A cognitive linguistic view of metaphor and therapeutic discourse by Zoltán Kövecses

this is a reproduction of http://www.clas.ufl.edu/ipsa/journal/2001_kovecses01.shtml and is not original to the author of this blog.

A cognitive linguistic view of metaphor and therapeutic discourse by Zoltán Kövecses In the cognitive linguistic view, three levels of metaphor can be usefully distinguished (see K–vecses, 2002, ch. 17): (1) the "supra-individual" level, (2) the individual level, and (3) the "sub-individual" level. I suggest that each "conceptual metaphor" (as Lakoff and Johnson, 1980, call metaphors of thought, not just of language) can be analyzed on these three levels. Most of the recent research in cognitive linguistics takes place on and is directed at one or several of these levels. In this paper, I will try to characterize the three levels, point out some common misunderstandings concerning metaphor analysis, and show some of the potential of this view of metaphor for psychotherapy. My goal in this paper is not to deal with any specific issues concerning the study of metaphor in psychotherapy (such as whether conceptual metaphor can be an indicator of good-outcome vs. poor-outcome therapy, the role of conceptual metaphor in dreams, etc.). This is a topic that others already started to explore (see, e.g., Angus, 1996; McMullen and Conway, 1996; Lakoff, 1997). Rather, my main goal here is to sketch a three-level view of metaphor that is based on what is known as the cognitive linguistic view (see Lakoff, 1993) and that is capable of offering a framework within which a variety of seemingly disparate issues related to metaphor in psychotherapy can be studied systematically. go >> keywords: ____ url: http://www.clas.ufl.edu/ipsa/journal/2001_kovecses01.shtml To cite this article, use this bibliographical entry: Kövecses, Zoltán. "A cognitive linguistic view of metaphor and therapeutic discourse"PSYART: A Hyperlink Journal for the Psychological Study of the Arts. Article ___. October 1, 2001. Available HTTP: http://www.clas.ufl.edu/ipsa/journal/2001_kovecses01.shtml, Dec. 31, 2001 [or whatever date you accessed the article]. Received: January 1, 2001 || Published: April 26, 2001 || Copyright © 2001 Zoltán Kövecses author info: Zoltán Kövecses zoltan@icsi.berkeley.edu Eötvös Loránd University Department of American Studies Ajtosi Durer sor 19-21, 1146 Budapest, Hungary

A cognitive linguistic view of metaphor and therapeutic discourse Zoltán Kövecses The supra-individual level Let me begin with the supra-individual level of analyzing metaphors. What "supra-individual" simply means is that there is a level of metaphor that is based on the conventionalized linguistic metaphors of a given language (such as English, Chinese, Zulu, Wolof, Hungarian, etc.). Consider some metaphorical expressions relating to anger in English (see Kövecses, 1986, 1990; Lakoff, 1987; Lakoff and Kövecses, 1987). People talk about "boiling" and "seething" with anger, the angry person being "insane" with anger, "unleashing" your anger, having a "stormy" meeting in the office, a comment "getting the boss going," two people "snarling at each other," etc. These are English words and phrases that are conventionally available to speakers to talk about their, or somebody else's, anger. For most therapists, this is what metaphors are--words and expressions that cannot be literally true when we use them about emotions, relationships, life, death, and other abstract topics, concepts, or domains, such as anger in the example above. Indeed, this is the most widespread view of metaphor that goes back to Aristotle, and therapists and analysts of various persuasions have made abundant use of this view for both theoretical and practical purposes. But the cognitive linguistic view goes way beyond the time-honored traditional conception of metaphor. Cognitive linguistic research begins where the traditional view ends, that is, with identifying conventionalized metaphorical linguistic expressions. Researchers within the cognitive linguistic paradigm typically collect conventionalized metaphorical expressions from dictionaries, thesauri, random other sources such as books, newspapers, magazines, and other news reports in the media, or from their own "mental lexicon" as native speakers of a language. They then analyze these collections of conventionalized metaphorical expressions by grouping them into what came to be called "conceptual metaphors." (Kövecses, 2000a, is a large-scale application of this procedure to the domain of emotions.) Conceptual metaphors have a concrete source and an abstract target domain (see Lakoff and Johnson, 1980; Kövecses, 2002). One such grouping of expressions that emerges from such an analysis of anger-related expressions is the following: "boil with anger," "be pissed off," "seethe with anger," "make one's blood boil," "simmer down," and many others (see Kövecses, 1986, 1990; Lakoff, 1987; Lakoff and Kövecses, 1987). What makes it possible to group these expressions