. Thus I pick out from some specimens given by O’Shea, p. 135 f. (girl, 2.9), the following words: pell (smell), teeze (sneeze), poke (smoke), tow (snow), and formulate the rule: s + a nasal became the voiceless stop corre[108]sponding to the nasal, a kind of assimilation, in which the place of articulation and the mouth-closure of the nasals were preserved, and the sound was made unvoiced and non-nasal as the s. In other combinations m and n were intact.

Some further faults are illustrated in Tony E.’s [tʃouz, pʌg, pus, tæm, pʌm, bæk, pi·z, nouʒ, ɔk, es, u·] for clothes, plug, push, tram, plum, black, please, nose, clock, yes, you.

Even when a sound by itself can be pronounced, the child often finds it hard to pronounce it when it forms part of a group of sounds. S is often dropped before another consonant, as in ‘tummy’ for ‘stomach.’ Other examples have already been given above. Hilary M. (2.0) had difficulty with lp and said [hæpl] for ‘help.’ She also said [ointən] for ‘ointment’; C. M. L. (2.3) said ‘sikkums’ for ‘sixpence.’ Tony E. (2.0) turns grannie into [nægi]. When initial consonant groups are simplified, it is generally, though not always, the stop that remains: b instead of bl-, br-, k instead of kr-, sk-, skr-, p instead of pl-, pr-, spr-, etc. For the groups occurring medially and finally no general rule seems possible.

To begin with, the child is unable to master long sequences of syllables; he prefers monosyllables and often emits them singly and separated by pauses. Even in words that to us are inseparable wholes some children will make breaks between syllables, e.g. Shef-field, Ing-land. But more often they will give only part of the word, generally the last syllable or syllables; hence we get pet-names like Bet or Beth for Elizabeth and forms like ‘tatoes’ for potatoes, ‘chine’ for machine, ‘tina’ for concertina, ‘tash’ for moustache, etc. Hilary M. (1.10) called an express-cart a press-cart, bananas and pyjamas nanas and jamas.

It is not, however, the production of long sequences of syllables in itself that is difficult to the child, for in its meaningless babbling it may begin very early to pronounce long strings of sounds without any break; but the difficulty is to remember what sounds have to be put together to bring about exactly this or that word. We grown-up people may experience just the same sort of difficulty if after hearing once the long name of a Bulgarian minister or a Sanskrit book we are required to repeat it at once. Hence we should not wonder at such pronunciations as [pekəlout] for petticoat or [efelənt] for elephant (Beth M., 2.6); Hilary M. called a caterpillar[109] a pillarcat. Other transpositions are serreval for several and ocken for uncle; cf. also wops for wasp.

To explain the frequent reduplications found in children’s language it is not necessary, as some learned authors have done, to refer to the great number of reduplicated words in the languages of primitive tribes and to see in the same phenomenon in our own children an atavistic return to primitive conditions, on the Häckelian assumption that the development of each individual has to pass rapidly through the same (‘phylogenetic’) stages as the whole lineage of his ancestors. It is simpler and more natural to refer these reduplications to the pleasure always felt in repeating the same muscular action until one is tired. The child will repeat over and over again the same movements of legs and arms, and we do the same when we wave our hand or a handkerchief or when we nod our head several times to signify assent, etc. When we laugh we repeat the same syllable consisting of h and a more or less indistinct vowel, and when we sing a melody without words we are apt to ‘reduplicate’ indefinitely. Thus also with the little ones. Apart from such words as papa and mamma, to which we shall have to revert in another chapter (VIII, § 8), children will often form words from those of their elders by repeating one syllable; cf. puff-puff, gee-gee. Tracy (p. 132) records pepe for ‘pencil,’ kaka for ‘Carrie.’ For a few weeks (1.11) Hilary M. reduplicated whole words, e.g. king-king, ring-ring (i.e. bell), water-water. Tony F. (1.10) uses [touto] for his own name. Hence pet-names like Dodo; they are extremely frequent in French—for instance, Fifine, Lolotte, Lolo, Mimi; the name Daudet has arisen in a similar way from Claudet, a diminutive of Claude.

It is a similar phenomenon (a kind of partial reduplication) when sounds at a distance affect one another, as when Hilary M. (2.0) said [gɔgi] for doggie, [bɔbin] for Dobbin, [dezmən di·n] for Jesmond Dene, [baikikl] for bicycle, [kekl] for kettle. Tracy (p. 133) mentions bopoo for ‘bottle,’ in which oo stands for the hollow sound of syllabic l. One correspondent mentions whoofing-cough for ‘whooping-cough’ (where the final sound has crept into the first word) and chicken-pops for ‘chicken-pox.’ Some children say ‘aneneme’ for anemone; and in S. L. (4.9) this caused a curious confusion during the recent war: “Mother, there must be two sorts of anenemies, flowers and Germans.”

Dr. Henry Bradley once told me that his youngest child had a difficulty with the name Connie, which was made alternatingly [tɔni] and [kɔŋi], in both cases with two consonants articulated at the same point. Similar instances are mentioned in German books on children’s language, thus gigarr for ‘zigarre,’ baibift[110] for ‘bleistift,’ autobobil (Meringer),[18] fotofafieren (Stern), ambam for ‘armband,’ dan for ‘dame,’ pap for ‘patte’ (Ronjat). I have given many Danish examples in my Danish book. Grammont’s child (see Mélanges linguistiques offerts à A. Meillet, 1902) carried through these changes in a most systematic way.

The time comes when the child corrects his mistakes—where it said ‘tat’ it now says ‘cat.’ Here there are two possibilities which both seem to occur in actual life. One is that the child hears the correct sound some time before he is able to imitate it correctly; he will thus still say t for k, though he may in some way object to other people saying ‘tum’ for ‘come.’ Passy relates how a little French girl would say tosson both for garçon and cochon; but she protested when anybody else said “C’est un petit cochon” in speaking about a boy, or vice versa. Such a child, as soon as it can produce the new sound, puts it correctly into all the places where it is required. This, I take it, is the ordinary procedure. Frans (my own boy) could not pronounce h and said an, on for the Danish pronouns han, hun; but when he began to pronounce this sound, he never misplaced it (2.4).

The other possibility is that the child learns how to pronounce the new sound at a time when its own acoustic impression is not yet quite settled; in that case there will be a period during which his use of the new sound is uncertain and fluctuating. When parents are in too great a hurry to get a child out of some false pronunciation, they may succeed in giving it a new sound, but the child will tend to introduce it in places where it does not belong. On the whole, it seems therefore the safest plan to leave it to the child itself to discover that its sound is not the correct one.

Sometimes a child will acquire a sound or a sound combination correctly and then lose it till it reappears a few months later. In an English family where there was no question of the influence of h-less servants, each child in succession passed through an h-less period, and one of the children, after pronouncing h correctly, lost the use of it altogether for two or three months. I have had similar experiences with Danish children. S. L. (ab. 2) said ‘bontin’ for bonnet; but five months earlier she had said bonnet correctly.

The path to perfection is not always a straight one. Tony E. in order to arrive at the correct pronunciation of please passed through the following stages: (1) [bi·], (2) [bli·], (3) [pi·z],[111] (4) [pwi·ʒ], (5) [beisk, meis, mais] and several other impossible forms. Tracy (p. 139) gives the following forms through which the boy A. (1.5) had to pass before being able to say pussy: pooheh, poofie, poopoohie, poofee. A French child had four forms [mèni, pèti, mèti, mèsi] before being able to say merci correctly (Grammont). A Danish child passed through bejab and vamb before pronouncing svamp (‘sponge’), etc.

It is certain that all this while the little brain is working, and even consciously working, though at first it has not sufficient command of speech to say anything about it. Meringer says that children do not practise, but that their new acquisitions of sounds happen at once without any visible preparation. He may be right in the main with regard to the learning of single sounds, though even there I incline to doubt the possibility of a universal rule; but Ronjat (p. 55) is certainly right as against Meringer with regard to the way in which children learn new and difficult combinations. Here they certainly do practise, and are proudly conscious of the happy results of their efforts. When Frans (2.11) mastered the combination fl, he was very proud, and asked his mother: “Mother, can you say flyve?”; then he came to me and told me that he could say bluse and flue, and when asked whether he could say blad, he answered: “No, not yet; Frans cannot say b-lad” (with a little interval between the b and the l). Five weeks later he said: “Mother, won’t you play upon the klaver (piano)?” and after a little while, “Frans can say kla so well.” About the same time he first mispronounced the word manchetter, and then (when I asked what he was saying, without telling him that anything was wrong) he gave it the correct sound, and I heard him afterwards in the adjoining room repeat the word to himself in a whisper.

How well children observe sounds is again seen by the way in which they will correct their elders if they give a pronunciation to which they are not accustomed—for instance, in a verse they have learnt by heart. Beth M (2.6) was never satisfied with her parents’ pronunciation of “What will you buy me when you get there?” She always insisted on their gabbling the first words as quickly as they could and then coming out with an emphatic there.

As to the differences in the tone of a voice, even a baby shows by his expression that he can distinguish clearly between what is said to him lovingly and what sharply, a long time before he understands a single word of what is said. Many children are[112] able at a very early age to hit off the exact note in which something is said or sung. Here is a story of a boy of more advanced age. In Copenhagen he had had his hair cut by a Swedish lady and did not like it. When he travelled with his mother to Norway, as soon as he entered the house, he broke out with a scream: “Mother, I hope I’m not going to have my hair cut?” He had noticed the Norwegian intonation, which is very like the Swedish, and it brought an unpleasant association of ideas.

In the preceding chapter, in order to simplify matters, we have dealt with sounds only, as if they were learnt by themselves and independently of the meanings attached to them. But that, of course, is only an abstraction: to the child, as well as to the grown-up, the two elements, the outer, phonetic element, and the inner element, the meaning, of a word are indissolubly connected, and the child has no interest, or very little interest, in trying to imitate the sounds of its parents except just in so far as these mean something. That words have a meaning, the child will begin to perceive at a very early age. Parents may of course deceive themselves and attribute to the child a more complete and exact understanding of speech than the child is capable of. That the child looks at its father when it hears the word ‘father,’ may mean at first nothing more than that it follows its mother’s glance; but naturally in this way it is prepared for actually associating the idea of ‘father’ with the sound. If the child learns the feat of lifting its arms when it is asked “How big is the boy?” it is not to be supposed that the single words of the sentence are understood, or that the child has any conception of size; he only knows that when this series of sounds is said he is admired if he lifts his arms up: and so the sentence as a whole has the effect of a word of command. A dog has the same degree of understanding. Hilary M. (1.0), when you said to her at any time the refrain “He greeted me so,” from “Here come three knights from Spain,” would bow and salute with her hand, as she had seen some children doing it when practising the song.

The understanding of what is said always precedes the power of saying the same thing oneself—often precedes it for an extraordinarily long time. One father notes that his little daughter of a year and seven months brings what is wanted and understands questions while she cannot say a word. It often happens that[114] parents some fine day come to regret what they have said in the presence of a child without suspecting how much it understands. “Little pitchers have long ears.”

One can, however, easily err in regard to the range and certainty of a child’s understanding. The Swiss philologist Tappolet noticed that his child of six months, when he said “Where is the window?” made vague movements towards the window. He made the experiment of repeating his question in French—with the same intonation as in German, and the child acted just as it had done before. It is, properly speaking, only when the child begins to talk that we can be at all sure what it has really understood, and even then it may at times be difficult to sound the depths of the child’s conception.

The child’s acquisition of the meaning of words is truly a highly complicated affair. How many things are comprehended under one word? The answer is not easy in all cases. The single Danish word tæppe covers all that is expressed in English by carpet, rug, blanket, counterpane, curtain (theatrical). And there is still more complication when we come to abstract ideas. The child has somehow to find out for himself with regard to his own language what ideas are considered to hang together and so come under the same word. He hears the word ‘chair’ applied to a particular chair, then to another chair that perhaps looks to him totally different, and again to a third: and it becomes his business to group these together.

What Stern tells about his own boy is certainly exceptional, perhaps unique. The boy ran to a door and said das? (‘That?’—his way of asking the name of a thing). They told him ‘tür.’ He then went to two other doors in the room, and each time the performance was repeated. He then did the same with the seven chairs in the room. Stern says, “As he thus makes sure that the objects that are alike to his eye and to his sense of touch have also the same name, he is on his way to general conceptions.” We should, however, be wary of attributing general ideas to little children.

In the first period we meet the same phenomena in the child’s acquisition of word-meanings that we found in his acquisition of sounds. A child develops conceptions of his own which are as unintelligible and strange to the uninitiated as his sounds.

Among the child’s first passions are animals and pictures of animals, but for a certain time it is quite arbitrary what animals are classed together under a particular name. A child of nine[115] months noticed that his grandfather’s dog said ‘bow-wow’ and fancied that anything not human could say (and therefore should be called) bow-wow—pigs and horses included. A little girl of two called a horse he (Danish hest) and divided the animal kingdom into two groups, (1) horses, including all four-footed things, even a tortoise, and (2) fishes (pronounced iz), including all that moved without use of feet, for example, birds and flies. A boy of 1.8 saw a picture of a Danish priest in a ruff and was told that it was a præst, which he rendered as bæp. Afterwards seeing a picture of an aunt with a white collar which recalled the priest’s ruff, he said again bæp, and this remained the name of the aunt, and even of another aunt, who was called ‘other bæp.’ These transferences are sometimes extraordinary. A boy who had had a pig drawn for him, the pig being called öf, at the age of 1.6 used öf (1) for a pig, (2) for drawing a pig, (3) for writing in general.

Such transferences may seem very absurd, but are not more so than some transferences occurring in the language of grown-up persons. The word Tripos passed from the sense of a three-legged stool to the man who sat on a three-legged stool to dispute with candidates for degrees at Cambridge. Then, as it was the duty of Mr. Tripos also to provide comic verses, these were called tripos verses, such verses being printed under that name till very near the end of the nineteenth century, though Mr. Tripos himself had disappeared long ago. And as the examination list was printed on the back of these verses, it was called the Tripos list, and it was no far cry to saying of a successful candidate, “he stands high on the Tripos,” which now came to mean the examination itself.

But to return to the classifications in the minds of the children. Hilary M. (1.6 to 2.0) used the word daisy (1) of the flower itself, (2) of any flower, (3) of any conventional flower in a pattern, (4) of any pattern. One of the first words she said was colour (1.4), and she got into a way of saying it when anything striking attracted her attention. Originally she heard the word of a bright patch of colour in a picture. The word was still in use at the age of two. For some months anything that moved was a fly, every man was a soldier, everybody that was not a man was a baby. S. L. (1.8) used bing (1) for a door, (2) for bricks or building with bricks. The connexion is through the bang of a door or a tumbling castle of bricks, but the name was transferred to the objects. It is curious that at 1.3 she had the word bang for anything dropped, but not bing; at 1.8 she had both, bing being specialized as above. From books about children’s language I quote two illustrations. Ronjat’s son used the word papement, which stands for ‘kaffemensch,’ in speaking about the[116] grocer’s boy who brought coffee; but as he had a kind of uniform with a flat cap, papement was also used of German and Russian officers in the illustrated papers. Hilde Stern (1.9) used bichu for drawer or chest of drawers; it originated in the word bücher (books), which was said when her picture-books were taken out of the drawer.

