Design for a Theory of Meaning

Home Page: Mark Turner

Design for a Theory of Meaning

Copyright © 1992 by Mark Turner
Published in W. Overton and D. Palermo, editors,
The Nature and Ontogenesis of Meaning,
Lawrence Erlbaum Associates, 1994, pages 91-107.

Panton metron anthropos -- Protagoras


In the fifth-century B. C., when the Attic philosopher Protagoras proposed that man is the measure of all things, he offered a design for a theory of meaning. The distinctive character of any theory of meaning congruent with this design is its conception of meaning not as a static property external to human beings but rather as an aspect of dynamic human thought grounded in human nature.

Since Protagoras' epistemology is known to us principally from dubious and hostile summaries in Plato's Theaetetus, we have no reliable idea of the specific character of his own specific theory of meaning. It is clear that he lost the epistemological contest of his age, and of ours. Socrates, Plato, and Aristotle rejected his fundamental stand, as did most later theorists of meaning. Our present century stands out as a particularly abysmal time for the premise of Protagoras, in which prominent theorists have attempted to explain the nature of meaning as quite independent of the human person. Reference theories of meaning tell us, for example, that the light is on and it doesn't matter whether anybody is home: meaning is conceived of in such theories as essentially anchored in states of affairs in an objective reality, with the consequence that the meaning of an utterance must be the reality to which it refers. This leaves the human person out of the loop altogether: a semantic express train shoots straight from the linguistic symbols to an objective reality without passing through the human brain, let alone stopping in the human brain, let alone taking its entire journey there.

Other contemporary theories, such as formalist theories of meaning, do make a stop in the human brain, but it is only a courtesy stop, which consists of acknowledging that the human brain is the site in which meaning is attributed but viewing this fact as incidental, a matter of mere implementational detail. For example, artificial intelligence theories commonly regard mind as a formal engine that performs formal computation over meaningless symbols that remain meaningless as they undergo manipulation but that receive a separate interpretation from some register of fixed interpretations. The combined competence of the formal engine and the fixed interpretations is called an interpreted formal system, otherwise known as a "semantic" engine. The formal, syntactic engine does all the work while the rules of interpretation go along for the syntactic ride, and the result is meaning. Meaning is thus a precipitate of a lot of formal computation, and the real job of a theory of meaning is to discover the nature of this formal computation. This view has its roots in the work of Alan Turing. It is familiar from the work of Alan Newell and Herbert Simon, who saw intelligence as an aspect of a physical symbol system, which is to say, as the result of manipulating symbols by means of formal rules. The tradition behind Newell and Simon stretches back through Alfred North Whitehead, Bertrand Russell, Gottlob Frege, the Wittgenstein of the Tractatus, Hobbes, Leibniz, and Descartes. Although numerous objections have been laid against the premises of this approach, including that formal syntactic work plus fixed interpretation is insufficient to generate semantics, that the world is not like a computer tape, and that the brain is not a Turing machine, nonetheless over the last several decades many thinkers have concluded that the brain simply must be a type of interpreted formal system, or at least that there is no plausible alternative hypothesis. (For a summary, see Edelman, 1989: 42.)

In practice, these traditions work from the largely unargued assumption that it is merely incidental that the interpreted formal system happens to be instantiated in a human brain or a human body. The human being is not the measure of all things, the interpreted formal system is. No accommodation is required in the theory or in the fact for the humanness of the brain and the humanness of the body in which the system happens to have been implemented. The human being is the platform on which a portable utility is installed. Questions about human nature are peripheral. Protagoras would not have approved.

Protagoras seems today to have come in for a revival, by which I mean that certain cognitive investigations of the nature of meaning appear to be taking as fundamental rather than incidental the fact that meaning is attributed by a human brain in a human body. (For surveys of these cognitive investigations, see, e. g., Edelman, 1987, 1989, 1992; Johnson, 1987; Lakoff, 1987, this volume; Lakoff & Turner, 1989; Turner, 1987, 1991.) A staggering mistake was made two and a half millennia ago in trivializing the premise of Protagoras, and we are only beginning to get over it. Of course, we will need to perform some crucial updating of its technical details. Whatever concept of the human person Protogoras had in mind, a modern concept would view the human person as patterns of dynamic activity in a human brain that has evolved to serve the human body of which it is a part. Culture, society, language, and the rest of human life are patterns in brains. Meaning is patterns in human brains.

If we embrace an updated version of the premise of Protagoras to view meaning as an aspect of neural activity in a brain that is structured in accord with the body it serves, the cost is high: we must cross off our list of conceptual instruments for building theories of meaning most of what has historically proved indispensable. The first thing to cross off our list is the partitioning of meaning into objective and subjective.


