There is a controlled association test conducted like this one in free association, except that the subject is required to respond to each stimulus word by a word standing in a specified relation to it. To one series of words he must respond by saying their opposites; to another, by mentioning a part of each object named; to another series, consisting of names of countries, he must respond by naming as quickly as possible the capital of each country named; and there are many tests of this sort, each dealing with some class of relationships which, being often observed, are easily handled {382} by a person of normal intelligence. The intelligent subject makes few errors in such a test, and responds in very quick time. Indeed, the remarkable fact is that he takes less time to respond in an easy controlled association test than in the free association test; which shows that the "control" acts not simply to limit the response, but also to facilitate it.

The "control" here is often called by the name of "mental set". It is a good example of a "reaction tendency". On being told you are to give opposites, you somehow set or adjust your mental machinery for making this type of response. The mental set thus thrown into action facilitates responses of the required type, while inhibiting other responses that would readily occur in the absence of any directive tendency. If the word "good" came as a stimulus word in a free association test, it might easily arouse the responses, "good day", "good night", "good boy", "good better", and many besides, since all of these combinations have been frequently used in the past; and the balance of frequency, recency and intensity might favor any one of these responses. But when the subject is set for opposites, the balance of these factors has little force as against the mental set. The mental set for opposites favors the revival of such combinations as "new--old", "good--bad", and such others of this class as have been noted and used in the subject's past experience.

Mental set is a selective factor, a factor of advantage. It does not supersede the previously formed associations, or work independently of them, but selects from among them the one which fits the present task. Does it get in its work after recall has done its part, or before? Does it wait till recall has brought up a number of responses, and then pick out the one that fills the bill? No, it often works much too quickly for that, giving the right response instantly; and introspection is often perfectly clear that none but the right {383} response is recalled at all. The selective influence of the mental set is exerted before recall; it facilitates the right recall and inhibits recall of any but the right response.

In controlled association, as in free association, only one of the facts previously linked with the stimulus is recalled at a time; but while in free association the factors of frequency, recency and intensity of past linkage determine which of the many possible facts shall be recalled, in controlled association the additional factor of mental set is present and has a controlling influence in determining which fact shall be recalled. Thus, in an opposites test, the stimulus word "good" promptly calls up the pair "good--bad", because the mental set for opposites gives this response a great advantage over "good night" and other responses which may have a very strong linkage with the stimulus word.

The mental set is itself a response to a stimulus. It is an inner response thrown into activity by some stimulus, such as the stimulus of being asked to give the opposites of a series of words that are presently to be shown or spoken. This inner response of getting ready for the task can be introspectively observed by a person who is new to this type of test. It may take the form of mentally running over examples of opposites--or whatever kind of responses are to be called for--or it may take the form of calling up some image or diagram or gesture that symbolizes the task. A visual image of the nose on the face may serve as a symbol of the part-whole relationship, a small circle inside a larger one may symbolize the relation of an object to a class of objects, and gesturing first to the right and then to the left may symbolize the relationship of opposites. But as the subject grows accustomed to a given task, these conscious symbols fade away, and nothing remains except a general "feeling of readiness" or of "knowing what you are {384} about". The mental set remains in force, however, and is no less efficient for becoming almost unconscious.