A warning is, however, necessary. When a grown-up person says that a child uses the same word to denote various things, he is apt to assume that the child gives a word two or three definite meanings, as he does. The process is rather in this way. A child has got a new toy, a horse, and at the same time has heard its elders use the word ‘horse,’ which it has imitated as well as it can. It now associates the word with the delight of playing with its toy. If the next day it says the same sound, and its friends give it the horse, the child gains the experience that the sound brings the fulfilment of its wish: but if it sets its eye on a china cow and utters the same sound, the father takes note that the sound also denotes a cow, while for the child it is perhaps a mere experiment—“Could not I get my wish for that nice thing fulfilled in the same way?” If it succeeds, the experiment may very well be repeated, and the more or less faulty imitation of the word ‘horse’ thus by the co-operation of those around it may become also firmly attached to ‘cow.’

When Elsa B. (1.10), on seeing the stopper of a bottle in the garden, came out with the word ‘beer,’ it would be rash to conclude (as her father did) that the word ‘beer’ to her meant a ‘stopper’: all we know is that her thoughts had taken that direction, and that some time before, on seeing a stopper, she had heard the word ‘beer.’

Parents sometimes unconsciously lead a child into error about the use of words. A little nephew of mine asked to taste his father’s beer, and when refused made so much to-do that the father said, “Come, let us have peace in the house.” Next day, under the same circumstances, the boy asked for ‘peace in the house,’ and this became the family name for beer. Not infrequently what is said on certain occasions is taken by the child to be the name of some object concerned; thus a sniff or some sound imitating it may come to mean a flower, and ‘hurrah’ a flag. S. L. from an early age was fond of flowers, and at 1.8 used ‘pretty’ or ‘pretty-pretty’ as a substantive instead of the word ‘flower,’ which she learnt at 1.10.

I may mention here that analogous mistakes may occur when missionaries or others write down words from foreign languages with which they are not familiar. In the oldest list of Greenlandic words (of 1587) there is thus a word panygmah given with[117] the signification ‘needle’; as a matter of fact it means ‘my daughter’s’: the Englishman pointed at the needle, but the Eskimo thought he wanted to know whom it belonged to. In an old list of words in the now extinct Polabian language we find “scumbe, yesterday, subuda, to-day, janidiglia, to-morrow”: the questions were put on a Saturday, and the Slav answered accordingly, for subuta (the same word as Sabbath) means Saturday, skumpe ‘fasting-day,’ and ja nedila ‘it is Sunday.’

According to O’Shea (p. 131) “a child was greatly impressed with the horns of a buck the first time he saw him. The father used the term ‘sheep’ several times while the creature was being inspected, and it was discovered afterwards that the child had made the association between the word and the animal’s horns, so now sheep signifies primarily horns, whether seen in pictures or in real life.” It is clear that mistakes of that kind will happen more readily if the word is said singly than when it is embodied in whole connected sentences: the latter method is on the whole preferable for many reasons.

A child is often faced by some linguistic usage which obliges him again and again to change his notions, widen them, narrow them, till he succeeds in giving words the same range of meaning that his elders give them.

Frequently, perhaps most frequently, a word is at first for the child a proper name. ‘Wood’ means not a wood in general, but the particular picture which has been pointed out to the child in the dining-room. The little girl who calls her mother’s black muff ‘muff,’ but refuses to transfer the word to her own white one, is at the same stage. Naturally, then, the word father when first heard is a proper name, the name of the child’s own father. But soon it must be extended to other individuals who have something or other in common with the child’s father. One child will use it of all men, another perhaps of all men with beards, while ‘lady’ is applied to all pictures of faces without beards; a third will apply the word to father, mother and grandfather. When the child itself applies the word to another man it is soon corrected, but at the same time it cannot avoid hearing another child call a strange man ‘father’ or getting to know that the gardener is Jack’s ‘father,’ etc. The word then comes to mean to the child ‘a grown-up person who goes with or belongs to a little one,’ and he will say, “See, there goes a dog with his father.” Or, he comes to know that the cat is the kittens’ father, and the dog the puppies’ father, and next day asks, “Wasps, are they the flies’[118] father, or are they perhaps their mother?” (as Frans did, 4.10). Finally, by such guessing and drawing conclusions he gains full understanding of the word, and is ready to make acquaintance later with its more remote applications, as ‘The King is the father of his people; Father O’Flynn; Boyle was the father of chemistry,’ etc.

Difficulties are caused to the child when its father puts himself on the child’s plane and calls his wife ‘mother’ just as he calls his own mother ‘mother,’ though at other moments the child hears him call her ‘grandmother’ or ‘grannie.’ Professor Sturtevant writes to me that a neighbour child, a girl of about five years, called out to him, “I saw your girl and your mother,” meaning ‘your daughter and your wife.’ In many families the words ‘sister’ (‘Sissie’) or ‘brother’ are used constantly instead of his or her real name. Here we see the reason why so often such names of relations change their meaning in the history of languages; G. vetter probably at first meant ‘father’s brother,’ as it corresponds to Latin patruus; G. base, from ‘father’s sister,’ came to mean also ‘mother’s sister,’ ‘niece’ and ‘cousin.’ The word that corresponds etymologically to our mother has come to mean ‘wife’ or ‘woman’ in Lithuanian and ‘sister’ in Albanian.

The same extension that we saw in the case of ‘father’ now may take place with real proper names. Tony E. (3.5), when a fresh charwoman came, told his mother not to have this Mary: the last charwoman’s name was Mary.[19] In exactly the same way a Danish child applied the name of their servant, Ingeborg, as a general word for servant: “Auntie’s Ingeborg is called Ann,” etc., and a German girl said viele Augusten for ‘many girls.’ This, of course, is the way in which doll has come to mean a ‘toy baby,’ and we use the same extension when we say of a statesman that he is no Bismarck, etc.

The association of a word with its meaning is accomplished for the child by a series of single incidents, and as many words are understood only by the help of the situation, it is natural that the exact force of many of them is not seized at once. A boy of 4.10, hearing that his father had seen the King, inquired, “Has he a head at both ends?”—his conception of a king being derived from playing-cards. Another child was born on what the Danes call Constitution Day, the consequence being that he confused birthday and Constitution Day, and would speak of “my Consti[119]tution Day,” and then his brother and sister also began to talk of their Constitution Day.

Hilary M. (2.0) and Murdoch D. (2.6) used dinner, breakfast and tea interchangeably—the words might be translated ‘meal.’ Other more or less similar confusions may be mentioned here. Tony F. (2.8) used the term sing for (1) reading, (2) singing, (3) any game in which his elders amused him. Hilary said indifferently, ‘Daddy, sing a story three bears,’ and ‘Daddy, tell a story three bears.’ She cannot remember which is knife and which is fork. Beth M. (2.6) always used can’t when she meant won’t. It meant simply refusal to do what she did not want to.

It is interesting to watch the way in which arithmetical notions grow in extent and clearness. Many children learn very early to say one, two, which is often said to them when they learn how to walk; but no ideas are associated with these syllables. In the same way many children are drilled to say three when the parents begin with one, two, etc. The idea of plurality is gradually developed, but a child may very well answer two when asked how many fingers papa has; Frans used the combinations some-two and some-three to express ‘more than one’ (2.4). At the age of 2.11 he was very fond of counting, but while he always got the first four numbers right, he would skip over 5 and 7; and when asked to count the apples in a bowl, he would say rapidly 1-2-3-4, even if there were only three, or stop at 3, even if there were five or more. At 3.4 he counted objects as far as 10 correctly, but might easily pass from 11 to 13, and if the things to be counted were not placed in a row he was apt to bungle by moving his fingers irregularly from one to another. When he was 3.8 he answered the question “What do 2 and 2 make?” quite correctly, but next day to the same question he answered “Three,” though in a doubtful tone of voice. This was in the spring, and next month I noted: “His sense of number is evidently weaker than it was: the open-air life makes him forget this as well as all the verses he knew by heart in the winter.” When the next winter came his counting exercises again amused him, but at first he was in a fix as before about any numbers after 6, although he could repeat the numbers till 10 without a mistake. He was fond of doing sums, and had initiated this game himself by asking: “Mother, if I have two apples and get one more, haven’t I then three?” His sense of numbers was so abstract that he was caught by a tricky question: “If you have two eyes and one nose, how many ears have you?” He answered at once, “Three!” A child thus seems to think in[120] abstract numbers, and as he learns his numbers as 1, 2, 3, 4, etc., not as one pear, two pears, three pears, one may well be skeptical about the justification for the recommendation made by many pedagogues that at an early stage of the school-life a child should learn to reckon with concrete things rather than with abstract numbers.

A child will usually be familiar with the sound of higher numerals long before it has any clear notion of what they mean. Frans (3.6) said, “They are coming by a train that is called four thirty-four,” and (4.4) he asked, “How much is twice hundred? Is that a thousand?”

A child’s ideas of time are necessarily extremely vague to begin with; it cannot connect very clear or very definite notions with the expressions it constantly hears others employ, such as ‘last Sunday,’ ‘a week ago,’ or ‘next year.’ The other day I heard a little girl say: “This is where we sat next time,” evidently meaning ‘last time.’ All observers of children mention the frequent confusion of words like to-morrow and yesterday, and the linguist remembers that Gothic gistradagis means ‘to-morrow,’ though it corresponds formally with E. yesterday and G. gestern.

Very small children will often say up both when they want to be taken up and when they want to be put down on the floor. This generally means nothing else than that they have not yet learnt the word down, and up to them simply is a means to obtain a change of position. In the same way a German child used hut auf for having the hat taken off as well as put on, but Meumann rightly interprets this as an undifferentiated desire to have something happen with the hat. But even with somewhat more advanced children there are curious confusions.

Hilary M. (2.0) is completely baffled by words of opposite meaning. She will say, “Daddy, my pinny is too hot; I must warm it at the fire.” She goes to the fire and comes back, saying, “That’s better; it’s quite cool now.” (The same confusion of hot and cold was also reported in the case of one Danish and one German child; cf. also Tracy, p. 134.) One morning while dressing she said, “What a nice windy day,” and an hour or two later, before she had been out, “What a nasty windy day.” She confuses good and naughty completely. Tony F. (2.5) says, “Turn the dark out.”

Sometimes a mere accidental likeness may prove too much for the child. When Hilary M. had a new doll (2.0) her mother said to her: “And is that your son?” Hilary was puzzled, and[121] looking out of the window at the sun, said: “No, that’s my sun.” It was very difficult to set her out of this confusion.[20] Her sister Beth (3.8), looking at a sunset, said: “That’s what you call a sunset; where Ireland (her sister) is (at school) it’s a summerset.” About the same time, when staying at Longwood Farm, she said: “I suppose if the trees were cut down it would be Shortwood Farm?”

An English friend writes to me: “I misunderstood the text, ‘And there fell from his eyes as it were scales,’ as I knew the word scales only in the sense ‘balances.’ The phenomenon seemed to me a strange one, but I did not question that it occurred, any more than I questioned other strange phenomena recounted in the Bible. In the lines of the hymn—

I supposed that the words ‘as little as my bed’ were descriptive of my future grave, and that it was my duty according to the hymn to fear the grave.”

Words with several meanings may cause children much difficulty. A Somerset child said, “Moses was not a good boy, and his mother smacked ’un and smacked ’un and smacked ’un till she couldn’t do it no more, and then she put ’un in the ark of bulrushes.” This puzzled the teacher till he looked at the passage in Exodus: “And when she could hide him no longer, she laid him in an ark of bulrushes.” Here, of course, we have technically two different words hide; but to the child the difficulty is practically as great where we have what is called one and the same word with two distinct meanings, or when a word is used figuratively.

The word ‘child’ means two different things, which in some languages are expressed by two distinct words. I remember my own astonishment at the age of nine when I heard my godmother talk of her children. “But you have no children.” “Yes, Clara and Eliza.” I knew them, of course, but they were grown up.

Take again the word old. A boy knew that he was three years, but could not be induced to say ‘three years old’; no, he is three years new, and his father too is new, as distinct from his grandmother, who he knows is old. A child asked, “Why have grand dukes and grand pianos got the same name?” (Glenconner, p. 21).

When Frans was told (4.4) “Your eyes are running,” he was much astonished, and asked, “Are they running away?”

[122]

Sometimes a child knows a word first in some secondary sense. When a country child first came to Copenhagen and saw a soldier, he said, “There is a tin-soldier” (2.0). Stern has a story about his daughter who was taken to the country and wished to pat the backs of the pigs, but was checked with the words, “Pigs always lie in dirt,” when she was suddenly struck with a new idea; “Ah, that is why they are called pigs, because they are so dirty: but what would people call them if they didn’t lie in the dirt?” History repeats itself: only the other day a teacher wrote to me that one of his pupils had begun his essay with the words: “Pigs are rightly called thus, for they are such swine.”

Words of similar sound are apt to be confused. Some children have had trouble till mature years with soldier and shoulder, hassock and cassock, diary and dairy. Lady Glenconner writes: “They almost invariably say ‘lemon’ [for melon], and if they make an effort to be more correct they still mispronounce it. ‘Don’t say melling.’ ‘Very well, then, mellum.’” Among other confusions mentioned in her book I may quote Portugal for ‘purgatory,’ King Solomon’s three hundred Columbines, David and his great friend Johnson, Cain and Mabel—all of them showing how words from spheres beyond the ordinary ken of children are assimilated to more familiar ones.

Schuchardt has a story of a little coloured boy in the West Indies who said, “It’s three hot in this room”: he had heard too = two and literally wanted to ‘go one better.’ According to Mr. James Payne, a boy for years substituted for the words ‘Hallowed be Thy name’ ‘Harold be Thy name.’ Many children imagine that there is a pole to mark where the North Pole is, and even (like Helen Keller) that polar bears climb the Pole.

This leads us naturally to what linguists call ‘popular etymology’—which is very frequent with children in all countries. I give a few examples from books. A four-year-old boy had heard several times about his nurse’s neuralgia, and finally said: “I don’t think it’s new ralgia, I call it old ralgia.” In this way anchovies are made into hamchovies, whirlwind into worldwind, and holiday into hollorday, a day to holloa. Professor Sturtevant writes: A boy of six or seven had frequently had his ear irrigated; when similar treatment was applied to his nose, he said that he had been ‘nosigated’—he had evidently given his own interpretation to the first syllable of irrigate.

There is an element of ‘popular etymology’ in the following joke which was made by one of the Glenconner children when four years old: “I suppose you wag along in the wagonette, the landau lands you at the door, and you sweep off in the brougham” (pronounced broom).

[123]

A class of words which presents grave difficulty to children are those whose meaning differs according to the situation, so that the child hears them now applied to one thing and now to another. That was the case with words like ‘father,’ and ‘mother.’ Another such word is ‘enemy.’ When Frans (4.5) played a war-game with Eggert, he could not get it into his head that he was Eggert’s enemy: no, it was only Eggert who was the enemy. A stronger case still is ‘home.’ When a child was asked if his grandmother had been at home, and answered: “No, grandmother was at grandfather’s,” it is clear that for him ‘at home’ meant merely ‘at my home.’ Such words may be called shifters. When Frans (3.6) heard it said that ‘the one’ (glove) was as good as ‘the other,’ he asked, “Which is the one, and which is the other?”—a question not easy to answer.

The most important class of shifters are the personal pronouns. The child hears the word ‘I’ meaning ‘Father,’ then again meaning ‘Mother,’ then again ‘Uncle Peter,’ and so on unendingly in the most confusing manner. Many people realize the difficulty thus presented to the child, and to obviate it will speak of themselves in the third person as ‘Father’ or ‘Grannie’ or ‘Mary,’ and instead of saying ‘you’ to the child, speak of it by its name. The child’s understanding of what is said is thus facilitated for the moment: but on the other hand the child in this way hears these little words less frequently and is slower in mastering them.

