or 
| J.—PSYCHOLOGY. 18é 
unverifiable conclusions arising from the applications of mathematics t 
_ psychological data. 
But in psychology an additional difficulty confronts us. We have to 
recognise that the data with which the ‘mathematical psychologist ’ 
operates are not measurements of the fundamental subject-matter of 
_ psychology—conscious mental processes. (Nor can they probably be 
_ measurements of unconscious processes, so long as the latter are regarded 
-asmentalin character.) For mental processes are not directly measurable : 
we can grade a series of conscious experiences according to their degree or 
amount, say of hue, brightness, loudness, pitch, temperature, extent, 
Se cmstion, clearness, pleasantness, etc.: we can say that one member of 
such a series has more or less of any one such character or quality than 
another member, or that the difference between two members of a series 
in respect to any one of these characters or qualities is greater or less 
_ than, or equal to, the difference between two other members. We cannot 
sony that “ whatever exists exists in some amount.” But the psychologist 
can only measure the amount of any conscious experience indirectly— 
_ either by reference to behaviour, i.e. to the organism’s physical response 
or expression, or by reference to the physical character of the relevant 
stimuli, in terms of objective standards of number, space and time which 
are immediately independent of actual subjective experience. 
Let us remember, then, that when we are attempting to measure any 
mental ability or character or quality by means of a test or series of tests, 
_ we are not directly measuring that mental ability or character or quality, 
_ but only the corresponding stimulus or the outward response or expression 
by which that mental ability or character or quality is manifested. We 
are, no doubt, justified in assuming a broad correlation between the speed, 
accuracy, amount, etc., of the response or expression of a mental ability 
or character or quality and the degree in which that mental ability or 
nie or quality is present. Even this broad assumption, however, 
‘sometimes unjustifiable, as in the case where too much of a given 
T onger to an improvement, in the corresponding pate inne ‘But we 
e certainly never justified in assuming that we can accurately measure 
“any mental process by measuring its objective response—that, for 
‘instance, twice the amount of the response necessarily means twice the 
quantity of the mental process of which the response is the expression. 
All that we are measuring is behaviowr—that is to say, something largely 
the efferent side, something largely physical and indescribable in terms 
of pure immediate experience, involving a complex of factors many of 
which, indeed, may be remote from those which we commonly believe, we 
‘are Measuring ; whereas what we ultimately aim to deal with in psychology 
1s experience—the meeting point of the afferent and efferent sides. 
In fact, then, the ‘ behaviourists’ are quite right when they insist 
that scientific. measurement is applicable only to the behaviour of the 
rganism. Where they are quite wrong is in their assumption that 
sonscious processes must necessarily be ousted from scientific psychology, 
-§* Measurement in Education,’ by Thorndike. xvii Yearbook of Nat. Soc. Stud. 
a? 
- 
