The Principles of Psychology, Volume 1 · William James
REACTION-TIME.
One of the lines of experimental investigation most diligently followed of late years is that of the ascertainment of the _time occupied by nervous events_. Helmholtz led off by discovering the rapidity of the current in the sciatic nerve of the frog. But the methods he used were soon applied to the sensory nerves and the centres, and the results caused much popular scientific admiration when described as measurements of the 'velocity of thought.' The phrase 'quick as thought' had from time immemorial signified all that was wonderful and elusive of determination in the line of speed; and the way in which Science laid her doomful hand upon this mystery reminded people of the day when Franklin first '_eripuit cœlo fulmen_,' foreshadowing the reign of a newer and colder race of gods. We shall take up the various operations measured, each in the chapter to which it more naturally pertains. I may say, however, immediately, that the phrase 'velocity of _thought_' is misleading, for it is by no means clear in any of the cases what particular act of thought occurs during the time which is measured. 'Velocity of nerve-action' is liable to the same criticism, for in most cases we do not know what particular nerve-processes occur. What the times in question really represent is the total duration of certain _reactions upon stimuli_. Certain of the conditions of the reaction are prepared beforehand; they consist in the assumption of those motor and sensory tensions which we name the expectant state. Just what happens during the actual time occupied by the reaction (in other words, just what is added to the pre-existent tensions to produce the actual discharge) is not made out at present, either from the neural or from the mental point of view.
The method is essentially the same in all these investigations. A signal of some sort is communicated to the subject, and at the same instant records itself on a time-registering apparatus. The subject then makes a muscular movement of some sort, which is the 'reaction,' and which also records itself automatically. The time found to have elapsed between the two records is the total time of that observation. The time-registering instruments are of various types. One type is that of the revolving drum covered with smoked paper, on which one electric pen traces a line which the signal breaks and the 'reaction' draws again; whilst another electric pen (connected with a pendulum or a rod of metal vibrating at a known rate) traces alongside of the former line a 'time-line' of which each undulation or link stands for a certain fraction of a second, and against which the break in the reaction-line can be measured. Compare Fig. 21, where the line is broken by the signal at the first arrow, and continued again by the reaction at the second. Ludwig's Kymograph, Marey's Chronograph are good examples of this type of instrument. Another type of instrument is represented by the stopwatch, of which the most perfect form is Hipp's Chronoscope. The hand on the dial measures intervals as short as 1/1000 of a second. The signal (by an appropriate electric connection) starts it; the reaction stops it; and by reading off its initial and terminal positions we have immediately and with no farther trouble the time we seek. A still simpler instrument, though one not very satisfactory in its working, is the 'psychodometer' of Exner & Obersteiner, of which I picture a modification devised by my colleague Professor H. P. Bowditch, which works very well.
1. The stimulus excites the peripheral sense-organ adequately for a current to pass into the sensory nerve; 2. The sensory nerve is traversed; 3. The transformation (or reflection) of the sensory into a motor current occurs in the centres; 4. The spinal cord and motor nerve are traversed; 5. The motor current excites the muscle to the contracting point. Time is also lost, of course, outside the muscle, in the joints, skin, etc., and between the parts of the apparatus; and when the stimulus which serves as signal is applied to the skin of the trunk or limbs, time is lost in the sensorial conduction through the spinal cord. The stage marked 3 is the only one that interests us here. The other stages answer to purely physiological processes, but stage 3 is psycho-physical; that is, it is a higher-central process, and has probably some sort of consciousness accompanying it. What sort?
