Concepts and reasoning must be independent of the sensory-motor system, since the sensory motor system, being embodied, cannot be a form of disembodied abstract symbol-manipulation.

Language too - if it was to fit the symbol-manipulation paradigm - had to be literal, independent of imagery, and independent of the sensory-motor system.

From this perspective, the brain could only be a means to implement abstract "mind" - wetware on which the "programs of the mind" happened to be implementable. Mind on this view does not arise from and is not shaped by the brain. Mind is a disembodied abstraction that our brains happen to be able to implement. These were not empirical results, but rather followed from philosophical assumptions.

In the mid-1970's, cognitive science was finally given a name and outfitted with a society and a journal. The people who formed the field accepted the symbol-manipulation paradigm. I was originally one of them (on the basis of my early work on generative semantics) and gave one of the invited inaugural lectures at the first meeting of the Cognitive Science Society. But just around the time that the field officially was recognized and organized around the symbol-manipulation paradigm, empirical results started coming in calling the paradigm itself into question.

This startling collection of results pointed toward the idea that mind was not disembodied - not characterizable in terms of the manipulation of meaningless symbols independent of the brain and body, that is, independent of the sensory- motor system and our functioning in the world. Mind instead is embodied, not in the trivial sense of being implementable in a brain, but in the crucial sense that conceptual structure and the mechanisms of reason arise ultimately and are shaped by from the sensory-motor system of the brain and body.

JB: Can you prove it?

LAKOFF: There is a huge body of work supporting this view. Here are some of the basic results that have interested me the most: The structure of the system of color categories is shaped by the neurophysiology of color vision, by our color cones and neural circuitry for color. Colors and color categories are not "out there" in the world but are interactional, a nontrivial product of wave length reflactances of objects and lighting conditions on the one hand, and our color cones and neural circuitry on the other. Color concepts and color-based inferences are thus structured by our bodies and brains.

Basic-level categories are structured in terms of gestalt perception, mental imagery, and motor schemas. In this way the body and the sensory-motor system of the brain enters centrally into our conceptual systems.

Spatial relations concepts in languages around the world (e.g, in, through, around in English, sini in Mixtec, mux in Cora, and so on) are composed of the same primitive "image-schemas", that is, schematic mental images. These, in turn, appear to arise from the structure of visual and motor systems. This forms the basis of an explanation of how we can fit language and reasoning to vision and movement.

Aspectual concepts (which characterize the structure of events) appear to arise from neural structures for motor control.

Categories make use of prototypes of many sorts to reason about the categories as a whole. Those prototypes are characterized partly in terms of sensory-motor information.

The conceptual and inferential system for reasoning about bodily movements can be performed by neural models that can model both motor control and inference. Abstract concepts are largely metaphorical, based on metaphors that make use of our sensory-motor capacities to perform abstract inferences. Thus, abstract reason, on a large scale, appears to arise from the body.

These are the results most striking to me. They require us to recognize the role of the body and brain in human reason and language. They thus run contrary to any notion of a disembodied mind. It was for such reasons that I abandoned my earlier work on generative semantics and started studying how mind and language are embodied. They are among the results that have led to a second-generation of cognitive science, the cognitive science of the embodied mind.

JB: Let's get back to my question about the difference between cognitive science and philosophy.

LAKOFF: OK. Cognitive science is the empirical study of the mind, unfettered by apriori philosophical assumptions. First-generation cognitive science, which posed a disembodied mind, was carrying out a philosophical program. Second-generation cognitive science, which is working out the nature of the mind as it really is - embodied! - had to overcome the built-in philosophy of earlier cognitive science.

JB: Does "second-generation cognitive science" presuppose a philosophy?

LAKOFF: We have argued that it does not, that it simply presupposes commitments to take empirical research seriously, seek the widest generalizations, and look for convergent evidence from many sources. That is just what science is committed to. The results about the embodied mind did not begin from, and does not presuppose, any particular philosophical theory of mind. Indeed, it has required separating out the old philosophy from the science.

JB: Where does this leave philosophy?

LAKOFF: In a position to start over from an empirically responsible position. Young philosophers should be thrilled. Philosophy is anything but dead. It has to be rethought taking the empirical results about the embodied mind into account. Philosophy considers the deepest questions of human existence. It is time to rethink them and that is an exciting prospect.

JB: What about the academic wars between postmodern and analytic philosophy? LAKOFF: The results suggest that both sides were insightful in some respects and mistaken in others. The postmodernists were right that some concepts can change over time and vary across cultures. But they were wrong in suggesting that they all concepts are like that. Thousands are not. They arise around the world in culture after culture from our common embodiment.

Postmodernists were right in observing that there are many places where the folk theory of essences fails. But they were wrong in suggesting that such a failure undercuts our conceptual systems and makes them arbitrary. The analytic tradition insightfully characterized the theory of speech acts. Although formal logic does not work for all, or even most, of reason, there are places where something akin to formal logic (much revised) does characterize certain limited aspects of reason. But the analytic tradition was wrong in certain of its central theses: the correspondence theory of truth, the theory of literal meaning, and the disembodied nature of reason.

The academic world is now in a position to transcend both positions, each having contributed something important and each needing revision.

JB: Is there an East Coast and West Coast divide?

LAKOFF: Dan Dennett referred to the "East Pole" and "West Pole" back in the early-to-mid 1980's, as if the proponents of the disembodied mind were all on the East Coast and the proponents of the embodied mind were all on the West Coast. Research on the embodied mind did tend to start on the West Coast, but even then the geographical characterization was oversimplified. By now, both positions are represented on both coasts and throughout the country. Cambridge and Princeton in the past have tended largely toward the old disembodied mind position, at least in certain fields. But there are so many interesting thinkers on both coasts and spread across the country that I think that any geographical divisions that still exist won't last long.

