Philology.

Philology. The meaning of this term has varied curiously. As we find it first employed by Plato, it meant the love of discussion, limited practically to the moral and social questions in which Plato delighted; and the method of such discussion was the Socratic one. At Alexandria the 'philologer' was busied with all the knowledge of his day brought together for the first time in the great library of the Ptolemies. Thus, Eratosthenes, who applied this term to himself, was specially famous as a writer on astronomy. But the great scholars of Alexandria applied themselves with especial eagerness to the study of the older Greek literature. Men like Zenodotus and Aristarchus compared the numerous MSS. of Homer, selected the best texts, made lists of difficult words, called glossæ—the earliest dictionaries—and were the founders of the science of criticism by establishing canons on which they rejected what they deemed spurious in the copies before them. Two centuries later at Rome Cicero, who uses the word not unfrequently in his letters, applies it to study in general. But ever after the Alexandrian time 'philology' was essentially the intelligent critical study of the traditional learning of the past. It widened again at the revival of learning to include the study of grammar, rhetoric, poetry, archaeology—in a word, all the 'humane' studies. And this wider sense clung to the word. Johnson in the last century defines a philologer as 'one whose chief study is language, a grammarian, a critic.' Yet Watts, whom he quotes as one of his authorities for the use of the word, says that 'studies called philological are history, language, grammar, rhetoric, poetry, and criticism.' In fact philology was the study which interpreted the best writings; and these writings were Greek and Latin; and in these languages were to be found the most noteworthy speculations on all matters of human interest. So philology had its special and its wider sense. Since the middle of the 19th century the use has been changing in a different direction. Through the discovery of Sanskrit the scientific comparison of Greek and Latin with other languages became possible. That study was called 'comparative philology.' But this term is cumbrous, and in England, and to some extent in France, it is customary to speak in this sense of 'philology' alone. So, whereas formerly philology meant pre-eminently the knowledge, grammatical, critical, exegetical, which enabled men to explain the most important literature of the world, it now is becoming limited to the study of language apart from the literature embodied in it—to the science which deals with the origin, development, and general character of the different families of speech throughout the world and of human speech as a whole: to which, therefore, the most highly cultivated languages may be not more important than languages which have hardly any literature or none at all: because such languages develop naturally, whereas literary languages are subject to artificial restraint. In Germany, it may be noted, this special study of language is called not 'Philologie' but 'Linguistik.'

It must not be supposed that this scientific study of language as an end in itself has superseded the older philology of a Bentley or a Porson. Rather, it has helped it by throwing light on the forms and meaning of words and phrases isolated or obscure in one language, but frequent and clear in another. But indeed the work of the critical scholar of the present day has been widened to a degree which could not have been realised three generations ago. First, the art of interpretation has been developed; the old storehouses of MSS. have been more fully searched; the principles of palæography are more widely known, so that we have many more scholars capable of dealing at first hand with MSS., of deciding upon their age, origin, and relative value. The sister-art of epigraphy, which deals with inscriptions, has ever-increasing material to work upon. Secondly, the science of archaeology has been almost re-created. Explorations in all parts of the Hellenic world—Attica, Delphi, Peloponnese, Cyprus, and Asia Minor—especially the recent search in the Acropolis at Athens, have profoundly affected our ideas of Greek art, both plastic and constructive. The long-continued excavations at Pompeii have shown us the domestic life of the 1st century in the most minute detail. There can be no doubt that we have yet much to learn, and that the zeal of modern students will be equal to the task. Again, historical inscriptions and coins have corrected and supplemented the statements of ancient writers, and have often given us unexpected and perfectly trustworthy information as to periods unrepresented in any literature. These and other auxiliary sciences have given new life and meaning to the scholar's work. They make it at once more satisfying and more difficult: there is so much more to be known. Formerly a really great scholar could master the whole domain of knowledge; now every man must depend for something on the labours of others.

