perhaps, return to the very spot from which we set out, without even so much knowledge, as to have the slightest guess, that we were again where we had been before.

To drop this allegory, however, it is very evident, that, though we should be capable of reasoning, even without language of any sort, and of reasoning sufficient to protect ourselves from obvious and familiar causes of injury, our reasonings, in such circumstances, must be very limited, and as little comparable to the reasoning of him who enjoys the advantage of all the new distinctions of a refined language, as the creeping of the diminutive insect to the soaring of the eagle. Both animals, indeed, are capable of advancing;-but the one passes from cloud to cloud, almost with the rapidity of the lightning, which is afterwards to flash from them, and the other takes half a day, to move over the few shrunk fibres of a withered leaf.

What must be the arithmetic of that people in South America, of whom Condamine tells us, whose whole numeration did not extend beyond three, and who had no resource afterwards, but to point first to their fingers and then to their hair! What the reasonings of arithmetic would be to such a people, every other species of reasoning would be to us, if our general vocabulary bore no greater proportion to the feelings that were to be expressed by it, than this very limited numeral vocabulary, to all the possible combinations of numbers!

The extent of error into which we should be likely to fall, in our classifications and reasonings in general, if our language were of this very imperfect kind, it is, of course, impossible for us, in our present circumstances, to guess; though we may derive some assistance, in our estimation of these possible absurdities, from facts of which voyagers occasionally tell us. I may take for an example a fact mentioned by Captain Cook, in describing the people of Wateeoo, a small island, on which he lighted in his voyage from New Zealand to the Friendly Islands. "The inhabitants," he says, were afraid to come near our cows and horses, nor did they form the least conception of their nature. But the sheep and goats did not surpass the limits of their ideas; for they gave us to understand, that they knew them to be birds.""It will appear rather incredible," he adds, "that human ignorance could ever make so strange a mistake; there not being the most distant sim

ilitude between a sheep or goat and any winged animal. But these people seemed to know nothing of the existence of any other land animals besides hogs, dogs, and birds. Our sheep and goats, they could see, were very different creatures from the two first; and, therefore, they inferred, that they must belong to the latter class, in which they knew that there is a considerable variety of species.""I wouid add," says Mr Stewart, who quotes this very striking fact, together with the judicious remark of Cook,-"[ would add, that the mistake of these islanders, perhaps, did not arise from their considering a sheep or goat, as bearing a more striking resemblance to a bird than to the two classes of quadrupeds with which they were acquainted, but from the want of a generic word, such as quadruped, comprehending these two species; which men in their situation would no more be led to form, than a person who had seen only one individual of each species would think of an appellative to express both, instead of applying a proper name to each. In consequence of the variety of birds, it appears that they had a generic name comprehending all of them, to which it was not unnatural for them to refer any new animal they met with."*

The observation of Mr. Stewart, with respect to the influence of a generic name on this seemingly very strange arrangement of these very rude zoologists, is ingenious and just. It must be remembered, however, in opposition to his general doctrine on the subject, that the application of the generic term, even in this very strange manner, is a proof, not that we are without general notions, but that we truly have general notions, that are independent ⚫ of the mere terms which express them. It was not merely because they had a generic term that they extended this term to the unknown sheep and goats, but because the sheep and goats coincided, in some measure, with the general notion expressed by the general term. Of this the most striking evidence is contained in the very statement of Captain Cook. The cows and horses, sheep and goats, were all equally unknown to the islanders. Why, then, did they not class the cows and horses with birds as much as the goats and sheep? As far as the mere possession of a generic word could have led to this application,—if a word alone were necessa

*Stewart's Elements, Part II. c. iv. sect. 1.

ry, it was common to all the new cases alike.

