Language and Ideology An interview with George Lakoff Roberta Pires de Oliveira Universidade Federal de Santa Catarina, Brazil 1. Preface This is a second interview with George Lakoff, whom we thank for the opportunity of clarifying some issues concerning the theme “language and ideology”, which was not the focus of our first interview. This new interview started with René Dirven’s suggestion of a re-edition of the 1998 interview, reformulated in such a way that the theme of the present book, “language and ideology”, would be highlighted. However, going through it, I realised the need for deeper explanations. Meanwhile the contributors to this volume read it, and some of them contributed topics to this new interview, sending me their own doubts and comments. Some Brazilian researchers have also contributed in the same way. Thus, the present interview brings together many voices. Some relevant fragments of the 1998 interview are here republished as a guide to the reader, since they are the background of the present interview. The new questions were answered by e-mail. Perhaps a face-to-face dialogue would have been better, but e-mails turned out to be the only possible vehicle of communication. My gratitude to all those who have helped. My special acknowledgement to René Dirven, without whom this new interview would not be, to Esra Sandikcioglu, who was my link with the other participants, and to Robson de Souza Bittencourt, who provided support at all steps of this new interview. The organisation, and the final form are my own responsibility. 24 ROBERTA PIRES DE OLIVEIRA To George Lakoff, I want to express my deepest gratitude for agreeing to be interviewed and for his democratic way of being. I also want to express my apologies to him for my stubbornness at certain points in the interview. 2. Lakoff’s contribution to the cognitive paradigm R: When Metaphors We Live By (Lakoff and Johnson 1980) appeared, there was a favourable environment for the study of metaphor. Ortony’s first edition of Metaphor and Thought was published in 1979. The philosophical rebirth of metaphor may, however, be drawn back to I. A. Richards (1936) and Max Black (1962, first published 1954) in the 50s. To what extent does cognitive semantics owe to the whole atmosphere which emerged from the refutation of the Logical Positivist approach to metaphor? L: I had read Black and I had no interest in what Black was doing. Black had accepted the basic tenets of analytic philosophy and he saw metaphor as external to ordinary everyday language and meaning, which was the heart of what I was interested in. Mark Johnson had studied with Paul Ricoeur. So he knew the Ricoeur tradition and the continental tradition and had come to the conclusion, through working with Ricoeur, that metaphor was central to thought.1 But I wasn’t at all influenced by that tradition. What influenced me was the discovery that ordinary everyday thought and language, and especially ordinary everyday thought, is structured metaphorically. It followed that the correspondence theory of truth and analytic philosophy in general was fundamentally mistaken. That was the major discovery. Max Black, as an analytic philosopher, hated that. He wrote a review of our book and he thought it was an awful book. If he was to maintain his fundamental philosophical assumptions, he had to believe that (Lakoff 1998: 88–89). R: The 1998 interview started with the historic roots of cognitive linguistics (hereafter CL). Let us now turn to the contemporary state of the art in the cognitive paradigm. How do you see your contribution to it? Is your own theoretical “evolution” representative of the cognitive paradigm? To what extent does your approach orient CL? 1. Paul Ricoeur has written a great deal on metaphor. He does account for Black’s influence in his work, see Ricoeur (1981). AN INTERVIEW WITH GEORGE LAKOFF 25 L: One of the wonderful things about CL is that it is not dominated by any one figure. I consider myself to be extremely fortunate to have had great colleagues and students, people who have had their own interests and their own ideas. That is what a lively discipline is like. My own theoretical evolution started with generative linguistics and generative semantics, then moved to CL, and now has evolved to the neural theory of language, which I see as part of CL. Langacker and Fauconnier also started with important careers in the generative tradition. Most other cognitive linguists did not, so far as I can tell. I am always trying to integrate everything within CL. Most of the others of my generation are busy enough doing the important work they do, without concentrating on such integration. But that does not necessarily mean that what I do orients the field. R: You have claimed that your research is connected to that of Fillmore’s, Fauconnier’s, Langacker’s. Could you be more specific about the contribution of each one of these authors? Let’s start with Langacker. How does what Langacker (1990, 1991, 1999) is doing relate to what you are doing? Has each one of you followed a different path (after finding some obstacle)? L: Langacker is interested in a broad set of issues: imagery in semantics, dynamic processes in semantics and in discourse, a wide range of semantic structures, polysemy, the way meaning is expressed in the languages of the world, and so on. He has worked more on these than I have, though I share his interests and respect his insights. Langacker is not interested in doing the kinds of thing that I am most interested in: precise formulations of metaphors and radial categories, the formulation of a theory of constructions, links to other branches of cognitive science, the neural theory of language, and applications to literature, politics, philosophy and mathematics. In some cases, we disagree as to particular analyses, but not all that often. We mostly just think about different issues in different ways. And Ron has developed his own vocabulary and notations, which are not always useful for what I am interested in. R: In what respect is Construction Grammar, as it is developed by Fillmore (1996), Sweetser (1996), Goldberg (1995, 1996a, 1996b), compatible with your approach? Where do you converge, what are the differences? 