;*L m l , gale ^(centennial publication?! LECTURES ON THE STUDY OF LANGUAGE gale 'Bicentennial publications With the approval of the President and Fellows of Yale University, a series of volumes has been prepared by a number of the Professors and In- structors, to be issued in connection with the Bicentennial Anniversary, as a partial indica- tion of the character of the studies in which the "University teachers are engaged. This series of volumes is respectfully dedicated to ] Changes which take place in the transmission of language from one individual to another ( 22) : [i] Changes due to defective perception by the ear in the transmission of speech to children. Partial sound deaf- ness. An " acoustic sound basis " should be assumed parallel to the physiological "basis of articulation." The loss of inaudible movements ( 23). The influence of spelling on pronunciation. The acquisition of a foreign idiom through ear and eye ( 24). [ii] Changes made by children in reproducing speech-sounds. Compensatory articulations and their effect on the future develop- ment of sounds. Variability in articulation of the adult as a factor in sound change ( 25). [iii] Changes in the reproduction of sounds by foreigners ( 26). Speech-mixture as an explanation of the differences in the Indo-European languages. Ascoli's proofs for " ethnological sound changes." The difficulty of meeting Ascoli's requirements hi the case of the prehistoric Indo-European languages. Hempl's discussion of the influence of race-mixture on language. Ratzel on the mechanism of migration and coloni- zation. The process of " infiltration." The origin of " contact languages." The present Indo-European languages may be survivals of such contact languages ( 27). THE NATURE OP PHONETIC LAWS. The difference of Les- kien's and Curtius' views regarding phonetic law. The neo-gram- marians insisted on a causal explanation of all exceptions, but did not demand such an explanation for the phonetic "laws." Most phonetic laws are formulaic statements of observed regularities in phonetic development and not statements of the observed se- quence of two phenomena which stand to each other in the rela- tion of cause and effect ( 28). The neo-grammarians contrast phonetic laws as the physiological element of linguistic changes with analogy formations as the psychological element. The mechanical conception of phonetic laws is based upon the regu- xvi CONTENTS LECTURE IV continued larity of phonetic changes, not upon their nature. The necessity of recognizing regularities which are not physiological (mechani- cal) but psychological (social) ( 29). The degree of uniformity of sound changes in the spoken language ( 30). The reason for relative uniformity of most sound changes is not a simultaneous change in the articulations of the members of a linguistic com- munity but social imitation. The spread of phonetic changes as of all other linguistic change rests on a psychological basis. Im- itation as a conservative and as a progressive factor ( 31). The explanation of the phenomena summarized in the " phonetic laws " of our grammars is to be sought in general psychological or physical laws ( 32). LECTURE V CHANGES IN LANGUAGE: III. SEMANTIC CHANGE 274-329 Every linguistic utterance must be examined (1) with respect to its phonetic form (morphological aspect), and (2) with respect to its significance (semantic aspect). What is linguistic form ? The means of linguistic expression ( 1). The hearer starts with the form, the speaker with the meaning. The study of the speech content in relation to the linguistic form with which it appears associated is the province of semasiology ( 2). Analysis by the speaker of a composite idea into elements. The grammatical "sentence " is the linguistic reflex of such an analysis. The syn- thesis of the hearer ( 3). What calls for expression in language? \_A~\ Things, qualities, actions, and states. Names and roots. The difference between logical and grammatical categories ( 4). A compound idea expressed by modifications of one name ( 5). [5] The attitude of the speaker. Its expression by inde- pendent words or as modifications of "names" ( 6). [C] The relation of the members of an utterance to each other ( 7). The polysynthetic character of the Indo-European languages and its consequences. The absence of a pure casus nominativus. The im- personal verbs as a compromise between grammar (Indo-European and Semitic) and logic ( 8). The terms " meaning " and " func- tion " are essentially synonymous, though the former is current with reference to independent units, the latter usually refers to significant parts of such units. The dynamic problems of lexicog- raphy and syntax are the same (9). The difference between xvii CONTENTS LECTURE V continued logical and psychological treatment of semantic problems (10). The composite nature of the psychical contents of words ( 11). The emotional element in words. Its sources. Its effect. Taboo on words. Changes and national differences in the valuation of words ( 12). The unification of the composite psychical content of a word by subordinating all elements to one dominant element. The variability of this dominant element. Words are not felt to be descriptive of the things for which they stand. The original process of naming. The bearing of etymology on lexicography ; " linguistic" and "philological" lexicography ( 13). DIVISIONS OF SEMANTIC CHANGE: [A~\ Oscillation resulting from the variability of the dominant element. Devices for fixing the domi- nant element ( 14). [.B] The fusion of two adjacent concepts into one, accompanied by the loss of the name of either concept. This is a process of abbreviation. Change in the semantic value of words due to a redistribution of the elements of the larger com- pound idea over the two component names ( 15). [C] The dis- solution of percepts. The change from synthetic (inflectional) structure to analytic structure. The progress from particular to general ideas and its reflex in the earlier and later meaning of "roots." The difference between indistinct notions and general concepts ( 16). [D] Semantic changes due to associative inter- ference ( 17). [.E] Semantic changes due to transfer. The use of concrete terms for abstract ideas. The transfer of terms from one sense sphere to another ( 18). The shifting of sense by a series of semantic changes. Changes in the meaning of one word frequently disturb other words which are in some way connected with it (19). INDEX . 