3i8 THE FACTORS OF THE MIND

trace of the verbal element, not only from the material or
instructions, but even from the inner mental processes of
the testees ?

(ii) Similarly, it may be doubted whether we are justified
in disallowing every negative correlation, however small the
probable error may be. After all, why must we so rigidly
assume that ' the primary factors only act positively unless
they are absent' ? ([84], p. 71). Why should we deny
that a group-factor may sometimes operate as an inter-
ference-factor ? Is it not possible that my life-long reliance
on verbal methods may actually handicap me when I deal
with tests of a visuo-kinaesthetic type ? Conversely, is it
not possible that an interest in problems of a performance
type might actually militate against an interest in numerical
problems ? In short, as I tried to show in my discussion of
* negative correlations between special educational abilities'
([SSl P- 58)j a psychology of intellectual abilities that is not
too intefiectualistic is bound to recognize that negative
values among the residual correlations may be truly * signi-
ficant,3 in every sense of the word.

Now, in permitting the group-factors to overlap, we shall
almost inevitably be reintroducing some small degree of
correlation between their saturations, and even between the
factor-measurements. Our analysis thus degenerates into
a c fractionating' type, and our selective operators are no
longer ' pure.3 x For practical purposes the oblique com-
ponents thus obtained may perhaps be of more significance
than the orthogonal; the former are relatively concrete,
the latter highly abstract. From a theoretical standpoint,
however, it might be argued that we ought rather to modify
our general factor so as to keep the secondary factors un-
correlated. And if we are to preserve the original specifica-
tion of the factors so far as possible, then presumably we
shall modify both, so that in the end everything will be as
little changed as possible. In any case, the final outcome
will be a more or less arbitrary rotation, chosen chiefly
because it suits our psychological preconceptions. As an
exploratory procedure, there can be no harm in such rota-
tions : but the mere fact that they lead to an elegant and
1 See above, p. 266.