Objective meaning is meaning that exists independently of any human agency in conceiving that meaning. On the premise of Protagoras, we must regard objective meaning as an oxymoron, since no meaning exists independently of its attribution by the human brain.

The premise of Protagoras does not require us to discard everything associated with the concept of objective meaning. It allows us, for example, to keep the sense that certain capacities for attributing meaning are virtually indispensable and not open to choice. Some parts of our human system for attributing meaning are directly susceptible to pressures of fitness. For example, it seems to be a very fit part of brain activity to:

--attribute a vertical up-down gradient to the environment;

--distinguish between the interior and the exterior of one's body, with the skin as boundary;

--partition the world into objects and actions;

--understand certain objects as agents; and

--attribute purposes to agents.

Inabilities to attribute meaning in these ways would count as fundamental deficits in the organism: someone lacking them would die. It is hard even to imagine that someone lacking them could count as a human being. Evolution is likely to be extremely conservative in regard to such features. On the other hand, pressures of fitness seem virtually indifferent to other capacities, such as synaesthesia in the form of connecting certain colors to certain sounds. (For a discussion of differential evolutionary pressures, see Wimsatt, 1986.)

Our ways of attributing meanings, and the actual meanings we attribute, are thus not undifferentiated. They are differentiated both by the degree to which they are susceptible to pressures of fitness and by the degree of fitness they in fact have. The premise of Protagoras forbids us from talking about objective meaning as part of a theory of meaning, but it allows us to talk about the fitness of a conceptual system, and even about the extremely high fitness of some of its components. We can therefore feel perfectly respectable in believing that certain meanings and patterns of meaning are just indispensable, which is not to say objective. Fitness is not a measure of correspondence to objective reality but rather a measure of success. By inference, fitness is a measure of capacity, provided good and bad luck average out for large numbers of cases. In principle, two unlike conceptual systems for attributing meaning could be equally fit.

Protagoras seems to have suggested that we view things in exactly this way. It is a weak citation, but in Plato's Theaetetus, we find Plato's Socrates trying to imagine what the ghost of Protagoras might say on this point. The conjured ghost of Protagoras, as invented by Socrates and then remembered by Plato, argues that, while it is beside the point to say that some understandings are truer than others, it is surely the case that some are simply practically better than others. A sick person may find food distasteful (or, I suppose, see the visual field in an odd way, or find something warm to be cold, or be unable to distinguish up from down, and so on). While it is beside the point to say that the sick person's understanding of the world is untrue, it is obvious that the sick person is messed up and isn't going to do very well. We want to avoid saying that the healthy person has true ideas, but we surely want to say that he has better ideas of what is going on. He will do a lot better. The case is the same, so this ghostly Protagoras argues, with respect to education: the teacher leads the student not from false to true concepts, but from unsound to sound concepts. Sound concepts are those that make us more competent.

More subtly, the premise of Protagoras does not require us to abandon the view that extremely fit concepts are true reflections of objective reality, it merely asks us to take this view as itself fit rather than true. We can thus keep the most commonsensical part of the objectivist view. Some ways of attributing meaning are virtually indispensable and not open to choice. To lack them is to suffer, perhaps fatally. Viewing reality as objective and objectively knowable is a fit habit, however embarrassing as a tenet of a theory of meaning.

Although Protagoras asks us to discard from any theory of meaning the notion of objective meaning, the alternative is not to embrace the notion of subjective meaning--meaning that is private to the individual and essentially arbitrary. Protagoras was much attacked on just this point. He was dismissed as a radical relativist.

If meaning is attributed by human brains, then the concept of "subjective" meaning is as mistaken as the concept of "objective" meaning. Human capacities for attributing meaning show regularities that keep them from being radically private or arbitrary not because they are grounded in objective meaning but because the human brain develops under what we might call necessary biology and necessary experience.

Necessary Biology and Necessary Experience

For phylogenetic reasons, we all share a certain neurobiological endowment. We share not only gross anatomical structure in the brain but also dispositions of neuroembryology (Edelman, 1987, 1989, 1992; Churchland, 1986). This brain, endowed with necessary biology, develops under what we might call necessary experience. Given the standard neurobiological endowment, it is not possible for an infant to fail to have early experience of the gravity vector, of force dynamics, of image-schemas such as the source-path-goal schema, and so on. This necessary neurobiological endowment and this necessary early experience, common to all, work to develop a brain that attributes meaning in certain ways.[1] Although those ways belong to the individual, they are not appropriately to be called subjective, since they are common to all of us.

