Go to Semiotic Review of Books Home Page
Go to SRB Highlights
Go to SRB Archives


SRB Archives

This review appeared in Volume 1 (3) of The Semiotic Review of Books.

Focusing on Metaphors

by John M. Kennedy

Metaphor and Symbolic Activity. A quarterly journal edited by Howard R. Pollio (U. of Tennessee) and published by Lawrence Erlbaum Associates Inc., 365 Broadway, Hillsdale, N.J., 07642. (ISBN 0-885-7253)

The journal Metaphor and Symbolic Activity is a significant recent addition to psychology and semiotics. Founded in 1986 and now beginning its fifth volume, it has passed the two-year milestone where many small journals stumble and fall. It has offered enough scholarly substance to have a presence that can be reviewed: it has a past, a present and I contend, despite some important, principled arguments, a future.

Metaphor and Symbolic Activity is a regular quarterly. It is an outgrowth of an "underground" publication, Metaphor Research Newsletter which was produced on mimeograph and photocopy machines for about five years, largely at Adelphi University, Long Island, by Robert R. Hoffman, a psychologist. The Newsletter was a quirky, informal publication with a sense of humour. Its popularity showed there might be a niche in academic psychology for a journal dedicated to studies of metaphor. The 1970s saw psychologists advance from studies of grammar and word use in a literal manner to problems of imagery of many kinds. Work on computer models of perception, action, logic and linguistic translation circled around this domain. The need for serious examination of the issues was evident to many.

A. Ortony and S. Vosniadou of lllinois, D.S. Palermo of Pennsylvania State, R. Honeck of Cincinnati, many of the Project Zero group at Harvard (including H. Gardner, E. Winner and D. Wolf), D. MacKay of UCLA and other notable psychologists were drawn into a field where psychologists had been largely absent. The main historical texts of experimental psychology (e.g. Boring, 1950) contain no references to metaphor whatsoever, but in the 1970's, attention began to be paid. The usual topic was verbal metaphor of the copula type X is a Y, though irony, personification and other problems were a respectable second. In the study of tropes in pictures, hyperbole in caricatures was virtually the only interest until the end of the 1970s for experimental psychologists. Then some of us brought oxymorons, prolepsis, metonymy, syllepsis and catachresis into the psychological literature on pictures and studies of metaphor (Kennedy, 1980; Kogan and Chadrow, 1986; Newton, 1985). The correspondence in the Newsletter on these topics ranged fairly wide, but it was largely the product of psychology departments in North America. It indicated how psychology could apply itself to metaphor. It was a start. The content of Metaphor and Symbolic Activity follows the topics evident in the Newsletter. The tone is now formal, thorough and correct, problems are tackled at length, and it is easier to see where dismaying flaws might be thought to lie in the entire enterprise.

The editor of the quarterly is Howard Pollio, a psychologist at the University of Tennessee, and the associate editors are a colleague, Michael K. Smith, and Hoffman. Pollio's influence on the journal's concept is flagged in its name. Pollio (1974) wrote The Psychology of Symbolic Activity a decade and a half ago and the ghostly legacy of a positivist approach to human cognition can be discerned behind the title. Psychology went through a horrendous phase in which operationism and behaviorism dictated what lay within the bounds of enquiry and what was said to be arrant spiritualist and teleological nonsense. I recall one title in the Psychological Bulletin in the mid-1960s which revealed the paranoia of the times; it was entitled "Nightmare Behaviour" as though nightmares, as private mental events, were unsuitable topics for study until the magical positivist tincture "behaviour" was added and all grew bright and clean. Similarly, Pollio argues that the phrase "symbolic activity" helps unenlightened scholars to realize the territory that is covered by the topic of figurative language (vide the Editorial by Pollio, Hoffman and Smith (1986) in the first issue).

My own view is that while the title assists the unwary it disappoints the initiate: it reveals a history of defenses against unsophisticated operationism. Operationism was a dead issue in philosophy within five years of its origins. It argued that science was only the result of its operations. Its fatal flaw was that it had no way of saying which operational measures were truly independent and which were measuring the same thing. It needed criteria that guided the use of operations and those criteria were at least as much the basis of science as the operations. But the weight of this dead hand swayed psychology for decades, and still does in some circles. It made out that the discussion of mental events and imagery was foolish prattle.

