altered according to the musical context. In fact, musical disorientation easily affects our whole system of tonal relations if we are introduced for a time into a realm of amusical tone-steps. The usual acoustical perspective can by such a process be so altered that even the octave may appear equal to the higher fifth.54 But waiving the question as to how inseparable is the fusion of a consonant interval we may recall that certain observed facts, notably those of synaesthesia, favor strongly the view that relatively inseparable associations can, under certain conditions, be formed by an individual, especially during childhood. C. S. Myers55 points out that in all his cases of colored audition it is the individual experience which seems to determine just what particular color will be associated with a given tone. A tone that is "seen" by one subject as green, appears to another as blue, and to another as yellow. "Two members of the same family, both endowed with synaesthesia, will hotly dispute the colors of certain words or vowels." Myers sees no possible explanation of these individual variations otherwise than as accidental personal associations. That this sort of association may be formed relatively late is indicated by the fact one subject reported on by Myers acquired, by playing with figures at the age of twelve, the rather troublesome habit of seeing with every letter of the alphabet its individual number. This association, acquired at random, has persisted for eighteen years, although it is not invariably present. Myers thinks that the general setting in which such connections are found must be a strong tendency to a certain kind of association,-"a 'sympathy,' the psychological and physiological basis of which is unknown to us." It seems to me that in any consonance theory some such 'sympathy' as this must be assumed as the basis of the greater affinity of the simpler related tones, but this does not prejudice the possibility that the actual realization of definite fusion developments may take place during the lifetime of an individual. Certainly the cases of L. E. Emerson, The Feeling Value of Unmusical Intervals, Harvard Psych. Studies, II, 269, 270. 55 5 C. S. Myers, A. Case of Synaesthesia, Brit. Journ. Psych. 1911, IV, 228-238. "Ibid., 238. Myers leave little doubt that synaesthesia, at least, though in general a congenital endowment, may in its particular manifestations be the result of the individual's life experience. To be sure, the "associations" in these are sensory rather than ideational. But exactly the same is to be said of that peculiar association of two tones which causes them to be heard as relatively unified. Stumpf argues57 that if mere being in consciousness innumerable times necessarily involved fusion, the person who heard music frequently should be less capable of differentiating the two notes of an octave than one who heard music seldom. But "fusion," according to his own definition, is something distinct from analysability; it is a qualitative unity that persists unaltered even after the ear has discriminated the two elements involved. If Stumpf maintained his own definition with strict consistency he would have to admit that the very practice which improves the musician's ear in discriminating the two component tones of an interval may be at the same time fusing the same two tones into a more perfect qualitative unity. It is only his failure to make clear the psychological difference between the two kinds of unity -the numerical "Einsheit" which disappears with analysis, and the qualitative “Einheit" which remains after analysis—that enables him to argue, now as if fusion were identical with difficulty of analysis, now as if it were something altogether distinct. The view to be maintained by the writer, that a tonal fusion means simply a highly permanent connection wrought under the direct influence of repeated sensory co-incidences, is rejected by Stumpf largely on "a priori" grounds; there is no record that this possibility was ever put by him to experimental test. The experiments to be reported in this paper were devised with the view to examining the characteristic changes effected by the frequent or prolonged hearing of certain intervals, it being my hope that an interpretation of these changes might throw light on the possibility of fusion changes denied by Stumpf. But before passing to the experimental results I shall state my own theoretical position more fully. "Tonpsychologie, II, 210. A GENETIC THEORY I shall begin with an hypothetical account of the genesis of consonance, the theoretical soundness of which will be tentatively assumed. After an examination of the historical and experimental facts to which it offers an explanation, I shall return to a discussion of theoretical points in detail. The tentative hypotheses with which we shall begin are: (1) Consonance and dissonance have to do directly with the degree to which an individual is able to synthesize A two tonal elements into a unified complex; this synthesis is unaffected by analysis, and the individual's ability to analyse is unaffected by synthesis. (2) Tonal synthesis may be conceived as an undertaking, attended by success or failure on the part of the hearing organism. Success is experienced as the characteristic quality of