ion that this is a rather artificial and unusual
type of mental performance; but in reality controlled association is a
very representative mental process, and enters very largely into all
forms of mental work. This is true in arithmetical work, for example.
A pair of numbers, such as 8 and 3, has been linked in past experience
with several responses; it means 83, it means 11, it means 5, and it
means 24. But if you are adding, it means 11, and no other response
occurs; if you are multiplying, it means 24, and only that response
occurs. The mental set for multiplying facilitates the responses of
the multiplication table and inhibits those of the addition table,
while the mental set for adding does the reverse. Rapid adding or
multiplying would be impossible without an efficient mental set. Thus
in arithmetic, as in the tests, the mental set is an inner response to
the _task_.

In reading, there is a mental set which is an inner response to the
_context_, and which determines which of the several well-known
meanings of a word shall actually be called to mind when the word is
read. Presented alone, a word may call up any of its meanings,
according to frequency, etc.; but in context it usually brings to mind
just the one meaning that fits the context. The same is true of
conversation.

The objective _situation_ arouses a mental set that controls both
thought and action. The situation of being in church, for example,
determines the meanings that are got from the words heard, and
controls the motor behavior to {385} fit the occasion. The subject,
observing the situation, adjusts himself to it, perhaps without any
conscious effort, and his adjustment facilitates appropriate mental
and motor reactions, while inhibiting others.

A _problem_ arouses a mental set directed towards solution of the
problem. A difficult problem, however, differs from a context or
familiar task or situation in this important respect, that the
appropriate response has not been previously linked with the present
stimulus, so that, in spite of ever so good a mental set, the right
response cannot immediately be recalled. One must _search_ for the
right response. Still, the mental set is useful here, in directing the
search, and keeping it from degenerating into an aimless running
hither and thither. Problem solution is so different a process from
smooth-running controlled association that it deserves separate
treatment, which will be given it a few ch