The study of philology, especially the study of grammar, after it was transplanted from Alexandria, flourished at Rome in the imperial time. It lingered on at Constantinople till the line of great grammarians ended in Priscian; while lexicography culminated later in the work of Photius and of Suidas. But it was in Italy that the study of antiquity became again a living thing: Italians in the 14th and 15th centuries discovered in Italy, Germany, and France works of the greatest Roman writers, such as Cicero and Lucretius, which had remained for generations buried in convents; it was to Italy that learned Greeks, especially after the sack of Constantinople, came to teach the Greek language at Florence and at Venice, and brought with them the MSS. of Thucydides and of Plato, to be translated by Valla and by Ficinus; and it was in Italy that the results of their labours were first published to the world at large by the great houses of the Manucci and the Ginnta. But with the exception of Faerno Italy had no great scholar. Yet by its connection with France it produced scholars in northern Europe. Budé was secretary to Louis XII., whose main interests were in Italy; Lambin visited Rome and Venice before he became the editor of Horace, of Lucretius, and of Cicero; while Muret, though born near Limoges, lived and worked at Rome from 1559 to his death in 1585, and at Rome he lodged for two years the young Joseph Scaliger, destined to become the foremost of European scholars. Italian by descent, but born in France, where he edited his Manilius and wrote his De Emendatione Temporum, the first great work in historical criticism, he was led by his conversion to Protestantism eventually to retire from France to Leyden, where he ended his days, the dictator of the world of letters. Isaac Casaubon, a leader in exegesis, as Scaliger was in criticism, though born at Geneva, was the son of a French Huguenot refugee, and in France he lived during the greater part of his life, till like Scaliger he found a securer home elsewhere—in Protestant England. Justus Lipsius, the third great scholar of the day, was a Belgian, professor successively at Jena, at Leyden, and at Louvain; he also had travelled in Italy.

Of the members of the older German school the most famous was Erasmus, by birth a Hollander, but the centre of a band of able scholars at Basel, where he spent the last sixteen years of his life, well known in England, and for a short time a professor at Cambridge, a man of vast learning, but not a master in criticism, 'the man of letters, the first who had appeared in Europe since the fall of the Roman empire' (Mark Pattison). He lived some three years in Italy, but gained, as he thought, nothing from it. Able scholars were Camerarius, professor at Leipzig, and Gruter of Antwerp, the first great collector of Greek and Latin inscriptions.

For different reasons France and Germany ceased to be the nurse of scholarship in the 17th century. The reign of Louis XIV. fostered modern rather than ancient literature; and Germany was the scene of furious war. But Scaliger's influence lived on in Holland. At Leyden, where he died, lived Daniel Heinsius and his son Nicolas, Gronow, conspicuous for his skill in appreciating MSS., and Clüwer, the first great writer on geography. At the same time Graefe was editing Cicero at Utrecht. Somewhat later in England lived the first of English scholars, Richard Bentley, in frequent correspondence with Dutch scholars, but owing nothing to them, a man whose astounding critical power could not always save him from errors due to his self-confidence. The only other Englishman whose fame, like Bentley's, has steadily grown with time was Richard Porson, professor of Greek at Cambridge at the end of the 18th century. Distinguished names in the school of Holland in the 18th century are Hemsterhuis and his pupils Ruhnken and Valckenaer. In Germany we find Ernesti at Leipzig, the editor of Cicero; his scholar, Heyne, the founder of the school of Göttingen; Reiske of Leipzig, skilled not only in Greek and Latin, but also in Arabic, whose edition of the Greek orators is still in use; Winckelmann, the first great writer on ancient art; Wolf, professor at Halle, best known as the great Homeric critic, but whose general power and method almost entitle him to a place by the side of Scaliger and Bentley. Scholars of a later date, whose lives extended far into the 19th century, are Immanuel Bekker, professor at Berlin, editor of Plato, Thucydides, the orators, and Aristophanes; Godfrey