Wundt has little difficulty in deciding that it is consciousness of a quite elaborate kind. He distinguishes between two stages in the conscious reception of an impression, calling one _perception_, and the other _apperception_, and likening the one to the mere entrance of an object into the periphery of the field of vision, and the other to its coming to occupy the focus or point of view. _Inattentive awareness_ of an object, and _attention_ to it, are, it seems to me, equivalents for perception and apperception, as Wundt uses the words. To these two forms of awareness of the impression Wundt adds the conscious volition to react, gives to the trio the name of 'psycho-physical' processes, and assumes that they actually follow upon each other in the succession in which they have been named.[108] So at least I understand him. The simplest way to determine the time taken up by this psycho-physical stage No. 3 would be to determine separately the duration of the several purely physical processes, 1, 2, 4, and 5, and to subtract them from the total reaction-time. Such attempts have been made.[109] But the data for calculation are too inaccurate for use, and, as Wundt himself admits,[110] the precise duration of stage 3 must at present be left enveloped with that of the other processes, in the total reaction-time.
My own belief is that no such succession of conscious feelings as Wundt describes takes place during stage 3. It is a process of central excitement and discharge, with which doubtless some feeling coexists, but _what_ feeling we cannot tell, because it is so fugitive and so immediately eclipsed by the more substantive and enduring memory of the impression as it came in, and of the executed movement of response. Feeling of the impression, attention to it, thought of the reaction, volition to react, _would_, undoubtedly, all be links of the process _under other conditions_,[111] and would lead to the same reaction--after an indefinitely longer time. But these other conditions are not those of the experiments we are discussing; and it is mythological psychology (of which we shall see many later examples) to conclude that because two mental processes lead to the same result they must be similar in their inward subjective constitution. The feeling of stage 3 is certainly no articulate perception. It can be nothing but the mere sense of a reflex discharge. _The reaction whose time is measured is_, in short, _a reflex action pure and simple, and not a psychic act_. A foregoing psychic condition is, it is true, a prerequisite for this reflex action.
The preparation of the attention and volition; the expectation of the signal and the readiness of the hand to move, the instant it shall come; the nervous tension in which the subject waits, are all conditions of the formation in him for the time being of a new path or arc of reflex discharge. The tract from the sense-organ which receives the stimulus, into the motor centre which discharges the reaction, is already tingling with premonitory innervation, is raised to such a pitch of heightened irritability by the expectant attention, that the signal is instantaneously sufficient to cause the overflow.[112] No other tract of the nervous system is, at the moment, in this hair-trigger condition. The consequence is that one sometimes responds to a _wrong_ signal, especially if it be an impression of the same _kind_ with the signal we expect.[113] But if by chance we are tired, or the signal is unexpectedly weak, and we do not react instantly, but only after an express perception that the signal has come, and an express volition, the time becomes quite disproportionately long (a second or more, according to Exner[114]), and we feel that the process is in nature altogether different.
In fact, the reaction-time experiments are a case to which we can immediately apply what we have just learned about the summation of stimuli. 'Expectant attention' is but the subjective name for what objectively is a partial stimulation of a certain pathway, the pathway from the 'centre' for the signal to that for the discharge. In Chapter XI we shall see that all attention involves excitement from within of the tract concerned in feeling the objects to which attention is given. The tract here is the excito-motor arc about to be traversed. The signal is but the spark from without which touches off a train already laid. The performance, under these conditions, exactly resembles any reflex action. The only difference is that whilst, in the ordinarily so-called reflex acts, the reflex arc is a permanent result of organic growth, it is here a transient result of previous cerebral conditions.[115]
I am happy to say that since the preceding paragraphs (and the notes thereto appertaining) were written, Wundt has himself become converted to the view which I defend. He now admits that in the shortest reactions "there is neither apperception nor will, but that they are merely _brain-reflexes due to practice_."[116] The means of his conversion are certain experiments performed in his laboratory by Herr L. Lange,[117] who was led to distinguish between two ways of setting the attention in reacting on a signal, and who found that they gave very different time-results. In the '_extreme sensorial_' way, as Lange calls it, of reacting, one keeps one's mind as intent as possible upon the expected signal, and 'purposely avoids'[118] thinking of the movement to be executed; in the '_extreme muscular_' way one 'does not think at all'[119] of the signal, but stands as ready as possible for the movement. The muscular reactions are much shorter than the sensorial ones, the average difference being in the neighborhood of a tenth of a second. Wundt accordingly calls them 'shortened reactions' and, with Lange, admits them to be mere reflexes; whilst the sensorial reactions he calls 'complete,' and holds to his original conception as far as they are concerned.