26 ROBERTA PIRES DE OLIVEIRA L: Construction grammar began with a collaboration by me and Fillmore in the late 1970s, growing out of his work on case grammar and frame semantics and my paper on “Linguistic Gestalts” (1977). The first major work on construction grammar was my case study of There-Constructions, done in 1983 and published in 1987 in Women, Fire, and Dangerous Things (1987) (henceforward: WFDT). The Fillmore-Kay-O’Connor paper came later, in 1988. Goldberg was a student of both mine and Fillmore’s — a brilliant and original student. Fillmore and I parted company intellectually over the issue of CL, specifically over the need for linguistics to incorporate and contribute to results of cognitive science. Fillmore sees himself as an ordinary working grammarian, not as a contributor to the theory of mind, nor as beholden to broader research on the mind and brain. Because of this difference in perspective, Fillmore has never learned metaphor analysis, nor the theory of radial categories, nor cognitive grammar, and has no interest in the neural theory of language. He sees himself as a generative linguist. Nonetheless, he is one the world’s most insightful linguists, and I and other cognitive linguists have profited enormously from his insights. But as Fillmore and Kay have moved away from cognitive reality and toward an HPSG-like theory, our views have diverged. I am still very much involved in developing construction grammar within the neural theory of language. R: Is there any incompatibility between your approach and Fauconnier’s (1997)?2 L: None that I know of. Though we do tend to focus on different issues. I accept and use the theory of mental spaces and I think it is deep and important. But, since I tend to be interested in details that he is not interested in, I don’t find his formalism useful in most cases. R: One could argue that Len Talmy (1996) is one of the founders of the cognitive paradigm. Where does he stand now in comparison with the above strands? L: Len has been one of the mainstays of the field since the mid-1970s. His interests are largely limited to the semantics of closed class elements and to 2. See Fauconnier and Sweetser (1996), Fauconnier and Turner (1994, 1996). AN INTERVIEW WITH GEORGE LAKOFF 27 differences in lexicalisation patterns across languages. He is not interested, say, in studying conceptual metaphor or the theory of constructions or the neural theory of language. Given what he is interested in, he has made enormous contributions. R: What is it that keeps all these strands in the cognitive paradigm together? L: A passion for studying all of language from a cognitive perspective, a genuine feeling of mutual respect, a realisation that no one person is going to be able to think about everything or get everything right, and a commitment to building a co-operative and open scientific community. Ultimately I think the Neural Theory of Grammar will be a unifying force, since it will provide a common vocabulary for all the different strands.3 3. Language, culture and cognition R: One of the topics of our first interview dealt with the relation between culture and cognition. Your answers support the hypothesis, empirically verified, as you claim, of a universal basis for cognition. R: Would you say we have universal concepts? Do we have universal feelings? In WFDT, you claim that anger, for instance, is universal and it may be explained by physiological reactions. L: Yes, there are universal concepts. There are universal metaphors, universal aspects of language, because we all have very similar bodies and our physical experiences in the world are very similar. Those are where universals come from. R: So, you are on the opposite side of the Whorfian hypothesis (1956) about language? L: No. The Whorfian hypothesis as it is usually stated is badly described. There is a large chapter on this in WFDT. Whorf is much more interesting than most people give him credit for and he said many, many more interesting 3. The reader may find more about the Neural Theory of Grammar on http://www.icsi. berkeley.edu/NTL/. 28 ROBERTA PIRES DE OLIVEIRA things than to propose linguistic determinism. Whorf had a theory of universals. Whorf taught Summer School courses in which he went through the kinds of universals of semantics that would show up in every language. So Whorf was not against the existence of universals. People describe him as if he were, but that’s not true. He was interested in many, many things. The use of metaphors in language. He thought incorrectly that there were languages that had no metaphors. He was wrong about it. But he was right about a great many things. He was right that there are differences in conceptual systems and that these differences do show up in different parts of language. And he was especially right that the morphological and grammatical parts of language function differently than the non-morphological and non-grammatical parts. They function more automatically, almost like reflexes. And therefore the kinds of concepts coded in those systems are automatic and unreflective. He said that this is important, and he was right. We are not against that. In fact, one of the things we are very much concerned with is developing a neural theory that will characterise the differences between those aspects of language that are subject to reflection and those that are not. We don’t have it yet (Lakoff 1998: 98–99). R: Many researchers both in cognitive anthropology and in language acquisition come up with strong evidence that different cultures use entirely different conceptualisations of space experience and space co-ordinates,4 thereby giving new credibility to the Whorfian relativity principle, in its classical interpretation: language, understood as a social phenomenon, guides our thoughts and perceptions. What would you say about this? To what extent is the Whorfian hypothesis relevant and even part of CL? L: It is not all one way or the other. As I say in the section on Whorf in WFDT and in Philosophy in the Flesh (Lakoff & Johnson 1999), there are both universal concepts and language-particular concepts. In some cases, like spatial relation concepts, there are universal primitives that combine in different ways in different languages. The work in cognitive anthropology tends to ignore research on universal primitives. 4. For cognitive anthropology see Levinson (1994) for Tzeltal. For language acquisition see Bowerman (1996a, 1996b). AN INTERVIEW WITH GEORGE LAKOFF 29 But there are also concepts that are just different across cultures, and there are even metaphors that differ across cultures, though as Grady (1997) has observed, the primary metaphors tend to be the same (but aren’t always). There is no such thing as “the” Whorfian Relativity Principle. There are about a dozen different dimensions to Whorf’s central idea, each with different empirical evidence for and against it. See WFDT, Chapter 18. R: Wierzbicka (1996) and some other researchers have been pursuing the hypothesis of universal concepts. In what ways is such a project related to your own approach? L: Not at all. Wierzbicka is a Leibnitzian. She begins with an a priori philosophical theory. She does not look empirically at the same range of data that cognitive linguists do. Her analyses sometimes capture some aspects of meaning, but they miss an awful lot. And they do not fit what we know of the mind and brain. I do not consider her a cognitive linguist at all. R: Most of your own work describes linguistic data from a synchronic viewpoint. Is it the search for universal concepts that justifies such a methodology? Does your methodology account for