The quarterly's content reveals our intellectual times just as much as the title. There is a definite emphasis on language as the heart of metaphor. The pages offer continued wide open debate on the relationship between language and metaphor, and serious reservations about psychological scientific method applied to the study of metaphor (Honeck, 1986; Gruber, 1988).

Even the initial editorial slips unguardedly into language-centred ideas on metaphor. Noting that metaphor is a metaphoric term for tropes of all kinds, it avows that the target of the journal is figurative language. In fast, metaphors can appear in any realm of communication. Mime, emblems, logos and pictures are full of metaphors. In principle, every kind of trope can be found in any representational medium. Semiotics, points the way out of language centred chauvinism-- in this regard.

The editors of the journal are wiser than their early statement of their mission. The quarterly has welcomed articles on metaphors in the deaf and the blind, metaphors from children stimulated by small models, and pictorial metaphors in art and nineteenth-century book illustration. Some of the contributors however contend strongly that language is still the sine qua non of metaphor. The most extreme version of this view is that metaphoric expression must ultimately be translated into language before it can truly be said to be metaphor. This view creates important difficulties for the study of tropes. Ultimately, it can lead to the conclusion that there is no metaphoric thought since thought is independent of language, language being thought's vehicle. If metaphor is just a garnish to language, it is no more important than rhyme. It is just the fuzz on the peach.

The past five years of the journal have seen the discussion of language's relation to metaphor swing to and fro. Some authors support the idea that metaphor is a decoration on language. Others follow George Lakoff, arguing metaphor is so pervasive in language that it offers the major connections between everyday ideas: metaphors, it is said, are what we are governed by. We think guided by metaphors in which for example, at least in some cultures, the moon is female and the sun is male. Some contend personification and human role-playing is the prototypical metaphor, not only the most frequent, but the basis for all the others.

There is a problem bedeviling all claims about the relationships between language and metaphor and it surfaces regularly in Metaphor and Symbolic Activity quite explicitly. It is even more basic than the question of the priority of thought or language. Some authors contend there is no principled distinction whatsoever between metaphor and ordinary use of language. David Olson (1988), for example, points out that language by its very nature applies to new instances. The broom standing in the corner may be a new use of stand but this could be meant quite literally by someone who took being erect to be the basis for standing.

To criticize the inability of the contributors to agree on the definition of their subject matter is not to condemn utterly the area of research. No one would reject Nature because its contributors had contradictory ideas about what is natural or living. Psychological Review is the champion of psychology journals, but its pages are full of conflict on what constitutes proper psychology. However, there are serious concerns about the continued existence of psychology today; as an independent discipline, now that the Maginot line of operationism has been seen to be largely irrelevant. In good measure perception psychology has fallen into the capable hands of robotics researchers and physiological specialists. Philosophy has reclaimed much of cognition and studies of classification. Communication is better investigated by semiotics and rhetoric than by psychology. Linguistics theory is no longer within psychology to any degree. Psychology is losing out to its companion disciplines. What does this portend for metaphor studies and this journal?

Metaphor and Symbolic Activity is largely written by psychologists; the methods of experimental psychology are everywhere in evidence in its pages. Serious attempts are made to standardize material, to be explicit about the exemplars being studied and how they have been culled from some corpus. There are two articles that underline this message especially clearly. One, by Richard P. Honeck, describes the many pitfalls confronting experimental psychologists when they study metaphor (Honeck, 1986). He points out that at present the materials used are usually unique to each study. The criteria used to select the materials are in his term "nonreplicable"; they become a bone of contention between different investigators. Honeck himself admits that his earlier studies used as examples of proverbs instances he would no longer accept. There is a great deal of room for guess-work.

In response, Katz, Paivio, Marschark and Clark (1988) present as standardized materials a set of 464 metaphors rated on 10 psychological dimensions. The set contains 204 metaphors culled from works of poetry and 260 generated by the experimenters. The ten dimensions include scales of comprehensibility, imagery, goodness of the metaphor and the number of alternative interpretations. These may prove extremely useful to investigators concerned that their interpretations shift with quite superficial changes in the material being used.