The facts, however, do not seem to me to warrant even this amount of fidelity to the original Wundtian position. When we begin to react in the 'extreme sensorial' way, Lange says that we get times so very long that they must be rejected from the count as non-typical. "Only after the reacter has succeeded by repeated and conscientious practice in bringing about an extremely precise co-ordination of his voluntary impulse with his sense-impression do we get times which can be regarded as typical sensorial reaction-times."[120] Now it seems to me that these excessive and 'untypical' times are probably the real 'complete times,' the only ones in which distinct processes of actual perception and volition occur (see above, pp. 88-9). The typical sensorial time which is attained by practice is probably another sort of reflex, less perfect than the reflexes prepared by straining one's attention towards the movement.[121] The times are much more variable in the sensorial way than in the muscular. The several muscular reactions differ little from each other. Only in them does the phenomenon occur of reacting on a false signal, or of reacting before the signal. Times intermediate between these two types occur according as the attention fails to turn itself exclusively to one of the extremes. It is obvious that Herr Lange's distinction between the two types of reaction is a highly important one, and that the 'extreme muscular method,' giving both the shortest times and the most constant ones, ought to be aimed at in all comparative investigations. Herr Lange's own muscular time averaged 0''.123; his sensorial time, 0''.230.
These reaction-time experiments are then in no sense measurements of the swiftness of _thought_. Only when we complicate them is there a chance for anything like an intellectual operation to occur. They may be complicated in various ways. The reaction may be withheld until the signal has consciously awakened a distinct idea (Wundt's discrimination-time, association-time) and then performed. Or there may be a variety of possible signals, each with a different reaction assigned to it, and the reacter may be uncertain which one he is about to receive. The reaction would then hardly seem to occur without a preliminary recognition and choice. We shall see, however, in the appropriate chapters, that the discrimination and choice involved in such a reaction are widely different from the intellectual operations of which we are ordinarily conscious under those names. Meanwhile the simple reaction-time remains as the starting point of all these superinduced complications. It is the fundamental physiological constant in all time-measurements. As such, its own variations have an interest, and must be briefly passed in review.[122] The reaction-time varies with the _individual_ and his _age_. An individual may have it particularly long in respect of signals of one sense (Buccola, p. 147), but not of others. Old and uncultivated people have it long (nearly a second, in an old pauper observed by Exner, Pflüger's Archiv, vii, 612-4). Children have it long (half a second, Herzen in Buccola, p. 152).
_Practice_ shortens it to a quantity which is for each individual a minimum beyond which no farther reduction can be made. The aforesaid old pauper's time was, after much practice, reduced to 0.1866 sec. (_loc. cit._ p. 626). _Fatigue_ lengthens it. _Concentration of attention_ shortens it. Details will be given in the chapter on Attention. The _nature of the signal_ makes it vary.[123] Wundt writes: "I found that the reaction-time for impressions on the skin with electric stimulus is less than for true touch-sensations, as the following averages show: Average. Average Variation Sound 0.167 sec. 0.0221 sec. Light 0.222 sec. 0.0219 sec. Electric skin-sensation 0.201 sec. 0.0115 sec. Touch-sensations 0.213 sec. 0.0134 sec. "I here bring together the averages which have been obtained by some other observers: Hirsch. Hankel. Exner. Sound 0.149 0.1505 0.1360 Light 0.200 0.2246 0.1506 Skin-sensation 0.182 0.1546 0.1337"[124] _Thermic_ reactions have been lately measured by A. Goldscheider and by Vintschgau (1887), who find them slower than reactions from touch. That from heat especially is very slow, more so than from cold, the differences (according to Goldscheider) depending on the nerve-terminations in the skin. _Gustatory_ reactions were measured by Vintschgau. They differed according to the substances used, running up to half a second as a maximum when identification took place. The mere perception of the presence of the substance on the tongue varied from 0''.159 to 0''.219 (Pflüger's Archiv, xiv, 529).