When all these

were equally unknown, there must have been some previous general notion of certain circumstances of resemblance in birds, with which the goats and sheep coincided more exactly than the cows and horses. Nor is it very difficult to guess what this previous notion was, the bulk of the different animals must have led to the distinction. The winged tribes with which they were acquainted, though they might perhaps approach, in some slight degree, to the stature of the smaller quadrupeds, could have no resemblance in this respect to the horses and cows. A bird, in their mental definition of it, was certainly a living thing, of certain various sizes familiar to them, and not a dog or a hog. A sheep, or a goat, was seen by them to be a living thing, not a' dog nor a hog, and of a size that implied no remarkable opposition to that involved in their silent, mental definition of a bird. In such circumstances, it was classed by them as a bird, with as much accuracy as is to be found in many of our systematic references, even in the present improved state of science and natural history,-in that, for example, which classes and ranks under one word, the whale that swims with the man that walks; or, to use a case still more analogous, even the ant that creeps with the gnat that flies,—and, with equal accuracy, they excluded the cows and horses that did not coincide with the general notion, of which a certain resemblance of size formed an essential part. The extension of the term to the one set of quadrupeds, and the exclusion of the other set, must have had some reason; and this reason, whatever it may have been, must have been some general feeling of resemblance of some sort, -a relative suggestion, intervening between the perception of the animals, and the application of the term.

218

LECTURE XLVIII.

ANALYSIS OF THE PROCESS OF REASONING.

GENTLEMEN, my last Lecture brought to a conclusion the remarks which I had to offer on that very interesting tribe of our suggestions of relation which constitute the feelings of resemblance, -a tribe, on the existence of which, as we have seen, all classification depends, and in a great measure the whole power of language, as an instrument or medium either of distinct thought in the mind of the individual, or of reciprocal communication of thought from mind to mind.

The examination of this species of relation, led us into one of the most memorable controversies in the whole science of Intellectual Philosophy; and though I knew well that there could be no reason to fear your adoption of the absurdities of Realism, and, therefore, did not think it necessary to occupy your time with any serious confutation of that obsolete hypothesis, I knew also too well the prevailing influence of the opposite error of Nominalism, and the high authorities which sanction it, not to think it necessary to put you fully on your guard against the fallacy of this system, by shewing you how incomplete it is, and, therefore, how unfit to be adopted as a narrative of the actual Process of Generalization.

This process I described, as involving, not two stages only, as the Nominalists contend, but three. In the first place, the perception or conception of the two or more external objects, or the conception of the two or more internal feelings that are afterwards classed together; in the second place, the feeling or general relative notion of the resemblance, which these separate objects bear to each other, in certain respects, the relative suggestion, in consequence of which alone we are led to class them together; in the

third place, the expression of this felt general resemblance, by a general term, as significant of that silent mental generalization which has already classed them together. The mental generalizing may, indeed, be considered as complete, before the invention of the general term; the term being of use, only as fixing and recording, or conveying to others the knowledge of that general notion or feeling of resemblance, which preceded the first use of the general word.

At the same time, however, that I exhibited to you,-as simply and forcibly as the complex nature of the process would allow me, the doctrine of general notions, as distinct mental affections of a peculiar species, arising from that susceptibility of the mind, by which we perceive, together with various other relations, the relations that constitute the resemblances of objects,-I took occasion to point out to you some errors of thought, and consequent improprieties of arrangement and expression, on the part of the Conceptualists, which I regarded as having had the chief effect in preventing the universal and ready adoption of this doctrine of the threefold nature of the process, as consisting in perceptions, relations, and verbal signs,-a doctrine, which, but for the almost universal prevalence of the opposite system of Nominalism, would have appeared to me to stand little in need of any argument in its support; since the fact of the extension of general terms only to certain objects, to the exclusion of others, seems, of itself, sufficiently to shew, that there is a certain general notion of resemblance, a peculiar state of mind,-intervening between the primary perceptions, and the use of the general term, which forms, as it were, the measure of adjustment of the particular objects,— that are arranged in the same class, if they agree with this general notion, and excluded, if they do not agree with it. An arrangement, without some principle of resemblance to direct the order in which objects are placed, seems to me absolutely unworthy of the name of an arrangement, and certainly could be but of very little aid to the memory,-even if it could be of any advantage to remember divisions, and subdivisions, that were founded upon nothing. The classifications, which our dictionaries form, according to the mere initial sounds of words,--which Dr Reid, in reference to works of this kind, calls a sort of modern categories,--would be far more philosophic, than a classification which implied no previous notion of