Copyright 1992 by the American Psychological Association Inc. 0022-006X/92/S3.00 Therapists' Intentional Use of Metaphor: Memorability, Clinical Impact, and Possible Epistemic/Motivational Functions Jack Martin Program in Counselling Psychology, Faculty of Education, Simon Eraser University Burnaby, British Columbia, Canada Anne L. Cummings and Ernest T. Hallberg Counselling Psychology Research Group, Althouse College, University of Western Ontario London, Ontario, Canada The memorability, clinical impact, and possible epistemic and motivational functions of thera- pists' intentional use of therapeutic metaphor were examined in 4 dyads of experiential psychother- apy. Clients tended to recall therapists' intentional metaphors approximately two thirds of the time, especially when these metaphors were developed collaboratively and repetitively. Clients rated therapy sessions in which they recalled therapists' intentional use of metaphors as more helpful than sessions in which they recalled therapeutic events other than therapists' intentional metaphors. Four distinctive epistemic and motivational functions of therapeutic metaphor were observed. Historically, metaphor has interested psychotherapists as a means by which therapists might assist clients to access intu- itive, unconscious processes and material (e.g., Freud, 1900/ 1965; Jung, 1961). However, recent constructivist theories of metaphor in psychology, linguistics, and philosophy have em- phasized mnemonic, epistemic, and motivational functions of metaphor as a stimulus to new learning, understanding, and development (cf. Haskell, 1987; Ortony, 1979). Several contem- porary theorists of psychotherapy have suggested that these constructive functions of metaphor might be important factors in the promotion of client change during psychotherapy (cf. Muran & DiGiuseppi, 1990). However, only a small number of studies have been conducted to examine empirically such theo- retical suggestions (e.g., Angus & Rennie, 1988,1989; McMul- len, 1985). To date, there are no empirical studies of metaphor in psycho- therapy that have used any sort of experimental manipulation: (a) to test any of the various mnemonic, epistemic, and motiva- tional claims that have been made concerning the functions of metaphor in psychotherapy or (b) to relate any of these func- tions to clinical impact. The specific impetus to the research reported was a series of studies conducted recently by Martin and Stelmaczonek (1988) and Martin, Paivio, and Labadie (1990). These researchers asked clients to recall important therapeutic events immedi- ately following therapy sessions and at follow-ups of several We gratefully acknowledge the assistance of Wyn Martin, Margaret Rossiter, Karen Sochaczevski, Tali Wettstein, and Grace Woo in the conduct of the study reported, as well as the support of the Social Sciences and Humanities Research Council of Canada (Grant 410-88- 0022). Correspondence concerning this article (and requests for an ex- tended report of this study) should be addressed to Jack Martin, Pro- gram in Counselling Psychology, Faculty of Education, Simon Fraser University, Burnaby, British Columbia, Canada V5A 1S6. months after the termination of therapy. Of particular interest was the finding that certain characteristics of therapeutic dis- course distinguished therapeutic events that clients recalled as important from temporally proximate, but distinct, events that clients did not recall. One of the most salient of these discourse characteristics was elaboration, the extension of therapeutic dis- course through the use of concrete, imagery-laden metaphor and illustration. Furthermore, it was therapists', not clients', use of such figurative language during therapeutic events that pre- dicted client recall. These findings provide the most direct (al- beit, post hoc) evidence yet available attesting to the mnemonic function of metaphor in actual psychotherapeutic interactions. The current study used an experimental manipulation in which experienced therapists were coached to use metaphor intentionally in an attempt to enhance clients' recall of thera- pist-judged-important therapeutic material from actual ther- apy sessions. Three specific questions were asked. First, to what extent would clients recall therapy events associated with thera- pists' intentional use of metaphor? Second, would clients rate sessions from which they recalled events associated with thera- pists' intentional use of metaphor as more helpful and effective overall than sessions from which they recalled therapeutic events other than those associated with therapists' intentional use of metaphor? Third, in addition to the possible mnemonic effects explored in the first two questions, what kinds of episte- mic and motivational effects might be associated with thera- pists' intentional use of metaphor during psychotherapy? Method Three therapists participated in the study (2 women, 1 man). All therapists were highly experienced and identified their therapeutic orientation as person-centered, experiential. One of the female thera- pists met with 2 clients. The 4 clients were highly educated and pre- sented with identity and relationship issues. All were informed volun- teers from the regular case loads of the participating therapists. 