Hermann, a scholar of unusual breadth, whose fame rests securely upon his work on Æschylus; Welcker, professor at Bonn, the first who combined the study of Greek art, literature, and mythology; K. O. Müller, whose services to the study of ancient history were cut short by his premature death; F. Ritschl, professor at Bonn, the restorer of the text of Plautus, as Lachmann at Berlin was the restorer of Lucretius. We have only of late suffered the loss of H. A. J. Munro of Cambridge, the inheritor of Lachmann's labour on Lucretius; of Madvig the Dane, a Latin scholar of eminent acuteness; and of Cobet, in whom the critical power of Holland seemed to be renewed again. (An excellent sketch of the history of classical philology, by Dr L. von Urlichs of Würzburg, will be found in the first volume of Iwan Müller's Handbuch der Klassischen Altertumswissenschaft.)

Speculations on the connection of Greek and Latin—e.g. that Latin was derived from some Greek dialect, and that both, as well as all the other languages of the earth, must be derived from Hebrew—are not wanting in the writings of the older scholars. It was reserved for an Englishman, Sir William Jones, in 1786, to point out that Sanskrit, Zend, Greek, Latin, Gothic, and Celtic belong to the same family; and for Franz Bopp, born at Mainz in 1791, to become the founder of the special science of comparative philology. In his 'conjugation-system,' published at Frankfurt in 1816, he worked out the details of the principle already established by tracing out the history of the verb inflections of the Greek, Latin, Old Persian, and Tenticic as compared with Sanskrit. His monumental work, the Comparative Grammar, appeared at different times between 1833 and 1852. In this he lays down the phonetic laws of the several languages, and traces their grammatical forms back to their common origin in a lost 'Indo-Germanic' speech. It is impossible here to describe the development of this work in the hands of the singularly able men who laboured at it in the same generation, such as Jakob Grimm, the founder of the scientific study of the Tenticic languages; Pott, the most learned and voluminous of writers; Benfey, the acute philologist and accomplished Sanskritist. The most successful application of the science to Greek was made somewhat later by Georg Curtius, and to Latin by Corssen, and in France by Michel Breal. The common principle of all these writers was that the never-ceasing change in every language is regulated by law; that in each language there is a regular sequence of sound, one passing into the other, not by chance or by the will of any speaker or speakers, but in a definite ascertainable course; and that only by the knowledge of these sequences, commonly but not very wisely termed 'laws,' can the science of language exist. One of the best known of these is the sound-change commonly called 'Grimm's Law,' which states the fact that whenever we find a k, t, or p in Sanskrit, Greek, Latin, and (inferentially) in the common parent-language, we shall find in English and most other Tenticic languages an h, th, f—e.g. kapδ-la (Lat. 'cord-') will appear as 'heart,' 'tres' as 'three,' 'pes' as 'foot'; that g, d, b will appear as k, t, p ('genus' = 'kin,' 'duo' = 'two'); and aspirates (gh, dh, bh) will appear as g, d, b (e.g. Ind.-Ger. 'ghans' = 'goose,' 'bhāgos' = 'beech'). Further, that a subsequent 'shifting' of the same kind, but much less complete, took place many centuries later in Germany itself, and produced there the changes which distinguished the dialects of South Germany ('High') from those of the north ('Low German,' or, as it is called in Germany itself, 'Platt-Deutsch') and from our own English—e.g. 'drei' from 'three,' 'zwei' from 'two,' &c. Such astounding uniformity of change over so large an area was well calculated to confirm the belief in the 'reign of law' in language. But that belief was certainly strengthened when in 1875 Karl Verner wrote a paper (published in Kuhn's Zeitschrift, vol. xxiii.) in which he showed that some apparent exceptions to Grimm's Law were really due to the operation of another and wider law (commonly called by his name) dating from variation of accent preserved by the old Teutonic from the parent-speech; in consequence of which at the present day a west Cumberland dalesman says 'fadder' and 'mudder,' but 'brother,' because in the parent-language (as attested by Sanskrit and Greek) the stress was laid on the last syllable in 'patér,' 'matér,' but on the first syllable in 'bhráter,' and every Englishman says 'land,' but 'earth' (in the oldest Teutonic 'hlūdā' and 'értha,' with the final vowel still surviving), because the suffix (Ind.-Ger. -to) which was common to both was accented in the first word, but not in the last; so also 'mind,' but 'growth' (suffix -ti), 'flood,' but 'death' (suffix -tu). Here the regularity seems little less than miraculous; and it was discoveries such as this which led about this time to the rise of what is sometimes called the 'new school' of philology, men who in reality only press to their furthest limit the principles of their predecessors. For example, Curtius and Schleicher held that the operation of unvarying law in certain cases did not exclude the possibility of 'sporadic changes'—i.e. of changes found in some words only, not in every word in which the sound so affected occurred: thus, 'lacrima' in Latin shows an l instead of the d of the original word, as proved by Greek δακρυ and our 'tear;' but they did not therefore think it necessary that every d should pass into l in Latin.