together is the fact that they all have to do with a hot fluid in a container. This is a concrete conceptual domain that we call the "source domain" of a metaphor. The target domain will be anger because it is the abstract emotion concept of anger that is the "target" of the linguistic expressions in the source; they can all have an anger-related meaning in the appropriate context. Such a pairing of the hot fluid in a container as source with anger as target leads us to establish the conceptual metaphor that we can put as anger is a hot fluid in a container. In other words, we can now use the term "metaphor" to refer to two "things": metaphorical linguistic expressions (such as "boil with anger") and conceptual metaphors (such as anger is a hot fluid in a container). I will use the term metaphor here in primarily in the second sense. Conceptual metaphors are constituted by what are termed "mappings." Mappings are fixed conceptual correspondences between a source and a target domain. (On mappings, see Lakoff, 1993 and Kövecses, 2002). Mappings provide a certain structure for the abstract domain to which a source domain applies. We can find out what the mappings are between a source and a target on the basis of the metaphorical linguistic expressions that guide us to recognize conceptual metaphors. Let us see the mappings, or correspondences, in the case of the anger is a hot fluid in a container metaphor. We find that the container in the source conceptually corresponds to the body of the angry person in the target (i.e., anger is seen as being inside the body); that the hot fluid corresponds to the anger emotion (such as in "boiling with anger"); that the intensity of the heat of the fluid corresponds to the intensity of anger (such as in "simmering down"); that the cause of the heat rising corresponds to the cause of anger (such as in "making someone boil"). We can spell out the mappings as follows: source: hot fluid in a container target: anger container à body hot fluid à anger intensity of heat à intensity of anger cause of heat à cause of anger What we see here is that the source is characterized by a simple, coherent, and tightly organized knowledge structure that is utilized in understanding, or making sense of, the target, that is, the anger emotion in the present example. This notion of imposing a coherent and tightly organized knowledge structure on the target by means of a set of mappings has, in my view, far-reaching implications for psychotherapy. I take it that at least a large part of psychotherapy involves achieving an understanding, or making sense of, in a coherent fashion of a difficult-to-handle emotion, a difficult life situation, a traumatic experience, and the like. It is precisely the same kind of cognitive function that conceptual metaphors have in "normal" cases for "normal" people, but this function is especially foregrounded in and relevant to the "deviant" cases encountered in psychotherapeutic and psychoanalytic practice. It is my assumption that the goal of therapy is to achieve an understanding of our problems associated with emotions, life situations, etc., and that treatment, cure, and "insight" crucially depend on finding such a coherent and tightly organized knowledge structure that is mapped onto a problematic target domain. Source domains come with, or imply, a great deal of knowledge that metaphor researchers often explore. In other words, in addition to the constituent elements that comprise simple and basic knowledge structures that make up source domains, the domains also give rise to what are called "metaphorical entailments" (see Lakoff, 1993 and Kövecses, 2002, ch. 8). These are conceptual implications or consequences that follow from basic knowledge structures. Metaphorical entailments also structure target domains. For example, one of the things we know about the behavior of hot fluids in closed containers is that the hot fluid produces steam ("be steaming with anger); another is that there is pressure on the wall of the container ("burst with anger); a third is that as a result of too much pressure the container can explode ("He blew up"); a fourth is that as a result of the explosion parts of the container go up in the air ("She blew her stack"), and so forth. The expressions in parentheses show how our folk theory of the behavior of hot fluids in a closed container structures the way we understand anger. However, only those entailments participate in this job that meet certain specific requirements. A number of such requirements have been offered in the cognitive linguistic literature (see Kövecses, 2002, chs. 7, 8, and 9). Each of these functions independently in accounting for the question of what gets mapped from the source to the target. One is the requirement that only those conceptual materials are mapped from the source that are consistent with the image-schematic structure of the target. This is called the "invariance principle" (see Lakoff, 1993, Turner, 1996, and Kövecses, 2002, ch. 8). Another is