The phenomenon involves a rapid alternation of motor discharges and resultant afferent impressions, for as long a time as it lasts. They must be associated in one definite order; and the order must once have been learned, i.e., it must have been picked out and held to more and more exclusively out of the many other random orders which first presented themselves. The random afferent impressions fell out, those that felt right were selected and grew together in the chain. A chain which we actively teach ourselves by stringing a lot of right-feeling impressions together differs in no essential respect from a chain which we passively learn from someone else who gives us impressions in a certain order. So to make our ideas more precise, let us take a particular concatenated movement for an example, and let it be the recitation of the alphabet, which someone in our childhood taught us to say by heart.
What we have seen so far is how the idea of the sound or articulatory feeling of A may make us say 'A,' that of B, 'B,' and so on. But what we now want to see is why the sensation that A is uttered should make us say 'B,' why the sensation that B is uttered should make us say 'C,' and so on.
Fig. 89.
To understand this we must recall what happened when we first learned the letters in their order. Someone repeated A, B, C, D to us over and over again, and we imitated the sounds. Sensory cells corresponding to each letter were awakened in succession in such wise that each one of them (by virtue of our second law) must have 'drained' the cell just previously excited and left a path by which that cell tended ever afterwards to discharge into the cell that drained it. Let Sa, Sb, Sc in figure 89 stand for three of these cells. Each later one of them, as it discharges motorwards, draws a current from the previous one, Sb from Sa, and Sc from Sb. Cell Sb having thus drained Sa, if Sa ever gets excited again, it tends to discharge into Sb; whilst Sc having drained Sb, Sb later discharges into Sc, etc., etc.—all through the dotted lines. Let now the idea of the letter A arise in the mind, or, in other words, let Sa be aroused: what happens? A current runs from Sa not only into the motor cell Ma for pronouncing that letter, but also into the cell Sb. When, a moment later, the effect of Ma's discharge comes back by the afferent nerve and re-excites Sa, this latter cell is inhibited from discharging again into Ma and reproducing the 'primordial motor circle' (which in this case would be the continued utterance of the letter A), by the fact that the process in Sb, already under headway and tending to discharge into its own motor associate Mb, is, under the existing conditions, the stronger drainage-channel for Sa's excitement. The result is that Mb discharges and the letter B is pronounced; whilst at the same time Sc receives some of Sb's overflow; and, a moment later when the sound of B enters the ear, discharges into the motor cell for pronouncing C, by a repetition of the same mechanism as before; and so on ad libitum. Figure 90 represents the entire set of processes involved.
Fig. 90.
The only thing that one does not immediately see is the reason why 'under the existing conditions' the path from Sa to Sb should be the stronger drainage-channel for Sa's excitement. If the cells and fibres in the figure constituted the entire brain we might suppose either a mechanical or a psychical reason. The mechanical reason might lie in a general law that cells like Sb and Mb, whose excitement is in a rising phase, are stronger drainers than cells like Ma, which have just discharged; or it might lie in the fact that an irradiation of the current beyond Sb into Sc and Mc has already begun also; and in a still farther law that drainage tends in the direction of the widest irradiations. Either of these suppositions would be a sufficient mechanical reason why, having once said A, we should not say it again. But we must not forget that the process has a psychical side, nor close our eyes to the possibility that the sort of feeling aroused by incipient currents may be the reason why certain of them are instantly inhibited and others helped to flow. There is no doubt that before we have uttered a single letter, the general intention to recite the alphabet is already there; nor is there any doubt that to that intention corresponds a widespread premonitory rising of tensions along the entire system of cells and fibres which are later to be aroused. So long as this rise of tensions feels good, so long every current which increases it is furthered, and every current which diminishes it is checked; and this may be the chief one of the 'existing conditions' which make the drainage-channel from Sa to Sb temporarily so strong.[511]
The new paths between the sensory cells of which we have studied the formation are paths of 'association,' and we now see why associations run always in the forward direction; why, for example, we cannot say the alphabet backward, and why, although Sb discharges into Sc, there is no tendency for Sc to discharge into Sb, or at least no more than for it to discharge into Sa.[512] The first-formed paths had, according to the principles which we invoked, to run from cells that had just discharged to those that were discharging; and now, to get currents to run the other way, we must go through a new learning of our letters with their order reversed. There will then be two sets of association-pathways, either of them possible, between the sensible cells. I represent them in Fig. 91, leaving out the motor features for simplicity's sake. The dotted lines are the paths in the backward direction, newly organized from the reception by the ear of the letters in the order C B A.