The 'new school'—Leskien, Brugmann, Osthoff, Paul (to mention a few only of the most conspicuous members)—hold that sound-change so far as it is due to physical causes is absolutely uniform in any one language or dialect at the same time; and that the new form produced by such change in every case drives out the old one. Briefly put, phonetic law is invariable; there is no such thing as sporadic change—change attacking a few words, and sparing others; throughout the whole speech-area all words in which a particular sound occurs are alike affected. This doctrine has commanded very general assent; it is tempting to the scientific mind; and its strength lies in the number of apparent exceptions which have been satisfactorily explained. Yet the a priori arguments against it are strong, and it is certain that it is incapable of absolute proof: you cannot prove a negative. But the good done by its supporters in every branch of philology has been immense. They protested, and rightly, against the habit, seen in the later writings of Curtius and other lesser men in other languages, of allowing obvious exceptions to ascertained phonetic law on the ground of identity of function. Be the principle of the new men right or wrong, its observance in practice is excellent. But this protest is only one of the services of the new school. They (especially Paul) have called attention to the inner side of language; the older writers had spent themselves on the outer side. For language is both a physical and a psychical product. The sounds which make language are due to the physical apparatus of speech. As such their nature can be exactly ascertained, and they are subject to changes which can be known and registered like the sequences of any physical science (see PHONETICS), and may be called the outer side of language. The inner development is due to the mind of man; and this, like all the other works of man, belongs to historical and not to physical science. Here the ruling principle is imitation. Every child learns every word it utters through imitation, and imitation of words heard from others or previously uttered by ourselves is the parent of the language of every grown-up man. That we are wholly unconscious of the process does not make the fact any less certain. One form of imitation especially active in speech is what grammarians call 'form-association.' Different forms of speech—e.g. preterite tenses of many verbs, or a particular case of many nouns—are alike each in their grammatical function, and tend to be thought of, and to be used, together. But these forms are nevertheless often very unlike; though their use may be the same they may have different origins—e.g. 'I swam,' 'I stood,' in English owe their distinctive form to a vowel-variation, but 'I heard' to the adding of a suffix. Now, the mind is constantly acting under the influence of analogy to reduce such variations, to do away with unnecessary differences, to reduce old forms to one level, or to make new forms on the analogy of old ones. Thus (see under article GRAMMAR), we now say 'we ran,' not 'we run' as our forefathers did, because the reason for the different vocalism of singular and plural is no longer discernible, and the unmeaning difference is 'levelled.' Again, we say 'I climbed,' not 'I clomb' any more, because the preterite in -ed is the simplest and commonest form of the tense, capable of being applied without difficulty when a new verb becomes necessary—e.g. 'we boycotted him.' This principle of 'analogy' (as for brevity it is often called) acts widely in every language; it conditions every new word we make, and it must have acted much more upon vernaculars which had no literature (a good example may be seen in the reconstruction of the old Latin verb in modern Italian), and still more upon wholly unwritten languages.

