SECTION J.—PSYCHOLOGY. 
SOME ISSUES IN THE THEORY OF “G” 
(INCLUDING THE LAW OF DIMINISHING RETURNS). 
ADDRESS BY 
Proressor C. SPEARMAN, Pu.D., F.B.S., 
PRESIDENT OF THE SECTION. 
I. Theory of ‘g.’ 
Tue following communication treats of certain points in a theory which 
has become known as that of Two Factors or of g. At the present 
time this theory has undergone very elaborate development. The 
mental testing from which it originated lay at first as a foreign intruder 
in the midst of general psychology. Its opponents—and these were not 
few—regarded it as an excrescence that should be forthwith cast out ; 
and even its best friends wondered how the general psychology was ever 
going to assimilate it. But, seemingly, neither of these solutions is 
happening to any great extent. The mental testing has waxed larger 
and established itself more firmly than ever without much assimilation 
with the current general doctrines ; indeed, it seems more likely, cuckoo- 
wise, to eject them from the psychological nest. In particular, the theory 
of g, which arose from the mental tests, has now managed to spread 
itself over the whole of the cognitive side of psychology, and not impossibly 
it will soon extend its scope over into the supplementary or orectic side. 
For the present I do not propose to try your patience by depicting 
the whole elaborate theory of g even in outline. Such an attempt is 
reserved for a work that will appear shortly. But a very few words may 
be allowed here to indicate its essential foundation as unwaveringly 
preserved from the very beginning. This consists in the theorem, that 
the measure of every different ability of any person can be resolved into two 
factors, of which the one is always the same, but the other always independent. 
Suppose, for instance, that any person undergoes a mental test and 
obtains seventeen marks forit. The theory asserts that this can actually be 
divided into two parts, say eleven and six, such that (on reduction 
to comparable units) the eleven re-occurs for this person in every other 
test however widely different, whereas the six is each time replaced 
by some other number independently. 
The establishment of this doctrine falls into three distinct phases. 
The first is to ascertain what are the conditions under which the measures 
of any ability admit of such division into two factors. We may note 
that this phase has often been erroneously called an assumption or 
hypothesis. It is really nothing of the kind, but simply a mathematical 
demonstration. Given the said conditions, then the divisibility into such 
two factors must necessarily occur, just as, given that a triangle has all 
ee ae 