_Olfactory_ reactions have been studied by Vintschgau, Buccola, and Beaunis. They are slow, averaging about half a second (cf. Beaunis, Recherches exp. sur l'Activité Cérébrale, 1884, p. 49 ff.). It will be observed that _sound_ is more promptly reacted on than either _sight_ or _touch. Taste_ and _smell_ are slower than either. One individual, who reacted to touch upon the tip of the tongue in 0''.125, took 0''.993 to react upon the taste of quinine applied to the same spot. In another, upon the base of the tongue, the reaction to touch being 0''.141, that to sugar was 0''.552 (Vintschgau, quoted by Buccola, p. 103). Buccola found the reaction to odors to vary from 0''.334 to 0''.681, according to the perfume used and the individual. The _intensity of the signal_ makes a difference. The intenser the stimulus the shorter the time. Herzen (Grundlinien einer allgem. Psychophysiologie, p. 101) compared the reaction from a _corn_ on the toe with that from the skin of the hand of the same subject. The two places were stimulated simultaneously, and the subject tried to react simultaneously with both hand and foot, but the foot always went quickest. When the sound skin of the foot was touched instead of the corn, it was the hand which always reacted first. Wundt tries to show that when the signal is made barely perceptible, the time is probably the same in all the senses, namely, about 0.332'' (Physiol. Psych., 2d ed., ii, 224).
Where the signal is of touch, the place to which it is applied makes a difference in the resultant reaction-time. G. S. Hall and V. Kries found (Archiv f. Anat. u. Physiol., 1879) that when the finger-tip was the place the reaction was shorter than when the middle of the upper arm was used, in spite of the greater length of nerve-trunk to be traversed in the latter case. This discovery invalidates the measurements of the rapidity of transmission of the current in human nerves, for they are all based on the method of comparing reaction-times from places near the root and near the extremity of a limb. The same observers found that signals seen by the periphery of the retina gave longer times than the same signals seen by direct vision. The _season_ makes a difference, the time being some hundredths of a second shorter on cold winter days (Vintschgau _apud_ Exner, Hermann's Hdbh., p. 270). _Intoxicants_ alter the time. _Coffee_ and _tea_ appear to shorten it. Small doses of _wine_ and _alcohol_ first shorten and then lengthen it; but the shortening stage tends to disappear if a large dose be given immediately. This, at least, is the report of two German observers. Dr. J. W. Warren, whose observations are more thorough than any previous ones, could find no very decided effects from ordinary doses (Journal of Physiology, viii, 311). _Morphia_ lengthens the time. _Amyl-nitrite_ lengthens it, but after the inhalation it may fall to less than the normal. Ether and chloroform lengthen it (for authorities, etc., see Buccola, p. 189).
Certain _diseased states_ naturally lengthen the time. The _hypnotic trance_ has no constant effect, sometimes shortening and sometimes lengthening it (Hall, Mind, viii, 170; James, Proc. Am. Soc. for Psych. Research, 246). The time taken to _inhibit_ a movement (e.g. to cease contraction of jaw-muscles) seems to be about the same as to produce one (Gad, Archiv f. (Anat. u.) Physiol., 1887, 468; Orchansky, _ibid._1889, 1885). An immense amount of work has been done on reaction-time, of which I have cited but a small part. It is a sort of work which appeals particularly to patient and exact minds, and they have not failed to profit by the opportunity.
Text from Project Gutenberg, public domain in the United States.