There are then two chief factors which act on language: one, the vis inertiae, which is the prime cause of phonetic change, and is in the main destructive, doing away with sounds or combinations of sounds which, owing to causes varying with nations and even with individuals, are inconvenient to produce; the other, the imitative habit of mind, which may destroy old forms, but is in the main reparative, giving new forms for those which through phonetic change had become obscure, and constantly producing new forms on the analogy of old ones to supply each want caused by the progress of the human mind. For except in very rare cases man turns old speech-materials to new uses. He does not invent absolutely new names for new things as they become known to him or are produced by him. A hippopotamus was called a 'river-horse;' the trees and beasts of Australia are known by the names of quite different species in Europe. A new invention is called by some descriptive name, as a 'photograph,' which has superseded the older but equally descriptive Talbotype and Daguerreotype: when Talbot and Daguerre were forgotten their names were less suggestive than terms with a meaning. Mesmerism in like manner will be superseded by hypnotism, there being a sufficient number of people able to understand the new name, and so to familiarise it to those to whom it means nothing. Similarly, old names may take a new value, becoming thereby really new words if the connection with the old thing is quite broken off. Thus, a 'bead' to us is a round object made of different materials, absolutely unconnected with the same word in Old English which meant a prayer: the link—counting prayers by means of a string of beads—being no longer a matter of common observation. Sometimes both the old and the new meaning co-exist: a man may be called a 'lion,' but a lion is a lion still. Lastly, new words expressing action—verbs and verbal nouns—constantly spring up, being commonly either more picturesque or giving some new circumstance. To murder is old, but to 'burke,' a verb coined from a once famous murderer, gives the added sense of smothering and hushing up; and the word seems likely to survive, though with its origin forgotten. To 'boycott' expresses a slightly new form of exclusive dealing, and as yet its history is remembered.

This slight sketch may suffice to throw some light on the nature of language. It is a work of man, the product of man's mind and vocal organs, just as a statue or a picture is the product of his mind and hands. But it differs from them in some important respects. First, it is not a finished product, permanent and unchangeable. It is subject to incessant change, involuntary on its physical side, partly voluntary on the mental side. It is a constant flux, a 'becoming,' not a 'being,' as Plato might have put it; this change is obscured for literary languages; it becomes clear at once to any one who will take the pains of looking at the development of English from Chaucer's day to ours. Old forms die out, and new ones take their place (see GRAMMAR). Old families of words cease to be used: new ones take their place or are produced under new needs—e.g. the Boycott family—boycotted, boycotting, boycotter, boycottee (extinct by this time), &c., all formed on the analogy of the words of other older families. A second point of difference is that language is not an end in itself as a picture is: it exists for an end, communication between man and man. So long as this is achieved the form which language takes is immaterial: it may change so long as it is intelligible. No doubt language once existent can serve other ends more or less connected with the first; but that communication was the first is undeniable.

But language is not the only means by which communication can be made. Animals, which have no language, certainly communicate. Man can communicate by gestures, by pictures—whether rude scratches made on the ground with a stick or the more polished drawings which developed into Egyptian hieroglyphics—or, lastly, by cries common to man and to beasts, the natural expressions of joy, fear, pain, &c. But such cries are not even the elements of language till they are consciously reproduced to express the feeling even when the stimulating cause may be absent. Language itself arises when two men connect the same feeling with the same expression of it, and so can communicate that to others. There is no reason to believe that any brute has ever attained even to the first of these last two steps. Their progress is arrested. A dog may bark to express delight, or to have a door opened to him, but he does no more than any dog could do 2000 years ago. Man can develop. It is not permissible, however, to lay down that the possession of speech is the barrier between man and the brute, and to settle thereby the question of the origin of man. Speech may be the clear differentia now. But it is at least conceivable that there may have been lost types between man and the common progenitor of man and the anthropoid ape with intermediate stages of speech-development. It is another matter to maintain that such development could have been produced by natural selection alone (see Wallace, Darwinism, p. 461, ff.).

