lowed him in the same field, have led to the most important consequences in philology, and have finally vindicated that science from the old charge against it, that it was either wholly sterile, or abounded only in empty and unprofitable speculations. We propose now to explain a few of the principles on which, by the best authorities, the vocalismus of the Indo-Germanic languages is considered to depend. Our illustrations shall be mainly directed to elucidate the affinity, which it is so pleasing to trace, between the classical languages and the mother-tongue transmitted to us by our Teutonic ancestors; and we shall seldom diverge into other regions, except so far as may be necessary for connexion and explanation. It is essential to observe, that no one Teutonic language in a modern form, is in itself a fit subject of etymological comparison with the languages of antiquity. The forms of speech now in use among us, are indeed bound by a "natural piety" to their immediate parents; and these again in their highest gradation can count kindred with the languages of Greece and Rome, Persia and Hindostan. But between the two extremes, the distance is too great to admit of an immediate communication such as can be safely relied on. A mere English or a mere German scholar, knows nothing genealogically of his own language. He cannot with certainty distinguish a compound word from a simple one, a derivative from a primitive, an exotic from one of native growth. Wherever, therefore, a philological comparison with the ancient languages has been attempted, either in England or in Germany, by those who knew no Teutonic lan guage but their vernacular tongue, their speculations have been always unsuccessful, and often ludicrous. In order to obtain the proper point of view from which the Teutonic formations can be compared with the classical, we must gain the vantageground afforded by the most ancient of the Teutonic chain, and thence direct our observations to the most primitive of the remoter languages. The Gothic, as might have been expected from its being the oldest Teutonic language known, has the signal advantage of exhibiting its elements, its vowels as well as its consonants, in a shape at once the most authentic and the most uniform: and to this standard it is always best to refer the other Teutonic languages before we compare them with those of an earlier date. Unfortunately the literary monuments of the Gothic, are too limited to furnish us with the counterparts or roots of all or even the greater part of modern vocables. But in the absence of this direct assistance, we are generally enabled to supply the defect by the laws of mutual comparison among the later languages: and if we have a given word in any two or three even of the most modern Teutonic forms, it is seldom difficult to approximate to what it must or would have been according to its Gothic type. The precaution we have in this respect recommended, of recurring to the most ancient Teutonic form in the first instance, is but ill observed by rash or inexperienced philologers: but it is never omitted with impunity. Let us now endeavour to explain some of the leading principles on which this part of the science seems to be founded. The single and short vowels of a language are alone considered to be primitive; its long vowels and diphthongs are regarded as derivative. The explanation, therefore, of any system of vowels, is intimately connected with prosody. But in modern languages great uncertainty and confusion has here been introduced; and for this reason among others, it becomes necessary to recur to the ancient forms of Teutonic speech in which the rhythmical quantity of the vowels was for the most part settled and certain. The primitive vowels of the IndoTeutonic languages are considered to be three in number, ă, ĭ, u; pronounced nearly after the Italian fashion. The other vowels, whether short or long, are held to be modifications of these. It is doubtful if this division is founded in universal nature: but it seems to be observable in the family of languages which we are here considering. The three vowels above enumerated are the only primitive vowels in Sanscrit and Gothic; and in the remaining languages there seems a law or tendency, according to which all the other vowel sounds seek to range themselves under one or other of three sections corresponding to the three simple vowels we have mentioned. All the roots of words, as appearing in their purest monosyllabic form, are necessarily characterised by one or other of these three vowels; and in the perfection of this theory, the vowel which characterises a root in any one language, ought to remain its charac-a." It will not from this proposition teristic in all the cognate languages into which it may pass, subject always to such known and definite modifications as the vowels are found to admit without changing their essential nature. The modifications of the vowels to which we have here alluded, consist of certain regular processes to be afterwards explained, by which they are severally augmented, weakened, or diphthongated. A root, which in Sanscrit is characterised by a, ought to retain the same vowel, in a pure or modified form, in