the historical dimensions of a given language as well as each word/concept’s complicated and multifaceted history? L: I work within a community of linguists who are very much concerned with history and cross-linguistic differences. My students and colleagues at Berkeley are very much concerned with such issues, and I depend on their insights. I just do not happen to do that research myself. The methodology of our community is particularly geared to the study of the history of both words, concepts, and grammatical constructions. The entire field of grammaticisation, a major emphasis in our department and our field as a whole, is about history. I can’t imagine how you could think that CL was not concerned with history. R: As I said, one does not find historical analyses or studies in your books. Let me focus on universal concepts and ideology. Couldn’t an appeal to universal concepts be used to justify ideological exclusions? L: It’s not an “appeal”. It’s an empirical matter. Just because many concepts are universal, it does not follow that all concepts are universal, by a long shot, as I have been saying for decades. Indeed, my writings on 30 ROBERTA PIRES DE OLIVEIRA politics stress differences in conceptual systems across cultures and subcultures. My “Metaphor and War” paper (1992) discussed differences between the conceptual systems of American policy makers and Saddam Hussein. My Moral Politics (1996) was entirely about differences in conceptual systems of liberals and conservatives in America. R: OK, but the hypothesis of universal concepts may underestimate the role of language, understood as a social institution, because language differences may turn out to be just “surface” differences, combinations of universal schemas. Liberals and Conservatives have different conceptual systems but they share the family conceptual metaphor: both of them see the power relation as a relation between parents and children. L: Your question was, “The hypothesis of universal concepts may underestimate the role of language, understood as a social institution, because language differences would turn out to be just “surface” differences, combinations of universal schemas.” This makes a number of false presuppositions. First, the issue of universal concepts is an empirical one, not an initial hypothesis, and it is clear, as I said, that many concepts seem to be universal and many concepts seem not to be. Second, it does not follow from the existence of many universals, that language differences would be “just ‘surface’ differences, combinations of universal schemas”. Differences in concepts across languages could still be major and profound. Moreover, even if they were “different combinations of universal schemas” that too could lead to profound differences, since it could lead to radically different inferences. Again, all this is an empirical issue. R: I agree. By ‘just surface differences’ I meant the possibility (an empirical one) that one could trace back cultural differences to combinations (perhaps algebraic ones) of primitive universal concepts and metaphors. For instance, the   metaphor would be a primitive universal concept. It may, however, combine with different metaphors and generate completely different conceptual systems. Indeed, in More than Cool Reason (1989), you and Turner describe the     metaphor as if it were not a historically mediated metaphor. In other words, as if it were not a cultural choice, whose roots are to be found in the nature of Western culture. The   metaphor is claimed to be in nature itself. It has no history. AN INTERVIEW WITH GEORGE LAKOFF 31 What guarantees that the cognitive approach is not taking its authors’ cultural models as universal ones? L: The guarantee of empirical constraints. We argued that on the basis of results in developmental psychology showing that children, by the age of three, tend to categorise animals, plants and objects in terms of their similarity to those children themselves. That is an empirical finding. It may be contradicted ultimately when children across many other cultures are studied. However, the prevalence of such a common “animacy hierarchy” in the languages of the world tends to support the developmental findings. As we point out, the common Great Chain of Being is not the same as the Elizabethan one, which was much more elaborated. As in all things, the question of whether a metaphor is universal or widespread across unrelated languages or cultures — or whether it is a historical innovation — is strictly an empirical issue. We were going with whatever empirical findings we knew about. The claim that the G C is just a matter of Western history, that it is not widespread in unrelated cultures and does not arise spontaneously in very young children, seems just to be an empirically false claim. R: Perhaps we could deepen the issue concerning the relation between concept, language and social world. Indeed this was one of the topics in our former interview. R: The relationship between concepts and language is certainly a vexed question in metaphor research. The thrust of your writings seems to rely upon the assumption that language reflects conceptual metaphors, i.e., that language is not independent of the mind, but reflects a perceptual and conceptual understanding of experience. Does the organisation of language, besides reflecting our conceptual systems, help shape them as well? L: Well, there is a difference between what we believe and what we talk about. And the reason is this: the things that are physically embodied are easier to study than interpersonal interaction. A child at birth interacts with its parents immediately. There’s personal interaction, physical interaction, every kind of interaction, right away. It’s not that interpersonal interaction is less important. It’s simply that we know less about how to describe it. We know less about how it functions in language and in reason at the present time. 32 ROBERTA PIRES DE OLIVEIRA R: But it is always possible within the framework of CL to trace moral and social concepts back to primitive bodily interactions. If you have an abstract or a highly cultural-dependent concept, let us say “democracy” or “love”, you can always trace it back to our bodies, isn’t that right? L: There is a difference, as I said before, between what we believe and what we write about. We write about something we have evidence for and we believe that culture plays a major role in language, although we don’t have a lot of evidence for that. Partly because the evidence has not been gathered together in a way that would be empirically adequate for this field. But we have no doubt that interpersonal relationships play a major role in language. Take, for example, the fact that children, when they are born, are able to imitate their parents, and they are able to get their parents to react positively. This is an interpersonal fact about human beings at birth. In order to imitate