A relatively new addition to the familiar Platonic–Cartesian dualisms at the heart of Western thought has emerged since the linguistic/discursive turn in philosophy and the social sciences—the separation of semiotic from somatic.¹ Semiotic has been characterized as necessarily representational and/or linguistic, while somatic refers to a wide range of corporeal processes and practices assumed to be separated from mind, language, and/or conscious thought. In this paper, and in contrast to Jackson (1983) and Csordas (1990), we propose instead that the semiotic can be somatic, a conception that requires liberating semiotic from a conflation with the representational. And, conversely, we argue that the somatic or corporeal is necessarily semiotic when it involves the agentic, meaning-making practices of social persons, as in domains such as the senses, emotions, and, our focus here—body movement.

     To articulate this in a theoretically fruitful way requires a 'new realist' approach to the philosophy of science and the concept of 'dynamic embodiment.'² These provide a means to recover human agency as one of a variety of causally empowered natural kinds in the world. As a result, materially grounded physical human being makes possible cultural being, enabling a concept of personhood that is simultaneously bio–psycho–social. We are thereby able to revive Marcel Mauss's (1979 [1935]) tripartite suggestion, using new conceptual resources. Varela used these 'new realist' resources in an earlier argument for dynamic embodiment as a theoretical means to get beyond the absent moving body in embodied social theory (Varela 1994, Farnell 2000).

     In Part I of this paper, we present a general framework for this perspective, which we are calling "a paradigm of dynamic embodiment." In Part II, we apply this paradigm ethnographically to illustrate how bringing semiosis and somatics together requires a robust conception of multisensory modalities. At this stage, it is primarily a sensitizing conception rather than a definitive one.

Part I: Reaching for a Paradigm: Dynamic Embodiment

For some time now, in anthropology, sociology, and psychology, it has been commonplace to understand that human activity in everyday life is best conceived as 'action', not 'behavior' (Harré and Secord 1972, Ardener 1973, Crick 1976, Taylor 1980, Williams 1991, Varela 1995). In anthropology, there is a further understanding that human action is best framed in accordance with the ideas of practice, discourse, and embodiment (see Hymes 1971, Bourdieu 1977, Jackson 1983, and Ortner 1984). By the 1990s, the third component, embodiment, was captured in part by Csordas's call for the adoption of a "paradigm of embodiment" (1990). The special feature of the new paradigm (that we will later refer to as the Csordas-Jackson paradigm)³ was that human action is centered in, and constituted by, human physical being. Following Merleau-Ponty, 'physical being' here refers explicitly to the subjective (lived body) and not the objective (mechanical) body. The 'lived body' means the body as human beings themselves perceive it—felt, experienced, and sensed.⁴ The thread tying these three perceptual processes together is the feeling of doing. In sociology, Shilling (1993) and Turner (1984) embraced this paradigm and its special features, as did Harré (1984, 1986a, 1994, 1998) and Shotter (1993) in psychology. We consider the paradigm articulated by these anthropologists, sociologists, and psychologists to represent the first somatic revolution. This was an important challenge to disembodied theories of human action; whether idealist (for example, classic Levi-Straussian structuralist anthropology) or reductionist (for example, the 'unconscious' in Freud's structural model of id, ego, and super ego as a deterministic system of bio–psycho–social forces).

     Note that the theoretical emphasis in both anthropological and sociological versions of a paradigm of embodiment is on the feeling of the doing and not the doing itself. There is, thus, an omission of Harré's insight that, "[i]f a thing . . . cannot move about then perhaps it is not a person" (Harré and Secord 1973: 110) and the realization of his point in Merleau-Ponty's comment that "no [human being] perceives except on condition of being a self of movement" (1968: 257). We are, therefore, entitled to be skeptical of the extent to which the Csordas-Jackson paradigm faithfully employs the existential-phenomenology of Merleau-Ponty. While it is certainly true that Merleau-Ponty's key idea of the perceived body or embodied consciousness has been a major source of the somatic turn in social scientific theory, it is important to recall that the central principle which underwrites his concepts of the 'lived-body,' 'intercorporeity,' and 'flesh' is the 'self of movement.'

Theoretical Enrichment

     What we discuss here is an enrichment of the Csordas-Jackson paradigm in light of the approach to embodiment taken by Williams and Farnell. They contribute the idea that human action is best understood as a dynamically embodied discursive practice, a move that we regard as a second somatic revolution. Ironically perhaps, the second somatic revolution predated the first, having been initiated by Williams's dissertation in 1975. Williams discovered Harré's causal powers theory and used it to ground a semiotic approach to the embodied signifying moving person known as 'semasiology' (see Williams 1982). Williams's and Farnell's approach draws upon certain aspects of Suzanne K. Langer's philosophy of art, as clearly seen in a brief excerpt from Langer's paper "The Dynamic Image," about observing human movement in an American dance studio:

The dance is an appearance. . . . It springs from what dancers do; yet it is something else. In watching a dance, you do not merely see what is physically before you—people running around—what you see is a display of interacting forces, by which the dance seems to be . . . driven. . . . [But] these powers [of the dancers] . . . that seem to operate in the dance are not [simply] the physical forces of the dancer's muscles. . . . [These human powers which] we seem to perceive most directly and convincingly are created [by dancers] for our perception, and exist only for it (Langer 1957: 5; emphasis provided).