333 xviii LECTURES ON THE STUDY OF LANGUAGE LECTURES THE STUDY OF LANGUAGE LECTURE I THE LEADING IDEAS OF LINGUISTIC SCIENCE DUEING THE NINETEENTH CENTURY X IT 1 is a common method to define a science by enumerating and describing the objects with which it deals, and (in a manner not unlike that of an actio finium regundorum in Roman law) by establishing 1 My plan was to sketch in this first lecture the beginnings of the most important currents of linguistic thought during the nineteenth cen- tury. The relative importance of these currents has been judged accord- ing to their bearing upon the linguistic investigations and problems of our own time. I have endeavored to seize an idea not when it occurred absolutely for the first time (the germs of almost all the leading ideas are much older than the century past), but where it appeared for the first time in such a shape as to exercise some effect upon the trend of investigation. Many ideas have been advanced at an unfavorable junc- ture and left no impress whatever, while these same ideas advanced again at some later point (often independently) have been received with enthusiasm. Again, a new idea being once advanced and having gained acceptance, space forbade the tracing out of its subsequent history in detail. A large mass of work of great value for the history of the study of language is thus necessarily omitted, for, as a rule, a new idea gains in importance and solidity by the careful elaboration of later scholars. Nowhere has it been my purpose to characterize investigators (in the biographical fashion of Achelis' Moderne Volkerkuude), but to trace what appeared to be leading ideas. 1 1 LECTURES ON THE STUDY OF LANGUAGE boundary lines between it and its neighbors. Such a method is unsatisfactory, 1 because it attempts a separa- tion which in reality does not exist, and which, if carried out in practice, would necessarily result in very serious disadvantages. For how often the same object or phe- nomenon is treated by more than one science. Theseus, we are told, erected upon the Isthmus of Corinth a column bearing upon one side the inscription : " Here is the Peloponnesus and not Attica," and upon the oppo- site side: "Here is Attica and not the Peloponnesus." But no scientific investigator may safely imitate him, for the very complexity of most objects which come under his observation forces him to concentrate upon them all the light from whatever quarter he may have to seek it. The secret of the success of one of the fore- most Greek archaeologists 2 lay to a great extent in his thorough training in architecture. The investigators of literature and of political economy are equally interested in the Homeric poems. 3 The study of Goethe and of Rousseau owes a valuable contribution to the pen of a physician, 4 and the requirements for the successful in- 1 Cf. e. g. Heeren, Geschichte des Stadiums der Klassischen Litteratur (1797), Einleitung, 1, vol. I, p. 1. L. Lange, Kleine Schriften (1887), p. 7 (in his inaugural address, 1855). Curtius, Philologie und Sprach- wissenschaft (1862), p. 3. W.Arnold, Ansiedelungen und Wanderungen deutscher Stamme (1874-5), p. 6. Scherer, Jenaer Litteratur Zeitung (1876), III, p. 472 = Kleine Schriften, I, p. 458. Poehlmann, Aus Alter- tum und Gegenwart (1895), p. 34. 2 Dorpfeld. 8 Poehlmann, Die Feldgemeinschaft bei Homer, in Zt. f. Sozial- und Wirtschaftsgeschichte, 1893 = Aus Altertum u. Gegenwart, (1895), p. 105 ; Znr geschichtlicheu Beurteilung Homers, in Sybel's Historische Zt., 1894 = Aus Altertnm u. Gegenwart, (1895) p. 56. * P. J. Mobius, tfber das Pathologische bei Goethe (1898) QCf. the reviews in Litterarisches Centralblatt (1898), col. 1902, and in Zt. f. Psy- chologie und Physiologic der Sinnesorgane, XX (1899), p. 221] and the same author's J. J. Rousseau's Krankheitsgeschichte (1889). Cf. a 2 HISTORICAL INTRODUCTION terpretation of the military commentaries of Caesar, of the medical treatises of Hippocrates or the Thucydidean description of the plague are obvious to every one. True specialization consists in making all information of whatever kind contribute toward a fuller under- standing of the one object under investigation. 1 But quite apart from being unsatisfactory and often harmful for practical purposes, this method, theoreti- cally also, suffers from an inherent and incurable weak- ness. 2 Such a definition, to be exact and exhaustive, must consist in a repetition of all the main results of a science. It necessarily presupposes these results already achieved. These it sets out to systematize. Thus, as Taine 3 has admirably put it, un systeme est une explication de V ensemble et indique une ceuvre faite* Sciences, however, are living and constantly developing. They show the same lack of symmetry in their develop- ment which is the characteristic sign of an organism during the period of its growth. To defy codification is the sign of a youthful science as it is the sign of a youthful religion. It is only when their vitality is spent that they submit to being embalmed and laid at rest in a final system. 2 Fortunately the unity and character of a science do not depend on the establishment of such boundary lines. As the true unity of a drama does not depend on the dramatis personae but rests upon the dramatic similar discussion of Kleist's Penthesilea by Krafft-Ebing and Roettken (Zt. f. vergleich. Litteraturgeschichte, N. F., VIII, p. 28). 1 Some good remarks against the narrow policy of the Chinese Wall in the Memoirs of Chief Justice Parsons, p. 153, and in Sir Edward Coke's Preface to Co. Rep., part III. 2 Prantl, Geschichte der Logik (1855), I, p. 1. For a fuller exposition cf. Wnndt, Philosoph. Studien, XII (1895), p. 1. 