r':'^:*^, %^M:.M-^ . ' V d^r S^s'iJ iV' _*• %t^ frnM'* ■J. %.' ■ ■ i ^- ■ '-'-'if'-:. ■afJ ■im^'-i *f U^W^riHl- .jS> fSUi ? ► The date shows when this volume was taken. To, renew this book copy the call No. and give to - the librarian^ li.My'09 22UyV9 ffii- fii1ViBl*iilHijyfafMIPrT JUL U 2006 HOME USE RULES/ All Books subject to Recall. Books not used for instruction or research are returnable within 4 weeks. Volumes of periodi- cals and of fjairiphlets are held in flie library ag^much as possible. I^or special purposes they aye given out for a lirdited time. Bbirrowers should _not use their library "^Idleges f pr thi^bene- ^^^^^^jg^gohs. "Books not needed during recess periods should' be returned to the library, or arrange- .ments made for their return during borrp^- er's absence.if wanted. Boo'ks needed "by niore than, one person are held on the reserve list. Books of special value and •gif^'bpQks, when' the giver wishes it, are not allowed to circulate. Readers are asked to report all cases "of books marked or mttti- lated, " - ■ Do not deface books by marks and writing. Cornell University Library P 543.D34I6 1882 Introduction to the study of language: 3 1924 026 454 854 Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/cletails/cu31924026454854 BIBLIOTHEK INDOGERMAmSCHER GfRAMMATIKEN BEARBEITET VON F, BUOHELER, B. DELBEUOK, E. POY, H. HTJBSOHMANN, A. LESKIEN, G. METEK, E. SIEYEES, H. WEBEE, ¥. D. WHITNEY, E. WINDISOH. BAND IV. INTRODUCTION TO THE STUDY OF LANGUAGE : A CRITICAL SURVEY OF THE HISTORY AND METHODS OF COMPARATIVE PHILOLOGY OF THE INDO-EUROPEAN LANGUAGES. HY B. Delbruck. LEIPZIG, DRUCK UND VERLAG VON BREITKOPF & HARTEL. 1882. INTRODUCTION TO THE STUDY OF LANGUAGE: A CRITICAL SURVEY OF THE HISTORY AND METHODS OF COMPARATIVE PHILOLOGY OF THE INDO- EUROPEAN LANGUAGES. BY B. DELBRdCK. ATITHOEIZED TEANSLATION, WITH A PBEFACE BY THE AUTHOR. THIS WORK IS COPYBISHT. LEIPZIG, BREITKOPF AND HABTEL. LONDON, TRtfBNER & CO. 57 AND 59, LUDGATE HILL, E. C. 8. ENTD STA. HALL. 1882. Entered according to Act of Congress, in the year 1882, by E. Channing in the office of the Librarian of Congress at Washington XC C. ^OR UNIVERSITY V LIBRARY z/ ^ J-.U. J l ■J■L■■■^■L I ■ J. ' L I i' I » M — — 1^ f ikri PREFACE. Ihe character of the present work is mainly determined by the circumstance that it is intended by the author to facili- tate the study of the "Grammars" which Breitkopf & Hartel are publishing, as well as the comprehension of comparative philology in its newest form. The field of this "Introduction" is no broader than that of the above-mentioned "Grammars". Wherever in the follow- ing pages language, language-development, phonetic laws etc. are discussed, Indo-European language, Indo-European lan- guage-development etc., must alone be understood. I have felt the less inclination to discuss questions which lie outside the Indo-European domain, as for example those suggested by universal philology, because in reality the influence of phi- losophical linguistic research upon the science founded by Bopp has always been of slight account, and is very trifling at present. In limiting myself to the departments of phonetics and inflection I have also followed the "Grammars", but I must acknowledge that I should not, perhaps, have practised this self-denial if I had not just shown, in the fourth volume of my syntactical investigations [Die Grundlagen der griechi- schen Syntax, Halle, 1879), how in my opinion the compara- tive syntax of the Indo-European languages should be treated. The book here laid before the public is divided into a his- torical and an analytical part. In the former the development of philology from Bopp's time to the present is roughly sketched. VI Pkeface. with the especial aim of showing clearly how the problems which particularly occupy our attention today have developed naturally from what preceded them ; in the second part the most important of these problems are discussed. In the fifth chapter (Agglutination Theory) I have expressed myself with more skep- ticism and reserve than formerly. But I have thought it best not to suppress my present views, since they are the result of mature deliberation, and I cannot believe that I shall ever have a more decided opinion on these points. In the sixth chapter (Phonetic Laws) I hope I have shown that the harmony between the re- presentatives of different tendencies is greater than the uninitiat- ed would be inclined to suppose, possibly greater than is recog- nized by many of these representatives themselves. All these discussions, as well as the historical part (chapters I — IV) and the seventh chapter (on the Separation of the Eaces) , are not written for the narrow circle of my fellow-workers in this field, but mainly for those who do not make a special study of comparative philology. For this reason I have made use, so far as I was able , of the exceedingly abundant philological literature, but have quoted it only to a small extent. In conclusion, I would like to say a word respecting my judgment of living scholars. In characterizing great men who are deceased , like Bopp and Schleicher, I have imposed no restraint upon myself, but to proceed in the same way vnth reference to the living seemed to me unfitting. I have accord- ingly never attempted to picture in full the distinguishing characteristics and achievements of any living philologist, while I have not hesitated to define my position in regard to single views of my contemporaries. To this point I would fain call the attention of those readers who perhaps might be in- clined to think that I have not done justice to this or that eminent living philologist. Jena. August 20, 1880. B. Delbruck. AUTHOR'S PREFACE TO THE ENGLISH TRANSLATION. In writing the present work, my only thought was to in- troduce some of my countrymen to the study of the Indo-Eu- ropean languages, and to call their attention to certain points in the history and present condition of this study. Accordingly, the preface, as well as my treatment of the subject, were de- stined for German readers. On this account it seems to me especially importantthat the English translation should be pre- ceded by a few words of explanation, or, I might almost say, of apology. Above all, I would deprecate any comparison of this short and unpretentious "Introduction" with the extensive works of Max MiJLLER, Whitney und Sayce. I by no means aim to furnish an explanation of the nature of language in general, as these scholars have done ; it is only my wish to show how a minute section of the whole linguistic mass has been studied, and in my opinion ought to be studied. It is true that in one chapter, that which treats of agglutination, I touch on more general ground, and it would perhaps have been better if here I had at least attempted to discuss my disagreement with Sayce's views somewhat in detail. I hope that in case of a second edition I shall be able to extend my original plan in this direction. VIII Author's Preface to the English Translation. Perhaps it may seem to the reader that not only the above- named English and American philologists have received insuf- ficient consideration, hut that the same is true of other for- eigners, as for instance those exact scholars of whose views the "Memoires de la soci6te de linguistique" may be consid- ered the exponent. I cannot wholly deny the justice of this criticism, but the following considerations will serve to ex- plain my course. This book was written with the aim (how far it has been attained, I will leave others to judge) of con- tributing to the history of the German mind. It is universally acknowledged, by those who have traced the history of Ger- man development, that there is an immense gulf between the views of the Germans of today and those prevalent up to the fourth or fifth decennium of this century. This difference of view is almost as great in scientific fields as in the domain of politics. One side of this mighty revolution can be concisely expressed in the statement that we have passed from a philo- sophical epoch into a historical one. I attempted to show (as no one to my knowledge had done before) that the science founded by Bopp stands in evident connection with the philo- sophical endeavors of German scholars , and also how it has come about that in linguistic science a sort of metaphysics has arisen, which is at present undergoing a process of dissolution. But at the same time I wished at least to intimate that it is wrong to undervalue endeavors of this nature, since the occa- sion for such investigations is found in the linguistic material itself, and will probably continue in the future. I would beg that my estimate of linguistic science and of the great philologists may be judged from this more general stand-point. It was not my intention to write a glorification of linguistic science, but to contribute toward a just estimate of it. My position with regard to the great philologists Bopp, Gkimm and others is as untrammeled as that we occupy to- Author's Preface to the English Translation. IX ward Shakespeare and Goethe. If a historian of literature asks whether Goethe possessed dramatic talent in the high- est and truest sense, no one will charge him with lack of reverence, but it will be recognized that he has only done his duty in proposing and answering this question. In the same sense I claim for myself the right to investigate what constitutes the actual power of that richly -endowed master to whom we owe the foundation of our science. Whoever reads with unprejudiced mind my sketch of Bopp and Schleicher will, I hope, be impressed with the fact that my pen was guided by both love of truth and a feeling of veneration for these great men. With the above remarks I would commend this book to the kind indulgence of the English and American public. TflANSLATOE'S PREFACE. When I began to translate this little book, shortly after its publication, I did not anticipate the various delays and inter- ruptions which have postponed the completion of my task for a whole year. So long a delay might be fatal to the useful- ness of a translation, in the case of a work which aimed at a systematic exposition of the whole science of comparative phi- lology, down to the latest development of its smallest detail, and the discussion of all disputed questions relative to both method and practice. A treatise with such an aim in view would require constant revision and extension, and would be completely antiquated in the course of a year. The present work, however, only proposes to exhibit the historical devel- opment of the science , and while discussing the chief prob- lems which now present themselves in this field, it does not claim to chronicle all the various attempts to solve them, nor to initiate the reader into the intricate details of a philological warfare which is today raging at its hottest. Therefore this little volume may be said to fulfil its avowed purpose as well at present as it did a year ago. I was impelled to undertake this translation by the con- sideration that I had never found a book which appeared to me to give so clear und succinct an account of the rise and devel- opment of comparative philology in Germany. It therefore Translatok's Peeface. XI seemed particularly desirable that this condensed sketch of the progress in linguistic methods should be made accessible to those who are not conversant with German philological liter- ature , more especially to those who are desirous of subse- quently devoting more attention to the subject. If this trans- lation shall serve to awaken or strengthen an interest in the science which owes its origin as well as many of its most able representatives to Germany, I shall feel amply repaid for any time and labor spent upon it. A few words will suffice in explanation of the method I have pursued. My first aim was to render the sense of the German with the utmost possible accuracy, so that if I have erred, it has been on the side of too close adherence to the text of the original. In those instances where trivial altera- tions have been made, this has occurred with the knowledge and consent of the author , and the same is true of the very few notes I have ventured to add , which are always desig- nated as the translator's. No one who has not made a similar attempt can realize the peculiar difficulties of transferring the German philological nomenclature to the English tongue, where certain of the technical terms, it is true, already have their recognized equivalents, but others are either differently rendered by different scholars, or are not represented at all in the language. In many cases where an important term could not be adequately translated, I have thought it only fair to in- troduce the German word in brackets. The titles of the Ger- man works quoted I have thought it more advisable to repeat in their original shape, since few of these works are translat- ed, and for purposes of reference the German title would be necessary. Whenever Prof. Whitney has been quoted, I have naturally referred to Ms own book, instead of to Jolly's Ger- man translation, and similarly, the extracts from Bopp's Ana- lytical Comparison appear in their original English form, as XII Teanslatoe's Preface. well as the remarks of Sir William Jones on page 1 . Two or three of the longer sentences from Curtius' Grundzuge have been quoted in the English translation, in which case the page of the translation has been added in square brackets. In my transcription of Sanskrit words I have adopted the method re- commended by Prof. Whitney in the "Proceedings of the Amer- ican Oriental Society" for Oct. 1880, and used by him in his Sanskrit Grammar, which forms the second volume of this se- ries; for Zend, Hubschmann's transcription, advocated by him in KuhrCs ZeitscJirift, 24, page 328 seq., has been em- ployed. In conclusion, I must express my heartiest gratitude to Prof. Delbeuck for the cordial sanction he has given to my undertaking, and above all, to Prof. Sievers, who was so kind as to read over the whole translation, and to offer many valuable hints and suggestions. Leipzig, Dec. 1881. E. Channing. CONTENTS. I. HISTOEICAL PART, Chapters I— IV. CHAPTER I : Fbanz Bopp, pages 1—26. 1. Bopp's views of the origin of Inflection , 3 — 16. Bopp at first follows Fkiedkioh Schleobl. Theory of the latter, 3 — 5. Bopp's theory in its first form, 6 — 9. Bopp's theory in its second form, 9 — 12. The third and final form, 12—15. 2. Bopp's method of comparing given languages, 16 — 25. Is Bopp's mode of view that of natural science? His general views of language, 17 — 20. Bopp's conception of phonetic laws : a) mechanical laws (i. e. law of gravity), 20 — 21 ; h) physical laws, 21 — 23. Incompleteness of Bopp's system of phonetics, — especially evident in his treatment of non-Indo-European languages, 23 — 24. Comprehensive estimate of Bopp, 25 — 26. CHAPTER II: Bopp's Contbmpokabibs and Svoobssors down to August Scbieiohbb, pages 26 — 40. Wilhblm von Humboldt, 26 — ^28. A. W. von Schlbgbi, 28 — 32. Schlegbl's position toward Bopp , 29. Lassen's critique , 30 — 31. Jacob Gkimm, 32 — 34. A. F. Pott, 35 — 36. Theodob Bbnpby, 36. Development of linguistic science down to Schleichbb, 37 — 40. Progress in knowledge, 37—39; in method, 39—40. CHAPTER III : August Sohleichee, pages 40—55. Hbgbl's influence, 40 — 42. Influence of natural science, 42 — 44. Schlbi- chbb's works , 44 — 45. Comparison hetween his views and Bopp's , 45 — 47. Schlbicheb's parent speech, 48 — 53. Schlbickeb not a scientist, hut a philo- logist, 54 — 55. CHAPTER IV : New Endeavors, pages 55—61. New endeavors are manifest in the following tendencies : 1) Less interest is felt in the history of inflection (i. c. the origin of forms in primitive periods of the parent speech), 56. 2) It is recognized that no composition of unfinished linguistic elements can take place in the individual languages, 57. 3) More rigorous demands are made upon phonetic laws. The view arises that phonetic laws admit of no exceptions, 60. . In consequence of this, attention is especially drawn to formations hy analogy, 60. Importance of modern languages, 61. XIV Contents. II. THEORETICAL PAET, Chapters V— VII. CHAPTER V: The Agglutination Theoky, pages 61—102. Improbability of the other two theories ■which have thus far appeared : a) the evolution theory, in the form given it hy "Westphal , 62 — 66; b) Ltjdwig's adaptation theory, 66 — 71. Bopp's theory supported by the analogy of other languages, 71 — 73. Thus the agglutination theory is recommended in principle. A special discussion follows : 1) roots, 73 — 85; 2) the noun, 85 — 92; 3) the verb, 92 — 99. Can we distinguish different periods in the primitive speech? 100. AscoLi's hypothesis, 100—101. Concluding observations, 101 — 102. CHAPTER VI : Phonetic Laws, pages 102 — 130. Sketch of the views of G, Cuktius. Three notions are prominent : 1) pho- netic laws ; 2) analogy ; 3) the influence of sense on sound. These are dis- cussed in inverse order : 1. Influence of sense on the preservation of sound, 106 — 107. 2. Action of analogy, 107 — 113. Demarcation of the field, 108. Classi- fication, 108—111. Practical problems, 111—112. 3. Phonetic laws, 113 — 130. Definition of the notion "phonetic change" as such, 113 — 115. The regularity of phonetic change cannot be inductively proved, 115 — 117. Hence a solution is sought deductively, 117 — 129. What are the causes of phonetic change ? — according to Cnniius, "Whitney, Ost- HOFF, 117 — 120. More general theory, in agreement with Benpet, 120. From this stand-point is shown : 1) how a language can split into dialects, 121 — 122 ; 2) how great a uniformity can be expected within the bounds of the dialect, 123 — 129. Fluctuations must be assumed for a state of [transition, 123 — 124; but no permanent lack of uniformity in the treatment of sounds, 125 — 129. Comprehensive remarks on the notion and significance of phonetic laws, 129—130. CHAPTER VII : The Sepabation op the Races, pages 130—139. Bopp's view, 131 — 132. Theory of a genealogical tree, 132—134. "Wave- theory", 134—136. Present state of the question, 136—139. INDEX, pages 140—142. CHAPTER I. FEANZ BOPP. When Feanz Bopp (bom in 1791), the founder of com- parative philology, hegan to devote his attention to Sanskrit, the statement that the language of the Brahmans was nearly related to the languages of Europe, especially to Latin and Greek, had been repeatedly made, and strengthened by a number of authentic proofs. Above all. Sir William Jones, the first president of a society organized in Calcutta for the exploration of Asia, had, as early as 1786, expressed himself on this point as follows : " The Sanskrit language, whatever may be its antiquity, is of 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 have been produced by accident ; so strong that no philologer could examine all the 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 Celtic, though blended with a different idiom, had the same origin with the Sanskrit." (Cf. Benfey^ Geschichte der Sprachvjissenschaft, page 348.) In the main coincident with the above, but less correct in one point, are the opening sentences of Friedkich Schlegbl's celebrated book on the language and wisdom of the Hindus [TJeber dieSprache und Weisheit derlndier, Heidelberg, 1808): Delbrdok, Introduction to the Study of Langn^ge. 1 2 Chapter I. "The old Indian Sanskrito ,\i. e. the refined or perfect, also called Gronthon^ i. e. the written or book-languagej bears the closest relationship to the Roman and Greek as well as to the Germanic and Persian languages. The resemblance is found not only in a large number of roots, which it has in common with them, but extends to the innermost structure and gram- mar. The agreement is therefore not an accidental one, such as could be explained through intermixture, but an essential one, which points to a common origin. On comparison it is further discovered that the Indian language is the elder, the others younger and derived from it." We cannot, therefore, say that Bopp was the discoverer of the Indo-European i) community of language, but to him is due the credit of having instituted a systematic comparison, ■which, starting from the forms of the verb, gradually extended over the whole language , and of thus demonstrating for all time what Jones , Schlegbl and others had only suspected or affirmed. This demonstration will, without doubt, be regarded in the future as the epoch-making achievement of Bopp's genius, but it is quite as certain that Bopp himself from the very begin- ning had in view not the comparison, but the explanation of forms, and that comparison was to him only a means to the attainment of this chief end. To illustrate by an example : he was not satisfied with the discovery, so all-important for the phonetics of each individual language, that dsmi, e{|j.i, sum, im, jesrm are all at bottom one and the same form ; but it was of greater interest to him to learn from what elements this form had arisen. Not a comparison of actual forms of speech, but an insight into the origin of inflection was the essential aim of his >work. 1) I have followed Prof. Whitney and others in preferring the term 'Indo-European" to "Indo-Germanie", which latter name cannot in English claim the excuse of preponderating usage alleged by Prof. Delbruck in support of its German equivalent. He says : "I use the name 'indogerma- nisch' (originated by Klaproth?) because, as far as I can see, it is the most common in Germany." The term "Aryan", so frequently employed by English philologists, I have rejected as being more properly applicable to the Indo-Iranian division of the family. [Translator.] Fkanz Bopp. 3 That this is really the case has been abundantly empha- sized by the older as well as the more recent critics of Bopp. It will suffice here to recall the well-known statement of Bopp's teacher Windischmann , namely, that Bopp's aim from the beginning was "to penetrate by way of linguistic investigation into the mystery of the human soul, and to gain some cognizance of its nature and laws"; and to quote a remark of Theodok Benfey : "I would therefore con- sider that the real task of this grand work [the Comparative Grammar] was to gain a knowledge of the origin of the gram- matical forms of the Indo-European languages; that their comparison was only a means to the attainment of this end, merely a method of discovering their fundamental forms ; and that, finally, the investigation of phonetic laws was the chief means of comparison, the only sure foundation for the proof of relationship, especially of the fundamental forms." Under these circumstances it seems to me expedient to speak first of Bopp's view of the origin of inflection, and after- wards to discuss his method of comparison. I. Bopp's views of the origin of inflection. Bopp's theories concerning the genesis of linguistic forms are not, as might be imagined, the pure result of his gram- matical analysis, but can be traced back in great part to older views and prejudices. Among these the theory of Friedkich ScHLEGEL, which is brought forward in his above-mentioned work Ueher die Sprache und Weisheit der Indier, plays an im- portant part. It seems to me necessary, therefore, to famil- iarize the reader with this theory at the outset. According to Feiedkich SchlegeIj there are two chief classes of languages ; first, those which characterize the minor shades of meaning by an inner change of the root, and sec- ondly, those which for this purpose affix actual words hav- ing in themselves the significance of plurality, past time, future obligation, or other comparative notions of the sort. The first class embraces the inflectional languages. Schlegel therefore understands by "inflection" the inner change of the root. He most emphatically opposes the view that the 4 Chapter I. inflectional forms could have been obtained by affixing pre- viously independent words ') : "In Greek there is at least a semblance of possibility that the inflectional syllables might have had their source in particles and auxiliary words which have melted into the word itself, although it would not be possible to carry out this hy- pothesis without having recourse to almost all the etymological artifices and juggleries which should all, without exception, be banished at the outset, if we are to view language and its origin scientifically, i. e. in a thoroughly historic light ; and even then this hypothesis could scarcely be carried out. But in Sanskrit the last semblance of such a possibility vanishes, and we are compelled to admit that the structure of the lan- guage is a thoroughly organic one, ramified by inflections or inner changes and variations of the root in all its significa- tions, and not a simple mechanical compositum formed by the affixion of words and particles, while the root itself re- mains barren and unchanged." (Page 41.) In this organic nature he finds the main advantage of the inflectional languages : "To this is due on one side the wealth, on the other the stability and durability of these languages, which can be said to have arisen organically, and to form an organic tissue ; so that centuries after, in languages which are separated by broad tracts of land, it is often possible vrith little pains to find the thread which extends through the wide -spread wealth of a whole word-family, and leads us back to the simple origin of the first root. On the other hand, in languages which instead of inflection have only affixes, the same cannot be said of the roots ; they are no fruitful seed, only a heap of atoms, as it were, which every chance wind can easily scatter or sweep together ; the connection is really no other than a purely mechanical one, by means of outward affixion. In their first origin these languages lack a germ of living development" etc. (Page 51.) ') Probably in this opposition he has in mind the school of Lemstep and ScHElD (v. below), hardly Horne Tooee (concerning whom cf. Max MiJLLEE, Lectures on the Science of Language, page 255) . Feanz Bopp. 5 If we ask how this explanation of inflection as an inner change of the root, which seems to us so wanting in precision and clearness, can have arisen in the mind of this gifted schol- ar, so much is plain at once, that it was not derived from immediate ohservation (for where could we observe such an organic growth?) ; it seems more prohable that it is really nothing but the necessary logical opposite of the theory which ScHLBGEL felt obliged to reject. In face of the absurdities of Lennep, Scheid & Co., by whom language was most stupidly cut to pieces and forcibly derived from purely imaginative roots, ScHLEGEL had evidently arrived at the conviction that it was impossible to approach the mystery of the development of linguistic forms by means of analysis. He therefore, inj opposition to the theory which explained the origin of language, by composition, preferred to postulate its development by means of organic growth, without very distinctly picturing to himself the nature and causes of this growth. He was per- haps strengthened in this view by another consideration. The relation existing between the Latin and Romanic languages (which his brother afterwards sought to characterize by the expressions "synthetic'' and "analytic") seemed to him the more remarkable from the fact that in Sanskrit he found, so to speak, a more Latin condition of things than in Latin itself. (Page 40.) If, he may have concluded, a language shows the less composition the more ancient it is, how can we sup- pose that the linguistic forms in oldest times originated en- tirely by means of composition'? Now it was quite in the spirit of the philosophers of the Romantic School, with whose train of thought and method of expression Schlegbl was familiar, that he characterized such a growth from within outwards as "organic", and at the same time regarded this organic growth, in comparison with com- position, as the higher and nobler process. Bopp, in his first publication [Conjugationssystem der Sanskritsprache, 1816) adopted fully this briefly -outlined theory of Schlegbl (although without mentioning the author's name) , which he afterwards stoutly opposed. But he extended it at once in one direction, by adding to the criterion of the g Chapter I. inner variation of the root the capacity to incorporate the substantive verb ') : "Among all the languages known to us", he says on page 7, "the sacred language of the Hindus shows itself one of the most capable of expressing the most varied conditions and relations in a truly organic way, by inner inflection and change of the stem- syllable . But in spite of this admirable flexibility, this language is sometimes fond of incorporating into the root the abstract verb, whereupon the stem - syllable and the abstract verb share the grammatical functions of the verb." This division of labor can be observed, for example, in the aorist, in the following manner. In the Sanskrit dgrausam, "I heard"; a characterizes past time ; the especial modification of the past which is peculiar to the aorist is intimated by the strengthening of the u in the root gru to au ; and the substan- tive verb is incorporated into the thus formed preterit, "so that, after the time-relations have been expressed in a purely organic way by inner variation of the root, person and num- ber are defined by inflection of the affixed auxiliary verb." (Page 18.) The incorporation of the substantive verb is sup- posed by Bopp to have taken place in the future and aorist in Sanskrit and Greek, in the Sanskrit precative, in the well- known perfect and imperfect formations of Latin, and (although he afterwards gave this up) in the passive endings of the same language. Bopp recognizes no other composition than that with as in his Conjugationssystem . To be sure he speaks of affixing the "characteristics of person" [ Personskennzeichen] M, S, T, but he does not recognize in these characteristics any re- mains of formerly independent words. On the other hand, he remarks expressly in another connection : " It is contrary to the spirit of the Sanskrit language to express any relation by affixing several letters which can be regarded as an individual word." (Page 30.) In the Conjugationssystem he leaves the origin of these "characteristics of person" just as much in the ') Bopp can have had this method of explanation alone in mind, when he says (Conjugationssystem, page 12) that in his labors he nevex leans upon the authority of another. Fkanz Bopp. 7 dark as the origin of the "interpolated" vowel t, which char- acterizes the optative. It would be interesting to discover what considerations induced Bopp to modify Schlegbl's definition of the idea of inflection. Fortunately there is sufficient material for this in Bopp's writings. But in order to make the passages in question intelligible, I must first say a word about the customary clas- sification of the parts of speech at the beginning of our cen- tury. There was at that time a general prejudice in favor of the theory that the sentence must be an image of the logical judgment; hence arose the opinion that, inasmuch as a judg- ment consists of three parts, subject, predicate and copula, [ the number of the parts of speech also must be neither larger nor smaller than three. It was naturally no easy matter to bring the traditional parts of speech under three heads, and this classification could not be carried out without sophistry. For instance, A. F. Beknhardi knew no better method of re- conciling his philosophical theory with his practical experience than by making out the following table : I. Parts of speech [Redetheile] : a. Substantives. b. Attributives. aa. Adjectives, bb. Participles, cc. Adverbs. c. The verb be. II. Smaller parts of speech [Redetheilchen] : a. Prepositions. b. Conjunctions. c. Original adverbs. III. Parts of speech and smaller parts of speech : Pronouns. GoTTFEiED Hermann is convinced, as well as Beenhaedi, that there can be but three parts of speech , and we find that Bopp was of the same opinion, as most clearly appears 8 Chapter I. from a remark in the English version of his first publication, Analytical Comparison^] etc., page 14 : "Potest unites in itself the three essential parts of speech, t being the subject, es the copula and^o< the attribute." Here it is especially noteworthy that not the verb as such, but only the verb he is regarded as the third part of speech. Gottfried Hermann says [De emendanda ratione graecae grammaticae, Leipzig, 1801, page 173) : "Est enim haec verbi vis, ut praedicatum subjecto tribuat atque adjungat. Hinc facile coUigitur proprie unum tantum- modo esse verbum idque est verbum esse. Caetera enim quae- cunque praeter hoc verbum verba reperiuntur, hanc naturam habent, ut praeterquam quod illud esse contineant, quo fit ut verba sint, adjunctam habeant etiam praedicati alicujus nota- tionem. Sic 'ire' 'stare', ut aliqua certe exempla aiferamus, significat 'euntem, stantem esse'." Bopp shared this opinion, as is sufficiently clear from the first words of his Conjugationssystem, which are as follows ■ "By the word 'verb' in its strictest sense is meant that part of speech which expresses the union of an object with a quality, and their relation to each other. According to this definition, the verb has in itself no actual significance, but is simply the grammatical bond between subject and predicate, through whose inner variation and change of form these mu- tual relations are indicated. In this sense there is but one verb, viz., the abstract verb, Se, esse" etc. Since, therefore, according to Bopp's view no predicate can exist except with the aid of the verb esse, and since, ac- cordingly, this predicate in point of meaning inheres in every so-called verb, to be consistent Bopp would necessarily find it natural that the verb as should be palpably and visibly re- presented in every verbal