they have to be able to project their bodies onto to their parents bodies, and they have some idea how to control their bodies, in the way that the parents are controlling their own bodies. But, that takes a remarkable amount of neural sophistication, which is the ability to project your body out to someone’s else. This is the basis of empathy. So it’s very important in learning motor programs, in learning all sort of things about having a function in the world. It would be silly to say that this capacity plays no role in conceptual systems. Not only it plays a role in conceptual systems, but it also plays a role in language. For example, as Claudia Brugman (1983, 1984) shows in her study of Mixtec, there you have a system of body part terms that express spatial relations. And the way they work is by the people projecting their bodies onto the things in space, and that capacity of projecting your body onto to something else or someone else, is necessary in order to understand those space structures in language. Now, that seems to be the same capacity as the interpersonal capacity, the capacity to imitate, that interpersonal capacity is also physical, it’s not a separation between the physical and the cultural, or the physical and the interpersonal; they are both one and the same. The interpersonal capacity is at the basis of linguistic capacity for conceptualising space (Lakoff 1998: 93–94). R: In your paper “The metaphor system for morality” (1996b), you claim that morality is bodily based. Its basis is in the promotion of the material well being. Among others, you claim that it is better to be strong than to be weak. Does such a natural morality system transcend history and ideology? AN INTERVIEW WITH GEORGE LAKOFF 33 Isn’t it based upon our WASP biases (White, male, Anglo-Saxon Protestant)? L: First, I am NOT an Anglo-Saxon Protestant!!!!! Nobody who knows me could mistake me for an Anglo-Saxon Protestant. My culture and values are not those of Anglo-Saxon Protestants (though some of my best friends are Anglo-Saxon Protestants, and I love and respect them). R: I am really sorry. The question was meant to be a generic one, I said “our WASP biases”. Let me re-state it: To what extent is the moral conceptual system based upon the researcher’s own biases? L: I deeply resent the sexism, racism, and cultural bias implicit in the question. The issue is ONLY whether the claims that I and other cognitive linguists have made are empirically well-founded, not whether they meet some a priori test of non-WASP-ness that is based on racism, sexism, and cultural bias. Third, the question makes other incorrect presuppositions. You portrayed the paper as saying that “the basis of morality is in the promotion of material well-being”. What I actually said was somewhat different: “Morality around the world has its basis in the promotion of the material well-being of others and the avoidance and prevention of material harm to others”. I did not say, “it is better to be strong than to be weak”. I said: Other things being equal, you are better off if you are: healthy rather than sick, rich rather than poor, strong rather than weak, safe rather than in danger, cared for rather than uncared for, cared about rather than ignored, happy rather than sad, disgusted or in pain, … (Lakoff 1998: 250) “Better off” in English does not mean morally “better”. It means that it is easier to survive, function effectively in the world, and flourish. In other words, these are material conditions that (other things being equal) increase the probability that one will survive, function effectively, and flourish. When I said that, all things being equal, one is better off if one is rich rather than poor, I was not referring necessarily to monetary wealth, but to whatever forms of wealth are valued in a given culture, whether beads, cows, chickens, children, land, or yams. To say that this IS true, is NOT to say that it 34 ROBERTA PIRES DE OLIVEIRA SHOULD BE true. Indeed, I do not believe that it should be true. However it is. In general, the material conditions that enhance survival and flourishing happen (it is an empirical finding) to form the basis for the metaphors for morality in cultures around the world — metaphors like   ,  ,   , and so on. Does the fact that such metaphors show up in cultures throughout the world mean that morality is the same in all cultures? Of course not. As I pointed out in Moral Politics and in Philosophy in the Flesh, these basic metaphors are given different priority in different moral systems and can be organised by family-based metaphors for morality. The same universal metaphorical building blocks can give rise to wildly different moral systems, and moral systems have elements beyond these universal building blocks. You asked, “Does such a natural morality system transcend history and ideology?” Yes and no. Yes, in the sense that many of the building blocks of moral systems come from natural experiences. No, in the sense that history certainly plays a major role. As for ideology, ideologies tell you what is right and wrong and hence are comprised to a considerable extent by moral conceptual systems. Moral systems do not arise from ideologies; they are part of what constitutes ideologies. You asked, “Isn’t it based upon our own biases (that of a white, male, Anglo-Saxon Protestant)?” Putting aside the racism and sexism of the question, the answer is no. It is an empirical issue. R: I am really sorry about our misunderstanding on this point. As a woman in a third-world Catholic chauvinist country I certainly understand your feelings. But my question was a theoretical one: isn’t it the case that the concepts you are identifying as universal, are seen as such because, being bodily based, they are a function of your own language and culture? Consider the universal concept of ‘empathy’: to what extent does the fact that you speak English (and analyse it as well) influence such a conclusion? L: First of all, morality as empathy occurs in cultures around the world, e.g., all Buddhist cultures from India to China to Japan to Vietnam. It is hardly just an English concept. Second, our best current understanding of the biological locus of empathy is in our mirror neurons in the prefrontal cortex, which fire when you either perform a complex action or perceive someone else performing the same action. This appears to be the basis for imitation in AN INTERVIEW WITH GEORGE LAKOFF 35 children, but also for being able to experience internally what you perceive in someone else. The earliest research on mirror neurons was done with macacque monkeys and has since been extended to human beings. The research was done in Italy, not in an English-speaking country. If the mirror neuron hypothesis is correct, then empathy is fundamentally physical, though it may play a role in culturally specific conceptual systems and play special roles in such systems. Empathy is a natural capacity that has