8 Essais de Critique et d'Histoire (5th ed.), Pre"f. p. vii. 3 LECTURES ON THE STUDY OF LANGUAGE action, 1 so the unity of a science consists in the uni- form final aim 2 toward which all individual investi- gations converge and which makes them all parts of one organic whole. It is this final goal which deter- mines the uniform point of vision from which all objects and phenomena are to be viewed, and thus gives inde- pendence, coherence, and individuality to a science. It determines the method, i. e. the manner in which all material should be treated. 3 As long as a science firmly adheres to its characteristic point of view, it will always be clearly kept distinct from its neighbors, 4 and, far from sacrificing its independence or individuality by hospitably receiving their aid, it will on the contrary gain thereby in strength. But as soon as it fails to keep its goal in clear view it begins to disintegrate, and unless it can find another central point around which its elements may crystallize, though it may linger for a while, it will cease to be productive. It is for this reason that many of the most important epochs of a science are not marked by external changes such as territorial expansion, but by internal transformations brought about by modifications of the point of vision. 3 The growth of a science is reflected in the chief ten- 1 Aristotle, Poet., VI. 10. Scherer (Zt. f. d. oesterreich. Gymnasien, XXIX, p. 125 [1878] = Kleine Schriften, I, p. 373, quoted by Ries, Was 1st Syntax? [1894], p. 95 and p. 161, note 75) speaks of the " hero " who is required to give unity to an investigation ; a happy comparison which also occurs in Humboldt (Ges. Werke, I, p. 310) and in Littre (Etudes et Glanures [1880], p. 2). 2 Cf. Aristotle, Poet., VI, 10, " rl> Se -re\os ptyiffrov air, etc., with which it had been analogi- cally connected. Although there are psychological processes producing sound change which do not fall under the head of analogy, yet those which do are numerous enough to make the introduction of analogy as a methodological principle an important step in the psychological interpretation of linguistic facts. A peculiar one-sidedness of the neo-grammarian movement must be noted here. It is the quiet ac- ceptance of all regular development, e. #., of a sound change regularly exhibited by a large mass of words, without inquiry into its cause, when at the same time a causal explanation for any irregularity was demanded. The Attic change of an (Ionic) rj to a after p, i, e, v was accepted without further explanation because the change is universal; but the 77 in ^0/377709 required causal ex- planation (due to analogy of crrparr}^}. In the dis- cussion which centred around the term "phonetic law" this discrimination in favor of majority changes, and their acceptance without genetic explanation, and against minority changes for which such explanation was demanded, has tended to obscure the true issue and re- tarded a settlement, as will appear hereafter. 3 36 The study of semantics, to which reference was made 1 Osthoff and Brugmann distinctly say (Morph. Untersuch., I, p. xiii): " Aller lautwandel, soweit er mechanisch vor sick gehl., vollzieht sich nach ausnahmslosen gesetzen." 2 Brugmann, Morph. Untersuch., I, p. 156, note. 8 Cf . below, Lecture IV. 70 HISTORICAL INTRODUCTION above, has its beginnings in lexicography. In his Lec- tures on the Latin Language (delivered for the last time during the winter term of 1826-27, and published mu6h later, in 1839, by his pupil Haase) K. Reisig had called attention to the importance of a scientific and systematic study of the meanings of words and devoted about a dozen pages l to a few suggestions concerning " the prin- ciples of semantic development," in which he states, among other things, that "the basis of the semantic de- velopment of words is the association of related ideas." His premature death, perhaps, prevented Reisig from penetrating farther into the subject which he had thus opened. 2 It was reserved for his pupil, Agathon 1 K. Reisig's Vorlesungen iiber lateinische Sprachwissenschaft (1839), p. 18, 20. "Das Wort betrachten wir in seiner Gestalt nach gewissen Grundsatzen, und daraus entsteht 1) die Etymologic, Formenlehre > demnach seine Verbindungen mit anderen Wortern, und dies bildet 2) die Syntax. Das Wort hat aber noch eine andere Eigenschaft an sich, die Bedeutung ; es giebt eine Gattung von Wortern, die in jeder Art der Rede in Anspruch genommen werden, deren Bedeutung aber weder in der Etymologie erortert werden kann, noch auch in der Syntax Plata findet, weil ihre Bedeutung weder von etymologischen noch von syntak- tischen Regeln abhangig ist. Lassen sich nun gewisse Grundsatze auf- stellen, welche von einer Menge von Wortern die Entwickelung ihrer Bedeutung und ihrer Anwendung zeigen, so entsteht noch ein integriren- der Theil der Grammatik namlich 3) die Bedeutungslehre, Semasiologie." P. 286-307 are devoted to Semasiology : 286-298 " Grundsatze fiir die Entwickelung der Bedeutung" and 298-307 " Grundsatze iiber die Wahl der WSrter nach ihrer Bedeutung." (This second half is purely stylistic.) P. 286, 171 " Die Entfaltung der Gedankenreihe in Betreff der Bedeutung der Worter ist ein anziehendes, anmuthiges Geschaft . . . Die Grundlagen der Ideenentwickelung in den Wortern ist die Gedanken- association in der Gemeinschaft der Vorstellungen." 2 The following passage from W. v. Schlegel, Reflexions, etc. (1832)> p. 42, is worth quoting : Les articles concernant les termes polystman- tiques sont particulierement peu satisfaisans : la re'daction de ces sortes d'articles est pourtant la pierre de touche d'un bon dictionnaire. II fant d'abord chercher la signification primitive ou fondamentale du mot, a laquelle toutes les autres doivent etre rameuees comme a leur centre 71 LECTURES ON THE STUDY OF LANGUAGE Benary (1834), to leave the purely lexicographical aspect of the semantic question and to give to it a much wider and at the same time profounder meaning. He was the first to distinguish clearly between the formal and the semantic side of a word, and this not only with reference to the word as a whole, but he applied this same distinction to the grammatical elements of which the word is made up, such as inflectional and formative affixes. These also, he points out, deserve a separate treatment from the point of view of their meaning, which should be fuller and more systematic than that which is now very grudgingly permitted them, tucked away in some syntactical paragraph where it is wholly out of place. 