143 144 BRIEF REPORTS After agreeing to participate in the study, all 3 therapists attended a single 3-hr training session conducted by J.M., at which (a) the study was described in detail as an investigation of clients' recall and use of therapeutic metaphors; (b) the nature of metaphor and its possible effects on clients' experiencing and awareness (within an experiential therapeutic framework) were explained, discussed, illustrated, and modeled; (c) the specific roles, tasks, and expectations of participating therapists and clients were described and discussed; and (d) therapists' concerns and questions were elaborated and resolved. Discussion during the training session revealed that all 3 therapists commonly used metaphor in their usual style of therapeutic interven- tion. For the purposes of the current study, they were asked to intro- duce and elaborate metaphors (at least one in each session of psycho- therapy, if possible) in a more purposeful manner, when they judged that such intentional use of metaphor would help promote client en- coding and recall of significant therapeutic material and would be appropriate to the current therapeutic context. Throughout the study, therapists recorded all sessions on audiotapes supplied to them. Immediately following each session, therapists and clients completed Episodic Memory Questionnaires (EMQs, see be- low) independently, in separate locations. Each completed EMQ was sealed in a separate, coded envelope by each respondent and collected later, together with the audiotape of the session, by research assistants. In all, 41 therapy sessions were conducted during the study (8,10,10, and 13 in each of the 4 dyads, respectively). The major portion of the EMQs consisted of five basic questions: (a) "Relax and think back to the session in which you just participated. . . . Try to remember some of the specific phrases or sentences that were spoken.. . . What were some of these phrases or sentences?" (b) "What was the most memorable event that occurred in this session? (Try to remember exact words, phrases, or sentences spoken during the event.)" (c) "Why do you remember this event?" (d) "What was the next most memorable event? and (e) "Why do you remember this event?" Participants then were asked to rate the helpfulness of the session on a 5-point scale, anchored by 1 (not at all helpful) and 5 (extremely help- ful). Therapists made this rating in response to the question: "How helpful was this session for your client?" Participants made one addi- tional rating of the session overall ("How would you rate this session overall?") on a 6-point scale, anchored by 1 (very poor) and 6 fyerygood). At the end of their EMQ forms, therapists were asked: "What, if any, metaphors did you attempt to initiate and/or develop intentionally in this session?" Results and Discussion The 3 therapists made intentional use of therapeutic meta- phor in 29 of the 41 sessions they conducted as part of the study. Two research assistants checked therapists' EMQ descriptions of intentional therapeutic metaphor and listened to audiotapes of therapy sessions to ensure that the metaphors identified as intentionally developed by the therapists fitted the operational definition of metaphor used in the study. In all cases, therapist- identified intentional metaphors were judged to meet the opera- tional criteria (i.e., "a metaphoric vehicle not typically asso- ciated with a particular topic is used to inform that topic"). For example, in the therapeutic metaphor, "Right now, you are be- ing driven by the red devil of perfection," current life experi- ences of a client were being informed by the vehicle, "driven by the red devil of perfection." The ground or relationship between vehicle and topic (assuming appropriate context and compre- hension) probably conveyed to the client something about the nature and severity of emotions currently being experienced and their possible origins. Interrater agreement between the two research assistants concerning the validity of therapist- identified intentional metaphors (according to our operational definition) was 100%. Although most sessions contained meta- phors other than those identified by the therapists as intention- ally developed, only therapist-identified intentional metaphors were relevant to our hypotheses. Our first question probed the extent to which clients would recall therapists' intentional use of metaphor. Clients recalled therapists' intentional use of metaphor in 19 of the 29 sessions (66%) in which therapists reported intentional use of metaphor (ranging from 55% to 100% across dyads). A client was consid- ered to have recalled a therapist's intentional use of metaphor in a particular