For example, we all recognize the eleven basic color categories and their central members; the neurobiological basis of this has been discovered. This capacity for attributing meaning comes exclusively from the individual brain--therefore it is not objective--but nothing about it is private or arbitrary--therefore it is not subjective. One can lack part of this capacity by being color-blind, but that is clearly a lack of what everybody else has (Berlin and Kay, 1969; Lakoff, 1987: 24-30; DeValois, Abramov, & Jacobs, 1966; DeValois & Jacobs, 1968; Gregory, 1990; Calvin, 1989: 146).

To take another example, the human brain knows the bilateral symmetry of our body and of our experience with the world. Our bodies are bilaterally symmetric as a consequence of our ancestors having had necessary experience over our phylogenetic history, in which they experienced our environment in a certain way, namely, as asymmetric in two dimensions but symmetric in the third. Our environments are asymmetric up-down (by virtue of gravity) and forward-back (by virtue of our directed movement forward), but they are symmetric right-left. What can happen from the right can happen from the left and conversely. What we can do to the right we can do to the left and conversely. We are endowed phylogenetically to recognize and exploit right-left symmetry in our environments. This experience is incorporated in our genetic endowment not by Lamarckian mechanisms but by mechanisms of selection for fitness: those organisms that are bilaterally symmetric and are endowed to recognize bilateral symmetry are fitter; their differential replication is higher, indeed so much higher that they are the only members of the species to survive.

Our biology carries the experience of our ancestors by means of the survival of some genes and the disappearance of others. Each of us is guided by the successes and failures of our ancestors through encoding in the gene pool by means of selection for fitness. Recognizing bilateral symmetry is part of this necessary phylogenetic biology that we all share. It is also part of necessary individual ontogenetic experience that we all share. We all have early experience of balance. We develop the capacity to attribute meaning according to this right-left symmetry, to see some things as in balance or in equilibrium. This has proven to be an extremely fit capacity, with the result that we all share this specific capacity for attributing meaning to our world and ourselves. It is just the wrong question to ask whether this meaning is objective or subjective. The capacity and the disposition to attribute meaning in this way are extremely fit and universal parts of our conceptual apparatus. They come from necessary biology and necessary experience.

The Metaphoric Basis of the Objective versus Subjective Distinction

The distinction between objective and subjective meaning, which the updated premise of Protagoras asks us to cross off our list, seems to derive from a conceptual metaphor in which we understand mind in terms of a container. In this basic metaphor, meaning is an object. Objective meaning is, metaphorically, located outside the mind-container. Metaphorically, we can have inside the mind-container a copy of this external objective meaning. If the copy is good, then we know objective meanings. We can all have copies, connected by virtue of their all having an external referent. In this metaphor, these internal copies cannot be connected to each other directly. They can be connected only through the intermediary of the objective referent, which is located outside the brain, and which is independent of the copies, and which has priority over the copies.

Under an updated version of the premise of Protagoras, this metaphor and its implications are swept aside. Meaning lies in patterns in human brains. Meaning lies inside the brain in a nonmetaphoric sense. No meaning lies outside the human brain. Meanings in different brains are connected not as copies of an external meaning, but because the brains share a biological and functional similarity in the activities of attributing meaning. We share genetic dispositions and we share universal early individual experience. As a result, we share ways of attributing meaning. We all have bodies asymmetric top to bottom and we all experience the gravity vector. We all have retinas and lateral geniculate nuclei that are structured neurobiologically in similar ways and we all experience the visible spectrum of light. We all have bodies that are right-left symmetric and we all experience moments of physical equilibrium. We all have bodies that bleed when their skin is punctured and we all have a great deal of early experience with objects going into or out of the body. And so on. By virtue of our shared phylogenetic past and our similar ontogenetic experiences, we share some fundamental ways of attributing meaning. It is merely misleading and confusing to ask whether the meaning, as attributed by these capacities, is objective or subjective. It is neither.

It is interesting to note that nothing is lost when we delete the distinction between objective meaning and subjective meaning from our list of useful conceptual instruments. Given the tenacity with which this distinction has been held, one would think that it does some profound conceptual work. On the contrary, if instead of thinking of a concept, idea, or model as the one that is objectively "most true," we instead think of it as merely the one that is "most fit," no functional capacity will be lost. There will be no consequence for our practical or scientific behavior. We will still inquire in the same way into the nature of things and still have the same standards for experiment, discovery, and belief. We will still be motivated to find the fittest ways of making sense of the world. True, our abandoning the distinction of objective versus subjective may change some of our metaphysical sense of our existence, but it will rob us of none of our knowledge or our practice.


