12. In languages, in which the coales cence between the verb and its adjuncts has taken place, and also the coalescence between nouns and its connective words, (Clubman, § 19), much greater liberty of inversion is practicable than in those in which such coalescence has not at all occurred, or but incompletely. In other words, where the noun, adnoun, and verb, admit of flexion, there the arrange ment depends in many instances more upon the sound than upon the sense ; and nearly in all cases may be made sub servient to the former. This gives such languages considerable advantage over those which admit of but few changes, so far as respects their modulation ; and further, the coalescence renders them much more forcible, where emphasis on any of the fractional parts is not required. Whenever flexion increases perspicuity, the advantage is decisive and obvious : with respect to modulation, though an object of some consequence, (since we may sometimes find the way to the head and heart by pleasing the ear) yet all cul tivated languages will be found to pos sess sufficient power of pleasing the na tive ear ; and among those who made sound so much an object, sense was often sacrificed to it : with respect to force, it may fairly be doubted whether the ad vantage of greater precision, by means of more accurate emphasis, does not coun terbalance it. 'kW are willing to admit on the whole, that the advantage is some what in flivour of those languages in which flexion is extensively adopted ; but we can by no means admit the opi nion of those who think it necessary to a perfect language. That language is not the most perfect, which enables us to ex press one thought in a great variety of ways, but that which enables us to ex press any thought with precision and perspicuity : and contemptible as our own nninflected language may appear to those who can think nothing good hut what accords with the objects of their early taste, we are disposed to believe that in its real powers it rises beyond all the ancient languages, and beyond most of the modern.
13. Before we leave the subject of oral language, we shall pay some attention to the three following inquiries ; whether words were originally imitative ; whether they were long ; and of what kind of ar ticulations they were composed. The latter of these are of importance in trae lug ihe gradation from hieroghypl deal to alphabetical writing. Words, in their present state, are simply arbitrary marks. The sound of some appears to be " an echo of the sense ;" but in the greater number of instances in which there is supposed to be this resemblance, very much may be attributed to the fancy of the observer. It is obvious, however, that some words are truly imitative, such e. g. as denote the various sounds of ani mals. When we carry our inquiries far ther back, we are led to suppose that the original words would be formed upon some resemblance, real or supposed, be tween their sound and the thing signi lied. What else, at first, could induce men to fix upon one sound rather than another ? Sensible objects were the first which obtained names ; and of these the number is considerable, which either emit some imitable sound, or perform such motions as are generally accompanied with sound. These would probably be denoted by words imitative of the sound, in the same manner as the Otaheitans gave to the gun the appellation of tick tick-boo, evidently imitating the cocking and report of the gun, and as we give the cuckwas its name from its note. With respect to qualities totally unconnected with sound, particularly mental qualities, this principle of imitation is not directly applicable : we immediately see the in congruity of sound and colour, for in stance, when we call to mind the idea of the blind man, that a scarlet colour was very much like the sound of a trumpet. Yet there can scarcely be a doubt that funded resemblances would, as much as real ones, direct the application of names. Some ingenious writers on this subject have observed certain letters ap plied to denote a certain class of ideas, which have some common features of re semblance, and have inferred that those letters were significant of that common feature ; e. g. that c denotes hollowness.
This particular coincidence arises proba bly from the circumstance, that the ori ginal word denoting hollowness, which has entered variously modified into the words in question, was c with some vocal sound: This appears to be the extent of the inference which may be justly drawn ; that it was so applied, but not that the sound was significant of the idea. We are accustomed to use sounds in particular connections with such regu. larity and constancy, that they appear to have a signification of themselves consi dered; but this inference arises from in attention to the matter of fact. Frequent ly, from our acquaintance with t he sense, we read a combination of words as the sense dictates, and suppose the imitation in the words, which really exists only in our mode of enunciation ; but these in stances, however just, afford no ground for argument in the present discussion, which refers only to single words: and with respect to them, we cannot but con fine the resemblance of their sound to their sense, to cases in which they de note sound or motion usually accompa 'Died with sound.
14. The chief importance of the inquiry, whether the original words of language were long, is principally confined to that language, in which the transition took place from hieroglyphics to letters. This is usually supposed to have been the Egyptian ; but as of this language only a few words are preserved in the Coptic, (of which however a large proportion are mo nosyllables) we may make the inquiry more general. Lord Monboddo supposes, that the first articulate sounds were imita tions of the cries of animals, and that con sequently they were of great length, " for such cries of almost all animals have a certain tract or extension : and that we may not think man an exception to this rule, we need only attend to the dumb persons among us, who utter inarticulate cries, sometimes very loud, but always of considerable length." Leaving the latter argument, which surely is nothing to the purpose, we may observe, that lithe cries of animals were imitated to denote those animals, great length of words was unne cessary and improbable : unnecessary, because one or two distinct articulations would usually answer every purpose ; im probable, because articulation is difficult. If we extend the principle of imitation far ther, and suppose the cries of animals imitated by man, in order to express feel ing merely, his cries would surely be un deserving the name of words,-and at any rate would throw no light on our inquiries. The theory of long words appears to de rive confirmation from the vocabularies of the North American Indians. For in stance, of three which are given by Mac kenzie, two appear to be composed of words of from two to seven syllables, with scarcely any words of one syllable. The third, however, is composed principally of words of °neer two syllables. With re spect to the former, even where the words actually denote sensible objects, our in. ference, that they are uncompounded, should be cautiously drawn. The moon is expressed by two words, tibiscapesim, and several others appear to be circumlocutions. The catholic savages on the river St. Lawrence call the priest, the master of life's man ; and it is very pro bable that, in uncultivated nations, names of new objects would, where possible, be formed rather by significant combinations of words in use, than by the formation of new words. Thus we learn from Mr. Parke, that the Mandingo nation use the tbllowing (among many) circumlocutions : fruit is eree-ding, child of the tree; finger, boullakon ding, cldhl of the hand or arm ; noon teeleekoniata, the oars ovcr head ; bro ther, ba ding kea, mother's male child; proud, telingabalid. straight-bodied ; angry, jusu bota, the heart conies out we think it almost unnecessary to remark, how much the last two instances countenance the po sitions before laid down, respecting the transference of names from external to internal things.