Two things, then, B and C, indistinguishable when compared together alone, may each contract adhesions with different associates, and the compounds thus formed may, as wholes, be judged very distinct. The effect of practice in increasing discrimination must then, in part be due to the reinforcing effect, upon an original slight difference between the terms, of additional differences between the diverse associates which they severally affect. Let B and C be the terms: If A contract adhesions with B, and C with D, AB may appear very distinct from CD, though B and C per se might have been almost identical.
To illustrate, how does one learn to distinguish claret from burgundy? Probably they have been drunk on different occasions. When we first drank claret we heard it called by that name, we were eating such and such a dinner, etc. Next time we drink it, a dim reminder of all those things chimes through us as we get the taste of the wine. When we try burgundy our first impression is that it is a kind of claret; but something falls short of full identification, and presently we hear it called burgundy. During the next few experiences, the discrimination may still be uncertain—"which," we ask ourselves, "of the two wines is this present specimen?" But at last the claret-flavor recalls pretty distinctly its own name, 'claret,' "that wine I drank at So-and-so's table," etc.; and the burgundy-flavor recalls the name burgundy and some one else's table. And only when this different setting has come to each is our discrimination between the two flavors solid and stable. After a while the tables and other parts of the setting, besides the name, grow so multifarious as not to come up distinctly into consciousness; but pari passu with this, the adhesion of each wine with its own name becomes more and more inveterate, and at last each flavor suggests instantly and certainly its own name and nothing else. The names differ far more than the flavors, and help to stretch these latter farther apart. Some such process as this must go on in all our experience. Beef and mutton, strawberries and raspberries, odor of rose and odor of violet, contract different adhesions which reinforce the differences already felt in the terms.
The reader may say that this has nothing to do with making us feel the difference between the two terms. It is merely fixing, identifying, and so to speak substantializing, the terms. But what we feel as their difference, we should feel, even though we were unable to name or otherwise identify the terms.
To which I reply that I believe that the difference is always concreted and made to seem more substantial by recognizing the terms. I went out for instance the other day and found that the snow just fallen had a very odd look, different from the common appearance of snow. I presently called it a 'micaceous' look; and it seemed to me as if, the moment I did so, the difference grew more distinct and fixed than it was before. The other connotations of the word 'micaceous' dragged the snow farther away from ordinary snow and seemed even to aggravate the peculiar look in question. I think some such effect as this on our way of feeling a difference will be very generally admitted to follow from naming the terms between which it obtains; although I admit myself that it is difficult to show coercively that naming or otherwise identifying any given pair of hardly distinguishable terms is essential to their being felt as different at first.[424]
I offer the explanation only as a partial one: it certainly is not complete. Take the way in which practice refines our local discrimination on the skin, for example. Two compass-points touching the palm of the hand must be kept, say, half an inch asunder in order not to be mistaken for one point. But at the end of an hour or so of practice with them we can distinguish them as two, even when less than a quarter of an inch apart. If the same two regions of the skin were constantly touched, in this experience, the explanation we have been considering would perfectly apply. Suppose a line a b c d e f of points upon the skin. Suppose the local difference of feeling between a and f to be so strong as to be instantly recognized when the points are simultaneously touched, but suppose that between c and d to be at first too small for this purpose. If we began by putting the compasses on a and f and gradually contracted their opening, the strong doubleness recognized at first would still be suggested, as the compass-points approached the positions c and d; for the point e would be so near f, and so like it, as not to be aroused without f also coming to mind. Similarly d would recall e and, more remotely, f. In such wise c—d would no longer be bare c—d, but something more like abc—def,—palpably differing impressions. But in actual experience the education can take place in a much less methodical way, and we learn at last to discriminate c and d without any constant adhesion being contracted between one of these spots and ab, and the other and ef. Volkmann's experiments show this. He and Fechner, prompted by Czermak's observation that the skin of the blind was twice as discriminative as that of seeing folks, sought by experiment to show the effects of practice upon themselves. They discovered that even within the limits of a single sitting the distances at which points were felt double might fall at the end to considerably less than half of their magnitude at the beginning; and that some, though not all, of this improved sensibility was retained next day. But they also found that exercising one part of the skin in this way improved the discrimination not only of the corresponding part of the opposite side of the body, but of the neighboring parts as well. Thus, at the beginning of an experimental sitting, the compass-points had to be a Paris line asunder, in order to be distinguished by the little-finger-tip. But after exercising the other fingers, it was found that the little-finger-tip could discriminate points only half a line apart.[425] The same relation existed betwixt divers points of the arm and hand.[426]
Here it is clear that the cause which I first suggested fails to apply, and that we must invoke another.
What are the exact experimental phenomena? The spots, as such, are not distinctly located, and the difference, as such, between their feelings, is not distinctly felt, until the interval is greater than the minimum required for the mere perception of their doubleness. What we first feel is a bluntness, then a suspicion of doubleness, which presently becomes a distinct doubleness, and at last two different-feeling and differently placed spots with a definite tract of space between them. Some of the places we try give us this latest stage of the perception immediately; some only give us the earliest; and between them are intermediary places. But as soon as the image of the doubleness as it is felt in the more discriminative places gets lodged in our memory, it helps us to find its like in places where otherwise we might have missed it, much as the recent hearing of an 'overtone' helps us to detect the latter in a compound sound (supra, [pp. 439-40]). A dim doubleness grows clearer by being assimilated to the image of a distincter doubleness felt a moment before. It is interpreted by means of the latter. And so is any difference, like any other sort of impression, more easily perceived when we carry in our mind to meet it a distinct image of what sort of a thing we are to look for, of what its nature is likely to be.[427]
These two processes, the reinforcement of the terms by disparate associates, and the filling of the memory with past differences, of similar direction with the present one, but of more conspicuous amount, are the only explanations I can offer of the effects of education in this line. What is accomplished by both processes is essentially the same thing: they make small differences affect us as if they were large ones—that large differences should affect us as they do remains an inexplicable fact. In principle these two processes ought to be sufficient to account for all possible cases. Whether in fact they are sufficient, whether there be no residual factor which we have failed to detect and analyze out, I will not presume to decide.