Page:Mind (Old Series) Volume 11.djvu/409

 408 B. BOSANQUET : COMPARISON, ETC. purpose reinforces and interprets the ruder shapes ; or, using the same phraseology as before, we may say that the ruder shapes are subsumed under those whose import is more unmistakeable. That is to say, we notice the latter first, and then, when we look at the former, the latter are supplied by " redintegration," and so enable us to pick out the characteristic outline in the ruder form. That this process is a copious source of fallacy as well as of dis- covery (for a rude outline is often of ambiguous interpretation) only shows that it must be carried into detail with great care. It is obviously the unconscious rudiment of the Method of Concomitant Variations, which is only an application of the principle that Identity cannot exhibit itself except in Difference. The problems of degrees of Difference and of Identity seem to me therefore to belong to logic and not to psychology. We are, as I believe, on the wrong track, if we try to refer these degrees to difficulty or delay experienced in the psychical process of making identifications or transitions. We can only speak of degree in reference to a standard, and this standard must be, I think, of the nature either of Quantity or of Kind. Under the head of Kind we may rank Purpose. If there is indiscernible identity on the one hand, or mere qualitative distinction on the other, then we have no question of degree. Two shades of green are more or less of the same ; but green and carmine would be, I presume (if pure, which no actual colour-sensation ever is), simply distinct, simply, that is, not the same, and no degree of difference can be assigned them as colour. Of course as light- stimuli they have a measurable relation. It would follow that to ask as regards any two given terms whether there is more identity or difference between them, is a question entirely relative to the standard which we select. Xo identity, and no difference, has an absolute value. The nearest approach to such an absolute value is relation to a quantitative scale. If the two terms can be referred to places on a scale, which is exhaustive and of which the intervals are truly equal, then we ^so facto judge whether the units which the two terms have in common (their identity) or those by which the one exceeds the other (their difference) make the greater sum. But this is all. In using such a scale we use it for a purpose or for an effect (e.</., in music and painting), and then the purpose or effect at once becomes the standard of identity and difference. We may say if we like that an aggregate of 100 sacks of corn compared with 49 sacks has more difference than identity. But if we only want 40, and are not bound or allowed to dispose of what we do not want, the two aggregates are identical for our purpose. Mill, in distinguishing analogy from induction (Lo<jic, ii. 85), comes very near attempting to balance likenesses as such against differences as such in a way which is purely chimerical. We can make nothing out of asking how much likeness or difference there is between two terms. The only fruitful question is tchat the likenesses, &c., are, and what they prove.