If some children soon learn to say ‘I’ while others speak of themselves by their name, the difference is not entirely due to the different mental powers of the children, but must be largely attributed to their elders’ habit of addressing them by their name or by the pronouns. But Germans would not be Germans, and philosophers would not be philosophers, if they did not make the most of the child’s use of ‘I,’ in which they see the first sign of self-consciousness. The elder Fichte, we are told, used to celebrate not his son’s birthday, but the day on which he first spoke of himself as ‘I.’ The sober truth is, I take it, that a boy who speaks of himself as ‘Jack’ can have just as full and strong a perception of himself as opposed to the rest of the world as one who has learnt the little linguistic trick of saying ‘I.’ But this does not suit some of the great psychologists, as seen from the following quotation: “The child uses no pronouns; it speaks of itself in the third person, because it has no idea of its ‘I’ (Ego) nor of its ‘Not-I,’ because it knows nothing of itself nor of others.”

[124]

It is not an uncommon case of confusion for a child to use ‘you’ and ‘your’ instead of ‘I,’ ‘me,’ and ‘mine.’ The child has noticed that ‘will you have?’ means ‘will Jack have?’ so that he looks on ‘you’ as synonymous with his own name. In some children this confusion may last for some months. It is in some cases connected with an inverted word-order, ‘do you’ meaning ‘I do’—an instance of ‘echoism’ (see below). Sometimes he will introduce a further complication by using the personal pronoun of the third person, as though he had started the sentence with ‘Jack’—then ‘you have his coat’ means ‘I have my coat.’ He may even speak of the person addressed as ‘I.’ ‘Will I tell a story?’ = ‘Will you tell a story?’ Frans was liable to use these confused forms between the ages of two and two and a-half, and I had to quicken his acquaintance with the right usage by refusing to understand him when he used the wrong. Beth M. (2.6) was very jealous about her elder sister touching any of her property, and if the latter sat on her chair, she would shriek out: “That’s your chair; that’s your chair.”

The forms I and me are a common source of difficulty to English children. Both Tony E. (2.7 to 3.0) and Hilary M. (2.0) use my for me; it is apparently a kind of blending of me and I; e.g. “Give Hilary medicine, make my better,” “Maggy is looking at my,” “Give it my.” See also O’Shea, p. 81: ‘my want to do this or that; my feel bad; that is my pencil; take my to bed.’

His and her are difficult to distinguish: “An ill lady, his legs were bad” (Tony E., 3.3).

C. M. L. (about the end of her second year) constantly used wour and wours for our and ours, the connexion being with we, as ‘your’ with you. In exactly the same way many Danish children say vos for os on account of vi. But all this really falls under our next chapter.

The number of words which the child has at command is constantly increasing, but not uniformly, as the increase is affected by the child’s health and the new experiences which life presents to him. In the beginning it is tolerably easy to count the words the child uses; later it becomes more difficult, as there are times when his command of speech grows with astonishing rapidity. There is great difference between individual children. Statistics have often been given of the extent of a child’s vocabulary at different ages, or of the results of comparing the vocabularies of a number of children.

[125]

An American child who was closely observed by his mother, Mrs. Winfield S. Hall, had in the tenth month 3 words, in the eleventh 12, in the twelfth 24, in the thirteenth 38, in the fourteenth 48, in the fifteenth 106, in the sixteenth 199, and in the seventeenth 232 words (Child Study Monthly, March 1897). During the first month after the same boy was six years old, slips of paper and pencils were distributed over the house and practically everything which the child said was written down. After two or three days these were collected and the words were put under their respective letters in a book kept for that purpose. New sets of papers were put in their places and other lists made. In addition to this, the record of his life during the past year was examined and all of his words not already listed were added. In this way his summer vocabulary was obtained; conversations on certain topics were also introduced to give him an opportunity to use words relating to such topics. The list is printed in the Journal of Childhood and Adolescence, January 1902, and is well worth looking through. It contains 2,688 words, apart from proper names and numerals. No doubt the child was really in command of words beyond that total.

This list perhaps is exceptional on account of the care with which it was compiled, but as a rule I am afraid that it is not wise to attach much importance to these tables of statistics. One is generally left in the dark whether the words counted are those that the child has understood, or those that it has actually used—two entirely different things. The passive or receptive knowledge of a language always goes far beyond the active or productive.

One also gets the impression that the observers have often counted up words without realizing the difficulties involved. What is to be counted as a word? Are I, me, we, us one word or four? Is teacup a new word for a child who already knows tea and cup? And so for all compounds. Is box (= a place at a theatre) the same word as box (= workbox)? Are the two thats in ‘that man that you see’ two words or one? It is clear that the process of counting involves so much that is arbitrary and uncertain that very little can be built on the statistics arrived at.

It is more interesting perhaps to determine what words at a given age a child does not know, or rather does not understand when he hears them or when they occur in his reading. I have myself collected such lists, and others have been given me by teachers, who have been astonished at words which their classes did not understand. A teacher can never be too cautious about assuming linguistic knowledge in his pupils—and this applies not only to foreign words, about which all teachers are on the alert,[126] but also to what seem to be quite everyday words of the language of the country.

In connexion with the growth of vocabulary one may ask how many words are possessed by the average grown-up man? Max Müller in his Lectures stated on the authority of an English clergyman that an English farm labourer has only about three hundred words at command. This is the most utter balderdash, but nevertheless it has often been repeated, even by such an authority on psychology as Wundt. A Danish boy can easily learn seven hundred English words in the first year of his study of the language—and are we to believe that a grown Englishman, even of the lowest class, has no greater stock than such a beginner? If you go through the list of 2,000 to 3,000 words used by the American boy of six referred to above, you will easily convince yourself that they would far from suffice for the rudest labourer. A Swedish dialectologist, after a minute investigation, found that the vocabulary of Swedish peasants amounted to at least 26,000 words, and his view has been confirmed by other investigators. This conclusion is not invalidated by the fact that Shakespeare in his works uses only about 20,000 words and Milton in his poems only about 8,000. It is easy to see what a vast number of words of daily life are seldom or never required by a poet, especially a poet like Milton, whose works are on elevated subjects. The words used by Zola or Kipling or Jack London would no doubt far exceed those used by Shakespeare and Milton.[21]

To sum up, then. There are only very few words that are explained to the child, and so long as it is quite small it will not even understand the explanations that might be given. Some it learns because, when the word is used, the object is at the same time pointed at, but most words it can only learn by drawing conclusions about their meaning from the situation in which they arise or from the context in which they are used. These conclusions, however, are very uncertain, or they may be correct for the particular occasion and not hold good on some other, to the child’s mind quite similar, occasion. Grown-up people are in the same position with regard to words they do not know, but which they come across in a book or newspaper, e.g. demise. The meanings of many words are at the same time extraordinarily vague and yet so strictly limited (at least in some respects) that the least deviation is felt as a mistake. Moreover, the child often learns a secondary or figurative meaning of a word before its simple[127] meaning. But gradually a high degree of accuracy is obtained, the fittest meanings surviving—that is (in this connexion) those that agree best with those of the surrounding society. And thus the individual is merged in society, and the social character of language asserts itself through the elimination of everything that is the exclusive property of one person only.

To learn a language it is not enough to know so many words. They must be connected according to the particular laws of the particular language. No one tells the child that the plural of ‘hand’ is hands, of ‘foot’ feet, of ‘man’ men, or that the past of ‘am’ is was, of ‘love’ loved; it is not informed when to say he and when him, or in what order words must stand. How can the little fellow learn all this, which when set forth in a grammar fills many pages and can only be explained by help of many learned words?

Many people will say it comes by ‘instinct,’ as if ‘instinct’ were not one of those fine words which are chiefly used to cover over what is not understood, because it says so precious little and seems to say so precious much. But when other people, using a more everyday expression, say that it all ‘comes quite of itself,’ I must strongly demur: so far is it from ‘coming of itself’ that it demands extraordinary labour on the child’s part. The countless grammatical mistakes made by a child in its early years are a tell-tale proof of the difficulty which this side of language presents to him—especially, of course, on account of the unsystematic character of our flexions and the irregularity of its so-called ‘rules’ of syntax.

At first each word has only one form for the child, but he soon discovers that grown-up people use many forms which resemble one another in different connexions, and he gets a sense of the purport of these forms, so as to be able to imitate them himself or even develop similar forms of his own. These latter forms are what linguists call analogy-formations: by analogy with ‘Jack’s hat’ and ‘father’s hat’ the child invents such as ‘uncle’s hat’ and ‘Charlie’s hat’—and inasmuch as these forms are ‘correct,’ no one can say on hearing them whether the child[129] has really invented them or has first heard them used by others. It is just on account of the fact that the forms developed on the spur of the moment by each individual are in the vast majority of instances perfectly identical with those used already by other people, that the principle of analogy comes to have such paramount importance in the life of language, for we are all thereby driven to apply it unhesitatingly to all those instances in which we have no ready-made form handy: without being conscious of it, each of us thus now and then really creates something never heard before by us or anybody else.

The -s of the possessive is so regular in English that it is not difficult for the child to attach it to all words as soon as the character of the termination has dawned upon him. But at first there is a time with many children in which words are put together without change, so that ‘Mother hat’ stands for ‘Mother’s hat’; cf. also sentences like “Baby want baby milk.”

After the s-form has been learnt, it is occasionally attached to pronouns, as you’s for ‘your,’ or more rarely I’s or me’s for ‘my.’

The -s is now in English added freely to whole groups of words, as in the King of England’s power, where the old construction was the King’s power of England, and in Beaumont and Fletcher’s plays (see on the historical development of this group genitive my ChE iii.). In Danish we have exactly the same construction, and Danish children will very frequently extend it, placing the -s at the end of a whole interrogative sentence, e.g., ‘Hvem er det da’s?’ (as if in English, ‘Who is it then’s,’ instead of ‘Whose is it then?’). Dr. H. Bradley once wrote to me: “One of your samples of children’s Danish is an exact parallel to a bit of child’s English that I noted long ago. My son, when a little boy, used to say ‘Who is that-’s’ (with a pause before the s) for ‘Whom does that belong to?’”

Irregular plurals are often regularized, gooses for ‘geese,’ tooths, knifes, etc. O’Shea mentions one child who inversely formed the plural chieves for chiefs on the analogy of thieves.

Sometimes the child becomes acquainted with the plural form first, and from it forms a singular. I have noticed this several times with Danish children, who had heard the irregular plural køer, ‘cows,’ and then would say en kø instead of en ko (while others from the singular ko form a regular plural koer). French children will say un chevau instead of un cheval.

In the comparison of adjectives analogy-formations are frequent with all children, e.g. the littlest, littler, goodest, baddest,[130] splendider, etc. One child is reported as saying quicklier, another as saying quickerly, instead of the received more quickly. A curious formation is “P’raps it was John, but p’rapser it was Mary.”

O’Shea (p. 108) notices a period of transition when the child may use the analogical form at one moment and the traditional one the next. Thus S. (4.0) will say better perhaps five times where he says gooder once, but in times of excitement he will revert to the latter form.

The child at first tends to treat all verbs on the analogy of love, loved, loved, or kiss, kissed, kissed, thus catched, buyed, frowed for ‘caught, bought, threw or thrown,’ etc., but gradually it learns the irregular forms, though in the beginning with a good deal of hesitation and confusion, as done for ‘did,’ hunged for ‘hung,’ etc. O’Shea gives among other sentences (p. 94): “I drunked my milk.” “Budd swunged on the rings.” “Grandpa boughted me a ring.” “I caughted him.” “Aunt Net camed to-day.” “He gaved it to me”—in all of which the irregular form has been supplemented with the regular ending.

A little Danish incident may be thus rendered in English. The child (4.6): “I have seed a chestnut.” “Where have you seen it?” He: “I seen it in the garden.” This shows the influence of the form last heard.

I once heard a French child say “Il a pleuvy” for ‘plu’ from ‘pleuvoir.’ Other analogical forms are prendu for ‘pris’; assire for ‘asseoir’ (from the participle assis), se taiser for ‘se taire’ (from the frequent injunction taisez-vous). Similar formations are frequent in all countries.

Do the little brains think about these different forms and their uses? Or is the learning of language performed as unconsciously as the circulation of the blood or the process of digestion? Clearly they do not think about grammatical forms in the way pursued in grammar-lessons, with all the forms of the same word arranged side by side of one another, with rules and exceptions. Still there is much to lead us to believe that the thing does not go of itself without some thinking over. The fact that in later years we speak our language without knowing how we do it, the right words and phrases coming to us no one knows how or whence, is no proof that it was always so. We ride a bicycle without giving a thought to the machine, look around us, talk with a friend,[131] etc., and yet there was a time when every movement had to be mastered by slow and painful efforts. There would be nothing strange in supposing that it is the same with the acquisition of language.

Of course, it would be idle to ask children straight out if they think about these things, and what they think. But now and then one notices something which shows that at an early age they think about points of grammar a good deal. When Frans was 2.9, he lay in bed not knowing that anyone was in the next room, and he was heard to say quite plainly: “Små hænder hedder det—lille hånd—små hænder—lille hænder, næ små hænder.” (“They are called small hands—little hand—small hands—little hands, no, small hands”: in Danish lille is not used with a plural noun.) Similar things have been related to me by other parents, one child, for instance, practising plural forms while turning over the leaves of a picture-book, and another one, who was corrected for saying nak instead of nikkede (‘nodded’), immediately retorted “Stikker stak, nikker nak,” thus showing on what analogy he had formed the new preterit. Frequently children, after giving a form which their own ears tell them is wrong, at once correct it: ‘I sticked it in—I stuck it in.’

A German child, not yet two, said: “Papa, hast du mir was mitgebringt—gebrungen—gebracht?” almost at a breath (Gabelentz), and another (2.5) said hausin, but then hesitated and added: “Man kann auch häuser sagen” (Meringer).

In the forming of words the child’s brain is just as active. In many cases, again, it will be impossible to distinguish between what the child has heard and merely copied and what it has itself fashioned to a given pattern. If a child, for example, uses the word ‘kindness,’ it is probable that he has heard it before, but it is not certain, because he might equally well have formed the word himself. If, however, we hear him say ‘kindhood,’ or ‘kindship,’ or ‘wideness,’ ‘broadness,’ ‘stupidness,’ we know for certain that he has made the word up himself, because the resultant differs from the form used in the language he hears around him. A child who does not know the word ‘spade’ may call the tool a digger; he may speak of a lamp as a shine. He may say it suns when the sun is shining (cf. it rains), or ask his mother to sauce his pudding. It is quite natural that the enormous number of nouns and verbs of exactly the same form in English (blossom, care, drink, end, fight, fish, ape, hand, dress, etc.) should induce children to make new verbs according to the same pattern;[132] I quote a few of the examples given by O’Shea: “I am going to basket these apples.” “I pailed him out” (took a turtle out of a washtub with a pail). “I needled him” (put a needle through a fly).

Other words are formed by means of derivative endings, as sorrified, lessoner (O’Shea 32), flyable (able to fly, Glenconner 3); “This tooth ought to come out, because it is crookening the others” (a ten-year-old, told me by Professor Ayres). Compound nouns, too, may be freely formed, such as wind-ship, eye-curtain (O’Shea), a fun-copy of Romeo and Juliet (travesty, Glenconner 19). Bryan L. (ab. 5) said springklers for chrysalises (‘because they wake up in the spring’).

Sometimes a child will make up a new word through ‘blending’ two, as when Hilary M. (1.8 to 2) spoke of rubbish = the rubber to polish the boots, or of the backet, from bat and racquet. Beth M. (2.0) used breakolate, from breakfast and chocolate, and Chally as a child’s name, a compound of two sisters, Charity and Sally.