This citation reveals that Langer's philosophical intuition led her to connect human action (movement) with the powers of causation. More than this, she recognized that such powers did not merely reside in the physical forces of the dancer's muscles but are the powers of dancers as embodied persons engaged in making artistic meaning through generating movement patterns.⁵

     Williams and Farnell have constructed their theoretical standpoint by generalizing Langer's reference to the dance to all forms of human movement:

Human beings everywhere engage in complex structured systems of bodily actions that are socially acquired and laden with cultural significance. Some structured movement systems, such as the martial arts, sporting activities, idioms of dancing, dramatic arts, ceremonials, and ritual events, involve highly deliberate choreographed movement. Other uses of body movement remain out of . . . focal awareness. . . . Examples include ways of eating, dressing, walking, and sitting as well as modes of physical labor. . . . Also out of focal awareness most of the time are the hand gestures, postures, facial expressions, and spatial orientations that accompany speech. (Farnell 1996a: 536).

In contrast to the theoretical approach championed by Csordas and Jackson, their interest is in the moving body, the doing itself, which may, of course, also be felt. Here we have with greater precision the distinction between the first and second somatic revolutions: a difference between the feeling of the body (moving or not) and the movement of the body itself.

     The enrichment at issue here stems from the following theoretical principle that unifies the concepts of action, discourse, and embodiment: the primacy of the signifying moving person. Starting with the premise that all human action is the discursive practice of persons, we contend that Williams and Farnell are proposing a way of interconnecting three kinds of body-referenced talk that is found in social theory and in everyday life: talk about the body, talk of the body, and 'talk' from the body, that is, in the medium of movement (see Farnell 1994, Varela 1995).

     We use the terms 'discourse,' 'discursive,' 'conversation,' and 'semiotic' interchangeably, together signifying some kind of semiosis in sociocultural life for the living and conduct of meaning. Summarily, we use the words 'talk' and/or 'discourse' as a means to embrace all them, but not in the service of a linguistic model or logocentricity. We appropriate them to gloss a semiosis whose medium of exchange is multimodal, and not simply spoken.⁶

     The Csordas-Jackson paradigm presupposes that these forms of social theoretical discourse are separate and alternative approaches to embodying social scientific theory. There is even the suggestion, and at times more than that, that the discourses are incommensurable. More specifically, we can observe that, in traditional disembodied social theory, there is talk about the observed body from an objectivist intellectualist standpoint (for example, symbolic/structural anthropology, psychoanalysis, Durkheimian sociology). In the predominant dissenting tradition of embodied social theory in the first somatic revolution, there is talk of the experienced body from a subjectivist lived standpoint (for example, the Jackson-Csordas paradigm). Finally, in dynamically embodied social theory, there is 'talk' from the moving body (an agentist enactment standpoint). Here we have the basis from which we can better identify the first somatic revolution in social science theory. The Csordas-Jackson paradigm was a revolt against the deterministic reduction of the human body to a mechanical system: behaviorism, psychoanalysis, and naturalistic sociology were different ways to theorize that reified conception of human somatics.

     Williams and Farnell propose instead that we conceptualize the three forms of body-referenced 'talk' as complementary moments of everyday social symbolic interaction (see Farnell 1994). Each of the three moments can now be regarded as situated options that persons may take up in reference to themselves and/or others as they contextually see fit, according to their ordinary and/or professional interests. The distinctions are illustrated via the following example:

	1. "I hit my thumb with the hammer." [Talk about the (objective) body] 2. "Wow—I nearly fainted from the pain!" [Talk of (the experience of) the body] 3. [Talk from the (moving) body: a transcription of a person hitting his thumb and reacting, including facial insert expression]                                      Figure 1. Three complementary situated options for body-referenced talk.

    Central here is the idea that the way human agency works is in terms of the signifying enactments of moving persons. The varied discursive practices of semiosis are performatively grounded in, and conventionally a structuring of, a suitable region of the mindful body that serves the purposes of meaning-centered sociocultural living—such regions as the mouth and lips in speech, the hands in sign languages, and the whole body in forms of dance, ceremony, or practical skills of various kinds (see Farnell 1999), in interactive and dialogic contexts as well as past, present, or future temporal frames. The human actions that constitute speech-act systems, action-sign systems, and any other form of semiosis are the creative outcome of a primary generative act—signifying enactments from the body (Farnell 1999, Williams 2003).

     The main thesis of our paper is that the paradigm proposed by Williams and Farnell is one way to realize the full significance of Merleau-Ponty's theoretical view of the human character of embodied consciousness. Another fruitful way to realize Merleau-Ponty's theoretical goals can be found in Ingold's "dwelling perspective" (2000). We contend that it is not enough to focus on the feeling of the doing, as is the case in the Csordas-Jackson paradigm, because Merleau-Ponty himself was especially interested in the doing of the body. To illustrate how this is the case, we briefly examine the concept of 'flesh' found in Merleau-Ponty's posthumous work The Visible and Invisible (1968).

Flesh: Embodiment in a New Key

     For Merleau-Ponty, overcoming Cartesian dualism (and its supposed philosophical heir, the Kantian transcendental ego) was the heart of the matter in the revolt against the intellectualism that gives us the objective body.⁷ Merleau-Ponty gave us, in effect, a post-Cartesian slogan, namely, instead of "I think, therefore I am," we have, "I can, therefore I am." We have here an incisive Faustian touch: in the beginning is the embodied act not the disembodied word or mind.⁸ This dispositional concept of the 'I can' clearly refers to the idea of agency as a power, but this is not the old-fashioned power of the will. In The Visible and the Invisible (1968) in which the new concept of 'flesh' is introduced, Merleau-Ponty makes a sustained effort to realize a key change in the philosophy of embodiment. Consider the following various characterizations of 'flesh.' It is, he declares,

"A general manner of being" [that is] (147)

"A power that is not a factual power" [thus] (160–61)

"A center" [that is] (145)

"A node of properties . . . [an] internal arrangement . . . extant by its own efficacy" [and so we have] (160–61)

"A pregnancy . . . a power of fecundity" [which is] (208)

"The formative medium of subject and object" (147).