form. Bopp did actually accept this consequence in a very remarkable sentence in the Analy- tical Comparison, J)age 14: *) Analytical Comparison of the Sanskrit, Greek, Latin and Teutonic Languages, showing the original identity of their grammatical Structure, by F. Bopp, published in the Annals of Oriental Literature, London, 1820. This is not merely a translation, but in many respects quite different from the German original. [Transl.] Fbanz Bopp. 9 "After these observations the reader will not be surprised if, in the languages which we are now comparing, he should meet with other verbs constructed in the same way as potest, or if he should discover that some tenses contain the substan- tive verb, while others have rejected it or perhaps never iised it. He will rather feel inclined to ask : 'Why do not all verbs in all tenses exhibit this compound structure?' — and the absence of the substantive verb he will perhaps consider as a kind of ellipsis." Whoever carefully weighs this extraordinary sentence, in which the solution of a difficulty is ingeniously thrown upon the reader, while he would naturally expect the author to solve it, will certainly agree with me when I assert that Bopp was led to seek the substantive verb in the occasionally appearing s of Indo-European forms, chiefly in consequence of his false theory concerning the three parts of speech. Accordingly, we can characterize Bopp's oldest theory of inflection, as we find it in the Oonjugationssystem, as the union of an apergu of Schlegbl with the traditional theory of the three parts of speech. In the above-mentioned English version (1819), the Ana- lytical Comparison, we find a very marked advance upon the view brought forward in the Conjugationssystem (1816). This progress can be briefly summed up as follows : the principle of composition, which up to this time was only applied in the case of the root as, is now recognized as the prevailing one. How Bopp arrived at this change of opinion can be traced out in his explanation of the notion "root", and in his hypothesis concerning the origin of the personal endings of the verb. First of all, in regard to the notion "root", it was pos- sible for Bopp to derive from the grammatical tradition pre- valent at his time the opinion which he here expressed and retained ever after, namely, that all words go back to mono- I syllabic elements. For Adelung had already declared that all the words of the German had their origin in monosyllabic constituent parts, which bear the name "root". (Cf. Adelung, Ueber den TJrsprung der Sprache und den Bau der Worter, be- 10 Chapter I. sonders des Deutschen, Leipzig, 1781, page 16seq.)i) Bopp found this view confirmed by an investigation of the Sans- krit root -indices, with which he became acquaiated in the edition of Carey and Wilkins. (Cf. A. W. von Schlegel, Indische Bibliothek, 1, 316 and 335.) He formulated his opinion in the Analytical Comparison, page 8, as follows : "The character of Sanskrit roots is not to be determined by the number of letters, but by that of syllables, of which they contain only one ; they are all monosyllabic, a few ex- cepted, which may justly be suspected of not being primitives." (Cf. also A. W. von Schlegel in the above article , page 336.) Now Bopp assumed for the roots of the kindred languages what was true of the Sanskrit roots, and accordingly made the statement : "Roots are monosyllables in Sanskrit and its kin- dred languages." With this conception of "root", Schlegel's idea of inflec- tion must naturally appear very questionable. For how can a monosyllabic root (especially if, as is obviously the case, the consonants remain intact) be inwardly inflected and al- tered to any considerable extent? The idea of the mono- syllabic nature of the root must necessarily strengthen that of composition in inflection , and it is therefore not surprising that Bopp's polemic against Schlegel had its starting-point just here. We find this polemic expressed in the following paragraph, page 10 : "If we can draw any conclusion from the fact that roots are monosyllables in Sanskrit and its kindred languages, it is this, that such languages cannot display any great facility of expressing grammatical modifications by the change of their original materials, without the help of foreign additions. We must expect that in this family of languages the principle of 1) It is not uninteresting to see what was the doctrine of Fhlda, a predecessor of Adelung [Sammlung und Ahstatnmung germaniseher Wur- zelworter, Halle, 1776), concerning the method of obtaining roots ; "Take from a single word its grammatical functions, its prefixes and suffixes, verbal, nominal, and those of gender, number, case, person, tense. Wherever at the beginning or end two consonants stand together, cast away the foremost and hindermost; the root, without losing any of its ohief significance, will become a single syllable." (Page 59. - Franz Bopp. 1 1 compounding words will extend to the first rudiments of speech, as to the persons, tenses of verbs, and cases of nouns &c. That this really is the case, I hope I shall be enabled to prove in this essay, in opposition to the opinion of a celebrated German author, who believes that the grammatical forms of the Sanskrit and its kindred languages consist merely of in- flections, or inner modifications of words." The second point is still more important, viz., the hypo- thesis which appears in the Analytical Comparison concerning the derivation of the personal suffixes from personal pronouns. The passage where this hypothesis is first introduced is so interesting that I quote it in full : "The indication of the persons of verbs in the Sanskrit language and those of the same origin Mr. F. Schxegbl con- siders as being produced by inflection ; but Scheiditjs shews very satisfactorily, with respect to the plural at least, that even the Greek verbs make use of pronouns, in compound structure with the root, to indicate the various persons. With respect to the singular, he would have succeeded much better if he had not limited himself to the corrupt form in to, terminating the third person of the present in ei, where I cannot perceive any pronoun incorporated, — but had extended his view to the form in [xt, terminating the third person in the Doric dialect with n. ScHBiDius commits another fault, namely, that in speaking of the pronouns he stops at the nominative, whilst the crude form of nouns may be better extracted from the oblique cases. In this way it is easy to discover that to is the radical ■ form of the Greek article, which is originally nothing more \ than a pronoun of the third person, and is used as such in i Homer. This to, bereft of the final vowel, becomes an essen- tial element of verbs in their third person, singular, dual and plural, as 8180T1 {!), SiSotov, 8Kovti. I have no doubt but it can be proved, with as much certainty at least as in the case of the Arabic, that Sanskrit verbs also form their persons by com- pounding the root with the pronouns, upon which subject I shall offer a few remarks in its proper place." (Page 11.) For these intended remarks, however, Bopp seems to have found no opportunity in the course of his discussion , and merely observes (page 16) : "In the present tense the pronominal 12 Chapter I. consonants M, S, T, of the singular number and of the third person plural, are articulated with a short «'", — from which we see that at that time he had not come to the conclusion, as he did later, that mi arose from ma. In the above exposition our attention is first of all attract- ed by the reference to Schbidius, who is said to have akeady established the principle of composition "very satisfactorily". He refers to the detailed treatment of the question contained in i. C. Valckenarii ohservationes acad. et Jo. Dan. a Lennep praeleciiones academicae rec. Everardus Scheidius [Trajecti ad Rhenum, 1790), page 275seq. Leaving it to the reader to enjoy the various etymological fantasies, I will only quote the words of ScHEiD which are of interest for the main question. They are as follows : "Memini equidem, quum ante hos octodecim, et quod excurrit, annos, contubemio fruerer viri summi, quem honoris causa nomino , Joannis Jacobi Schultensii , inter familiares sermones, quibus de linguarum indole agebatur, narrare Schul- tensium, virum suavissimum et harum rerum elegantissimum arbitrum, Lennepio placuisse, ut, quemadmodum in verbis orientalium, adformantes, quae dicuntur, temporis praeteriti proprie essent syllabae literaeve, a pronominibus antiquis quasi resectae : ita et in Graecorum verborum temporibus per- sonisque eadem fuisset sermonis ratio." We see from this passage that Bopp's view of the per- sonal endings was finally suggested by Hebrew grammar. Now that the principle of composition was once recom- mended in this way, it is no wonder that it was also applied in other cases than in the tenses compounded with as, and in the personal suffixes, — so, for instance, in the optative, whose I is first explained in the Analytical Comparison , page 23, as the verb "wish", "desire". Of real inflection in Schlegel's sense of the term Bopp in the Analytical Comparison retains only certain vowel - changes (so the ai of the middle voice, which he did not then explain by means of composition, as he did later), and reduplication. (Pages 12 and 34.) After Bopp's view had been formulated in the two ways above mentioned, in the Conjugationssystem and the Analytical Franz Bopp. 13 Comparison, it assumed at length a third and final shape, which was first introduced in a series of academical essays, and at last appeared in the Comparative Grammar, and which chiefly differs from the second form in more and more exclu- sively emphasizing the principle of composition, as well as applying it to those departments of grammar which had not heen treated in the Conjugationssystem and the Analytical Comparison. This theory is now intelligible without further preface, and we can sum it up briefly as follows : The words of the Indo-European languages must be derived from roots, which are all monosyllabic. There are two classes of roots, viz., verbal roots, from which spring verbs and nouns , and pronominal roots , from which pro- nouns , primitive prepositions , conjunctions and particles have their origin. (Cf. beside the Vergleichende Grammatih, § 107, also Ahhandlungen der Berliner Akademie, 1831, page 13 seq.) The case -endings are (at least for the most part i) pro- nouns by origin. Thus the s of the nominative is derived from the pronoun sa ; the m of the accusative recalls the Sanskrit pronominal stem i~ma; the T- sound of the ablative comes from the same pronominal stem ta to which the neuter d in id also owes its origin, etc. (Cf. Abh. der AAad., 1826, page 98.) The personal endings of the verb are derived from the pronouns of the first, second and third person ; mi is a weak- ening of the syllable ma , "which in Sanskrit and Zend forms the underlying theme for the oblique cases of the simple pro- noun". From mi is further derived m. In the plural ending mas, is found either as, the plural characteristic of nouns, or the pronominal element sma. The v of the dual is only a cor- ruption of the plural m. The endings of the second person in similar fashion go back to tva, those of the third person to ta (for nti V. below, page 15) . Bopp does not express a confident 1) "For the most part", because a few endings {os and sam) are not Considered as accounted for, and sometimes a symbolical explanation {v. below, page 15) is attempted. 14 Chapter I. opinion in regard to the middle endings, but he thinks it probable that they are due to the doubling of the correspond- ing active endings. As for the characteristics of the present stem, like vo in CsoyvufAi, it is most probable that the greater part of these are pronouns. The augment, which is mentioned in connection with the imperfect, is considered by Bopp [Vgl. Gr., § 537, and eren earlier in the Analytical Comparison, page 27) as identical with a privative, and is therefore regarded by him as a nega- tion of present time. But he |also admits the possibility of connecting it directly with the pronominal stem a "that", to which, moreover, he regards the negative particle itself as related. In the S-aorist the s belongs to the substantive verb, and the explanation of the composition is that the imperfect of as (but without the augment) forms the end of it. "I recognize", he says in § 542, "in this s the substantive verb, with the im- perfect of which the first form [of the aorist] wholly coincides, except that the a of asam etc. is lost". The sya of the S-fu- ture, such as dasydti, Bopp regards as the future of as, which is lost in its isolated use. Beside this, he thinks it probable that all verbs once possessed a future formed by means of ya, and that this ya itself, as well as the sign of the optative, comes from the root I "wish". In the aya of causatives he discovers the verb i "go" (as well as ya "go" in the ya of the Sanskrit passive) , and in the s of desideratives the substantive verb. The same composition is met with in cextain formations of the individual languages, e. g. ama-m, in which the root hhu can be recognized ; ama-rem, where we find the root as, etc.i) (Cf. Vgl. (??•.,§ 521.) Finally, the stemforming suffixes are partly of pronominal, partly of verbal origin (e. g. da\tar "giver" means really "he who walks through the action of giving", from da "give" and tar "walk through") . ') On the other hand, Bopp does not assume that new root-words could arise in an individual language. (Cf. preface to the third section of the Vgl.Gr., 1st edition, page XIV.) Franz Bopp. 15 Beside this explanation by composition, a second is some- times brought forward, the symbolical. Thus the following remark is made concerning the dual: "The dual, inasmuch as it is founded upon a clearer observation than the more in- definite plurality, prefers the fullest endings, as conducing to stronger emphasis and a more vivid personification." ( Vgl. Gr., § 206.) The same is true of the feminine, "which in Sanskrit prefers a luxuriant fullness of form, in the stem as well as in the case-endings." (§ 113.) The n is also symbolic in the third person plural -nti, which is supposed to be formed from Why the insertion of a nasal. He considers that this in- sertion is the least strange of admixtures, and the nearest approach to the simple lengthening of an already existing vowel. (§ 236; cf. also § 226.) If we compare this final aspect of Bopp's views with the preceding one , we observe that Schxegel's influence has dwindled down to a slight remnant. For the ai of the middle endings, in which Bopp formerly saw an inner inflection of the root, is now rather explained by composition, and there- fore reduplication alone remains as a sort of inner modifica- tion of the root. (And even this reduplication , which per- haps was originally the repeated root , cannot be called an "inner" change in any strict sense of the word.) Accordingly it was natural that Bopp should formally de- clare his disagreement with Fr. Schlegbl, by a keen polemic in the Comparative Grammar. The passage referred to is as follows : "By 'inflection' Fr. v. Schlbgel understands the inner change of the root -sound, or the inner modification of the root, to which he opposes affixion from without. But if the Greek 8i8a)[i,i, Soiooj, 8o&7jao(jie&a come from 8o or 8u), what else are the forms [ii, aou, dTjaofisSa except evident external additions to the root , which in its interior is either not changed at all, or only in the quantity of the vowel? If, then, we are to understand by 'inflection' an inner modification of the root , Sanskrit , Greek etc. scarcely exhibit any inflec- tion at all, with the exception of reduplication, which is de- rived from the resources of the root itself. If, on the other 16 Chapter I. hand, drjaofieda is an inner modification of the root 80, simply because it is connected with it, is adjacent to it, and together with it represents a whole, then the notion of sea and main land could as appropriately (represent a modification of the sea, or the reverse." We can characterize the theory of Bopp, as developed above, leaving out the slight symbolical addition, as the com- position or agglutination theory i) . I will not attempt here a more detailed criticism of the agglutination theory, but will leave it for the fifth chapter. I would like , however, to call attention again to the fact that Bopp's explanations have not, as has been sup- posed , spontaneously arisen as the natural consequence of comparison, but that they have grown out of various and in- dependent views and conclusions. For in addition to the sug- gestive stimulus resulting from the details of the investigation itself, Bopp had also in mind bits of the learned tradition of former times , as for instance the prejudice in favor of the threefold nature of the parts of speech, which seeMtts to have first given rise to the idea that the substantive verb is to be recognized ia the shape of various s's in the verbal forms; further, the transmitted theory that roots are to be regarded as monosyllabic ; and finally, the tradition derived from He- brew grammar, that we have to recognize affixed pronouns in the personal suffixes of the verb. 7 II. Bopp's method of comparing given languages. Having discussed Bopp's theory of inflection in the first section of this chapter, I will now treat of his comparison of given individual languages. Of course it cannot be my aim to record the results which have been attained through Bopp's comparison of the Indo-European languages ; I will simply attempt to describe the method which Bopp employed. We must not, however, expect from Bopp a systematic answer, which shall comprehend all separate instances, either on this point or any other. Bopp's method of demonstration is 1) It -was so named first by Lassen, with the intention of easting a slur upon it. (Cf. Pott, JEtymologische Forschungen, l^t edition, 1, 179.) Fbanz Bopp. 17 exactly the opposite of Humboldt's. While Wilhelm von Humboldt is never weary of expounding generalities, and exerts himself at every turn to subordinate details to ideas, Bopp occupies himself chiefly with individual points in Ian- \ guage , and very seldom intersperses general observations, such as could be termed "philosophical". It is as impossible to obtain a theory and systematic method for linguistic science from Bopp's Comparative Grammar, as it would be to extract grammatical paradigms from Humboldt's "Introduction to the Kawi Language". Under these circumstances we must in- vestigate with caution Bopp's theoretical views in regard to the forces at work in language, — that is, we must be careful, where he uses certain terms with an easy carelessness, not to examine their significance and breadth of application as inexorably as if we were setting up a system of terminology. I feel, therefore, as if the fairest way of proceeding were to frame our question thus : what are the general views constitut- ing the standpoint from which Bopp was accustomed to judge the processes of language ? — and to answer the question as fol- lows : his general views had a coloring of natural science, be- neath which, however, the old philological background had not yet vanished. His fondness for the terminology of natural science is at once apparent when he attempts to describe his method of treating language in contrast to that of former schol- ars. He aims at a comparative " dissection " [ Zergliederung ] of language ; systematic comparison of languages is a "lan- guage-anatomy" ; we have to deal with an "anatomical dissec- tion" or "chemical decomposition" of the body of language, or, to use another figure, with the "physics" or "physiology" of language. This coloring is very prominent in the first sen- tence of the preface to the Vergleichende Grammatik : "In this book my aim is a comparative, comprehensive description of the organism of the languages mentioned in the title, an investigation of their physical and mechanical laws, and the origin of the forms characterizing grammatical rela- tions." What is meant by "physical and mechanical laws" in this sentence, the author has himself explained in reply to inquiry, as Breal informs us in the French translation of Bopp's Com- Delbbdck, Introduction to the Study of Language. 2 18 Chapter I. parative Grammar. By "physical laws" is meant what we now call "phonetic laws"; by "mechanical laws" the rules which Bopp believed he had established concerning the relative weight of vowels and syllables, of which we shaU speak later. The meaning of "organism" and "organic" is shown by one or two passages in the Vergleichende Grammatik. In the pre- face to Heft 2, 1=* edition, page VII, we read: "The inflec- tions make up the true organism of a language"; and on the other hand he speaks of "languages with monosyllabic roots, without the capacity of composition, and hence without organ- ism, without grammar." (§ 108.) "Organism" of a language is accordingly nothing but the grammatical "arrangement" [ Einrichtung ] , which is founded on agglutination (preface to the first volume of the Vgl. Gr. , page IV) ; "organic" is every- thing which is in accordance with this arrangement , and "inorganic" what is at variance with it. We can therefore say "original" instead of "organic", and "not original" instead of "inorganic". So, for example, the v of the ending [atjv is said to be "organic, i. e. not a later, meaningless addition, but intentionally employed, and inherited from the prim- itive period of our branch of language"; 'on the contrary,, the [J.I of TuiTTOi[ii is considered "inorganic", because the opta- tive, in all languages where it exists as a separate form, has the short endings, even in the first person, with th* single exception of Greek. Everything is "inorganic"" which cannot, according to the view of the grammarian in ques- tion, be derived from the original structure of the Indo-Eu- ropean. We see that the terms "mechanical", "physical", "or- ganic" are not used strictly in the sense they possess in natural science, yet we can conclude from their application that Bopp's conception of language was of a kind of organic body. He uses this very word in the Vocalismus, page 1 : "Languages must be regarded as organic bodies [orga- nische Naturkorper ] , formed in accordance with definite laws; having a life-giving principle within, they develop and then gradually die out, after losing consciousness of their true na- ture, and throwing aside, or mutilating, or misusing (i. e. ap- plying to uses to which they were not primarily adapted) their Franz Bopp. 19 members or forms, which, were originally significant, but have gradually become a more external mass." This sentence introduces us to two new trains of thought. In the first place , I would call the attention of the reader to the remark that language in the course of time loses conscious- ness of its own nature. Here a mental activity is ascribed to language ; it is referred to as if it were a thinking being. Nor is this an isolated instance. In other passages Bopp speaks of the spirit or genius of language, and recognizes in its procedure certain tendencies and aims. Sometimes, in- stead of language as a whole, an individual form is regarded as a thinking being. So for example in the Vffl. Gr., P* edi- tion, page 516, the Slavonic stem sj'o is said to be "no longer conscious of its composition , which was handed down from the primitive period of the language." These expressions are metaphors, — very natural ones, too, and probably, if any one had called his attention to the point, Bopp would have acknowledged that in reality these psychical activities take place, not in language, but in speaking individuals ; yet it is important to call attention here to the first beginnings of a mode of view which with Schleichbe rose to a conscious hypostasizing of the notion "language". In the next place, in the sentence above quoted the ex- pression "die out" is noteworthy. According to Bopp, all ex- ternal changes which we observe in the Indo-European lan- guages betoken not development, but disease, mutilation and decline. We become acquainted with languages, not in their ascending development, but after they have passed the goal set for them. That is, we find them in a state "where they might still perfect themselves syntactically, but where , gram- matically considered, they have lost more or less of what belonged to that perfect arrangement, in virtue of which the separate members were in accurate proportion to each other, and all derivative formations were still connected, by a visible and unimpaired bond, with that from which they ori- ginated." {Vocalismus, page 2.) As long as the meaning of the composition continues to be felt in a grammatical form, it offers opposition to any change. But the farther languages are separated from their source, the more love of euphony gains 2* 20 Chapter I. in influence. [Ahh. der Berl. Akad., 1824, page 119.) This view has also been extended and systematized by Schleichbk. Having thus briefly characterized Bopp's fundamental views , I will now give a more detailed account of his ideas concerning changes in language, and will classify them in accordance with the categories introduced by Bopp himself: mechanical and physical laws. The effect of Bopp's so-called "'mechanical laws" is espe- cially visible in the changes which the weight of the personal ! endings produces in the stem. A light ending follows a heavy ' form of the stem, e. g. emi "I go", from i "go"; but before a i heavy ending a light stem-form alone is permitted , e. g. imds "we go". The same law accounts for the German Ab- laut, which is preserved to the present day in weiss and ids- sen. These facts, which were first formulated by Bopp, we now explain in a different manner, by ascribing the weaken- ing of certain syllables no longer to a law of relative weight, but to the power exercised by the accent of the following syllable. Beside the influence of the weight of the personal endings, Bopp recognizes another action of this law of gravity, which will be apparent from the following examples. It is the task of the stem-syllables to carry the formative syllables, and it sometimes happens that a stem -syllable is not strong enough for this purpose. We have such an instance in l^he Sanskrit imperative cin-A "gather", from ci; Bopp here re- marks that the sign nu is/'onTy. able to carry the ending hi iwhen the u is supported 'Sy two preceding consonants, as for example in apnuhi. " But where the u is.^^1^ preceded by a single consonant, it has become incapable of carrying the ending hi, hence cinu 'gather', from ci." (§ 451.) In a" similar manner Bopp explains the circumstance that the per- fect endings appear greatly mutilated in comparison with those of the present. Since in the perfect the root has also the reduplication- syllable to carry, it is , so to speak, claimed by both sides at once, and is therefore no longer in a condition (To lift a heavy ending. It is clear that this sec- ond law of gravity, whose action Bopp discovers in several other instances, is in direct contradiction to the first, and it Fkanz Bopp. 21 is now universally acknowledged that the idea expressed in this law suffers from a metaphorical obscurity. I have intimated above that the mechanical laws can no longer be understood and accepted by us in the same manner as by Bopp, and will pass to the '^physical laws", which we are now accustomed to call "phonetic laws". In order to appre- ciate Bopp's stand-point in this connection, it is important to come to a clear understanding of the possible method of estab- lishing phonetic laws. Whoever compared Sanskrit with an- other Indo-European tongue, the Greek, for instance, was of necessity impressed with the fact that there exist in both lan- guages words and formations which completely coincide. No one could avoid noticing, for example, that the Skr. matdr and Gr. [atjttjp, Skr. ddma and Gr. 86(^0?, Skr. pitdr and Gr. Tca-njp were the same words, and that the inflectional endings of the verb agree in the main in the two languages. The re- cognition of this agreement rested upon immediate evidence, and could not be further demonstrated. From comparison it was possible to deduce the rule that certain sounds of the Greek corresponded to certain sounds of the Sanskrit, m to p., ^ to t, etc. Yet after collecting a very few words, it immediately be- came plain that the same sound of the Sanskrit was not al- ways represented by the same sound of the Greek. So for example in ddma 8o[j.og, dddami Si'Stop-i, the Greek 8 cor- responded to the Sanskrit d; but in the pair duhitdr &uyaTirjp, which no one wished to separate, the Sanskrit d was repre- sented by a Greek 0. As a result of such observations, it was necessary to adoptthe conclusion that these rules admit of excep- tions, and to say accordingly : "Usually Sanskrit d corresponds to a Greek 8, but often also to a Greek 0." Now two positions are conceivable in relation to such a rule. We can either start with the theoretical conviction that laws admit of no excep- tions, and feel ourselves bound to investigate the causes which produce the so-caUed "exceptions"; or we can content our- selves in the wording of our rules with the expressions "usu- ally" and "often". And this latter is on the whole Bopp's stand-point. "We must expect to find no laws in language", he remarks, "which offer more resistance than the shores of rivers and seas". {Vocalismus, page 15.) In other passages he 22 Chapter I. adopts the same convenient view, at least for part of the phonetic processes observed by him, his opinion being that there are two sorts of euphonic change in language; "one, which is elevated to a universal law, appears in like form on every like provocation, while the other, which has not become a law, occurs only occasionally." {Vgl. Gr., 1^* edition, § 236, note.) That the latter class of phenomena in Bopp's opinion covers a broader ground than the former, is soon evident. He frequently claims for language the right to depart from the existing law with "a certain freedom". That vowels should be lengthened without cause, extensive mutilations take place without conceivable provocation (as for example that itUTtr^v should be a mutilated form of exocp&Yjv) , and that the same phonetic group should pass into widely differing formations in the same linguistic period, appears to him not at all extra- ordinary. For instance, he assumes that the pronominal stem sma in Gothic appears in six different forms, as nsa, sva, , nka, nqva, mma and s. (§ 167.) When he was unable to find in the same language an analogy for a phonetic change which seemed probable to him, he had recourse to another ; for ex- ample, in order to confirm the assertion that the I of the Slavonic participles was derived from i, he referred to the Bengali. The X of SsStoxa he traces back to an «, but in tixutpa this x. has become A, "as it were in the spirit of the Germanic Law of Permutation of Cor\so-a.dia.\:& [Lautverschiebungsgesetz]'\ and this h together with the preceding tenuis or media has become an aspirate. (§569.) Even the admission of wholly isolated cases of change does not terrify him. Bopp but seldom claims infallibility for a phonetic law. An interesting example of the sort occurs in his article on the demonstrative pronoun and the origin of case-signs. [Abh. der Berl. Akad.