The pure spectral colors observed in surrounding darkness have the strongest affective tone. These colors are, however, generally very different from those of the natural objects to which accompanying feelings might refer.
There is just as little justification for the attempts to derive tonal feelings from such ideas exclusively. It can not be doubted that familiar musical ideas may be aroused through a single tone; still, on the other hand, the constancy with which certain tonal qualities are chosen to express particular feelings, as, for example, deep tones to express grave and sad feelings, can be understood only under the condition that the corresponding affective quality belongs to the simple tonal sensation. The circle in which the argument moves is still more obvious when the affective tones of sensations of taste, smell, and the general sense are derived from the accompanying ideas. When, for example, the agreeable or disagreeable tone of a taste-sensation is increased by the recollection of the same impression as experienced before, this can be possible only under the condition that the earlier impression was itself agreeable or disagreeable.
3. The varieties of simple sense-feelings are exceedingly numerous.
The feelings corresponding to a particular sensational system also form a system, since, in general, a change in the quality or intensity of the affective tone runs parallel to every change in the quality or intensity of the sensations. [p. 78]
At the same time these changes in the affective systems are essentially different from the corresponding changes in the sensational systems, so that it is impossible to regard the affective tone as a third determinant of sensations, analogous to quality and intensity. If the intensity of a sensation is varied, the affective tone may change not only in intensity, but also in quality; and if the quality of the sensation is varied, the affective tone usually changes in quality and intensity both. For example, increase the sensation sweet in intensity and it changes gradually from agreeable to disagreeable. Or, gradually substitute for a sweet sensation one of sour or bitter, keeping the intensity constant, it will be observed that, for equal intensifies, sour and, more especially, bitter produce a much stronger feeling than sweet. In general, then, every in sensation is essentially accompanied by a twofold change in feeling. The way in which changes in the quality and intensity of affective tones are related to each other follows the principle already stated (p. 33) that every series of affective changes in one dimension ranges between opposites, not,,as is the case with the corresponding sensational changes, between greatest differences.
4. In accordance with this principle, the greatest qualitative differences in sensations correspond to the greatest opposites in affective quality, and to maxima of affective intensity which are either equal or at least approximately equal, according to the special pecularities of the qualitative opposites. The middle point between these two opposites corresponds to an absence of all intensity, so far as only the single dimension to which the opposites belong is concerned. This absence of intensity can be observed only when the corresponding sensational system is absolutely one-dimensional. In all other cases, a point which is a neutral middle for one particular series of sensational differences, belongs at the same time to another [p. 79] sensational dimension or even to a number of such dimensions, each of which it has a definite affective value. Thus, for example, spectral yellow and blue are opposite colors which have corresponding opposite affective tones. In passing gradually along the color-line from one of these to the other, green would be the neutral middle between them.
But green itself stands in affective contrast with its opposite color, purple; and, furthermore, it is, like every saturated color, one extremity of a series made up of the transitional stages of a single color-tone to white. Again, the system of simple tonal sensations forms a continuity of only one dimension, but in this case more than in others it is impossible to isolate the corresponding affective tones through abstraction, as we did the pure sensations, because in actual experience we always have, not only intermediate stages between tones of different pitch, but also transitions between absolutely simple tones and noises made up of a profusion of simple tones. The result of these conditions is that every many-dimensional sensational system has a corresponding complex system of affective tones, in which every point generally belongs at once to several dimensions, so that the feeling corresponding to a given sensation is a resultant of the affective elements due to its position in various dimensions of the sensational system.