Notice that no literary journal on metaphor would ever assert a need for standardized materials, nor would it ever deem it worthwhile to establish that a group of students found "the bird is nature's airplane" to be high in comprehensibility. These are, however, quite useful tasks for psychologists. What can they make of these tasks? One illustrative kind of discussion that continues apace is the matter of reaction time and mental processes.

If metaphors are indeed different from standard language of a literal kind, and if we must take novel metaphors as violations of standard language meanings, then surely language users need something to alert them to take the language as metaphoric. The implication is that the utterance must be analyzed first for its literal meaning and then, if that is rejected, this is a cue to begin searching for a possible metaphoric meaning. When we hear "The brain soaks up information", this theory of analysis holds: we first reject the literal claim that the brain can act like a sponge, then we accept the metaphoric version where the message is that the brain acquires information swiftly, as rapidly and naturally as a sponge absorbs and holds liquid. This theory of orderly analysis, proceeding from literal to metaphoric bases, can be tested with methods well worked-out in experimental psychology (Hoffman and Kemper, 1987). Subjects can be asked to respond true or false when a sentence is presented on a screen and reaction times to literal interpretations of sentences can be shown to be much faster than metaphoric interpretations of the same sentences. If the sentences are idioms, then literal and metaphoric interpretations should be equally fast. If the sentences are dead metaphors, then literal interpretations should be extremely slow. The literal interpretation of an idiom like "He fell in love" is almost inaccessible in normal conversation. If we asked the speaker "did he break his leg?" the most likely initial response is befuddlement.

In sum, experimental psychology can bring a saddle-bag full of methods to the study of metaphor, there are sensible questions about metaphor where the methods are godsend, and standardized and often rather hum-drum materials are useful in this regard. The psychology of metaphor advances apace and this quarterly is a testament to the current interest in assessing how metaphors are understood.

If Metaphor and Symbolic Activity were restricted to psychological reaction-time studies I think it would be very thin indeed. These studies can only be undertaken usefully when the theories and experiences being assessed are carefully worked out (Doyle, 1988) and when distinctions between different kinds of metaphors are reasonably effective and clear. Has "The brain sponges up information" become an idiom? Is it virtually a dead metaphor? The prediction being used in the study rests on this prior classification. It is vitally necessary for psychology to be open to the philosophy and rhetoric of communication studies of many kinds before applying the tools it has developed. Psychology is inherently a second-order discipline: it studies the "psychology of ---". The hyphen precedes the psychological theory. Sensibly, the editors of the quarterly deliberately take this broader view of their mandate; they are open to assessments of the real uses of metaphors.

In their first issue, Barbara Tomlinson (1986) studied how writers described their craft in metaphors. For Philip Roth writing could follow something "simmering", for lonescu images "surface", for John Brunner piots "grow", and for John Updike "you're spinning out the reel". This kind of article surveys metaphors describing a skill. The quarterly's contributors also contend with metaphors used by scientists as the basis for disciplined enquiry, using the known to give shape to the unknown. This kind of article deals with metaphors at work. The raison d'ëtre of metaphors is suggested by articles like these and they provide a motivation for attempting the kind of detailed somewhat abstracted scouting that laboratory psychologists ask undergraduates to endure. I am struck by a line of discussion that is now quite active and reflected in resent issues of Metaphor and Symbolic Activity: McMullen (1989) points out that metaphor often plays a role in psychotherapy. She argues that successful therapy consistently shows the elaboration of themes in "bursts" of figurative language or the gradual emergence of a metaphor, a central metaphor giving rise to related, subsidiary figures, and to the use of metaphor to describe the positive changes that occur during therapy -- untying what was knotted, cleaning out what was rotten, getting in touch with what was disconnected, keeping above minor problems and irrations, finding the lost, getting warm, releasing.