session if the client's responses to Questions 1,2, and/or 4 on the client's EMQ for that session contained any of the exact vehicles (or obvious synonyms for these vehicles) of the intentional metaphors reported by the therapist for that same session. Independent judgments of two coders who scored the entire data set showed agreement concerning the presence or absence of client recall in 27 of the 29 possible cases (93%). The two cases of disagreement were easily resolved through brief discussion between the two coders. The percentages and frequencies of client recall of therapists' intentional use of metaphor may be inflated slightly. Therapists reported two or three instances of intentional use of metaphor in 9 of the 19 sessions in which clients recalled the metaphors reported by therapists. To count as recall, a client's EMQ re- sponses had only to contain the vehicle of one of these thera- pist-reported metaphors. However, in 3 of these 9 sessions, clients recalled all the metaphors that therapists used. In the six remaining instances of therapists' multiple use of metaphor within a single session, the audiotapes of the sessions revealed (a) that only one of the two or three metaphors the therapist reported actually was developed much beyond inclusion in a single utterance or (b) the two or three metaphors that were developed tended to merge together into a single, multi-vehicle metaphor. A second source of possible inflation in these results relates to the number of separate client responses to Questions 1, 2, and/or 4 on the clients' EMQs. The mean number of client responses (i.e., complete, unique "thought units" or "full sen- tence equivalents") to Question 1 was three. Of course, clients almost always gave a single response to each of Questions 2 and 4. Consequently, clients had an average of five chances to recall therapists' intentional metaphors in a given session. However, this source of possible inflation is not as extreme as it might seem. In only 13% of the cases of client recall of therapists' intentional metaphors did the recall arise from client responses to Question 1 on the EMQs. Audiotapes of the 19 sessions from which therapeutic meta- phors were reported by therapists and recalled by clients re- vealed that in almost all these cases (with only two exceptions) the metaphors recalled were initiated or developed through the explicit collaborative participation of both partners to the therapeutic dialogue. In some instances a single fully and col- laboratively developed metaphor actually was used in 2 or more sessions (also see Angus & Rennie, 1988,1989). However, even when developed collaboratively and repetitively across sessions, intentional therapeutic metaphors still accounted for less than 10% of the total dialogue in any given therapy session. (A complete listing of the actual therapeutic metaphors devel- oped intentionally by therapists in this study may be obtained from Jack Martin.) BRIEF REPORTS 145 The second question concerned the possible clinical impact of clients' recall of therapists' intentional use of therapeutic metaphors. Our prediction was that clients would rate sessions from which they recalled events associated with therapists' in- tentional use of therapeutic metaphor as more helpful and ef- fective overall than sessions from which they recalled events other than those associated with therapists' intentional meta- phoric interventions. We reasoned that clients recall of inten- tional therapeutic metaphors would indicate that such meta- phors had helped clients to encode and recall therapist-judged- significant therapeutic material, thereby permitting clients to benefit more from therapy sessions in which such encoding and recall occurred. Small samples and occasional lack of variability in client ratings within dyads did not permit us to test this prediction statistically within individual dyads or to include a within-dyad factor in our analyses. We therefore used a t test (for a difference between two independent means) to compare the mean client rating of session helpfulness ("How helpful was this session?") for all 19 sessions from which clients recalled therapists' inten- tional metaphors with the mean client rating of session helpful- ness for the 10 sessions from which clients recalled therapeutic events other than therapists' intentional metaphors. We thus considered therapy session as the unit of analysis for the pur- poses of this comparison and considered the two types of ther- apy sessions being compared as independent types. Of course, individual sessions from the same therapy dyad are not really independent of each other. However, we had unequal N & and interval data, and approximately the same number of therapy sessions from each of the 4 dyads contributed to both the sam- ples of sessions being compared. The result of this comparison was statistically reliable, t(27) = 3.45, one-tailed p < .001. The mean helpfulness rating that clients gave to sessions for which they recalled