Greek, Gothic, German, and English, and so also of i and u. To a great extent, accordingly, a uniformity in preserving the radical vowel through the different languages is found to prevail; though there are anomalies, real or apparent, which break in upon the rule. But these are much less numerous than they appear: they generally occur in circumstances where the word or language exhibiting them has otherwise been corrupted; and in many cases they vanish altogether before a more searching examination into the facts. It is thought that even the slight and superficial explanations which we are about to offer, will give clearness and consistency to much that may formerly have appeared to our readers capricious and obscure. by Bopp. "The Sanscrit," he observes, "has no short e and o, or at least no letters for these sounds; and its short a is represented in cognate Greek words for the most part by &, more rarely by, and least frequently by be inferred, that there is no distinction between the Greek vowels a, &, ; but the circumstance that these different vowels indiscriminately represent one simple sound in so important a cognate language as the Sanscrit, entitles us to believe that a peculiar affinity subsists between them. Whether it is to be explained by supposing a finer and more chromatic subdivision of sounds in the Greek, or a scantier vocalic alphabet in the Sanscrit, the existence of the rule is of obvious value in confirming the close connexion of these vowels with each other, and enabling us to compare them with sounds of the same class in other kindred languages. We shall offer, by way of illustration, a few examples where the Sanscrit a passes into each of the Greek vowels we have mentioned. In some of our instances the Sanscrit vowel is long; but this does not destroy. the affinity, though it disturbs the precise correspondence. We shall now notice the three primitive vowels A, I, U, in their order. I. Of these, the vowel A is the most important and remarkable, both from the frequency of its occurrence, and the multiplicity of its modifications. The roots characterised by this vowel are, in all the Indo-Teutonic languages, the most numerous of any. It seems the most easy to our lips, and the most agreeable to our ear. Its diversity also in point of aspect or utterance, is peculiar; and we are tempted to say of it, analogously to what has been said of the colour of green, that it imitates in its various changes all the shades of sound, from the lightest and liveliest to the most deep and sombre. A most important law for comparative philology has here been announced We may here observe, that in the Latin language also the short e and o are used indiscriminately as belonging to the A class, and as corresponding to the Sanscrit a. Our examples shall include Latin as well as Greek words. The rule, we repeat, is, that the Sanscrit a corresponds not only to the Greek and Latin a, but also to the short e and o, and, of these languages. We have various instances of this in the numerals of these different languages. Thus, the Sanscrit panchan, five, corresponds to the Greek ; shash, six, to the Greek i, and Latin sex; saptan, seven, to ira and septem; ashtan, eight, to ox, octo; nawan, nine, to -Fa, and novem; daçan, ten, to dina and decem. Many miscellaneous words may likewise be noticed. Danta-s, the Sanscrit for a tooth, assumes in Greek an internal, in -des, o-davros, besides prefixing a vowel. In Latin it takes the e, in dens, dentis. Pada-s, S. for a foot, undergoing the same changes, becomes Tus, Todos; pes, pedis. Naman, S. for aname, seems to correspond to voμa, as much as to nomen, though we sus lowed him in the same field, have led to the most important consequences in philology, and have finally vindicated that science from the old charge against it, that it was either wholly sterile, or abounded only in empty and unprofitable speculations. We propose now to explain a few of the principles on which, by the best authorities, the vocalismus of the Ins do-Germanic languages is considered to depend. Our illustrations shall be mainly directed to elucidate the affinity, which it is so pleasing to trace, between the classical languages and the mother-tongue transmitted to us by our Teutonic ancestors; and we shall seldom diverge into other regions, except so far as may be necessary for connexion and explanation. A It is essential to observe, that no one Teutonic language in a modern form, is in itself a fit subject of etymological comparison with the languages of antiquity. The forms of speech now in use among us, are indeed bound by a "natural piety" to their immediate parents; and these again in their highest gradation can count kindred with the languages of Greece and Rome, Persia and Hindostan. But between the two extremes, the distance is too great to admit of an immediate communication such as can be safely relied on. mere English or a mere German scholar, knows nothing genealogically of his own language. He cannot with certainty distinguish a compound word from a simple one, a derivative from a primitive, an exotic from one of native growth. Wherever, therefore, a philological comparison with the ancient languages has been attempted, either in England