The next conceptual instrument to cross off our list under the guidance of Protagoras is an ancient distinction which lives with us strongly in the present, the distinction between the mind and the body. This distinction comes in two avatars. The first, in which the body is viewed as a machine inhabited by an incorporeal ghost called the mind, belonged to Plato, Boethius, and Descartes, but is no longer taken seriously.

The second avatar is fundamental to the dominant theories of meaning from Plato to current artificial intelligence approaches. It is doubtful that Protagoras would have had any sympathy with it. In this view, the workings of the mind are disembodied; the brain is of course connected at its far margins to sensory receptors and motor neurons, but if we shave off this negligible surface, then we are left with the essence of mind: a structural and relational conceptual arena in which thought is configured, essentially free of the lowly body. Different theories take different views as to what is configured within this arena, but the usual candidates include varieties of formalism, such as recursive function theory, geometrical thinking, predicate calculus, logic of some stripe, or certainly something that is assumed to transcend the body and its experience.

This view of meaning as disembodied is contradicted by the growing neurobiological and cognitive research which suggests that the human mind is very deeply constrained, structured, and limited by the human body in which it resides and the human environment in which it lives (Edelman, 1987, 1989, 1992; Johnson, 1987; Lakoff, this volume). The human body incorporates the experience over time of the human species within its highly specific environment. This specificity structures the brain, constraining and guiding mental operations. A surgically removed brain in a vat may lack its fleshy house but is not disembodied, for the brain carries in its structure and operation the nature of the body it serves. The sensory sheets of the central nervous system, for example, do not stop at the surface periphery of the brain. Rather, they are projected with topological integrity deep into the brain. An attempt to remove them would destroy the most inward reaches of the brain. The pure mind depends upon the daily body. As Patricia Churchland writes in Neurophilosophy:

One of the most promising and puzzling discoveries about the organization of nervous systems is that many structures abide by a principle of topographic mapping, whereby neighborhood relations of cells at one periphery are preserved in the arrangement of cells at other locations in the projection system. If we think of the neurons at the sensory periphery as forming a receptor sheet, then deformed versions of that sheet are represented in a large number of CNS [Central Nervous System] regions. Whatever the functional principle served, it must be served successfully, since wide areas of the cerebral cortex as well as other structures abound in such maps, and new research is turning up more all the time. (Churchland, 1986: 119-120.)

The brain is not an unconstrained general-purpose machine. It has some highly specific purposes and values and corresponding structure (Edelman, 1992). As an example of this specificity, consider that everyone has a right-left symmetric environment, a right-left symmetric body, and a right-left symmetric brain, with the result that having a human body means having a human brain that is disposed to attribute meaning according to its understanding of balance and equilibrium. Such bodily understanding is available to be projected metaphorically to impart structure and meaning to more abstract concepts, such as our concept of justice. Although different societies can have different concepts, conceptual categories, and ways of naming, nonetheless the image-schemas of balance and equilibrium are to be found at work in the structuring of concepts wherever one cares to look. For example, our concept of balance in a line of poetry is a metaphoric projection from our supposedly low-level embodied knowledge of balance and force-dynamics (Turner, 1992, chapter 4; Johnson, 1987).

The premise of Protagoras would suggest that some patterns in the brain are inherently meaningful to human beings with human bodies because those patterns correspond to human bodily situations.[2] Patterns in the brain not inherently meaningful would derive their meaning through links to patterns that are inherently meaningful. In other words, patterns that are inherently meaningful could be present in other concepts, such as abstract concepts, so that even the most refined parts of our conceptual apparatus for attributing meaning could be ineradicably embodied.


A view of the nature and ontogenesis of meaning as ineradicably embodied would blur in two profound ways the distinction between culture and biology, requiring us to cross this distinction off our list of respectable conceptual instruments. The first blurring is now apparent: if meaning is structured and guided by the mapping of the body in the brain, then it is not possible to separate human culture from human bodies. Culture is patterns of activity in brains; brains are structured in accord with their bodies; therefore culture, which is activity in brains, is structured in accord with the bodies in which it resides. Conversely, brains are in various ways developed under cultural experience, such as experience of language. A certain amount of our actual neurobiology is inseparable from culture.