If we are to appreciate what Merleau-Ponty is after here, we must first set aside the positivist rejection of the concept of causal relations between entities, with its preference for the concept of the mere regularity of events. We can replace it with a realist understanding of science in which causation is not merely correlation but rather refers to causal entities as "powerful particulars." In action in the real world, causal relations are forceful particulars at work. Causation is, as Merleau-Ponty would have it, a power of fecundity or, as it is ordinarily expressed in the philosophy of science, the power of production. Efficacy is making things happen. And, in nature, there are different kinds of efficacious things: physical, chemical, biological, and, yes, human.

     Consider two ideas in The Visible and the Invisible that would seem to underwrite the theme of the six quotations listed above, that 'flesh' refers to the primordial reality of causal powers in nature. The first and central idea is seen in the remarkable comment by Merleau-Ponty that,

[i]n a sense, if we were to make completely explicit the architectonics of the human body, its ontological framework, and how it sees itself and hears itself, we would see that the structure of its mute world is such that all the possibilities of language are already given in it. (1968: 155; emphasis provided)

The ontological framework is perhaps the very idea of human flesh, that human bodies are agentic persons. In this regard, note Merleau-Ponty's second idea:

But the actual body I call mine [is] this sentinel standing quietly at the command of my words and acts. (1968: 254; emphasis provided)

We have here the proposal that the embodiment of agentive persons enables them to use their physical being (or to use themselves)—via the acquisition of techniques, skills, and rules—to move meaningfully throughout the worlds of nature and culture. Contrary to Merleau-Ponty's way of stating it, but in keeping with the import of what he says, we can phrase this as follows: instead of saying "the subject is his body," we could say "the subject becomes his/her body." Initially, the subject is simply the organism, for, indeed, since the 'body' is the indexical site of the person (Urciuoli 1995, Varela 1995), developmentally, the organism becomes a body as the individual is becoming a person. Harré has brought all of these considerations together, particularly the connection of discourse and causal powers, in this statement:

The conversational world [of embodied persons] like the physical world evolves under the influence of real powers and forces, dispositional properties of the utterances that are the real substrate of all interchanges. (Mühlhäusler and Harré 1990: 24)

We interpret the term "utterances" here to refer to multimodal resources, including not only words but also gestures, facial expressions, torso movements, and other physical actions.

     It is revealing and intriguing to note that in The Phenomenology of Perception, Merleau-Ponty declared that "[t]here is no experience without speech as the purely lived through has no part in the discursive life of man" (1962: 337). Madison, the eminent Merleau-Pontian scholar, has indeed claimed that Merleau-Ponty abandoned the primacy of perception for the primacy of language. He sums up this change by saying, "Being that can be perceived is language" (Madison 1992: 92). It is, therefore, not the case that the Cartesian claim that 'minds intend' can be overcome by the phenomenological counterclaim that it is 'bodies' which 'intend' (the alleged "purely lived through"), not ethereal 'minds.' We would argue that neither minds nor bodies intend, only people do, because, as embodied persons, they are causally empowered to engage in social and reflexive commentary with the primary resources of vocal and kinetic systems of semiosis provided by their cultural ways of being human.

     Merleau-Ponty's last work allows us to assert with some confidence that Williams's principle of the primacy of the signifying moving person is theoretically consistent with the notion of flesh as that "primordial pregnancy," nature's "hidden powers of vegetation" (1968: 9), as summarized above. Farnell, Williams, and Merleau-Ponty are all emphasizing the embodiment of the doing and not simply the feeling of that embodied doing. The upshot of this special theoretical focus is the understanding that human physical being is moving being. It is in this precise sense that it can be said that the paradigm of dynamic embodiment stands today as one way to realize the full significance of Merleau-Ponty's philosophy of embodied consciousness.

Part II: Kinesthetic Sense and Dynamically Embodied Action

A second somatic revolution can thus be said to reside in a theory of dynamic embodiment—a theory of Moving Being—a principle of which is that the somatic is necessarily semiotic. In this section of the paper, we will try to link the somatic and the semiotic via the notion that semiosis is multisensory.

     The familiar Western taxonomy of the five senses, in which vision is accorded pride of place as "the noblest of the senses," has a venerable history going back to Plato and Aristotle. As several scholars have noted (for example, Classen 1993, 1997; Herzfeld 2001; Ingold 2000; Seremetakis 1994; and Stoller 1989), vision is closely followed by hearing, both of which are deemed superior to the lower, more animalistic "contact" senses of touch, taste, and smell. As Herzfeld (2001) notes, this hierarchy was readily mapped onto nineteenth century evolutionism in both popular and scholarly thinking in the West, as the racist tendencies of an earlier anthropology associated the "lower" senses with the "lower" races. For example, the early nineteenth century, pre-Darwinian natural historian and embryologist Lorenz Oken mapped this sensory hierarchy onto the conventional racist ordering of human groups in a 'taxonomy by fives' (Gould 1985: 204–5) that placed the European "eye-man" at the top, followed by the Asiatic "ear-man," the indigenous American "nose-man," the Australian "tongue man," and the African "skin-man":

1. The skin-man is the black, African

2. The tongue man is the brown, Australian-Malaysian

3. The nose man is the red, American

4. The ear man is the yellow, Asiatic-Mongolian

5. The eye-man is the white, European

This was elaborated into a classification of the entire animal kingdom according to the same principle of the five senses:

	Table 1. Oken's Path of Progress by Wheels of Five (after Gould 1985: 206)9

Kinesthesia

     Notably absent from this conventional taxonomy, however, is kinesthesia, our sensory awareness of the position and movement of the body. We ask the reader to please close your eyes, lift your arm—move it around and ask yourself how you know where your arm is located. This is kinesthesia: literally, 'movement' (kinetic) + 'sensitivity' (aesthesia). It is this kinesthetic sense that provides information on the whole repertory of our motor actions, from the raising of an arm, to walking, even to the turn of the eyeballs and swallowing. Physiologically speaking, (that is, in the discourse of Western natural sciences), kinesthetic sensations are registered by receptors in the muscles, tendons, and joints of the body. As the muscles function when we move bodily parts, various patterns of pressures on these receptors provide essential information for the guiding of motor action.