the view that what gets mapped depends on the primary metaphors that make up complex ones; the primary metaphors determine entailments (see Grady, 1997). A third possible requirement suggests that each source domain is associated with a main meaning focus (or foci) and it is this that determines what gets mapped from the source; items outside the main meaning focus do not get mapped onto the target domain (see Kövecses, 2000b and 2002). As a final feature of the supra-individual level, we observe that many of the same metaphors that are identified on the basis of language can also be found in all kinds of cultural institutions (as these are broadly conceived), such as art, science, politics, sports, and so forth (see Lakoff, 1993 and Kövecses, 2002, ch. 5). These are real-world enactments of metaphors identified initially in language. Thus, in addition to the linguistic dimension, this gives an important cultural dimension to the supra-individual level. Metaphors can thus be said to pervade and structure many aspects of not only language but also cultural reality. Do they also pervade and structure the thought, the conceptual system of people? The individual level Metaphor can only be important for psychotherapy and psychoanalysis if it is in the mind, not only in language. The metaphors found on the supra-individual level are mainly based on the analysis of (decontextualized) linguistic expressions. But the question arises: Does this, or can this, analysis reveal anything about metaphors in the heads of individual speakers? In particular: Do people actually have the metaphors in their conceptual system that cognitive linguists discover on the basis of their linguistic analyses? Or, to put the same question bluntly: Are conceptual metaphors indeed conceptual? The breakthrough in answering these questions came with Ray Gibbs's (e.g., 1990 and 1994) psycholinguistic work on metaphor. In a variety of mental imagery tasks, he convincingly showed that conceptual metaphors indeed exist in the heads of individual speakers. He asked subjects to form mental images of such anger-related idioms as blow one's stack, flip one's lid, and hit the ceiling. In the experiments, people's images were highly uniform and consistent about what they imagined: a container with heated fluid inside that explodes as a result of too much pressure inside the container. Why was this so? This is only possible, Gibbs concluded, if people's images are constrained by something in their conceptual system. This something can only be the conceptual metaphor anger is a hot fluid in a container. That is, what Gibbs showed was that the metaphors discovered by cognitive linguists actually exist in the heads of speakers. However, the same research also shows that the match between the supra-individual and the individual levels is not perfect or complete. The incompleteness of the fit can come from a variety of factors (see Kövecses, 2002, ch. 17). The entire range of metaphors at the supra-individual level is not utilized by every single speaker of a language. The individual level is the level at which individual speakers of a given language use the metaphors that are available to them at the supra-individual level in actual communicative situations, but this is also the level at which they create new metaphors. This is the level that is characterized by such issues as the selection of metaphors for particular communicative purposes, how people think on-line using metaphors, how the context of communication constrains the use of metaphors, and how metaphors can organize or otherwise structure actual texts or discourses. Given all these possibilities for divergence from the metaphors that are identifiable at the supra-individual level, it is obvious that the individual level is crucially important for both the theory and practice of psychotherapy and psychoanalysis. Not all the metaphors that have been, or could be, identified at the supra-individual level are available to all speakers of a language. Both individuals and social groups vary in the kinds of metaphors they use and they also often invent new conceptual metaphors. Patients often invent new, that is, unconventional conceptual metaphors in interaction with therapists and analysts. These metaphors may relate to a number of different topics, or target domains; for example, their emotional states, marital problems, self-esteem, the relationship with the therapist, and many others. Therapists and analysts also often create new metaphors that concern the same (but also often additional) issues as the patients. One class of additional issues concerns, of course, the nature of the therapeutic or analytic process; that is to say, the theory of both psychotherapy and psychoanalysis. When people engage in on-line thinking in the course of communication, such as during the therapeutic interview, they commonly create what are termed "blends" (see, for example, Fauconnier, 1997; Turner, 1996; and Coulson, 2001). This process involves blending