The second blurring is equally evident, having to do not with the mapping of the body in the brain but simply with the fact that the instantiation of culture can only be neurobiological. Meanings reside in the brain. Imagine you are microscopic and inside a brain, gazing about, trying to distinguish between biology and culture. You could not meaningfully point to one synapse and say, "see that, that's biology," and then point to a different synapse and say, "see that, that's culture." It's an absurd and untenable distinction. To the extent that there is culture, it just is neurobiology; it just consists of neurobiological events and structures. If, while you are in miniature and touring the brain, you observed some neurobiological activity involved in, say, language, you could not say in a meaningful fashion, "see that, it's just biology, culture is not involved." At the level of what exists, culture cannot be distinguished from neurobiology.


Among the largest and most influential concepts to be crossed off our list of instruments for constructing a theory of meaning is the distinction between innate and acquired. The innate versus acquired distinction lines up with some related distinctions: innate is internal, genetic, fundamental, manifested early, and so on, while acquired is external, experiential, relatively optional, late in development, and so on.

The distinction of innate versus acquired derives from a metaphoric view of the mind as a medium to which the world can add. In specific metaphors, this medium is specified as a wax tablet of a certain nature, upon which the world imprints, or as a sheet of paper bearing certain innate inscriptions and upon which the world additionally writes, or as an engine of certain capacities which are augmented by the world through experience and learning.

What is missing from the distinction of innate versus acquired is the recognition of how regular our experience is. Is there necessary experience? The answer is clearly yes. Some of our experience is so constant and universal that it must count as innate, in a certain sense. Capacities that are absolutely indispensable to the phenotype do not have to be carried by the genotype, or at least not exclusively by the genotype, provided that they arise from experience that can be absolutely counted upon. For example, consider our capacity to use the image schema of a path from a source to a goal. We use this image-schema to attribute meaning to many things, from the street we live on to the direction of an argument. A human being who lacked this image-schema of SOURCE-PATH-GOAL would be so incapacitated as to seem fundamentally deficient. Now imagine that the SOURCE-PATH-GOAL schema arises only through early experience. This would present no problem, because it is a matter of necessary early experience. Conceivably, severe sensory deprivation of a developing child could prevent it from "acquiring" the SOURCE-PATH-GOAL schema. But this would not mean that, because the schema is "acquired," it is therefore optional or exterior to the organism or anything but fundamental to the organism.

What is fundamental, criterial, or indispensable to the organism need not be carried genetically, or at least not exclusively so, provided that it is carried by necessary experience. A great deal of the development of the brain as a system for attributing meaning depends upon early experience, but much of this early experience counts as predictable nourishment rather than as optional acquisition. Our fitness in attributing meaning is a matter not just of genetic endowment, but of the regularity of the environment we encounter.

One student of evolutionary biology, William Wimsatt, has recently proposed that we replace the distinction of innate versus acquired with an alternative concept that saves what is useful about the distinction but discards what is specious (Wimsatt, 1986). Instead of disputing whether a capacity is innate or acquired, we should consider its degree of "generative entrenchment." A feature or capacity is more generatively entrenched the more other features or capacities depend upon it. Evolution is likely to be extremely conservative for capacities that are deeply generatively entrenched, and less so for capacities that have low generative entrenchment. Deeply generatively entrenched capacities are, on average, likely to appear earlier in development. But deeply generatively entrenched capacities need not be innate, in the sense of interior to the organism, or independent of experience, or carried exclusively by genetic material. They can equally come from the stability of the environment. Wimsatt writes of the imprinting of the greylag goose upon its mother:

not only is the imprinting mechanism of the greylag goose at birth "innate" (as on the standard ethological accounts), but the object of imprinting is also "innate." When the infant goose extricates itself from its shell, it imprints upon and follows any moving object. . . . In their natural environment, however, there is a very high probability that the young goose will properly imprint on its mother and will, in short order, learn to distinguish her cries and her appearance from that of other female greylag geese nearby. The family structure and behavior of the mother greylag goose at the time of birth makes it almost a certainty that the baby geese will imprint properly. . . . Thus the correct information (that a close moving object first detected at birth is mother) is reliably present in the environment. (Wimsatt, 1986: 200-201).

In short, this capacity of the greylag goose for operating in its world is "innate" but comes from experience. In this sense, necessary experience of a regular environment is innate and acquired. The distinction of innate versus acquired fails.