The role of metaphor in literature is a longstanding and steady focus of interest, but a narrow one compared to the breadth of the everyday work of metaphors. What Metaphor and Symbolic Activity is doing for psychology is to offer a platform for the examination of metaphor in many corners of the human marketplace. The arts, the sciences, and the therapeutic professions seem to me to be an excellent start. There is good reason to suppose that profit will come from studies of metaphor in each of the main frames for human intelligence: the arts, the sciences, politics (Howe, 1988), finance, philosophy, mathematics, domesticity, education, agriculture and industry. For example, articles about economics may be full of metaphor. Certainly in practice journalism about business is. Salesmen, advertisers and hucksters most definitely are fountains of tropes. As the most fundamental challenge for metaphor studies is to distinguish what is literal from what is metaphoric, so too the challenge for the law studying advertisements is to distinguish the truth from acceptable metaphoric embellishment by means of hyperbole (applied to one's own products), irony (applied to competitors) and emblems (applied to associate your product with wonders like the Wild West or sporting idols).

Alongside George Lakoff's belief that ordinary life is shot through with influential metaphors, I think we can legitimately envisage metaphors that guide the serious student of science, the dedicated educator, the active businessman, the archaeologist (Larsen, 1987) and the artist. Perhaps too, there are major metaphors guiding many philosophers -- most especially philosophers of mind (Kearns, 1987) who envisage clockwork, statues worked by water power, "levels" of consciousness and, today, "the language of thought". I wonder though whether mathematics have been shorn of almost all metaphors. By definition, when a matter can be expressed in an equation, we deem that to be literal description. A scientific metaphor becomes a mathematical model or diagram once all the key parameters are specified.

One more stream of activity in Metaphor and Symbolic Activity deserves mention. From the very first issue, and regularly every year, the periodical has given important space to studies on children and metaphor. The contrast of the conceptions guiding this research is as extreme as in any domain of enquiry into metaphor and the empirical work as provocative. One view is that the child's mind makes metaphoric connections freely: the lion is thought to be male and deer is female, to follow a stereotype common in childhood. For some theorists, childhood metaphors are rooted in perceptual affinities that are mental modules independent of general semantics. A sharp figure is better labelled with the meaningless work TAKETE than by MALUMA, another meaningless word, in this view. Some theorists even entertain quasireligious, Jungian, anthropological notions about the mind of the child despite all the obvious difficulties in making this kind of interpretation stick in the face of alternatives. For other, the distinction between metaphor and literal arises only when the child deems some uses of a communication to be standard. Metaphor, for these theorists, follows the child's understanding of constraints, not the rather sloppy inability to notice distinctions in the first place. In my opinion, one of the most interesting contributions in this vein has been by Cathy Dent (1987). She argues for developmental studies of perception and metaphor, following Kelly and Keil (1987) who offer a theory of the kinds of concepts children can use to make metaphors at different ages. Dent contends many "natural kinds" can be distinguished by perception, and artful perceptual violations of natural kinds (like a toy deer wearing a tutu) can lead children as young as 4 to produce metaphors like "a deer is a dancer!".

Articles in Metaphor and Symbolic Activity often assess metaphors in ways that make tropes out to be minor grace notes, passing fancies. If this holds up, psychological inquiry into metaphor is an enterprise under threat. Some of the problems are fairly straightforward, to do with method, and disagreements about what constitutes a trope and what defines particular kinds of figurative communication. These are the kinds of problems of definition, however, faced by every scholarly enterprise. Some of the problems are those of intellectual isolation. For decades psychologists were safe behind the walls of positivism and rejected the bulk of Gestalt theory, semiotics, virtually all of structuralism and all deconstrustionist attempts to introduce new views of the relation between text, meaning and context. Articles in Metaphor and Symbolic Activity occasionally allude to the fast that its enterprise is not yet much more than contemporary cognitive psychology and its allied disciplines. But I imagine it is not much threatened by their rival schools without, compared to what may be more serious- the complications of the debate within its covers. Is metaphor no more than decorative language? Is it trivial? Is all language always metaphoric?