     The perception of spatial movement and orientation of the body as a whole also involve a fluid-filled receptor system located in the vestibules of the inner ear. More than balance, this structure provides the means by which we are aware of being tilted, shaken, or whirled about, and how, most of the time, we know "which way is up"!

     The exclusion of kinesthesia from the Western taxonomy of the senses—this (ab)sense, as it were—is particularly interesting because scholars of perception as diverse as Descartes, Dewey, Gibson, and Merleau-Ponty all acknowledge body movement as the unexamined ground of all sensory perception. One is led to ask why, then, has kinesthesia been excluded from consideration?

Theories of Perception

     If the senses are the means by which we experience the world, then any theory of the senses assumes a theory of perception by means of which such experience is possible. The classical two-stage representational theory of perception—for example, that of Descartes—contains a foundational but unexamined assumption that perception is built out of sensations (Harré 1986a: 147). In the first stage, a causal relation is supposed to obtain between a world-state and a sensation. In the second stage, the sensation is reworked in some cognitive process to yield the percept—a mental awareness. This representational tradition thus institutionalizes the separation of inside/outside, mind/body, and reason/feeling. Harré concludes that foundational to four centuries of perception theory is the notion that "percepts are cognitively transformed sensations and the basis of perception is an awareness of states of the brain that are the remote effects of physical causes" (Harré 1986a: 155).

     In contrast, James Gibson (1966, 1979) provides us with an anti-Cartesian ecological approach to perception. As Ingold has succinctly summarized, Gibson argues that, instead of

thinking of perception as the computational activity of a mind within a body we should think of it as the exploratory activity of the whole organism within its environmental setting in active participation through practical bodily engagement. As such it does not yield images or representations. It rather guides the organism along in the furtherance of its project. The perceptually astute organism is one whose movements are closely tuned and ever responsive to environmental perturbations. (Ingold 2000: 260)

Such a conception situates bodily action at the heart of our being-in-the-world, rather than merely a means to mental representations of the world.

     Merleau-Ponty's position accords with Gibson's in many respects, but Merleau-Ponty takes this one step further back by positing our immersion in a preobjectively given life-world that is ontologically prior to perceiving objects in the environment. As Ingold so aptly puts it, "[T]he world of our experience is a world suspended in movement that is continually coming into being as we—through our own movement—contribute to its formation" (2000: 242).

     Despite this common acknowledgment of body movement as the ground for the very possibility of experience, absent from the table is any discussion of bodily movement in and of itself as a sensory modality, and therefore as a potential resource for meaning making or semiosis. Perhaps, as beings "continually on the move actively exploring the environment in the practical pursuit of [our] life in the world" (Ingold 2000: 261), our own bodily movement has become an unexamined common sense, its very familiarity conspiring to hide it from us analytically. More than this, however, its omission stems from metatheoretical problems with a viable concept of embodied personhood as dynamically embodied.

     Problems that stem from the Platonic–Cartesian legacy continue to arise in attempts to embody social theory and practice according to the philosophical foundations of the first somatic revolution. This is especially salient when addressing the senses. For example, during a recent discussion of the responses a person might have to viewing a powerful work of art or a museum exhibit, an anthropological colleague considered it unproblematic to say that the emotional charge of such objects operates "presemiosis." "Objects," she said, "cannot be reduced to what they can be said to signify—to do so is a semiotic reduction." This statement reflects Michael Jackson's (1983) position in which he rejects semiotic processes as necessarily representational, (formally) cognitive, and linguistic, in favor of a phenomenologically inspired 'radical empiricism' wherein sensory experience and perception are thought to afford a pre- or nonlinguistic, precultural mode of experiencing the world. We are thus left with a residual positivism grounded in empiricism/experientialism. This formulation does not transcend the problem of Cartesian body/mind dualism; it merely entrenches the bifurcation, by swinging the pendulum over to 'the body' (see Farnell 1994, Varela 1994, 1995).

     Thomas Csordas (1990) moderates Jackson's position with the important corrective that Merleau-Ponty's concept of 'preobjective' does not mean 'precultural' or 'prelinguistic,' but rather 'prereflective'—not thought about. In Gilbert Ryle's (1949) terms, this would be "knowing how" rather than "knowing that." However, Csordas likewise limits the concept of semiotic to representational signs and symbols, which, he maintains, reduces embodied experience to language, or discourse, or representation (Csordas 1999: 183). He proposes that we embrace Merleau-Ponty's preobjective being-in-the world as a dialogical partner to representation: "The equation is that semiotics gives us textuality in order to understand representation, phenomenology gives us embodiment in order to understand being-in-the-world" (1999: 184). In so doing, Csordas seems to accept the dualism on which the separation of a representational mind from an experiential body is predicated. Csordas's work thus remains rooted in the spirit of the Cartesian tradition, although that is certainly not his intent.