To eliminate the distinction of innate versus acquired might prove particularly unsettling at the present stage of inquiries into the nature of human thought and language because eliminating the distinction necessarily impeaches the hypothesis that human beings are born with an innate language module whose operation is independent of experience and autonomous from other cognitive capacities. Were we to grant for the sake of discussion that some of our linguistic capacities are fundamental, universal, interior to the organism, and incapable of being developed through the ontogenetic experience of the individual alone, this would not entail that these linguistic capacities must be carried exclusively or even largely by genetic material. The manifestation of these capacities could perfectly well depend upon experience, provided that the experience was necessary, which is to say, that it came from stable parts of our interaction with the environment, such as the experience of bounded interiors, of the gravity vector, of balance and equilibrium, of force dynamics, of image schemas like SOURCE-PATH-GOAL, and so on.


The distinction between innate and acquired is usually correlated with the distinction between genetics and experience, but this is misleading, since most of what is carried genetically is present because of experience. Through mechanisms of selection, genetic material embodies the history of the experience of its ancestors. The gene pool reflects bad experience by lacking the genetic material that disposed its phenotypes to bad experience; it reflects successful experience by carrying the genetic material that disposed its phenotypes to successful experience. We still have the sense of smell because our ancestors had certain kinds of experience as a result of having the sense of smell. In this way, the phylogenetic endowment reflects the differential experience of evolutionarily prior organisms. Our genes, for example, carry the instructions for constructing a body that is bilaterally symmetric because our ancestors in the phylogenetic tree experienced the world as right-left symmetric. Those organisms best adapted to this right-left symmetry were selected for. Bilaterally symmetric bodies have been selected for through the interaction of our ancestors and their environments. In this sense, our genetic material, and our bodies, reflect the experience of our species. Our bodies incorporate the history of our species with the environment. The genes that build these bodies incorporate the history of our species with the environment.

Blurring the distinction between genetics and experience will require us to reconceive in a second way the hypothesis that human beings are born with an innate language module whose operation is independent of experience and autonomous from other cognitive capacities. Even if such a language module exists, and even if it is carried entirely in genetic material, in such a way that universal capacities of language are not in the least developed under the influence of ontogenetic experience, nonetheless those linguistic capacities would still, in a fundamental sense, result from experience, the experience of the individual organisms in the phylogenetic tree behind us. The genetic material disposing us to the possession of a language module would have been subject to selection. The mechanism of this selection, natural and sexual, would have been exactly the experiential interaction between the organism and its environment.

The dispute over the extent to which language is innate versus acquired is likely to be with us for a very long time, first because the distinction itself is conceptually confused and so prevents us from thinking clearly about the issue, but second because we have failed to take into account the regularity of human experience. To the extent that universal linguistic capacity could depend upon individual ontogenetic experience, that ontogenetic experience would have to be universal. Experience that is ontogenetically universal for us is virtually certain to have been ontogenetically universal for our ancestors. This sharing of universal experience between ourselves and our ancestors leads to a difficulty in analyzing the origins of linguistic capacities: the genetic basis that could lie behind these capacities and the individual experiential basis that could lie behind these capacities just are equivalent. For a great many basic things, the average experience of ancestors, as represented in the genome, will be exactly the same experience to be had by the individual developing organism, and not just one such developing organism, but every such developing organism. We see the particular human linguistic capacity, but it will be hard to deduce its source, because the genetic basis it could have and the experiential basis it could have are not distinguishable. Do you have a capacity because you developed it in response to certain experience, or do you have it because every one of your ancestors had that experience and so evolution, through differential selection of capacity, developed it genetically for you and placed it into your genes? Or is it a combination of both? Logically, any of the three answers is possible, and the actual answer about the genesis of a specific capacity can only come, then, from empirical work, much of it necessarily in the sciences of genetics, neuroembryology, and developmental neurobiology. In consequence, the debate over the source of linguistic capacity in human beings cannot be resolved through "in principle" reasoning. A great deal of practical, empirical work lies before us, and it will simply have to be done.

It is important to realize that the question of whether a linguistic capacity is autonomous from other capacities--such as our capacities for seeing, for hearing, for understanding force dynamics, for understanding spatial relations, and so on--is not a question of innate versus acquired or of genetics versus experience. There is no reason to conclude that a linguistic capacity to which our genes might dispose us must be autonomous from nonlinguistic capacities. The phylogenetic development of linguistic capacity could depend upon nonlinguistic capacities; the ontogenetic development of linguistic capacity could depend upon nonlinguistic capacity; and the actual operation of linguistic capacity in the individual could involve or depend upon nonlinguistic capacity. Evolution the tinker seems always to work with and build upon what it has. My own guess--and it can count only as a guess at this stage in our research--is that various capacities for attributing meaning to the world must have been in place before the development of a linguistic capacity, and that the linguistic capacity depends upon them to some considerable extent. These capacities include image-schematic, visual, tactile, auditory, proprioceptive, sensorimotor, and other capacities (Edelman, 1992). I think it just improbable that whatever genetic endowment we have for a linguistic capacity was not constructed in part by borrowing strength from these prior capacities, but only empirical work can tell. The argument from the paucity of linguistic data in the early life of the individual to the assertion of an innate, autonomous language module does not follow. In this argument, it is asserted that children do not have sufficiently rich linguistic experience to allow them to develop the rich linguistic capacity they in fact do develop, and therefore that this linguistic competence must be carried genetically. What is missing from this argument is a consideration of the possible contribution to linguistic development of nonlinguistic experience. The brain may well draw on nonlinguistic capacities, as reflected in the genome or as developed through early experience, in arriving at a linguistic competence.

