Yet the

parts of space are

sive.

236. In order to a proper appreciation of this passage it is essential to bear in mind that Hume, so far as the usages of language would allow him, ignores all such differences in modes of consciousness as the Germans indicate by the distinction between Empfindung' and 'Vorstellung,' and by that between 'Anschauung' and 'Begriff;' or, more properly, that he expressly merges them in a mode of consciousness for which, according to the most consistent account that can co-existent be gathered from him, the most natural term would be not succes feeling." It is true that Hume himself, admitting tinction in the degree of vivacity with which this consciousness is at different times presented, inclines to restrict the term 'feeling' to its more vivacious stage, and to use 'perception' as the more general term, applicable whatever the degree of vivacity may be. We must not allow him, however, in using this term to gain the advantage of a meaning which popular theory does, but his does not, attach to it. 'Perception' with him covers 'idea' as well as 'impression; ' but nothing can be said of idea that cannot be said of impression, save that it is less lively, nor of impression that cannot be said of idea, save that it is more so. It is this explicit reduction of all consciousness virtually, if not in name, to feeling that brings to the surface the difficulties latent in Locke's idealism.' These we have already traced at large; but they may be summed up in the question, How can feelings, as 'particular in time' or (which is the same) in perpetual flux,' constitute or represent a world of permanent relations ?3 The difficulty becomes more obvious, though not more real, when the relations in question are not merely themselves permanent, like those between natural phenomena, but are 'relations between permanent parts,' like those of space. It is for this reason that its doctrine about geometry has always been found the most easily assailable point of the 'sensational' philosophy. Locke distinguishes the ideas of space and of duration as got, the one from the permanent parts of space,' the other from the fleeting and perpetually perishing parts of succession." He afterwards prefers the term 'expan

sion' to space, as the opposite of duration, because it brings out more clearly the distinction of a relation between permanent parts from that between 'fleeting successive parts which never exist together.' How, then, can a consciousness consisting simply of fleeting successive parts' either be or represent that of which the differentia is that its parts are permanent and co-exist?

not make space a

tion' with

out being

account of

237. If this crux had been fairly faced by Hume, he must have seen that the only way in which he could consistently deal with it was by radically altering, with whatever consequence to the sciences, Locke's account of space. As it was, he did not face it, but-whether intentionally or only in effect --disguised it by availing himself of the received usages of Hume canlanguage, which roughly represent a theory the exact opposite of his own, to cover the incompatibility between the 'percepestablished view of the nature of space, and his own reduction of it to feeling. A very little examination of the passage, false to quoted at large above, will show that while in it a profession his own is made of identifying extension and a certain sort of per- perception; ception with each other, its effect is not really to reduce extension to such a perception as Hume elsewhere explains all perceptions to be, but to transfer the recognised properties of extension which with such reduction would disappear, to something which for the time he chooses to reckon a perception, but which he can only so reckon at the cost of contradicting his whole method of dealing with the ideas of God, the soul, and the world. The passage, in fact, is merely one sample of the continued shuffle by which Hume on the one hand ascribes to feeling that intelligible content which it only derives from relation to objects of thought, and on the other disposes of these objects because they are not feelings.

238. The table, which just now appears to me, is only a perception, and all its qualities are qualities of a perception. Now, the most obvious of all its qualities is extension. The perception consists of parts. These parts are so situated as to afford us the notion of distance and contiguity, of length, breadth, and thickness,' &c., &c. If, now, throughout this statement (as according to Hume's doctrine we are entitled to do) we write feeling for 'perception' and 'notion,' it will appear that this table is a feeling, which has another feeling, called extension, as one of its qualities; and that this latter feeling consists of parts. These, in turn, must be themselves

as appears feeling

if we put

for per

ception' in the pas

sages in question.

To mako sense of

them, we

must take perception

to mean

perceived thing,

=

feelings, since the parts of which a perception consists must be themselves perceived, and, being perceived, must, according to Hume, be themselves perceptions which feelings. These feelings, again, afford us other feelings of certain relations -distance and contiguity, &c.-feelings which, as Hume's doctrine allows of no distinction between the feeling and that of which it is the feeling, must be themselves relations. Thus it would seem that a feeling may have another feeling as one of its qualities; that the feeling, which is thus a quality, has other feelings as its co-existent parts; and that the feelings which are parts afford us' other feelings which are relations. Is that sense or nonsense?