^ \^2^.) There it is of great importance to him to prove that the article sa, 6, can never have had a nominative -s, and while rejecting the assumption that the s could have fallen away in Sanskrit and Greek, he adduces the infallibility of certain phonetic laws in the following expressive terms : "But we must not overlook the fact that such elimina- tions /"^Ssc/Jeifww^ewy usually, if not always, occur in numbers and according to rule, rather than in single instances and arbi- Fkanz Bopp. 23 trarily ; and if the spirit of a language at any period of its history conceives a hatred for any letter as the terminal pillar of a word, it removes it wherever it occurs, so that not a single such letter remains to give ground for the suppo- sition that a similar one ever existed. In this way a phonetic law raged in Greek against the letter x, and eradicated it in every case where it stood as 'final consonant, in spite of the importance and extent of its grammatical r6le, which we can clearly recognize hy comparison with the kindred languages. On the other hand, a has always been a favorite final letter to Greek ears, and as readily as it has allowed itself to be drop- ped out in the middle between two vowels, just so persistently does it appear at the end, wherever the researches of compar- ative philology lead us to expect it." We see from these quotations, which could be increased ad infinitum^ that Bopp did ascribe infallibility to a phonetic law in single cases, where the facts seemed to prompt it, but by no means as a general rule ; on the contrary, he granted to language the freedom of occasionally emancipating itself from the existing laws. It is universally acknowledged (even by those scholars who do not advocate the principle that phonetic laws admit of no exception) that Bopp has left the greatest task for his successors in the department of phonetics. The impression that the words compared were identical was, as already intimated, always decisive for him, and the sounds had to adapt themselves to this impression ; in his assertions about sounds, he did not give sufficient weight to the modify- ing influence arising from a comparison of the fate of the same sounds witnessed elsewhere. It is to August Fried- rich Pott that the great credit is due of having filled up this gap. This want of method in Bopp's investigations was not so palpably evident in the Indo-European domain, because there a great number of forms and words really exist in which the same sound appears in the same position, and because in the discovery of hidden resemblances Bopp was guided with wonderful coiTectness by the deep insight of his genius ; it became very conspicuous, however, when Bopp undertook to introduce into his comparison languages whose relationship 24 Chapter I. to our linguistic branch was not established, — I refer to the Malay-Polynesian. I think it is now universally acknowl- edged by philologists that these languages have nothing in common with the Sanskritic languages, but Bopp was under the impression that they stood in a daughterly relation to the Sanskrit, and attempted to establish this relationship in the same way as he had that of the Indo-European languages in his Comparative Grammar, — so far, that is, as was per- mitted by the character ot these tongues, which "have under- gone a total dissolution of their original structure". Here, also, he formed no tables of phonetic correspondence, but compared words which seemed to him identical (e. g. num- bers), and tried to account for the phonetic changes in each separate instance. His method was naturally more ar- bitrary here, where he had to work with an entirely an- tagonistic material, than within the Indo-European do- main. As an example, I will take the word po. which has the meaning "night". Bopp says in regard to it [Ueber die Verwandtschaft der malayisch-polynesischen Sprachen mit den indisch-europdischen, Abh. der Berl. Akad., 1840, page 172) : "The usual appellation of 'night' in the South Sea languages, i. e. in the New Zealand, Tahiti and Hawaian tongues, is po, which, echo-like, reproduces only the last syllable of the Sanskrit ksapas, hsa'po." Now there is another word ho "day", which, as he says on page 218, might have been derived from the Sanskrit divas, divo. "But if", Bopp continues, "there should prove to be a connection between the Tongan bo and the above-mentioned po, which in the South Sea languages signifies 'night', we should be obliged to give up connecting this po with the Sanskrit kmpns, and to assume that this po has lost an epithet which in the Tongan language changes 'day' to 'night', and characterizes the latter as 'black' or 'dark day'." After what I have said of Bopp's relation to phonetics, it is not necessary to occupy ourselves any further with such vagaries, as it will be clear from the preceding that the failure of this undertaking in the field of the Malay-Polynesian does not manifest a constitutional blemish in linguistic Feanz Bopp. 25 science as a whole, but simply a lack inBopp's method, which was subsequently supplied. Yet it was very natural that Bopp's ideas concerning phonetic change and phonetic laws should be rather latitu- dinarian. Bopp was no natural philosopher, but a philolo- gist, who was occupied with grammars his whole life long. To a natural philosopher, it is true, the idea that a law can have exceptions at will is ridiculous or repugnant ; but this view was quite common in philological theory and prac- tice. In all grammars the mass of "irregular" words was at least as great as that of the "regular" ones, and a rule with- out exceptions actually excited suspicion. Such traditional opinions, indeed, only die out in the course of generations. What Bopp achieved, as already remarked, was the estab- lishment of, an independent theory concerning the origin of inflection , and the scientific demonstration of the original community of the Indo-European languages. Now that we have introduced the reader to Bopp's labors in both fields, we are able to declare briefly and comprehen- sively what mental peculiarity is especially prominent in the writings of this great scholar. When we hear that a single individual has treated com- paratively the Sanskrit, Old Persian, Zend, Armenian, Greek, Italic, Celtic, Slavonic and Germanic languages, and has even passed on beyond this immense field to the languages of the South Sea, we are easily inclined to ascribe to him an unusual, nay, an extraordinary amount of learning. But on a nearer examination we readily see that learning is not really a qual- ity which is especially characteristic of Bopp. He certainly learned a great deal in the course of an industrious life, but he was not one of the men whose learning appals us, as is the case with A. W. v. Schlegel. He possessed (from a phi- lological point of view) but scanty knowledge of many lan- guages in the elucidation of which he acquired immortal honor, as for example the Slavonic and Celtic ; and with regard to 26 Chapter II. certain traditional details, as for instance the rules of Latin grammar, he was occasionally more indifferent than could be desired. For example, he had no objection to giving his Sans- krit dictionary the title : "Glossarium sanscritum a Franzisco Bopp", and preferred to construe postquam with the pluper- fect! Whatever did not seem to him to contribute to the explanation of forms and the comprehension of the primitive condition of language was comparatively indifferent to him. Nor is it wholly correct that Bopp, as is often asserted, invented the method of linguistic comparison. Bopp is incom- parable in his power of recognizing the former unity of what has been separated , but he has introduced no methodic art which could be learned from him in turn. Indeed, his weak point lies on just this methodic side, as has been shown above. Bopp's greatness consists in something else, something which is independent of learning and method, namely, in what we call genius. His Comparative Grammar is based, upon a series of discoveries which were not due to learning and experience, but to a gift of nature which we cannot analyze. Of course I do not mean to say that Bopp was not greatly indebted to his learning and his logical mind, but simply that a happy intuition plays a much more important part with him than with other distinguished philologists, as for in- stance with AtTGTJST Schleicher. CHAPTER II. BOPP'S OONTEMPOEAEIES AND SUOOESSOES DOWN TO AUGUST SOHLEIOHEE. Bopp was independent , but not solitary in his depart- ment. At the same period Wilhelm von Humboldt, August WiLHELM VON ScHLBGEL and Jacob Grimm were working in BOPP'S CONTEMPORABIES AND SUCCESSORS DOWN TO A. ScHLEICHEE. 27 closely adjoining fields. I will try to estimate the influence which these men exerted on the science of which Bopp was the founder. Of WiLHELM VON HuMBOLDT, Bopp never speaks without an expression of reverence. It will suffice to quote the words with which he closes the preface to the second part of the Comparative Grammar : "As to this idea [regarding the declension of adjec- tives], which has already been touched upon elsewhere, I have had the happiness of learning the favorable judgment, above all precious to me, of my lamented patron W. v. Hum- boldt, in whom philology has recently lost its fairest orna- ment. While still overwhelmed by grief at this severe loss, I cannot refrain from here paying the tribute of most heartfelt reverence and admiration to the renowned memory of this great man, since I have been deeply impressed by his brilliant writings in the field of philosophical and historical linguistic research, as well as by the instructive and delightful inter- course I had with him, both in person and by letter." Yet I cannot discover that W. v. Humboldt exerted any considerable influence upon Bopp. Humboldt's many-sided nature, with its capacity for uniting and reconciling the most endless variety of conceptions and aspirations, was not adapt- ed to change the current of a mind of such great and simple power as Bopp's. There is nothing more difficult than to clearly define in what the influence consists which Humboldt exerted upon Indo-European linguistic research. It is not easy within this domain to point to a field where he was pioneer, to definitely name a theory which he established, to mention a mode of view which can be wholly traced to him ; yet not only Bopp, but also other representatives of the science, as Pott, Schleicher and Cubtius, acknowledge themselves Hxbiboldt's grateful pupils. To the question, how Humboldt influenced these men, I think we must answer : chiefly through the totality of his own being. His lofty and disinterested love of truth ; his endeavor not to lose the whole from sight while considering details, nor details while considering the whole, and thus to avoid the dangers of specialism as well as those of the previous universal grammar ; the just balance of his judg- 28 Chapter II. ment; his broad mental culture, and his noble humanity, — all these qualities have a strengthening and elevating effect, upon any other scholar who approaches Wilhelm von Hum- boldt, and this sort of influence I think Humboldt will still retain for a long time to come , and will continue to exert even upon those who can make nothing of his theories. Posterity has taken a less friendly position in relation to August Wilhelm von Schlegel than toward Wilhelm von Humboldt. I think it is not sufficiently known outside of philological circles that the translator of Shakespeare was also the founder of Sanskrit philology. A. W. v. Schlegel was in his forty-eighth year when he began to occupy himself with Sanskrit, but his admirable industry, and a gift he had of fam- iliarizing himself with new subjects, which had been strength- ened by practice in many directions, made him in a short time master of the vast difficulties which then stood in the way of the study of Indian literature. With admiration we see how rightly he at once defined the tasks which were to be accomplished : "If the study of Indian literature is to thrive", he says in the Indische Bibliothek^ 1, page 22, "the principles of classical philology must be applied to it, and that, too, with the most scientific acuteness. It is in vain to object that the learned Brahmans possess the knowledge of their old books through unbroken transmission ; that for them Sanskrit is still a living language, and that accordingly we should go to school to them alone. With the Greeks the case was the same before the de- struction of Constantinople ; the knowledge of a Laskaris, of a Demetrius Chalkondylas, in regard to the ancient literature of their race, was certainly of value; yet the scholars of the West did well not to confine themselves to it. However, the acquaintance with Latin literature, which had never wholly died out, gave a tolerable preparation in Europe for the read- ing of Greek. Here , on the contrary, we come into a com- pletely new circle of ideas. We must learn to understand the written monuments of India both as Brahmans and as Euro- pean critics. The Homeric questions of today were not more foreign to those learned Greeks than the investigations regarding the origin of the Indian religion and legislation, the BOPP's CONTEMPORAKIESAND SUCCESSORS DOWNTOA. ScHLEICHEE. 29 gradual development of mythology, its unity and its contra- dictions, its cosmogonic, physical or historical significance, or finally, regarding the intermixtures of subsequent fraud, would be to the scholars of India. The same tasks belong to the editor of Indian books as to the classical philologist, viz. : proof of the genuineness or spuriousness of entire works and single passages; comparison of manuscripts, choice of readings, and sometimes conjectural criticism ; and finally, the employ- ment of all the artifices of the most sharp-sighted hermeneu- tics", etc. A. W. V. ScHLEGEL let deeds follow in the wake of his article. His editions, according to the opinion of competent judges, accomplished all that was possible at that time, and formed the beginning of Indian philology. A. W. V. Schxegel's position toward Bopp was at first a friendly one. It was he who first (in the Heidelberger Jahr- bilcher, Sept., 1815, No. 56) announced to the public what it had to expect from Bopp ; he reviewed Bopp's edition of the Nala with appreciation and good will, and declared in 1827, in the first letter to Heeren [IndischeBibliothek, 2, page 385), that Bopp and he since their acquaintance, begun in Paris in 1812, "had always worked for the same aim in friendly emu- lation and harmony". Subsequently the relation was changed, and in place of the "friendly emulation" grew up one of those literary enmities which were a vital necessity to A. W. V. ScHLEGEL. A thorough polemical discussion never took place between ScHLEGBL and Bopp ; there were only single sharp epigrams by A. W. V. ScHLEGEL, which were answered by Bopp. The difference was in relation to two fields, Sanskrit philology and linguistics. Bopp found time amid his vast comparative labors to gather the necessary materials for the study of Sanskrit, and to bring out the edition of Nala, a glossary, and above all, his Sanskrit Grammar in several different forms. And in this latter he was guilty of an omission which A. W. v. Schle- GEL could not pardon him. Bopp never made a special study of the native Indian grammarians, but what he could use from them he took at second hand, i. e., from the grammars of his English predecessors, contenting himself with penetrating into 30 Chapter II. the sacred language of India by direct observation and com- parative analysis. Now there is certainly no doubt that Schlb- GBL was quite right in theory, when he demanded that the native masters of Indian grammar should not be neglected ; but it is also just as certain that Bopp was guided by a cor- rect feeling. It would have cost him years to familiarize him- self with the Indian grammarians, with the aids then at his command, and Benfey justly remarks [Geschichte der Sprach- wissenschaft, page 389) that it is questionable whether this eminently philological task would have been exactly adapted to Bopp. In the other field, that of comparative philology, A. W. V. ScHLEGEL felt himself called upon as it were to defend the honor of the family. The brother took it very ill that Bopp separated himself more and more from the theory of Feibdeich ScHLEGEL. He regarded himself as the natural defender of the "organic" view, upon which Bopp's "agglutination theory" was gaining ground in so threatening a manner. Schlegel unfortunately did not get any farther than the announcement of a great philological work , which was to bear the title : "Etymologicum novum sive synopsis linguarum, qua exponi- tur parallelismus linguae Brachmanum sacrae cum lingua Graeca etLatina; cum reliquiis linguae Etruscae, Oscae ceter- arumque indigenarum veteris Italiae dialectorum; denique cum diversis populorum Teutonicorum liuguis, Gothica, Sax- onica, Francica, Alemannica, Scandica , Belgica." However, there exists a comprehensive and detailed critique of Bopp's grammatical works by Schlbgel's confidential pupil Christian Lassen, from which we can form some idea of how Bopp was judged in Schlbgel's circle. The tone in which Lassen writes is that of the cold but just judge. What is praiseworthy is appropriately emphasized, what is mistaken is censured se- riously, and only on the mention of the agglutination theory does animosity appear. The passage in question is as follows [Indische Bibl., 3, page 78) : "I had intended to speak against the agglutination theory, which again recurs in this connection, but since I know that Herr von Schlegel intends to discuss this point, I will gladly impose a voluntary silence upon myself in regard to the matter. BOPP'S CONTEMPOHABIES AND SUCCESSORS DOWN TO A. SCHLEICHEK. 3 1 which well deserves to be treated by his superior hand. I will therefore simply mention that according to Herr Bopp's view the characteristic letters of the personal endings are really af- fixed pronouns, and that the origin of many tenses is sought for in the incorporated substantive verb [as) . This word plays in general, in the book in question, the r6le of the old 'every- where-and-nowhere', and transforms itself in Protean fashion into the most diverse forms. Although the preparations in which Herr Bopp dishes up this small word as seldom appear to me particularly tasteful, yet out of gratitude for his former meritorious efforts, I will call his attention to an unknown form of this verb, concerning which I am rather at a loss, — without meaning to assert thereby that it could not be employed by others for the most unexpected derivations. This form is as (for ast], the third person sin^lar of the imperfect active (Panini, VII, 3, 97). The shortness of the form renders it very convenient for derivations, just as for word -comparison no words are so useful as the short Chinese ones, because it is only necessary to leave a vowel out of account, and to change one consonant into another, in order to manufacture Finnish, Koptic and Iroquois at will. But m'C reach the culmination of the agglutination theory in the derivation of the simple aug- ment from alpha privative. Of all the extraordinary qualities which have been ascribed to the primeval race, this logic is the most remarkable, namely, that they said 'I do not see', instead of 'I saw' ! As applied to pedagogy, this modus operandi would have to be expressed as follows : 'Begin the education of your children by cutting off their heads'. A verb is first deprived of its meaning, in order to construct a new form from it." This critique of Lassen excited great indignation among Bopp's friends, but it had no permanent influence, because it was devoid of positive statements, such as could have replaced Bopp's agglutination theory. Nor was this lack openly sup- plied at any subsequent time, either by A. W. v. Schlbgel or any of his adherents. Thus Schlegel's opposition was gradually forgotten, and Bopp's theories maintained undisput- ed possession of the field. Not until the appearance of Westphal's grammatical works did Schlegel's view expe- 32 Chapter II. rience a sort of renaissance. Of these we shall have to speak later. Hence we see that Schlbgel's influence upon comparative philology could hardly be called a directly promotive one. But indirectly it has been not inconsiderable. Since Schlbgel gave a powerful impulse to Sanskrit study, a portion of the gratitude is due to him which comparative research owes to Sanskrit philology. Powerful and direct, however, was the influence of Jakob Gbimm. Jakob Grimm stands wholly independent beside Bopp. When the first volimie of the German Grammar came out, in 1819, Bopp had only published his Conjugationssystem, and a critique of Forster's Sanskrit Grammar in the Heidelberger Jahrbilcher . Both of these were quoted and utilized by Grimm, but the whole frame-work of his Grammar dates back to the pre-Boppian period. We learn from Gkimm himself in what his epoch-making achievement consists : "I have been seized with a strong impulse", he says in the preface to the first edition of his Grammar, "to under- take a historical grammar of the German language, even if, as a first attempt, it should soon be surpassed by future writings. During my careful reading of Old German sources, I daily discovered forms and perfections which we are accustomed to ascribe with envy to the Greeks and Romans, when we con- template the constitution of our present speech ; traces which had here remained in ruins, as if turned to stone, became gradually plain to me, and the phonetic changes were explain- ed when the new took its place beside the intermediate, and the intermediate joined hands with the old. But at the same time there appeared the most surprising resemblances between all the sister dialects, as well as relations, hitherto overlooked, between their differences. It seemed of great importance to establish and illustrate this progressive, continuous connection down to the smallest detail ; I have thought out the execution of this plan so completely that what I am at present able to accomplish falls far short of it." The opinion of competent judges has long since, in con- nection with the above words, summed up Grimm's especial merits in the sentence : Grimm is the creator of historical BOPP'S CONTEMPOKARIES AND SUCCESSOBS DOWN TO A. ScHLEICHER. 33 grammar. The German Grammar had a powerful influence upon his contemporaries. In the first place, a deep impression was made by the indescribable richness of material, in com- parison with which the school-boy rules of Greek and Latin grammar appear paltry. It is Grimm's Grammar which first teaches us that complete induction is necessary to the estab- lishment of a law. His method increased the esteem for what can be called the "natural condition" [Naturzustand] of language, securing to the so-called "dialects" their proper position beside the written language , not simply in the field of German, but also in that of other languages, as we can see from the words of Ahrens, who in the dedication of his work on the Greek dialects gratefully mentions the man "qui conspicuo Grammaticae Diutiscae exemplo docuit , dia- lectorum secundum aetates vel stirpes diversarum diligenti et sagaci comparatione quam possit in secreta linguarum pene- trari". Of especial influence upon philologists was the so-called "Law of Permutation of Consonants" [Lautverschiehung], which goes by Grimm's name, although already proclaimed by K.ASK in its main features. While Bopp's researches chiefly aimed at the comparison and explanation of forms, so that in his system the importance of phonetic observations was not emphasized, Eask and Grimm, by means of the law of Lautverschiehung , established the fact that the changes of sounds, or, as it was then expressed, of letters, into one another take place in accordance with laws, and above all, that a fixed historical relation can be observed between the sounds of the German on the one side and of the classical languages on the other. How influential the discovery of the law of iawfeer- schiehung grew to be we are taught by A. F. Pott, the creator; of the phonetics of the Indo-European languages : "It is by no means the least among the excellent services Grimm has rendered to special and general linguistics, that he reinstated the letters in their natural rights, which had hitherto been curtailed by linguistic science, and raised them to the same plane on which they stand in language itself. Grimm's historical exposition of the phonetic changes in the Germanic languages has alone more value than many a philosophical Delbeuck, Introduction to the Study of Language. 3 34 Chapter II. system of philology full of one-sided and empty abstractions ; for in it is demonstrated with sufficient clearness that the letter, as the palpable linguistic element, which although it is not constant, yet moves in a comparatively quiet path, is a more certain [!] clew in the dark labyrinth of etymology than the meaning of words, which is often subject to bold trans- formations ; in it is also taught that philology, especially com- parative philology, has no firm foundation when it lacks an accurate historical knowledge of the letters ; finally, it shows with astounding clearness that even in the case of the simple letters no arbitrary lawlessness reigns (as, indeed, is never the case anywhere in language , the idea being only a dream of idle ignorance), but a reasonable freedom, i. e. limitation through special laws founded on the nature of the sounds themselves." [Etymologische Forschungen, 1, page XII.) Perhaps the opinion is not without foundation that beside Eopp no one has exerted such an influence on comparative philology as Jakob.Grimm (although he was never a compar- ative philologist in the sense that Bopp was, and did not al- ways derive the benefit from Bopp's works which they might have afforded him); at all events, we can assert that he made the worthiest returns for the priceless gifts which accrued to German grammar from Bopp. The immense importance of the investigations begun by Bopp and Grimm could not remain concealed from their con- temporaries, for in truth we can just as well (as Corssbn later expressed it) refuse recognition to the sunlight as to the chief results of comparative philology. But the consequences, that is, as far as the metamorphosis of classical study was concern- ed, were accepted but slowly. Eminent investigators like BuTTMANN went on cultivating their land without looking over the hedge of the neighbor who had discovered a new and bet- ter method of husbandry ; and pedagogues who felt themselves called upon to defend the existing order of things complained of the youths who presumed to metamorphose everything which had hitherto been held as true, but from whose labors BOPP'S CONTEMPORAEIES ADN SUCCESSORS DOWN TO A. ScHLEICHEE. 35 nothing resulted for Greek and Latin grammar but the "ever- lasting locative". {AUffemeine Schulzeitunff, July, 18S3.) All these scholars, who remained conservative either from love of ease or from prejudice, found it difficult to resist the vigorous attacks of the man vs^ho is universally recognized as the most prominent of Bopp's successors, August Friedkich Pott, whose great work : Etymologische Forschungen auf dem Ge- biete der Indogermanischen Sprachen mit besonderem Bezug auf die Lautumwandlung im Sanskrit, Oriechischen, Lateinischen, Littauischen und Gothischen (Lemgo 1833 — 1836) was the foundation of scientific phonetics. Pott recognized that after the works of Bopp and Grimm a sure key to etymology must now be found in phonetics (v. the interesting passage in Etym. Forsch., 2, page 349), and com- petent judges have pronounced Pott especially fitted by nature for the accomplishment of this task, — so far as it is possible to speak of the "accomplishment" of tasks which are in their very nature endless. He showed himself, as Renan expressed it, "un esprit a la fois severe et hardi", as richly endowed with constructive fancy as with controlling judgment. To him are due not only a very large number of the etymologies which are held to be correct, but also the first comparative phonetic tables which embrace the whole extent of the compared lan- guages. In my opinion it will be the verdict of the future that Pott occasionally, misled by his fancy, took the liberty of mak- ing unwarranted assumptions, as for instance in regard to the analysis of roots, a point in which Cuetius has victoriously opposed him ; but that on the whole, he more than any other man has contributed to the establishment of fixed laws for phonetic changes, and that in consideration of this fact Pott's Etymological Investigations must be counted among the fun- damental works of comparative grammar, which are entitled to the next place after those of Bopp and Geimm. In regard to the origin of inflectional forms. Pott follows Bopp, express- ing the opinion that Bopp has made the subject of inflection so transparent and clear, that, with the exception of some un- solved minor difficulties , we can sufficiently comprehend its nature and character from an etymological point of view. He, as well as Bopp, considers the principle of com- 3* 36 Chapter II. position the chief agent in inflection, although without wholly rejecting the symbolical explanation. "Language - designation [Sprachbezeichnung]^\ he re- marks, "is either symbolical or kyriological. In declension the variation according to gender [Motion], and the designa- tion of gender are often symbolical ; on the contrary, the mode of expressing case and number is mostly kyriological." (2, page 621.) The inflectional endings of the verb he regards in the main as Bopp does, yet it is worthy of mention that he does not favor the latter's symbolical explanation of the n of the third person plural in anti, but considers it as a pronominal stem (as Schlbichek also did later) , and that he declared the first person plural masi was derived from "I" and "thou". (2, page 710.) He is therefore as decidedly an adherent of the explanation by agglutination as Bopp, even although, as we shall see later, he was inclined to reject the historico-linguistic consequences of Bopp's theory. Among Pott's contemporaries and successors Theodor Benfbt must be particularly mentioned, who, on the whole an adherent of Bopp, in the very first years of his appearance before the public displayed an independent activity in sev- eral different directions. His Greek root- lexicon [Wurzel- lexicon, Berlin, 1839), the forerunner of a Greek grammar which was planned on a grand scale but not subsequently car- ried out, exhibited not only an astonishing copiousness of con- tents, but also the richest power of combination ; yet it cannot be regarded as an advance upon Bopp's stand-point in its con- ception of phonetic changes. Benfey's theory concerning pri- mary verbs, with which he would fain replace what is usually called "root", and concerning the derivation of stem-forming suffixes, will be further discussed below. Here we will only mention the great honor he gained by his works in the field of Indian philology, especially by his edition of the Sa- ma Veda, Leipzig, 1848. His glossary to the Sama Veda was the first work which supplied philologists with reliable material from the Vedic language for convenient use, and has exerted the most salutary influence upon etymological study. BOPP'S CONTEMPORARIES AND SUCCESSORS DOWN TO A. ScHLEICHER. 37 This reference to a book which apx^eared in 1848 brings us to the next period, which must be discussed in a different manner from that which has hitherto been admissible. For in the era which must now be mentioned, if not described in detail, a great number of scholars appear, whose achieve- ments are so mutually involved that it will be wiser to shape our discussion no longer according to persons, but according to the tendencies and aims which now manifest themselves. ScHLEicHBK alouc, who Unites a number of these efforts and brings them to a certain conclusion, will claim an individual treatment. In the period between Pott's Etymologische Forschungen and Schleicher's Compendium a very considerable extension of our knowledge took place, and to this fact we must first turn our attention. Perhaps no extension of knowledge was ever fraught with higher results to philology than that which occur- red in the Indian field. Our acquaintance with Indian liter- ature began with the Indian middle ages, and not till after- wards, when (from about the year 1840) Vedic studies began to flourish, did Indian antiquity become known to us. Through the works of Rosen, Roth, Beneey, Wbstbkgaard, Mullek, KuHN, AtTFRECHT and others, a mass of new and reliable ma- terial was in a comparatively short time brought before the etymologists, who had hitherto been rather scantily provided with Indian lexicographic aids. Wilson's Lexicon (concerning which, V. the preface to the first volume of Bohtlingk & Roth's dictionary, as well as the article of Schlegel in the Indische Bibliothek, 1, page 295 seq.) was far from being a historically arranged lexicon, and the Indian lists of roots are an aid which is attended by peculiar dangers. Even if we could assume that the lists which the Indian grammarians drew up were made and transmitted with perfect accuracy, they could be used only with caution for etymological compar- isons , for the manner in which the Indian scholars denote the meanings is different from that to which we are accustom- ed. When they append to a root the locative of a substan- tive, to determine its meaning, they do not always intend to denote the individual sense, but often only the general cate- gory of meaning under which the verb falls. On this account 38 Chapter II. the critical editor of these lists (Wbstbegaard, Radices linguae sanscritae, Bonn, 1841) uttered the following warning against a too confident use of them : "Ceterum puto cavendum esse, ne ilia grammaticorum de potestate radicum deer eta nimis urgeantur, nam illis nihil va- gius, nihil magis dubium et ambiguum esse potest; sic, ut unum modo exemplum afferam, vocula quae gatau est, unum- quemque motum ut eundi, currendi, volandi etc. indicat, quin etiam exprimit mutationem, quam subit lac coagulando, et nescio quam multas alias." But the assumption made above will not hold good. Na- turally all the roots are not correctly given. If, there- fore , we would proceed prudently, we cannot feel confidence in a root until we have authentic proof of its occurrence in the literature (which we do not possess in the case of many) , un- less some reason can be found why a root would naturally be wanting in the written language, as for instance is the case ■with pard = TtspSofAai. But beside this, they are not correctly transmitted, having been exposed to all the injuries which time is wont to cause to literary products. And this corruption has not only aifected the roots (Westbrgaard, page IX, men- tions no less than 130 roots figuring erroneously with his predecessors, part of which had been employed for compari- sons) , but also the meanings given. We can see how much opportunity there was for error, and as a matter of fact, there has been much harm done through etymological use of un- authenticated roots and erroneously assumed meanings. That this source of error is now closed, is due to the la- bors of the above - mentioned men , first of all , however, to the Sanskrit lexicon of Bohtlingk and Roth, that in- comparable master - work , which was almost as epoch- making for linguistic science in general as for Sanskrit phi- lology. Beside the Sanskrit, the Slavo-Lithuanian and Celtic languages (the latter of which was assumed even by Pott in his Etym. Forsch., 2, page 478, to belong to another family than the Indo - European , but to have been mingled with the latter in pre-historic times) were the object of attention and study. Yet we can say that at the period which now occupies BoPP'g CONTEMPOKABIES AND SUCCESSORS DOWN TO A. ScHLEICHEE. 39 US, the Sanskrit, the classic and the Germanic languages al- ways took the lead. Not only the extension of knowledge, but also the posi- tion taken with regard to phonetic laws, seems characteristic of this period. What I mean is clearly illustrated by a passage from CuRTius' remarks on the extent of phonetic laws [Ueber die Tragweite der Lautgesetze, Berichte der phil.