If we were to grant that the linguistic capacity is innate, it would not follow that it is independent of ontogenetic experience: the capacity of the greylag goose to recognize its mother is innate and depends absolutely upon reliable ontogenetic experience, predictable from the regularity of the environment. Nor would it follow that the linguistic capacity would be autonomous of nonlinguistic capacities: it may well have arisen genetically by drawing upon nonlinguistic capacities as they are carried in genetic material.


Under Protagoras's design for a theory of meaning, any capacity of mind has a claim to be respected within theories of meaning. But under objectivist designs, only mental capacities imagined as leading to so-called objective meaning have any claim on our respect. The appallingly influential objectivist design to be found in Descartes's Discours de la Méthode Pour bien conduire sa raison, et chercher la verité dans les sciences (1637), for example, views mind as dramatically separate from anything bodily, and considers the recognition of meaning as the exclusive province of the conscious mind and its supposed arsenal of capacities for objective knowledge (Wilkes, 1988: 22). In such an objectivist design, imaginative human mental capacities we associate with poetry are effectively dismissed from playing any role in theories of meaning. They are instead to be studied within the branches of rhetoric and poetics, and viewed as peripheral, optional, decorative, and seconday to the science of mind. As a pedagogical consequence, it is assumed that their study can be safely eliminated from the curriculum of those concerned with entailment and class inclusion, syntax and reference. The study of the theory of meaning in our century has largely ignored these rhetorical, figural, and narrative capacities of mind, at least until roughly the last decade.

If we look back to classical antiquity and its discussion of meaning and language, we see many marvelous claims. Someone like Demetrius, for example, feels comfortable explaining in his work On Style that

--Everyday language is widely and ineradicably metaphoric.[3]

--The conceptual metaphoric mappings we see manifested in language are asymmetric.[4]

--Metaphoric conceptions can be as "true" as any others.[5]

--We understand linguistic constructions to some extent in terms of what we might now call image-schemas.[6]

In Longinus, we can find the view that:

--Figural understanding is all the more powerful when it is so automatic that we do not recognize its figural nature.[7]

This view that our conceptual system and our consequent languages are fundamentally and ineradicably dependent upon rhetorical or poetic capacities sounds new because it is new to us. Education has split the scientific study of the mind, meaning, and language off from the rhetorical and literary study of the imaginative mind; it has done this on the pretext that it is safe to do so, because rhetorical and poetic capacities of mind are merely acquired extras, subjective, optional, and peripheral to any study of meaning: certainly they cannot be innate; certainly they cannot be fundamental to the capacity of the organism. The study of meaning has been prosecuted as the study of literal meaning, on the view that objective meaning is literal. Figural meaning has been characterized as unimportant.

But if we adopt the updated premise of Protagoras, we must abandon this split between the scientific study of mind and the rhetorical study of mind. The human mind attributes meaning. We should look at how it actually manages to do so, without preconceptions about how it ought to do so. When we begin that investigation, we find something quite interesting. Rhetorical powers of mind, such as the figural and the narrative, are not peripheral. Indeed, they appear to be fundamental, ineradicable, and indispensable to the organism. They appear to be ubiquitous in the meanings we do attribute. They appear in fact to play a dominant role in capacities that were once thought to be where all the action is: categorization, deduction, reasoning, argument. These rhetorical powers of mind have a claim as good as any to be called literal and innate, in the sense that they are basic, ineradicable, and indispensable to the normal functioning of the human organism.

I have sketched the outlines of such a cognitive rhetorical inquiry into the nature and ontogenesis of meaning in Reading Minds: The Study of English in the Age of Cognitive Science (Turner, 1991). Cognitive rhetoric is compatible with the premise of Protagoras in viewing the human person as the measure of all things. The human person has a human brain in a human body in a human environment which it must find meaningful if it is to survive. In consequence, the human person finds certain patterned experience inherently meaningful. The human person also has certain fundamental conceptual capacities for attributing meaning, such as the metaphoric, the metonymic, and the narrative. Far from trivial or peripheral, these capacities appear to be indispensable to the organism; they necessarily evolve ontogenetically according to necessary biology and necessary experience. Our job, as theorists of meaning, is to discover how they evolve and how they work.