to be developed like other natural capacities. If you don’t hear any music throughout your childhood, you won’t become a musician. If you are not treated empathetically and are not brought up in a culture of empathy, you will not develop your empathic capacity. R: Let us suppose you are right about empathy being not only a universal bodily-based concept, but also a neurological one. How would it relate to moral issues? Would the natural concept of empathy be sufficient to guarantee a society ruling out the extermination of “otherness”? L: Of course not. Moral systems are very complex. Empathy is just one part. Other metaphorical concepts like M S or M A or the M O may play a greater role, one that overrides empathy. R: This is the explanation for the fact that “natural” morality is often overridden, right? L: The term “natural morality” is yours, not mine. As I said, and discussed at great length in Moral Politics, the capacity for empathy has to be developed through the experience of empathy. Particular moral systems, in households or wider cultures, can lead to the lack of the development of the capacity for empathy. Here it is important to refer to Moral Politics, where I discuss the   versus   moral systems. The   system promotes empathy and responsibility (toward both others and yourself). The   system promotes  ,   and the  . The   metaphor derives from the experience that you are better off if you have power over others than if others have power over you. In the   metaphor, the hierarchy of power should reflect the hierarchy of morality. Thus, in the 36 ROBERTA PIRES DE OLIVEIRA   metaphor, your group is more moral than other groups — and hence should exert power over them. In extreme forms, other groups are less than human. Horrible examples abound in recent history.    can be seen as either defining ends (in conservative political systems) or means (in both conservative and militant leftist systems). The Khmer Rouge saw those who disagreed with their political and economic views as lower in the Moral Order and subject to torture and execution. The Nazis saw Jews as lower on the Moral Order and less than human. The same has occurred, via   , on both the left and the right — in Yugoslavia, Argentina, East Timor, China, and elsewhere. Empathy suffers under either neglect or under a form of a   . 4. Language and ideology R: We are now on the theme of “language and ideology”, a vast and turbulent area, where hardly any basic agreement over the definition of each of the concepts the theme mobilises has been reached. R: Let me focus on the topic of language itself first. In several of your writings, we read that language is a reflection of our cognitive structures, which are bodily based. But isn’t there something — call it “la langue” — which has some kind of autonomy and which precedes us as individuals, a public treasure, so to say? L: What is language itself? Suppose you’d subtract all of phonetics, everything to do with actual sound systems, the auditory system, the acoustic systems, and you subtract everything that has to do with semantics, that is, argument structure, hierarchical semantic structure, and, you know, suppose you subtract everything that has to do with attention, and so on, you have almost nothing left. That is, what we see is an organisation of cognitive faculties on the phonological side, and on the semantics and pragmatics side, and the attention side. The functional side has to do with attention and memory, and so on. All those things come together to structure what language is. There are only particular ways in which these can be put together to structure what language is. I don’t see anything in the language AN INTERVIEW WITH GEORGE LAKOFF 37 that is truly independent of all these things. I don’t see any phenomenon at all that is independent of all these things. R: If I understood you correctly, language is not a social institution. What is language, for you? L: I don’t know what language is, if it’s not a relationship between the phonological means of expression — and, in sign language that would include hands as well, — and concepts. The way which you express in phonological form what you conceptualise, and that doesn’t exist independently of the kinds of ideas expressed, nor it exists independently of the phonology and the actual phonetics. And phonology doesn’t exist independently of phonetics. Given language as a means of expression, social considerations of course enter in at every level, since expression always occurs in a social and interpersonal context. The reason is that we use our conceptual systems to function socially and to comprehend social life. Since language reflects our conceptual systems, it will reflect the social aspects of our conceptual systems. Thus, seeing language from a cognitive perspective entails seeing language from a social perspective. In addition, since language is a tool for expression and communication, it can be used for social ends and as a marker of social status. In order to study the use of language for social ends, one must have a conceptual system characterising what “social” means. Here cognition enters the study of social issues once more (Lakoff 1998: 109–110). R: How do you relate this concept of language, a link between a material means of expression and concepts, and ideology? Is ‘ideology’ a synonym for ‘system of ideas’? L: That’s not quite right. Any ideology is a conceptual system of a particular kind, including a moral system. However, ideologies have both conscious and unconscious aspects. If you ask someone with a political ideology what she believes, she will give a list of beliefs and perhaps some generalisations. A cognitive linguist, looking at what she says, will most likely pick out unconscious frames and metaphors lying behind her conscious beliefs. To me, that is the interesting part of ideologies — the hidden, unconscious part. It is there that cognitive linguists have a contribution to make. 38 ROBERTA PIRES DE OLIVEIRA R: Do you believe ideology is bodily-based? If so, in what sense? L: The conceptual building blocks, including primary metaphorical concepts, are ultimately bodily based, but not in any simple direct way. Suppose concepts A and B each have bodily groundings A’ and B’. Now a complex concept C formed from both A and B, will be partially grounded through bodily experiences A’ and B’, but there will be no single bodily experience grounding the entire concept C, since A’ does not ground concept B and B’ does not ground concept A. In short, there is no simple-minded yes or no answer, but only such a complex answer. 