-histor. Classe der Konigl. Sachs. Gesellschaft der Wissenschaften , 1870), which is as follows : "Since the first bold onset of the founders of our science, a younger generation, from 1840 or thereabouts, has had for its watch- word: 'strictest regard for phonetic laws'. The abus- es of which even meritorious scholars had been guilty, through the assumption of weakenings, corruptions, mutilations etc., had engendered a well-founded distrust, which inevitably led to a greater strictness and conservatism in this respect. The results of this tendency, which in this sense is a more rig- orous one, we can certainly call beneficent. More accurate observation of the phonetic changes and their causes, more careful separation of the individual languages, linguistic pe- riods and linguistic varieties, more definite insight into the origin of many sounds and sound-groups have been attained. In this respect we can see much farther and more clearly than twenty years ago, as is most evident from the fact that many an airy assertion formerly propounded has been recog- nized as impossible even by its originators." Finally , we must regard as especially important the attempt to separate the individiial languages more strictly from each other. Bopp did not scruple to confirm an asserted phon- etic change in Latin by a reference to the Armenian. Such freedom could from this time forth no longer be tolerated. Each separate language must be recognized in its own pe- culiarities. In this direction the works of Geoeg Curtius (of whose labors in the general Indo-European field we shall speak later) were of great influence, i. e., his investiga- tions concerning the formation of moods and tenses in Greek and Latin, and his Grundzuge der griechischen Etymologie. It was the aim of this latter book to record the sure gain accruing 1o Greek etymology from linguistic comparison, and this task 40 Chapter III. has been, to use Ascoli's words, executed with that masterly- power in the use of positive, creative criticism which charac- terizes the author. Less fortunate were Corssen's efforts in the Italic field. Benfby [Orient 8( Occident, 1, page 230 seq.) has justly censured the individualizing style of this scholar, whose method of observation must necessarily cause that to be regarded as individually Italic which certainly had belonged to the Indo-European parent speech. Yet it is impossible to deny that Corssen, especially in the first edition of his work, where comparison is not so prominent a feature, contributed to the better knowledge of the Italic languages in a way to deserve very considerable credit. (Cf. on this point Ascoli's admirable verdict in the Kritische Studien, page IX.) Many of the attempts of this period (not all, for Bbnfet's school has always gone its own way) are, to a certain extent, summed up in Schleicher's Compendium. It therefore seems to me expedient just here to devote to Schleicher a some- what more detailed consideration. CHAPTER III. AUGUST SOHLEIOHEE. On our first acquaintance with the works of August Schleicher (born 1821, died 1868) we are compelled to ob- serve that an influence, recognized by himself, was exerted upon this scholar from two fields of science which lie outside the domain of philology, viz., Hegel's philosophy, of which he was an adherent in his youth, and modern natural science, for which in the latter part of his life he showed a passionate predilection i) . Let us try to define the nature and strength ') Although the inclination appeared much earlier, — cf. Formen- lehre d&r kirchenslmvischen Sprache, preface, page VI, note. August Schleicher. 41 of this influence in general, before we follow Schleichek into the details of his investigations. At the outset, in the introduction to his first great work, the Sprachvergleichende Untersuchungen (Bonn, 1848), Schlei- cher shows himself an adherent of Hegel, as we can see from the ideas he introduces there. Language (as he explains in detail) is made up of mean- ing and relation. The former is contained in the root, the latter in the formative syllables. Therefore there can be three and only three classes of languages. Either the meaning alone is designated , as occurs in the isolating languages ; or the sound showing the relation [Beziehung slant] \i affixed to the sound showing the meaning [ Bedeutungslaut] , as happens in the agglutinating languages ; or, finally, the two varieties of sounds form the closest union, as in the inflectional languages. A fourth case is not possible, since the Beziehungslaut cannot stand alone. Now three periods of development must corre- spond to these three norms of the system. We are accordingly compelled to assume that the isolating languages represent the oldest form, that from these have arisen the agglutinating, and from these in turn the inflectional languages, so that the last stage of the process contains the two previous ones. But Schleicher argues further that our actual experience is not in harmony with this theoretical construction, for we find the languages which come within the circle of our experience not in a state of development, but of decomposition ; higher forms do not arise before our eyes, but existing ones fall to pieces. Yet since philosophical construction and the result of obser- vation must both claim credence, the inevitable conclusion is that the two processes in question must be located in differ- ent periods. Languages were formed in pre-historic times, and are destroyed in historic ones. The making of languages and of history are activities of the human mind which mutual- ly exclude each other. The above is a condensed reproduction of Schleicher's reasoning, which recurs, at least partially, in his later works, and was not wholly set aside even by the leaning to natural science which was so strong in his latter years. This is not the place to criticise these views, whose 42 Chapter HI. weakness is self-evident at the present day, but it may be interesting" to observe in bow far Schleicher shows himself dependent upon Hegel. This dependence is evidently slight as regards material. In the first place, the division of lan- guage into the three groups mentioned above was not derived from Hegel, but from experience. Schleicher had worked, it out for himself under the guidance of Fribdrich Schlegel and Wilhelm v. Humboldt. (Cf. the Beitrage of Ktjhn and Schleicher, 1, page 3, note.) Further, the opinion that in- flection sprang from composition followed from Bopp's form- analysis, which Schleicher on the whole favored; and Bopp had likewise originated the theory that it is only possible to observe languages (at least the Indo-European) while in a state of decline. We can therefore recognize the material influence of Hegel only when Schleicher assumes that in the develop- ment of mankind we must distinguish between a pre-historic period, during which the mind was still in dreamy thraldom, and a historical period, in which it awakes to freedom. This classification of human development into a pre-historic and a historical period (language being perfected in the pre-historic period) was always retained by Schleicher, and it is not im- probable that this view was induced by Hegel. While, therefore, but little of the subject-matter in Schlei- cher's writings could be pronounced Hegelian, yet in the early work which was mentioned above, the influence of Hegel is unmistakable in the formulation of the thoughts and the struc- ture of the argument. This influence waned with Schleicher's growing maturity, yet we can still feel it in his later works, and trace it here and there, particularly in his terminology. We accordingly come to the conclusion that the influence of Hegel's philosophy on Schleicher was only a moderate and comparatively superficial one. Schleicher took a different position from most philolo- gists in regard to the natural sciences, inasmuch as he really possessed considerable knowledge of them. He was especially versed in botany. According to scientists who knew him, he was celebrated for his admirable preparations for the micro- scope, as well as for certain productions of horticultural art. As the years went on, these studies and favorite pursuits gained August Schleicher. 43 an ever greater influence upon his philological views. When he walked up and down in his beloved garden, and analyzed forms of speech, the thought must have often occurred to him that the analyzer of word-forms and the analyzer of plants have in reality the same profession ; and when he contemplated the law-ahiding nature of linguistic development, which it was his most earnest endeavour to demonstrate, the idea seemed to him very natural that language was nothing less than an organic being. In his methodical mind these thoughts and impressions took the shape of a serious system, whose axioms are as follows : / Language is a natural organism ; it lives like other organ- isms, although its mode of action is not that of man. The ; science of this organism belongs to the natural sciences, and the method by which it must be treated is that of natural 'science. \ Schleicher set great value upon these axioms, and I would venture to assert that if he had been asked in his last years what in his own opinion constituted his chief service to science, he would have answered, the application of the method of natural science to philology. The judgment of the majority of his contemporaries was difi'erent, and at the pres- ent day it is almost unanimously agreed that Schleicher's three axioms cannot find approval. . Bopp had already applied 1 the expression "organism" to language, but he had simply ' meant that language is not arbitrarily manufactured. Such a figure can be tolerated, but when the metaphor is taken lit- erally, the contradiction becomes evident. Language is not a being, but the utterance of beings ; accordingly, if we are to use the phraseology of natural science, it is not an organism, • but a function. . It will also be found extremely difficult to classify philology with the natural sciences. Since language is manifested in human society, the science of language can- not belong to the natural sciences, at least, if this name is used in the accepted technical sense. . And finally, as regards the method, I feel certain that there is no single method adapted to all natural sciences. Por one part of the natural sciences the application of mathematics is characteristic ; for another, experiment ; for a third, to which biology belongs, 44 Chapteb III. the so-called "genetic" method. And to this last the philolo- gical method certainly bears some resemblance, inasmuch as in both sciences we strive to understand the historical devel- opment of their objects. However, it is not my intention to discuss these views more in detail. There is no necessity, for my immediate pur- pose, of showing whether Schleicher's views are right or wrong, but only how they arose and worked within him. It cannot be denied that Schleicher's fondness for natural science is more plainly traceable in his chief works than Hegelianism. But we can only judge from detailed ob- servation how deep these influences were in separate instances. Accordingly I now pass to a critical survey of Schleicher's philological achievements and views. In Schleicher's first works we can still plainly discern the philosophical atmosphere in which they originated, inas- much as they aim less at a through investigation of details than a systematic survey of a broad field. For in the first part of his Sprachvergleichende Untersuchungen he traces certain influences of / (the so-called "zetacism") through as many lan- guages as possible, and in the second part [Die Sprachen Eu- ropds] he gives the outline of a system of linguistics. Very similar in character is a much later work. Die Unterschei- dung von Nomen und Verbum in der lautlichen Form [Sachs. Ges. d. Wiss., Leipzig, 1865). In addition to these general studies Schleicher began very early to appropriate to him- self a special field, the Slavonic languages, and here he has acquired a reputation of which no change in time or opinion can rob him, Schleicher stands beside Miklosich in this field somewhat as Bopp does beside Grimm in the Germanic one. He more than any one else has helped to illumine the Slavonic languages by the light of comparison. In his Lithua- nian studies he brought a wholly new material within the reach of science, by collecting the Lithuanian forms here and there, as a botanist does his specimens, and preserving them for all time in the herbarium of his grammar. In consequence of his academical duties (in Bonn, Prague and Jena) he was also compelled to devote his constant. attention to the other Indo- European languages, and thus was prepared, in the broadest August Schleicher. 45 imaginable way, for the chief work of his life, his Compen- dium der vergleichenden Grammatik der indogermanischen Sprachen (Weimar, 1861), which we must regard as the crowning glory of his career, since an early death called him away from still greater plans. Schleicher's Compendium is the conclusion of a period in the history of philology, the heginning of which was formed by Bopp's works. This is the reason why the general impres- sion produced by the Comparative Grammar on the one hand and the Compendium on the other is so utterly different. Bopp was obliged to prove the essential identity of the Indo- European languages, while Schleicher regarded it as establish- ed; it wasBopp's place to acquire, Schleicher's to organize. Bopp's attention was especially occupied with what was com- mon to all Indo-European language; it was Schleicher's task to make the individual languages stand out clearly on this common background. Hence the comparative Grammar is a connected narrative, while the Compendium could without much trouble be resolved into a number of separate gram- mars. The author of the Grammar in his demonstration of single points employs principally the form of investigation, which he handles with a natural charm of manner ; in the Compendium, on the contrary, we find almost exclusively the concise and monotonous style of assertion. The older work may be compared with the exposition of an interesting law- suit, the younger with the paragraphs of a code of laws . The difference is less striking when we compare the views which are brought forward in the two books. In the first place, Bopp's theory of the origin of inflection was in the main adopted by Schleicher, although he formulated it dif- ferently. Like Bopp he regarded roots, which in his opinion were inevitably monosyllabic, as the constituent elements of the Indo-European. Like Bopp he distinguished two classes of roots (although he considered it probable, contrary to Bopp, that the so-called "pronominal" roots were derived from the verbal ones). Like Bopp he recognized affixed pronouns in the stem and word-forming suffixes. Only on single points was his opinion different. For instance, in the explanation of the middle endings, where Bopp was uncertain, he expressed 45 Chapter III. himself decidedly in favor of the theory of composition, which he carried out to the smallest detail. He followed Pott in his view of the plural endings of the active ; he held that the element of the optative was not the root I or i, but the pro- nominal root ya (without, indeed, informing us how the mean- ing of the optative could be explained under this assumption) ; and in the subjunctive, which Bopp had not regarded with certainty as a special mood, he discovered a pronominal root a. There does seem to be a marked difference in the inter- pretation given to the idea "inflection", which Schleicher in the Compendium, § 2, defines as follows : "The essence of in- flection lies in the vocalism". These words, which at first sight are very striking, must be understood thus. Schleichek recognises two classes of languages in which the forms ori- ginate by means of composition, the agglutinating and the in- flectional. He finds the peculiarity of the latter in their ability to change the root-vowel for the purpose of expressing rela- tion; so e. g. si(i.i is compounded of i and jxi, the t being changed to si in order to express the relation. The inflectional languages accordingly possess the principle of composition, and beside this the capacity of changing their root-vowel as just shown. But in his definition Schleicher mentions only the latter of these distinguishing qualities. We can readily perceive that beneath this form of the definition there lurks a remainder of Schlegel's conception of inflection , with which at an earlier period Schleicher's views were more in accord ; but this remainder is so trifling in its actual significance that it may be overlooked, and we can justly characterize Schleicher as an adherent of Bopp's agglutination theory. Schleicher also agreed with Bopp in assuming that not the primitive period alone had the capacity to produce new formations by agglutination, but that composition might also appear in the individual languages, in the same way as in the parent speech, as for instance in the Latin perfect. The difference seems widest in the department of phonet- ics, but even here it is not a diff'erence of principle. In prin- ciple Schleicher's stand-point was the same as Bopp's, since it was also his opinion that the phonetic changes of language August Schleicher. 47 do not exhibit development, but decline, and since he admits occasional (although much rarer) exceptions to the prevailing phonetic laws. But there is a very considerable difference in degree between the phonetic systems of the two scholars. What a stately appearance Schleichek's Lautlehre presents, occupying, as it does, half of the whole Compendium, compared with Bopp's scanty and unevenly written chapter, which bears the title "System of writing and phonetics" [Schrift- und Laut- system] ! It was Schleicher's task to sift down and turn to account the great mass of detailed investigations which had been undertaken since Bopp's time by Pott, Benfey, Kuhn, CxjRTius and others. In his treatment of the subject we can observe the progress intimated above. The diiferences of the separate languages are taken into account, all related instan- ces are carefully placed side by side, and the probability of each single instance measured from the result obtained. Thus Schleicher established a long series of carefully weighed, well-grounded phonetic laws, which were destined to serve as a regulating principle for every etymologist, and he has undeniably won great credit by this task of sifting and ar- ranging. Of course all such laws have only a provisional value. For since obvious etymologies form the material from which the phonetic laws are derived, and this material can perpet- ually increase and change , it is always possible that new phonetic laws should be recognized, or old ones transformed. This idea, whose correctness has been amply confirmed by ex- perience (for how much that is new has been discovered by FiCK alone !) was not sufficiently appreciated by Schleicher. This was probably owing to the fact that he himself, with his methodical mind, had no conception of that combining fancy which is indispensable to the discovery of new etymologies, and therefore undervalued etymologizing in general. We conclude from the foregoing that in all important s points which have hitherto been mentioned, the difference ^ between Bopp and Schleicher cannot be called a diiference of principle. But one point still remains, which at all events i brings Schleicher's originality into the clearest light, — I 48 Chapter HI. refer to the reconstructed Indo-European parent speech [ TJr- sprache]. I find the earliest mention of this parent speech in the preface to the Formenlehre der kirchenslawischen Sprache, where we read : "In comparing the linguistic forms of two kindred lan- guages, I try first of all to trace hack both the compared forms to their probable fundamental from, i. e., the shape which they ought to have, leaving out of account the later changes ; or at any rate, to bring them upon the same phonetic plane. Now since the oldest languages of our family (even the Sans- krit) do not exhibit their oldest phonetic form, and since the different languages are known to us in very different degrees of age, this difference in age must first be eliminated as far as possible, before there can be any comparison ; the given quan- tities must be reduced to common terms before we can com- pare them, whether the expression thus obtained be the old- est form which can be deduced for both the compared lan- guages, or the oldest form of one of them." Hence, in comparing two languages, we can either reduce the fofm of one language to that of the other (e. g., Slavonic pekqMa to a Sanskrit * pacantyasya , — v. the work quoted above), or trace both forms back to a common primitive form. The first method, so far as I can see , has very seldom been actually applied by Schleicher ; the second, on the other hand, if for "the comparison of two languages" we substitute the words "comparison of all Indo-European languages", contains the following rule for the construction of fundamental Indo- ^ European forms : from a form which appears in all languages, subtract all that is due to the special development of the indi- vidual languages, and what remains will be the primitive form. An example will make these directions clear. "Field"- in Sanskrit is djras, in Greek aypo?, in Latin ager, in Gothip akrs. Now we know that in Gothic g has become k, and that an a was lost before the s ; thus we obtain from the Gothic the primitive form agras. We know further that the Greek o is derived from a, so that we likewise obtain agras, and so with each language in turn. Hence agras may be re- garded as the primitive form, and by a similar process we de- du.ce the accusative agram, genitive agrasya, ablative agrat, August Schleichek. 49 nominative plural agrasas etc., as well as a large number of pronouns, prepositions etc. All these forms together make up the Indo-European parent speech ; or, expressed in historical style: the parent speech is the language which was spoken immediately before the first separation of the primitive Indo- European race. Schleichek did not always content himself with this simple and clear notion of the parent speech, for he often as- cribes to it a quality which cannot be derived from the pre- vious definition, — the quality of complete integrity of its original structure. An example will best explain what is meant. The nominative of the word for "mother" is in Sanskrit mata, in Greek [J.rjtrip, in Lithuanian mote, in Old Slavonic mati, in Old High German muoter. Nowhere does an s appear in the nominative. Accordingly, by comparing the separate forms we can only obtain the form nidtar or mata (the latter if we assume that the r, as for example in jJ-v^'^Trjp, was in the individual languages introduced into the nomina- tive from the oblique cases), but not the form matars, as Schleichek does. He assumed this form because matar is the stem, and s the suffix of the nominative, and he felt convinced that in the parent speech so-called "phonetic laws", mutual in- fluences of sounds, and similar phenomena, were not yet in existence. But this supposition is quite arbitrary, for if the primitive speech was spoken by human beings, it must have shared the fate of all language, viz., change in phonetic and morphological constitution. There is, then, no reason why we should not ascribe to the parent speech forms like matar or mata. It is true that in a still older period the form may have been matar s , as Schleichek assumes, but then it would be necessary to distinguish the different periods of the primitive language, so that we should not put older and younger forms upon the same plane, as Schleicher seems to have done. The failure to make this distinction has undeniably introduc- ed a certain ambiguity into Schleicher's notion of the pa- rent speech. In the following discussion I venture to. leave this difficulty out of account, and will understand the term "parent speech" only in the sense previously defined, i. e. in the sense originally intended by Schleicher. Delbkuok, Introdnction to tie Study of Language. 4 50 Chapter III. Is it, now, Schleicher's opinion that a historical reality must be ascribed to the forms of the parent speech, taken in this sense? I think the reader of the Compendium will be inclined to answer this question in the affirmative, and will be somewhat surprised to find in the appendices [Chrestoma- thie, page 342) the following remark: "The assumption of these primitive forms does not necessarily imply the assertion that they ever possessed an actual existence." In order to ex- plain this apparent contradiction I will choose the form of in- dependent discussion, proceeding in the following order : I will first formulate the objections which have been brought against Schleicher's parent speech (in the sense described above), and then try to determine their true value. The first difficulty is naturally due to the demand that in the case of a certain form each individual language shall be consulted. This demand, however, can be satisfied only in the rarest instances, for how few are the words and forms which we can trace through all the languages ! But in actual practice this objection has little weight. For we must con- sider that it is possible to point out in all languages quite a number of inflectional suffixes, or at least traces of them; and since we know the phonetic laws which would come in play, we can say in the case of a good many word-stems what their form must have been in a given individual language. A second objection is of a more serious nature. Is it real- ly possible to define the point where the development of each separate language began? Can we determine with certainty whether a certain modification of sound or form belonged to the primitive speech or originated in the individual language ? Schleicher had definite views on this point. For instance, he thought it possible to assert that the parent speech possessed the following sounds : Vowels ; a i u aa ai au aa ai au Consonants : h g gh t d dh P h bh J s V n m r August Schleicher. 51 How did he reach this conclusion? In separate fields the way had been prepared for him, as in the field of the a- vowels. It is well known that the Indo-Iranian group of the Indo- European possesses no S and o, but exhibits an a where the other languages have these vowels. Bopp was only at first of the opinion that S and o originally belonged also to the Sans- krit, and were subsequently lost ; he then adopted the view of Geimm (Grammar, I, 2"* edition, page 594), who in con- nection with his Gothic researches denied that ^ and o were original, so that for the Indo-European there remained three simple primitive vowels, a, i, u. This assumption also gained favor in consequence of the high esteem which the number three is wont to enjoy; Pott, for instance, begins the section on vowels in his Etymologische Forschungen with the remark : "It seems to follow from historical and physico-philosophical grounds that language possesses but three simple fundamental vowel sounds, viz., a, «, ti." Thus the hypothesis of Grimm seemed to find confirmation in all directions, and was accepted by ScHLBiCHBE. i He assumed that the primitive language agreed with the Sanskrit in the simplicity of its vocalism, while the more diversified Greek exhibited a condition of greater development or deterioration. For the consonants, however, an opposite conclusion was reached. The cerebrals of Sanskrit had been early regarded with suspicion, the as- sumption being made that the Hindus had obtained these ex- traordinary sounds from barbarian aborigines; the palatals also were found in many cases to be younger than the guttu- rals, as for instance in reduplication [cakara from kar) . In this point, therefore, the Greek appeared to have preserved the original condition of things, while the Sanskrit had dete- riorated, and the main conclusion was that the rich and diver- sified phonetic material, which the individual languages either exhibit or must have once exhibited, arose by means of various processes of division and multiplication out of a limited and simple phonetic material in the primitive speech. From the analogy of this result Schleicher drew the further conclusion that the phonetic condition at a still earlier period must have been yet more simple : "At an earlier period in the life of the Indo-European 52 Chaptek ni. parent speech, the three aspirates and the three diphthongs with a were wanting ; in the original condition of language, before it had become inflectional, there were no diphthongs at all. Accordingly, the Indo-European probably possessed at first six momentary sounds [momentane haute], viz., three surds and three sonants ; six consonantal duratives [Dauerlaute], viz., three spirants and three so-called 'liquids', i. e. the two nasals m, n, and r [I being a sub-variety of r) ; and six vowels. In a later stage of the language, shortly before the first separa- tion, there were nine momentary and nine vowel sounds. The symmetry of the numerical relations apparent in the number of the sounds must not be overlooked." [Compendium, § 1, note 1.) This theory of development offers a broad field for criti- cism. In the first place, the general statements must be put aside as not conclusive. For the remark that in earliest times the phonetic condition must have been a very simple one, can with equal justice be confronted with a contrary assertion. We notice, indeed, that the individual languages often lose in phonetic volume. Why should it be impossible to assume that the parent speech was richer than any of its daughters? And the symmetry of numerical relation so emphasized by Schlei- OHBK would only be of value if it could be shown to result from the nature of the human vocal organs, which is not the case. Therefore the special grounds in each individual instance can alone decide; and these seem to speak against the ten- dency of Schleicher's hypotheses. On the contrary, it seems as if we must rather assert that the primitive language re- sembled the Greek most closely in the diversity of its vocalism, and the Sanskrit in the manifoldness of its consonantism. For instance, we can iu my opinion declare with certainty that two sets of Ks existed in the parent speech, one of which was liable to palatalization; further, that the parent speech possessed an e [ox a, if that style of writing is preferred) . What is true of e must probably be assumed for o also, and the words of the primitive speech which were reconstructed by ScHLEicHEK would, according to these views, present a very different aspect. Schleicher once indulged in the pleasantry of writing a fable in the Indo-European parent speech, which AuGCST Schleicher. 53 he entitled : Avis akvasas ha ("The sheep and the horses") . According to the newer theories this title would run: Ovis ekivos &2^ (the ^2 ™ this case designating the palatal k of the parent speech) . "He saw" would be represented no longer by dadar/na, but by dedork^e; the accusative of a participle "rid- ing" not by vaffhantam, but by veg\ontm (where the m is syl- labic), etc. Ten years from now the transcription will perhaps have assumed a different coloring. It accordingly follows that "parent speech" is nothing but a formal expression for the changing views of scholars in regard to the nature and extent of the linguistic material brought by the individual languages out of the common tongue. With this definition of "parent speech" the question of the historical value of the reconstructed forms is settled. It can neither be doubted nor denied that the primitive language possessed a great number of words which were capable of grammatical inflection, and a series of unin- flected words. But whether it wore exactly the aspect claimed for it by present investigation, whose stand-point is mirrored in these reconstructions, it is of course impossible to deter- mine. The use and significance of these forms can now be defin- ed. They add no new material to. our knowledge, but they bring more clearly before us what is already known. They have the same value for linguistic research which curves or similar illustrative aids have for statistics, and form besides a means of demonstration whose usefulness must not be under- valued. At the same time, the necessity of postulating funda- mental forms impels the investigator to always inquire whether each particular form which is under consideration must be regarded as original or as a new formation, and does not suffer him to rest satisfied without a thorough mastery of all difficul- ties, phonetic or otherwise. We see from the whole previous discussion that Schlbi- .CHER does not essentially differ from Bopp in principle. His iiiViduality consists in his method of demonstration by means of the parent speech. 