properties of the source with properties of the target. However, this is part of a broader phenomenon than metaphor. We do not need metaphorical source and target domains to get conceptual blends; people often use blends on-line, or in real time, in the course of working conceptually with input domains of any kind. But, obviously, we are more interested in metaphorical than in nonmetaphorical blends here. A nice example of a metaphorical blend relating to anger is provided by Turner and Fauconnier (2000). The example comes from their reanalysis of the anger is a hot fluid in a container metaphor by Lakoff and Kövecses (1987). Take the sentence "God, he was so mad I could see the smoke coming out of his ears." This is a novel elaboration of a conventional conceptual metaphor. In it, an element of the source is blended with an element of the target. There are no ears in the source and there is no smoke in the target, but in the blend both are present at the same time in the form of smoke coming out of his ears. A frame is created with smoke and ears in it that is novel with respect to both the source and the target. One important challenge facing cognitive linguists interested in psychotherapy is to see to what extent and under what circumstances conceptual blending is characteristic of psychotherapeutic texts of any kind. The use of conventional metaphors and the creation of new ones also depend on the context of communication, as broadly conceived. The kinds of personal concerns speakers have in their lives, their life histories, and even the physical context (such as the particular season in which they communicate) can significantly contribute to choosing, or arriving at, the metaphors they use in connection with a given topic. To what extent is this so in psychotherapy and psychoanalysis? As I understand it, one of the key ideas of psychoanalysis is that life history (and especially early childhood experiences) has a profound influence on our mental and emotional health. In a similar vein, some analysts suggest that significant earlier life experiences can function as source domains for significant later life experiences (see, e.g., Borbely, 1998). Since these experiences differ from person to person, we get a large number of, in our terms, unconventional source domains to conceptualize ourselves and our mental and emotional condition. If this is true, we can regard a large part of the therapeutic and analytic interview as an attempt to recover a source domain (i.e., a coherent knowledge structure that makes sense of a problem), together with its appropriate target domain (the problem itself). However, this is not to claim that in therapeutic and analytic discourse these individually unique conceptual metaphors that are based on particular life experiences dominate over the conventional ones isolated at the supra-individual level (many of which have been uncovered by cognitive linguists). What type of metaphor--the unique individual or the conventional ones--is more pervasive and more important in therapeutic and analytic discourse is an open empirical question. We do not have (and I feel that at this stage of research cannot have) any statistical evidence to decide on this issue. What we know is that both types are used, and sometimes people use novel extensions of conventional metaphors in the course of therapy. One example for the latter possibility is mentioned by McMullen and Conway (1996). They quote a patient as saying the following: "I felt like I was a stick of dynamite with a fuse about a quarter of an inch long" (McMullen and Conway, 1996, p. 67). What this sentence is based on is the completely conventional conceptual metaphor The angry person is a pressurized container, of which the metaphor anger is a hot fluid in a container is a version. In the example, the patient created several novel metaphorical entailments to conceptualize his or her condition. First, he or she viewed herself as a stick of dynamite, an extremely dangerous explosive that can cause severe injury to others, and second, he or she is on a quarter of inch long fuse, which presents emotional outbursts sudden, unpredictable, and dangerous. Now it is obvious even on the basis of this brief example that patients think of their problematic condition or situation in terms of extended, complex metaphorical images (such as the angry person is a pressurized container conceptual metaphor), rather than just using certain words and phrases metaphorically (which is in line with the traditional view of metaphor). When I talk to therapists and analysts about the function and importance of metaphor in their work, I often have the impression that many of these professionals interpret metaphor in the traditional sense of metaphor being a word or statement that is used nonliterally. (There are some notable exceptions to this. See, for example, the work by Linda McMullen, Lynne Angus and their associates, such as McMullen and Conway, 1996, Angus 1996, and the work done by the editors of PSYART.) I