5. Ideology and scientific research R: In many passages of our first interview you claim that from a scientific viewpoint the cognitive approach is the “best” available one at present. R: CL holds that the basis of meaning is embodiment. I believe this is a very interesting and a powerful hypothesis, but sometimes I have the impression that the body is just an organism. L: CL is not just a nice idea. There is evidence for it. If you look at the meaning of colour, you have to look at the physiology of colour vision, at the neural physiology of colour vision. It shows you that there is no colour out there in the world. So if you cannot be an objectivist about this, you are no longer an empiricist about colours. But colours do come out of your body so you can not be subjectivist about colour concepts either. There is no way of being either of the traditional things, just taking the small amount of data about the nature of colour. Colour is a very simple example, because it avoids a lot of the complexity, and you can see right there that both rationalists and empiricists are wrong about colour. And that is not simply a matter of having a nice idea that concepts are embodied. Colour is embodied. The same is true with basic-level categorisation and spatial relations and all sorts of other things including metaphor. So, once you see that, it is the evidence that is compelling. It is not just a nice idea. R: Well, I could always reply that evidence is a function of the theory, data is also a function of the model that you are building in. AN INTERVIEW WITH GEORGE LAKOFF 39 L: I refer you to Chapter 6 of Philosophy in the Flesh. There we discuss the idea of “convergent evidence” which is required in any scientific theory. Convergent evidence is evidence from multiple methodologies, each with different assumptions. The assumptions of various sources of evidence, being different, cancel each other out. That way the evidence is not a function of any specific model or methodology. A good example is evolution, where you look at the geological record, carbon dating evidence, morphological evidence from specimens, and DNA evidence. Similarly, in metaphor studies you look at inferential evidence, polysemy evidence, psychological experiments, historical change evidence, gesture evidence, discourse evidence, and so on (Lakoff 1998: 116–117). R: Someone reading our former interview may get the impression both that CL was discovered (it has no history), and all its achievements are empirical findings. Isn’t such a posture a way of turning a blind eye to the contributions of many philosophers and linguists who have been anticipating central tenets and findings of the cognitive theory of metaphor? L: After the fact, one can find quotes here and there that we didn’t know about that talked about something like conceptual metaphor, but which are usually so vague you can’t tell. I don’t really see detailed cross-domain mappings that are experientially grounded in any earlier material. I don’t see spatial relations primitives (image schemas and force dynamic schemas) before Talmy, Langacker, etc. I don’t see basic-level concepts before Brown, Berlin, and Rosch. Some things really are discovered. R: Isn’t that parallel to the claim that there are objective facts out there, and, by consequence, objective criteria for choosing the best theory. Aren’t you claiming that there are non-ideological criteria? To what extent is the choice for the best theory a non-ideological move? L: I am saying no such thing. Again I refer you to Chapter 6 of Philosophy in the Flesh. We are making three assumptions, assumptions that any scientist would have to make: to take evidence seriously, to look for convergent evidence, and to seek generalisations. I am not saying this is objective. It is just what defines reasonable scientific practice. I’m not saying it always leads to the best theory. That depends on all sorts of things, like how smart and imaginative the scientists are in finding convergent evidence and in formulating hypotheses to explain them. It also depends on history — the 40 ROBERTA PIRES DE OLIVEIRA history of science. Without knowledge of neuroscience, neural computation, and child development, we would have no theory of metaphor learning and no explanation for primary metaphor. In short, current theories of metaphor were impossible three decades ago, for historical reasons. R: Let’s talk about CL’s “discoveries”. Why are they “discoveries” rather than “conclusions”? Is it a way of putting CL in the “natural sciences”? Would you say human affairs may be studied with the natural sciences methodology? L: I do not accept the old-fashioned distinction between the so-called “hard” sciences and the humanities and social sciences. I think scientific findings are possible in humanistic and social subject matter. I do not think that all of the humanities is “reducible” to scientific questions — only some aspects. For example, in More Than Cool Reason, Turner and I argue that certain interpretations of poems use the same conceptual metaphors that exist in our everyday conceptual systems. This is a scientific claim and I think it is a reasonable application of cognitive science to the humanities. In my “Metaphor and War” paper, I claimed that American foreign policy is based on certain conceptual metaphors. I think this is a reasonable application of cognitive science to the social sciences. I think there are a lot more. R: In your book Moral Politics you affirm that from a non-ideological viewpoint the nurturance metaphor is better than the authoritative one. Your argumentation implies two strong claims: first, there are better metaphors; second, there is a neutral place — which seems to be science — from where one may point to better metaphors. Isn’t it dangerous to believe that scientific knowledge may be used to help decide moral issues? L: Let’s take these one at a time. First, I do not believe in the so-called “scientific method”. I think there are necessary ingredients to good science, like the three I mentioned above, e.g., convergent evidence. Second, suppose that science CAN bear on moral issues. Then it would be dangerous NOT to use it. The question then is whether science can have moral implications. Third, cognitive science CAN tell us about the precise nature of moral conceptual systems and their implications. It would be immoral NOT to use such knowledge. THIS knowledge does not tell us what to do, but it does clarify the bases on which we act. AN INTERVIEW WITH GEORGE LAKOFF 41 Fourth, some metaphors ARE better than others for certain purposes. The reason is that conceptual metaphors preserve inference and they may have entailments that are not metaphorical and that can be checked out empirically. Scientific metaphors are a good example. Einstein’s metaphor that the force of gravity isn’t a force at all but is the curvature of space-time has had a great deal of success. I think that the common neuroscience metaphor that networks of neurons are electrical circuits is a pretty apt one, which has also had a lot of success. The metaphor that electricity is a fluid has done pretty well — at least in restricted cases. Fifth, I argue at the end of Moral Politics that there are scientific reasons to choose   over   : (1) it’s better for raising children (all major child development research paradigms agree), (2) it is consistent with the way the mind works, while    isn’t, and (3)   is in accord with basic human flourishing and    isn’t. As I said, the danger is in ignoring what cognitive science and other sciences can tell you. 6. Cognitive linguistics and the social dimension R: Would you say that changing our ordinary metaphors is a way of changing our world view. Like, instead of seeing people as numbers we should try to see people as individual beings … L: That’s true. I think it’s possible that once you understand your own metaphors that there are certain possibilities for changing your world view. I think that’s true, but I think that is not new. I think every therapist knows this. R: Yes, but there is a danger in this position. L: But there is nothing dangerous in my description of it. R: The danger is not in the description, but in the prescription. 