1 commnn ; il faut observer raffinite" des idees et retracer le passage graduel et nuance de 1'une a 1'autre ; il faut expliquer les transitions brusques et inattendues : ce sont pour la plupart des expressions origi- nairement figurees et devenues peu a peu des mots propres, lorsque la metaphore ou Pallusion qui leur avait donne' naissance a ete' obliteree par le temps. Quelquefois une seule se'rie ne suffit pas: il faut revenir plusieurs fois a la tige commune, pour suivre les ramifications divergentes. 1 Benary in Jahrbiicher f. wissenschaftliche Kritik, Juli 1834, col. 66 ff. : " Das Wort erscheint in der Sprache als Ausdruck des Gedankens und wie sein Ursprung rein der Form nach in der Laut- und Sylbeulehre gezeigt wird, so wird seine Genesis als Bedeutungstrager in der Bedeu- tungslehre und endlich sein Begriff als Darstellung des vollstandigen Gedankens in der Syntax dargethan. So zerfallt uns die ganze Gram- matik in drei Theile. I. Das Wort als Form : (a) die Elementarlehre (Laut Sylbe Wort). (b) die Flexionslehre. (c) die Ableitungs- und Composition slehre, sammtlich nurformell, die Lantveranderungen und Erweiterungen, abgesehen von ihrer Geltung und ihrem Werthe als Begriffstrager. II. Die Bedeutungslehre : (a) Die Wurzel als allgemeines, nnentwickeltes, in ihrer Bewegung in sich (die verschiedenen Bedeutungen der Wurzel) und in ihrem Fortgangzura System des Wortes: 1. Das Begriffswort (als Adjectivum, Substantivum, Verbum). 72 HISTORICAL INTRODUCTION Here, then, is the germ of a healthy reaction against the one-sided formal treatment of Bopp and his follow- ers. Its growth has been slow, and only during the last decades has this side of the question begun to receive adequate treatment. 1 But wherever these problems were touched at all they demanded a psychological solution. 37 The same is finally true of syntax when freed from the 2. Das Formwort (als Pronomen, Zahlwort, Partikel). (b) Die grammatische Form : 1. Die Verhaltnisse der Dinge zu einander (Genus, Casus, Numerus). 2. Die Verhaltnisse der Eigenschaften (die Gradation). 3. Die Verhaltnisse der Handlung (die Zeit, die Modi, die Genera). (c) die Ableituug und Composition. III. Das Wort als Gedanke (Syntaxis). (a) der einfache Satz Taxis. (b) der Nebensatz Parataxis. (c) die Periode Syntaxis. Von den gewohnlichen besseren Eintheilungen . . . weichen wir also darin ab, dass wir einen zweiteu Theil der Gesammtgrammatik in der Bedeutungslehre vindiciren. Dies hat einst mein unvergesslicher Lehrer Keisig in seinen Vorlesungen iiber lateinische Sprachwissenschaft gethan, doch nur ahueiid ; denn dass er nur den ganz ausserlichen Theil der Synonymik (diese fallt ihrer grammatischen, nicht lexicalischen Seite nach bei uns in den Fortgang der Wurzel zum System des Wortes) und einige ganz specielle Falle in diesen Theil zog, ohne wie wir das ganze grammatische Feld in ihr Gebiet in vollstandiger verniinftiger Gliederung aufzunehmen, das wissen mit mir seine ehemaligen zahlreichen Zuhorer; den Mangel fiihlte er, auszufiillen war ihm der gerade in der regsten Zeit der Entwicklung der Sprachwissenschaft starb uicht vergonnt. Die Wichtigkeit dieses Theiles macht sich aber vor allem bei Behandlung der Syntax kund. Hier werden gewohnlich die Formen, mit Vorausse- tzung der Kenntniss ihrer Bedeutung, ohne weiteres aufgenommen, oder diese nachtraglich an Orten abgehandelt wo sie ihrer Natur nach fremd und storend sind, wie etwa die Bedeutung des Conjunctivs bei der Lehre von den hypothetischen Satzen, etc." This extract will show how deeply Benary had penetrated into these problems. The last clauses sound like an anticipation of some parts of Ries' Was ist Syntax"? (1894). 1 Cf. Paul, Sitz. Ber. d. bayer. Akad. (1894), p. 88; (1897), p. 692. 73 LECTURES ON THE STUDY OF LANGUAGE tutelage of logic. Syntax has been the last of all de- partments of grammar to receive psychological treat- ment; in fact, while since Humboldt's time sporadic attempts to treat syntactical problems psychologically have not been wholly wanting, it is only now beginning to receive it at the hand of the professed syntacticians. Being first, in the time of Bopp, unduly neglected, historical syntax, which was inaugurated by Lange (1852), l and comparative syntax, begun by Delbriick and Windisch (1871), 2 have claimed the attention of scholars ; and even Paul, who perhaps more than any one else has been instrumental in gaining recognition for the psychological method 3 among philologists, had no chap- ter on syntax in the first edition (1880) of his Principien, a lacuna filled in the second edition (1886) by two im- portant chapters, while Ziemer 4 had meanwhile (1882) 1 In the very important address before the meeting of German philo- logists at Gottingen in 1852 (Verhandl. der XIII. Versammluug deut. Philol. etc. [1853], p. 96= Kleine Schriften, I, p. 39.) 2 Syntaktische Forschungen (1871-88) in five volumes, forerunners, as it were, to Delbriick's Vergleichende Syntax (1893 ff.). 8 The third [improved] edition of Paul's Principien appeared in 1898 ; an English translation was made by H. A. Strong from the second German edition in 1889. This is out of print. It contains a preface by B. I. Wheeler. The " Introduction to the Study of the History of Language by H. A. Strong, W. S. Logeman, and B. I. Wheeler" (1891) is, according to the preface, "an attempt ... to enable students to grasp the main points of the contents of one of the most important philological works which have been published during the last ten or twenty years. Paul's Principien der Sprachgeschichte. Wundt's Volker psychologic : Die Sprache (1900), in two volumes, is now the most important contribution to the psychological study of language. Cf. also Wegener's Untersuch- ungen liber die Grundfragen des Sprachlebens (1885, an elaboration of two lectures delivered in 1883 and 1884 respectively) ; and Reichel's Sprachpsychologische Studien, 1897. (1. Die deutsche Wortstellung in der Gegenwart. 2. Die deutsche Betonung in der Gegenwart. 