     In a similar vein, Johannes Birringer, in his book Performance on the Edge (2000), says of a powerful performance, "[W]e find [it] impossible to grasp, except emotionally, viscerally as it sometimes happens when we witness a powerful performance we don't understand but for its bodily and affective impact on us." Birringer thereby restricts what we might mean by 'understanding' to one type only—to the self-conscious, theoretical articulations of a propositional kind of which we, as language-using creatures, are capable—again, Ryle's "knowing that." We would want to argue instead that "knowing how" surely involves knowing how to respond emotionally. This capitulation to the old logical positivist strictures around what will count as 'knowledge' or 'understanding' unwittingly perpetuates mind/body dualism by privileging its second half. We will try to show that these contributions fail to articulate an adequate concept of embodied personhood because they presuppose impoverished notions of semiosis and language. At the heart of the problem is limiting the concept of 'sign' (and therefore semiosis) to the representational, that is, as "standing for" something outside itself.

     Our proposed solution to the problem of whether sensory modalities operate prior to, or separate from, semiotic meaning making is simply to dispense with the dichotomy and the concept of personhood upon which it is predicated. Instead of restricting semiosis to representational signs and symbols, we propose a multisensory semiosis loosely defined as processes of agentic embodied meaning making afforded by the modalities of taste, hearing, touch, pain, smell, sight, and kinesthesia in various relationships with talk and other bodily action. The post-Cartesian move is to view such somato-sensory semiotic modalities as providing human beings with resources for meaningful action that frequently elide spoken expression, but which are never separate from the nature, powers, and capacities of linguistically capable agents (Williams 1982, Farnell 1999). In addition to an anti-Cartesian theory of perception mentioned earlier, this move requires an updated and enriched nonrepresentational view of language and semiosis, together with a concept of 'sensory act.'

A Wittgensteinian Move

     In contemporary linguistic anthropology, the nonrepresentational view of language articulated by the later Wittgenstein, in addition to discourse-centered approaches to culture (see Farnell and Graham 1998) have developed or considerably modified concepts of language that gave rise to Peircian and Saussurean semiotics. An important development in this line of inquiry was to separate what Silverstein (1976) called the "semantico-referential" function of speech—the naming function—from a "representational" theory of language in the sense of inferring accompanying mental representations or images. Current work in semiotically informed linguistic anthropology recognizes that the "representational" or "semantico-referential" function of vocal signs is only one among many. The same is true of signs in other modalities such as Williams's action signs (1975, 2003). We can thereby relieve meaning from being fixed to a semantico-referential function (that is, symbols that name or stand for something) and add creative and presupposed indexical aspects of sign functions into the analytic frame (see Urciuoli 1995, 1996). A contemporary approach to semiosis would also find meaning in 'contexts of use' and 'dialogic recontextualization' (see Duranti and Goodwin 1992). This allows the kinds of practical activities of special interest to Bourdieu (1977), for example, to be included in the realm of the joint construction of social action as signifying acts of embodied discursive practices.

     Let us take ordinary walking as an example. The mundane activity of walking is not an action sign that "stands for" anything outside itself—it does not normally carry semantico-referential meaning. But that does not make it meaningless. To argue in a behavioristic manner that "I'm just walking; it doesn't mean anything" is to decontextualize the act and reduce action to gross physical movement (Best 1978). One may be walking across the road to the post office, or on the way home, or for exercise, or for the sheer joy of walking in the afternoon sunshine (because be-ing matters—it is a human value). All these actions are semiotic in the sense of being meaningful, intelligent activities (Ingold 1993). Walking as an 'action sign' thus takes its meaning from the social and physical context in which the walking occurs, from its place within a system of signs (as Saussure emphasized). Action signs, like vocal signs also take part in deictic (space/time) reference, indexicality, and performativity. These are, in turn, embedded within larger performance spaces of all kinds (for example, living spaces, village plazas, courtrooms, and so forth). They are also related in numerous ways and at several levels to other action signs.

     One's walking may also carry indexical meaning—the way I am walking may index my gender, age, class, sexual orientation, or ethnicity. Since styles of walking are shaped socially, as Marcel Mauss observed, others can use the way I walk to position me socially, as I can use it to position myself. Although walking is normally outside one's focal awareness, it is always available for focal attention if necessary. In Northern Ireland, for example, careful reading of the walk, posture, eye gaze, and clothing of other persons (a practice called "telling") determines whether a person is identified as Catholic or Protestant and therefore evaluated as someone worthy of "talk" (that is, social interaction), or not. As Bill Kelleher (2003) notes, in this tension-ridden context, attention to ways of walking and accompanying bodily practices has become important. When social borders of any kind must be crossed, it seems that habitual actions take center stage instead of remaining out of awareness. We maintain that this kind of reconceptualization of semiosis can usefully apply to signifying acts in all kinds of modalities other than speech, without reducing embodied experience to propositional language or ignoring prereflective aspects of "knowing how."

     The Wittgensteinian philosopher of human movement David Best reminds us how readily we fall into Cartesian traps in our discourse when he says, "[T]o describe an action as thoughtful is not to say that the physical behavior is accompanied or preceded by an inner mental event: it is to describe the kind of action it is" (1992: 201). Active engagement in any activity is thinking, which is not to say that one cannot also be reflective and think about the activity when one is not engaged in it.

     Likewise, to describe a sensory experience as meaningful is not to say that the physical sensory response is accompanied by an inner mental event or external signified that is its significance; it is to describe the kind of sensory experience it is. Active engagement in sensory experience is meaningful. The signifying here is not some semantico-referential meaning outside the sensory act; it is meaningful because it is understood at some level, and therefore a semiosis—a meaning-making process—is at work. Sensory acts make sense without necessarily being thought about—that is, engaging in reflective, abstract, critical, propositional, or theoretical thought.

     This formulation retains the spirit of Merleau-Ponty's "preobjective," without getting tangled in problematic subject/object talk.¹⁰ This is not to say that one cannot also be reflective and think about the meaning of sensory experience either at the time or later. It is also worth remembering that, in the midst of social interaction, spoken discourse too is most often used without thinking about it.