42 ROBERTA PIRES DE OLIVEIRA L: It is not a prescription. I mean the prescription is simply the same one that Socrates had: it is better to know yourself than not to know yourself. That’s the only prescription and then you make your choice. I am not suggesting that we should manipulate people to change their metaphors at all. In fact, if you know yourself, then you are less subject to that manipulation. There is no danger at all (Lakoff 1998: 116). R: So, some metaphors should be overcome. What criteria do you rely upon in order to justify the claim for better metaphors? L: Just as I said before. Conceptual metaphors preserve inferences and in context the inferences can be about non-metaphorical things. Some scientific metaphors are just scientifically inadequate, like Chomsky’s metaphor that a language is a set of strings, of meaningless symbols, and a grammar is a set of autonomous rules for generating them. The metaphor just doesn’t fit the facts of language. Some metaphors are harmful, like the American foreign policy metaphor that      and that   . It’s a disastrous metaphor. Or the metaphor that       . Another moral disaster. The same for    . It’s ecologically immoral. R: You still owe me an explanation about the criteria you are using to justify your claim that the above metaphors are harmful or wrong. But you are certainly right about their being harmful. Thus describing our common metaphors is not only a way of making ourselves aware of ideologies — something you have already stressed — but also a way of combating them, right? L: I agree entirely. R: Does the description of our common folk metaphors really uncover our deep metaphysical beliefs? We still talk of the sun rising and setting, as if we didn’t know better. L: Conceptual metaphor is primarily a matter of reason, not just speech. The question is: Do we reason on the basis of conceptual metaphor, do we act on those conclusions, and does it matter? The case of sunrise and sunset is trivial. When we are not thinking about physics, we automatically conceptualise the sun as rising and setting and it doesn’t hurt anything. But when George W. Bush uses the metaphor that     in AN INTERVIEW WITH GEORGE LAKOFF 43 the world to justify opposition to gun control, that matters. When conservatives use the metaphor that     and    , that matters. When the head of the American National Security Agency uses the conceptual metaphor that     , arguing against trying to prevent the East Timor massacre on the grounds that, like the mess in his daughter’s college dorm room, not all messes are worth trying to clean up — that matters. It is more than a bit dangerous to ignore our metaphor systems. R: In Brazil, the liberal policy is imposing upon us the metaphor    . A student of mine, Lunardi (2000), shows that the authority metaphor for language — that is her terminology — is dangerous precisely because second language teachers ignore that their acting is oriented by it. Do you think CL may help research in applied linguistics, especially in the area of second language learning? How? What would you say about an applied CL? L: I am engaged in starting a political think-tank to apply CL to politics. I think that the metaphors for    can help create a better society. I think new foreign policy metaphors need to be found. Yes, I think there can and should be an applied CL. I also think CL could be of enormous use in second language learning, in mathematics teaching, in therapy, and in the understanding of social and political life. R: What about a cognitive sociolinguistics? L: There is a cognitive sociolinguistics in existence, in my Moral Politics, in Steven Winter’s new book on law and CL, A Clearing in the Forest, in the dissertation by Pamela Morgan on political speeches and the conceptual system of business schools, and in the dissertation by Nancy Urban on business metaphors being used to restructure education. R: What is, in your opinion, the best metaphor to characterise the “social” role of CL? If there is one. L: It is vital. CL is not merely an academic discipline that studies language from a cognitive perspective. Rather it provides a methodology for understanding the conceptual basis of harmful social and political policies and allows us to articulate better the moral basis of more helpful social and 44 ROBERTA PIRES DE OLIVEIRA political policies. It can also help one comprehend one’s own life and one’s interactions with others. It is a great aid in following the Socratic advice to “Know thyself”. R: Thank you very much, George. Acknowledgments We would like to thank the organisers of the first volume of Fórum Lingüístico and Prof. George Lakoff for authorising the re-publication of fragments of Lakoff’s interview. Although I interviewed Lakoff in 1997, during the 5th International Cognitive Linguistics Conference in Amsterdam, the interview was published in 1998, and may be found in Lakoff (1998). References Berlin, Brent and Kay, Paul. 1969. Basic Color Terms. Their Universality and Evolution. Berkeley, Los Angeles: University of California Press. Berlin, Brent, Breedlove, Dennis E. and Raven, Peter H. 1974. Principles of Tzeltal Plant Classification. New York: Academic Press. Black, Max. 1962. Metaphor. Models and Metaphors. New York: Ithaca, 25–47. ———. 1993. “More about metaphor”. In A. Ortony (ed.), Metaphor and Thought. Cambridge: Cambridge University Press, 19–71. Bowerman, Melissa. 1996a. “Learning how to structure space for language: A crosslinguistic perspective”. In P. Bloom, M. Peterson, L. Nadel and M. Garrett (eds), Language and Space. Cambridge, MA: MIT Press. ———. 1996b. “The origins of children’s spatial semantic categories: Cognitive vs. linguistics determinants?” In J. J. Gumperz and S. C. Levinson (eds), Rethinking Linguistic Relativity. Cambridge: Cambridge University Press. Brown, Cecil H. 1990. “A survey of category types in natural language”. In: S. L. Tsohatzidis (ed.), Meanings and Prototypes. Studies in Linguistic Categorization. London: Routledge, 17–47. Brugman, Claudia. 