would like to suggest that psychotherapy and psychoanalysis must go beyond this limited interpretation of metaphor and embrace the cognitive linguistic view, which opens up new possibilities in understanding the nature of therapeutic discourse and people with mental and emotional problems. However, some other professionals did talk about extended, complex "images" in therapy and their importance. But, in talking to them, I felt that they had only a vague notion in mind concerning these images; their notion lacked all the structure that the concepts of source and target domain, conceptual mappings, metaphorical entailments, conceptual blending, and so on, can provide for the interpretation of therapeutic discourse. My suggestion here is that by bringing in these notions we can achieve greater precision and, hopefully, greater depth in diagnosis, interpretation, and treatment than before. The novel use of conventional conceptual metaphors does not only occur in therapeutic discourse. It is, of course, part and parcel of poetic texts, as in the fragment below by Adrienne Rich, who writes: Fantasies of murder: not enough: to kill is to cut off from pain. but the killer goes on hurting Not enough. When I dream of meeting the enemy, this is my dream: white acetylene ripples from my body effortlessly released perfectly trained on the true enemy raking his body down to the thread of existence burning away his lie leaving him in a new world; a changed man. The poetic image is again based on a version of the angry person is a pressurized container metaphor. What is novel here is that the dangerous substance under pressure is acetylene and that it is deliberately directed at the target of anger. The more general point is that poetic texts and therapeutic discourse are not wholly different kinds of talk; they often share resources for creativity. The shared resources enhance the possibility for poetry and arts in general to contribute to therapeutic treatment. The point here is not that this potential for contribution has not been recognized before (that it has been is obvious from various forms of play, art, music, dance, etc. therapy), but that with the cognitive linguistic view of metaphor we could attempt to possibly answer such deep questions as why and how these nontraditional forms of treatment work, and perhaps we could even suggest new forms of treatment based on the theory of conceptual metaphor. (There are several such attempts already on the Internet, ranging from plausible candidates through naïve attempts to downright nonsense.) As noted above, conceptual metaphors consist of sets of mappings, or correspondences, that map a coherent and tightly organized concrete knowledge structure and some of its entailments onto an abstract target domain. Individuals may differ in whether or not they make use of all the mappings of a metaphor that are associated with it supra-individually when they use a particular metaphor in particular communicative situations. As a limiting case, this can happen (and it can even happen in poetic texts) and all mappings may occasionally be utilized. But more often than not only a selection of conventional mappings is utilized in actual speech situations, depending on one's communicative needs. Thus it is not the case that all of the mappings arrived at by cognitive linguists at the supra-individual level are activated by individual speakers in the course of on-line thinking and communication in the real world. However, it is a working hypothesis of this study that there is one (or a limited number of) conceptual metaphor(s) whose mappings are utilized more or less fully by the patient and/or the therapist in the course of making sense of and making coherent a problematic area in a patient's life. To find the appropriate conceptual metaphor in the course of therapeutic treatment means in part to begin to see as coherent a problematic area of experience that was not coherent beforehand. The more of the mappings of a conceptual metaphor are applicable, the more successful the metaphor, and, I believe, the treatment is. The sub-individual level What I call the "sub-individual" level of metaphor is the level at which the conceptualization of a domain (the target) by means of another conceptual domain (the source) is made natural and motivated for speakers. Since the bringing together of the two domains into a conceptual metaphor is often motivated by sensori-motor experiences and human beings (no matter which language they speak) share these experiences, this is a level of the universal aspects of metaphor. The most obvious cases in which two different kinds of experience are seen as being in correlation are those that involve the human physiology. Bodily experiences are often correlated with certain abstract or subjective experiences, and this gives rise to conceptual metaphors that we find natural and well-motivated. Consider again the kindred metaphors anger is a hot fluid in a container and the angry person is a pressurized