     Gibson and Merleau-Ponty both connect the sensory with action. This is captured in Merleau-Ponty's statement that "my gaze, my touch and all my other senses are together the powers of one and the same body integrated into one and the same action" (1962: 317–18). He spoke of the bodily synergy of the senses in their convergent striving toward a common goal. It would be a mistake then, to separate kinesthesia as a sensory experiential 'feeling of doing,' from bodily movement as physical action, for how can one act purposefully without experiencing the position of one's body parts and the dynamic feeling of doing that informs and assigns meaning to the action? "Knowing how" to engage in action requires skills that may be out of focal awareness, once learned.

     For example, when learning a new phrase of danced movement from an Egyptian dance I am studying, I might ask the teacher to "explain" how to perform a subtle hip action that I have observed and tried to perform but cannot yet reproduce accurately. This distinct action sign has no name in this dance tradition. My teacher says, "It goes like this," as she repeats the action more slowly and carefully, adding "See, it's this [pointing gesture] part of your hip leading—yaam da da, yaam da da," and the syllables create a rhythm that echoes the dynamics and timing of the action as she performs it again before I try once more. Our point is that there are very few spoken language concepts involved here, but what is going on is not preconceptual, prelinguistic, prereflective, or "representational" Why? Because I have had to focus my attention (my kinesthetic awareness, not my eyes) on the front side of my hip and learn to carve a shape in the space surrounding it with that part of the hip. I've had to "draw" two horizontal circles clockwise in space. However confusing the process may sound in words, this is a person acting, not a mind thinking while the body experiences, and this point cannot be overstressed (Farnell 1996a: 318). Once learned, my performance of the actions no longer requires my focal awareness.

     Several forms of sensory awareness are interwoven in any action. Digging, for example, requires perception of things in the environment—seeing and touching the spade one is picking up in order to dig. There is kinesthetic awareness of one's body and bodily movement in one's conscious action of picking up the spade. This is most likely to be an out-of-focal awareness of one's acting because attention will probably be focused on where and how one is going to act with the spade, that is, on the soil. An experienced gardener may also pay attention to the smell, color, texture, and even taste of the soil. Also required is a cultural understanding of the activity of digging and some learned skill. For example, Mauss observed that, during the first World War, English troops did not know how to use French spades, a fact requiring eight thousand spades to be changed whenever the French troops were relieved, and vice versa! These multisensory forms of awareness, typical of all skilled action, cannot be reduced to the others, whether reducing action to cognition or cognition to action or "experience," and none is foundational for the others (Woodruff-Smith 1988). The multisensory semiosis at work when action signs (gestures) and vocal signs (speech) complement each other is well illustrated by the following ethnographic example of a metaphorically laden conversation.

Ethnographic Case 1: Talking from the Body

     Benjamin Lee Whorf, the famous American linguistic anthropologist, said of English speakers, "[W]e are more apt to make a grasping gesture when we speak of grasping an elusive idea than when we speak of grasping a door knob" (1956: 157). In identifying spatialized metaphors as an organizing principle in English and other European languages, Whorf notes the way gestures integrate with such spoken metaphors. English, he observes, is a language that systematically turns abstract concepts about intangible matters such as 'time' or 'ideas' into nouns, which then are handled discursively as if they are tangible. Thus, we frequently talk about ideas as if they are physical objects—I can "hold several ideas at once," I might "pass some of my ideas on to you," and I can "twist your ideas around," and so on.¹¹ As Whorf puts it,

Many of the gestures made by English speaking people serve to illustrate, by a movement in space, not a real spatial reference but one of the nonspatial references that our language handles by metaphors of imaginary space. The gesture seeks to make a metaphorical and somewhat unclear reference more clear. (1956: 157)

     We see this principle at work in the first segment transcribed in Figure 2, as an American English-speaking professor, during an informal interview with two students, employs vocal signs to speak of "conceiving of a project." This is accompanied by an action sign in which she uses both hands symmetrically to make a small horizontal circle in the space just in front of her head.¹² She repeats the same action sign in the next sentence when she says "having an idea." The two intangible nouns 'a project' and 'an idea' have been metaphorically transformed into visible, tangible objects, metaphorically enclosed within the circular space circumscribed with her hands. Just as Whorf observed, she has used body movement in corporeal space to create a visual representation of two nonspatial, nontangible referents—'a project' and 'an idea'—that English speakers conceive of in spatialized terms.

	Figure 2. An American English-speaking professor gestures as she talks informally to her students about doing field research.

     In addition, the movement path of the action sign—this tracing of a circular pathway through space with both hands—is iconic of the notion of process involved in the verbs 'to conceive' and 'to have an idea.' The structure of the action sign thus simultaneously mirrors an English speaker's commonsense metalinguistic notion of basic language structure—that nouns are 'things' and verb are 'doing words.' The moving hands are iconic of the verbs (to conceive, to have), while the nouns become the imaginary objects thus circumscribed (a project, an idea).

     A similar integration of vocal signs and action signs happens later in the narrative when the speaker says "fit into . . . your interests" (Figure 2). Her left hand becomes a metaphorical container for the intangible noun (the interests), while, with her right hand, she takes up the action of the verb and spatial preposition "fit into" by moving her hand back and forth as if stuffing the container with 'interests.'

     A later phrase in the narrative illustrates how English speakers structure concepts of 'time' through spatialized metaphors. We 'objectify' time, another intangible of the experienced physical world, by using a noun form—'the time'—to which we assign the properties of length and substance. Hence, we speak of a 'long' or 'short' time and divide its 'length' into units we call weeks, days, hours, minutes, and seconds. We talk about 'not having enough time,' of spending and wasting 'it,' and so forth. Consistent with this principle, the professor uses vocal signs to say "it was the first extended time I'd spent in Korea." These words are accompanied by an action sign in which she draws a horizontal line with both hands across the space in front of her torso. Palms are facing each other as the hands separate, moving to left and right sides as they create a visual representation of 'time' as a length.