3. Spar- samkeit. 4. Begriindung der Normalsprache.) * Junggrammatische Streifziige im Gebiete der Syntax (1882), an en- largement of his ' Programm' (Colberg, 1879). 74 made the psychological element in the formation of syn- tactical constructions the topic of a short monograph. To any one looking over the syntactical literature since then, it will become at once apparent what an important and even dominating part the psychological method is destined to play in further syntactical investigations. 1 38 Aside from the psychological aspect of linguistic phe- nomena the chief importance of the essay with which Steinthal and Lazarus opened their Zeitschrift fur Volkerpsychologie und Sprachwissenschaft (1860) lies in recognizing the fact that the psychical phenomena which manifest themselves in speech are not wholly identical with those exhibited by the single individual, and in grouping them together with the other similar psychical phenomena, namely, those of belief and custom. True, there are earlier passages which lay stress on the social character of speech, as when Humboldt 2 said, in 1821, that " speech is not a free product of the individ- ual, but always belongs to the nation as a whole," and elsewhere speaks of "the weakness of the individual against the power of language ; " 3 moreover the ground for such views had been prepared by the collectivism of the Romanticists to which allusion has been made above (p. 56). But a definite statement we first find in Stein- thai 4 (1855) : " In our discussion of speech and grammar ... we have never left the domain of psychology. 1 Cf. e. g. Brugmann, Griech. Gramm. 8 1900, p. 364, note 1. 2 Ges. Werke, III, p. 260. 8 Ges. Werke, VI, p. 65 : " Wenn man bedenkt, wie anf die jedesmalige Generation in einem Volke alles dasjenige bindend einwirkt, was die Sprache desselben alle vorigen Jabrbunderte hindurch erfahren hat, und wie damit nur die Kraft der einzeluen Generation in Beriihrnng tritt, und diese nicht einmal rein ... so wird klar wie gering eigentlich die Kraft des einzelnen gegen die Macbt der Sprache ist." * Grammatik, Logik und Psychologie (1855), p. 388. 75 LECTURES ON THE STUDY OF LANGUAGE Nor need we do so in discussing the differences of languages. We only leave one department of it, to which to-day, to be sure, psychology is still confined, and pass into another, which belongs no less to psy- chology, although it has until now only been touched very casually. For the psychology of to-day is individ- ual, i. e., its object is the individual soul as it manifests itself generally in every being which possesses a soul, in every man, and, to a certain extent, also in the animal. Now it is the important fate of the human soul not to exist as an independent individual, but to exist in a member of some community, who from the very begin- ning, both in body and soul, forms part of some people (VolK). And for this reason individual psychology strongly demands a supplement, namely, social psychol- ogy (Volkerpzydiologie).^ By birth every man belongs to some society which materially influences his psychical development. So that the individual cannot be fully comprehended without reference to the community within which he was born and lives." This supplement to the psychology of the individual (without which it must remain one-sided) cannot, however, be found in simple additions dealing with the relation of the individ- ual to the community, but it demands that the commu- nity as such and contrasted with the individual should be made the subject of investigation. 2 "For within the 1 In the next paragraph he points out that it is hardly possible as yet to speak of " social psychology " as a new science on account of the scantiness of material, and he refers to a few stray suggestions in Her- bart's Pyschologie (Introduction), Carl Hitter's Erdkunde (I, p. 19), and an essay by Lazarus in the Deutsches Museum for 1851. 2 Zt. f. Volkerpsychol. u. Sprachw., I (1860), p. 5. After quoting Her- bart (Lehrb. z. Psychol., 2 240) that all psychology which considers man as isolated is one-sided, he continues : " Die Sache ist nun aber damit nicht abgethan, dass man diese Einseitigkeit hinterher durch gewisse 76 HISTORICAL INTRODUCTION community there arises a peculiar set of psychical phe- nomena, which do not really concern man as an individ- ual and do not emanate from him as such. We deal here not so much with conditions in man as with condi- tions between men ; with phenomena in which he is not directly but indirectly concerned, because he is a part of the whole which experiences them. In brief, we deal with the collective soul which is not identical with the sum total of all the individual souls that make up a social group." It must be stated here, once for all, that neither Lazarus nor Steinthal has for a moment assumed that there existed a substantial substratum for their social psyche. In fact, they took pains to forestall the very argument which Paul l urges against them along this line: "At first glance the use of the term psychol- ogy for the phenomena of the life of a social body . . . might be open to criticism for the reason that we cannot imagine a psyche (in the proper sense) of a social body, and that, for this reason, the substance which must be assumed as the substratum of the psychical activity appears to be wanting. But on closer inspection it will readily be seen that a knowledge of the soul, Zusatze, dnrch eine gewisse Rucksicht anf die Verhaltnisse des Menschen in der Gesellschaft, zu erganzen sucht ; sondern diese Erganzung ist iiber- haupt nur erst dann moglich, wenn zuvor der Mensch als gesellschaftliches Wesen, d. h. wenn die menschliche Gesellschaft, also ein ganz anderer Gegenstand als der einzelne Mensch, zum Gegenstande einer besonderen Untersnchung gemacht ist. Denn innerhalb des Menschen- Vereines treten ganz eigenthiimliche psychologische Verhaltnisse, Ereignisse und Scho- pfungen hervor, welche gar nicht den Menschen als Einzelnen betreffen, nicht von ihm als solchen ausgehen. Es sind nicht mehr sowohl Verhalt- nisse in Menschen, als zwischen Menschen ; es sind Schicksale, denen er nicht nnmittelbar nnterliegt, sondern nur mittelbar, weil er zu einem Ganzen gehort, welches dieselben erfahrt. Knrz, es handelt sich nm den Geist einer Gesammtheit, der noch verschieden ist von alien zu derselben gehorenden einzelnen Geistern nnd der sie alle beherracht." 