Introspection

   1) Experimental Introspection Is the One Reliable Method of Knowing Ourselves

   When we are trying to understand the mental processes of a child or a dog or an insect as shown by conduct and action, the outward signs of mental processes, . . . we must always fall back upon experimental introspection . . . [;] we cannot imagine processes in another mind that we do not find in our own. Experimental introspection is thus our one reliable method of knowing ourselves; it is the sole gateway to psychology. (Titchener, 1914, p. 32)

   2) The Limitation of Introspection

   There is a somewhat misleading point of view that one's own experience provides a sufficient understanding of mental life for scientific purposes. Indeed, early in the history of experimental psychology, the main method for studying cognition was introspection. By observing one's own mind, the argument went, one could say how one carried out cognitive activities. . . .

   Yet introspection failed to be a good technique for the elucidation of mental processes in general. There are two simple reasons for this. First, so many things which we can do seem to be quite unrelated to conscious experience. Someone asks you your name. You do not know how you retrieve it, yet obviously there is some process by which the retrieval occurs. In the same way, when someone speaks to you, you understand what they say, but you do not know how you came to understand. Yet somehow processes take place in which words are picked out from the jumble of sound waves which reach your ears, in-built knowledge of syntax and semantics gives it meaning, and the significance of the message comes to be appreciated. Clearly, introspection is not of much use here, but it is undeniable that understanding language is as much a part of mental life as is thinking.

   As if these arguments were not enough, it is also the case that introspective data are notoriously difficult to evaluate. Because it is private to the experiencer, and experience may be difficult to convey in words to somebody else. Many early introspective protocols were very confusing to read and, even worse, the kinds of introspection reported tended to conform to the theoretical categories used in different laboratories. Clearly, what was needed was both a change in experimental method and a different (non-subjective) theoretical framework to describe mental life. (Sanford, 1987, pp. 2-3)