container. What could possibly motivate the emergence of these particular metaphors? Why don't we get anger is a book or anger is softly falling snow? The answer is that there seems to be nothing in our experience of anger that would make the latter metaphors natural for us, whereas in the case the former two we do seem to have the right kind of motivation (see Kövecses, 2000a, c). For one thing, we commonly associate the experience of anger with an increase in body heat. The correlation between the two is so taken-for-granted that it has become conventionalized in the vocabulary of many languages, including English. In English, we can describe someone who is easily angered as being a hothead and we can say that someone is hot under the collar. These are metonymic, rather than metaphoric, expressions where body heat stands for anger. We find a similar situation with the related metaphor the angry person is a pressurized container. Here there are several metonymies that show the close correlation of anger with felt bodily (either blood or muscular) pressure in our folk theory of anger. Thus people often use such metonymic expressions as "Don't get a hernia!" "He was blue in the face" or "She'll have a cow" in talking about anger. (The third example also shows that metonymy often combines with metaphor, in that the "pressure" part of the image in giving birth to a cow is metonymic but the giving birth to a "cow" part of the image is metaphoric.) The notion of correlation brings with it an important implication for the study of the relationship between metaphor and metonymy. It is that correlation in experience brings together two (no matter how) distant domains of experience in a single one. If we characterize metaphor as involving two distant domains and metonymy as involving a single domain, then we should regard correlation as a metonymic relationship in our conceptual system (see Kövecses, 2002, ch. 11). In it, one domain correlates with, thus metonymically stands for, another domain. The implication of this is that correlation-based metaphors can all be seen as having a metonymic basis. Thus in this view, metonymy is a bridge between experiencing two domains simultaneously, on the one hand, and seeing them as metaphorically related (a-as-b), on the other. But the question is: Has anyone ever come up with any real evidence independent of linguistic claims about such correlations? The answer is yes. Ekman, Levenson, and their colleagues conducted several experiments that show that abstract domains such as emotions regularly correlate with physiological changes in the body (see, e.g., Ekman, et al. 1983 and the references in Kövecses, 2000a, c). For example, anger has been shown to be correlated with an increase in skin temperature, blood pressure, and other autonomic nervous system activities. These changes make anger different from other emotions, which are characterized by a different ANS profile. These studies provide independent (i.e., nonlinguistic) motivation for the existence of the anger is a hot fluid and the angry person is a pressurized container metaphors that were discussed as a test case for the three-level view of metaphor above. Similar to this one, many other metaphors could be characterized at each of the supra-individual, individual, and sub-individual levels. This is not to claim, however, that each and every conceptual metaphor is based on such correlations in experience. In this respect again, the study of metaphors in therapeutic discourse offers fascinating possibilities for future research. What we called "individually unique metaphors" above cannot obviously be based on universal bodily experience. The motivation of these metaphors seems to come from individual life experiences and personal concerns and preoccupations in life. These metaphors are meaningful only to the people who have had those experiences and have the preoccupations on which the metaphors are based. In a way, the study of these metaphors presents a greater challenge to anyone interested in metaphor in therapy than the study of conventional metaphors. While in the case of the conventional metaphors cognitive linguistic descriptions may provide help for the researcher (and the therapist), in the case of individually unique metaphors only intensive deep interviews about the life and preoccupations of the patient might help. Because of these possibilities, the sub-individual level of metaphor is only partially universal--to the degree to which motivation is based on correlations in universal human experience. When it comes to individually unique metaphors, only individually unique personal motivation can provide motivation and hence meaningfulness for a large proportion of the metaphors used in therapeutic and analytic discourse. Where do metaphors "reside" in the human organism? The most natural location for metaphors, and especially for simple, or primary, metaphors (see, e.g., Grady, 1997), is in the brain. Given a source and a target domain, if one domain is