§ 4. Experiment and Observation.—To experiment is to observe under conditions which we have ourselves prearranged. The prearrangement is intended to simplify the issue that is to be decided, by excluding irrelevant conditions. In this wide sense psychology has always been to some extent experimental. What is especially modern is the introduction of apparatus and of exact measurement, such as are employed by the physical sciences. Experiment may be used in connexion with any of the modes of observation which we have described. It generally involves more than one of them, and often all three. The primary question may be, what kind of object will be presented under certain assignable conditions. A simple illustration is afforded by the old Aristotelian experiment of holding an object between the second finger and the forefinger of the hand, not in their usual position, but with the second finger crossing backwards over the forefinger. Under these circumstances, there arises a perception of doubleness, so that we appear to be touching two distinct objects instead of one. Here the question is, what object do we perceive under the given conditions? Is it single or double? We may also put a question to introspection proper, and ask how far our mental attitude resembles that which exists in ordinary cases in which two objects are perceived by touch, e.g., when two opposite sides of the same finger are touched. In my own case, for instance, I find that when two opposite sides of the same finger are touched, the appearance of doubleness is more definite and unmistakable. With the crossed fingers there is a certain sense of strangeness and hesitancy which is absent in the ordinary perception of doubleness. Another case in which the primary question relates to the presented object is that of our stock example—the stereoscope. Here the conditions of perception are prearranged by means of a special apparatus, and the question is, what, under these conditions, is the nature of the object apprehended? Here, too, the introspective inquiry may be also raised, if we ask whether our apprehension of the object is direct or due to a process of inference. It is also possible to make experiments in which the primary issue is introspective. Thus, we may attempt to will something which we know to bo impossible, in order to find out whether we can do so or not. Or again, we may deliberately attempt to attend simultaneously to two disconnected objects, with the view of discovering whether attention can be so divided.

Finally, we may experiment on the connexion between a mental state and its external manifestation. In this way, it is possible to discover many subtle signs and symptoms of mental process which evade ordinary observation.

For instance, variations in the circulation of the blood, and in respiration, and in muscular power, accompanying various phases of emotion, may be accurately measured by physical apparatus. In principle, this kind of experiment often occurs in ordinary life. Whenever we say a thing or do a thing to a person, in order to see how he will take it, we are performing a psychological experiment.

It is clear that the experimental method does not disclose any essentially new source of psychological data. It is only observation under test conditions, deliberately prearranged for the purpose of settling a definite question. It is not quite accurate to define it merely as observation under test conditions. For test conditions may arise in the ordinary course of things, without any deliberate prearrangement on our part. All pathological cases come under this head. In such cases as those of Laura Bridgman or Helen Keller we have an opportunity of observing, under test conditions, what the sense of touch alone can effect, in the absence of sight, hearing, smell, and taste. But the test conditions are such as could not be prearranged by the psychologist. He is not permitted to make people blind and deaf from their birth in order to watch the consequences.

The experimental method has often great advantages; but it has also certain drawbacks. The very conditions which we wish to investigate are often such as occur only in the normal course of mental life, and are interfered with by artificial arrangements. For instance, experiments on the association of ideas labour under this defect. The question which interests us is how the succession of ideas is determined in ordinary thinking. But experiment subjects the mind to conditions which are quite remote from those of the normal flow of thought. In experiment, isolated words or other objects are successively presented to a person, and he is called on to name the first idea which each of them suggests to him. Thus, continuity of interest, which is allimportant in ordinary thinking, is excluded. Another question in which the experimental method is seriously defective is that relating to the mental imagery accompanying the use of words. When we deliberately select a word, and ask ourselves what imagery it calls up in our minds, we are by the very process of our inquiry interfering with the result. We are looking for mental imagery, and we have no right to affirm that the imagery which we find would be present if we had not been looking for it. The only safe course in such a case as this is to cultivate the habit of watchfulness, so that we may frequently catch ourselves in the act of using words in a natural manner in the ordinary course of thought. This perpetual readiness to notice what is taking place in our own minds, without deliberately resolving to do so, on this or that special occasion, is at once a most difficult and a most necessary equipment of the introspective psychologist. The special function of the experimental method has been well stated by one of its most enthusiastic advocates, Professor Titchener. "An experiment is a trial, test, or observation, carefully made under certain special conditions: the object of the conditions being (1) to render it possible for any one who will to repeat the test, in the exact manner in which it was first performed, and (2) to help the observer to rule out disturbing influences during his observation, and so to get at the desired result in a pure form. If we say precisely how we have worked, other investigators can go through the same processes, and judge whether our conclusions are right or wrong; and if we do the work in a fitting place, with, fitting instruments, without hurry or interruption, guarding against any influence which is foreign to the matter in hand, and which might conceivably alter our observation, we may be sure of obtaining 'pure' results, results which follow directly from the conditions laid down by us, and are not due to the operation of any unforeseen or unregulated causes. Experiment thus secures accuracy of observation, and the connection of every result with its own conditions; while it enables observers in all parts of the world to work together upon one and the same psychological problem."*