Semantic and Pragmatic Functions

     One is led to ask why the speaker adds visual-kinetic material to the utterances and what functions these action signs serve when the words alone would seem adequate to convey the semantic content. Answers to these questions emerge if one looks at the semantic and pragmatic functions of the utterances. Since what is defined is only accessible through metaphor, these are not new descriptions of something previously discerned: they create the meaning. The professor creates the metaphorical vehicle of these spatialized metaphors within tangible space. This assists the speaker's clarity of conceptualization by making the intangible visible in the physical, corporeal space in front of her own body. She actively shapes this dynamically embodied space, acting the verbs and making the objects (nouns). This is experienced largely through her kinesthetic sense, since we don't usually watch our own gestural production (Euro-Americans usually look at co-participants, as the transcript shows). For the co-participants in this speech event—the students—the action signs provide visual references that likewise lend support to and clarify the meaning of the spatialized metaphors. As Whorf puts it, "The gesture seeks to make a metaphorical and somewhat unclear reference more clear" (1956: 157). The action signs thus provide visual-kinetic metaphors that complement the vocal metaphors and vice versa, in a semantic gestalt that functions as a pragmatic aid to both conceptualization and communication.

The Spatial Location of Gestures and Cultural Concepts of Body/Mind

     An additional constituent feature of the action signs in Figure 2 is their spatial location in relation to the rest of the professor's body. She locates her action signs close to her head, thereby utilizing the English speaker's conventional notion of where in the body 'thinking' is located. It is interesting to note in passing that signs for know, think, understand, and idea in American Sign Language, used by members of the Deaf community in the United States, are also located close to the head, corresponding closely to the coexpressive actions signs employed by American English speakers (see Farnell 1995: 252). That such a location is a cultural construction can be illustrated through a comparison with utterances from Nakota (Siouan) speech events in which a Nakota sign talker utilizes similar metaphorical content about person attributes, but with interesting contrasts in the corporeal location of action signs relating to 'thinking' and 'mind' (see Farnell 2001: 409–13).

Ethnographic Case 2: Cashinahua Concepts of Person

     Finally, a brief summary of Kensinger's (1991) ethnographic account of Cashinahua (Peru/Western Brazil) personhood and knowledge provides an example of an alternative taxonomy of the senses and embodied personhood that helps stretch the Western imagination as to how an agent-centered, sensory semiosis might operate as dynamically embodied action in a non-Cartesian environment.

     Kensinger embarked upon an ethnographic quest into what counted as "knowledge" among the Cashinahuas—its location, constitution, and acquisition. A wise man, he learned, has knowledge (una) throughout his body—"his whole body knows," they say. Una is that which one's body learns from experience. When asked where specifically a wise man had knowledge, Kensinger's consultants listed his hands, his skin, his eyes, his ears, his genitals, and his liver. When asked, "Does his brain have knowledge," they responded, "It doesn't."

Hand Knowledge. All knowledge associated with physical labor is located in the hands because they are the body parts most directly involved in work. Kensinger explains:

[W]hen a man chops down a tree to clear a garden in the forest, he learns something about the nature of the tree and of his tool, about the force needed to make the cuts and the direction in which the tree falls with reference to the placement of the cut and about whether it falls cleanly to the ground with reference to the surrounding brush and trees, and more. (Kensinger 1991: 39)

This knowledge resides in the hands, say the Cashinahua, because they held the axe that cut the tree, causing it to fall and thus are the conduit by which the knowledge entered the body.

     Knowledge learned by and associated with men's hands involves hunting; fishing; and making gardens tools, bows and arrows, feather headdresses, and other objects. Women's hands know planting and harvesting gardens, cooking, weaving, and making baskets, pottery, and other objects.

Skin Knowledge. Besides hand knowledge, successful hunting also involves knowledge of the behavioral characteristics of the animals hunted, based on observation. Contrary to our expectations, this is classified by the Cashinahua as "skin knowledge," as is all knowledge of the natural world. One learns about things like sun, wind, water, and rain through the sensations they produce on the surfaces of the body. It is in this sense that knowledge of the natural world is skin knowledge.

     When Kensinger asked why knowledge of animal behavior was not eye knowledge, since it came from observation, he was told that it was knowledge of the jungle's "body spirit" (yuda bake yushin). This opened up a whole second classificatory system that constitutes the Cashinahua notion of person, according to which every human being consists of a body (yuda), plus a series of at least five spirits. Although Kensinger's consultants disagreed on exactly how many spirits a person has, they all listed at least the following:

Yuda bake yushin—body child spirit

Bedu yushin—eye spirit

Nama yushin—dream spirit

Pui yushin—faeces spirit

Isun yushin—urine spirit

Kensinger discusses only the first two in his account. We learn that the "body child spirit" encases a person's body like an outer skin. It is not really visible—it consists of a person's aura, an indicator of the state of a person's vitality and health, or lack thereof, and a person's sheer physical presence. Although ephemeral, intangible, and invisible, the body takes on a different aura in the absence of the body child spirit, as when a person dies. Yuda bake yushin also refers to a person's reflection in water or a mirror, as well as a person's shadow. All living things, including people, animals, vegetation, and all other aspects of nature are said to have "body spirits." When one sees a person or thing, one can be said to see its body spirit.

