244 THE FACTORS OF THE MIND

sphere of work; and, however it is solved, much the same
headings, I imagine, perhaps with slightly modified labels,
will in practice still be used.

A similar but more dangerous instance of such reification is
the attribution of unconscious behaviour—i.e. of acts carried
out without the person being aware of them—to the un-
conscious * parts/ ' regions,' ' tendencies,' or c wishes ' of
the mind—in short, to * non-cognitive mental factors' still
regarded as concrete entities—a notion which seems usually
associated with a strong belief in physical and tempera-
mental ' types.' Everyone who has read the case-reports
compiled by the young student will have noted how a
smattering of psycho-analytic theory, reinterpreted, if he is
up-to-date, in quasi-factorial language, can lead to absurdly
artificial explanations couched in highly speculative terms,
and suggest a wholly misguided treatment.

There is, beyond question, no richer field awaiting the
factor-analyst than that of so-called instinctive and un-
conscious behaviour. A beginning was made many years
ago. But the few recent attempts to supplement psycho-
analysis by factor-analysis have tended to reinforce rather
than to refine the crude notions that have hitherto obtained
in regard to * Kretschmerian types,' * Freudian factors,'
and the like, in this particular field. Most medical psycho-
logists, it is true, still bluntly profess to " believe only in
clinical research, not in statistical research " (and by clinical
research they seem primarily to mean drawing conclusions
from one or two cases only instead of from a number).
Statistical investigators have lent substance to the implied
reproach by supposing that, if they take a hundred cases
instead of one, they can safely substitute rough impression-
istic assessments for data gleaned from methodical examina-
tion and prolonged case-histories. The two methods of
approach must supplement each other; either alone will
be highly equivocal. And in this branch of psychology the
revised view of * factors ' that I have advocated above would,
I am convinced, lead to more trustworthy descriptions of
the intricate influences at work, and display clear issues for
joint research.

At the same time, the attempt to apply factorial con-