therapists' intentional use of metaphor was 4.89 (SD = .31), compared with a mean rating of 4.20 (SD = .98) given to sessions from which events other than therapists' intentional use of metaphor were recalled. A similar test conducted on clients' overall rat- ings of sessions ("How would you rate this session overall?") revealed no statistically reliable difference between these two types of sessions, t(27) = 1.42, one-tailed p = .07. Nonetheless, even with our small sample size and low statistical power, this latter test approached statistical significance in the direction hypothesized. Therapist ratings were not analyzed in the ab- sence of specific hypotheses concerning them. Our final question concerned possible epistemic and moti- vational effects of therapists' intentional use of therapeutic met- aphor. Four effects were identified (two epistemic and two mo- tivational) from participants' responses to Questions 3 and 5 on the EMQs ("Why do you recall this event?) for those sessions from which clients recalled therapists' intentional metaphors. The two epistemic functions identified were (a) enhanced emo- tional awareness and understanding, and (b) conceptual "bridging." Both these functions may relate to the potential of metaphoric communications to capture verbally significant ele- ments of affective, experiential content inaccessible to more literal communications (cf. Haskell, 1987; Ortony, 1979). The two motivational functions identified were (a) enhanced rela- tionship with therapist, and (b) goal clarification. Eighty-six percent of participant responses to EMQ Questions 3 and 5 were judged by J.M. and a research assistant (working in collab- oration) to fall within one or more of the foregoing four catego- ries. We interpret our results to suggest that therapists can make intentional use of metaphor during psychotherapy to enhance clients' encoding and recall of therapist-judged-significant ther- apeutic events and that such encoding and recall may be asso- ciated with clients' evaluations of individual therapy sessions as helpful. In addition to these mnemonic effects of intentional use of metaphor by therapists, a small number of distinctive epistemic and motivational functions of therapists' use of meta- phor were noted, particularly with respect to clients' emo- tional/conceptual understanding and the therapeutic alliance. Because of the small number of dyads, the unusually high educational level of the clients, and the single therapeutic orien- tation of participating therapists, it would be highly inappropri- ate to interpret these results as more than exploratory and sug- gestive. Also, recall data in this study are perhaps inflated by the procedures employed. At the same time, it also is important to emphasize that only 10% of the therapeutic dialogue in the sessions studied was related to therapists' intentional use of metaphor. Consequently, the fact that clients recalled a full 66% of such metaphors clearly warrants future, larger scale studies with more diverse populations and procedures. Given the extreme difficulties associated with the conduct of a nontrivial, experimental manipulation in a naturally occur- ring psychotherapeutic context, we think our initial results will be of interest to both scholars and practitioners of psychother- apy and to psycholinguists and cognitive psychologists inter- ested in ecologically valid experimentation. References Angus, L. E., & Rennie, D. L. (1988). Therapist participation in meta- phor generation: Collaborative and noncollaborative style. Psycho- therapy, 25, 552-560. Angus, L. E., & Rennie, D. L. (1989). Envisioning the representational world: The client's experience of metaphoric expression in psycho- therapy. Psychotherapy, 26, 372-379. Freud, S. (1965). The interpretation of dreams. New York: Avon. (Origi- nal work published 1900) Haskell, R. E. (Ed.). (1987). Cognition and symbolic structures: The psychology of metaphoric transformation. Norwood, NJ: Ablex. Jung, C. (1961). Memories, dreams, reflections. New York: Vintage Books. Martin, J., Paivio, S., & Labadie, D. (1990). Memory-enhancing charac- teristics of client-recalled important events in cognitive and experi- ential therapy. Counseling Psychology Quarterly, 3, 239-256. Martin, J., & Stelmaczonek, K. (1988). Participants' identification and recall of important events in counseling. Journal of Counseling Psy- chology, 35, 385-390. McMullen, L. M. (1985). Methods for studying the use of novel figura- tive language in psychotherapy. Psychotherapy, 22, 610-619. Muran, J. C, & DiGiuseppi, R. A. (1990). Towards a cognitive formula- tion of metaphor use in psychotherapy. Clinical Psychology Review, 10, 69-85. Ortony, A. (Ed.). (1979). Metaphor and thought. New York: Cambridge University Press. Received February 13,1991 Revision received June 25,1991 Accepted July 2,1991
Somatoform and Other Psychosomatic Disorders: A Dialogue Between Contemporary Psychodynamic Psychotherapy and Cognitive Behavioral Therapy Perspectives