consonance, failure as that of dissonance. (3) The inherent feeling value of a particular interval is a function of two factors-success and degree of difficulty. It is the barely successful synthesis, in which the individual accomplishes with difficulty the unification of a manifold, that he finds the keenest pleasure. If the synthesis is effected without effort he has little awareness of his accomplishment, and therefore little pleasure. If, on the other hand, in spite of his synthesizing activity, he is baffled by the complexity of the manifold to be unified, he experiences acute displeasure. (4) A certain initial disposition to synthesis is pre-determined for a given interval by the simplicity of its numerical ratio. That is to say, we must pre-suppose a "sympathy" such as that which Myers postulates in the case of synaesthesia,—a more or A. I shall use the expression "synthesis" instead of "fusion" in order to avoid the ambiguity which has come to attach to the latter term. By "synthesis" I mean simply a kind of fusion that is wholly independent of analysability. + less strong tendency to association, the strength of the tendency being proportionate to the simplicity of the ratio. (5) The actual degree to which any given interval realizes its possibilities of synthesis depends upon the number of times the interval has been attentively heard by an individual or by his ancestors.A (6) Between the distinct qualities of consonance and dissonance lies an intermediate zone of neutrality such as, for the present-day ear, 4:7 relation. In such cases neither success nor failure of synthesis is pronounced. It is characteristic of this zone that change due to frequency of hearing is more marked than for either the decided consonances or dissonances. Also we are likely to find the greatest individual differences in judgment of an interval when it has reached this stage. Our task will be to show that the above assumptions, taken together, are the ones which best satisfy the facts. Now the facts with which we are concerned are of two kinds, those derived from the actual development of musical feeling, and those derived from experiments designed to test the conditions of consonance. We may examine the two kinds in order. A. The addition of the phrase "or by his ancestors" may seem to imply an anti-Weissman point of view on the question of acquired characteristics. I should be unwilling to admit, in the present state of that controversy, that such an implication would necessarily invalidate my position, but I do think it possible to account for the transmission, here assumed, as a purely social one. If thirds are acquired as a consonance by one generation they will be heard so much the more by the next, and will therefore become more completely consonant. We may accordingly interpret the degree of synthesis of a given interval either in Lamarckian fashion, as being in part a congenital association resulting from race experience, or, in accordance with Weissmannian principles, as resulting entirely from an individual experience which has had the benefit of the accumulations of the race. I am indebted to Professor B. M. Anderson for the suggestion that such accumulation of increments of advance as is here involved may be plausibly assumed in any society where the musical profession tends to run in families. The son, on account of his father's tutoring, reaches at an earlier age the point attained by his father, the grandson 'surpasses the son, and so on. EVIDENCE FROM THE HISTORY OF MUSIC Historical evidence does not go back far enough to either prove or disprove the statement that the octave was originally a dissonance. To be sure, we have what at first might be taken as evidence to the contrary in the findings of von Hornbostel and others58 to the effect that our present-day "primitive" peoples of Africa and Asia make abundant use of parallel octaves, and in a large number of cases, even of fifths and fourths. But it hardly needs to be argued that the conditions here are far from original. Even Stumpf, who makes much of anthropological evidence as proof that the present-day fusion grades were an original possession of the race, makes the following admission 59 "Wir müssen freilich auch schlieszen, dasz das meiste, was bei den Naturvölkern an Musik beobachtet wird, keineswegs die allerprimitivsten Zustände darstellt, sondern mindestens in Hinsicht . . . des ganzen Aufbaues doch schon vielfach eine lange Geschichte hinter sich hat." And C. S. Myers, 60 speaking of the contamination of primitive music, expresses the opinion that unpolluted aboriginal music can hardly be found nowadays owing to the almost universal influence of the trader and the missionary. He says: He says: "We have evidence of these conditions in the influence of Arabic and Portuguese tunes introduced into Africa, and in the spread of favorite native airs throughout North America and throughout Australia." If we examine still more rudimentary conditions we again find that such evidence as is obtainable is inconclusive. Experiments have been made1 to see whether a bird, by hearing its melody played after it in another key, could be induced to transpose. No variation on the part of the bird could be noticed. The 58 E. M. von Hornbostel, Zeitschrift f. Angew. Psych. III, 465-487. |