When Dennett first made that distinction, the great revolutions in neuroscience and neural modeling were just starting. Cognitive linguistics was just coming into existence. Metaphors We Live By had barely come out and Women, Fire, And Dangerous Things had not yet been written. Nor had Edelman's Bright Air, Brilliant Fire nor Damasio's Descartes Error, nor Regier's The Human Semantic Potential, nor the various books by Pat and Paul Churchland. Over the past decade and a half, neuroscience and neural computation have changed the landscape of cognitive science and they will change it even more in next decade or two. Those changes will inevitably move us further toward an appreciation of the embodiment of mind. You cannot think anything without using the neural system of your brain. The fine structure of neural connections in the brain, their connections to the rest of the body, and the nature of neural computation will keep being developed. The more we discover about the details, the more we will come to understand the detailed nature of how reason and the conceptual systems in which we reason are embodied.

The idea of disembodied reason was an apriori philosophical idea. It lasted 2500 years. I can't imagine it lasting another 30 years in serious scientific circles.

JB: And what do we have to look forward to?

LAKOFF: Cognitive science and neuroscience are triggering a philosophical revolution. Philosophy In The Flesh is just part of the first wave. Over the next decade or two, the neural theory of language should develop sufficiently to replace the old view of language as meaningless disembodied symbol manipulation that one finds in the old Chomskyan tradition. But the biggest, and one of the most important, changes will come in our understanding of mathematics.

The precursor of that change was Stanislas Dehaene's The Number Sense, which reviewed the evidence from neuroscience, child development, and animal research indicating that we (and certain other animals) have evolved with a part of our brains dedicated to enumeration and simple arithmetic up to a small number of objects (around four). Rafael Núñez and I begin with those findings and ask how sophisticated arithmetic (with the laws of arithmetic) developed, that is, how could ordinary conceptual mechanisms for human thought have given rise to mathematics?

Our answer is that the ordinary embodied mind, with its image schemas, conceptual metaphors, and mental spaces, has the capacity to create the most sophisticated of mathematics via using everyday conceptual mechanisms. Dehaene stopped with simple arithmetic. We go on to show that set theory, symbolic logic, algebra, analytic geometry, trigonometry, calculus, and complex numbers can all be accounted for using those everyday conceptual mechanisms. Moreover, we show that conceptual metaphor is at the heart of the development of complex mathematics. This is not hard to see. Think of the number line. It is the result of a metaphor that Numbers Are Points on a Line. Numbers don't have to be thought of as points on a line. Arithmetic works perfectly well without being thought of in terms of geometry. But if you use that metaphor, much more interesting mathematics results. Or take the idea, in set-theoretical foundations for arithmetic, that Numbers Are Sets, with zero as the empty set, one as the set containing the empty set, and so on. That's a metaphor too. Numbers don't have to be thought of as being sets. Arithmetic went on perfectly well for 2000 years without numbers being conceptualized as sets. But if you use that metaphor, then interesting mathematics results. There is a third less well-known metaphor for numbers, that Numbers Are Values of Strategies in combinatorial game theory. So which is it? Are numbers points? Are they sets? Are numbers fundamentally just values of strategies in combinatorial games?

These metaphors for numbers are part of the mathematics, and you make a choice each time depending on the kind of mathematics you want to be doing. The moral is simple: Conceptual metaphor is central to conceptualization of number in mathematics of any complexity at all. It's a perfectly sensible idea. Conceptual metaphors are cross-domain mappings that preserve inferential structure. Mathematical metaphors are what provide the links across different branches of mathematics. One of our most interesting results concerns the conceptualization of infinity. There are many concepts that involve infinity: points at infinity in projective and inversive geometry, infinite sets, infinite unions, mathematical induction, transfinite numbers, infinite sequences, infinite decimals, infinite sums, limits, least upper bounds, and infinitesimals. Núñez and I have found that all of these concepts can be conceptualized as special cases of one simple Basic Metaphor of Infinity. The idea of "actual infinity"-of infinity not just as going on and on, but as a thing- is metaphorical, but the metaphor, as we show turns out to quite simple and exists outside of mathematics. What mathematicians have done is to provide elaborate carefully devised special cases of this basic metaphorical idea.

What we conclude is that mathematics as we know it is a product of the human body and brain; it is not part of the objective structure of the universe - this or any other. What our results appear to disprove is what we call the Romance of Mathematics, the idea that mathematics exists independently of beings with bodies and brains and that mathematics structures the universe independently of any embodied beings to create the mathematics. This does not, of course, result in the idea that mathematics is an arbitrary product of culture as some postmodern theorists would have it. It simply says that it is a stable product of our brains, our bodies, our experience in the world, and aspects of culture. The explanation of why mathematics "works so well" is simple: it is the result of tens of thousands of very smart people observing the world carefully and adapting or creating mathematics to fit their observations. It is also the result of a mathematical evolution: a lot of mathematics invented to fit the world turned out not to. The forms of mathematics that work in the world are the result of such an evolutionary process.

It is important to know that we create mathematics and to understand just what mechanisms of the embodied mind make mathematics possible. It gives us a more realistic appreciation of our role in the universe. We, with our physical bodies and brains, are the source of reason, the source of mathematics, the source of ideas. We are not mere vehicles for disembodied concepts, disembodied reason, and disembodied mathematics floating out there in the universe. That makes each embodied human being (the only kind) infinitely valuable - a source not a vessel. It makes bodies infinitely valuable - the source of all concepts, reason, and mathematics.

For two millenia, we have been progressively devaluing human life by underestimating the value of human bodies. We can hope that the next millenium, in which the embodiment of mind will come to be fully appreciated, will be more humanistic.