The premise of Protagoras is a design for a theory of meaning, and, as such, it is a design for the invention of the university, or at least that part of the university that considers human beings, the brain, and meaning. On this design, it would be best to bring together as one division of the university all those studies concerned with the way in which the human brain attributes meaning. This would include cognitive linguistics and cognitive psychology, rhetoric and neurobiology, indeed any field of study concerned with the interpretive or creative mind. Under the premise of Protagoras, this division would be called "the humanities," since it would concern whatever belongs to human beings, the measure of all things.

Such a definition of the humanities may seem strange to us now, but it would not have seemed strange when the humanities were originally carved out as the study of what belongs to human beings, by contrast with the study of divinity, which was already established as the study of what belonged to divine beings. Under the premise of Protagoras, the study of the human brain is part of the humanities, as is the study of human language, human psychology, human development, and human neuroscience. If calling this organization "the humanities" seems objectionable, we can give it another name. But if the intellectual organization itself seems conceptually strange rather than merely unfamiliar, then we have some re-thinking to do. The university as we know it is organized according to a rejection of the premise of Protagoras. If that premise manages, in the next age, to come back into its own, it will mean a changed view not only of what belongs to a theory of meaning, but also of how a university should be organized. It is just such a re-organization, I think, that is most likely to lead us to the discovery of the nature and ontogenesis of meaning.


Berlin, Brent and Kay, Paul. (1969). Basic Color Terms: Their Universality and Evolutions. Berkeley: University of California Press.

Calvin, William H. (1989). The Cerebral Symphony: Seashore Reflections on the Structure of Consciousness. New York: Bantam.

Churchland, Patricia. (1986). Neurophilosophy: Toward a Unified Science of the Mind/Brain. Cambridge: MIT.

Cowan, Jack D. and Sharp, David H. (1988). Neural Nets and Artificial Intelligence. In Stephen R. Graubard, (Ed.), The Artificial Intelligence Debate: False Starts, Real Foundations (pp. 85-121). Cambridge: MIT Press.

Demetrius. (1932). On Style, section 2:86, translated by W. Rhys Roberts. Cambridge: Harvard [Loeb].

DeValois, R. L., Abramov, I., and Jacobs, G. H. (1966). Analysis of Response Patterns of LGN Cells. Journal of the Optical Society of America, 56, PAGES

DeValois, R. L. & Jacobs, G. H. (1968). "Primate Color Vision" Science 162, 533-540.

Edelman, Gerald. (1992). Bright Air, Brilliant Fire. New York: Basic Books.

Edelman, Gerald. (1989). The Remembered Present: A Biological Theory of Consciousness. New York: Basic Books.

Edelman, Gerald. (1987). Neural Darwinism: The Theory of Neuronal Group Selection. New York: Basic Books.

Gregory, Richard. (1990). Eye and Brain: The Psychology of Seeing, fourth edition. Princeton: Princeton University Press. [First edition London: Weidenfeld & Nicholson, 1966].

Johnson, Mark. (1987). The Body in the Mind. Chicago: University of Chicago Press.

Lakoff, George. (1987). Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: University of Chicago Press.

Lakoff, George & Turner, Mark. (1989). More than Cool Reason: A Field Guide to Poetic Metaphor. Chicago: University of Chicago Press.

Longinus. (1932). On the Sublime, section 17, translated by W. Hamilton Fyfe. Cambridge: Harvard [Loeb].

Turner, Mark. (1991). Reading Minds: The Study of English in the Age of Cognitive Science Princeton: Princeton University Press.

Turner, Mark. (1987). Death is the Mother of Beauty: Mind, Metaphor, Criticism. Chicago: University of Chicago Press.

Wilkes, Kathleen V. (1988). --, yìshì, duh, um, and consciousness. In A. J. Marcel & E. Bisiach (Eds.), Consciousness in Contemporary Science, pp. 16-41. Oxford: Clarendon.

Wimsatt, William C. (1986). Developmental Constraints, Generative Entrenchment, and the Innate-Acquired Distinction. In W. Bechtel (Ed.), Integrating Scientific Disciplines, pp. 185-208. Dordrecht: Martinus Nijhoff.

Home Page: Mark Turner