or in Germany, by those who knew no Teutonic language but their vernacular tongue, their speculations have been always unsuccessful, and often ludicrous. In order to obtain the proper point of view from which the Teutonic for mations can be compared with the classical, we must gain the vantageground afforded by the most ancient of the Teutonic chain, and thence direct our observations to the most primitive of the remoter languages. The Gothic, as might have been expected from its being the oldest Teutonic language known, has the signal advantage of exhibiting its elements, its vowels as well as its consonants, in a shape at once the most authentic and the most uniform: and to this standard it is always best to refer the other Teutonic languages before we compare them with those of an earlier date. Unfortunately the literary monuments of the Gothic, are too limited to furnish us with the counterparts or roots of all or even the greater part of modern vocables. But in the absence of this direct assistance, we are generally enabled to supply the defect by the laws of mutual comparison among the later languages: and if we have a given word in any two or three even of the most modern Teutonic forms, it is seldom difficult to approximate to what it must or would have been according to its Gothic type. The precaution we have in this respect recommended, of recurring to the most ancient Teutonic form in the first instance, is but ill observed by rash or inexperienced philologers: but it is never omitted with impunity. Let us now endeavour to explain some of the leading principles on which this part of the science seems to be founded. The single and short vowels of a language are alone considered to be primitive; its long vowels and diphthongs are regarded as derivative. The explanation, therefore, of any system of vowels, is intimately connected with prosody. But in modern languages great uncertainty and confusion has here been introduced; and for this reason among others, it be comes necessary to recur to the ancient forms of Teutonic speech in which the rhythmical quantity of the vowels was for the most part settled and certain. The primitive vowels of the IndoTeutonic languages are considered to be three in number, ă, Ĭ, ŭ; pronounced nearly after the Italian fashion. The other vowels, whether short or long, are held to be modifications of these. It is doubtful if this division is founded in universal nature: but it seems to be observable in the family of languages which we are here considering. The three vowels above enumerated are the only primitive vowels in Sanscrit and Gothic; and in the remaining languages there seems a law or tendency, according to which all the other vowel sounds seek to range themselves under one or other of three sections corresponding to the three simple vowels we have mentioned. All the roots of words, as appearing in their purest monosyllabic form, are necessarily characterised by one or other of these three vowels; and in the perfection of this theory, the vowel which characterises a root in any one language, ought to remain its characteristic in all the cognate languages into which it may pass, subject always to such known and definite modifications as the vowels are found to admit without changing their essential nature. The modifications of the vowels to which we have here alluded, consist of certain regular processes to be afterwards explained, by which they are severally augmented, weakened, or diphthongated. A root, which in Sanscrit is characterised by a, ought to retain the same vowel, in a pure or modified form, in Greek, Gothic, German, and English, and so also of i and u. To a great extent, accordingly, a uniformity in preserving the radical vowel through the different languages is found to prevail; though there are anomalies, real or apparent, which break in upon the rule. But these are much less numerous than they appear: they generally occur in circumstances where the word or language exhibiting them has otherwise been corrupted; and in many cases they vanish altogether before a more searching examination into the facts. It is thought that even the slight and superficial explanations which we are about to offer, will give clearness and consistency to much that may formerly have appeared to our readers capricious and obscure. We shall now notice the three primitive vowels A, I, U, in their order. I. Of these, the vowel A is the most important and remarkable, both from the frequency of its occurrence, and the multiplicity of its modifications. The roots characterised by this vowel are, in all the Indo-Teutonic languages, the most numerous of any. It seems the most easy to our lips, and the most agreeable to our ear. Its diversity also in point of aspect or utterance, is peculiar; and we are tempted to say of it, analogously to what has been said of the colour of green, that it imitates in its various changes all the shades of sound, from the lightest and liveliest to the most deep and sombre. A most important law for comparative philology has here been announced by Bopp. "The Sanscrit," he ob- We may here observe, that in the Latin language also the short e and o are used indiscriminately as belonging to the A class, and