I personally feel a strong tie to Metaphor and Symbolic Activity, partly because I have helped it as a reviewer and in the editorial process, but also because of the issues it discusses. I think we can safely reject the view that metaphors are just gloss. I also reject the following kindred belittling arguments. All language involves (dead) metaphors, and so (novel) metaphors cannot be distinguished from the rest of language. Remove the bracketed terms and there would be a problem. Retain them and the distinction is saved. Metaphor is not primarily decorative language. First, metaphor is common in every medium of communication. A political cartoon showing a political leader on the hot seat is a metaphor. Recently (June 1990) a blind woman drew for me her conception of marriage: a corral with a padlock on the inside. This is a pictorial metaphor that came untutored (and unrequested), from a person quite inexperienced with pictures. It can be contended that metaphor studies are not about arcane clever decorations. Metaphor comes fairly readily to novices. It emerges from the intent to communicate and an understanding of how an item in the medium is likely to be taken by the recipient. The item has a standard use, plus it reveals a superordinate intension to provide an additional use. The extra use supplements, but does not replace, the standard use. Metaphor is a special-purpose communication in that it does not correct standard use, i.e. it does not offer a new, corrected standard. Everyday psychology has a safe place in metaphor study because ordinary people find metaphors useful.

Likewise, I think metaphor involves cognition and psychology has a role to play in metaphor, since it is not just language. We can, l suggest, think metaphorically. The importance of metaphor lies in practical everyday uses of metaphor to think. Hyperbolic claims that all of thinking is guided by metaphors we live by may be misguided, and anyway they are incidental to my point here. Metaphor is, in my view, not entirely guided in early life by personification. The child's mind has far more innate ideas than just notions of persons and social relations. Metaphor is a way of asking people to realize what is important for particular purposes. "Surgeons are butchers" invites us to make a class-inclusion of the form "all surgeons are butchers, even if not all butchers are surgeons". The reason is that our purpose for the moment is to draw attention, for example, to the surgeon's indifference to their patients as people, sentient, and capable of rational sensitivity. For the typical surgeon, the metaphor argues, your feelings are unimportant. All metaphors change a category we take to be standard, while asking for the classicisation to be pro tem. Since metaphors deal with categories, they are not like rhymes or other decorations in language. By being about categories, they can be expressed via language, but also they belong in all the routes to our understanding of kinds. By being pro tem, they are relevant to a special occasion and the purpose central to the occasion. By drawing attention to features relevant for the moment, they bring ideas forward that would otherwise be neglected. They change one item from its primary category to another category. They force the item to change its relevance to the superordinate ideas considerably, while having a smaller but notable effect on the superinordinate category being applied: "butchers" changes a bit if surgeons belong under this category, much as Picasso's reputation would change somewhat if we learned he had painted all of Braque's most regarded works in addition to all of the corpus for which he is now known.

The current wave of interest in theory and experiment on metaphor offers continual challenges to each of my contentions. If the debates come down on the wrong side, Metaphor and Symbolic Activity will be lightweight and its future will be in jeopardy. At present, there are however many questions about definitions, development, language and thought still with committed proponents on every side. The debates have come on strong since Miall (1982), reviewed by Nuessel (1986), Honeck and Hoffman (1980) and Ortony (1979), edited volumes of papers which I think caught the first serious surge of interest that combined psychology and philosophy in this area. The field of honour is now Metaphor and Symbolic Activity, although psychology articles on related subjects occasionally appear elsewhere, in such journals as Perception, Child Development, the Journal of Verbal Learning and Verbal Behaviour, Memory and Cognition, Cognitive Psychology, Ecological Psychology and Journal of Psycholinguistic Research. I confess, when the quarterly began, l wondered whether there were enough of us to make a go of it. Five years later, it seems there are, despite the arguments that may undermine the whole works. The quarterly's pages juggle ideas and methods provocatively. The most unkind critic would grant that while the issues are not showing signs of being settled, the dangers -- that the issues are trivial and the methods too weak to bring evidence to bear on worthwhile issues- are being discussed in interesting ways. Despite the seriousness of the charge that psychological issues in metaphor studies will prove to be tiny affairs, mere fleabites in language and cognition, the journal is open to proponents of this view, and I think the outcome will be a clearer view of significant uses of metaphor.


Dent, C.H. (1987) "Developmental studies of perception and metaphor. The twain shall meet," Metaphor and Symbolic Activity, 2, 53-72.

Doyle, C.L. (1988) "Living and knowing an imaginary world: A discussion of articles by Wallace, Davis and Franklin. "Metaphor and Symbolic Activity, 3, 175-182.

Boring, E.G. (1950) A history of experimental psychology (2nd. Ed.) Appleton-Century Crofts, New York.