The different languages of the world can be classified according to their principle of formation; and within the classes so reached different families of languages may be distinguished in which a common origin can be proved for the different languages of each family. Only the briefest sketch can be attempted here. We find two main classes: I. Those languages which show no signs (or hardly any signs) of inflection (see GRAMMAR)—e.g. in which the plural of 'man' is ex- pressed not by vowel-change (as our 'men') nor by an added suffix (as in Latin 'homin-es') which has no independent value, but by such a combination as our 'man-kind,' where each part can be used independently. Such languages are the Chinese, the Tibetan, and those of Annam and the neighbouring states; they are commonly called Isolating. II. Those which possess some degree of flexion—i.e. elements which have lost their independent meaning, and are mere grammatical machinery to make nouns, as ter in 'pa-ter,' or cases of nouns, as s in 'father-s,' or persons of verbs, as s or th in 'gives' or 'giveth.' Such elements, however, are only the worn-out remnants of words compounded with other words (see GRAMMAR); and the languages of this type may vary very much according to the degree of obscurity in the character of the compound. Some, like the Mongolian, the Finnish, the Hungarian, the Turkish, make very long compounds, yet the original elements, though not all capable of separate use, remain quite distinct and recognisable in each word. These languages used to be called 'agglutinative,' as different from the specially 'inflectional' Sanskrit, Latin, &c.; but even these show signs of the phonetic change which produced 'inflections.' The North American languages incorporate different elements which are each barely recognisable in the compound; the principle of composition is not really different. The Dravidian family of languages in South India—Tamil, Telugu, &c.—is also of the agglutinative sort. The languages of the purely inflectional type—i.e. those where the different elements of the original compound are so firmly welded together as to be frequently indistinguishable—are (I.) the Semitic family, comprising Hebrew and the closely-connected Phoenician; Aramaic, spoken in Mesopotamia, Syria, and in later days in Palestine; Arabic; and some Abyssinian languages. This family is remarkable because of its trilateral roots—i.e. expressions of the several ideas by three unchanging consonants, the relation of the various derived forms of the same idea being expressed by vowel-change. (II.) The Indo-Germanic or Aryan family. The first term is most likely to survive; it denotes the extreme limits of the area over which the languages spread—from Sanskrit India to Germanic (or Teutonic) Iceland. The term Aryan has not been adopted much outside of England. The chief languages of this family are (1) Sanskrit, of which the oldest remains are the Vedic hymns, with the cognate Old Persian and Zend, the language of the Avesta. (2) Armenian, as yet imperfectly known, with records dating from the 5th century A.D. (3) Greek, with its numerous dialects. (4) Albanian, proved to belong to this family by Bopp, and lately investigated by G. Meyer and others, but possessing no ancient records. (5) Italic, including Latin, and the Umbrian and Oscan dialects; from these are descended the modern (so-called) Romance languages—the Italian, Spanish, Portuguese, French, Wallachian, and the speech of certain Alpine districts (Grisons, &c.). (6) Celtic, including the ancient speech of Gaul, with its surviving remnant the Bas Breton; the now extinct Cornish; the Welsh, still likely to survive; the Erse of Ireland; the Gaelic of the Highlands of Scotland (the records of these three date from about the 8th century); and the Manx. (7) The Teutonic, now more commonly called Germanic, comprising Gothic, the language into which Ulfilas translated the Gospels in the 4th century A.D.; the Scandinavian, of which a very old form is isolated in Iceland, more modern forms in Norway, Denmark, and Sweden; the Anglo-Saxon; the Old Frisian; the Old Saxon of the 'Heliand,' the parent of the Platt-Deutsch languages of North Germany; the Lower Franconian, whence comes the Dutch; the Franconian of Mid Germany; and the Old High German, spoken in different dialects from South to Middle Germany, whence is derived the modern literary German. (8) The Slavonic, to which belong Bohemian, Polish, Old Bulgarian, and the parents of the Russian and numerous dialects of south-eastern Europe; together with the Lithuanian, Old Prussian, and Lettish languages of the south Baltic, which, though their records are late, yet preserve strikingly early linguistic features. See ARYAN LANGUAGES.