We are so accustomed to see sentences in writing or print with a little space left after each word, that we have got altogether wrong conceptions of language as it is spoken. Here words follow one another without the least pause till the speaker hesitates for a word or has come to the end of what he has to say. ‘Not at all’ sounds like ‘not a tall.’ It therefore requires in many cases a great deal of comparison and analysis on the part of the child to find out what is one and what two or three words. We have seen before that the question ‘How big is the boy?’ is to the child a single expression, beyond his powers of analysis, and to a much later age it is the same with other phrases. The child, then, may make false divisions, and either treat a group of words as one word or one word as a group of words. A girl (2.6) used the term ‘Tanobijeu’ whenever she wished her younger brother to get out of her way. Her parents finally discovered that she had caught up and shortened a phrase that some older children had used—‘’Tend to your own business’ (O’Shea).

A child, addressing her cousin as ‘Aunt Katie,’ was told “I am not Aunt Katie, I am merely Katie.” Next day she said: “Good-morning, Aunt merely-Katie” (translated). A child who had been praised with the words, ‘You are a good boy,’ said to his mother, “You’re a good boy, mother” (2.8).

Cecil H. (4.0) came back from a party and said that she had been given something very nice to eat. “What was it?”[133] “Rats.” “No, no.” “Well, it was mice then.” She had been asked if she would have ‘some-ice,’ and had taken it to be ‘some mice.’ S. L. (2.6) constantly used ‘ababana’ for ‘banana’; the form seems to have come from the question “Will you have a banana?” but was used in such a sentence as “May I have an ababana?” Children will often say napple for apple through a misdivision of an-apple, and normous for enormous; cf. Ch. X § 2.

A few examples may be added from children’s speech in other countries. Ronjat’s child said nésey for ‘échelle,’ starting from u'ne‿échelle; Grammont’s child said un tarbre, starting from cet arbre, and ce nos for ‘cet os,’ from un os; a German child said motel for ‘hotel,’ starting from the combination ‘im‿(h)‿otel’ (Stern). Many German children say arrhöe, because they take the first syllable of ‘diarrhöe’ as the feminine article. A Dutch child heard the phrase ‘’k weet ’t niet’ (‘I don’t know’), and said “Papa, hij kweet ’t niet” (Van Ginneken). A Danish child heard his father say, “Jeg skal op i ministeriet” (“I’m going to the Government office”), and took the first syllable as min (my); consequently he asked, “Skal du i dinisteriet?” A French child was told that they expected Munkácsy (the celebrated painter, in French pronounced as Mon-), and asked his aunt: “Est-ce que ton Kácsy ne viendra pas?” Antoinette K. (7.), in reply to “C’est bien, je te félicite,” said, “Eh bien, moi je ne te fais pas licite.”

The German ‘Ich habe antgewortet’ is obviously on the analogy of angenommen, etc. (Meringer). Danish children not unfrequently take the verb telefonere as two words, and in the interrogative form will place the personal pronoun in the middle of it, ‘Tele hun fonerer?’ (‘Does she telephone?’) A girl asked to see ele mer fant (as if in English she had said ‘ele more phant’). Cf. ‘Give me more handier-cap’ for ‘Give me a greater handicap’—in a foot-race (O’Shea 108).

In the first period the child knows nothing of grammar: it does not connect words together, far less form sentences, but each word stands by itself. ‘Up’ means what we should express by a whole sentence, ‘I want to get up,’ or ‘Lift me up’; ‘Hat’ means ‘Put on my hat,’ or ‘I want to put my hat on,’ or ‘I have my hat on,’ or ‘Mamma has a new hat on’; ‘Father’ can be either ‘Here comes Father,’ or ‘This is Father,’ or ‘He is called Father,’ or ‘I want Father to come to me,’ or ‘I want this or that from Father.’ This particular group of sounds is vaguely associated with the mental picture of the person in question,[134] and is uttered at the sight of him or at the mere wish to see him or something else in connexion with him.

When we say that such a word means what we should express by a whole sentence, this does not amount to saying that the child’s ‘Up’ is a sentence, or a sentence-word, as many of those who have written about these questions have said. We might just as well assert that clapping our hands is a sentence, because it expresses the same idea (or the same frame of mind) that is otherwise expressed by the whole sentence ‘This is splendid.’ The word ‘sentence’ presupposes a certain grammatical structure, which is wanting in the child’s utterance.

Many investigators have asserted that the child’s first utterances are not means of imparting information, but always an expression of the child’s wishes and requirements. This is certainly somewhat of an exaggeration, since the child quite clearly can make known its joy at seeing a hat or a plaything, or at merely being able to recognize it and remember the word for it; but the statement still contains a great deal of truth, for without strong feelings a child would not say much, and it is a great stimulus to talk that he very soon discovers that he gets his wishes fulfilled more easily when he makes them known by means of certain sounds.

Frans (1.7) was accustomed to express his longings in general by help of a long m with rising tone, while at the same time stretching out his hand towards the particular thing that he longed for. This he did, for example, at dinner, when he wanted water. One day his mother said, “Now see if you can say vand (water),” and at once he said what was an approach to the word, and was delighted at getting something to drink by that means. A moment later he repeated what he had said, and was inexpressibly delighted to have found the password which at once brought him something to drink. This was repeated several times. Next day, when his father was pouring out water for himself, the boy again said ‘van,’ ‘van,’ and was duly rewarded. He had not heard the word during the intervening twenty-four hours, and nothing had been done to remind him of it. After some repetitions (for he only got a few drops at a time) he pronounced the word for the first time quite correctly. The day after, the same thing occurred; the word was never heard but at dinner. When he became rather a nuisance with his constant cries for water, his mother said: “Say please”—and immediately came his “Bebe vand” (“Water, please”)—his first attempt to put two words together.

Later—in this formless period—the child puts more and more words together, often in quite haphazard order: ‘My go snow’[135] (‘I want to go out into the snow’), etc. A Danish child of 2.1 said the Danish words (imperfectly pronounced, of course) corresponding to “Oh papa lamp mother boom,” when his mother had struck his father’s lamp with a bang. Another child said “Papa hen corn cap” when he saw his father give corn to the hens out of his cap.

When Frans was 1.10, passing a post-office (which Danes call ‘posthouse’), he said of his own accord the Danish words for ‘post, house, bring, letter’(a pause between the successive words)—I suppose that the day before he had heard a sentence in which these words occurred. In the same month, when he had thrown a ball a long way, he said what would be in English ‘dat was good.’ This was not a sentence which he had put together for himself, but a mere repetition of what had been said to him, clearly conceived as a whole, and equivalent to ‘bravo.’ Sentences of this kind, however, though taken as units, prepare the way for the understanding of the words ‘that’ and ‘was’ when they turn up in other connexions.

One thing which plays a great rôle in children’s acquisition of language, and especially in their early attempts to form sentences, is Echoism: the fact that children echo what is said to them. When one is learning a foreign language, it is an excellent method to try to imitate to oneself in silence every sentence which one hears spoken by a native. By that means the turns of phrases, the order of words, the intonation of the sentence are firmly fixed in the memory—so that they can be recalled when required, or rather recur to one quite spontaneously without an effort. What the grown man does of conscious purpose our children to a large extent do without a thought—that is, they repeat aloud what they have just heard, either the whole, if it is a very short sentence, or more commonly the conclusion, as much of it as they can retain in their short memories. The result is a matter of chance—it need not always have a meaning or consist of entire words. Much, clearly, is repeated without being understood, much, again, without being more than half understood. Take, for example (translated):

Shall I carry you?—Frans (1.9): Carry you.

Shall Mother carry Frans?—Carry Frans.

The sky is so blue.—So boo.

I shall take an umbrella.—Take rella.

Though this feature in a child’s mental history has been often noticed, no one seems to have seen its full significance. One of the acutest observers (Meumann, p. 28) even says that it has no importance in the development of the child’s speech. On the contrary, I think that Echoism explains very much indeed. First let us bear in mind the mutilated forms of words which a child[136] uses: ’chine for machine, ’gar for cigar, Trix for Beatrix, etc. Then a child’s frequent use of an indirect form of question rather than direct, ‘Why you smoke, Father?’ which can hardly be explained except as an echo of sentences like ‘Tell me why you smoke.’ This plays a greater rôle in Danish than in English, and the corresponding form of the sentence has been frequently remarked by Danish parents. Another feature which is nearly constant with Danish children at the age when echoing is habitual is the inverted word order: this is used after an initial adverb (nu kommer hun, etc.), but the child will use it in all cases (kommer hun, etc.). Further, the extremely frequent use of the infinitive, because the child hears it towards the end of a sentence, where it is dependent on a preceding can, or may, or must. ‘Not eat that’ is a child’s echo of ‘You mustn’t eat that.’ In German this has become the ordinary form of official order: “Nicht hinauslehnen” (“Do not lean out of the window”).

Most children learn to say ‘no’ before they can say ‘yes’—simply because negation is a stronger expression of feeling than affirmation. Many little children use nenenene (short ĕ) as a natural expression of fretfulness and discomfort. It is perhaps so natural that it need not be learnt: there is good reason for the fact that in so many languages words of negation begin with n (or m). Sometimes the n is heard without a vowel: it is only the gesture of ‘turning up one’s nose’ made audible.

At first the child does not express what it is that it does not want—it merely puts it away with its hand, pushes away, for example, what is too hot for it. But when it begins to express in words what it is that it will not have, it does so often in the form ‘Bread no,’ often with a pause between the words, as two separate utterances, as when we might say, in our fuller forms of expression: ‘Do you offer me bread? I won’t hear of it.’ So with verbs: ‘I sleep no.’ Thus with many Danish children, and I find the same phenomenon mentioned with regard to children of different nations. Tracy says (p. 136): “Negation was expressed by an affirmative sentence, with an emphatic no tacked on at the end, exactly as the deaf-mutes do.” The blind-deaf Helen Keller, when she felt her little sister’s mouth and her mother spelt ‘teeth’ to her, answered: “Baby teeth—no, baby eat—no,” i.e., baby cannot eat because she has no teeth. In the same way, in German, ‘Stul nei nei—schossel,’ i.e., I won’t sit on the chair, but in your lap, and in French, ‘Papa abeié ato non, iaian abeié non,’ i.e., Papa n’est pas encore habillé, Suzanne n’est pas[137] habillée (Stern, 189, 203). It seems thus that this mode of expression will crop up everywhere as an emphatic negation.

Interrogative sentences come generally rather early—it would be better to say questions, because at first they do not take the form of interrogative sentences, the interrogation being expressed by bearing, look or gesture: when it begins to be expressed by intonation we are on the way to question expressed in speech. Some of the earliest questions have to do with place: ‘Where is...?’ The child very often hears such sentences as ‘Where is its little nose?’ which are not really meant as questions; we may also remark that questions of this type are of great practical importance for the little thing, who soon uses them to beg for something which has been taken away from him or is out of his reach. Other early questions are ‘What’s that?’ and ‘Who?’

Later—generally, it would seem, at the close of the third year—questions with ‘why’ crop up: these are of the utmost importance for the child’s understanding of the whole world and its manifold occurrences, and, however tiresome they may be when they come in long strings, no one who wishes well to his child will venture to discourage them. Questions about time, such as ‘When? How long?’ appear much later, owing to the child’s difficulty in acquiring exact ideas about time.

Children often find a difficulty in double questions, and when asked ‘Will you have brown bread or white?’ merely answer the last word with ‘Yes.’ So in reply to ‘Is that red or yellow?’ ‘Yes’ means ‘yellow’ (taken from a child of 4.11). I think this is an instance of the short memories of children, who have already at the end of the question forgotten the beginning, but Professor Mawer thinks that the real difficulty here is in making a choice: they cannot decide between alternatives: usually they are silent, and if they say ‘Yes’ it only means that they do not want to go without both or feel that they must say something.

Prepositions are of very late growth in a child’s language. Much attention has been given to the point, and Stern has collected statistics of the ages at which various children have first used prepositions: the earliest age is 1.10, the average age is 2.3. It does not, however, seem to me to be a matter of much interest how early an individual word of some particular grammatical class is first used; it is much more interesting to follow up the gradual growth of the child’s command of this class and to see how the first inevitable mistakes and confusions arise in the little brain. Stern makes the interesting remark that when the[138] tendency to use prepositions first appears, it grows far more rapidly than the power to discriminate one preposition from another; with his own children there came a time when they employed the same word as a sort of universal preposition in all relations. Hilda used von, Eva auf. I have never observed anything corresponding to this among Danish children.

All children start by putting the words for the most important concepts together without connective words, so ‘Leave go bedroom’ (‘May I have leave to go into the bedroom?’), ‘Out road’ (‘I am going out on the road’). The first use of prepositions is always in set phrases learnt as wholes, like ‘go to school,’ ‘go to pieces,’ ‘lie in bed,’ ‘at dinner.’ Not till later comes the power of using prepositions in free combinations, and it is then that mistakes appear. Nor is this surprising, since in all languages prepositional usage contains much that is peculiar and arbitrary, chiefly because when we once pass beyond a few quite clear applications of time and place, the relations to be expressed become so vague and indefinite, that logically one preposition might often seem just as right as another, although usage has laid down a fast law that this preposition must be used in this case and that in another. I noted down a great number of mistakes my own boy made in these words, but in all cases I was able to find some synonymous or antonymous expression in which the preposition used would have been the correct one, and which may have been vaguely before his mind.

The multiple meanings of prepositions sometimes have strange results. A little girl was in her bath, and hearing her mother say: “I will wash you in a moment,” answered: “No, you must wash me in the bath”! She was led astray by the two uses of in. We know of the child at school who was asked “What is an average?” and said: “What the hen lays eggs on.” Even men of science are similarly led astray by prepositions. It is perfectly natural to say that something has passed over the threshold of consciousness: the metaphor is from the way in which you enter a house by stepping over the threshold. If the metaphor were kept, the opposite situation would be expressed by the statement that such and such a thing is outside the threshold of consciousness. But psychologists, in the thoughtless way of little children, take under to be always the opposite of over, and so speak of things ‘lying under (or below) the threshold of our consciousness,’ and have even invented a Latin word for the unconscious, viz. subliminal.[22]

[139]

Children may use verbs with an object which require a preposition (‘Will you wait me?’), or which are only used intransitively (‘Will you jump me?’), or they may mix up an infinitival with a direct construction (‘Could you hear me sneezed?’). But it is surely needless to multiply examples.

When many years ago, in my Progress in Language, I spoke of the advantages, even to natives, of simplicity in linguistic structure, Professor Herman Möller, in a learned review, objected to me that to the adult learning a foreign tongue the chief difficulty consists in “the countless chicaneries due to the tyrannical and capricious usage, whose tricks there is no calculating; but these offer to the native child no such difficulty as morphology may,” and again, in speaking of the choice of various prepositions, which is far from easy to the foreigner, he says: “But any considerable mental exertion on the part of the native child learning its mother-tongue is here, of course, out of the question.” Such assertions as these cannot be founded on actual observation; at any rate, it is my experience in listening to children’s talk that long after they have reached the point where they make hardly any mistake in pronunciation and verbal forms, etc., they are still capable of using many turns of speech which are utterly opposed to the spirit of the language, and which are in the main of the same kind as those which foreigners are apt to fall into. Many of the child’s mistakes are due to mixtures or blendings of two turns of expression, and not a few of them may be logically justified. But learning a language implies among other things learning what you may not say in the language, even though no reasonable ground can be given for the prohibition.

