weeping, and a few demonstrative ejaculations denoting spatial proximity or remoteness. It is from such lowly beginnings that language has arisen; and when we seek to account for the genesis of human speech, we must forget for a while the numberless delicate contrivances by which civilized man gives unerring expression to every shade of thought. We must forget our nouns and verbs, our inflections, prepositions, and auxiliaries, and search for the meaning latent in simple, ungarnished roots. We must turn our eyes from the architecture of the sacred edifice, that we may scrutinize with unbiassed vision the rugged blocks out of which it has been reared.

But before we can properly arrive at a final residuum, one further element must be eliminated from the problem. The fact that all abstract words whatever, as well as all concrete words denoting supersensuous objects or conceptions, have been formed by the potent agency of metaphor out of words with a purely physical import, is too thoroughly established to need more than a passing illustration. That soul and sea are derived from the same root, that ghost, yeast, gust, gas, and geyser have a common origin, that words like intellect, conceive, admire, permit, contrition, attention, etc., are built up out of wholly material notions, has been proved and illustrated over and over again. From the bare notion of "moving," we get not only words signifying "furniture" (Fr. meubles), the Skr. sarit, "a river," sara, "sap," drapsa, "a drop," but also age and eternity, and the most abstract of the many names for the Deity.

The explanations of the word God in our popular dictionaries are one and all untenable. Grimm has shown the common derivation from good to be impossible; and the identification with Persian Khoda, from Zend quadata, Skr. svadata, Lat. a se datus, has met, at Aufrecht's hands, with no better fate. Müller more plausibly suggests that it was formerly a heathen name for the Deity, which passed into Christian usage, as Lat. Deus became Fr. Dieu.* There can, I think, be little doubt that God is iden

Bunsen, Outlines, I. 78; Grimm, Deutsche Mythologie, p. 12; Müller, Lectures, II. 302. — Donaldson (New Cratylus, p. 710) would connect God with kaλós, and Oeós with rioŋue; but this is arrant guesswork, and the latter name was long ago proved to be identical with Lat. deus, Skr. devas, from div, "to shine," being a personification of daylight. See Lassen, Indische Alterthumskunde, I. 756.

tical with Wodan, the name of the great Northern deity. Guillaume and William, guerre and war, guardian and warden, guile and wile, will at once occur to the reader as analogous instances; and a similar change is seen in the Armorican guasta and Ital. guastare, as compared with Lat. vastare, Eng. waste; and in the Eng. quick, Goth. quivs, Lat. vivus. Paulus Diaconus tells us that the Lombards pronounced Wodan as Guodan.* The initial digamma, consisting of a combined guttural and labial (F, kv, gu, qu, hw, sw, etc.), is resolved into its constituent elements, one of which is usually eliminated, so that Guod is differentiated into God and Wod. The surviving element is frequently vocalized, becoming a mere breathing, as in Odin, with which we may compare the change from Georgian ghwino to Lat. vinum and Gr. oivos. Now F-Odin is derived by Grimm from wuotan, Lat. vadere, Eng. wade, "transmeare, cum impetu ferri," and thus signifies the Prime Mover, “qui omnia permeat," the pervading source of the dynamic phenomena of the universe; or, as Dante says: "La gloria di Colui che tutto muove." Strictly in keeping with this signification is Odin's character as the storm-spirit or inciting genius of the wind, in which he answers to the Greek Hermes, or Vedic Sarameyas, "Son of the Dawn." This is at bottom a purely sensuous group of notions. Yet, as developed by modern philosophy, the conception of Odin is far loftier and more adequate than those expressed by Ocós, Jupiter, Dyaus; ‡ as the solar ray (Tiws= Zeus), though doubtless the immediate source of all life and motion in our little group of worlds, is nevertheless itself but a limited manifestation of that limitless Power which is revealed in

* And we find Ludoguicus for Ludovicus in Erchempert, Hist. Langob. 11, Pertz, III. 245. In Germany and Lorraine we have the town names of Godesberg, Gudenberg, Godensholt, and Vaudemont, all derived from Wodan. In the Westphalian dialect, Wednesday ("day of Wodan ") is called Godenstag or Gunstag; in NetherRhenish, Gudenstag; in Flemish, Goensdag. See Thorpe, Northern Mythol. I. 229; Taylor, Words and Places, 323; and cf. Grimm, Gesch. der Deutschen Sprache, 296. The Westphalian Saxons wrote both Guodan and Gudan. In view of such a convergence of proofs, I am surprised that this etymology has not been sooner insisted

on.

For the explanation of Hermes and his correlatives, see Müller, Lect. II. 492; Cox, Manual, 85, fol'g; Baring-Gould, Curious Myths, II. 152, fol'g; Mackay, Religious Development of the Greeks and Hebrews, I. 260-273.

For the physical meaning latent in 717′ see Oort, Worship of Baalim, Chap. II.; and cf. Mackay, op. cit. II. 414–426.

ceaseless rhythmical activity throughout the length and breadth of the Kosmos.

The word truth affords another illustration, which may even now be worth citing, since writers are still to be found wasting eloquence over Horne Tooke's derivation from troweth, and his sophistical inference τὸ ἀληθές οὐ φύσει ἀλλὰ νόμῳ. Its correlative in Sanskrit is dhruwa, from dhru, "to place or establish "; whence also Goth. triggvus, "secure," and Germ. trauen and treu.* While, according to Pott, the Lat. verus, Germ. wahr, Welsh gwir, Gael. fior, Sclavon. viera, is from the same root which appears in Germ. wehren, O. H. G. warjan, Welsh gwared, Ital. guardare, and means that which is "covered," ergo sheltered, or secure from attack.