activated other metaphorically-connected domains will also be activated. This shows that metaphors not only have linguistic and psychological reality but are also real in our neuro-anatomy. But metaphors have further bodily motivation. As Lakoff and Johnson (1999) observe, we have three ways in which simple, or primary, metaphors are embodied: (1), as we just saw, the correlations are embodied in our neuro-anatomy; (2) the source domains arise from the sensorimotor experiences of the human body; and (3) we repeatedly experience in the world situations in which source and target domains are connected. Conclusion The cognitive linguistic view of metaphor that has been discussed in this paper works on three levels: the supra-individual level corresponding to how a given language and culture reflects metaphorical patterns, the individual level corresponding to the metaphorical cognitive system as used by individual speakers of a language, and the sub-individual level corresponding to universal aspects of various kinds of embodiment. The main aim of the present paper was to show that this three-level view could offer us an integrated framework within which a number of questions and issues concerning metaphor in psychotherapy can be brought together. My aim was not to offer solutions to specific questions concerning the role of metaphor in psychotherapy; instead, my intention was to outline a framework in which a number of apparently disparate and independent metaphor-related issues potentially relevant to psychotherapy can be systematically and meaningfully discussed. This view, I contend, may have important implications for psychotherapy and psychoanalysis. I believe it is especially the study of what I called the "individual" level that may bring new insights to the field. I have mentioned some of these possibilities briefly in the paper, but the agenda is, I feel, much longer. If cognitive linguists and therapists are willing to listen to and learn from each other, the promise can be fulfilled and, as a result, people in trouble can be helped better. References Angus, Lynne E. 1996. An intensive analysis of metaphor themes in psychotherapy. In J.S. Mio and A.N. Katz (eds.), Metaphor: Implications and applications. (Pp. 73-84). Mahwah, NJ: Lawrence Erlbaum. Borbely, Antal F. 1998. A psychoanalytic concept of metaphor. International Journal of Psychoanalysis 79, 923-936. Coulson, Seana. 2001. Semantic Leaps. Cambridge and New York: Cambridge University Press. Ekman, Paul., R. W. Levenson, and W. V. Friesen. 1983. Autonomic nervous system activity distinguishes among emotions. Science 221: 1208-1210. Fauconnier, Gilles. 1997. Mappings in Language and Thought. Cambridge and New York: Cambridge University Press. Gibbs, Raymond W. 1990. Psycholinguistic studies on the conceptual basis of idiomaticity. Cognitive Linguistics 1, 417-451. Gibbs, Raymond W. 1994. The Poetics of Mind. Cambridge and New York: Cambridge University Press. Grady, Joseph E. 1997. theories are buildings revisited. Cognitive Linguistics 8: 267-290. Kövecses, Zoltán. 1986. Metaphors of Anger, Pride, and Love. Amsterdam: Benjamins. Kövecses, Zoltán. 1990. Emotion Concepts. Berlin and New York: Springer-Verlag. Kövecses, Zoltán. 2000a. Metaphor and Emotion. Cambridge and New York: Cambridge University Press. Kövecses, Zoltán. 2000b. The scope of metaphor. In A. Barcelona (ed.), Metaphor and Metonymy at the Crossroads. Berlin: Mouton de Gruyter. 79-92. Kövecses, Zoltán. 2000c. The concept of anger: Universal or culture-specific? Psychopathology 33, 159-170. Kövecses, Zoltán. 2002. Metaphor. A Practical Introduction. Oxford and New York: Oxford University Press. Lakoff, George and Mark Johnson. 1980. Metaphors We Live By. Chicago: The University of Chicago Press. Lakoff, George and Mark Johnson. 1999. Philosophy in the Flesh. New York: Basic Books. Lakoff, George and Zoltán Kövecses. 1987. The cognitive model of anger inherent in American English. In D. Holland and N. Quinn (eds.), Cultural Models in Language and Thought. Cambridge and New York: Cambridge University Press. Lakoff, George. 1997. How unconscious metaphorical thought shapes dreams. In D.J. Stein (ed.), Cognitive Science and the Unconscious. (Pp. 89-120). Washington DC: American Psychiatric Press. Lakoff, George. 1987. Women, Fire, and Dangerous Things. Chicago: The University of Chicago Press. Lakoff, George. 1993. The contemporary theory of metaphor. In A. Ortony, ed. Metaphor and Thought, 202-251, Cambridge and New York: Cambridge University Press. McMullen, Linda M. and John B. Conway. 1996. Conceptualizing the figurative expressions of psychotherapy clients. In J.S. Mio and A.N. Katz (eds.), Metaphor: Implications and applications. (Pp. 59-71). Mahwah, NJ: Lawrence Erlbaum. Turner, Mark and Gilles Fauconnier. 2000. Metaphor, metonymy, and binding. In A. Barcelona (ed.), Metaphor and Metonymy at the Crossroads. 133-145. Berlin: Mouton de Gruyter. Turner, Mark. 1996. The Literary Mind. Oxford and New York: Oxford University Press.