How does it happen that children in general learn their mother-tongue so well? That this is a problem becomes clear when we contrast a child’s first acquisition of its mother-tongue with the later acquisition of any foreign tongue. The contrast is indeed striking and manifold: here we have a quite little child, without experience or prepossessions; there a bigger child, or it may be a grown-up person with all sorts of knowledge and powers: here a haphazard method of procedure; there the whole task laid out in a system (for even in the schoolbooks that do not follow the old grammatical system there is a certain definite order of progress from more elementary to more difficult matters): here no professional teachers, but chance parents, brothers and sisters, nursery-maids and playmates; there teachers trained for many years specially to teach languages: here only oral instruction; there not only that, but reading-books, dictionaries and other assistance. And yet this is the result: here complete and exact command of the language as a native speaks it, however stupid the children; there, in most cases, even with people otherwise highly gifted, a defective and inexact command of the language. On what does this difference depend?

The problem has never been elucidated or canvassed from all sides, but here and there one finds a partial answer, often given out to be a complete answer. Often one side of the question only is considered, that which relates to sounds, as if the whole problem had been solved when one had found a reason for children acquiring a better pronunciation of their mother-tongue than one generally gets in later life of a foreign speech.

Many people accordingly tell us that children’s organs of speech are especially flexible, but that this suppleness of the tongue and lips is lost in later life. This explanation, however, does not hold[141] water, as is shown sufficiently by the countless mistakes in sound made by children. If their organs were as flexible as is pretended, they could learn sounds correctly at once, while as a matter of fact it takes a long time before all the sounds and groups of sounds are imitated with tolerable accuracy. Suppleness is not something which is original, but something acquired later, and acquired with no small difficulty, and then only with regard to the sounds of one’s own language, and not universally.

The same applies to the second answer (given by Bremer, Deutsche Phonetik, 2), namely, that the child’s ear is especially sensitive to impressions. The ear also requires development, since at first it can scarcely detect a number of nuances which we grown-up people hear most distinctly.

Some people say that the reason why a child learns its native language so well is that it has no established habits to contend against. But that is not right either: as any good observer can see, the process by which the child acquires sounds is pursued through a continuous struggle against bad habits which it has acquired at an earlier stage and which may often have rooted themselves remarkably firmly.

Sweet (H 19) says among other things that the conditions of learning vernacular sounds are so favourable because the child has nothing else to do at the time. On the contrary, one may say that the child has an enormous deal to do while it is learning the language; it is at that time active beyond all belief: in a short time it subdues wider tracts than it ever does later in a much longer time. The more wonderful is it that along with those tasks it finds strength to learn its mother-tongue and its many refinements and crooked turns.

Some point to heredity and say that a child learns that language most easily which it is disposed beforehand to learn by its ancestry, or in other words that there are inherited convolutions of the brain which take in this language better than any other. Perhaps there is something in this, but we have no definite, carefully ascertained facts. Against the theory stands the fact that the children of immigrants acquire the language of their foster-country to all appearance just as surely and quickly as children of the same age whose forefathers have been in the country for ages. This may be observed in England, in Denmark, and still more in North America. Environment clearly has greater influence than descent.

The real answer in my opinion (which is not claimed to be absolutely new in every respect) lies partly in the child itself, partly in the behaviour towards it of the people around it. In the first place, the time of learning the mother-tongue is the most favourable of all, namely, the first years of life. If one assumes[142] that mental endowment means the capacity for development, without doubt all children are best endowed in their first years: from birth onwards there is a steady decline in the power of grasping what is new and of accommodating oneself to it. With some this decline is a very rapid one—they quickly become fossilized and unable to make a change in their habits; with others one can notice a happy power of development even in old age; but no one keeps very long in its full range the adaptability of his first years.

Further, we must remember that the child has far more abundant opportunities of hearing his mother-tongue than one gets, as a rule, with any language one learns later. He hears it from morning to night, and, be it noted, in its genuine shape, with the right pronunciation, right intonation, right use of words and right syntax: the language comes to him as a fresh, ever-bubbling spring. Even before he begins to say anything himself, his first understanding of the language is made easier by the habit that mothers and nurses have of repeating the same phrases with slight alterations, and at the same time doing the thing which they are talking about. “Now we must wash the little face, now we must wash the little forehead, now we must wash the little nose, now we must wash the little chin, now we must wash the little ear,” etc. If men had to attend to their children, they would never use so many words—but in that case the child would scarcely learn to understand and talk as soon as it does when it is cared for by women.[23]

Then the child has, as it were, private lessons in its mother-tongue all the year round. There is nothing of the kind in the learning of a language later, when at most one has six hours a week and generally shares them with others. The child has another priceless advantage: he hears the language in all possible situations and under such conditions that language and situation ever correspond exactly to one another and mutually illustrate one another. Gesture and facial expression harmonize with the words[143] uttered and keep the child to a right understanding. Here there is nothing unnatural, such as is often the case in a language-lesson in later years, when one talks about ice and snow in June or excessive heat in January. And what the child hears is just what immediately concerns him and interests him, and again and again his own attempts at speech lead to the fulfilment of his dearest wishes, so that his command of language has great practical advantages for him.

Along with what he himself sees the use of, he hears a great deal which does not directly concern him, but goes into the little brain and is stored up there to turn up again later. Nothing is heard but leaves its traces, and at times one is astonished to discover what has been preserved, and with what exactness. One day, when Frans was 4.11 old, he suddenly said: “Yesterday—isn’t there some who say yesterday?” (giving yesterday with the correct English pronunciation), and when I said that it was an English word, he went on: “Yes, it is Mrs. B.: she often says like that, yesterday.” Now, it was three weeks since that lady had called at the house and talked English. It is a well-known fact that hypnotized persons can sometimes say whole sentences in a language which they do not know, but have merely heard in childhood. In books about children’s language there are many remarkable accounts of such linguistic memories which had lain buried for long stretches of time. A child who had spent the first eighteen months of its life in Silesia and then came to Berlin, where it had no opportunity of hearing the Silesian pronunciation, at the age of five suddenly came out with a number of Silesian expressions, which could not after the most careful investigation be traced to any other source than to the time before it could talk (Stern, 257 ff.). Grammont has a story of a little French girl, whose nurse had talked French with a strong Italian accent; the child did not begin to speak till a month after this nurse had left, but pronounced many words with Italian sounds, and some of these peculiarities stuck to the child till the age of three.

We may also remark that the baby’s teachers, though, regarded as teachers of language, they may not be absolutely ideal, still have some advantages over those one encounters as a rule later in life. The relation between them and the child is far more cordial and personal, just because they are not teachers first and foremost. They are immensely interested in every little advance the child makes. The most awkward attempt meets with sympathy, often with admiration, while its defects and imperfections never expose it to a breath of unkind criticism. There is a Slavonic proverb, “If you wish to talk well, you must murder the language first.” But this is very often overlooked by teachers of language, who[144] demand faultless accuracy from the beginning, and often keep their pupils grinding so long at some little part of the subject that their desire to learn the language is weakened or gone for good. There is nothing of this sort in the child’s first learning of his language.

It is here that our distinction between the two periods comes in, that of the child’s own separate ‘little language’ and that of the common or social language. In the first period the little one is the centre of a narrow circle of his own, which waits for each little syllable that falls from his lips as though it were a grain of gold. What teachers of languages in later years would rejoice at hearing such forms as we saw before used in the time of the child’s ‘little language,’ fant or vat or ham for ‘elephant’? But the mother really does rejoice: she laughs and exults when he can use these syllables about his toy-elephant, she throws the cloak of her love over the defects and mistakes in the little one’s imitations of words, she remembers again and again what his strange sounds stand for, and her eager sympathy transforms the first and most difficult steps on the path of language to the merriest game.

It would not do, however, for the child’s ‘little language’ and its dreadful mistakes to become fixed. This might easily happen, if the child were never out of the narrow circle of its own family, which knows and recognizes its ‘little language.’ But this is stopped because it comes more and more into contact with others—uncles and aunts, and especially little cousins and playmates: more and more often it happens that the mutilated words are not understood, and are corrected and made fun of, and the child is incited in this way to steady improvement: the ‘little language’ gradually gives place to the ‘common language,’ as the child becomes a member of a social group larger than that of his own little home.

We have now probably found the chief reasons why a child learns his mother-tongue better than even a grown-up person who has been for a long time in a foreign country learns the language of his environment. But it is also a contributory reason that the child’s linguistic needs, to begin with, are far more limited than those of the man who wishes to be able to talk about anything, or at any rate about something. Much more is also linguistically required of the latter, and he must have recourse to language to get all his needs satisfied, while the baby is well looked after even if it says nothing but wawawawa. So the baby has longer time to store up his impressions and continue his experiments, until by trying again and again he at length gets his lesson learnt in all its tiny details, while the man in the foreign country,[145] who must make himself understood, as a rule goes on trying only till he has acquired a form of speech which he finds natives understand: at this point he will generally stop, at any rate as far as pronunciation and the construction of sentences are concerned (while his vocabulary may be largely increased). But this ‘just recognizable’ language is incorrect in thousands of small details, and, inasmuch as bad little habits quickly become fixed, the kind of language is produced which we know so well in the case of resident foreigners—who need hardly open their lips before everyone knows they are not natives, and before a practised ear can detect the country they hail from.[24]

An important factor in the acquisition of language which we have not considered is naturally the individuality of the child. Parents are apt to draw conclusions as to the abilities of their young hopeful from the rapidity with which he learns to talk; but those who are in despair because their Tommy cannot say a single word when their neighbours’ Harry can say a great deal may take comfort. Slowness in talking may of course mean deficiency of ability, or even idiocy, but not necessarily. A child who chatters early may remain a chatterer all his life, and children whose motto is ‘Slow and sure’ may turn out the deepest, most independent and most trustworthy characters in the end. There are some children who cannot be made to say a single word for a long time, and then suddenly come out with a whole sentence, which shows how much has been quietly fructifying in their brain. Carlyle was one of these: after eleven months of taciturnity he heard a child cry, and astonished all by saying, “What ails wee Jock?” Edmund Gosse has a similar story of his own childhood, and other examples have been recorded elsewhere (Meringer, 194; Stern, 257).

[146]

The linguistic development of an individual child is not always in a steady rising line, but in a series of waves. A child who seems to have a boundless power of acquiring language suddenly stands still or even goes back for a short time. The cause may be sickness, cutting teeth, learning to walk, or often a removal to new surroundings or an open-air life in summer. Under such circumstances even the word ‘I’ may be lost for a time.

Some children develop very rapidly for some years until they have reached a certain point, where they stop altogether, while others retain the power to develop steadily to a much later age. It is the same with some races: negro children in American schools may, while they are little, be up to the standard of their white schoolfellows, whom they cannot cope with in later life.

The two sexes differ very greatly in regard to speech—as in regard to most other things. Little girls, on the average, learn to talk earlier and more quickly than boys; they outstrip them in talking correctly; their pronunciation is not spoilt by the many bad habits and awkwardnesses so often found in boys. It has been proved by statistics in many countries that there are far more stammerers and bad speakers among boys and men than among girls and women. The general receptivity of women, their great power of, and pleasure in, imitation, their histrionic talent, if one may so say—all this is a help to them at an early age, so that they can get into other people’s way of talking with greater agility than boys of the same age.

Everything that is conventional in language, everything in which the only thing of importance is to be in agreement with those around you, is the girls’ strong point. Boys may often show a certain reluctance to do exactly as others do: the peculiarities of their ‘little language’ are retained by them longer than by girls, and they will sometimes steadily refuse to correct their own abnormalities, which is very seldom the case with girls. Gaucherie and originality thus are two points between which the speech of boys is constantly oscillating. Cf. below, Ch. XIII.

The expression “mother-tongue” should not be understood too literally: the language which the child acquires naturally is not, or not always, his mother’s language. When a mother speaks with a foreign accent or in a pronounced dialect, her children as a rule speak their language as correctly as other children, or keep only the slightest tinge of their mother’s peculiarities. I have seen this very distinctly in many Danish families, in which the mother has kept up her Norwegian language all her life, and in[147] which the children have spoken pure Danish. Thus also in two families I know, in which a strong Swedish accent in one mother, and an unmistakable American pronunciation in the other, have not prevented the children from speaking Danish exactly as if their mothers had been born and bred in Denmark. I cannot, therefore, agree with Passy, who says that the child learns his mother’s sound system (Ch § 32), or with Dauzat’s dictum to the same effect (V 20). The father, as a rule, has still less influence; but what is decisive is the speech of those with whom the child comes in closest contact from the age of three or so, thus frequently servants, but even more effectually playfellows of his own age or rather slightly older than himself, with whom he is constantly thrown together for hours at a time and whose prattle is constantly in his ears at the most impressionable age, while he may not see and hear his father and mother except for a short time every day, at meals and on such occasions. It is also a well-known fact that the children of Danish parents in Greenland often learn the Eskimo language before Danish; and Meinhof says that German children in the African colonies will often learn the language of the natives earlier than German (MSA 139).

This is by no means depreciating the mother’s influence, which is strong indeed, but chiefly in the first period, that of the child’s ‘little language.’ But that is the time when the child’s imitative power is weakest. His exact attention to the minutiæ of language dates from the time when he is thrown into a wider circle and has to make himself understood by many, so that his language becomes really identical with that of the community, where formerly he and his mother would rest contented with what they, but hardly anyone else, could understand.

The influence of children on children cannot be overestimated.[25] Boys at school make fun of any peculiarities of speech noticed in schoolfellows who come from some other part of the country. Kipling tells us in Stalky and Co. how Stalky and Beetle carefully kicked McTurk out of his Irish dialect. When I read this, I was vividly reminded of the identical method my new friends applied to me when at the age of ten I was transplanted from Jutland to a school in Seeland and excited their merriment through some Jutlandish expressions and intonations. And so we may say that the most important factor in spreading the common or standard language is children themselves.

It often happens that children who are compelled at home to talk without any admixture of dialect talk pure dialect when playing with their schoolfellows out of doors. They can keep the[148] two forms of speech distinct. In the same way they can learn two languages less closely connected. At times this results in very strange blendings, at least for a time; but many children will easily pass from one language to the other without mixing them up, especially if they come in contact with the two languages in different surroundings or on the lips of different people.

It is, of course, an advantage for a child to be familiar with two languages: but without doubt the advantage may be, and generally is, purchased too dear. First of all the child in question hardly learns either of the two languages as perfectly as he would have done if he had limited himself to one. It may seem, on the surface, as if he talked just like a native, but he does not really command the fine points of the language. Has any bilingual child ever developed into a great artist in speech, a poet or orator?

Secondly, the brain effort required to master two languages instead of one certainly diminishes the child’s power of learning other things which might and ought to be learnt. Schuchardt rightly remarks that if a bilingual man has two strings to his bow, both are rather slack, and that the three souls which the ancient Roman said he possessed, owing to his being able to talk three different languages, were probably very indifferent souls after all. A native of Luxemburg, where it is usual for children to talk both French and German, says that few Luxemburgers talk both languages perfectly. “Germans often say to us: ‘You speak German remarkably well for a Frenchman,’ and French people will say, ‘They are Germans who speak our language excellently.’ Nevertheless, we never speak either language as fluently as the natives. The worst of the system is, that instead of learning things necessary to us we must spend our time and energy in learning to express the same thought in two or three languages at the same time.”[26]

The child takes delight in making meaningless sounds long after it has learnt to talk the language of its elders. At 2.2 Frans amused himself with long series of such sounds, uttered with the most confiding look and proper intonation, and it was a joy to him when I replied with similar sounds. He kept up this game for years. Once (4.11) after such a performance he asked me: “Is that English?”—“No.”—“Why not?”—“Because I understand English, but I do not understand what you say.” An hour later he came back and asked: “Father, do you know all languages?”—“No, there are many I don’t know.”—“Do you[149] know German?”—“Yes.” (Frans looked rather crestfallen: the servants had often said of his invented language that he was talking German. So he went on) “Do you know Japanese?”—“No.”—(Delighted) “So remember when I say something you don’t understand, it’s Japanese.”