1983. “The use of body-part terms as locatives in Chalcatongo Mixtec”. In Report no. 4 of the Survey of California and Other Indian Languages. Berkeley: University of California, 235–90. ———. 1984. Metaphor in the Elaboration of Grammatical Categories in Mixtec. Linguistics Department. University of California, Berkeley. Brugman, Claudia and Lakoff, George. 1988. “Cognitive topology and lexical networks”. In S. L. Small, G. W. Cottrell, and M. K. Tanenhaus (eds), Lexical Ambiguity Resolution: Perspectives from Psycholinguistics, Neuropsychology, and Artificial Intelligence. San Mateo, CA: Morgan Kaufman, 477–508. AN INTERVIEW WITH GEORGE LAKOFF 45 Fauconnier, Gilles. 1997. Mappings in Thought and Language. New York: Cambridge University Press. Fauconnier, Gilles and Sweetser, Eve (eds). 1996. Spaces, Worlds and Grammars. Chicago: University of Chicago Press. Fauconnier, Gilles and Turner, Mark. 1994. Conceptual projection and middle space. Technical Report 9401 Department of Cognitive Science. ———. 1996. “Blending as a central process of grammar”. In A. E. Goldberg (ed.), Conceptual Structure, Discourse and Language. Stanford, CA: CSLI Publications, 113–130. Fillmore, Charles J. 1996. “The pragmatics of constructions”. In D. I. Slobin, J. Gehard, A. Kyratzis and J. Guo (eds), SocialInteraction,SocialContext,andLanguage:Essays in Honor of Susan Ervin-Tripp. Mahwah, N.J.: Lawrence Erlbaum Associates, 53–69. Fillmore, Charles J., Kay, Paul and O’Connor, Catherine. 1988. “Regularity and idiomaticity in grammatical construction: The case of ‘let alone’”. Language 64: 501–538. Goldberg, Adele E. 1995. Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press. ———. 1996a. “Jackendoff and construction-based grammar”. Cognitive Linguistics 7: 3–19. Goldberg, Adele E. (ed.) 1996b. Conceptual Structure, Discourse and Language. Stanford, CA: CSLI Publications. Grady, Joseph E. 1997. “Theories are buildings revisited”. Cognitive Linguistics 8–4: 267–290. Johnson, Mark (ed.). 1981. Philosophical Perspectives on Metaphor. Minneapolis: University of Minnesota Press. Lakoff, George. 1977. “Linguistic gestalts”. CLS 13: 236–287. ———. 1986. “The meaning of literal”. Metaphor and Symbolic Activity 1,4: 291–296. ———. 1987. Women, Fire and Dangerous Things. What categories reveal about the mind. Chicago: University of Chicago Press. ———. 1990. “The invariance hypothesis: Is abstract reason based on image-schemas?” Cognitive Linguistics 1,1: 39–74. ———. 1992a. “Metaphor and semantics”. In W. Brigth (ed.), International Encyclopedia of Linguistics. Oxford: Oxford University Press, 417–419. ———. 1992b. “Metaphor and war: The metaphor system used to justify war in the Gulf”. In M. Pütz (ed.), Thirty Years of Linguistic Evolution. Studies in honour of René Dirven on the occasion of his sixtieth birthday. Amsterdam and Philadelphia: John Benjamins, 463–481. ———. 1993. “The contemporary theory of metaphor”. In: A. Ortony (ed.), 202–250. ———. 1996a. Moral Politics: What Conservatives know that Liberals don’t. Chicago: University of Chicago Press. ———. 1996b. “The metaphor system for morality”. In A. E. Goldberg (ed.), 249–266. ———. 1998. “Cognitive semantics. In the heart of language”. Fórum Lingüístico, Volume 1,1: 83–118. 46 ROBERTA PIRES DE OLIVEIRA Lakoff, George and Johnson, Mark. 1980. Metaphors we Live by. Chicago: University of Chicago Press. ———. 1999. Philosophy in the Flesh: The embodied mind and its challenge to Western thought. Chicago: University of Chicago Press. Lakoff, George and Turner, Mark. 1989. More than Cool Reason: A field guide to poetic metaphor. Chicago: University of Chicago Press. Langacker, Ronald W. 1990. Foundations of Cognitive Grammar. Vol.2 Descriptive application. Stanford, CA: Stanford University Press. ———. 1991. Concept, Image, and Symbol: The cognitive basis of grammar. (Cognitive Linguistics Research: 1) Berlin: Mouton de Gruyter. ———. 1999. Grammar and Conceptualization. (Cognitive Linguistics Research: 14). Berlin: Mouton de Gruyter. Levinson, Stephen C. 1994. “Vision, shape, and linguistic description: Tzeltal body-part terminology and object description”. Linguistics 32: 791–855. Lowe, John, Baker, C. F. and Fillmore, Charles J. 1997. “A frame-semantic approach to semantic annotation”. Proceedings of the SIGLEX Workshop ‘Tagging Text with Lexical Semantics: Why, what and how’. Washington, D. C., April 4–5: available from: (www.icsi.berkeley.edu/~framenet). Lunardi, Angelita Mendes. 2000. Professores de língua e a metáfora da língua. M. A. thesis in linguistics. Universidade Federal de Santa Catarina. Morgan, Pamela S. 1997. “Self-presentation in a speech of Newt Gingrich”. Pragmatics 7,3: 275–308. ———. 1998. Beyond the issues: A linguistic and conceptual study of American public discourse. Unpublished Ph.D. dissertation. University of California, Berkeley. Ortony, Andrew (ed.) 1993. Metaphor and Though. Cambridge: Cambridge University Press. Pires de Oliveira, R. 1997. “‘A manhã é uma esponja’: um Estudo sobre a Engenhosidade Semântica”. D.E.L.T.A., vol. 13, no. 2: 247–274. Richards, I. A. 1936.The Philosophy of Rhetoric. New York and Oxford: Oxford University Press. Ricoeur, Paul. 1981 “The metaphorical process as cognition, imagination and feeling”. In M. Johnson (ed.), Philosophical Perspectives on Metaphor. Minneapolis: University of Minnesota Press. Rosch, Eleanor. 1977. “Human categorization”. In N. Warren (ed.), Studies in Crosscultural Psychology, Vol. I. London etc.: Academic Press, 1–49. –1978. “Principles of Categorization”. In E. Rosch and B. B. Lloyd (eds), Cognition and Categorization. Hillsdale, J. J.: Lawrence Erlbaum, 27–48. Sweetser, Eve E. 1996. Mental Spaces and the Grammar of Conditional Constructions. In G. Fauconnier and E. Sweetser (eds), 318–333. Sweeter, Eve E. and Fauconnier, Gilles. 1996. “Cognitive links and domains: Basic aspects of mental space theory”. In G. Fauconnier and E. Sweetser (eds), Spaces, Worlds and Grammars. Chicago: Chicago University Press: 1–28. AN INTERVIEW WITH GEORGE LAKOFF 47 Talmy, Leonard. 1996. “Fictive motion in language and ‘ception’”. In P. Bloom, M. A. Peterson, L. Nadel and M. F. Garrett (eds), Language and Space. Cambridge, MA, and London: MIT Press, 211–276. Urban, Nacy Y. 1999. The school business: Rethinking educational reform. Unpublished PhD. dissertation. Berkeley: University of California. Wierzbicka, Anna. 1994. “Cognitive domains and the structure of the lexicon”. In L. Hirschfeld and S. A. Gelman (eds), Mapping the Mind. Domain Specificity in Cognition and Culture. Cambridge: Cambridge University Press, 431–452. ———. 1996. Semantics: Primes and Universals. New York and Oxford: Oxford University Press. Whorf, Benjamin Lee. 1956. Language, Thought and Reality: Selected writings. Edited and with introduction by John B. Carroll. Cambridge, Mass.: MIT Press. Winter, Steve. 1995. “A clearing in the forest”. Metaphor and Symbolic Activity 10,3: 223–245.