54 Chapter III. If we recognize this fact, we can at once answer the question, how far the scientific tendency of Schlbichee was materially influenced by natural science. Of course this ques- tion can only be asked in regard to that part of his system in which he differs from Bopp and other philologists, that is, in regard to the parent speech. And it really would seem as if he owed this to natural science. For Schleicher himself in- troduces his parent speech with the following remark [Formen- lehre der Mrchenslawischen Sprache, page VII) : "This procedure, like the method of linguistic science in general, agrees with the method of the natural sciences, of which linguistic science forms a part. The comparative ana- tomist never compares the form of the skull of two animals by taking the skull of a new-born specimen of the one sort, and the skull of an adult of the other ; if the needful material is wanting, as is often the case in fossil remains, he does just what we do ; according to known laws he reconstructs what is lacking, on the same plane of age with the specimen before him." Nevertheless, I cannot believe that Schleichee was in- spired by the anatomist to emulate his example ; it seems to me more probable that he sought among the scientists for analogies to his own procedure, after it was already complete. For on the one hand, the reduction of forms of different ages to the same plane of age seems to me so natural a thought, and one which would so easily suggest itself from the very task in question, that I should hardly suppose it would have originated in a foreign field ; and on the other hand, I should be inclined, despite my slight acquaintance with anatomy, to assert that in the passage above quoted.ScHLEicHER views the proceedings of the anatomist through philological spectacles. I come to the conclusion, therefore, that in this point also nothing has been borrowed from natural science ; nor has any transfer of the method of natural science taken place. Indeed, the whole idea of a transfer of methods strikes me as rather extraordinary. How can it be possible, after a method of pro- cedure has adapted itself to the peculiarities of one object, to apply it with profit to a different one? Schleicher himself New Endeavors. 55 has not done this. Little as he would himself acknowledge it, he, as well as Bopp and Grimm , Pott and Curtius, is, in the essence of his being, — a philologist. CHAPTER IV. HEW ENDEAVOES. We cannot regard the Compendium as a summary of the collective philological labors of the period. For — to mention only what is most important — Pott displayed his peculiar and partially isolated activity in other paths than Schleicher, and Benfey and his school also formed a separate group. Pott could not become reconciled to Schleicher's reconstruc- tions, and Benfey and his adherents held that Schleicher, by his phonetic laws, wrongfully fettered the movement of sounds. Schleicher himself considered that Curtius and CoRSSEN stood nearest to him in the philological field. Corssen's importance was at that time over-estimated (as re- marked above, page 39) , but we may justly regard Schleicher and CuRTiTJS as the chief representatives of a tendency which has ex,erted and still exerts a powerful influence, not only on the study of the individual languages, but upon all views re- garding the aim and method of linguistic science. For the labors of Curtius were not confined to the Greek field, but in his essay on the chronology of Indo-European linguistic research he furnished an amplified and improved sequel to that portion of the Compendium which treated of the parent speech ; in this publication, he made the attempt (as Schlei- cher had not yet done) to follow out the historical develop- ment of the original Indo-European language. After the tendency described above had for a number of years held the prominent position of interest, the objections which had been raised against it from different directions began to make themselves more strongly felt than hitherto ; 56 Chapter IV. fresh scruples appeared on individual points ; attention was gradually withdrawn, as if wearied, from certain sections which had hitherto claimed especial notice, while other por- tions, which had remained in the shade, became more plainly discernible ; in short, a new tendency began to gain ground, which was partly a continuation of the previous endeavors, partly an attempt to oppose, ameliorate and expand them. The impulses to this new movement did not proceed from one point, but from many, both simultaneously and succes- sively, so that it would almost seem as if it were more correct to speak not of one tendency, but of several diverging ones. Yet I think that the common element in these new efforts is the truly important and essential thing, and with this I will endeavor to acquaint the reader in a few words. I have shown that from the first it was the prevailing / tendency of Bopp's works to explain the origin of grammati- ' cal forms. In the beginning Bopp sought this explanation in each individual language, and said, for example, that the aorist was formed in Sanskrit by composition with as, in Greek similarly by composition with i?, and so on. Now the more plainly it was recognized that the explanation of the forms must be sought, not in the individual languages, but in the common speech [ Gesammtsprache] from which they have all sprung, the more this latter came into the foreground ; and it is therefore perfectly consistent that the primitive speech plays so important a part precisely in the works of Schleicher and CuRTius, who, like Bopp, aim at an explanation of the gram- matical forms. On the other hand, it was also natural that in course of time the objections which could be brought against analyzing a form belonging to the primitive language became more strongly emphasized. The realistic age, which prefers to hold itself aloof from things which cannot be known, has become more and more conscious of the hypothetical nature of such analyses, and we can accordingly assert that among a not inconsiderable number of philologists, all glottogonic hy- potheses, i. e. all attempts to explain the forms of the parent speech and to build up a history of inflection upon them, have come into disfavor. (Meanwhile, Schleicher's parent speech as defined above can of course be defended throughout.) New Endeatoes. 57 lu illustration of tliis view I will quote a few words of Johannes Schmidt, who, after emphasizing the difficulties attendant on the customary explanation of the optative (by composition with the root i or «/«), expresses himself as fol- lows [Kuhiis Zeitschrift, 24, page 320) : "I do not feel myself under obligation to propose a new explanation. It is the task of Indo-European linguistic science to demonstrate what the forms of the parent speech were, and by what methods those of the individual languages have sprung from them. We are in most cases as incapable of interpreting the significant value of the formative elements which are affixed to the so-called root, and for the same rea- son incapable, as the one-sided Grreek grammar was of ex- plaining the elements of Greek words. In this field the re- cognition of ignorance increases from year to year, as befits a healthy science." Another consideration is closely connected with the above. In order to explain the forms of the primitive speech, it is customary to'analyze them into the parts from which they are supposed to have arisen, e. g. dasyati "he will give" into da-sya-ti. Now the same process of analysis can also be em- ployed in those languages which, like the Sanskrit and Greek, have remained upon a tolerably ancient plane of development. Thus Bopp resolves a form like 8o&7)oo'[Ae&a into So-&rj-ao-[j,£&a. Can it, now, be assumed that the affixion of these elements first took place in Greek? Certainly not. The more thorough- ly the comparison of the Indo-European languages has been prosecuted, the plainer the following principle has become : inflection was completed in the parent speech, and only fin- ished forms were transmitted to the individual languages. If this is correct (and who can doubt it?), the question at once arises : how, then, are new formations possible in the separate languages? The credit of proposing this question is due to Mergxjet [Die Entvnckelung der lateinischen Formenbildung , Berlin, 1870) i), and that of answering it, to those scholars who have given particular prominence to the idea of forma- >) V. an essay by the same author, in Neue JahrlUcher fur Philologie und Fadagogik, pages 109, 145 seq. 58 Chapter IV. tion by analogy, especially Whitney, Scheree and Leskien. (Cf. MiSTELi, Lautgesetz und Analogie, in Steinthal's Zeit- schrift fur Volkerpsychologie und Sprachwissenschaft, 11, page 365 seq.). Since in a finished language it is no longer possible to construct new forms by affixion of tbe constituent elements, unless these elements are themselves finished words, all other new formations can only be due to analogy. New formations are imitative formations. With this view, the prin- ciple of analogy came naturally into the foreground in the ex- planation of forms, and many individual forms, as for instance the Latin imperfect, future etc., had to be regarded otherwise than hitherto. Another incitement to progress appeared in the field of phonetics. After the Sanskrit, Greek and Italic languages had for a long time taken the lead, and this, too, in their earliest stage of development, the more modern idioms grad- ually became the object of increased attention. Now these (for instance, the Slavonic and Romanic languages) possess a wealth of phonetic development, joined with a delicate pre- cision in phonetic distinctions, which is fairly astounding. At the same time, the study of the physiology of sounds /^iaw^- physiologie]^ which was again prosecuted with especial zeal, brought to light the manifold richness of their phonetic struc- ture. What was the objection to ascribing to earlier periods the same manifoldness of sound which we observe in the lan- guages of today? Why should it not also be assumed for the common Indo-European speech? Under this hypothesis, rule and order were discovered at many points where hitherto "exceptions" and irregularities had been admitted. The as- sumption of e for the parent speech comes under this head. This hypothesis (already mentioned above in connection with Schxeichee) was reached in the following manner. It is a fact that an e in other Indo-European languages often corre- sponds to the Indo-Iranian « ; e. g. the e in cpspw andyero to the a in hhdrami, the e in re and que to the a in <;«!. Now Cur- Tius showed that the European tongues usually agree in hav- ing e where the Indo-Iranian has a, and since he was as little in doubt as Bopp, Grimm, Pott, Schleicher and others that the a was the original vowel of the two, he assumed that this New Endeavors. 59 primitive a had become e in the European parent speech, (Ueher die Spaltung des a-Lautes etc., in the Berichte der Ko- nigl. sacks. Ges. der Wiss., 1864, pages 9 — 42.) This viewy which long prevailed, sustained its first shock from the discov- ery that the Armenian shares this e. Ought this language to be reckoned among the European ones, or must ^e drop the pre-supposed originality of the «, and assume that an e or a existed in the parent speech, and became a in those few lan- guages which do not retain it, which a is not distinguished (at least in the written language) from other varieties of a , — so that the lack of originality would fall, not on the Euro- pean, but on the Indo-Iranian side? The truth of this assumption appears more than probable when we consider the relation of the Indian (and Iranian) pal- atals to the following vowels. We observe that in number- less instances a palatal springs from a guttural through the in- fluence of a following i (so, for example, ojiyan belongs to the positive ugrds) . We often find that a has the same influence as i, but only the a which corresponds to e in the other lan- guages, as for instance the a of the reduplication-syllable in cakara, from kar "make", which corresponds to e in Greek, German etc., or the a of ca, which answers to a Greek and Latin e, and so on. It follows from these facts, which could be multiplied indefinitely, that the above-mentioned a of the Sanskrit must have borne some resemblance to t, and hence that it must have been e or « , which proves the existence of an e in all Indo-European languages. Now if it can be proved (and it can) that this palatalization which we find in Sanskrit must have extended back into the parent speech, the origi- nality of this e {a) is as surely demonstrated as it is possible to demonstrate any assertions of this nature. Without enter- ing here into further details, I will refer the reader to an es- say by Johannes Schmidt {Kuhn's Zeitschrift, 25, page 1 seq.), comprehending the previous labors of Ascoli, Brugman, FiCK, CoLLiTZ and others, on the whole question of the two guttural series and the European e, in which he discusses the tenability of the new theory with the strictest attention to detail. Beside the discovery of the Indo-European e-vowel we 60 Chapter IV. may range the very probable supposition (originating with Osthoff) that a ''sonant" (i. e. syllabic) r, such as we are acquainted with in Sanskrit, and which is represented in Greek by ap (pa), already existed in the primitive speech. [Zur Frage desUrsprungs der germanischen n-DecUnation, inPaulSfBraune^s Beitrdge, 3, page 1 seq.) According to this view, the a of the Greek aorist erpaTtov can no longer be regarded as an isolated remnant of a former condition of the language, whose preser- vation is due to a certain confusion of the linguistic instinct (CxTUTius, Grundzuge, 5*^ edition, page 52) , but pa is the reg- ular representative of the Indo-European ?--vowel, which is appropriate to the aorist. No less important is Betjgmast's assumption of a syllabic nasal [Nasalis sonans in der indoger- manischen Grundsprache, inCuRTitrs'-S'[j.i, SotTjp etc., and these words naturally formed an associated mass in the mind of the speak- er ; but a root oo or So did not exist in the language of the Greeks. In other cases, on the contrary, even in languages of as great antiquity as the Greek, kindred words are no longer held together by phonetic similarity. The Hindu may have still been conscious of a connection between a^us (cuxuc) and aqvas (iTuiro?), but the Greek certainly could no longer feel the slightest connection between (uxu? and tirTroc. Now the modern languages only differ from the Sanskrit, Greek etc. in The Agglutination Theory. 77 this, that the relation which we find exists in Greek in the case of coxu? and i^rTto? has with them become much more frequent. Although it is thus clear that it is unscientific to speak of roots in the individual languages, it is nevertheless probable that owing to their convenience they will not disappear from practical use in linguistic science. And there is really no ob- jection to the employment of illustrative aids, so long as they are not confused with realities. In postulating these roots the form is naturally of little moment. Whether we say' tpep, or in Curtius' foot-steps, and undertook a very comprehensive analysis of roots in the section of his root-lexicon which bears the title : "Roots and root-determinatives". There he arrives at the following general result : "The primitive root can consist of: 1) a single vowel [a, i, u] ; 2) «-vowel -|- consonant [ad, dp, as) ; 3) consonant or double-consonant -j- a-vowel [da, pa, sa ; sta, spa, sna) . All roots which have other or fuller forms either arose from the primitive roots by phonetic weakening (e. g. M from ka, gi from ga , tu from ta) , or are further formations from them by means of the affixion of determinatives." He tries to adduce an empirical proof for this assertion, by showing that all, or almost all roots whose form does not come under the three categories mentioned above, can without difficulty be traced back in form and meaning to roots which are in conformity with these three norms. To show how he conducts this process I will give an ex- ample : DELBKfioK, Introduction to the Study of Language. 6 82 Chapter V. ka "sound". A«, ha-n canere, "sound", "resound". ka-k "laugh". ka-t "be noisy", "chatter". ka-r "call", "name". kar-k, hra-h "resound", 'laugh", "crash", = kru-k id. kar-d, kra-d "rustle", "resound". kra-p "be noisy", "wail", "be wretched", cf. Skr. karuna "wretched". kru "hear", cf. Aryan krat-u "insight". [kru-k "cry", "crow", "croak", probably de- rived from krak.) kru-s "hear". ka-s "point out", "extol", "praise". kas "cough". Jm "cry", "howl". kv^k "cry", "howl". ku-g "whine", "chirp". ku-d "be noisy", "revile". FiCK has recently [Bezzenber gov's Beitrage, 1, page 1 seq.) modified this theory to a considerable degree, and now discov- ers remnants of syllables in all the "determinatives" assumed by him : "If forms like mak, star, dam were formed by the compo- sition of the primary roots ma, sta, da with a second member, it is quite impossible to doubt that the products of this com- position must have been originally ma-ka, sta-ra, da-ma ; for since elements like k, r , m., i. e. simple consonants, do not exist in the Indo-European , they can never have been em- ployed for purposes of derivation". Therefore it is Fick's opinion that (to repeat the example given above) gdmati must be resolved into gama-ti, and that gama must be regarded as a dissyllabic secondary root, which was formed from the original root by the affixion of ma. Before Fick, AscoliI) (in his Studj 'ariosem,itici, 1865) 1) It is only on account of Fick's detailed demonstration that I have judged it expedient to bring his arguments first before the attention of the reader. The Agglutination Theoey. 83 had brought forward essentially the same arguments, which he has recently taken up again in his introductory letter on the palaeontological reconstruction of language, in his Kritir sche Sttcdien ("Weimar, 1878). There he says (page XXXI) : "We find at the same time that very many radical combi- nations in the Indo-European lexicon, instead of remaining true to their old significance as genuine first elements, genuine roots or original monosyllables, allow of an accurate analysis, by which they are found to be compositions of a really ori- ginal monosyllable with one or more affixed elements, of deriv- ative, determinative or supplemental nature, as we are pleas- ed to call them. Thus these apparent roots are reductions of dissyllabic (or even trisyllabic) aggregates, reductions or inner kernels, which never possessed an actual, independent life, but were only obtained through the union of the old aggre- gates with new accessory elements of difierent derivative or inflectional meaning. So it is true that in the language of the Indo-Europeans before their separation the phonetic group SKID ('cut', 'split', Latin scid-^ Zend scid- etc.) , with the vowel «', did really exist, but we find at the same time unmistak- able successors of the synonymous SKAD (Zend skenda etc.), and of the also S3monymous SKA (SAK-A ; Sanskrit cha, Lat- in sec-) ; and we must in fact go back from skid to a ska-da. For 'run' the Indo-Europeans before their separation had a phonetic group DRAM (Skr. dram, Gr. 5pE[x-) , which, however, is really DRAMA ; DRA occurs in the synonymous dra of the Sanskrit and Greek (l-8pa-v) ; a third synonym, the Sanskrit dru [drava-ii] certainly cannot claim that its u is original. The accessory element of DRAM appears in TRAM (TRA-MA ; Lat. trem- etc.), Avhose true radical foundation occurs again in the synonymous group TRAS (TRA-SA ; Skr. tras, Gr. rpso-, Tps(u), and also in TRAP (TRA-PA; e. g. in Lat. trep- idus). Similarly, for Skr. krt 'cut' (cf. xsiptu) we must go back to KARTA (= KARA), or for Zend stayj-ra 'what of- fers resistance, stands firm' to STA-KA, and so on in count- less other cases." In our judgment of these views we must bear in mind the following considerations. From the existence of both yug and yu, dram and dra etc. , we are easily led to suppose that roots have 6* 84 Chaptee V. been amplified by the affixion of new elements. That these elements originally consisted not of single sounds, but of syl- lables, is also a very natural assumption. There are therefore no valid objections to bring against the supposition that dis- syllabic roots may have existed by the side of monosyllabic ones . We do, indeed, meet with great difficulties in single in- stances. For example, there may be a difference of opinion as to the method of explaining the second a of the Sanskrit pre- sent gdmati. Are we to assume that gama in gdmati is the an- cient dissyllabic root-form, or had gama already become gam in the pre-inflectional period, and was the present gam-a-ti then derived from it, with a suffix «, whose existence Tick does deny, but which, as I shall show below, it is impossible to avoid assuming? It seems to me that the greatest difficulty lies in the meaning. Can we assume that the oldest roots had a meaning so general and so indefinite as "sound"? Is there not, on the contrary, every probability that such conspicuous phenomena as rustling, singing, laughing etc. (the designa- tions for which Fick derives from the designation for "sound") were the first to find expression in language? But it is not my aim here to speak more in detail of these investigations , which are only in their infancy. I merely wished to show by what arguments a modem philologist can arrive at the postulation of dissyllabic roots by the side of monosyllabic ones. I would like, in conclusion, to say a word of Schleicher's opinion concerning the vowel in the root. We have derived from the Indian grammarians the view that diphthongs can be formed from the primitive vowels i and u by means of strengthening, and have also followed their example, for the must part, in ascribing the simple vowels to the roots; e. g. i "go", not ai (or ei) ; ruk "shine", not rauk (or reuk) . But here comes a difficulty. If from the present forms eimi, imds we extract the root «', we ought consistently to assume a root s for dsmi, smds. (Cf. Begemann, Das schwache Prateritum, Berlin, 1873, page IX seq.) Now root-forms such as s, pt, hhs had certainly made their appearance in the words of the prim- itive speech (cf. Brtjgman, Morphologische Untersuchungen, 1, page 11) ; but it is impossible to believe that they could The Agglutination Theory. 85 have been present in the root-period as independent linguistic elements. We are rather forced to assume for the root-period the forms as, pat, bhas, or es etc., and therefore not i, hut ai (or ei] . This involves an inversion of the previous theory of vowel-strengthening, vchich in fact has taken place in the case of several scholars. (Cf. Paul & Braunb, Beitrdge, 6, page 408.) Yet a systematic demonstration is still wanting. For our present pur-pose, we can at least draw the conclusion that Schleichee's view of the root-vowel cannot stand as an in- evitable and final result. So much for the form of roots. The most important points for the reader to bear in mind are the following. We have only words. We extract the roots from them by grammatical operations. But in these we can err, and opinions may change as to what is correct and what false. The same is accordingly true of the form of roots as of the form of the words in Schlei- cher's parent speech. If Bopp's analysis in general holds good, it is certain that so-called "roots" were the words of the primitive speech in a period previous to inflection ; but the form attributed to the individual roots merely exhibits the ' opinion of scholars regarding the method of analyzing the transmitted words of the Indo-European languages. II. THE NOUN. A. Stem-forming suffixes. ■ It is well known that in the Indo-European there are noun-forms originating from the immediate affixion of the case-sign to the root, e.g. duc-s ; while the majority have cer- tain elements between the root and the case-sign, which we call stem-forming suffixes. These consist now of a simple vowel, now of a consonant and vowel, like ta, ma, ra, or of a vowel and consonant, like as, or they have a fuller form, like tar, tama, mant etc. Bopp's opinion concerning those suffixes ( which consist simply of a, i or u was at first undecided, and somewhat in accord with Schlegel's view, as we see from a passage in an academical essay of July 28, 1831 (page 15) : "The meagre form of these suffixes leads us to easily over- look the ancient composition, in the case of those verbal roots 86 Chapter V. which by their agency are made into words, introduced to life and clothed with personality. It may be preferable to regard these sounds as the feet, with which a root is endowed, or which have grown to it in order that it may move upon them in its declension ; they may also be regarded as spiritual ema- nations of the root, which have come forth, no matter how, from its inner being, and have but the semblance of indivual- ity, since they are really one with the root, or merely its organically developed flower or fruit. But I prefer the ex- planation which is the simplest, and which is supported by the genesis of other linguistic families i) ; and since nothing is more natural than that word-formation, like grammar in general, should, on the whole, depend upon the union of one significant element with another, it seems to me hardly pos- sible to doubt that the a. for instance, iu ^ dam-a 'subduing', 'subduer' is intended to represent the person who possesses or exercises that quality which is designated by the root ^ dam ; ^ dam-a is therefore as it were a third person of the verb, in a nominal (i. e. substantive or adjective) state, independent of time-determinations." This theory is brought forward with greater certainty (as remarked above) in the Comparative Grammar, where the majority of the stem-forming suffixes are derived from pro- nouns, while he attempts to trace a portion (e.g. -tar] back to predicative roots. Pott follows Bopp's opinion in the main [Etym. Forschungen^ 1^* edition, 2, pages 454 seq.). Schlei- cher and CuRTius differ from him in giving up the derivation from predicative roots; for example, they would assert that tar was made up of the two pronominal roots ta and ra. (Cf. also KxJHN in his Zeitschrift, 14, page 229.) Scherer, on the other hand, took up arms for the predicative roots, and was in favor of granting much wider scope to this kind of deriva- tion than Bopp did, so that he considers it possible, for ex- ample, to connect the suffix va with the root av "satisfy one's self", "fill". It seems a matter of course to the adherents of Bopp's agglutination theory that in attempting to explain the stem- ') Previously (page 14) the Semitic family had been brought up for comparison. ' The Agglutination Theory. 87 forming suffixes recourse should be had to Bopp's two root- classes, or to one of them. I must, however, confess (in agreement with Scherer) that I can only form a clear idea of the derivation of suffixes from predicative roots, since we have an excellent analogy before us in support of this deriv- ation, in the shape of our German suffixes -bar, -heit, -thum. It is true , the assumption that pronouns are discernible in many suffixes is favored by the identity of form, or the simi- larity they bear to pronominal roots, but it is difficult to find the connection of meaning. We can say that the pronoun be- tokens the person or thing in general, which is afterwards more accurately defined by the predicative root to which the pronoun is affixed (so Windisch in Cvb.iius^ Studien, 2, page 402) ; or that the pronoun points, like an article, to the com- pleted word (so CuRTiTJS in his Chronologie) ; but it must al- ways appear strange that so many suffixes with almost the same meaning appear side by side, and that it is impossible to discover in these suffixes aught of the specific sense of the pronouns. Under these circumstances, we cannot be surprised that attempts have been made to explain the stem-forming suffixes in a different manner, as by Benfby, and also, with exclusive attention to certain forms of suffixes, by Scherer and Tick. Benfey has expounded his theory in several different places : in his essays in the Kieler Monatsschrift of the year 1854 ; in his short Sanskrit Grammar ; in various passages in his periodical, Orient und Occident; and in the briefest and clearest form, in fo article in the ninth volume oi Kuhn' s Zeit- schrift, entitled : Ein Ahschnitt aus meiner Vorlesung uber ver- gleichende Grammatik der indogermanischen Sprachen. The practical application of the theory can be seen most readily in Leo Meyer's Vergleichende Grammatik der griechischen und lateinischen Sprache, 2°* volume, Berlin, 1865. This theory of Benfey can be summed up as follows. The suffixes, which in the transmitted languages have such varying forms, were not different in the beginning; on the contrary, it is very probable that all, or almost all, must be derived from one fundamental form ant, which appears in the present active participle. But this ant itself is a metamor- gg Chapter V. phosis of the third person plural in anti. Accordingly, from hhdranti "they carry" came bharant- "carrying", and from this, Iharor- "carrier" etc. For the original ant has undergone a long succession of phonetic changes ; ant by weakening became at^ then was corrupted into aw, and farther into a; at was transformed into as, and an into ar ; a was changed to i, and so stems in it, in and is arose ; further, through af&xion of "the pronominal theme a" were obtained anta, ata, ana, ara, asa and isa ; and so on. Those suffixes, also, which begin with V or m, like vant and mant, probably belong with the above in respect to origin. For perhaps vant was derived from a third person plural vanti, which belongs to a perfect with v. But this perfect with v is compounded with hhu "be", and the v is the last remnant of bahhuva. The suffix mant, on the other hand, is supposed to have sprung from tmant, which came from tvant; but tvant itself is perhaps a participle from tu "be strong". (Cf. Benpey, Kurze Sanskrit Grammatik, § 336, re- mark, page 212.) This tvant then became difi'erentiated in course of time, so that it developed into tva on the one hand and mana on the other. If, now, all these assertions could be proved, and all, or almost all suffixes accordingly traced back to ant, which in turn comes from the third person plural, it would at the same time be demonstrated that all nouns are derived from verbs, and thus the hypothesis of the primary verbs, which was men- tioned above (page 78), would be justified. Against this theory, as we have just sketched it, we find the following weighty arguments : First : It is impossible to see clearly how the participle could have arisen from the third person plural. It would be easier to comprehend how the reverse could take place. (V. below, under C. in the following section.) Secondly: In the changes which the suffixes undergo, phonetic processes are assumed which cannot elsewhere be shown to occur. It is also a questionable assumption that one and the same form could have developed under like conditions into two wholly different shapes, as e. g. tvant into tva on the one hand and mana on the other. Thirdly : If all nouns can be finally traced back to for- The Agglutination Theory. 89 mations with ant^ we must assume that the suffixless nouns so frequent in the oldest Indian literature, like dvis^ ud etc., once possessed suffixes and then lost them (of course at a very an- cient period) . Bbnfey does make this assumption, but so far as I see it can be supported by nothing except the very neces- sity of his system. In conclusion, it must be remarked that after all it is impossible to derive all suffixes from ant^ and that Benfey himself has to make occasional use of pronouns as one source of suffixes, i) For the above reasons I cannot agree with Benfey's view, but it is self-evident that in rejecting the hypothesis as a whole, we do not necessarily give up every derivation of one suffix from another. Whether such a derivation shall be as- sumed or not, must be especially considered in each individual instance. ScHERBE, whose general views regarding suffixes have already been mentioned, originated the hypothesis that a number of suffixes were really signs of the locative, that is, that the stems formed by them were locatives. Thus he ob- serves in regard to the suffix a : "Those who assert that a gives a substantial sense to the root, that it is the universal «V, or in regard to persons the universal Ae, move in such a dizzy height of abstractions that I cannot follow them. All my ideas of language rebel against this. I regard the a of stem-formation as no other than the a of word-formation, i. e. declension. We know its locative meaning and prepositional application, which starts from the idea of union with anything. But what is the simplest and clearest way of denoting the possessor or accomplisher of an action, state or quality? What more so than to say that he is in this action, this state, this quality, he is united with them?" [ZMr Geschichte der deutschen Sprache^ 1^* edition, page 331.) I object to the above theory that the accomplis^ier of the action, the possessor or exerciser of the quality is really not expressed at all (for a hhar-a would thus mean "in carrying", but not "one who is engaged in carrying") ; and above all, I 1) A detailed criticism of Benfey's view, with which the above remarks are in harmony, is given by Zimmer, Die Nominalsuffixe a und a (Stras- bourg, 1876). 