It is the same everywhere. Hawthorne writes: “Pearl mumbled something into his ear, that sounded, indeed, like human language, but was only such gibberish as children may be heard amusing themselves with, by the hour together” (The Scarlet Letter, 173). And R. L. Stevenson: “Children prefer the shadow to the substance. When they might be speaking intelligibly together, they chatter senseless gibberish by the hour, and are quite happy because they are making believe to speak French” (Virginibus P., 236; cf. Glenconner, p. 40; Stern, pp. 76, 91, 103). Meringer’s boy (2.1) took the music-book and sang a tune of his own making with incomprehensible words.

Children also take delight in varying the sounds of real words, introducing, for instance, alliterations, as “Sing a song of sixpence, A socket full of sye,” etc. Frans at 2.3 amused himself by rounding all his vowels (o for a, y for i), and at 3.1 by making all words of a verse line he had learnt begin with d, then the same words begin with t. O’Shea (p. 32) says that “most children find pleasure in the production of variations upon some of their familiar words. Their purpose seems to be to test their ability to be original. The performance of an unusual act affords pleasure in linguistics as in other matters. H., learning the word dessert, to illustrate, plays with it for a time and exhibits it in a dozen or more variations—dĭssert, dishert, dĕsot, des'sert, and so on.”

Rhythm and rime appeal strongly to the children’s minds. One English observer says that “a child in its third year will copy the rhythm of songs and verses it has heard in nonsense words.” The same thing is noted by Meringer (p. 116) and Stern (p. 103). Tony E. (2.10) suddenly made up the rime “My mover, I lov-er,” and Gordon M. (2.6) never tired of repeating a phrase of his own composition, “Custard over mustard.” A Danish girl of 3.1 is reported as having a “curious knack of twisting all words into rimes: bestemor hestemor prestemor, Gudrun sludrun pludrun, etc.”

Children, as we have seen, at first employ play-language for its own sake, with no arrière-pensée, but as they get older they may see that such language has the advantage of not being understood by their elders, and so they may develop a ‘secret language’[150] consciously. Some such languages are confined to one school, others may be in common use among children of a certain age all over a country. ‘M-gibberish’ and ‘S-gibberish’ consist in inserting m and s, as in goming mout tomdaym or gosings outs tosdays for ‘going out to-day’; ‘Marrowskying’ or ‘Hospital Greek’ transfers the initial letters of words, as renty of plain for ‘plenty of rain,’ flutterby for ‘butterfly’; ‘Ziph’ or ‘Hypernese’ (at Winchester) substitutes wa for the first of two initial consonants and inserts p or g, making ‘breeches’ into wareechepes and ‘penny’ into pegennepy. From my own boyhood in Denmark I remember two languages of this sort, in which a sentence like ‘du er et lille asen’ became dupu erper etpet lilpillepe apasenpen and durbe erbe erbe lirbelerbe arbeserbe respectively. Closely corresponding languages, with insertion of p and addition of -erbse, are found in Germany; in Holland we find ‘de schoone Mei’ made into depé schoopóonepé Meipéi, besides an -erwi-taal with a variation in which the ending is -erf. In France such a language is called javanais; ‘je vais bien’ is made into je-de-que vais-dai-qai bien-den-qen. In Savoy the cowherds put deg after each syllable and thus make ‘a-te kogneu se vaçhi’ (‘as-tu connu ce vacher?’ in the local dialect) into a-degá te-dege ko-dego gnu-degu sé-degé va-dega chi-degi? Nay, even among the Maoris of New Zealand there is a similar secret language, in which instead of ‘kei te, haere au ki reira’ is said te-kei te-i-te te-haere-te-re te-a te-u te-ki te-re-te-i-te-ra. Human nature is pretty much the same everywhere.[27]

Do children really create new words? This question has been much discussed, but even those who are most skeptical in that respect incline to allow them this power in the case of words which imitate sounds. Nevertheless, it should be remembered that the majority of onomatopœic words heard from children are not their own invention, but are acquired by them in the same way as other words. Hence it is that such words have different forms in different languages. Thus to English cockadoodledoo corresponds French coquerico, German kikeriki and Danish kykeliky, to E. quack-quack, F. cancan, Dan. raprap, etc. These words are an imperfect representation of the birds’ natural cry, but from their likeness to it they are easier for the child to seize than an entirely arbitrary name such as duck.

But, side by side with these, children do invent forms of their own, though the latter generally disappear quickly in favour of the[151] traditional forms. Thus Frans (2.3) had coined the word vakvak, which his mother had heard sometimes without understanding what he meant, when one day he pointed at some crows while repeating the same word; but when his mother told him that these birds were called krager, he took hold of this word with eagerness and repeated it several times, evidently recognizing it as a better name than his own. A little boy of 2.1 called soda-water ft, another boy said ging or gingging for a clock, also for the railway train, while his brother said dann for a bell or clock; a little girl (1.9) said pooh (whispered) for ‘match, cigar, pipe,’ and gagag for ‘hen,’ etc.

When once formed, such words may be transferred to other things, where the sound plays no longer any rôle. This may be illustrated through two extensions of the same word bŏom or bom, used by two children first to express the sound of something falling on the floor; then Ellen K. (1.9) used it for a ‘blow,’ and finally for anything disagreeable, e.g. soap in the eyes, while Kaare G. (1.8), after seeing a plate smashed, used the word for a broken plate and afterwards for anything broken, a hole in a dress, etc., also when a button had come off or when anything else was defective in any way.

Do children themselves create words—apart from onomatopœic words? To me there is no doubt that they do. Frans invented many words at his games that had no connexion, or very little connexion, with existing words. He was playing with a little twig when I suddenly heard him exclaim: “This is called lampetine,” but a little while afterwards he said lanketine, and then again lampetine, and then he said, varying the play, “Now it is kluatine and traniklualalilua” (3.6). A month later I write: “He is never at a loss for a self-invented word; for instance, when he has made a figure with his bricks which resembles nothing whatever, he will say, ‘That shall be lindam.’” When he played at trains in the garden, there were many stations with fanciful names, and at one time he and two cousins had a word kukukounen which they repeated constantly and thought great fun, but whose inner meaning I never succeeded in discovering. An English friend writes about his daughter: “When she was about two and a quarter she would often use some nonsense word in the middle of a perfectly intelligible sentence. When you asked her its meaning she would explain it by another equally unintelligible, and so on through a series as long as you cared to make it.” At 2.10 she pretended she had lost her bricks, and when you showed her that they were just by her, she insisted that they were not ‘bricks’ at all, but mums.

[152]

In all accounts of children’s talk you find words which cannot be referred back to the normal language, but which have cropped up from some unsounded depth of the child’s soul. I give a few from notes sent to me by Danish friends: goi ‘comb,’ putput ‘stocking, or any other piece of garment,’ i-a-a ‘chocolate,’ gön ‘water to drink, milk’ (kept apart from the usual word vand for water, which she used only for water to wash in), hesh ‘newspaper, book.’ Some such words have become famous in psychological literature because they were observed by Darwin and Taine. Among less famous instances from other books I may mention tibu ‘bird’ (Strümpel), adi ‘cake’ (Ament), be’lum-be’lum ‘toy with two men turning about,’ wakaka ‘soldier,’ nda ‘jar,’ pamma ‘pencil,’ bium ‘stocking’ (Meringer).

An American correspondent writes that his boy was fond of pushing a stick over the carpet after the manner of a carpet-sweeper and called the operation jazing. He coined the word borkens as a name for a particular sort of blocks with which he was accustomed to play. He was a nervous child and his imagination created objects of terror that haunted him in the dark, and to these he gave the name of Boons. This name may, however, be derived from baboons. Mr. Harold Palmer tells me that his daughter (whose native language was French) at an early age used ['fu'wɛ] for ‘soap’ and [dɛ'dɛtʃ] for ‘horse, wooden horse, merry-go-round.’

Dr. F. Poulsen, in his book Rejser og rids (Copenhagen, 1920), says about his two-year-old daughter that when she gets hold of her mother’s fur-collar she will pet it and lavish on it all kinds of tender self-invented names, such as apu or a-fo-me-me. The latter word, “which has all the melodious euphony and vague signification of primitive language,” is applied to anything that is rare and funny and worth rejoicing at. On a summer day’s excursion there was one new a-fo-me-me after the other.

In spite of all this, a point on which all the most distinguished investigators of children’s language of late years are agreed is that children never invent words. Wundt goes so far as to say that “the child’s language is the result of the child’s environment, the child being essentially a passive instrument in the matter” (S 1. 196)—one of the most wrong-headed sentences I have ever read in the works of a great scientist. Meumann says: “Preyer and after him almost every careful observer among child-psychologists have strongly held the view that it is impossible to speak of a child inventing a word.” Similarly Meringer, L 220, Stern, 126, 273, 337 ff., Bloomfield, SL 12.

These investigators seem to have been led astray by expressions such as ‘shape out of nothing,’ ‘invent,’ ‘original creation’[153] (Urschöpfung), and to have taken this doctrinaire attitude in partial defiance of the facts they have themselves advanced. Expressions like those adduced occur over and over again in their discussions, and Meumann says openly: “Invention demands a methodical proceeding with intention, a conception of an end to be realized.” Of course, if that is necessary it is clear that we can speak of invention of words in the case of a chemist seeking a word for a new substance, and not in the case of a tiny child. But are there not many inventions in the technical world, which we do not hesitate to call inventions, which have come about more or less by chance? Wasn’t it so probably with gunpowder? According to the story it certainly was so with blotting-paper: the foreman who had forgotten to add size to a portion of writing-paper was dismissed, but the manufacturer who saw that the paper thus spoilt could be turned to account instead of the sand hitherto used made a fortune. So according to Meumann blotting-paper has never been ‘invented.’ If in order to acknowledge a child’s creation of a word we are to postulate that it has been produced out of nothing, what about bicycles, fountain-pens, typewriters—each of which was something existing before, carried just a little further? Are they on that account not inventions? One would think not, when one reads these writers on children’s language, for as soon as the least approximation to a word in the normal language is discovered, the child is denied both ‘invention’ and ‘the speech-forming faculty’! Thus Stern (p. 338) says that his daughter in her second year used some words which might be taken as proof of the power to create words, but for the fact that it was here possible to show how these ‘new’ words had grown out of normal words. Eischei, for instance, was used as a verb meaning ‘go, walk,’ but it originated in the words eins, zwei (one, two) which were said when the child was taught to walk. Other examples are given comparable to those mentioned above (106, 115) as mutilations of the first period. Now, even if all those words given by myself and others as original inventions of children could be proved to be similar perversions of ‘real’ words (which is not likely), I should not hesitate to speak of a word-creating faculty, for eischei, ‘to walk,’ is both in form and still more in meaning far enough from eins, zwei to be reckoned a totally new word.

We can divide words ‘invented’ by children into three classes:

[154]

But the three classes cannot always be kept apart, especially when the child imitates the grown-up person’s sound so badly or seizes the meaning so imperfectly that very little is left of what the grown-up person gives. As a rule, the self-created words will be very short-lived; still, there are exceptions.

O’Shea’s account of one of these words is very instructive. “She had also a few words of her own coining which were attached spontaneously to objects, and these her elders took up, and they became fixed in her vocabulary for a considerable period. A word resembling Ndobbin was employed for every sort of thing which she used for food. The word came originally from an accidental combination of sounds made while she was eating. By the aid of the people about her in responding to this term and repeating it, she ‘selected’ it and for a time used it purposefully. She employed it at the outset for a specific article of food; then her elders extended it to other articles, and this aided her in making the extension herself. Once started in this process, she extended the term to many objects associated with her food, even objects as remote from her original experience as dining-room, high-chair, kitchen, and even apple and plum trees” (O’Shea, 27).

To Class A I assign most of the words already given as the child’s creations, whether the child be great or small.

Class B is that which is most sparsely represented. A child in Finland often heard the well-known line about King Karl (Charles XII), “Han stod i rök och damm” (“He stood in smoke and dust”), and taking rö to be the adjective meaning ‘red,’ imagined the remaining syllables, which he heard as kordamm, to be the name of some piece of garment. This amused his parents so much that kordamm became the name of a dressing-gown in that family.

To Class C, where the child contributes only the sound and the older people give a meaning to what on the child’s side was meaningless—a process that reminds one of the invention of blotting-paper—belong some of the best-known words, which require a separate section.

In the nurseries of all countries a little comedy has in all ages been played—the baby lies and babbles his ‘mamama’ or ‘amama’ or ‘papapa’ or ‘apapa’ or ‘bababa’ or ‘ababab’ without associating the slightest meaning with his mouth-games, and his grown-up friends, in their joy over the precocious child, assign to these syllables a rational sense, accustomed as they are themselves to the fact of an uttered sound having a content, a thought, an idea, corresponding to it. So we get a whole class[155] of words, distinguished by a simplicity of sound-formation—never two consonants together, generally the same consonant repeated with an a between, frequently also with an a at the end—words found in many languages, often in different forms, but with essentially the same meaning.

First we have words for ‘mother.’ It is very natural that the mother who is greeted by her happy child with the sound ‘mama’ should take it as though the child were calling her ‘mama,’ and since she frequently comes to the cradle when she hears the sound, the child himself does learn to use these syllables when he wants to call her. In this way they become a recognized word for the idea ‘mother’—now with the stress on the first syllable, now on the second. In French we get a nasal vowel either in the last syllable only or in both syllables. At times we have only one syllable, ma. When once these syllables have become a regular word they follow the speech laws which govern other words; thus among other forms we get the German muhme, the meaning of which (‘aunt’) is explained as in the words mentioned, p. 118. In very early times ma in our group of languages was supplied with a termination, so that we get the form underlying Greek mētēr, Lat. mater (whence Fr. mère, etc.), our own mother, G. mutter, etc. These words became the recognized grown-up words, while mama itself was only used in the intimacy of the family. It depends on fashion, however, how ‘high up’ mama can be used: in some countries and in some periods children are allowed to use it longer than in others.

The forms mama and ma are not the only ones for ‘mother.’ The child’s am has also been seized and maintained by the grown-ups. The Albanian word for ‘mother’ is ama, the Old Norse word for ‘grandmother’ is amma. The Latin am-ita, formed from am with a termination added, came to mean ‘aunt’ and became in OFr. ante, whence E. aunt and Modern Fr. tante. In Semitic languages the words for ‘mother’ also have a vowel before m: Assyrian ummu, Hebrew ’êm, etc.

Baba, too, is found in the sense ‘mother,’ especially in Slavonic languages, though it has here developed various derivative meanings, ‘old woman,’ ‘grandmother,’ or ‘midwife.’ In Tonga we have bama ‘mother.’

Forms with n are also found for ‘mother’; so Sanskrit naná, Albanian nane. Here we have also Gr. nannē ‘aunt’ and Lat. nonna; the latter ceased in the early Middle Ages to mean ‘grandmother’ and became a respectful way of addressing women of a certain age, whence we know it as nun, the feminine counterpart of ‘monk.’ From less known languages I may mention Greenlandic a'na·na ‘mother,’ 'a·na ‘grandmother.’

[156]

Now we come to words meaning ‘father,’ and quite naturally, where the sound-groups containing m have already been interpreted in the sense ‘mother,’ a word for ‘father’ will be sought in the syllables with p. It is no doubt frequently noticed in the nursery that the baby says mama where one expected papa, and vice versa; but at last he learns to deal out the syllables ‘rightly,’ as we say. The history of the forms papa, pappa and pa is analogous to the history of the m syllables already traced. We have the same extension of the sound by tr in the word pater, which according to recognized laws of sound-change is found in the French père, the English father, the Danish fader, the German vater, etc. Philologists no longer, fortunately, derive these words from a root pa ‘to protect,’ and see therein a proof of the ‘highly moral spirit’ of our aboriginal ancestors, as Fick and others did. Papa, as we know, also became an honourable title for a reverend ecclesiastic, and hence comes the name which we have in the form Pope.