It being thus apparent that metaphor may educe the grandest abstract conceptions from the humblest material notions, we have now only to inquire how the roots expressive of such notions were produced. Of the various recondite, mystical, or incognoscible theories which have been framed, chiefly in Germany, concerning the origin of language, we have not time to speak. Max Müller's doctrine, that roots are "phonetic types," must be pronounced vague and inadequate. The term "phonetic type" can mean nothing but a representation, in vocal sounds, of an objective phenomenon which invites attention, or of a subjective feeling which demands expression. To say that a root is a "phonetic type," is therefore merely to state the problem without solving it. To explain the genesis of language, it is not enough to say that, as every substance when smitten rings response to the blow, so the human mind attunes itself to concord with the sensory percussion from without. It must be shown why this is possible, and how it comes to pass. It must be shown in what way mere vocal utterances can become the fitting signs of external and internal phenomena; by what subtle alchemy a rhythmic pulsation of the air is transmuted into the nimble messenger of thought and feeling. In the present state of our knowledge, amid the confusion of ideas in which this problem is entangled, its solution may be impracticable, and must needs be imperfect. Yet a theory which would fain satisfy

Garnett, p. 28. In the Semitic languages also, "l'idée du vrai se tire de la solidité, de la stabilité." Renan, Langues Sémitiques, I. 23.

us by a mere phrase, like "phonetic type," is, as Professor Whitney (p. 427) observes, utterly unphilosophical and infertile.

To explain the process whereby articulate sounds have been erected into symbols of mental and physical phenomena, a legitimate hypothesis has been framed, according to which a large class of words were originally mere representations of the sounds given out by animate or inanimate objects. Sounds uttered in imitative response to external sounds or noises passed into currency as vocal signs of the phenomena whereof the archetypal sounds were chiefly characteristic; and the concrete material designations thus produced were afterwards metaphorically applied to the expression of abstract and immaterial conceptions; while all words not thus originated were mere interjectional utterances, serving to give vent to the feelings aroused by the sight of external objects.

The Bow-wow Theory-though it has been received by many eminent philologists with indignant ridicule and foolish appeals to sentimental prejudice—is nevertheless the only hypothesis yet proposed which alleges a vera causa. Every one knows that many words have come from onomatopoeic roots, though few are aware to what an extent the process may be traced. Not only is it that words like cackle, caw, croon, snarl, grunt, chirp, twitter, sigh, groan, shriek, sneeze, crash, bang, whiz, puff, snap, crunch, buzz, hiss, hum, twang, jingle, chink, rustle, whisper, clatter, gurgle, bubble, rumble, etc., are manifest imitations, but also many words indicative of no sound or noise whatever, many words expressive of purely abstract notions, may safely be referred to an imitative source. They who have carefully noted the wanton freaks which metaphor delights in, and who know, moreover, that all the indigenous words in all the Aryan languages have arisen from a few hundred primitive roots, will not be surprised to find a single word begetting legions of offspring whose resemblance to each other and to their sire is far from obvious. Every one has read of the Arabian genie who became in succession a wolf, a cock, an eagle, a pomegranate, and lastly a raging fire which burnt up the daughter of a sultan. The metamorphoses of language are not less marvellous. We have already seen glory becoming the badge of hopeless servi tude. We may next contemplate a venerable root, mar, — ex

pressive of "grinding down," and doubtless imitated from the sound of that mechanical process,-producing in divers languages at least one hundred fifty distinct words (to say nothing of minor modifications), among which, in English alone, are comprised terms so different as mill, meal, maul, moil, mould, mortal, murder, member, meer, marble, Mark, milk, blaspheme, blame, smart, mellow, melt, malt, mild, ambrosia, travail, remorse, mallet, memory, and martyr.*

In like manner numerous words, retaining naught that is exclamatory in their forms or meanings, have nevertheless been developed from interjections. From the root ach are derived words expressive both of emotional discomfort and of material sharpness as well as swiftness; among which, in various languages, are Gr. ä«w, äкavla, äxos, Eng. ache, A. S. ege, egeslich, Icel. ecki, Germ. ekel, ecke, jucken, Eng. itch, Icel. eggia, Lat. acuo, acus, acies, Gr. ἀκωκή, ἄκρος, ὠκύς = Skr. acus, which brings us to açva = iππоs = equus, wrongly claimed as due to direct imitation. † Onomatopoeia is therefore far from being a sterile principle; nor is it easy to indorse the objection that imitative sounds, even in their crudest shape, are not yet language. Müller tells us that "there are cockatoos who, when they see cocks and hens, will begin to cackle, as if to inform us of what they see." The inference, that an imitation of cackling cannot be language, proves somewhat too much; for we cannot consent to banish a considerable number of decorous words and pithy sentences from the category of language, merely because certain parrots have a habit of repeating them. So, when the Englishman at dinner in China, craving information as to the nature of the dish before him, said interrogatively, Quack, quack and received the astounding answer, Bow-wow! Müller doubts whether such talk as this deserves the name of language. But the doubt rests upon a confusion of ideas. When the duck says quack, no language is used, because the sound in question corresponds to no general concept. But when man in framing language says quack, he is performing a

* See the details in Müller, II. 331-350; to which I have added travail, derived on Garnett's authority, through the French, from Kymric trafael = tra + mael, “overwork."

† Compare Curtius, Griechische Etymologie, p. 122, with Garnett, p. 64, and Müller, II. 75.