90 Chapter V. would state that I am persuaded with Kuhn (in his Zeitschrift, 18, page 365 seq.) that a locative suffix a, such as Schbrer assumes, cannot be shown to exist. Nor does it seem to me that ScHEEER has in general made it appear probable that declension was prior to stem- formation, so that I am not pre- pared to accept the explanation that a stem-forming suffix was derived from a locative. Finally, Fick (who must be mentioned third in this con- nection) disputed the entire existence of an a-suffix, in an essay in Bezzenberger's Beitragie (1, page 1 seq.). He starts with the assumption that those stems to which the suffix a was previously ascribed, are at bottom identical with certain present stems, as e.g. od[io? with the present stem 8sp.o- in SsfjLOixev. Then, in accordance with the root-theory described above, he resolves these present stems in a difi'erent manner from the usual one, separating Ssjio- not into Ssji-o, but into 8£-[io, Indo-European da-ma ; and by adopting a similar divi- sion in every case, he becomes convinced that a normal stem- forming suffix a never existed. But this conclusion leads to the greatest difficulties . Consider for example the following : are we really to resolve the roots" a® "refresh", as "be", an "breathe", am "oppress", and a number of others of like for- mation, into a-va etc., assuming a as their foundation, and therefore as their simplest root-form? Under this supposition the oldest language could hardly be characterized as intellig- ible. If Pick's method were mathematically certain, it would be impossible to avoid adopting this extraordinary conclusion, as Bbzzenberger does [Gott.Gel.Anz., 1879, article 18, page 558) ; but as it is, the correctness of the method must be doubted, in consequence of a result so difficult to accept. So I cannot make up my mind to withhold the name of suffix from the element a ; and M^e shall see below that the occur- rence of a in tense-formation is also no sufficient ground for denying that it can possess the quality of a noun-suffix. I must therefore acknowledge that none of the theories mentioned above is more to my taste than that of Bopp. Whether, indeed, it ever will be possible to attain to more than a certain degree of probability in this field, can be reason- ably doubted. The Agglutination Thegey. 91 I must remark 'particularly, in conclusion, that in the in- dividual languages exactly the same is true of the reality of stems as of the reality of roots. Stems can have existed only in the primitive speech, before the development of cases. If, notwithstanding, we postulate noun-stems in Greek, Latin etc., this occurs merely from practical considerations. B. Case-formation. If in our consideration of the Indo-European cases we use the analogy of the declension in the Finno-Tartaric lan- guages, we easily arrive at two suppositions, which are also recommended by their naturalness, viz., the supposition that I once, in the Indo-European itself, every case had only one / and the same sign in all numbers ; and that there was a gen- eral plural sign. But it is not possible, by means of the case-suffixes which actually exist in the Indo-European lan- guages, to establish the correctness of these two suppositions, by which several scholars, consciously or unconsciously, have been influenced in their attempts to explain the case-suffixes . We not only find the most various signs for the same case in different numbers (e. g. as and sya in the genitive singular, and am in the genitive plural), but there are also different signs for a case in one and the same number (e.g. in the lo- cative singular) ; and Schleichek, despite all his efforts, is by no means able to prove the former existence of the plural s in all the cases of the plural. On the other hand, it cannot be denied that there is something to be said in favor of the two above-mentioned hypotheses , and it is therefore natural to suppose that the original form of the Indo-European declen- sion has been distorted almost beyond recognition. Reasons for such a distortion could be easily found. It is more than probable that the Indo-European originally possessed many more cases than those we now find in the noun-declension of the Sanskrit, and it is therefore possible that where we think we discover several endings of one and the same case, there were originally several cases, and that the endings are lost which would give us the missing parallels to those still pre- served. 92 Chapter V. In such a hopeless state of affairs it does not seem expe- dient to examine the attempted explanations in detail ; I will content myself with briefly indicating the two main tenden- cies which can be followed in the explanation. We can either assume that the case-suffixes were affixed in the beginning in order to denote something similar to the present cases, and that they contain pronominal, or pronominal and preposition- al elements; or we can assume that stem -forming suffixes developed into case-suffixes, so that, for example, the geni- tive in -sya would be nothing but a stem used as an adjective. This latter opinion is adopted by Curtius for some cases, by Abel Bbegaigne [Mem. de la soc. de linguistique, 2, page 358 seq.) for the majority, by Ludwig for all. I cannot see what serious objection there can be to grant- ing a certain amount of latitude to both theories [as Curtius does) , but the uncertainty is here so great in every case, that after repeated consideration of the whole question (to which I have been constantly led by my syntactical labors) , I have arrived at no other solution than an ever recurring non liquet. III. THE VERB. In the present investigation the attempt of course will not be made to give a history of the origin of the verbal system, so that in this connection much of what has been discussed in Curtius' Chronologie, and recently in my Grundlage der grie- chischen Syntax [Synt. Forschungen , 4) , can be passed over in silence. The question here is only this, how far the agglu- tination theory can be applied in the case of the verb. I shall therefore treat only of: A, the tense-stems ; B, the mode- stems ; and C, the personal endings. A. The tense-stems. Among the tense-stems we have first to consider the manifold form of the present stem. Of the syllables which are characteristic of the present stems Bopp speaks in his Conjugationssystem, page 61, as fol- lows : "In Greek, as in Sanskrit, certain accidental letters are appended to the roots, which are only retained in certain ten- The Agglutination Theory. 93 ses, and disappear in the remainder. We might, as in Sans- krit, make this the basis of classification into different conju- gations, which would then mostly coincide, in their character- istics, with the Sanskrit ones." What Bopp says in the Comparative Grammar, § 495, shows a great advance upon this former stand-point. The pas- sage in question is as follows : "It is hardly possible to say anything positive concerning the origin of these syllables. It seems to me most probable that the majority are pronouns, by means of which the action or quality, which was expressed in the root in the abstract, becomes something concrete ; for example, the expression of the notion 'love' becomes an ex- pression for the person who loves. But this person is more strictly defined by the personal ending, whether it is 'I', 'thou' or 'he'." Here we find an intimation of what Bbnfey and Kuhn afterwards announced with respect to the present stem with nu, viz., that this is really a noun-stem, and that therefore the present-stem dhrsnu in dhrmumds "we dare" is nothing but the adjective dhrsnus 'Taold". This explanation was then extended to other present stems, especially to those which end in a. Ac- cording to this theory, we see in the o/e of Xsyo-jisv, Xsys-Te; (psuYo-|J.ev, cpsoYS-TE not a union-vowel [BindevocaT\ , which is interpolated for euphonic reasons, or which (as Pott assumed) represents the copula, but the noun-suffix a of which we have spoken above . Whether the same view shall hold good for all present stems is a question on which opinions differ. Curtius, for example, sees in the present sign ya the verb ya "go" ; others the noun-suffix ia. (Cf. Beugman, Zur Oeschichte der prdsensstammhildenden Suf/ixe, in the Sprachwissenschaftliche Abhandlungen, hervorgegangen aus Georg Curtius^ grammati- scher Gesellschaft zu Leipzig, Leipzig, printed by Hirzel in 1879.) At all events, according to this theory the great major- . ity of present stems would be really noun-stems , with the ! personal endings appended to them in the same way as to roots, so that, for instance, the same element would exist in ayo-ixev as in ayo-? "driver", and an original ageti would re- ally mean : "he is driver". FicK opposed this view in two articles in the first volume 94 Chapter V. of Bezzenbeeger's Beitrdge, one of which has already been mentioned. He first establishes the fact that noun-stems and tense-stems often coincide (overlooking, at that time, the dif- ference of vocalism, such as exists between 8o|xo-? and 8e|io-|j.£v, which must certainly be traced back to the parent speech) , and concludes from this that it is unlawful in such cases to speak of especial noun-suffixes. Now from the simple fact of the identity of noun-stems and tense-stems it is impossible to draw this conclusion, for this identity may have arisen from the subsequent assimilation of the independently formed noun-stem to the tense-stem. But this identity is not Fickl's only ground for his objection to certain nominal stem-forming suffixes ; in addition to this, he seems to be influenced by the idea that the tense-stems were always prior to the others. I say "seems", because, so far as I can see, he has not expressed himself clearly on this point ; yet we can find a number of in- dications which tend in that direction, as for example : "Ipoc, [i.aj(7] and Poaxo? are nothing but the verbal forms used as nouns" ; or : "the proof that the so-called nominal a-stems are identical with verbal a-stems", — in which clause it must be noticed that only the noun-stems, which Fick in general handles with a certain irony, receive the epithet "so-called". He further speaks of the "nominal shading of e into o" (page 14) ; he accordingly looks upon the vowel of the verb as ori- ginal. If, now, the verb-stems are prior to the noun-stems, the question naturally arises, whence do these elements of the verb originate, which may not receive the name of suffix '? For the suffix a Fick made the previously mentioned attempt at an explanation (page 82), but for ia (which he treats in the se- cond article) such an attempt is wanting. Accordingly, before we can pass definite judgment on Fick's actual theory, we must wait until he has perfected his system in this direction. At the point which present investigation has reached, the affair seems to me to take the following shape. It is obvious i, that the prototypes of certain tense-stems and certain noun- stems are the same. Whether, now, we are to assume that these prototypes possessed a character which was neither ver- bal nor nominal, i. e. such a sense as we ascribe to roots (which is Schleicher's opinion) ; or that they were originally nouns, The Agglutination Theory. 95 which adapted themselves to the verhal system; or verbal stems which were used as nouns, — this is a question which each one must answer in accordance with the idea which he has formed of the development of Indo-European inflection. I pass on to the aorist and future. As shown above, it was principally in consequence of a scholastic error regarding the three parts of speech that Bopp was led to his hypothesis that the root as inheres in the s-ao- rist and the future. The origin of the hypothesis cannot, therefore, be quoted in defence of its correctness. Let us now consider whether other reasons can be adduced in support of it. Bopp finds such a ground in the circumstance that the s appears twice in one form of the Sanskrit aorist, e. g. in dya- sisam from ya "go", which did, indeed , favor the assumption that the s belonged to a verb. Brtjgman (Cuetius' Studien, 9, page 312) objects to this view, first, that it is difficult to see what purpose the reduplication can serve here, and secondly, that from the. stand-point of the Sanskrit forms an easier and more natural explanation is ofi'ered. There are in Sanskrit the aorists ayasam dyasis dyaslt, and dvedisam dvedis dvedU. Is it not very natural that after the analogy of dvedisam, a first per- son dydsisam should be formed to dyasis'! I consider this sup- position especially probable, because the existence of this aorist is only proved for the Sanskrit. i) I cannot, therefore, regard it as an established fact that the double s of dydsisam has any weight in fa.vor of Bopp's hypothesis. On the other hand, it cannot be denied that Bopp's hypo- thesis possesses considerable intrinsic probability. For it is a very natural supposition that beside the direct inflection of a verb, the indirect one, formed by affixing forms of the auxil- iary verb as , could also be employed. (Various views can meanwhile exist concerning the nature and significance of this composition; cf. Cuetius, Ohronologie, pages 55 and 64.) This assumption cannot, indeed, be proved, and it is 1) Bezzenbeeger, SeitrSge, 3, page 159, note, is of a different opinion. But cf. Brugman's reply, Morph. Unters., 3, page 83, note. 96 Chapter V. therefore not surprising that another has been brought for- ward, namely, by Ascoli (cf. Cuetius as quoted above, and Kuhris Zeitschrift, 16, page 148), who is of opinion that the aorist stem, as well as the present stems (discussed on page 93) , has, perhaps, a nominal character. But the aorist stem by no means furnishes so plausible a foundation for the hypothesis as the present stems, and accordingly this supposition seems to me improbable. The future is, in the main, subject to the same judgment as the »S'-aorist. B. The mode-stems. Johannes Schmidt has demonstrated (v. Kuhris Zeit- [schrift, 24, page 303 seq.) that the sign of the optative in \ Indo-European was ia and t, with the distinction that ia occurs wherever the syllable has the main accent, and l where this is not the case. Accordingly, we shall have to assume that ia is the original form of the mode-element, and I a contraction of it. Can, now, this ia be considered identical with the Sans- krit verb «/a? This view, which is on the whole that of Bopp, is opposed by a weighty objection (also emphasized by Schmidt) in respect to the meaning involved. The first person can be explained very well in this way, but not the second and third ; thus it seems as if hanyas, under this supposition, can only mean "thou wishest to kill", and not what it in reality signi- fies, — "I wish thee to kill". The question arises, however, whether we shall allow the whole hypothesis to be shipwreck- ed on this difficulty. We could perhaps assume that the mean- ing of the first person influenced that of the second and third, or we could regard the future significance as the original one, and derive the wish from it. (Cf. Synt. Forsch., 4, page 115 seq.) Postponing to another opportunity the further discussion of this difficult question, I will content myseK with having here intimated the various possible methods of explanation. In regard to the subjunctive, of which a is the well- known sign, ScHERBR sees in the a of hdnati "he shall kill" the same a in which he recognizes a locative suffix. Such an explanation would, it is true, correspond to the sense of the subjunctive (for hana would then mean "for killing") ; but as The Agglutination Theory. 97 already remarked (p. 89), I cannot grant the existence of an Indo-European locative suffix a. Scherer's attempt cannot, therefore, bring into disrepute the view of Curtius, that the ^subjunctive is in form nothing but an indicative ; i. e. the sub- junctive hdnati has the same formation as the indicative bhdr- ' ati. Curtius explains the meaning of such indicatives as originally durative, and tries to derive from this the notion of the subjunctive, on which point I have expressed my agree- ment with him in Synt. Forsch., 1. I will now acknowledge, however, that there is no necessity of assuming such an inter- mediate stage of meaning, and I would therefore prefer not to make that the basis of a derivation of the subjunctive from the indicative ; but the external similarity of forms like hdnati and hhdrati still seems to me a very strong ground for assum- ing their original identity. I am inclined to agree with Cur- tius in regarding the subjunctive with a as a sort of formation by analogy. In the distribution of meaning to the different persons the same difficulty exists in the subjunctive as in the optative. C The personal endings. I have previously (page 71 seq.) characterized as probable the assumption that pronouns inhere in the personal endings of the verb, and shall not at present return to the arguments for and against agglutination ; I will simply bring forward what seems worth discussing within the bounds of the theory itself. In the first place, it must be remarked that not all the scholars who consider affixion as on the whole probable are willing to accept it in the case of all the persons. There is a I division especially in regard to the explanation of the third / person plural of the active. The resemblance between the present active participle and this person is so conspicuous that there is a strong inclination to seek for some genetic relation between the two forms. Benfey made this attempt by deriv- ing -ant from -anti. I have already (page 88) declared that I am not of his opinion. Ascoli and Brugman have adopted the opposite method; the latter says [Morpholog. Unters., 1, page 137): Deibkuok, Introduction to the Study of Language. 7 98 Chapter V. "Who knows that bhdranti is not the stem of the participle (bhdrant), which our Indo-European forefathers used as third person plural, and to which later, although still in the period of the primitive speech, they affixed -* after the analogy of hilar ati'C ■ It is difficult to decide whether the greater probability lies on the side of this view or that of Pott (i. e. that two pro- ) nominal stems, na and ta, are contained in the ending -nti]^ leaving Bopp's theory wholly out of account, according to which n indicates the plural in a symbolic way. Scherer goes far- ther than the above scholars, and considers that the third per- son singular is also of nominal origin, i. e. that it is the locative of a participle . But there is no participle which stands in so close a relation to the third person singular as the pres- ent active participle does to the third person plural, so that the customary explanation seems to me the most natural, ac- cording to which the stem ta (which adapted itself to mi and si in respect to form, as well as in its lack of distinction in gender) inheres in the suffix ti. (Cf. also Kuhn in his Zeit- schrift, 18, page 402 seq.) It therefore seems probable that the three endings of the singular and the first two of the plural (the dual we leave out of the discussion) must be regarded as pronominal roots (which combine with the verb in a more general sense than could be expressed by one of the later cases) , while the possibility must be held in reserve that the third person plural was originally nominal (like the Latin amamini) , being subsequently added to the system of endings, and assimilated to the other forms, i) All suppositions respecting the processes of composition, change and mutilation which the personal endings probably underwent in the parent language are amenable to grave ob- jections. If we assume — to give merely one example — that si was derived from tva, there is no proof that this cannot have taken place ; but neither can any analogous process in the prim- itive speech be quoted in support of this assumption, which 1) This supposition is made in regard to the imperative suffix -tdt, which was explained as an ablative, first by Scherek, and afterwards by Brugman [Morph. Unters., 1, page 163). Yet the transition from the ablative meaning to the imperative is difficult to find. The Agglutination Theoky. 99 rests simply upon the intrinsic probability of the supposition that all the suffixes of the second person belong to one stem. Now this probability is not so overwhelmingly great as to ex- clude all doubt. For why, Brugman asks [Morph. TJnters., 1, page 135), would it not be just as possible to assume two stems for the pronoun of the second person, as for the pro- noun of the first person, where the attempt would certainly not be made to trace back forms like nas and vaydm to the same stem-form? Equally unsatisfactory is the explanation of the middle endings by means of the double affixion of pronouns. It is true that their connection with the active endings is indubi- table, but the method of development of the separate middle forms can hardly be established with certainty. The follow- ing, difficulty must be especially considered. Schleichbk, and CuRTius explain the separate forms independently, assuming that the process of composition and mutilation has taken place in the case of each one. But is it not quite as natural to as- sume that the like endings are partly due to a process of bor- rowing? The other theory', which discovers a vowel-strength- ening in the ai of the middle, cannot command our unqual- ified approval. I must accordingly hold to the opinion express- ed in the Synt. Forsch., 4, page 69, viz. : that none of the proffered explanations is secure enough for us to be able to erect hypothetical structures , syntactical or otherwise, upon it. And the same is true with regard to the other questions which come up in this connection. In each separate instance we seem to find that the means at our command are not suffi- cient to enable us to choose with certainty between the differ- ent possibilities of development which are open to us. We must also bear in mind that the forms which we deduce by comparing the individual Indo-European languages havcja long period of development behind them , a development which has perhaps so metamorphosed the forms in question as to render it impossible to recognize their original character. 100 Chapter V. We have already found in our discussion of the notion "root" that there are two periods to he distinguished in the history of the Indo-European, viz., the pre-inflectional or root- period, and the inflectional period. Bopp, it is true, did not express this idea in direct terms , and Pott even rejected it (although inconsistently, as we have seen) , but we have shown above (page 76) that it is the inevitable consequence of Bopp's analyses. The inflection itself, however, cannot have attained its completeness in a moment, but must have developed by difierent stages, and hence the inflectional period must be sub- divided. Credit is above all due to Curtitjs for having, in his Chronologie, brought into especial prominence the idea that in the development of language, just as in the stratification of ; rocks, difierent layers must be recognized. But it is another question whether he (or any other, as ScHEEEE, for example) has succeeded in defining with any degree of probability the periods through which the formation of Indo-European inflection has actually passed. As may be inferred from the opinions expressed in this chapter, I do not feel myself in a position to discuss this question. Every hy- pothetical structure presupposes the existence of a number of single hypotheses, which may be regarded in themselves as securely established, and can then serve as support for the less certain ones. Now after having adopted a more or less skeptical stand-point in regard to each of the individual form- analyses, I must draw the conclusion that no structure can be reared on such a foundation. I must therefore confine myself to the assertion that inflection undoubtedly developed grad- ually, and not instantaneously, while I question whether the material we possess is sufficient to enable us to define the pe- riods of its development. The affair would, indeed, present a different aspect if we were in condition to amass new material ; and Ascoli has made this attempt. This distinguished philologist, who is at home in the Semitic as well as Indo-European field, assumes that the Indo-European and Semitic parent speech were derived from a common source, and that they even possess certain noun-stems and the rudiments of declension in common. Were this assumption correct, it would prove that the inflec- The Agglutination Theoky. 101 tion of the Indo-European began with the formation of noun- stems. I have too little familiarity with the Semitic field to pass judgment on AscoLi's reasoning, and must therefore, to my regret, content myself with referring the reader to Ascoli's own demonstration (most accessible in Kritische Studien, page 21). Having concluded our special discussion, we will now turn back to the beginning of this chapter, and inquire : has the agglutination theory been verified in individual cases'? I can scarcely believe that the patient reader, who has followed me through the whole of the above demonstration, will reply with a confident "yes". For in the individual analyses a cer- tain probability, at best, and not infrequently an empty "non liquef has been the result. Accordingly, at the end of a long and toilsome pilgrimage we find ourselves no nearer the goal. / Even now we cannot go beyond our previous assertion, that I the principle of agglutination is the only one which furnishes an intelligible explanation of the forms. There is nothing else we have met with which deserves the name of principle, certainly not the so-called ''symbolical'' explanation, in which Eopp in some cases takes refuge, and for which Pott exhibits a still greater partiality. I do not feel competent to consider this method of explanation more in detail at present. For so far as I can see, it is so subjective that a discussion ^ro and con cannot be instituted. Since, now, after our whole discussion the principle of agglutination is all that survives, the question arises whether it would not be better to relinquish philological metaphysics altogether, and confine ourselves to what can be really known ; that is, whether we shall not define as the task of Indo-Euro- pean philology the deduction of the fundamental forms (in Schleicher's sense), and the explanation of the individual forms from these. As we saw above, Johannes Schmidt has expressed an opinion which tends in this direction, and many philologists certainly agree with him. But I do not believe that this view will become general. The attempts to analyze the parts of speech do not, after all. 102 Chapter V. depend upon the arbitrary decisions and fancies of scholars, but are founded upon certain linguistic facts (as, for example, the resemblance of the personal and stem-forming suffixes to certain pronouns , and the like) , and therefore will probably be repeated in future. Whether, indeed, in after times a more satisfactory result will be attained, it is not the province of the present to decide. CHAPTEE VI. PHONETIC LAWS.i) After having briefly shown in Part First how the notion of phonetic law has been constantly increasing in importance in the field of philology, I now pass to its discussion, which I shall handle in the following manner : I shall first explain the stand-point of Geokg Curtius, and then append my own treatment of the subject. I do not aim to say anything new in this discussion, but will merely strive to give a brief though comprehensive outline of what has been said by others. In order to fully appreciate the stand-point of G. Cue- Tius, we must remember (what may be easily forgotten now- a-days, when Cuetius' principles are often opposed as being too lax) that his especial endeavor was to prove that a more rigorous order exists in the realm of sounds than his pre- decessors had succeeded in establishing, and thus to place etymology on a surer foundation. In the Grundzuge, 5"* edi- tion, page 80 [English translation, 1, page 104], he says : 1 The more recent literature on this subject may be found in the detailed and instructive essay of Misteli on "Phonetic Laws and Analogy" (Lautgesetz und Analogie), in Lazarus & Steinthal's ZeiUchrift fur VSlkerpsychologie, 11, page 365 seq. In the following pages not only what is quoted by Misteli, but also the essays of Benfey and his adherents, on the same topic, are especially taken into account. Phonetic Laws. 103 "If there really had occurred in the history of language such very sporadic variations and completely diseased and un- accountable corruptions of sound as are confidently assumed by many scholars, we should be obliged to renounce etymolo- gising altogether. For it is only what is regular, and intern- ally coherent, that can be scientifically investigated ; what is arbitrary can at most be guessed at, never decided with cer- tainty. The case is however, I believe, not quite so bad as that", but (page 81) "it is precisely in the life of sounds that fixed laws may be most surely discovered, which act almost with the consistency of the forces of nature". CuRTius, therefore, although he does distinguish an irreg- ular or sporadic substitution of sounds [ Lautvertretung] , in opposition to the regular one, would by no means assert that phonetic change is in part exempt from all laws, and given over to chance and arbitrary action. "It is needless to say", as he remarks farther on (page 90), "that we do not regard either the one or the other class of phonetic change as accidental, but rather start with the opinion that laws penetrate this pho- netic side of the language, as they do the whole." How it is possible, despite this regularity, which prevails throughout language, for corruptions and abnormal changes to occur in the substitution of sounds, will become clear to us if we consider more closely the nature of this regularity. In the first place, CuktiuS sees in all phonetic movement a pervading tendency or inclination. The fundamental ten- dency of phonetic change is a descending, diminishing one, or as CuKTius prefers to call it, "disintegration" [VerwitterungJ. "For in fact it is very natural to make a comparison with the stones, which are gradually diminished and wasted by atmo- spheric influences, yet in spite of this retain their core so per- sistently." (Page 409 [trans., 2, page 4].) Of course in the case of sounds the cause of the diminution does not lie in the action of external forces , but depends upon human conve- nience, which ever strives to make the pronunciation easier and easier. "Convenience is and remains the chief factor in pho- netic change under all circumstances." (Page 23, note.) But this convenience displays itself chiefly in two tendencies. First, there is an inclination to exchange the less convenient 104 Chapter VI. ; place of articulation for the more convenient, and therefore, ' since the place of articulation is more inconvenient the farther back it is situated, the inclination to form the sounds farther forward in the mouth can be established as a general tendency in phonetic change. So p arises in place of k, but not k in place of ^. Secondly, the sound which in its nature is more difficult to pronounce is replaced by the more easily pronounce- able one, and accordingly the so-called "explosive sounds" [Explosivlaute] pass over into the "fricative sounds" [Fricativ- laute], while the opposite process does not take place. So t becomes s, but s does not become t. All phonetic change, even the sporadic, comes under these chief norms, whose validity CuRTius tries to establish in special instances. "Even in the case of sporadic substitution of sounds, we must be guided by the principle that only a transition of the stronger sound into the weaker may be expected, and not the reverse." (Page 43 7.) Phonetic change, therefore, may not overstep the general conditions imposed by these norms, although we must allow it a certain freedom of motion within these limits. Thus the original a of the Eiu'opean languages is represented, now by a, now by e or o, without definite discoverable grounds for this change of coloring ; the Indo-European guttural tenuis appears in Greek now as x, now as tt, now as t, also without the pos- sibility of adducing satisfactory reasons for this divergence ; and in addition to these irregularities, which can always be brought into some system or ranged under some category, there are isolated abnormities, as for example, when an s at the beginning of a word regularly falls out in Greek, but a ou? is preserved by the side of u?, and many similar phe- nomena, with which every one is acquainted through practical experience. CuRTius by no means regards as wholly inexplicable this great mass of isolated exceptions, irregularities, corruptions and arrested forms, however he may christen them; he seeks, indeed, to discover the forces which can interrupt the normal course of phonetic change. Of such forces he mentions two : effort to preserve the significant sounds or syllables, and analogy. The first point he has especially treated in his remarks on the range of phonetic laws, particularly in Greek Phonetic Laws. 105 and Latin [Ber. der phil.