Side by side with the p forms we have forms in b—Italian babbo, Bulgarian babá, Serbian bába, Turkish baba. Beginning with the vowel we have the Semitic forms ab, abu and finally abba, which is well known, since through Greek abbas it has become the name for a spiritual father in all European languages, our form being Abbot.

Again, we have some names for ‘father’ with dental sounds: Sanskrit tatá, Russian tata, tyatya, Welsh tat, etc. The English dad, now so universal, is sometimes considered to have been borrowed from this Welsh word, which in certain connexions has an initial d, but no doubt it had an independent origin. In Slavonic languages déd is extensively used for ‘grandfather’ or ‘old man.’ Thus also deite, teite in German dialects. Tata ‘father’ is found in Congo and other African languages, also (tatta) in Negro-English (Surinam). And just as words for ‘mother’ change their meaning from ‘mother’ to ‘aunt,’ so these forms in some languages come to mean ‘uncle’: Gr. theios (whence Italian zio), Lithuanian dede, Russian dyadya.

With an initial vowel we get the form atta, in Greek used in addressing old people, in Gothic the ordinary word for ‘father,’ which with a termination added gives the proper name Attila, originally ‘little father’; with another ending we have Russian otec. Outside our own family of languages we find, for instance, Magyar atya, Turkish ata, Basque aita, Greenlandic a'ta·ta ‘father,’ while in the last-mentioned language a·ta means ‘grandfather.’[28]

[157]

The nurse, too, comes in for her share in these names, as she too is greeted by the child’s babbling and is tempted to take it as the child’s name for her; thus we get the German and Scandinavian amme, Polish niania, Russian nyanya, cf. our Nanny. These words cannot be kept distinct from names for ‘aunt,’ cf. amita above, and in Sanskrit we find mama for ‘uncle.’

It is perhaps more doubtful if we can find a name for the child itself which has arisen in the same way; the nearest example is the Engl. babe, baby, German bube (with u as in muhme above); but babe has also been explained as a word derived normally from OFr. baube, from Lat. balbus ‘stammering.’ When the name Bab or Babs (Babbe in a Danish family) becomes the pet-name for a little girl, this has no doubt come from an interpretation put on her own meaningless sounds. Ital. bambo (bambino) certainly belongs here. We may here mention also some terms for ‘doll,’ Lat. pupa or puppa, G. puppe; with a derivative ending we have Fr. poupée, E. puppet (Chaucer, A 3254, popelote). These words have a rich semantic development, cf. pupa (Dan. puppe, etc.) ‘chrysalis,’ and the diminutive Lat. pupillus, pupilla, which was used for ‘a little child, minor,’ whence E. pupil ‘disciple,’ but also for the little child seen in the eye, whence E. (and other languages) pupil, ‘central opening of the eye.’

A child has another main interest—that is, in its food, the breast, the bottle, etc. In many countries it has been observed that very early a child uses a long m (without a vowel) as a sign that it wants something, but we can hardly be right in supposing that the sound is originally meant by children in this sense. They do not use it consciously till they see that grown-up people on hearing the sound come up and find out what the child wants. And it is the same with the developed forms which are uttered by the child in its joy at getting something to eat, and which are therefore interpreted as the child’s expression for food: am, mam, mammam, or the same words with a final a—that is, really the same groups of sounds which came to stand for ‘mother.’ The determination of a particular form to a particular meaning is always due to the adults, who, however, can subsequently teach it to the child. Under this heading comes the sound ham, which Taine observed to be one child’s expression for hunger or thirst (h mute?), and similarly the word mum, meaning ‘something to eat,’ invented,[158] as we are told, by Darwin’s son and often uttered with a rising intonation, as in a question, ‘Will you give me something to eat?’ Lindner’s child (1.5) is said to have used papp for everything eatable and mem or möm for anything drinkable. In normal language we have forms like Sanskrit māmsa (Gothic mimz) and mās ‘flesh,’ our own meat (which formerly, like Dan. mad, meant any kind of food), German mus ‘jam’ (whence also gemüse), and finally Lat. mandere and manducare, ‘to chew’ (whence Fr. manger)—all developments of this childish ma(m).

As the child’s first nourishment is its mother’s breast, its joyous mamama can also be taken to mean the breast. So we have the Latin mamma (with a diminutive ending mammilla, whence Fr. mamelle), and with the other labial sound Engl. pap, Norwegian and Swed. dial. pappe, Lat. papilla; with a different vowel, It. poppa, Fr. poupe, ‘teat of an animal, formerly also of a woman’; with b, G. bübbi, obsolete E. bubby; with a dental, E. teat (G. zitze), Ital. tetta, Dan. titte, Swed. dial. tatte. Further we have words like E. pap ‘soft food,’ Latin papare ‘to eat,’ orig. ‘to suck,’ and some G. forms for the same, pappen, pampen, pampfen. Perhaps the beginning of the word milk goes back to the baby’s ma applied to the mother’s breast or milk; the latter half may then be connected with Lat. lac. In Greenlandic we have ama·ma ‘suckle.’

Inseparable from these words is the sound, a long m or am, which expresses the child’s delight over something that tastes good; it has by-forms in the Scotch nyam or nyamnyam, the English seaman’s term yam ‘to eat,’ and with two dentals the French nanan ‘sweetmeats.’ Some linguists will have it that the Latin amo ‘I love’ is derived from this am, which expresses pleasurable satisfaction. When a father tells me that his son (1.10) uses the wonderful words nananæi for ‘chocolate’ and jajajaja for picture-book, we have no doubt here also a case of a grown person’s interpretation of the originally meaningless sounds of a child.

Another meaning that grown-up people may attach to syllables uttered by the child is that of ‘good-bye,’ as in English tata, which has now been incorporated in the ordinary language.[29] Stern probably is right when he thinks that the French adieu would not have been accepted so commonly in Germany and other countries if it had not accommodated itself so easily, especially in the form commonly used in German, ade, to the child’s natural word.

[159]

There are some words for ‘bed, sleep’ which clearly belong to this class: Tuscan nanna ‘cradle,’ Sp. hacer la nana ‘go to sleep,’ E. bye-bye (possibly associated with good-bye, instead of which is also said byebye); Stern mentions baba (Berlin), beibei (Russian), bobo (Malay), but bischbisch, which he also gives here, is evidently (like the Danish visse) imitative of the sound used for hushing.

Words of this class stand in a way outside the common words of a language, owing to their origin and their being continually new-created. One cannot therefore deduce laws of sound-change from them in their original shape; and it is equally wrong to use them as evidence for an original kinship between different families of language and to count them as loan-words, as is frequently done (for example, when the Slavonic baba is said to be borrowed from Turkish). The English papa and mam(m)a, and the same words in German and Danish, Italian, etc., are almost always regarded as borrowed from French; but Cauer rightly points out that Nausikaa (Odyssey 6. 57) addresses her father as pappa fil, and Homer cannot be suspected of borrowing from French. Still, it is true that fashion may play a part in deciding how long children may be permitted to say papa and mamma, and a French fashion may in this respect have spread to other European countries, especially in the seventeenth century. We may not find these words in early use in the literatures of the different countries, but this is no proof that the words were not used in the nursery. As soon as a word of this class has somewhere got a special application, this can very well pass as a loan-word from land to land—as we saw in the case of the words abbot and pope. And it may be granted with respect to the primary use of the words that there are certain national or quasi-national customs which determine what grown people expect to hear from babies, so that one nation expects and recognizes papa, another dad, a third atta, for the meaning ‘father.’

When the child hands something to somebody or reaches out for something he will generally say something, and if, as often happens, this is ta or da, it will be taken by its parents and others as a real word, different according to the language they speak; in England as there or thanks, in Denmark as tak ‘thanks’[30] or tag ‘take,’ in Germany as da ‘there,’ in France as tiens ‘hold,’ in Russia as day ‘give,’ in Italy as to, (= togli) ‘take.’ The form tê in Homer is interpreted by some as an imperative of teinō ‘stretch.’ These instances, however, are slightly different[160] in character from those discussed in the main part of this chapter.[31]

We all know that in historical times languages have been constantly changing, and we have much indirect evidence that in prehistoric times they did the same thing. But when it is asked if these changes, unavoidable as they seem to be, are to be ascribed primarily to children and their defective imitation of the speech of their elders, or if children’s language in general plays no part at all in the history of language, we find linguists expressing quite contrary views, without the question having ever been really thoroughly investigated.

Some hold that the child acquires its language with such perfection that it cannot be held responsible for the changes recorded in the history of languages: others, on the contrary, hold that the most important source of these changes is to be found in the transmission of the language to new generations. How undecided the attitude even of the foremost linguists may be towards the question is perhaps best seen in the views expressed at different times by Sweet. In 1882 he reproaches Paul with paying attention only to the shiftings going on in the pronunciation of the same individual, and not acknowledging “the much more potent cause of change which exists in the fact that one generation can learn the sounds of the preceding one by imitation only. It is an open question whether the modifications made by the individual in a sound he has once learnt, independently of imitation of those around him, are not too infinitesimal to have any appreciable effect” (CP 153). In the same spirit he asserted in 1899 that the process of learning our own language in childhood is a very slow one, “and the results are always imperfect.... If languages were learnt perfectly by the children of each generation, then languages would not change: English children would still speak a language as old at least as ‘Anglo-Saxon,’ and there would be[162] no such languages as French and Italian. The changes in languages are simply slight mistakes, which in the course of generations completely alter the character of the language” (PS 75). But only one year later, in 1900, he maintains that the child’s imitation “is in most cases practically perfect”—“the main cause of sound-change must therefore be sought elsewhere. The real cause of sound-change seems to be organic shifting—failure to hit the mark, the result either of carelessness or sloth ... a slight deviation from the pronunciation learnt in infancy may easily pass unheeded, especially by those who make the same change in their own pronunciation” (H 19 f.). By the term “organic shifting” Sweet evidently, as seen from his preface, meant shifting in the pronunciation of the adult, thus a modification of the sound learnt ‘perfectly’ in childhood. Paul, who in the first edition (1880) of his Prinzipien der Sprachgeschichte did not mention the influence of children, in all the following editions (2nd, 1886, p. 58; 3rd, 1898, p. 58; 4th, 1909, p. 63) expressly says that “die hauptveranlassung zum lautwandel in der übertragung der laute auf neue individuen liegt,” while the shiftings within the same generation are very slight. Paul thus modified his view in the opposite direction of Sweet[32]—and did so under the influence of Sweet’s criticism of his own first view!

When one finds scholars expressing themselves in this manner and giving hardly any reasons for their views, one is tempted to believe that the question is perhaps insoluble, that it is a mere toss-up, or that in the sentence “children’s imitation is nearly perfect” the stress may be laid, according to taste, now on the word nearly, and now on the word perfect. I am, however, convinced that we can get a little farther, though only by breaking up the question, instead of treating it as one vague and indeterminate whole.

Among recent writers Meringer has gone furthest into the question, adhering in the main to the general view that, just as in other fields, social, economic, etc., it is grown-up men who take the lead in new developments, so it is grown-up men, and not women or children, who carry things forward in the field of[163] language. In one place he justifies his standpoint by a reference to a special case, and I will take this as the starting-point of my own consideration of the question. He says: “It can be shown by various examples that they [changes in language] are decidedly not due to children. In Ionic, Attic and Lesbian Greek the words for ‘hundreds’ are formed in -kosioi (diakósioi, etc.), while elsewhere (in Doric and Bœotian) they appear as -kátioi. How does the o arise in -kósioi? It is generally said that it comes from o in the ‘tens’ in the termination -konta. Can it be children who have formed the words for hundreds on the model of the words for tens, children under six years old, who are just learning to talk? Such children generally have other things to attend to than to practise themselves in numerals above a hundred.” Similar formations are adduced from Latin, and it is stated that the personal pronouns are especially subject to change, but children do not use the personal pronouns till an age when they are already in firm possession of the language. Meringer then draws the conclusion that the share which children take in bringing about linguistic change is a very small one.

Now, I should like first to remark that even if it is possible to point to certain changes in language which cannot be ascribed to little children, this proves nothing with regard to the very numerous changes which lie outside these limits. And next, that all the cases here mentioned are examples of formation by analogy. But from the very nature of the case, the conditions requisite for the occurrence of such formations are exactly the same in the case of adults and in that of the children. For what are the conditions? Some one feels an impulse to express something, and at the moment has not got the traditional form at command, and so is driven to evolve a form of his own from the rest of the linguistic material. It makes no difference whether he has never heard a form used by other people which expresses what he wants, or whether he has heard the traditional form, but has not got it ready at hand at the moment. The method of procedure is exactly the same whether it takes place in a three-year-old or in an eighty-three-year-old brain: it is therefore senseless to put the question whether formations by analogy are or are not due to children. A formation by analogy is by definition a non-traditional form. It is therefore idle to ask if it is due to the fact that the language is transmitted from generation to generation and to the child’s imperfect repetition of what has been transmitted to it, and Meringer’s argument thus breaks down in every respect.

It must not, of course, be overlooked that children naturally come to invent more formations by analogy than grown-up people,[164] because the latter in many cases have heard the older forms so often that they find a place in their speech without any effort being required to recall them. But that does not touch the problem under discussion; besides, formations by analogy are unavoidable and indispensable, in the talk of all, even of the most ‘grown-up’: one cannot, indeed, move in language without having recourse to forms and constructions that are not directly and fully transmitted to us: speech is not alone reproduction, but just as much new-production, because no situation and no impulse to communication is in every detail exactly the same as what has occurred on earlier occasions.

If, leaving the field of analogical changes, we begin to inquire whether the purely phonetic changes can or must be ascribed to the fact that a new generation has to learn the mother-tongue by imitation, we shall first have to examine an interesting theory in which the question is answered in the affirmative, at least with regard to those phonetic changes which are gradual and not brought about all at once; thus, when in one particular language one vowel, say [e·], is pronounced more and more closely till finally it becomes [i·], as has happened in E. see, formerly pronounced [se·] with the same vowel as in G. see, now [si·]. E. Herzog maintains that such changes happen through transference to new generations, even granted that the children imitate the sound of the grown-up people perfectly. For, it is said, children with their little mouths cannot produce acoustically the same sound as adults, except by a different position of the speech-organs; this position they keep for the rest of their lives, so that when they are grown-up and their mouth is of full size they produce a rather different sound from that previously heard—which altered sound is again imitated by the next generation with yet another position of the organs, and so on. This continuous play of generation v. generation may be illustrated in this way:

[165]

It is, however, easy to prove that this theory cannot be correct. (1) It is quite certain that the increase in size of the mouth is far less important than is generally supposed (see my Fonetik, p. 379 ff., PhG, p. 80 ff.; cf. above, V § 1). (2) It cannot be proved that people, after once learning one definite way of producing a sound, go on producing it in exactly the same way, even if the acoustic result is a different one. It is much more probable that each individual is constantly adapting himself to the sounds heard from those around him, even if this adaptation is neither as quick nor perhaps as perfect as that of children, who can very rapidly accommodate their speech to the dialect of new surroundings: if very far-reaching changes are rare in the case of grown-up people, this proves nothing against such small adaptations as are here presupposed. In favour of the continual regulation of the sound through the ear may be adduced the fact that adults who become perfectly deaf and thus lose the control of sounds through hearing may come to speak in such a way that their words can hardly be understood by others. (3) The theory in question also views the relations between successive generations in a way that is far removed from the realities of life: from the wording one might easily imagine that there were living together at any given time only individuals of ages separated by, say, thirty years’ distance, while the truth of the matter is that a child is normally surrounded by people of all ages and learns its language more or less from all of them, from Grannie down to little Dick from over the way, and that (as has already been remarked) its chief teachers are its own brothers and sisters and other playmates of about the same age as itself. If the theory were correct, there would at any rate be a marked difference in vowel-sounds between anyone and his grandfather, or, still more, great-grandfather: but nothing of the kind has ever been described. (4) The chief argument, however, against the theory is this, that were it true, then all shiftings of sounds at all times and in all languages would proceed in exactly the same direction. But this is emphatically contradicted by the history of language. The long a in English in one period was rounded and raised into o, as in OE. stan, na, ham, which have become stone, no, home; but when a few centuries later new long a’s had entered the language, they followed the opposite direction towards e, now [ei], as in name, male, take. Similarly in Danish, where an old stratum of long a’s have become å, as in ål, gås, while a later stratum tends rather towards [æ], as in the present pronunciation of gade, hale, etc. At the same time the long a in Swedish tends towards the rounded pronunciation (cf. Fr. âme, pas): in one sister language we thus witness a repetition of the old shifting, in the other a[166] tendency in the opposite direction. And it is the same with all those languages which we can pursue far enough back: they all present the same picture of varying vowel shiftings in different directions, which is totally incompatible with Herzog’s view.