1 Principien (3d ed.), 6, especially p. 11. 77 LECTURES ON THE STUDY OF LANGUAGE namely, of its substance and quality, is by no means the aim or even the chief task of psychology. This consists in the study of the psychical processes and of their de- velopment, i. e., in the discovery of laws according to which the psychical activity of man . . . takes place." 1 And they proceed then to divide psychology into two parts, namely, " Seelenlehre, " which has to do with the substance and quality of the psyche, and " Geisteslehre, " 1 Zt. f. Volkerps. and Sprachw., I (1860), p. 27 f. (under the caption : "Der Volksgeist keine Substanz , aber ein Subject"). "Zunachst konnte die Anwendung des Begriffes der Psychologie auf das Volker- leben, d. h. die Griindung einer solchen Wissenschaft, Zweifel gegen sich dadurch erregen, dass, well eine Psyche des Volkes im eigentlichen Sinne des Wortes undenkbar ist, die Substanz, welche als Trager der Thiitigkeit gedacht werden muss, zu fehlen scheint. Fassen wir aber die Sache naher ins Auge, so leuchtet bald ein, dass die Erkenntniss der Seele, d. h. der Substanz nnd Qualitat derselben, keineswegs das Ziel oder auch nur das Wesentliche der Aufgabe ist, welche die Psychologie zu losen hat. Vielmehr besteht diese wesentlich in der Darstellung des psychischen Processes und Progresses, also in der Entdeckung der Gesetze, nach denen jede innere Thiitigkeit des Menschen . . . vor sich geht, uud in der Auf- findung der Ursachen und Bedingungen jedes Fortschrittes und jeder Erhebung in dieser Thatigkeit. Wir konnten deshalb, da man in unserer Sprache fast allgemein und sicher den Unterschied zwischen Seele und Geist darin begreift, das jene eine Substanz, ein reales Etwas, dieser aber mehr die blosse Thatigkeit bedeutet die Psychologie in Seelenlehre und Geisteslehre unterscheiden, so dass jene, welche mehr das Wesen oder die Substanz und Qnalitat der Seele an sich betrachtet, eigentlich einen Theil der Metaphysik oder Naturphilosophie, diese aber (die Geisteslehre), welche die Thatigkeiten der Seele und deren Gesetze betrachtet, die eigentliche Psychologie ausmacht. Demgemass ist leicht ersichtlich, wie von einer Volkerpsychologie, analog der individuellen Psychologie, die Rede sein kann : namlich als Volksgeistlehre in dem eben bezeichneten engeren Sinne." Could anything be plainer ? Similar and equally defi- nite is the statement in Philologie, Geschichte und Psychologie (1864), p. 37, " Da es keine substantielle Volksseele giebt, sondern der Trager des Volksgeistes nur die zum bestimmten Volke gehorigen Individuen sind, etc." A reference to Waitz (Anthropologie, I, p. 388) is added : " Was als die Begabnng and Entwicklung eines Volkes erscheint, ist der Hauptsache nach bedingt von der Wechselwirkung der Individuen." 78 HISTORICAL INTRODUCTION which investigates the psychical activity, its manifesta- tions and their laws. To the former they assign a place in metaphysics, while the latter comprises the "real" (or, as we might better say, " empirical ") psychology. Wundt 1 has called attention to this advance of Lazarus and Steinthal beyond their teacher Herbart, and pointed out how in this passage they have practically outlined the position of modern psychology, which, as an empir- ical 2 science, has to do with psychical states and pro- cesses alone, while questions relating to a soul-substance come before the forum of metaphysics. As soon, there- fore, as we confine ourselves to empirical psychology and to the investigation of psychical states and processes (which, of course, never occur without some physical substratum), the question is not whether there exists a soul-substance in either individual or social body, but the vital question is whether there are certain psychical phenomena for the occurrence of which the association 1 Ueber Ziele nnd Wege der Volkerpsychologie (Philosoph. Studien, IV, p. 1 ff.) ; cf . now also Volkerpsychologie : Die Sprache, I, p. 17. It seems to me that the argumentation of Paul in the introduction of his Principien is fully met by Wundt's exposition in the two papers just cited. They do not only discuss and rectify some important points in Lazarus and Stein- thal's program (prefixed to the first volume of their Zt. f. Volkerpsych. u. Sprachwiss. ) , but they do the same for certain parts of Paul's Introduction, notably regarding his transcendental psychology ( 4, p. 6) and his con- ception of psychology as a normative science ( 1, pp. 2, 3) which makes a " Principienlehre " necessary. Wundt's articles furnish altogether the clearest exposition and criticism of all controversial points to which the assumption of a " Volkerpsychologie " may give rise. 2 For the " Actualitatsbegriff der Seele " in empirical psychology, cf. e. g. A. Hofler's Psychologic (1897), 1, p. 1 : " Gegenstand der Psycholo- gie : die psychologischen Erscheinungen. . . . Mit den Namen ' Seele,' ' Psyche,' verbindet der gegenwartige Sprachgebrauch bald die Bedeu- tung eines Inbfgriffs psychischer Erscheinuugen, bald die eines 'Tragers' psychischer Erscheinungen. Letzterer Begriff, der der ' Seelen-Substanz,' gehort der metaphysischen Psychologic an." 79 LECTURES ON THE STUDY OF LANGUAGE of a number of individuals is essential, and which must be contrasted, therefore, with those psychological phe- nomena for which this is not the case. It is not denied that the former (no less than the latter) must always manifest themselves in an individual ; all that is main- tained is, that they are not products of one individual but of many individuals, working together in uninten- tional social co-operation. The creative power of a social unit is not equal to the sum of the creative power of each of its members taken separately, but exceeds it. Concerning intentional literary collaboration, Brander Matthews l once wrote that " when two men have worked together honestly and heartily in the inventing, the developing, the constructing, the writing, and the revis- ing of a book or a play, it is often impossible for either partner to pick out his own share; certain things he may recognize as his own, and certain other things he may credit frankly to his ally; but the rest was the result of the collaboration itself, contributed by both parties together and not by either separately." Exactly the same takes place in all unintentional social co- operation, because all the other members of a community may share at any time the new invention (purposely or unwittingly made) of one of their number. 