90 THE FACTORS OF THE MIND

their amounts can in theory be translated into terms of
mass, length, and time. Intellectual output we can express
as the production of such and such a change against such
and such a resistance within such and such a time : but the
changes and the resistances can no longer be literally
described in terms of length, mass, and time ; they can only
be figuratively represented in terms of such units,

This will become clear if we ask what further assumptions would
have to be made in order that the amounts of mental work accom-
plished in performing a series of tests could be equated with energy
in the literal sense. Can we, for example, treat the marks as general-
ized co-ordinates representing the results of the generalized forces
acting on the system in question, namely, the testee 1l Let us
extend the principle of the self-registering barometer, and imagine
a mechanism working somewhat on the following plan. To repre-
sent the n tests let there be n sliding pieces, moving independently
along fixed bars that carry a scale. The mass of each piece (m)
must be made proportional to the difficulty of the test it represents.
Each piece as it moves will register the state of the changing mental
system in such a way that the distance (d) of its outer end from
zero on the scale will be proportional to the mark obtained in the
corresponding tests. The mental work done in all the tests may
then be described in terms of the forces required to move the
sliding pieces.

1 It will be seen that my development of the underlying analogy is based
on the suggestion that each test performance can, as a first step, be given the
dimensions of a velocity or rate—i.e. be measured by extent of change
or amount of output per unit of time—and that each test is then to be
weighted according to its difficulty. We might also attempt to give it the
dimensions of work. But we should then be confronted with the fact that, in
mental processes, the velocities apparently vanish with the driving forces, i.e.
when equilibrium is restored. Anything comparable to the inertia of masses
(which brings acceleration into such prominence in dynamics) apparently
plays but a minor part in prolonged mental processes: each isolated problem
no doubt requires its own little spurt, or rather starts its own oscillating waves
of attentive effort; but, as a first approximation, we have to assume that the
smoothed * curve of work' is flat. Thus the cyclic modifications which
constitute a mental process are perhaps analogous, not so much to the changes
considered in dynamics as to those considered in thermo-dynamics.

Let me add, to avoid misunderstanding, that I am here speaking of
cognitive processes only. The factor-analysis of conative processes, par-
ticularly those that arise in response to emotional or moral situations, leads
also to descriptions in terms of independent * forces.' But once again these
forces have the characteristics of relations, rather than of entities or properties