-hist. Classe der Konigl. sacks. Ges. der Wissenschaften, 1870, July 1). Ctjrtius tries to show in this essay that sounds and syllables which are felt to contain the chief significance oppose disintegration longer than others, and that accordingly the importance of the sounds must not be neglected in our judgment of phonetic change. What he says about the i of the optative may serve as an example : "The Greeks in general had a strong inclination to drop the last sound of the diphthongs ending in i, before vowels ; hence we find aiu, so, oo) for the older ay ami, ttosu) frequently for Ttoiiu), •etc. They followed the same tendency in the geni- tive singular, where at an early period oio was contracted to 00 and further to ou, Doric and Aeolic to. On the other hand, the ot remained undisturbed in optative forms like Soitjv^ Xi- Yoisv, YsvofaTo, TcoiotYjv. Aaj(07)V = Xaj(oi[ii. is only transmitted to us as an Aeolic form (Ahrbns, page 133). It was evidently I more necessary to save the mode-sign than the i of the geni- ' tive. The latter case could still be plainly recognized without I, even after contraction had taken place, while the optative formations would be almost unrecognizable without this i, or at all events very unlike the other forms of the mode", etc. The second point, analogy, has not been comprehensively treated by Curtius, but like other philologists, he has occa- sionally employed analogy as a principle in explanation. He by no means fails to observe that an important influence is exerted upon the whole theory of language by the conception which is formed of the working of analogy. In this connection, a sentence of the article mentioned above (of the year 1870), page 2, is of especial interest : "Two fundamental notions are of the highest importance for linguistic research, that of analogy, and that of phonetic laws. I think I can hardly be mistaken in asserting that the difference of opinion which exists concerning individual ques- tions depends in large measure upon the latitude allowed to each of these notions in the life of language." By means of these two mental forces, i. e. perception of 'the significant value of a sound, and of the power of analogy, it is possible to explain many of the existing deviations, al- though by no means all. According to Curtius' view, quite a 106 Chapter VI. large number remain, and I will call particular attention to one point, which seems to me of prime importance. Cuktius not seldom assumes that from one and the same sound, or one and the same group of sounds, diiFerent results can arise under precisely similar circumstances. The declension of the com- parative affords an example. From the it.zit^o'iaoz, which we must presuppose, could arise either [isiCooo?, with retention of the a, and from this [j,s{Coo<; ; or [let'Covoc (perhaps through the intermediate form [asi^ovvo?) , with retention of the v. (Cf. JEr- Vduterungen zu Curtius' griech. Schulgr., page 68.) It seems to me that such "doublets" (as Brbal calls them, in Mem. de la soc. de Unguistique de Paris, 1, page 162 seq.) can only he ex- plained by Curtius under the supposition that the speakers chose freely, and of course unconsciously, between existing possibilities. They are of especial importance in forming a conception of the phonetic laws, as we shall see later. In this system of Curtius, which we have roughly out- lined above, although not accurately portrayed, three notions are especially prominent : phonetic laws , analogy, and the preservation of sound on account of sense. I will discuss these three notions in inverse order. In regard to the last point, the influence of sense on sound, I cannot convince myself that Curtius' view is the correct one. There is an objection to it on general grounds. It seems to me we are not justified in assuming that the Hindus and Greeks had a perception, which we have ceased to possess, of the significance of the individual sound in a linguistic form. For they, as well as we, had only completed words, which were transmitted to them from generation to genera- tion ; and that primeval period in which, according to the BoppiAN assumption, the Indo-European forms were con- structed, by the composition of significant elements, lay for them, no less than for us, in the twilight of the past, whence no enlightening raiy could reach them. It also seems as if, in single points, more plausible explanations might be found for several of the phenomena discussed by Curtius. Thus in my opinion the preservation of the i in the optative, referred to above, may be more correctly regarded as due to the in- Phonetic Laws. 107 fluence of analogy. It seems, in fact, the most natural assumption that 8ot'7)v remained (i. e. did not become Sotjv) because it formed part of a series Sotfiev , SoTte etc. The case is the same with the is of the genitive in noctis (cf. Cur- Tius in the above article, page 22), which was retained in con- sequence of the innumerable ««'s in the genitive, while no sim- ilar analogy prevailed to an equal extent in the nominative ; the same is also true with regard to the i of tpuXa^i, which was more protected than that of evt, etc. Different explanations may be found for other points introduced by Curtius ; thus, as he himself intimates, in explaining the different forms of prepositions, we must take into account the difference of ac- cent," according as they are or are not used as proclitics. In this case, since the accent is a very important factor in the phonetic aspect of a word, we must seek the explanation on the phonetic side. Of course I am unable to solve many of the unexplained difficulties which Curtius brings forward in the article mentioned above ; but I can at least assert that no conclusive proof has yet been given that those sounds which were felt to contain the chief significance were occasionally preserved, in direct opposition to prevalent phonetic laws. I am therefore of the opinion that we have not yet the right to admit this idea into the repertoire of philology. Analogy^ the second of the notions emphasized by Cur- tius, has already been mentioned in its historical develop- ment. I will here repeat that this principle was not ignored in earlier times i) , but that lately it has been much more fre- quently applied, owing to various causes, among which are the example of modem tongues , the conviction that the new formations of individual languages depend upon imitative for- mation, and above all, the attempt to establish exceptionless rules in the case of phonetic change. The question arises 1) In MiSTELl's article, Benfey might have been! quoted, as well as Pott and Curtius, since as early as the year 1865, in the Orient und Occi- dent, 3, page 225, he spoke as follows regarding the Vedic language : 'It is not without a purpose that throughout this essay I have called attention to the examples of false analogy, by which the Vedic language is forced into the njost diverse channels." 108 Chapter VI. whether and in what way this frequency of application can be justified, and whether it is possible to set certain limits to the employment of the principle of analogy, and within these to hold fast to certain distinctions and divisions. In regard to the first point, the demarcation of the field, so far as I can see, no practical directions have hitherto been given. It is true that Misteli in the article quoted above, page 410, laid down the principle that not too many and not too complicated workings of analogy must be assumed for the individual instance ; but this general direction is no help in the individual instance, since in each case the question "what is many?" and "what is complicated?" will find various answers. Another suggestion seems more plausible at the first glance. It is natural to assume that the forms which exercise the attractive force (i. e. produce the analogy) must be more numerous than the attracted ones. But on closer examination this reasoning is found to be invalid. In my opinion, at least, Bexjgman is right when he argues (in Kuhris Zeitschrifi, 24, page 50, and Morph. Unters., 1, page 82 seq.) that the action of analogy takes place gradually, that is, one form may attract a second to it, these two a third , fourth and fifth, and these further the following ones, up to the thousandth, etc., so that we can easily imagine that a mere handful of forms may have served as the model for thousands. Such cases do actually occur ; thus , Brugman adduces the fact, already established by other scholars, that "four Old Slavonic verbs, /esm, verm, damt anijami, have brought it to pass that in New Slovenian and New Servian the verbs of all the conjugational classes end in -m in the first person singular", — and similar examples. (Cf. Morph. Vhterstich., 1, page 83.) It seems to me, therefore, that hardly any practical sug- gestions have been offered in regard to the boundaries within which the action of analogy takes place. Perhaps it would be easier to say something of the various kinds of formation by analogy. Since a formation by analogy is a change of form which occurs in consequence of an asso- ciation of ideas, we can make a classification from three fol- lowing points of view : from the nature of the psychic processes which play a part in such a formation ; from the constitution Phonetic Laws. 109 of the words in question; and from the result attained by the action of analogy. I will discuss these three points briefly in the above order. First, in regard to the classification according to the psychic processes : much that Misteli has brought forward on this subject may serve to introduce its discussion, which has not yet begun in earnest. I will only emphasize one point here : it is important to distinguish whether a transfer of form has taken place of itself, so to speak (as is the case in the greater majority of instances) , or whether the speaker, finding the form which is demanded by the phonetic laws for some reason inconvenient , seeks for some other formation, and as the result of this search a transfer of form takes place. An example of the latter sort is the Latin dative and ablative plu- ral in -abus, which frequently occurs in deabus, fliabus and libertabus, and in isolated instances in other words. As is most clearly shown by the passages in Nbue's Formenlehre der lateinischen Sprache, 2"^* edition, 1, page 22, these dative-ab- lative forms arose where a distinction from the corresponding forms of the masculine was needed. There was no objection to saying di deaeque, deorum dearumque, deos deasque ; but what should be said in the dative and ablative, — dis disque ? There was a similar drawback to the use of flia in wills or other provisions quae pertinent ad necessitatem Juris. Suppose, for example, provision must be made in case a son or sons, a daughter or daughters are living. Should it be worded : "flio seuJiUis, filia seufiliis exstantibus" "i It was evidently in such and similar predicaments that the forms in -ahus originated, and the process can hardly have been other than the follow- ing : another form is sought instead of dis and filiis, which in special cases could not be employed ; and this form is suggest- ed in consequence of the connection subsisting in the mind of the speaker between the SBTi:iesJiliae,Jiliarum,fiUis,Jilias, and duae, duMrum, duabus, duas. In ordinary speech the ablative duabus can exercise no attractive force on mensis and the rest, because their is is protected by connection with the Is of the second declension, which has the same significance. Not until this connection is for some special reason dissolved, does exert its attractive force. The old grammarians are 110 Chapter VI. therefore quite correct in saying that the forms deabus etc. were created differentiae causa ; but the impulse toward differ- entiation was not able to evolve new and original formations, only imitative ones, after existing models. This impulse to differentiate can accordingly be classed among the motives which are active in the constmction of forms by analogy. (Cf. MiSTELi as quoted above, page 472.) We find a second ground of classification in the constitu- tion of the words in question, that is, in the conditions which must be present in the words before any action of analogy can take place. Under this head we must ask first of all whether words connected only by sound, and also whether words con- nected only by sense, can influence each other through the working of analogy. I should be inclined to answer the first question in the negative, the second in the affirmative. To illustrate the first, Misteli gives a good example (page 434), which I will repeat here : "Although za&i'Cu), exa&ioa forms in the future xa&iu), -%i- et?, -&iet, as if xa& were the root and iZfa the ending, as in PaSt'Cu), pa5iou[i.ai, so that scarcely a shadow (in the t) of the root sed remains ; yet despite the identity of ending, xa&iCe, xa&tCov, -&tCu), -fti^ouv, -%iQaic, have not the remotest connec- tion with, for example, a itpoppi^s, irpoppiCov, -pt^u), -pt'Cwv, -piCoi? ; the gulf between noun and verb cannot be bridged over by any amount of phonetic identity, and it is only be- cause we regard this as self-evident that we can speak of purely phonetic analogy." As to the second point, it is at least clear that endings whose function is identical enter into association, even with- out phonetic similarity ; thus, (XYtovoic arises from analogy with the dative plural in -oi?, while there is no seductive sim- ilarity of form between ai (in 017(001) and 01?. Whether the same can be observed in word-stems (e. g. whether the form of the adjective "good" can influence the form of the adjective "bad", or the like) must be more accurately investigated. Caeolina MiCHAELis [Studien zur romanischen Wortschopfung, page 35) assumes such an attractive force in the case of the, Italian ffreve, which would accordingly owe its e to the influence of the e of leve. In the second place, we must remark that in Phonetic Laws. HI inflected words the associative action can start either from the word-stem or from the endings, and in this connection a di- stinction must be drawn between material dinii formal analogi- cal construction. 1) An example of »ia) To show the progress which has taken place in the strict administra- tion of phonetic laws in all departments of comparative linguistics, I will quote some remarks of two scholars who agree in their decided opposition to the new school of grammarians jjunggrammatische Schule], — remarks of Bezzenbergek und E. K^H^f. Bezzenberger expresses himself as follows in a review of AscoLi's Kritische Studien : "On page 404, note 2, Ascoli asks, in connection with the discussion in which he tries to ascribe to the 'original instrumental suffix -tra' a IIQ Chapter IV. be doubted that all scholars who have devoted any serious at- tention to phonetics have consciously or unconsciously been influenced by the idea that the moving spring of all changes is neither arbitrary nor accidental, but prevailingly regular. Yet on the other side the fact must be admitted that even in those fields where the work has gone on unceasingly for many years, much that is obscure still remains ; and although it is to be hoped that more difficulties will be successfully overcome i) , progeny of somewhat surprising dimensions within the bounds of the Latin and Romanic languages : 'Or will FiCK really assert that -9Xo (-bloj is radically different from -bro 1 Will he, for example, separate *tpu8-Xa from the Hesyohian cpuxpa ? Can we separate latibulum and latebra ?' I have not asked FiCK what position he takes with regard to these questions, but I earnestly hope that he answers them all in the affirmative ; and who eould blame him if he should ask in turn : 'Can we identify latibulum with late- bra'!' ASCOLI in my opinion is at fault when he says : ' — a primary suf- fix, which would stand isolated, like a Greco-Italic -rlhld . I have already- said elsewhere, and repeat it here, that the Slavonic -dlo- corresponds exactly to the Greek -ftXo- and the Latin -bulo-. , and if on the part of certain German scholars it has been preached, on the one hand, that phonetic laws admit of no exceptions, and on the other hand the Polish radio is pronounced equivalent to the Greek apotpov, this is merely one of the many instances of thoughtlessness exhibited by these very methodical 'investigators'." {Gstt. gel. Anz., 1S79, article 18.) E. Kuhn's remarks are as follows [K. Z., 25, page 327): "What is the relation between Sanskrit kumbha and Zend yumbat The Aryan primitive form of both was khumbha ; from this was derived without difficulty the Zend yumba , and it is well known that the Sanskrit kumbha has lost its first aspiration. The absence of aspiration in stambh, stigh etc. as opposed to stha etc. is explained by the same rule. The whole question of the aspirated tenues needs a thorough revision, and now-a-days many who do not exactly advocate the fashionable folly of the infallibility of phonetic laws will find little probability in Schlei- cher's assumption, according to which precisely the oldest examples of this phonetic class owe their origin to a wholly sporadic phonetic change." 1) A suggestive and valuable collection of such irregularities in the substitution of sounds, the cause of which is unknown, has been made by CURXIUS in his Grundziige, 5th edition, page 429 seq. Whoever will reach these difficulties must, in accordance with the above, attempt it in three ways, by investigating ; 1. Whether there is any borrowing. This is the case e. g. with -/.io- ■taTM by the side of oxiovaxai, Tdyoc by the side of otsyoc, and the like. 2. Whether there is any action of analogy. Under this head belongs e. g. the dative of the participle XeYOVTi, which was prevented from becom- ing Xeyouoi by its connection with Xsyontoj, \i•(Q•^xa etc. The same is true of xipaTi and Ttavxi. In dvt'i the t was probably preserved because ott' so often occurs. Phonetic Laws. 1 \ 7 no one can indulge in the delusion that it will ever be possible, in any language, to fully and entirely penetrate to phonetic change in its essence, and to view it in all its parts and develop- ments. We are forced to make the confession : it cannot be proved hy induction that phonetic laws admit of no exceptions. , We accordingly find ourselves compelled to seek a solu- tion of the problem deductively, by considering from what causes and in what manner languages change, and above all, how it can be explained that different dialects arise from a ho- mogeneous speech. In accomplishing this task we shall at the same time answer the question whether phonetic laws in themselves admit of exceptions or not. This point also I will discuss in connection with the theory of Georg Cuktius ; yet I must first mention a mode of view which was formerly cus- tomary, but has now-a-days been almost entirely thrust into the background. In the pre-BoppiAN period it was customary to derive the difference of languages from the difference of the human vo- cal organs, and to explain this in great part from differences of climate. How often has a comparison been confidently made between the alleged harshness of the Doric dialect, with the wild, mountainous nature of the Laconian landscape, and the alleged softness of the Ionic, with the mild breezes of the coast district of Asia Minor ! Whitney, in his Language and the Study of Language, pages 152, 153 spoke very decidedly against this old assumption, which, however, has lately been revived by Osthopf, who says : "The formation of man's vocal organs, as well as that of all his physical organs, is especially dependent upon the con- ditions of climate and civilization under which he lives. Al- though it is generally known, for example, that the differing climate of a mountainous and a flat country causes a different devielopment of lungs, breast and larynx in the inhabitants, 3. Whether two sounds are concealed under one sign. This is prob- ably the case with /; with the consonantal f becomes as (a), while " before the semi-vocalic / falls out. Only before the consonantal f does the so-called "prothetic" e appear, as in iepsT). It is to be hoped that through such treatment the lists drawn up by Ctjrtius will be sensibly diminished. 11^ Chapter VI. yet it is a fact hitherto too little heeded in linguistic science, that everywhere under identical or similar conditions of cli- mate and civilization, identical or similar phonetic tendencies are accustomed to manifest themselves in the language or dia- lect. I regret that I cannot here adduce sufficient examples to establish this principle. I vrill only remind the reader that in the Caucasus, for instance, neighboring nations, even when no original relationship exists between them, the Indo-Euro- pean Armenians and Iranians , and the non-Indo-European Georgians and others, possess in the main an almost identical vowel and consonant system. It has been convincingly proved, above all by recent investigations in various fields, that within the limits of one and the same language an almost regular gradation prevails, or formerly prevailed, between the single dialects which constitute the common speech ; for example, in the Germanic group, from the Alemannic of the Alps to the Low Saxon on the Baltic and North seas. I can hardly ima- gine that the regularity of climatic gradation covering the same area should not stand in some causal relation to this gra- dation of dialect." {Das physiol. und psychol. Moment in der sprachlichen Formenhildung , page 19, in the Sammlung ge- meinverstdndlicher wissenschaftlicher Vortriige, herausgeg . von Hud. Virchow und Fr. von Holtzendorff, Heft 327.) It is perhaps impossible to definitely answer the question whether climate and customs also have an effect upon pho- netic change, which is all that Osthoit claims. It will cer- tainly be granted in general terms that the climate cannot fail to have some influence upon the vocal organs, as it does upon the whole body ; but on the other hand it must be con- fessed that physiologists have not observed such a difference of the organs as would explain the difference in the pronun- ciation of the separate sounds. The similarity which, accord- ing to Osthoff's statement, exists between neighboring lan- guages, could perhaps also be explained by an influence ex- erted within historical times (for example, the Germans who live in Kurlaiid have acquired something of the pronunciation of the Lettish people) ; and above all, the numerous changes of habitation made by the nations of every period are strong evidence against the theory. Should an influence of the cli- Phonetic Laws. 119 mate upon phonetic change ever be demonstrated, a natural influence upon the formation of sounds would thereby be prov- ed, which would then have to be distinguished from a social or historical one. I am not able to give a satisfactory answer to this question, which has thus been agitated anew by Ost- HOFF, and will accordingly pass on to the theory of Geoeg CURTIUS. CuETius, as we remarked above (page 103), regards as the chief cause of phonetic change the attempt to make the , task easier, the love of convenience which is characteristic of I the human race, and Whitney agrees with him in the main. The latter scholar says in his Language and the Study of Lan- guage, page 70 : "All articulate sounds are produced by an eifort, by ex- penditure of muscular energy, in the lungs, throat and mouth. This effort, like every other which man makes, he has an instinctive disposition to seek relief from, to avoid: we may r call it laziness, or we may call it economy ; " It is the result of this love of convenience, or this care- , lessness, that no generation speaks words precisely as they \ were spoken by the preceding one ; only the fact that lan- guage is destined to be a medium of communication, and Jre- gard for its intelligibleness (says Whitney), keep this care- lessness within bounds. The following considerations have especial weight against this theory.') It seems to me very doubtful ^whether we have the right to assume that love of ease plays so predominant a part in human society. Would it not be possible to assert, on the other hand, that most men will exert themselves to imi- tate as accurately as possible what they have heard spoken, /because they are afraid of making themselves ridiculous by deviating from the rest of mankind? — and further, that in \ speaking, not only what is convenient is aimed at, but quite ( as much what is pleasing? (cf. Benfey, Gbttinger Nachrichten, 1877, No. 21, page 550) — and that the impulse to consult convenience is opposed in a very effective and thorough man- 1) See also Leskien in the Jenaer Liter atwzeitung, 1875, No. 6.'' 120 Chapter VI. ner by these and other conceivable motives? Perhaps still more weight must be given to an objection derived from prac- tical experience ; it was raised by Ascoli, one of the most prom- inent masters of empiricism. Ascoli asserts that in the lan- guages which come under our observation, innumerable cases of phonetic transfer are found which cannot be explained from the principle of "weakening" or "making easier", as Curtius expresses it ; and Curtius himself is by no means disposed to regard this objection as wholly unjustified (cf. Grundzuge^ page 410), — indeed, in one important point he now assumes with Ascoli a phonetic change opposed to the general prin- ciple to which pe usually adheres, i. e. the change of oa into TT in Greek. •'' Under these circumstances it would be desirable to find a more general theory, in which, in addition to the desire for convenience, the other imaginable motives of change might find their place. This theory will be easily formed if we first ask the question, whether the changes which are here men- tioned make their appearance all at once, among all the mem- bers of a community using a common language , or whether they start from an individual, or several individuals, and spread in different directions. It is only necessary to ask this question to answer it. If here, as well as in the whole of the following discussion , we disregard the possible influence of the climate, about which I can assert nothing definite, it is then clear that changes in pronunciation begin with the single individual, and are propagated by imitation throughout gi-oups and masses. The final cause of all linguistic change, there- fore, can only lie in the fact that the single individual does not circulate the language imparted to him precisely as he re- ceived it, but always individualizes what was transmitted to him, whether from love of convenience, or from an aesthetic impulse, or because his ear, in spite of every effort, could not accurately enough grasp it , and his mouth reproduce it , or from some other cause. Now the equalizing tendency of uni- versal linguistic custom continually exercises a counter-check upon these innovations, so that change in the phonetic form of language is a result of these individualizing and equaliz- ing forces. (Cf. especially Benfby, as quoted above.) Phonetic La ws . 121 The following will serve as further illustration of these general statements. We must be on our guard not to magnify the sphere of action of the individual (even leaving out of con- sideration the counter-influence of society) . In the first place, we must consider that in the transfer of sounds practical inter- ests hardly ever come into play, as may ,be the case in the transfer of words. It may happen that the chief of a warlike race suddenly issues the command that the appellations cor- responding to certain ideas shall be changed, in order that the spies of the enemy may not understand the conversation of the warriors ; or a prominent statesman or poet may for some reason bring forward a forgotten word and suddenly reinstate it in favor, — but in the field of sounds there seems to be no occasion for such a violent and arbitrary encroachment of the individual. Then we must not forget that the sounds of lan- guage (or a part of them) are arranged in series in the mind of the speaker, and that the change of one sound must inevi- tably induce a corresponding change of the remaining mem- bers of its series. If the pronunciation of h is changed in a certain way, the corresponding change of the remaining gut- turals occurs spontaneously, and thus a considerable portion of the sounds are excluded from the possibility of an individ- ualizing change. It would perhaps be advantageous if more weight were given to this idea, in our observations of phonet- ic change, than has hitherto been the case. Finally, we must regard it as certain that all (or nearly all) these acts take place unconsciously. How true this asser- tion is with regard to our language of today we can easily convince ourselves by experiment. Most people do not know how they speak, and it often requires the greatest pains to convince them that they really possess certain fine shades of pronunciation which an experienced observer detects in their speech. After the above remarks, we can comprehend the deriva- tion of various languages from one, as well as the relative uni- formity within the bounds of one language. The first point, the derivation of various languages from one, demands no detailed consideration. If we imagine a little 122 Chaptek VI. community of men, say a hundred souls, who live together within a small territory, the impulses, proceeding from single individuals, to introduce innovations , will be readily and quickly counterbalanced by the habit and inclination of the remainder, and the process of leveling will take place without difficulty. If, now, we suppose a larger mass of men in a wider domain, still forming a community united for purposes of intercourse, the process will be a different one. The level- ing will occur, but in each individual instance it will require more time than in the case of the smaller community, and there will always be marked differences between the separate natural groups of speakers, since some will still speak in the old way, while others employ the new. The leveling pro- cess will not come to a stand-still until there is a cessation of intercourse ; the boundary of speech will then be formed in connection with the boundary of intercourse. Various histor- ical complications may naturally occur in the formation of this boundary of intercourse. The following is a simple case. A tribe settles on the shore of a large stream, and subsequently a portion wanders over to the opposite shore. The intercourse naturally continues for a number of years, but gradually the ties which bind the wanderers to their old kinsmen become loosened, meetings take place only on rare occasions, and the linguistic impulses no longer cross the boundaries. Thus there is opportunity on both sides for the formation of a new lan- guage, which can develop more or less quickly according to circumstances. The case is more complicated if we assume that a portion of the emigrating party returns after a number of years ; their language, if it has not yet gained a strong in- dependent development, will perhaps be wholly absorbed by the old language ; or it will retain its individuality by means of one or more peculiarities ; or if the difference is already too great, a linguistic island will be formed, which may remain for centuries, until at length intercourse produces uniformity. ') But it is neither practicable nor necessary to bring forward in detail the endless variety of historical possibilities. In all 1) This would be the place to mention the mixed languages [Misch- sprachen], if a thorough treatment of them were in existence. Phonetic Laws. 123 cases the principle will evidently be found true that no unity of speech can exist where there is no unity of intercourse. It is more difficult to answer the question, how great the uniformity will be within the bounds of a homogeneous lan- guage. In the first place, it is clear that the different individ- uals of a linguistic community can never speak exactly alike ; we must therefore confess at the outset that a homogeneous language in its strictest sense can only exist in the individual, or among a limited number of individuals, and the question which occupies us will accordingly be more accurately worded ; as follows : can it be expected, in the case of the single individ- ual, that phonetic change will take place in a perfectly uniform ) and regular manner? As we should expect, it is precisely in answering this question that the difference of stand-point becomes manifest. But in one respect perfect harmony seems to reign. So far as I see, it is universally admitted (or should be admitted) that in the passage from one pronunciation of a sound to another a state of Jtuctuation can arise, in which the same individual speaks now in one way, now in another. Sibvbks, for example [Lautphysiologie, page 127), says in regard to this point: "The spontaneous construction of new phonetic forms naturally has its starting-point in the single individual, or a series of individuals, and it is only by subsequent imitation that these innovations are gradually transferred to the whole linguistic community to which these individuals belong. The complete adjustment between the colliding forms, the old and the new, may in some cases require a long time. For a certain space both forms will be used interchangeably ; they will also be differently employed according to the position of the sound, until finally the new phonetic form wholly supplants the older." At the same time, Sieveks mentions some instances of such fluctuation derived from practical observation : "Examples of fluctuation between two forms are found in many North (Grerman dialects, which use sonant and surd mediae ') without 1) The reader must bear in mind that Sievees' classification of the "Gerauschlaute" (i. e. all except the vowels, liquids and nasals) is as follows : 124 Chapter VI. distinction. The same is true of different dialects of the Armenian, while in those of Middle and South Germany, on the contrary, the surd mediae have for a long time held ex- clusive sway." Brxjgman's arguments [Kuhris ZeitscJtrift, 24, page 6) are quite similar, except that he would allow only a short duration to such transition-periods, whereas Sieveks assumes that they may occupy a long time. It is evident that it would be vain to dispute about such very elastic terms as "long" and "short". It is of far greater importance to collect further facts from living languages, in order to draw conclusions from them with regard to the ancient languages. In the Greek field we might bring up the unstable rhotacism of the old Elian inscriptions , where , as is well known, xoti; and toTp etc. stand side by side. How great the difference of pronunciation really was between a and p, un- fortunately cannot be established ; perhaps at the end of a word instead of a a sound like the Sanskrit visarga was spoken, and possibly the terminal p was not dental, but guttural, in which case the actual difference between the two sounds must have been very trifling. It is quite impossible to assume that two phonetic forms, between which the speaking individual vibrates, can differ to any considerable extent in an Indo- European language, while in other linguistic fields, for ex- ample, in the languages of the American aborigines, this dif- ference may be quite a wide one.') Meanwhile, whatever may be the judgment in regard to these single points, all agree in the theoretical recognition of the possibility of transition-periods. On the contrary, opinions differ as to the propriety of assuming Avithin the domain of phonetic change a permanent lack of uniformity in the treat- 1. Explosive sounds. a. Surds (tenues and surd mediae). b. Sonants {sonant mediae). 