286 



SCIENCE. 



[Vol. XIV. No. 351 



ond) ; the longest, for "sing-dance," "mountain-level" (from i.ioo 

 to 1.400 seconds). 



III. Instead of accepting the first association formed with the 

 call-word, the experimenter requires a word bearing a definite re- 

 lation to that word. What this relation is to be is announced just 

 before the call-word itself. The relation is always such as to 

 admit of several replies. Such general rubrics as, given a country 

 to name a city in it, or given a general term to name a particular 

 instance of it, sufficiently well describe what was wanted. How- 

 ever, the same rubric was not repeatedly given in succession, as 

 Cattell had done (for this gives too much scope for preparing the 

 answer), but all kinds of relations were employed in an arbitrary 

 order. This " limited " association occupies M .970 of a second 

 {v, .200), and R 1.103 seconds (v, .210) ; or .125 and .155 of a sec- 

 ond longer than II. The shortest times were found, rather sug- 

 gestively, in naming an instance of "a German wine, — Rudi- 

 sheimer ; " " a number between 10 and 4, — 6 ; " " a Greek poet, — 

 Homer;" — all between .450 and .600 of a second. The longest 

 were " beast of the desert, — lion ; " " French author, — Voltaire ; " 

 "a drama of Goethe, — Gbtz ; " — between 1.200 and 1.500 sec- 

 onds. 



IV. Here the associations were still further limited, there being 

 in each case only one cortect association : it is in the nature of 

 question and answer. Though, of course, the question was not 

 asked in full, it was easily understood as such ; nor was the query 

 such as to immediately suggest an evident answer. The average 

 time was, for M, .808 of a second {v, .180) ; for R, .889 of a sec- 

 ond (v, .140) ; or .162 and .214 of a second shorter than III. It 

 thus takes longer to name a drama of Schiller, than t/ie _firsi drama. 

 of Schiller. Times between .400 and .600 of asecond were, " Three 

 times four, — Twelve ; " " On what river is Cologne ? — Rhine ; " 

 "In what season is June.' — Summer ;"" In what continent is 

 India? — Asia." The longest times, 1. 100 to 1.300 seconds, were 

 needed to answer the following questions : " By what author is 

 Hamlet ? — Shakspeare ; " " What is the capital of Baden .' — Karls- 

 ruhe;" "What is the color of ice.' — White;" "Who was the 

 teacher of Plato ? — Socrates." These times seem t j be influenced 

 by momentary fluctuations of the mind as well as the intrinsic 

 difliiculty of the question. 



V. As in IV., there is but one correct answer : but this is not 

 obtained simply as an act of memory, but some process of com- 

 parison and judgment must be gone through after the question is 

 proposed ; e.g., " Which is larger, — a lion or a mouse .' " " Who 

 is greater, — Hume or Kant .' " First the general nature of the 

 comparison is announced, then the special terms to be compared. 

 An actual test shows no difference in time when the correct answer 

 was the first and when it was the second of the two terms. M's 

 average time was .906 of a second (v, .180) ; R's, 1.079 C'^- -220) 

 seconds. Most quickly answered (.600-.800 of a second) were, 

 " What smells better, — cloves or violets? — Violets;" "Who is 

 greater, — Virgil or Ovid ? — Virgil ; " " What is prettier, — woods 

 or mountain ? — Mountain." It took longest (i. 200-1. 500 seconds) 

 to answer, " What is healthier, — swimming or dancing ? — Swim- 

 ming ; " " What of Goethe do you know better, — drama or lyric ? — 

 Ljric;" "What is more difficult, — physics or chemistry? — 

 Chemistry." 



VI. This variation consisted in employing the same general line 

 of questions as in the foregoing, but preceding the question with a 

 series of about a dozen words of the same general class as the two 

 to be compared, and mentioning the two among the number. 

 Thus, "Apples, pears, cherries, nuts, peaches, grapes, strawberries, 

 dates, figs, raisins: which do you like better, — grapes or cher- 

 ries? — Cherries." The questions were prepared in advance, and 

 read as monotonously as possible. The result was, for M, .694 of 

 a second (v, .130) ; for R, .659 of a second (v, .160) ; or .212 and 

 .420 of a second shorter than V., — certainly a striking result. 

 Among the shortest times (.400 to .600 of a second) were " [men- 

 tioning twelve composers] Who is greater, — Gliick or Bach? — 

 Bach;" " [twelve capitals] Which is more important, — Rome or 

 Madrid? — Rome." Among longest times (.800 to i.ooo second) 

 were " [ten classical dramas] Which is more taking, — Gotz or 

 Tasso ? — Gotz ; " " [ten colors] Which goes better with blue, — 

 yellow or green ? — Yellow." 