That there was such a language as the hypothetical parent of all these Aryan languages is certain (though such knowledge brings us no nearer to the one original language—if one there was—spoken by primeval man). We can recover its character with certainty, for we know its suffixes both formative and inflectional; its vocabulary we know in part only. Now, if there was a language there must have been a people to speak that language—no doubt with many dialects, whence eventually sprang the derived languages which we know, and probably many others which are lost. Dialectal variation is the inevitable condition of all languages where there is no common literary tongue: that tends to level the speech of the whole area: and where there is great facility of communication throughout the area the dialects may die out, as they are rapidly doing in England. But though we assert with confidence that there was a common Indo-Germanic language, spoken by an Indo-Germanic people, we do not maintain that the people who speak the several languages derived from that the common parent are wholly of the same race. England, as we know, is inhabited by the descendants of Celts, of various nationalities enrolled in the Roman legions, of Angles, Saxons, Jutes, Frisians, of Norsemen, and of Franco-Normans; all of whom speak the same English language. But that fact does not make us doubt that there was once a common Tentic language spoken (in various dialects) by the different members of a Teutonic race which occupied Britain. It is extremely probable that there has been a like history in other lands where other Indo-Germanic languages are now spoken. That does not in the least disprove the existence of an Indo-Germanic race, speaking different Indo-Germanic dialects, settling themselves in the different countries of Europe and western Asia, and mixing their blood with that of the races whom they found already there. The proportion of Indo-Germanic blood in any given nation may be considerable: it may be very little. It is possible that in some nations it may be nothing at all: some stronger, but less civilised race may have overpowered the Indo-Germanic stock, but taken their language. Such loss has its parallels in history—e.g. when the Norsemen conquered part of France, but lost their own language. This would explain the fact that races of marked racial difference—with short skulls, and dark hair and colouring—are found speaking cognate languages with men of long skulls, ruddy colour, and light hair. Such an inconsistency has seemed to some anthropologists to absolutely destroy the value of language as a test of race. It does nothing of the kind. Linguistic arguments are as valuable as anthropologic ones; but neither give conclusive proof, only indications. The two sciences should work independently. Good anthropologists, such as Pösché and Penka, may be bad linguists; but results drawn from harmony of the strongest evidence on either side may be fruitful. Most, however, of these problems will probably never admit of certain solution. The evidence which now would place the habitat of the parent-race in northern Europe is only somewhat more weighty than that which formerly placed it in Pamir.

See especially Paul's Principien der Sprachgeschichte, translated by Strong (1888); and the History of Language (based on the same book), by Strong, Logeman, and Wheeler (1891). Whitney's Life and Growth of Language (Inter. Sc. series), Max-Müller's Essays on the Science of Language, and Sayce's Introduction to the Science of Language are useful; the latter contains a full list of authorities. By far the best book on the history of the Indo-Germanic languages in their earlier stage is Brugmann's Comparative Grammar (trans. vol. i. by Wright, vol. ii. by Conway and Rouse, 1888). Here will be found mentioned all the recent works of any importance in this department. A short book referring chiefly to Greek and Latin is Victor Henri's Précis de Grammaire Comparée (1888), also translated. For Romance languages the latest and most complete work is Gröber's great Grundriss der Romanischen Philologie (Strasburg, 1886-88), in which the different Romance languages are treated by the most competent authorities. Diez's Grammatik der Romanischen Sprachen, though somewhat out of date, is clear and good. A similar book on the Germanic languages is Paul's Grundriss der Germanischen Philologie (1889 et seq.). Special students of English should use Sievers-Cook's Grammar of Old English (Ginn, Heath, & Co. 1885), Sweet's History of English Sounds (1888), and Skeat's Principles of English Etymology (1887 and 1891). For Phonetics, Sweet's Handbook of Phonetics should be consulted. All questions on the early history of the Indo-Germanic people and its probable habitat are exhaustively treated in the Prehistoric Antiquities of the Aryan Peoples (Schrader-Jevons, 1890). Isaac Taylor's Origin of the Aryans (1890) is a useful smaller book on the same subject. The grammars and other works dealing with the modern languages of Europe are too numerous to be mentioned here.