We shall do well to put aside such artificial theories and look soberly at the facts. When some sounds in one century go one way, and in another, another, while at times they remain long unchanged, it all rests on this, that for human habits of this sort there is no standard measure. Set a man to saw a hundred logs, measuring No. 2 by No. 1, No. 3 by No. 2, and so on, and you will see considerable deviations from the original measure—perhaps all going in the same direction, so that No. 100 is very much longer than No. 1 as the result of the sum of a great many small deviations—perhaps all going in the opposite direction; but it is also possible that in a certain series he was inclined to make the logs too long, and in the next series too short, the two sets of deviations about balancing one another.

It is much the same with the formation of speech sounds: at one moment, for some reason or other, in a particular mood, in order to lend authority or distinction to our words, we may happen to lower the jaw a little more, or to thrust the tongue a little more forward than usual, or inversely, under the influence of fatigue or laziness, or to sneer at someone else, or because we have a cigar or potato in our mouth, the movements of the jaw or of the tongue may fall short of what they usually are. We have all the while a sort of conception of an average pronunciation, of a normal degree of opening or of protrusion, which we aim at, but it is nothing very fixed, and the only measure at our disposal is that we are or are not understood. What is understood is all right: what does not meet this requirement must be repeated with greater correctness as an answer to ‘I beg your pardon?’

Everyone thinks that he talks to-day just as he did yesterday, and, of course, he does so in nearly every point. But no one knows if he pronounces his mother-tongue in every respect in the same manner as he did twenty years ago. May we not suppose that what happens with faces happens here also? One lives with a friend day in and day out, and he appears to be just what he was years ago, but someone who returns home after a long absence is at once struck by the changes which have gradually accumulated in the interval.

Changes in the sounds of a language are not, indeed, so rapid as those in the appearance of an individual, for the simple reason that it is not enough for one man to alter his pronunciation,[167] many must co-operate: the social nature and social aim of language has the natural consequence that all must combine in the same movement, or else one neutralizes the changes introduced by the other; each individual also is continually under the influence of his fellows, and involuntarily fashions his pronunciation according to the impression he is constantly receiving of other people’s sounds. But as regards those little gradual shiftings of sounds which take place in spite of all this control and its conservative influence, changes in which the sound and the articulation alter simultaneously, I cannot see that the transmission of the language to a new generation need exert any essential influence: we may imagine them being brought about equally well in a society which for hundreds of years consisted of the same adults who never died and had no issue.

While in the shiftings mentioned in the last paragraphs articulation and acoustic impression went side by side, it is different with some shiftings in which the old sound and the new resemble one another to the ear, but differ in the position of the organs and the articulations. For instance, when [þ] as in E. thick becomes [f] and [ð] as in E. mother becomes [v], one can hardly conceive the change taking place in the pronunciation of people who have learnt the right sound as children. It is very natural, on the other hand, that children should imitate the harder sound by giving the easier, which is very like it, and which they have to use in many other words: forms like fru for through, wiv, muvver for with, mother, are frequent in the mouths of children long before they begin to make their appearance in the speech of adults, where they are now beginning to be very frequent in the Cockney dialect. (Cf. MEG i. 13. 9.) The same transition is met with in Old Fr., where we have muef from modu, nif from nidu, fief from feodu, seif, now soif, from site, estrif (E. strife) from stridh, glaive from gladiu, parvis from paradis, and possibly avoutre from adulteru, poveir, now pouvoir, from potere. In Old Gothonic we have the transition from þ to f before l, as in Goth. þlaqus = MHG. vlach, Goth. þlaihan = OHG. flêhan, þliuhan = OHG. fliohan; cf. also E. file, G. feile = ON. þēl, OE. þengel and fengel ‘prince,’ and probably G. finster, cf. OHG. dinstar (with d from þ), OE. þeostre. In Latin we have the same transition, e.g. in fumus, corresponding to Sansk. dhumás, Gr. thumós.[34]

[168]

The change from the back-open consonant [x]—the sound in G. buch and Scotch loch—to f, which has taken place in enough, cough, etc., is of the same kind. Here clearly we have no gradual passage, but a jump, which could hardly take place in the case of those who had already learnt how to pronounce the back sound, but is easily conceivable as a case of defective imitation on the part of a new generation. I suppose that the same remark holds good with regard to the change from kw to p, which is found in some languages, for instance, Gr. hippos, corresponding to Lat. equus, Gr. hepomai = Lat. sequor, hêpar = Lat. jecur; Rumanian apa from Lat. aqua, Welsh map, ‘son’ = Gaelic mac, pedwar = Ir. cathir, ‘four,’ etc. In France I have heard children say [pizin] and [pidin] for cuisine.

There is an important class of sound changes which have this in common with the class just treated, that the changes take place suddenly, without an intermediate stage being possible, as in the changes considered in IX § 4. I refer to those cases of assimilation, loss of consonants in heavy groups and transposition (metathesis), with which students of language are familiar in all languages. Instances abound in the speech of all children; see above, V § 4.

If now we dared to assert that such pronunciations are never heard from people who have passed their babyhood, we should here have found a field in which children have exercised a great influence on the development of language: but of course we cannot say anything of the sort. Any attentive observer can testify to the frequency of such mispronunciations in the speech of grown-up people. In many cases they are noticed neither by the speaker nor by the hearer, in many they may be noticed, but are considered too unimportant to be corrected, and finally, in some cases the speaker stops to repeat what he wanted to say in a corrected form. Now it would not obviously do, from their frequency in adult speech, to draw the inference: “These changes are not to be ascribed to children,” because from their frequent appearance on the lips of the children one could equally well infer: “They are not to be ascribed to grown-up people.” When we find in Latin impotens and immeritus with m side by side with indignus and insolitus with n, or when English handkerchief is pronounced with [ŋk] instead of the original [ndk], the change is not to be charged against children or grown-up people exclusively, but against both parties together: and so when t is lost in waistcoat [weskət], or postman or castle, or k in asked. There[169] is certainly this difference, that when the change is made by older people, we get in the speech of the same individual first the heavier and then the easier form, while the child may take up the easier pronunciation first, because it hears the [n] before a lip consonant as [m], and before a back consonant as [ŋ], or because it fails altogether to hear the middle consonant in waistcoat, postman, castle and asked. But all this is clearly of purely theoretical interest, and the result remains that the influence of the two classes, adults and children, cannot possibly be separated in this domain.[35]

Next we come to those changes which result in what one may call ‘stump-words.’ There is no doubt that words may undergo violent shortenings both by children and adults, but here I believe we can more or less definitely distinguish between their respective contributions to the development of language. If it is the end of the word that is kept, while the beginning is dropped, it is probable that the mutilation is due to children, who, as we have seen (VII § 7), echo the conclusion of what is said to them and forget the beginning or fail altogether to apprehend it. So we get a number of mutilated Christian names, which can then be used by grown-up people as pet-names. Examples are Bert for Herbert or Albert, Bella for Arabella, Sander for Alexander, Lottie for Charlotte, Trix for Beatrix, and with childlike sound-substitution Bess (and Bet, Betty) for Elizabeth. Similarly in other languages, from Danish I may mention Bine for Jakobine, Line for Karoline, Stine for Kristine, Dres for Andres: there are many others.

If this way of shortening a word is natural to a child who hears the word for the first time and is not able to remember the beginning when he comes to the end of it, it is quite different when others clip words which they know perfectly well: they will naturally keep the beginning and stop before they are half through the word, as soon as they are sure that their hearers understand what is alluded to. Dr. Johnson was not the only one who “had a way of contracting the names of his friends, as Beauclerc, Beau; Boswell, Bozzy; Langton, Lanky; Murphy, Mur; Sheridan, Sherry; and Goldsmith, Goldy, which Gold[170]smith resented” (Boswell, Life, ed. P. Fitzgerald, 1900, i. 486). Thackeray constantly says Pen for Arthur Pendennis, Cos for Costigan, Fo for Foker, Pop for Popjoy, old Col for Colchicum. In the beginning of the last century Napoleon Bonaparte was generally called Nap or Boney; later we have such shortened names of public characters as Dizzy for Disraeli, Pam for Palmerston, Labby for Labouchere, etc. These evidently are due to adults, and so are a great many other clippings, some of which have completely ousted the original long words, such as mob for mobile, brig for brigantine, fad for fadaise, cab for cabriolet, navvy for navigator, while others are still felt as abbreviations, such as photo for photograph, pub for public-house, caps for capital letters, spec for speculation, sov for sovereign, zep for Zeppelin, divvy for dividend, hip for hypochondria, the Cri and the Pavvy for the Criterion and the Pavilion, and many other clippings of words which are evidently far above the level of very small children. The same is true of the abbreviations in which school and college slang abounds, words like Gym(nastics), undergrad(uate), trig(onometry), lab(oratory), matric(ulation), prep(aration), the Guv for the governor, etc. The same remark is true of similar clippings in other languages, such as kilo for kilogram, G. ober for oberkellner, French aristo(crate), réac(tionnaire), college terms like desse for descriptive (géométrie d.), philo for philosophie, preu for premier, seu for second; Danish numerals like tres for tresindstyve (60), halvfjerds(indstyve), firs(indstyve). We are certainly justified in extending the principle that abbreviation through throwing away the end of the word is due to those who have previously mastered the full form, to the numerous instances of shortened Christian names like Fred for Frederick, Em for Emily, Alec for Alexander, Di for Diana, Vic for Victoria, etc. In other languages we find similar clippings of names more or less carried through systematically, e.g. Greek Zeuxis for Zeuxippos, Old High German Wolfo for Wolfbrand, Wolfgang, etc., Icelandic Sigga for Sigríðr, Siggi for Sigurðr, etc.

I see a corroboration of my theory in the fact that there are hardly any family names shortened by throwing away the beginning: children as a rule have no use for family names.[36] The rule, however, is not laid down as absolute, but only as holding in the main. Some of the exceptions are easily accounted for. ’Cello for violoncello undoubtedly is an adults’ word, originating[171] in France or Italy: but here evidently it would not do to take the beginning, for then there would be confusion with violin (violon). Phone for telephone: the beginning might just as well stand for telegraph. Van for caravan: here the beginning would be identical with car. Bus, which made its appearance immediately after the first omnibus was started in the streets of London (1829), probably was thought expressive of the sound of these vehicles and suggested bustle. But bacco (baccer, baccy) for tobacco and taters for potatoes belong to a different sphere altogether: they are not clippings of the usual sort, but purely phonetic developments, in which the first vowel has been dropped in rapid pronunciation (as in I s’pose), and the initial voiceless stop has then become inaudible; Dickens similarly writes ’tickerlerly as a vulgar pronunciation of particularly.[37]

Some of the most typical childish sound-substitutions can hardly be supposed to leave any traces in language as permanently spoken, because they are always thoroughly corrected by the children themselves at an early age; among these I reckon the almost universal pronunciation of t instead of k. When, therefore, we do find that in some words a t has taken the place of an earlier k, we must look for some more specific cause of the change: but this may, in some cases at any rate, be found in a tendency of children’s speech which is totally independent of the inability to pronounce the sound of k at an early age, and is, indeed, in no way to be reckoned among phonetic tendencies, namely, the confusion resulting from an association of two words of similar sound (cf. above, p. 122). This, I take it, is the explanation of the word mate in the sense ‘husband or wife,’ which has replaced the earlier make: a confusion was here natural, because the word mate, ‘companion,’ was similar not only in sound, but also in signification. The older name for the ‘soft roe’ of fishes was milk (as Dan. mælk, G. milch), but from the fifteenth century milt has been substituted for it, as if it were the same organ as the milt, ‘the spleen.’ Children will associate words of similar sound even in cases where there is no connecting link in their significations; thus we have bat for earlier bak, bakke (the animal, vespertilio), though the other word bat, ‘a stick,’ is far removed in sense.

I think we must explain the following cases of isolated sound-substitution as due to the same confusion with unconnected words in the minds of children hearing the new words for the first time: trunk in the sense of ‘proboscis of an elephant,’ formerly trump, from Fr. trompe, confused with trunk, ‘stem of a tree’; stark-naked, formerly start-naked, from start, ‘tail,’ confused with stark, ‘stiff’; vent, ‘air-hole,’ from Fr. fente, confused with vent,[173] ‘breath’ (for this v cannot be due to the Southern dialectal transition from f, as in vat from fat, for that transition does not, as a rule, take place in French loans); cocoa for cacao, confused with coconut; match, from Fr. mèche, by confusion with the other match; chine, ‘rim of cask,’ from chime, cf. G. kimme, ‘border,’ confused with chine, ‘backbone.’ I give some of these examples with a little diffidence, though I have no doubt of the general principle of childish confusion of unrelated words as one of the sources of irregularities in the development of sounds.

These substitutions cannot of course be separated from instances of ‘popular etymology,’ as when the phrase to curry favour was substituted for the former to curry favel, where favel means ‘a fallow horse,’ as the type of fraud or duplicity (cf. G. den fahlen hengst reiten, ‘to act deceitfully,’ einen auf einem fahlen pferde ertappen, ‘to catch someone lying’).

We now come to the phenomenon for which I have ventured to coin the term ‘metanalysis,’ by which I mean that words or word-groups are by a new generation analyzed differently from the analysis of a former age. Each child has to find out for himself, in hearing the connected speech of other people, where one word ends and the next one begins, or what belongs to the kernel and what to the ending of a word, etc. (VII § 6). In most cases he will arrive at the same analysis as the former generation, but now and then he will put the boundaries in another place than formerly, and the new analysis may become general. A naddre (the ME. form for OE. an nædre) thus became an adder, a napron became an apron, an nauger: an auger, a numpire: an umpire; and in psychologically the same way an ewte (older form evete, OE. efete) became a newt: metanalysis accordingly sometimes shortens and sometimes lengthens a word. Riding as a name of one of the three districts of Yorkshire is due to a metanalysis of North Thriding (ON. þriðjungr, ‘third part’), as well as of East Thriding, West Thriding, after the sound of th had been assimilated to the preceding t.

One of the most frequent forms of metanalysis consists in the subtraction of an s, which originally belonged to the kernel of a word, but is mistaken for the plural ending; in this way we have pea instead of the earlier peas, pease, cherry for ME. cherris, Fr. cerise, asset from assets, Fr. assez, etc. Cf. also the vulgar Chinee, Portuguee, etc.[38]