October ii, i! 



•] 



SCIENCE 



253 



ment as soon as an expected sensation has been received, Lange 

 had found a great and constant difference according as the atten- 

 tion was fixed upon the sensation, the expected sound, or upon the 

 movement of re-action. In the first case the subject distinctly 

 waits for the sensation, appreciates it, and then proceeds to move 

 the finger : it is a " sensory " re-action. In the second case the 

 impression is tal<en in almost automatically, and the desire is to 

 have the finger move the moment any impression is felt : it is a 

 " motor " re-action. The sensory is always longer than the motor 

 re-action. Lange found in three observers sensory times of .230, 

 .223, and .224 of a second ; and motor times of .123, .125, and .137 

 of a second, — an average difference of one-tenth of a second. 



I. Miinsterberg's sensory time is .162, his motor .120, of a 

 second, — a much smaller difference of only .042 of a second. It is 

 to be noted that the motor times of all the observers agree remark- 

 ably well, while their sensory times show individual variations. 

 The explanation of these facts will be attempted after the results 

 of certain other experiments have been given. 



II. The next complication consisted in calling out in an arbi- 

 trary order " one," " two," " three," " four," or " five," to which the 

 thumb, forefinger, middle finger, ring-finger, and little finger were 

 to respond respectively. This is more complicated, and involves 

 the association of " one " with a movement of the thumb ; and so 

 on. As before, one may fix the attention upon the expected sound 

 or upon the movement. The sensory time for the entire process 

 was .383 of a second ; the motor, .289 of a second, — a difference 

 of .094 of a second. When making a sensory re-action, the sound 

 is always appreciated ; and errors, i.e., raising the wrong finger, 

 never occur. In motor re-actions they occasionally occur, the 

 error invariably consisting in raising a neighboring finger. 



III. Here the re-action was the same; but, instead of the words 

 "one," "two," "three," "four," "five," the Latin declension — 

 luptis, lupi, lupo, lupitni, lupe — was associated with the five 

 fingers, the process being the same as before, but the association 

 more artificial. The result was, for sensory re-actions, .465 of a 

 second; for motor, .355 of a second, — a difference of .110 of a 

 second. Only a single false re-action was made. 



IV. The movements of the five fingers were associated respec- 

 tively with the five members of the three following declensions of 

 German pronouns : ich, 7neiner, mir, mich, wir; du, deiner, dir, 

 dich, ihr ; der, des, dem, den, die. This is more complex not only 

 by the change between three series,, but by the relatively minute 

 and irregular distinctions between the words. The sensory time 

 was .688 of a second ; the motor, .430 of a second, — a difference 

 of .258 of a second. Here errors occurred in the motor re-actions 

 ten per cent of the time, showing the increased facility of confusion. 

 It was noticed, too, that the second finger was often erroneously 

 raised in answer to du, apparently on account of its position in the 

 familiar series ich, du, er. 



V. Here the elements of the process are changed. If a noun 

 is called, the thumb is to be raised ; if an adjective, the fore- 

 finger ; if a pronoun, the middle finger; if a number, the ring- 

 finger ; if a verb, the little finger. This very artificial relation was 

 first thoroughly learned by going over the list, raising each finger 

 as the class to which it corresponded was mentioned. All the 

 words were monosyllables, and new words were being constantly 

 used, no word occurring twice. The sensory time was .712 of a 

 second ; the motor, .432 of a second, — a difference of .280 of a 

 second. Here errors are very frequent (thirty per cent), but are con- 

 fined to the motor re-actions. 



VI. This series was just like the former except that the cate- 

 gories were " a city," " a river," " an animal," " a plant," " an ele- 

 ment ; " such as "London," "Rhine," "dog," "rose," "gold." 

 The sensory time was .S93 of a second ; the motor, .432, — a dif- 

 ference of .461 of a second. Errors occurred in twelve per cent of 

 the motor re-actions. 



VII. Here the categories were still more difficult ; viz., " an 

 author," " a musician," " a naturalist," " a philosopher," " a states- 

 man or general." Only in a few very prominent cases is this decis- 

 ion easy. The average time was, for sensory re-actions, 1.122 

 seconds; for motor, .437 of a second, — a difference of .685 of a 

 second. Errors occurred in twenty- five per cent of the (motor) re- 

 actions. 