39 The problems of linguistic science present, therefore, two sides, one dealing with the phonetic and semantic development of speech in the individual, the other with the manner in which forms and meanings spread over a definite area and are accepted by a certain community. Of these two aspects the former has received fuller treatment than the latter, because our psychology has been pre-eminently a psychology of the individual rather than of social bodies. But inasmuch as every 1 With my Friends (1891), p. 2. 80 HISTORICAL INTRODUCTION social phenomenon owes its existence not only to indi- vidual creation but also to communal acceptance (in fact, it does not become a social phenomenon except by such acceptance), it can easily be seen that a proper valuation and understanding of linguistic facts can only be ob- tained by treating them as social phenomena. There are especially two points which must be viewed in this light. The first is the uniformity and regularity which . is so prominent a factor in language. 1 This uniformity is not due to the fact that many independent individuals simultaneously chance to coincide in certain innova- tions, but it is the result of social imitation or sugges- tion by which some individual innovations are contin- ually gaining social currency while the majority fail of acceptance. In the discussion of phonetic law this prob- lem will be more fully treated. The second point is the freedom of the individual in his use of language and the influence of language on the individual's mental economy. The individual does not create his lan- guage, but he receives in childhood a ready-made set of symbols which he must henceforth use as best he can. And in gradually appropriating these definite symbols during the formative period of his mental life, they are used as a supporting trellis around which the latter grows up. The forms of every language represent cer- tain characteristic groups of associations, relations, emo- tions ; and the child, in learning to use them intelligently, is forced to arrange his mental contents in the same groups in which preceding generations arranged theirs. For this reason language serves as the most important assimilative factor by which minds of new generations are forced into uniformity with those of their ancestors. The social value of language lies in this fact, that it 1 Cf. the discussion of this point by Wundt, Philos. Studien, IV, p. 25 f. 6 81 LECTURES ON THE STUDY OF LANGUAGE makes psychical heredity possible. Through physiolog- ical heredity ancestral traits can be transmitted to those succeeding generations only which are connected by ties of blood. Through language it is possible for pre- vious generations to affect the mental economy of members of later generations with whom they have no physical connection whatever. In this sense language represents first and foremost communal thought, and it becomes only secondarily the vehicle for individual thought, which may often find the limitations of lan- guage irksome and the set of symbols which it provides inadequate for its purpose. This conflict between the individual demands and the communal means of expres- sion leads to constant minute semantic changes in the use of the old material in order to adapt it to new usage. But the very fact that each innovation must await the approval and acceptance of the community delays all linguistic change and causes it to proceed by almost im- perceptible steps. 1 40 Linguistic science, dealing with the dynamic prob- lems of language, presupposes the data of historical grammar. The historical aspect of the facts of grammar is as essential as the historical aspect in every depart- ment of civilization, but it is not final, because it offers no clew regarding the connection of the successive facts which it chronicles. Their explanation can only be furnished by psychology. It is the purpose of linguis- tics to resolve the highly complicated phenomena of language into their component elements and thus to cor- relate them with the simple psychical phenomena which form the subject of psychological investigation. The particular linguistic phenomenon finds, then, its expla- nation in the general psychological law. There are no 1 Cf. Jodl, Lehrb. d. Psychol., p. 593. 82 HISTORICAL INTRODUCTION specific laws for language, any more than there are specific laws of art or mythology. But there are general psychological laws which have been deduced from the observation of kindred phenomena of the various depart- ments of intellectual manifestation, and in these an ex- planation must be sought for each particular fact. As Scherer very truly said in a discussion of the science of history : l " History is the science of the life of nations. . . . Nations are the primary objects of our observa- tion, and observation is the first step toward the discov- ery of laws. The whole national life must be divided into different sections, and the phenomena within each of these must be studied. Classifications of these phe- nomena, and descriptions of each class, genus, and spe- cies, mark the beginnings of our investigation. Ques- tions for their causes and effects will of necessity lead to a union of the various departments of national life. ... A demand for an explanation of these effects will finally force the investigator to appeal to psychology in order to solve the final problems." Linguistics is the psychological study of the facts of language, as the science of religion is the psychological study of the facts of communal belief, and sociology the psychological study of the facts of communal institutions. So conceived, linguistics is not a separate science, to be contrasted on the one hand with psychology, on the other with descrip- tive historical grammar, but it forms part of the general field of psychology. Certain facts, either of language, or of belief, or of communal life, present themselves to the observer. To chronicle these and to give them a preliminary classification is the first step of their scien- tific investigation. To explain their sequence is the 1 Zt. f. d. oster Gymnas., XVII (1866), p. 264 = Kleine Schriften, I, p. 170. 