VII. Here we combine III. and V. Instead of first asking, 

 " What is a drama of Goethe?" (III.), and then "Which is the 

 finest of Goethe's dramas, — Gotz, Faust, etc.?" (VI.), we ask at 

 once, " Which is the finest of Goethe's dramas? " the subject hav- 

 ing first to recall what the dramas are, and then to make his choice. 

 For this, M requires .962 {v, .180), and R 1.137 {v, .160), seconds ; 

 or only .008 and .034 of a second more than in III. The shortest 

 times (.600-.700 of a second) were in answering " What is the 

 pleasantest odor ? — Rose ; " " Which is the most important German 

 river ? — Rhine : " the longest (1.400 to 1.600), in answering " Who 

 is the most difficult Greek author? — Pindar;" "Who is your 

 favorite French writer ? — Corneille." 



VIII. By adding a comparison to the process in VII., we have 

 the scheme of VIII. : e.g., " Which is the more westerly, — Beilin. 

 or the most important German river? — [answer] Rhine;" or, 

 " What letter comes first in the alphabet, — L, or the initial letter 

 of the name of the prettiest tree? — [answer] Beech." This very 

 complex process engaged M for 1.844 seconds {v, .370), and R for 

 1.866 seconds {v, .340). Here the variations in the difficulty of the 

 questions and the alertness of the individual become rather impor- 

 tant. A still more complicated scheme was attempted, but the 

 results proved too variable. The questions were of this type '• 

 " What is more impressive, — Shakspeare's finest drama, or Wag- 

 ner's finest opera? [answer] Lohengrin." The result for M was 

 2.197 seconds {v, .970); for R, 2.847 seconds {v, .720). It is the 

 scheme of VIII. with case VII. in each term of the comparison. 



IX. This bears the same relation to VIII. that VII. does to IV. 

 The type of query would be, " Which lies more westerly, — Berlin, 

 or the river on which Cologne is situated [it is a comparison in 

 which one term is reached by the substitution of a concrete for a 

 generally described term]?" or, again, "Which is less, — 15. or 

 20 — 8 ? " " Who lived later, — Klapstich, or the author of ' Lear '?'' 

 The times were, for M, 1.291 seconds {tj, .180) ; forR, 1.337 seconds 

 (v. .230). These times are .553 and .529 of a second less thark 

 VIII., although the differences between VII. and IV., which have 

 the same relation pair to pair, were only .154 and .248 of a second. 



X. This is IX. preceded by a series of words of the same cate- 

 gory as the terms of the comparison : e.g., the experimenter first 

 calls twelve names of parts of the body, and then, " Which is 

 larger, — the hand, or that with which one smells? " " [ten colors] 

 Which is lighter, — blue, or the color of sulphur? " Here we have 

 the same process of shortening the time as in VI. The times are 

 1. 153 seconds for M \V, .170), and 1.145 seconds for R {v, .210) ; or 

 .138 and .192 of a second less than IX. The results of the ten 

 cases are based upon about 800 experiments. * 



Finding the current theory' entirely unable to explain these re- 

 sults. Dr. MUnsterberg attempts an explanation upon the principle 

 " that a stimulus begins to have an effect before the latter is con- 

 sciously perceived." and that " the process consists in the re-awak- 

 ening of the effects of previous stimuli, so that it is shortened by 

 any circumstances tending to call up these reproductions in advance 

 of the stimulus." We must note that all these associations are 

 really of the form of judgments. When we are asked to associate 

 a word with a given word, we really do not answer with the first 

 impression that may be passing through our minds at that particu- 

 lar moment ; but we give a word called up by the former, and the 

 relation to which we more or less clearly recognize. The question 

 is, " Name a word standing in some relation to the following 

 word ; " and our answer is, " This word answers this description." 

 Every such judgment, again, is the assertion of identity between twcK 

 objects standing in different associative relations. When we say 

 the beech is the prettiest tree, we mean that the thing that from 

 one aspect we call and recognize as a beech, and the concept which 

 from another aspect we describe as the prettiest tree, are one and 

 the same. 



The mind can be placed in more or less favorable attitude for 

 the calling-up of these association ties, and thus the time as meas- 

 ured be shortened ; and again the calling-up of certain associative 



1 The theory as represented by Wundt would require the limited ; 

 take longer than the unlimited, because it involves the latter and something more ; 

 but IV. requires less time than III. or II. Again, as VIl. is composed of III. and 

 v., we should get the times by the addition of these two times ; but this gives far too 

 large results. Other discrepancies could easily be pointed out. 