2. Spirants. a. Surds. b. Sonants. [Transl.] ') It is impossible, however, to pronounce with certainty upon the dialect of Elis, because inaccuracies and arbitrary changes may have taken place in transferring the sound to the written character. Phonetic L a ws . 125 ment of one and the same sound. The chief questions to be discussed in this connection are the following : First : Can it be assumed that a phonetic change appears in one series of words, and not in others? Secondly : Can it be assumed that one and the same word- form may, by a phonetic process, deyelop into permanently dif- ferent forms '! The first question was formerly answered with an unques- tioning affirmative whenever it was practically applied ; thus Bopp found no difficulty in assuming that although the s of the aorist was in Greek regularly represented by a, yet by an exception it appeared as x in r^xa, I9rjxa, eScoxa ; nor did he trotible himself to discover any special reason for this re- markable exception. In proportion as phonetic change receiv- ed more critical attention, such assumptions were naturally regarded with more and more suspicion, and they are resolute- ly rejected on principle by a number of philologists. The subject has recently been theoretically treated by Brugman, Kuhn's Zeitschrift, 24, page 4, and Bezzbnbbrgbr, Gott. gel. Aiizeigen, May 21, 1879. The possibility of a dissimilar treatment of sounds in dif- ferent words might be explained from two stand-pointsi In the first place, the assumption might be made that every pho- netic change begins with a definite word, and is propagated farther from this starting-point, so that it proceeds, for ex- ample, from one substantive to others, from these to adjectives and participles, and thence to the verb. Under this supposi- tion it would be easy to imagine that certain words should not be affected by a phonetic change ; that, for instance, in the case of prepositions, adverbs and other comparatively isolated words, the ancient phonetic condition should be retained, while in other words a new one should prevail. But evidently such an assumption would not correspond to the facts, and Brugman seems to me to be right in describing the process as follows : "If at a definite period a number of individuals change an r, which they have hitherto regularly pronounced as den- tal, into a guttural r, or if they give a deeper coloring to their pronunciation of a before I, which they hVd previously uttered 126 Chaptee VI. with a lighter timbre, this change does not begin with single definite words, so that what was at first only applicable to these single words is gradually transfen-ed to others, — but the change begins with the organs of speech themselves, and we must expect that every r hitherto spoken as a dental, in what- ever word and whatever category of words it stands, and sim- ilarly every a ,before /, which was hitherto spoken with a lighter timbre, will experience the change when it passes through the vocal organs." Only the expression "the change begins with the organs of speech themselves" is not a happy one, since it can be un- derstood to mean that every phonetic change begins with a physical change in the larynx etc. I would therefore rather say : the change begins with the pronunciation of the sounds. It is not only shown by experience with popular dialects that this is actually the case, but we must consider that it is only tmder the supposition of a uniform and consistent pro- nunciation of sounds that the acquisition of a foreign language is explicable. Another way of explaining the matter is intimated by Bezzenbekgbe, page 652 of the article quoted above. He in- trenches himself on the unassailable ground that diiferent phonetic tendencies can arise at two points of the same lin- guistic territory ; thus in one part a certain ^-sound can be- come s, in the other sz (i. e. sh). Now in his opinion, an ad- justment takes place between the difi^erent tendencies, and the result is that in certain words sz appears, in others s. Thus in Lithuanian sz has become the regular representative of one of the Indo-European ^-sounds, but in visas and sauja s has become established in its stead. I cannot accept this view. It would recommend itself more strongly than it does (although even then it would not be the only possible expla- nation) if it were an actual fact that the two different sounds, so to speak, share the word-material of the language between them. But this is not a fact, for (to speak in Bezzenbergbr's language) one tendency has almost wholly outweighed the other, of which on|y scanty traces remain. How extraordinary that the speakers, who in a hundred cases employ sz, in one or two condescendecT to adopt s ! — and why did this happen Phonetic Laws. 1 27 in precisely these cases ? Is it not much more natural to as- sume that the isolated exception to the empirical rule owes its origin not at all to phonetic change as such, hut to some other cause, even although (as in the present instance) we cannot succeed in determining what this special cause is'/ Bezzbnberger in the ahove article did give a somewhat dif- ferent tui'n to the same supposition, by assuming that two pho- netic tendencies, starting from different points and then com- ing together, might so adjust themselves that one phonetic form should appear in one category of forms, the other in an- other. Thus (if I understand him aright) when in Middle Prankish the tenuis t appears throughout changed by permu- tation [verschoien] to 2, with the exception of the neuter t in dat, wat^ it, allet and dit, he explains the process as follows : the tendency came from one direction to retain the t, from an- other to change it, and the condition in which the Middle Prankish is transmitted to us represents a compromise be- tween the two tendencies. But Paxjl has shown, in Paul & Braune's Beitrdge, 6, page 554, that this case must be ex- plained in a different manner. The retained (s of the Middle Prankish stand at the end of the syllable, and it is probable that the terminal ^'s in general were not changed according to phonetic laws, so that an original inflection Vikefat, fuzzes must be assumed. Now/a^ was attracted by fuzzes, and be- came fuz , but where there were no such attracting oblique cases the t remained, i. e. in dut, wat, it, dit and ullet. But even if the explanation were doubtful in single in- stances, I should still be inclined to refuse acceptance to Bez- zenberger's view, from the general ground alone that it pre- supposes a too great exercise of the reflective powers on the part of the speaker. I am therefore of the opinion that an- other explanation must likewise be sought for the few analogous cases which Bezzbnberger adduces in addition to the above. It accordingly seems to me that the first of our questions must be answered in the negative. I am of the same opinion with regard to the second ques- tion, viz.: can it be assumed that one and the same word-form may, by a phonetic process, develop into permanently different forms? I find myself compelled, with Leskien, Osthoff, 128 Chaptek VI. Brugman and others to answer this question also in the neg- ative, although two scholars so eminent and so often differ- ing in opinion as Benfby and Curtius answer it in the affirm- ative. A classical example for this assumption are the twin-forms [isiCovoi;, fAst'Couc, which are supposed to have arisen from the common primitive form p-etXovoo?. MetCovao? became (according to Curtius) on the one hand (isiCoao? (0,== nasal vowel), then fisi'Coao?, [XEi'Coo?, and finally fxstCoui; ; on the other hand it be- came [isiCovvo?, then [ieiCovo?. In order to form a clear idea of the process of development, we might assume that this double treatment originated at one point of the linguistic domain. Then it would be necessary to imagine that the single individ- ual tried first one, then the other change of the group ws, and retained both in his memory. Now it seems to me impossible to find any reason why a speaking individual should waver in this fashion, and so obstinately continue to waver ; and it is still more difficult to comprehend what could induce the rest to follow his example, when all the time the meaning of the two forms was identical, so that it could not be of the slight- est practical interest to distinguish them. I cannot, according- ly, believe that processes like the above-mentioned actually take place in language. The other possible assumption seems to me quite as little justifiable, viz., that the impulse to sep- arate the forms arose from two different points in the linguis- tic domain. It might, indeed, be assumed that (xsiCovoc per- haps originated in the West, and jasiCou? in the East, and that then an interchange took place between the two halves of the language. That the further development of a phonetic group should be different at the different points of a linguistic domain is naturally not beyond the range of possibility ; but that the germs of difference could unfold as must be assumed in the case of [iefCoo; by the side of [Ast'Covoc, would only be possible if the adjusting process were interrupted, that is, if the intercourse were no longer unbroken. Under this supposi- tion, it would be easy to understand how two dialects might be formed, one of which should have the form jj-sfCouc, the other the form |j,eiCovo? ; but why these two dialects should borrow each other's form is left unexplained. Phonetic Laws. 129 From whichever side I view the matter, I am unable to comprehend how jj^t'Cou? and [lefCovo? could have both been formed in the same dialect from [isfCovoos, and am accordingly of the opinion that only the form |xe(Coui; arose by a phonetic process from the primitive form jist'Cooo? (Sanskrit mdhlyasas) , while [istCovo? is a formation by analogy from the nominative (lei'Cwv. How the n can be explained in the latter is, indeed, a point where there is room for disagreement. (V. the detail- ed discussion of Beugman in Kuhris Zeitschrift, 24, page i seq.) ^ We are now prepared to give a comprehensive answer to the question proposed in the beginning, viz., do phonetic laws as such admit of no exceptions? We have seen where we may expect to meet with such laws. Certainly not in the collective mass of any existing speech, whether it be a popular dialect or a literary language. For it is not probable that all the individuals within a lin- guistic community will speak precisely alike. Therefore we can only expect to find these laws in the case of the single in- dividual, or rather, if we wish to be quite exact, only in the average speech of an individual at any one moment. Now from what an individual speaks or would speak at a definite moment of his life, if he allowed the whole mass of his voca- bulary to pass through his vocal organs, we must first subtract all that can be regarded as borrowed (in the broadest sense], and then all phonetic formations which depend upon the action of analogy. When this is done, the form which remains is the result of phonetic change alone. Here, and only here — leaving out of account the possible fluctuations of a transi- tional stage — we may expect complete uniformity in the treat- ment of all analogous cases, and in this sense we must assert that phonetic laws as such admit of no exceptions. At the same time, it must be confessed that complete uni- formity of phonetic change exists nowhere in the Avorld of ac- tual fact ; but there are sufficient grounds for assuming that regularly 'occurring phonetic change is one of the factors to whose united action the empirical form of language is due. In single instances, it is true, it will only be possible to ap- proximately reproduce this factor in its puyity. Delbbhck, Iiitrodnctioir%> the Study of Langtiage. 9 1 30 Chaptee VII. We can see at once from the above discussion whether and how far we are able to speak of "laws", or still more, of "natural laws", within the field of phonetics. It has been shown that the phonetic laws which we pos- tulate are nothing but uniformities which appear in a certain language and period, for which alone they are valid. Whether the expression "law" is really applicable here is doubtful. Yet I avoid entering upon a discussion of the notion "law", as em- ployed in natural science and statistics, because I find that the term "phonetic law" has become so fixed by usage that it cannot be eradicated, and furthermore, because I can propose no better expression in its stead. It is also A harmless term, if we keep in mind that it can have no other sense than that defined above. I cannot approve of characterizing phonetic laws as "nat- ural laws". These historical uniformities can evidently bear no resemblance to chemical or physical laws. Language is a result of human action , and consequently phonetic laws are not based upon the regularity of natural processes, but upon that of apparently arbitrary human activities. CHAPTER Vn. THE SEPAEATIOIf OP THE EAOES. As we mentioned on page 1, Sir William JoNiis, as early as the year '1786, remarked that every philologist who com- pares Sanskrit, Greek and Latin necessarily arrives at the con- clusion that these three languages must be derived from a common source, which perhaps no longer exists, while there are no such decisive grounds for assuming the same relation for Gothic and Celtic. We found that Feiedrich Schlegel took a backward step in comparison with Jones, since he came to the conclusion that the Sanskrit language is the older, the The Sepabation of the Races. 131 others younger and derived from it. EvenBopp, in the begin- ning of his literary career, does not always express himself correctly; thus in his Conjugationssystem^ pa-ge 9, he speaks of the languages which "spring from the Sanskrit, or with it from a common mother"; but later he rightly characterizes the relation as a sisterly one. He is also on his guard not to over- estimate the originality and antiquity of Sanskrit. Thus in the first edition of the Comparative Grammar there is a note (subsequently omitted) to § 605, which runs as follows : "In my Conjugationssystem, and in the Annals of Oriental Literature (London, 1820), I have called attention to the fact that the Sanskrit second person plural tutupd is a mutilated form, and in the earlier sections of this book allusion has often been made to the fact that in single instances the Sanskrit is at a disadvantage compared with its European sister-idioms. It therefore surprised me that Prof. Hofek in his work Bei- trage etc., page 40, made the sweeping assertion that the new investigators have not succeeded in ' wholly emancipating themselves from the unhappy delusion that the Sanskrit has preserved its original perfection of structure with inviolable fidelity'. I for my part never ascribed to the Sanskrit such fidelity to its original structure, and it has always been a pleasure to me to call attention to the cases in which it must yield the palm to its European sisters" etc. Bopp has no fixed name for the one ancestral speech [Stammsprache] from which the individual languages were derived. He speaks of the one ancestral speech, of the period of linguistic unity, of the primitive period of language, of the primeval formative period, etc. This one ancestral speech, which no longer exists, was in Bopp's opinion essentially sim- ilar to its sister languages. It is especially worthy of mention that he did not claim that it was incapable of change. Instead he assumes "that at the time of the identity of those languages which are now separated, many disturbances had akeady taken place in the organism of that one ancestral speech". (§ 673.) Thus he assumes that in oldest times the feminine in a had an -s in the nominative, but had already lost it in the peri^i of linguistic unity. I cannot find that Bopp expressed apy conjecture regarding the home of the race which spoke 9* 132 Chapter VII. < this primitive language, and he has in general no inclination to view things from an ethnological stand-point. The ethno- logical point of view was first emphasized by Kuhn in the Osterprogramm des Berliner Realgymnasiums for 1845. (Cf. Webbe, Indische Studien, 1, page 323.) According to B6pp, the individual languages have freed themselves from the "primitive home" by an "individualizing" process. The expression "separation of languages" [Sprach- trennung] also occurs (§ 493) . Of the nearer or more remote relationship, i. e. of the order followed in the separation of the languages, Bopp's opinion was as follows: in Asia the Sanskrit and Medo- Persian are intimately connected; in Europe the Greek and Latin. In regard to the position of the Slavonic Bopp's opinion changed in the course of time. First (Vergl. Gram., 1=* edition, page 760) he considered the Lith- uanian, Slavonic and German as "triplets" ; later ( Ueber die Sprache der alien Preussen, Ahh. der Berl. Akad., 1853, page 80) he defined his view thus: "The separation of the Slavo- Lithuanian idioms from the Asiatic sister-language, whether we call this Sanskrit or leave it without a name, is of later date than that of the classic, Germanic and Celtic languages, yet prior to the bifurcation of the Asiatic portion of our lin- guistic domain into the Medo-Persian and Indian branches." He did not assume a special relationship between the lan- guages of the Celts and Romans. ScHXEiCHER was the first to establish a formal system of ramification for the Indo-European languages (under the figure of a genealogical tree). He agreed with Bopp in his assump- tion of a closer relationship between the Indian and Iranian branches (which is, indeed, irrefutable), and between the Italic and Greek languages, but differed from him in regard to the position of the Slavo-Lithuanian. He attempted to prove that the similarity of phonetic structure, which indubi- tably exists between the Asiatic languages and the Slavo-Lith- uanian, does not date from primitive times, but originated in each group individually. Thus he assumes that the word for "hundred" in the parent speech was kantam, and that .from this, after the separation of the primitive race into two, gatam The Sepaeation of the Races. 133 was developed in the Asiatic division, and sicto in the Slavonic, quite independently of each other ; so that the similarity be- tween p and s in this word, in which the Greek and Latin have preserved the old k, could not furnish any basis for genealo- gical conclusions. (C{. Beitraffe, 1, T^age 107.) Accordingly, he wholly separates the Slavo-Lithuanian from the Asiatic divi- sion, and with Jacob Grimm places it with the Germanic group. The chief proof of the close relationship of these lan- guages consists in their agreement in the dative plural, where they exhibit an m, while the other languages have bh (e. g. Slavonic vlukomii and Gothic vulfam, but Sanskrit vfkehhyas) . Further, since SoHtiBicHBR places the Celtic with the Italic [Beitrdge, 1, 437), he obtains the following three groups: 1) Asiatic; 2) Slavo - Germanic ; 3) Greco -Italo- Celtic. He defined the historical relation between these groups according to the fidelity with which each (in his opinion) has retained the primitive type. This fidelity seemed to him least in the ' Slavo-Germanic branch ; he therefore assumed that this divi- sion was first separated from the primitive race, and then the Greco-Italo-Celtic, so that the Asiatic group alone remained. It is plain, however, that this chronological classification depends upon a very questionable line of argument. The more advanced phonetic decay of the Slavo-Germanic (if, indeed, it can be regarded as proved) may be simply owing to the fact that the Slavo-Germanic has developed more quickly than its sister-tongues. Schleicher does not, therefore, adduce suf- ficient grounds for dividing the Slavo-Germanic from the great< European mass to which it geographically belongs. That it also belongs there from linguistic considerations was shown by Lottner in Kuhris Zeitschrift, 7, page 18 seq. He I establishes two great groups, the Asiatic and the European, j the latter being especially characterized by a common I in op- I position to the Asiatic r (e. g. iro^o, Gothic Jilu, as opposed to Sanskrit ^wrii) . A further characteristic was added byG. Cur- / Tixjs, in the e which appears uniformly in many positions, in I opposition to the Asiatic a (e. g. tpspw, fero, Gothic haira^ i. e. bira, as opposed to bhdrami). Thus the supposition seemed very probable that the Indo-Europeans, who spoke a uniform language while they were together, first split apart into Euro- 134 Chapter VII. I peans on the one hand and Asiatics on the other, and that ! after the separation certain peculiarities were developed in hoth groups, as, for example, the European e, which sub- sequently clung to all the subdivisions of the main group. For the European branch it seemed necessary to make two such subdivisions, the northern and southern, of which the former was again divided into Slavonic and Germanic, the latter into Greek, Italic and Celtic. The Greek was here the hardest to dispose of. Some scholars assumed that the Celtic first freed itself from the South-European mass, after which the Greek and Italic re- mained together for a while ; others (like Schleickek) advo- I cated the closer community of the Italic and Celtic ; others, finally, divided the Greek wholly from Europe, and trans- ferred it to Asia. This is the decision of Geassmann [Kuhn's Zeitschrift, 12, page 119), who speaks with great certainty of the many phenomena "in which the far-reaching harmony be- tween a Greek and Aryan (pre-Brahmanic) nature becomes evident to us in language, poesy, mythology and life, and bears witness to the powerful intellectual development which the ancestral Greco-Aryan race passed through after its sepa- ration from the other branches." Sonne expresses the same opinion in his apparently forgotten article i) : Zur ethnogra- pMschen Stellung der Griechen, Wismar, 1869. All these hypotheses, so far as they involve the idea of the separation of races or languages, were opposed by Johan- nes Schmidt, in an essay on the relationship of the Indo-Eu- ropean languages [Die Verwandtschaftsverhdltnisse der indo- germanischen Sprachen, Weimar, 1872). Johannes Schmidt starts from the same point where Schleichbe's opposition to Bopp began, namely, the relation of the Slavo-Lithuanian to the Asiatic, but considers Bopp essentially in the right. It is, indeed, very remarkable that in both groups the Jc of hantam becomes a sibilant (or something similar) , while the k of ka 1) I take the occasion to quote a sentence from this article : "But if in Sanskrit the verb of the main sentence assumes an unac- cented form relatively to every preceding objective determination, I think we must recognize in this phenomenon, which is so thoroughly opposed to our European ideas, a remnant of pro-ethnical accentuation." (Page 3.) The Separation OF THE Races. 135 "who" remains. Should not this remarkahle agreement be ex- plained as a result of common development, and is not Schlei- cher's assumption of historical accident inadmissible? If, however, Bopp's view is correct, there is no break between Asia and Europe, but only a "continuous transition" [konti- nuierliche Vermittelung]. And Schmidt finds the same state of affairs in Europe. He recognizes that Greek, Italic and Celtic are intimately connected ; but they do not form a historically distinct group, for as the Italic occupies an intermediate posi- tion between Greek and Celtic, the Celtic, on the other hand, is intermediate between Italic and Germanic, and further, the Germanic between Celtic and Slavonic, etc. Thus we can compare the Indo-European languages to a great chain of dif- ferent rings, so linked together that it has neither beginning nor end. If we begin arbitrarily with the Indo-Iranian, the next ring will be the Slavo-Lithuanian, then the Germanic, Celtic, Italic, until the Greek is finally interlinked with the Indo-Iranian. The Armenian, which has been more accurate- ly investigated only within the last few years, would take its place between the Indo-Iranian and Greek. / It will readily be seen that this transition or "wave-theory" (as its originator christens it, since the progressive movement within the bounds of language can be compared with the mo- tion of the waves) agrees with the ramification theory in giv- ing weight to the points of agreement (some of which have been mentioned) between the separate Indo-European lan- guages, but differs from it in assuming a continous transition in place of ramification. We accordingly must first examine this assumption. I am of opinion that the transition theory is , untenable, if it is understood in the sense that a continuous transition takes place between all Indo-European languages, as they are historically transmitted to us. Against it we have the fact that the separate languages form independent unities, each shut off from the others. It is true that we may be in doubt under which group single dialects (e. g. within the Germanic family) are to be ranged ; but with the chief lan- guages, as for instance the Germanic in its relation to the Sla- vonic, the case is different. There could never be a doubt whether a certain linguistic mass were Slavonic or Germanic ; 136 Chapter VII. fixed boundaries exist between Germanic and Slavonic, as well as between the other chief languages. We are accord- ingly led to suppose that formerly, when the Germanic was spoken by fewer people, it constituted an uninterrupted field of intercourse, within which the separate Germanic dialects were developed in the course of time. The same is true of the other languages. And even if we were willing to make the assumption (which it seems to me cannot be proved, in spite of the ingenuity expended upon it) that the neighboring do- mains of two adjacent languages, like the Slavonic and Ger- manic, stand in closer relation than those more remote from each other, this would only prove that single peculiarities of the former boundary-region had passed over into the two di- vided territories, and that the position of the parts of each domain had suffered no great displacement ; the assumption would still remain possible that the separate Indo-Eiuropean languages have been divided from each other for a long pe- riod by boundaries preventing intercourse. The transition hy- , pothesis must therefore be understood in the sense that in primitive times the languages did indeed form one connected whole, in the manner described by Schmidt, but that then boundaries preventing intercourse were formed, and thus a separate life began, which subsequently gained a rich historical development. This modification of Schmidt's hypothesis, which evidently recommends itself by its vmiversal historical probability, is due to Leskien [Die Declination im Slawisch- Litauischen und Germanischen, Leipzig, 1876). It would ac- cordingly seem that the transition and ramification hypotheses do not unconditionally exclude each other, but are to a cer- tain extent compatible. Unfortunately an objection must be noticed, which pro- ceeds from the stand-point of more recent investigations, and is opposed to both the ramification and the transition hypo- thesis. That is, it has been discovered, by the investigations of the last few years, that the data from which it was- customary to draw conclusions in regard to the closer relationship of in- dividual languages are not so decisive as was hitherto assumed. In general, it is clear that not every point of identity be- tween two languages can be regarded as an argument for an The Separation of the Races. 1 37 original community of life. If, for example, some languages / have lost the augment, which is still possessed by others, of , course it does not follow that this loss necessarily took place during the common life of these languages. It must also be admitted that identity of vocabulary (unless this appears to an overwhelming extent) cannot be used to prove an original community of life, because the possibility always remains that a word which we only find in certain languages existed also in the others, although it has been effaced by the ravages of time. Our material is sensibly diminished by these consider- ations, so that, strictly speaking, we have as conclusive e\i-( dence only those new formations which are developed in com- mon. Under this head were ranked until recently the division of the unitary Indo-European k into k and s [sz] in both the Asiatic and Slavo-Lithuanian families ; the e of the European languages ; the r in the middle and passive of Italic and Celtic ; and the m in the Slavo-Lithuanian and Germanic da- tive plural. But another explanation for these facts has very \ recently presented itself. It is often assumed (as remarked above) that these cases are not examples of new formations in the individual languages, but that the manifoldness must be traced back to the primitive speech. Fick took the lead with his paper on the linguistic unity of the Indo-Europeans of Europe [Die Spracheinheit der Indogerm^nen Europds, Got- tingen, 1873), in which, following AscoLi, he showed that the two sounds of the Asiatic and Slavo-Lithuanian which were previously supposed to have originated from k were really the regular representatives of two different Indo-European ^'s (v. above, page 52). Then followed the very probable theory (also referred to above) that e belonged to the primitive speech ; further, that the r of the middle and passive in Italic and Celtic may possibly stand in connection with the r of the In- dian -»*e, -^ate etc. (cf. Windisch . Beitrage von Kuhn -und Schleicher, 8, page 465, note) ; and that the m of Slavonic and Germanic perhaps belongfed originally not to the 5A-suffix, but to another. If, now, this whole mode of reasoning is justified (as I assume), then from such differences as these, which reach back into the primitive speech, no conclusions can be drawn 138 Chapter VII. respecting the successive ramifications of the Indo-European languages, and it is necessary to adopt a skeptical position with regard to all the groupings hitherto attempted, with the single exception of the Asiatic group, which is held together by the common change of the old e into a. In fact, I consider this stand-point the correct one at the present stage of investigation , and accordingly I think that our assertions in regard to the whole question of the mutual relation of the separate Indo-European languages must be re- duced to the following. It is very probable that the primitive speech was not entirely homogeneous, as there was formerly an inclination to suppose. For if we are right in assuming that this speech passed through a development of centuries, the primitive race must have been very numerous at the time the inflection was fully perfected, and[therefore differences in speak- ing must have already begun to manifest themselves within its limits, as described in general terms above (pages 52 and 59) . These differences are the germs of some of the differences which we observe in the Indo-European languages. Others were added to these, after the primitive speech had divided into various individual languages. It is possible that the fore- fathers of the later Greek, Italic and Celtic nations were formerly settled beside each other in the way we are led to suppose from their 'present geograpTiical position; but it is also possible that great displacements of the races have occur- red, which render their former situation obscure. We will therefore content ourselves for the moment with acknowledging an original community of the Indo-European languages, but must abstain from classifying them into groups, with the ex- ception qf the Indo-Iranian. This is true with regard to the Greco-Italic unity so often assumed. It is impossible to affirm with certainty that this unity did not exist, but it is equally impossible to assert that it can be demonstrated. Of the reasons adduced in its favor M 1) Schmidt has very properly not introduced the word-comparisons of MoMMSEN, as they prove nothing. For a part of the words in question can also be found in other languages (as Mommsen himself acknowledges in the later editions of his Roman History) , and the others (like milium, rapa, vinum) are possibly or probably borrowed words. The Sepabation of the Races. 1 39 (Schmidt, page 19), the only ones which concern us in the present state of investigation are the two following : the fact I that Greek and Latin are the only languages which -have I feminines of the second declension ; and the agreement in the j accentuation. However, if it is true, as I have tried to prove ( in Synt. Forsch., 4, page 6 seq., that the masculines in -to. of I the first declension were transferred from feminines to mas- culines only in the independent life of the Greek language, then an analogous process may be suspected for the above- mentioned class of words; and in regard to the laws of accent, it is a question whether it is not possible to find traces of an older accentuation in Italic, which prevented the "three-syl- lable law" from gaining the supremacy in a pre-Italic period. At all events, a hypothesis of such significance as that of an original Greco - Italic unity cannot be founded upon a ques- tionable assumption. Whethef the future will attain to more definite results, remains to be proved. In the mean time, it will be well for historical investigators to abstain from making use of such lin- guistic and ethnological groups as the Greco-Italo-Celtic, Slavo-Germanic etc. INDEX. Abstract verb, v. Substantive verb. Adaptation theory of A. Ludwig, 66 seq. Adelung, 9, 80. Agglutination theory : a name for Bopp's explanation of inflectional forms, 16 ; agglutination in the in- dividual languages, 14, 46, 57; judgment of this theory, 61 se- %i •f .*«r. .*?- *i?if