. It is to be noted that the cases I., II., and III. involve associa- 

 tions of a finger-movement with but one word : it is an " antici- 

 pated " association. In the other cases a more or less wide range 

 of words is to be re-acted upon by the same movement : it is a 

 " free " association. It will have been noticed, too, at what a rapid 

 rate the difference between sensory and motor times increases as 

 the processes become more complex, this difference being sixteen 

 times as great in VII. as in I. 



The explanation of the shortening of the re-action time by the 

 motor form of re-action is comparatively simple in Case I. We 

 need only assume, quite naturally, that the fixation of the attention 

 upon the movement really gets ready the innervation (as it were, 

 lights the match beforehand), and is thus immediately ready to 

 make the movement (to set afire the train of powder). But in the 

 following cases not only does this explanation become doubtfijl 

 (for, inasmuch as it is not known which finger is to be moved, only 

 a general, unspecialized innervation to move a finger can be an- 

 ticipated), but it can only account for .042 of a second of difference, 

 while the real difference progressively rises to sixteen times that 

 amount. We can be quite sure, then, that the shortening takes 

 place in the purely mental process of recognizing a given word as 

 an instance of a more or less general class, and of appreciating that 

 this class is to be represented by a certain movement. While in 

 the four last cases the sensory times rose from .688 to .712, to .893, 

 to I.I 12 seconds, the motor lime practically remained unchanged, — 

 .430, .432, .432, and .437 of a second. The increase in the sensory 

 time indicates that the processes are becoming mentally more 

 complex. It is more difficult to recognize that a given word (heard 

 only once during the experiments) is a certain part of speech than 

 to recognize a wOrd as one of the same three, ich, du, der, or 

 7neiner, deiner, des, and so on ; still more difficult to recognize a 

 concept as belonging to one of five well-known general categories ; 

 and most difficult to place a man in one of five special, somewhat 

 closely related professions. But why should these differences dis- 

 appear by simply fixing the attention upon the movement to be 

 executed } Indeed, according to a current theory, of which Wundt 

 is the acknowledged champion, and which Dr. Mlinsterberg fiercely 

 combats, turnmg the attention towards an act shortens the time of 

 its accomplishment ; fixing the attention upon the mental, sensory 

 part of the process should shorten the time. This apperception 

 theory, that conceives the mind as a point in which only a single 

 act has room at a given mpment, and through which accordingly 

 the several elements of a complicated process must pass seriatim, 

 gives no satisfactory explanation. Dr. Miinsterberg regards the 

 true explanation to lie in the fact that in the motor re-actions the 

 several parts of the mental process overlap in time. In the motor 

 re-action we have before us, as it.were, five possible movements, 

 each (aided, perhaps, by unconscious tentative movements) ready 

 to be made, and five lines of association along one of which the 

 impulse is coming. The moment the word is sounded, it is referred 

 to the " third-finger-moving category," or whichever it may be, — 

 the intermediate acts of recognizing, let us say, that the word was 

 " frog," that a frog is an animal (and not one of the other four 

 classes), and that when an animal's name is called we must raise 

 the third finger ; which acts are gone through consciously and 

 successively, in the sensory re-action being performed almost 

 simultaneously and automatically, or at least subconsciously. This, 

 in Cases IV., V., VI., and VII., would be about the same process, 

 the tracts of association (cortical fibre-connections to concretize the 

 conception) being about equally much used in each case, since 

 their entire use was *hat brought about by the experiments them- 

 selves. We see, too, why it is natural that in the sensory cases 

 errors did not arise, but that in the motor re-action an impulse 

 could readily be switched into a neighboring association-tract. Dr. 

 Miinsterberg regards the motor form of re-action as the one more 

 closely corresponding to natual, every-day processes ; the sensory 

 re action being a mere artificial, experimental result. When we 

 act and speak, the movement results before we have consciously 

 appreciated the excitation, analyzed it, and referred it to certain 

 categories. It seems to be referred to certain definitely established 

 trains of thought, the reasons for doing so never consciously ap- 

 pearing. 



While this explanation is not entirely adequate, it has the advan- 