83 LECTURES ON THE STUDY OF LANGUAGE second step, and, since they are all essentially psychical phenomena, this is the province of psychology. Psychol- ogy is the clearing house into which all individual obser- vations made within a limited field are finally passed in order that similar phenomena of different departments may be assembled and correlated. Only through this psychological analysis and the comparison of like ele- ments in the different manifestations of intellectual life the individual fact is seen in its proper perspective and receives its adequate setting and explanation. 41 The study of the spoken language will, therefore, derive much help from being combined with the study of kindred phenomena. Of these the language of chil- dren has received the greatest attention, upon the prin- ciple that as in physiology so here the ontogenetic development in the individual might be considered as an abbreviated repetition of the phylogenetic development in the race. 1 This principle, however, cannot be trans- ferred directly from the physiological facts of embry- ology to the psychical facts of linguistic development, because language in the child never develops freely, but its natural growth is continually interfered with. When Ament, for instance, speaks of infantile sound changes, he compares the transformations of Louise's first "word" mammamm to m6mi and finally to mdma with the sound changes which in the course of thou- sands of years transform the words of a people. But these two transformations have really nothing whatever to do with each other. For Louise's changes are simply due to the clearer perception and more successful imita- tion of the same model which her nurse continued to speak before her. The three forms are successive stages 1 Ament, Die Entwicklung von Sprechen und Denken beim Kinde (1899). This contains a good historical introduction and bibliography. 84 HISTORICAL INTRODUCTION in an attempt to imitate a sound, comparable to the different forms by which a foreign word may be repre- sented in the native idiom. 1 Gesture language differs from both speaking and writing, in that in it the movements themselves serve as symbols and convey meaning, while in speaking and writing different results of such movements (namely, resultant sounds or tracings) form the vehicle of com- munication. The fullest discussion of gesture lan- guages and their bearing on speech is now found in Wundt's Vdlkerpsychologie. 2 Writing, on the other hand, has received but very scanty treatment, although it is in most essential respects an exact parallel to speak- ing. 3 In both cases a special motor centre is charged with directing the movements, here of the vocal organs, there of the hand. In both cases these movements are only indirectly concerned in the conveyance of thought. It is therefore to be expected that writing and speaking should show many points of similarity. As a matter of fact the form of handwriting is as characteristic for different nations as is their idiom. 4 What Preyer calls 5 the famity type of handwriting is comparable to the family type of speech. 6 As in speaking, so in writing, there is a certain margin within which variations are permitted and unheeded by the hearer or reader, and these variations occur not only in different individuals, but often in the same individual. Preyer discusses at 1 This was written before I saw Wundt's criticism of Ament's position, Volkerpsychologie, Die Sprache, I, p. 296, with the note. 2 Die Sprache, I, p. 131 (with literature). 8 Preyer, Zur Psychologic des Schreibens (1895), p. 38 : "Das Schreiben selbst ist im bnchstablichen Sinne eine Art Fingersprache." 4 Preyer, Zur Psychologie des Schreibens (1895), p. 2. 6 Ibid., p. 3. 6 Passy, in Phonet. Stud., I, p. 19-20. 85 LECTURES ON THE STUDY OF LANGUAGE some length the use of such polymorphous letters (as e and e) by the same writers. 1 Like the phonetic varia- tions, these graphic changes are due to cerebral causes rather than to the writing material or the muscles. 2 Preyer even alludes to cases of graphic mixture, 3 in which, by a kind of associative interference, those, for instance, who work intently and for a long time with figures (like mathematicians) assimilate the form of their letters to the form of somewhat similar figures (as B to 13, gb to 96, etc.). In view of the fact that the movements of the hand in writing are not only less complicated than the movements of the vocal organs in speaking, but also more easily registered, observations touching the changes in the individual's mode of writ- ing as well as in the forms of letters of successive periods should be made which would throw valuable light on similar phonetic variations. 1 Preyer, Zur Psych, d. Schreib., p. 100. 2 Ibid., p. 33-37. 8 Ibid., p. 133. 86 LECTURE II ON THE CLASSIFICATION OF SIMILAR SPEECH INTO DIALECTS AND LANGUAGES AND ON THE NATURE OF INFERRED PARENT LAN- GUAGES NORMATIVE or didactic grammar sets up a certain standard as correct. This standard 1 is obtained partly by philosophical, chiefly logical, considerations as to the manner in which language ought to meet adequately the demands made upon it, partly by eliminating geographi- cal differences among the "natural" speakers and unit- ing what is common to most of them, partly by recog- nizing some one geographical area and its speakers as a model which the rest should imitate, and partly by meas- uring the correctness of current speech by the standard of a more or less arbitrarily chosen past period, often termed "classical." The first of these methods is well illustrated by the distinction of the Greek grammarians between eXXTjwcr/io?, when denned as equivalent to op66rr]