as corresponding to the Sanscrit a. Our examples shall include Latin as well as Greek words. The rule, we repeat, is, that the Sanscrit a corresponds not only to the Greek and Latin a, but also to the short e and o, s and o, of these languages. We have various instances of this in the numerals of these different languages. Thus, the Sanscrit panchan, five, corresponds to the Greek Tr; shash, six, to the Greek , and Latin sex; saptan, seven, to ire and septem ; ashtan, eight, to oxTMw, octo; nawan, nine, to "-"F", and novem ; daçan, ten, to da and decem. Many miscellaneous words may likewise be noticed. Danta-s, the Sanscrit for a tooth, assumes in Greek an internal, in o-des, o-dovros, besides prefixing a vowel. In Latin it takes the e, in dens, dentis. Pada-s, S. for a foot, undergoing the same changes, becomes Tus, Todos; pes, pedis. Naman, S. for aname, seems to correspond to o-voμa, as much as to nomen, though we sus οστέον, Οδ. ofis, pect that this vocable, which is found with little difference in all the IndoTeutonic languages, is referable to a double root; the Greek vu, or Gothic niman on the one hand, and the Latin nosco, for gnosco, on the other. The resemblance between the different words for name, would thus be accidental and not real. Mathu, S., a sweet drink, is the Greek μsv. Dakshina-s, S., on the right side, corresponds to dis, dexter. Janu-s, S., the knee, becomes you, genu. Jănu-s, S., birth, becomes you, yevsa, genus. Awi-s, S., a sheep, is ovis. Nawa-s, S., new, is vi Fos, novus. Asthi, S., a bone, is Asti, S., the third person singular of the substantive verb, becomes T, est. Ad, S., the root of the verb to eat, becomes sow, edo. Wah, S., to carry, becomes Foxtw, veho. Pari, S., around, is the Greek περι. Many more such instances might be given if they were necessary. We might also give several instances where the Sanscrit a remains unchanged in Greek or Latin. We shall content ourselves with one or two of the more remarkable. Thus laghu-s, S., light, slight, becomes in Greek -λaxus, without a change of the characteristic vowel; while in Latin it is le-vis, with a change of the a, into e. In chatur, S., four, the converse takes place; the vowel remaining unchanged in the Latin quatuor, while it passes into in the Greek TTTags, with the additional change of the initial consonant from a guttural to a dental. Açru, S., a tear, corresponds apparently to the Greek Sangu, and Latin lacruma, where no change takes place in the characteristic vowel, but a consonant is prefixed; being, however, a different consonant in the two kindred tongues. E The mutual connexion thus traced between the Greek vowels,,,, and, in like manner, between the corresponding Latin vowels, is an important element in classical etymology itself. In particular, it helps to explain, and enables us with more confidence to infer, the interchange that so frequently takes place between and, two sounds which do not, at first sight, appear to be closely allied, but which we thus know to be united together by their common relation to the simple a. We now see, in a clearer light, the easy transition which connects such cognates as the following :—λeyw, 2070s; yEvos, youos; you, genu; Eg, Pogia; doxiw, deceo, do, decus; (compare, in signification, seem and seemly,) veos, novus; modos, pedis; vd, spondeo; daμaw, domo, gaxus, brevis; saa, oliva, &c. We may here remind our readers that an extensive class of Greek verbs are conjugated upon the principle that a, &, are cognate and interchangeable, these different vowels being used to characterise the different parts of the same verb. We select a few examples from Pott* and Donaldson; † and in some cases, where the conjugation is incomplete, supply the defect by referring to a derivative word. Thus, στέλλω, εσταλην, (στολος); βαλλω, (βελος), βεβολημαι, τέμνω, εταμον, (τομη) 5 τρέπω, ετραπον, τέτροφα, στρέφω, εστρα φήν, εστροφα; πέρδων, επαρδον, πεπορδα, &c. Having thus shown the affinity, and, in a certain sense, the identity of the short vowels, ă, ě, ŏ, in the classical languages, we shall now give examples in which Greek and Latin words with these vowels, are represented by Teutonic counterparts having the vowel a. From what has been said, it must be apparent that such Teutonic words are as properly allied to those which, in Greek and Latin, have the vowels ĕ and ŏ, as to those which have, as their characteristic, the more obvious, but, in those languages, the less frequent form of ǎ. It is remarkable that the English dialects often imitate the classical languages in using o for a, the o being merely a broader expression of the a. Our readers will here, at the same time, remember the consonantal rules formerly explained, according to which the Greek and Latin y and ♪ correspond to the Gothic K and T; the Greek and Latin T,, to the Gothic н, TH, F, and the Greek and Latin x or h, 2, 4, to the Gothic G, D, B. Keeping both sets of principles in view, then, the Greek aro, Latin ab, are literally represented by the Gothic AF, English off, of. Ad, L., is at E. The Greek ava is the Gothic ANA, English on. The classical oxTM, octo, * Pott's Etymologische Forschungen. 2 vols. Lemgo, 1833-6. † Donaldson's New Cratylus. Cambridge, 1839. |