The Evolution of Mental Metaphors in Psychology: A 90-Year Retrospective

It seems plausible that the conception of the mind has evolved over the first hundred years ofpsychology in America. In this research, we studied this evolution by tracing changes in the kinds of metaphors used by psychologists to describe mental phenomena . A corpus ofmetaphors from 1894 to the present was collected and examined. The corpus consisted of all metaphors for mental phenomena used in the first issue of Psychological Review in each decade, beginning with the inception of the journal in 1894 and continuing with 1905, 1915, and so on through 1975. These nine issues yielded 265 mental metaphors, which were categorized according to the type of analogical domain from which the comparison was drawn . The chief finding was that the nature of the mental metaphors. changed over time. Spatial metaphors and animate-being metaphors predominated in the early stages, then declined in favor of systems metaphors, often taken from mathematics and the physical sciences. A secondary finding was that the numbers of mental metaphors varied . Metaphors for mental phenomena were more prevalent in the early and late stages of the corpus than in the middle stages (1935 to 1955) . These patterns are interpreted in terms of conceptual evolution in psychologists' models of the mind. In this article we examine historical changes in the metaphors used by American psychologists to describe mental processes. Our aim is to use changes in metaphoric language to trace changes in the models of the mind that psychologists have held . It is by now accepted that researchers bring to their field of study a theoretical framework-which may be more or less explicitly conscious-in terms of which they construe the phenomena they observe (Koestler, 1964). Moreover, these frameworks change over history, sometimes quite rapidly (Kuhn, 1962) . Cognitive psychology during the past hundred years seems a prime example of a field in which conceptual change has been rapid and extensive. The turn-ofFebruary 1985 • American Psychologist V SO, N 1925 b t92 me~p ~ ~~, Inc. 0003-066X/85/$00.75 Bolt, Beranek and Newman, Inc . Cambridge, Massachusetts MRC Applied Psychology Unit Cambridge, England the-century dialectic between structuralism and . functionalism was followed by the schools of Gestalt psychology, behaviorism, and still later by information processing . Some of these changes are recorded in major position papers or books, such as Titchener's (1915) primer. Yet such position statements are not always available, and when we do find them they may constitute later codifications of principles rather than reflections of the actual development of the ideas. It would be useful to have a more immediate method for tracing changes in the zeitgeist . The kinds of metaphoric language used in articles through . this period of history in psychology may provide just such a measure . To infer underlying models from the metaphors used in writing about psychology requires making the assumption that metaphorical language reflects underlying metaphoric thought, that is, that it reflects a genuine mapping of significant relationships from the analogical domain to the target domain (the domain to be explained ; e.g., Lakoff & Johnson, 1980; Reddy, 1979). Certainly some caution is warranted here. The mere existence of metaphorical language cannot be taken to indicate an underlying conceptual model corresponding to the metaphor. Nevertheless, studies of the history of science have produced persuasive evidence that scientists use metaphor in the invention and organization of ideas (e .g ., Darden, 1980; Hesse, 1966; Hoffman, 1980;, Hofstaedter, 1981 ; Koestler, 1964). More importantly, the firsthand introspective reports of working scientists also stress the importance of metaphor in their creative thinking (e .g., Glashow, 1980; Kepler, 1620/ 1969; Oppenheimer, 1956) . Further; there are experimental demonstrations that the inferences people draw in problem-solving tasks are affected by the analogies they bring to the domain (Gentner & Gentner, 1983; Gick & Holyoak, 1980) . More specifically, it has been argued persuasively that metaphors from other domains have played a role in the shaping of psychological theory. Roediger (1979, 1980) noted several distinct metaphors for human memory, such as Freud's (1952) house model