METAPHYSICAL STATUS OF FACTORS 229

temporal dependence is what he seeks to formulate when he
talks of causal factors. His motives are clear. He is think-
ing primarily about the outer world ; and it would perplex
him if he were obliged, as the scientist and philosopher are
obliged, to keep thinking also about the validity of his think-
ing. Consequently, for the logical concepts of ' ground and
consequence/ of reason and conclusion, he substitutes the
more concrete notions of 6 cause and effect.'

Nevertheless, the language of causation, particularly in
psychology, brings with it two misleading implications.
First of all, causal knowledge is supposed to be certain
knowledge: vere scire est per cans as scire.. But the very-
fact that our data, by hypothesis, supply knowledge of
antecedent conditions only—often of remotely antecedent
conditions at that—implies that our so-called * causes,'
as known, can cover only the incomplete ' ground,5 not the
sum total of necessary and sufficient conditions, as is erron-
ously assumed ; and the shortcoming proves particularly
baffling in the study of mental behaviour, since here future
possibilities—for example, what a man desires or intends to
do—are frequently required to explain his actions, and may
explain them far more clearly than present or past conditions
by themselves. It follows that all inferences from causes,
particularly where the * ground ' is highly complex, and
most of all where it is partly mental, are bound to be, not
certain, as is commonly imagined, but merely probable.
Indeed, when a psychological determinist—an upholder of
psycho-analysis or behaviourism, for example—trusts dog-
matically to the postulate of complete causal determination,
the conclusions that he proceeds to deduce are likely to have
a low rather than a high probability. And much the same
danger besets the factorist: no predictions are so con-
fidently offered by the educational or vocational psycholo-
gist as those which are based on some such factor as g or
general intelligence. Yet again and again his forecasts
are falsified. A broader clinical approach, though yielding
more tentative and provisional predictions, would have
saved him from many pronouncements, that may seem to
possess precision of a narrow mathematical sort, and yet
ultimately serve only to discredit his methods in the eyes