STATUS OF FACTORS IN PSYCHOLOGY 9

About the genetic factors that influence mental ability
and temperament comparatively little is known as yet.
Children undoubtedly resemble their parents in regard to
general intelligence and many other mental factors, and
that in a degree that cannot be wholly explained by post-
natal influences. Yet the relation between the observable
phenomena is exceedingly indirect, and typical of the remote
and complex type of causal determination with which the
correlationist has to deal in psychology. The most that we
can say about mental inheritance with any assurance is that
each individual apparently receives through his two parents
a very large sample of a still larger number of unit-deter-
miners ; that this sample is mainly but by no means
entirely random (certain groups of determiners, for ex-
ample, being always carried on the same chromosome) ;
that his subsequent development must involve a further
sampling of this sample (or rather of its possible effects) ;
and that his mental reaction in any given situation must
depend on yet another process of selecting or sampling
whatever tendencies have thus developed or survived.
It follows that, with few exceptions, the overt mental
types, which are all that the psychologist can detect with
his tests and rating scales, are related only in a very remote
and indirect fashion to inherited types or tendencies:
they are, as the biologist would say, phenotypes rather than
genotypes. Indeed, if there were any likelihood of estab-
lishing mental genotypes, factor-analysis, I imagine, would
hardly be the main line of approach which the genetic
psychologist would adopt in his endeavours to discover
them.

The Criticism of Factors.—In seeking to demonstrate the
existence of the mental factors I have described, different
psychologists have employed different modes of calculation.
As a consequence, they have reached somewhat discrepant
conclusions. Each, therefore, has been tempted to criticize
any method yielding results a little different from his own.
So far, however, the validity of factor-analysis as such has
not been seriously questioned. The non-statistical psycho-
logist, it is true, is always a little dubious of statistical
demonstrations; but no systematic refutation of the pro-