Gruber, H.E. (1988) "Coping with multiplicity and ambiguity of meaning in works of art" Metaphor and Symbolic Activity, 3, 183-189.

Hoffman, R.R. and Kemper, S. (1987) "What could reaction-time studies be telling us about metaphor comprehension? "Metaphor and Symbolic Activity, 2,149-186

Honeck, R.P. (1986) "Verbal materials in research on figurative language." Metaphor and Symbolic Activity, 1, 25-42.

Honeck, R. and Hoffman, R. (Eds.)(1980) Cognition and figurative language Erlbaum, Hillsdale, N.J.

Howe, N. (1988) "Metaphor in contemporary American political discourse" Metaphor and Symbolic Activity, 3, 87-104.

Kearns, M.S. (1987) "Metaphors and the humanizing of the mind." Metaphor and Symbolic Activity, 2, 115-138.

Katz, A.N., Paivio, A.,Marschark, M. and Clark, J.M. (1988) "Norms for 204 literary and 260 nonliterary metaphors on 10 psychological dimensions" Metaphor and Symbolic Activity, 3, 191-214.

Kelly, M.H. and Keil, F.C. (1987)" Metaphor comprehension and knowledge of semantic domains" Metaphor and Symbolic Activity, 2, 33-52.

Kennedy, J.M. (1980) "Blind people recognizing and making haptic pictures" In M.A. Hagen (Ed.) The perception of pictures Vol. 2. Academic Press, New York.

Kogan, N. And Chadrow, M. (1986) "Children's comprehension of metaphor in the pictorial and verbal modality" International Journal of Behaviourial Development, 9, 285-295.

Larsen, S.F. (1987) "Remembering and the archaeology metaphor" Metaphor and Symbolic Activity, 2, 149-186.

Mackay, D.G. (1986) P"rototypicality among metaphors: On the relative frequency of personification and spatial metaphors in literature written for children versus adults" Metaphor and Symbolic Activity, 1,87-108.

McMullen, L.M.(1989) "Use of figurative language in successful and unsuccessful cases of psychotherapy: Three comparisons," Metaphor and Symbolic Activity, 4, 203-226.

Miall, D.S. (Ed.)(1982) Metaphor: Problems and Perspectives Harvester Press, Brighton, Sussex.

Newton, D.P. (1985) "Children's perception of pictorial metaphor" Educational Psychology, 5, 179-185.

Nuessel, F. (1986) "Metaphor: Problems and Perspectives" Ed. D.S.Maill. Book Review. Metaphor and Symbolic Activity, 1, 81-85.

Olson, D.R. (1988) "Or what's a metaphor?" Metaphor and Symbolic Activity, 3, 215-222.

Ortony, A. (Ed.)(1989) Metaphor and Thought Cambridge U. Press, Cambridge.

Pollio, H.R. (1974) The psychology of symbolic activity. Addison-Wesley, Reading, Massachusetts.

Pollio, H.R., Hoffman, R.R. and Smith, M.K. (1986) "Editorial" Metaphor and Symbolic Activity, 1, 1-4.

Tomlinson, B. (1986) "Cooking, mining, gardening, hunting: Metaphorical stories writers tell about their composing prowess" Metaphor and Symbolic Activity, 1, 57-80.

John M. Kennedy is Professor of Psychology at the University of Toronto, Vice-President of the Toronto Semiotic Circle, and a member of the senior editorial board of SRB. He is the author of "Tactile pictures", in M. Heller and W. Schiff (Eds.) Touch perception, in press. As an experimental psychologist, his work has been concerned with pictures chiefly, for two decades. He theorizes that outline drawings show relief but not chiaroscuro to the human visual system, innately. He argues perspective is based on an amodal sense of direction. Further, he contends metaphors in pictures can be understood without explanation by peoples unfamiliar with pictures. He has demonstrated the applicability of these ideas to tactile pictures, being tested with congenitally and early blind people. His numerous publications in this domain include "Haptic pictures" (1982), "What can we learn about pictures from the blind?' (1983) and Vision and Metaphors: Empirical Investigations (1984).

Go to Semiotic Review of Books Home Page
Go to SRB Highlights
Go to SRB Archives