239. To this a follower of Hume, if he could be brought to admit the legitimacy of depriving his master of the benefit of synonyms, might probably reply, that the apparent nonsense only arises from our being unaccustomed to such use of the term 'feeling;' that the table is a 'bundle of feelings,' actual and possible, of which the actual one of sight suggests a lively expectation, easily confused with the presence, of the others belonging to the other senses; that any one of these may be considered a quality of the total impression formed by all; that the feeling thus considered, if it happens to be visual, may not improperly be said to consist of other feelings, as a whole consists of parts, since it is the result of impressions on different parts of the retina, and from a different point of view even itself to be the relation between the parts, just as naturally as a mutual feeling of friendship may be said either to consist of the loves of the two parties to the friendship, or to constitute the relation between them. Such language represents those modern adaptations of Hume, which retain his identification of the real with the felt but ignore his restrictions on the felt. Undoubtedly, if Hume allowed us to drop the distinction between feeling as it might be for a merely feeling consciousness, and feeling as it is for a thinking consciousness, the objection to his speaking of feeling in those terms, in which it must be spoken of if extension is to be a feeling, would disappear; but so, likewise, would the objection to speaking of thought as constitutive of reality. To appreciate his view we must take feeling not as we really know it for we cannot know it except under those conditions of self-consciousness, the logical categories, which in his attempt to get at feeling, pure and simple, Hume is consistent

enough to exclude-but as it becomes upon exclusion of all determination by objects which Hume reckons fictitious. What it would thus become positively we of course cannot say, for of the unknowable nothing can be said; but we can decide negatively what it cannot be. Can that in any case be said of it, which must be said of it if a feeling may be extended, and if extension is a feeling? Can it be such a quality of an object, so consisting of parts, and such a relation, as we have found that Hume takes it to be in his account of the perception of this table?

can only

4

mean as

fictions.'

240. After having taken leave throughout the earlier which it part of the Treatise on Human Nature' to speak in the ordinary way of objects and their qualities-and otherwise the result of course he could not have spoken at all--in the fourth of certain book he seems for the first time to become aware that his doctrine did not authorise such language. To perceive qualities of an object is to be conscious of relation between a subject and object, of which neither perishes with the moment of perception. Such consciousness is self-consciousness, and cannot be reduced to any natural observ able event, since it is consciousness of that of which we cannot say 'Lo, here,' or 'Lo, there,' 'it is now but was not then,' or 'it was then but is not now.' It is therefore something which the spirit of the Lockeian philosophy cannot assimilate, and which Hume, as the most consistent exponent of that spirit, most consistently tried to get rid of. The subject as self, the object as body, he professes to reduce to figures of speech, to be accounted for as the result of certain 'propensities to feign:' nor will he allow that any impression or idea (and impressions and ideas with him, be it remembered, exhaust our consciousness) carries with it a reference to an object other than itself, any more than do pleasure or pain to which 'in their nature' all perceptions correspond.' He cannot, indeed, avoid speaking of the consciousness thus reduced to the level of simple pain and pleasure, as being that which in fact it can only be when determined by relation to a self-conscious subject, i.e. as

Every impression, external and internal, passions, affections, sensations, pains, and pleasures, are originally on the same footing; and, whatever other differences we may observe among them, appear, all of them, in their true colours, as impressions or perceptions.' P. 480.

All sensations are felt by the mind such as they really are; and, when we doubt whether they present themselves as distinct objects or as mere impres sions, the difficulty is not concerning their nature, but concerning their relations and situation.' P. 480.

If felt

have

itself an object; but he is so far faithful in his attempt to avoid such determination, that he does not reckon the object more permanent than the impression. It, too, is a 'perishing existence.' As the impression disappears with a 'turn of the eye in its socket,' so does the object, which really is the impression, and cannot appear other than it is any more than a feeling can be felt to be what it is not.1

241. Such being the only possible object, how can thing is no qualities of it be perceived? We cannot here find refuge more than feeling, in any such propensity to feign as that which, according to how can it Hume, leads us to endow objects with a continued existqualities? ence, distinct from our perceptions.' If such propensities can give rise to impressions at all, it can only be to impressions of reflection, and it cannot be in virtue of them that extension, an impression of sensation, is given as a quality of an object. Now if there is any meaning in the phrase 'qualities of an object,' it implies that the qualities co-exist with each other and the object. Feelings, then, which are felt as qualities of another feeling must co-exist with, i.e. (according to Hume) be felt at the same time as, it and each other. Thus, if an impression of sight be the supposed object, no feeling that occurs after this impression has disappeared can be a quality of it. Accordingly, when Hume speaks of extension being seen as one of the qualities of this table, he is only entitled to mean that it is one among several feelings, experienced at one and the same time, which together constitute the table. Whatever is not so experienced, whether extension or anything else, can be no quality of that 'perception.' How much of the perception, then, will survive? Can any feelings, strictly speaking, be cotemporaneous ? Those received through different senses, as Hume is careful to show, may be; e.g. the smell, taste, and colour of a fruit. In regard to them, therefore, we may waive the difficulty, How can feelings successive to each other be yet co-existent qualities ? but only to find ourselves in another as to what the object may be of which the cotemporaneous feelings are qualities. It cannot, according to Hume, be

See above, paragraph 208, with the passages there cited.

The taste and smell of any fruit are inseparable from its other qualities of colour and tangibility, and

'tis certain they are always co-existent.

Nor are they only co-existent in general, but also cotemporary in their appearance in the mind.' P. 521. (Contrast p. 370, where existence and appearance are identified.)