ABSTRACT
Two reasons persuaded us to continue our study with the remembrance of an M-shaped figure built up from two serial configurations, one descending and the other ascending, and consisting of a total of eleven elements (Figure 4). The first was subjective or moral: our
immediate reaction to the unexpected discovery that the memory can improve after a lapse of seven or eight months was one of such disbelief ('the thing is too good to be true') that a control experiment seemed essential. The second was a wish to isolate the factors involved. The M-shaped series, in our view, constitutes an operation in the full sense of the term; it calls for reversibility in the form of a co-ordination between the ascending (E < F, G ... X) and the descending branches (E > D, C ... A), and we assumed that what improvements in the memory might occur in the course of six months would be due to a spontaneous development of this schema. But we realized that other factors might be responsible as well; for instance, progress in the direction of Gestalten and 'good forms'. Now, the M-shaped figure, while representing just such a good (and even symmetrical) form, clearly involves greater operational difficulties than the figure discussed in Chapter 1, since to construct it on paper or materially means passing from one series of relations A > B > C ... > M (midpoint) to a series of reciprocal relations M < N < 0 < P, etc., which, as we shall see, presents a special problem.
