syllables for instance, then not only have the pauses to be lengthened, but the eye-stretches must be shortened as well. Yet the amount of print actually in focus and so distinctly perceptible is the same in each case. In reading 'sense' then a portion of what the eye takes in extends beyond the focus of distinct vision. Like the single letters at the first, several words or syllables at the last, are apprehended-in virtue of their general form or of a few salient traits-as a single whole. Indeed adequate apprehension of this sort, in the case of a coherent context, is possible when its distance from the eye exceeds the limits of exact definition altogether. But at the ordinary range of reading, when a portion at any rate of what the coup d'œil takes in is distinctly seen, more is read and more quickly. Here the part in the margin of the field of vision is usually mainly to the right of the fixation point, shewing the influence of the prior context in extending the span of apprehension. The child learning to read begins by reading aloud syllable by syllable. But the spoken syllable and the syllable as heard are already integrated into one complex whole: the new task then is simply to associate this whole with its visual symbols. Both for articulation and for audition, a series of syllables, always remains, as at first, a temporal series. Vision, however, has here the same superiority over movement and hearing as it has elsewhere over movement and touch: it can take in several syllables at once, although they can be heard or spoken only one at a time. At first, of course, this superiority does not count; but eventually it becomes easy to read far faster than one can speak, faster even than one can distinctly hear. There is evidence—perhaps not all that one could wish-to shew that "rapid readers not only do their work in less time but do superior work. They retain more of the substance of what is read than do slow readers'." No doubt because, in general, they concentrated their attention more, and being also more intelligent, integrated' better than the slower readers. Before proceeding, let us here note that in what is called endophasia or 'internal speech' there are three main types of verbal imagery, the motor, the auditory and the visual: words, that is to say, are 'mentally' spoken or heard or seen. For the entirely illiterate 1 Quantz, "Problems in the Psychology of Reading," Psychological ReviewMonograph Supplements, ii. (1897), p. 49. internal speech of the visual sort is, of course, impossible; and it is, in fact, usually absent in most people. It is so not merely because the race as a whole, and they as individuals, mastered speech before beginning to read at all, but also because they speak so much oftener than they read. Usually the motor and the auditory type are combined, the dominance of the motor being specially apparent in the reading of young children and the comparatively illiterate, who either speak aloud or whisper while they read; but this trait becomes less and less marked with increasing culture. Among thoroughly cultured persons a few cases of the exclusively visual type are found and still more of the combined visual-motor1. It seems further not unlikely that as moderate practice banishes articulation from reading and as frequent reading leads to an increasing prominence of visual word-imagery, both audition and articulation may for some fade out more or less entirely, and the visual word alone remain prominent. The few investigations that have been made bear out this conjecture: the fastest readers seem to be visualisers. The most perfect kind of integration would in this way be attained. The advantage which vision secures us in taking in the tout ensemble of things it seems also to secure in dealing with thought as a whole, when this is visualised in symbols. Herein perhaps lies the secret of Bacon's saying that writing makes an exact man, for in setting out our thoughts in black and white we secure a survey of them that internal audition alone can never give us. APPENDIX 'Age' and 'Strength' of Associations. §6. A somewhat paradoxical situation is brought to light when the method of saving and the method of scoring are used together. In the experiments by Jost, mentioned above, two series of verses, S1, S, were repeated thirty times; after an interval of twenty-four hours S1 was tested by the first method and S, by the second. 1 Cf. G. Saint-Paul, Le langage intérieur, 1904, pp. 200 f. 2 Cf. W. B. Secor, "Visual Reading: A Study in Mental Imagery," Am. J. of Psych. xi. (1899), pp. 225 ff.; Quantz, op. cit. pp. 46 ff. 3 § 1, p. 225. Two new series, S1, S., were then taken: S, was repeated four times, and after an interval of a minute tested by the first method; S, was then repeated in like manner, and tested after the same interval by the second method. This procedure was renewed day after day-in varying order-till records of twenty cases of both old and new series tested by each method were obtained. It was then found (by the method of saving) that an old series (an S1) required on an average 5.85 repetitions for relearning, and a new series (an S1) 9'6. But (by the method of scoring) it was found that a new series (an S.) yielded 27 hits,' with an average time of about 1 seconds for each, while an old series (an S1) yielded only 9 'hits,' with an average time of 4 seconds for each. Thus one may be able to reproduce relatively little of a given subject-matter, and yet require only a few repetitions in order to learn it off anew; on the other hand, one may know relatively much, and still find many more repetitions requisite for such complete learning. The 'age' of the associations is then important. Other things being equal, we may conclude that each fresh repetition effects more for older associations than for more recent ones. It might be supposed that the strength of the old associations was more uniform and on the average greater than the strength of the new; so that while none of the old were far below the threshold, few, if any, were above it; whereas more of the new might be above the threshold though the majority had lapsed entirely. And the latter would certainly be the case if the subject of experiment tried to make sure of a few 'hits,' and paid no attention to the rest of the series. Due care was, however, taken that the ends of the experiment should not in this way be defeated. Also, there is ample evidence to show that the supposed greater uniformity in strength of old associations is not, in fact, the rule, We seem left, then, to conjecture that the difference is the effect of the process of assimilation working subconsciously-that psychical aspect of nervous growth which Professor James has aptly characterized by saying that "we learn to skate in summer and to swim in winter." It continually happens that we can recognise connexions that we are quite unable to reproduce. To the diminished 'strength' of an association, as tested by the method of scoring, there may then quite well be an equivalent set-off in more developed assimilation. As a seed germinates it has less latent energy, but this is replaced by growth W. P. 16 in root and stem: similar relations may obtain when an old association is said merely to lose strength.' On the other hand -within the range of the primary memory-image—we can often reproduce what after a longer interval we should fail to recognise. We seem warranted, then, in concluding that this conception of 'association-strength' so freely used by G. E. Müller and his co-workers, requires more analysis than it has yet received. The two factors which their methods disclose in it appear to confirm the distinction we have already made between impressions and free ideas. They help us also to understand, further, the superiority of distributed over cumulated repetition, of 'inwardly digesting' over 'cram'.' 1 There is a most interesting article by P. B. Ballard dealing with many of the topics of this chapter that I have unhappily overlooked. It is entitled "Obliviscence and Reminiscence"; see British fl. of Psychology, Monograph Supplements, ii. 1913. CHAPTER X FEELING Introductory. § 1. Such summary survey of the more elementary facts of cognition, as our plan of exposition called for, is here at an end. So far the most conspicuous factors at work have been those concerned in the formation of what might be termed the ideational mechanism. In dealing with the higher processes of thought we shall have to take still more account of the voluntary activity, which we have seen to be essential even in the lower processes of perception and ideation, and also of the part played by language in perfecting the higher, intellectual, processes. But it seems preferable, before entering upon these topics, to explore also the affective and conative constituents of mind in their more elementary phases, so as to complete in outline our description of mind below what we may call the stage of understanding or reason. We have found that psychical life consists in the main of a continuous alternation of predominantly receptive and predominantly reactive consciousness. In its earliest form experience is simply an interplay of sensation and movement. At a later stage, we find that in the receptive or cognitive phase ideation is added to perception; and that in the active phase, thought, poetic fancy, &c.—or the voluntary direction and control of the ideational trains-are added to the voluntary direction and control of the sense-organs and of the muscles. At this higher level also it is possible that either form of receptive consciousness may lead to either form of active: sensations may lead to thought rather than to action in the restricted sense; and ideas apart from sensations may prompt to action rather than to thought. There is a further complication still not only |