Eye Spirit—Eye Knowledge. To see the true nature of people and the things that make up the natural world, however, one must also understand the bedu yushin—eye spirit, sometimes also called the "real spirit" (yushin kuin). The eye spirit dwells in a person's eyes, leaving the body during unconsciousness and hallucinogenic experiences to travel in the world of spirits. The knowledge gained in these travels is called bedu unaya—eye knowledge. It is only with the eye spirit that one can truly and fully see persons or objects in both their physical and spiritual substances, that is, their bodies and body spirits. Without the eye spirit, a person can only know the surface of things, that is, their skin and thus skin knowledge.

Ear Knowledge. Social knowledge is gained though and resides in the ears, a connection which comes from the centrality of language in social discourse; although speech (hancha) comes from the mouth, knowledge comes from hearing. There are two kinds of hearing, soft and hard. Soft hearing involves listening and absorbing facts about social matters—social awareness. Hard hearing requires digging beneath the surface to consider motivations, consequences, and so forth. Although both kinds of hearing involve knowledge, it is principally hard hearing that is involved when they say a person knows a lot or that they have much ear knowledge. Hard hearing results from both listening and thinking. Social misfits and persons who flaunt social conventions are said to be "deaf "or to "have hard ears" or "his ears are without holes"—they are people without ear knowledge.

     Kensinger was never able to find out where 'thinking' takes place. Several people said it takes place within the ears; others located it in the heart, the liver, or the whole body. Others found his questions incomprehensible or silly. Although he pressed the question, his informants consistently rejected the brain (mapu) or the place between the ears as the locus of thought.

Genital Knowledge. For the Cashinahua, the genitals are the locus of knowledge of mortality and immortality, of the life force. The sexual act is brief and fleeting, they explain, but through it one reproduces oneself. Children are the product of the genitals and genital activity and give one immortality by enduring beyond one's own lifetime.

Liver Knowledge. The liver provides knowledge of emotions. It is considered the locus of feeling joy, sorrow, fear, distrust, hope, and pleasure. A generous, pleasant person has a "sweet liver" or "his/her liver knows a lot"; a stingy person with a nasty disposition who always is gloomy and foresees disaster has a "bitter liver." A person with a bitter liver only knows a little. Liver knowledge is expressed in behavior and demeanor on the surface of the body. A happy disposition produces a "sweet face"; a grumpy disposition, a "bitter face." One can say of a person "Her liver has a lot of knowledge. Her face is very sweet. Her whole body is very sweet; it always makes us very happy."

     Cashinahua consultants consistently rejected any separation of mind and body. They insisted, instead, that different kinds of knowledge are gained through, and reside in, different parts of the body. In sum, a wise person is one who has a lot of una (knowledge): their hands know (they are skilled workers); their skin knows (they have an extensive and intimate knowledge of their physical surroundings). Their eyes give them knowledge of the spiritual world. Knowledge of their mortality and immortality resides in their genitals. Their liver provides them with the full range of emotions. A truly knowledgeable person is one whose whole body knows. Knowledge is derived from activity and in turn generates activity. It is in action not contemplation that knowledge is both gained and given expression. A wise Cashinahuan person is not only one who knows based on past experience but also one whose knowledge continues to increase as it is put into action. Knowledge is alive—it lives and grows in a body that acts, thinks, and feels.

     Part Two of this paper begins and ends with reference to contrasting taxonomies of the bodily senses and personhood in order to draw attention to the fact that our actions and experiences are shaped by such socially constructed, often normative, pretheoretical assumptions. To include them in analyses is not to fall into the trap of separating 'ideas' from 'action' but to recognize the interdependence between "knowing how" and "knowing that."

Concluding Remarks

We have suggested that, in the social sciences and prior to the first somatic revolution, the body was conceived, for the most part, as a deterministic object. For example, in the case of classical Freudian psychoanalysis and 'the unconscious,' agency was given to biology as the psychological unconscious, and not to the person; in the case of classical sociological theory, agency was given to society rather than the person. This was classic social scientific "talk about the body."

     The first somatic revolution, beginning in the 1980s and exemplified in the work of Csordas, Jackson, Turner, Shilling, and others, recovered a discourse of the feeling of the body, inspired by Merleau-Ponty's existential phenomenology. This we have called "talk of the body."

     We have proposed a second somatic revolution, beginning with Williams's dissertation on the 'anthropology of human movement' in 1975 (to date, unpublished). This introduced the fundamental conception of what we are calling "talk from the body" and the concept of "a paradigm of dynamic embodiment." A special feature of this paradigm is that the three forms of body referenced "talk" (about, of, and from) are complementary moments of everyday human action (dialogic, interactive, and temporal): the three moments are situated options that persons may take up as they contextually see fit.

     Part One of our paper illustrates how Williams's principle of the primacy of the embodied, signifying person is theoretically consistent with a deeper reading of Merleau-Ponty's conception of "flesh." In that combination, we have a theory of embodiment that is post-Cartesian, a dream that was the explicit though unfulfilled intention of Merleau-Ponty's last work, The Visible and the Invisible (1968).

     Part Two takes issue with the Cartesian residue from the first somatic revolution found in the separation of semiotic from somatic. Using ethnohistorical data from a Western taxonomy of the senses, ethnographic data from Euro-Americans in the United States, and taxonomic data on personhood/senses from the Cashinahua people of Brazil, we illustrate how the semiotic can be usefully liberated when necessary from a conflation with representational and/or the symbolic. We have shown how the semiotic can indeed be somatic through dynamically embodied acts, and how the somatic is necessarily semiotic when it involves the agentic, meaning-making practices of social persons as they move about. The signifying here is not some semantico-referential meaning outside the act; it is meaningful because it is understood as such by the agent, and, therefore, a semiosis is at work.

     We have also argued that the social sphere is most often multisensory and predicated upon dynamic embodiment, that is, body movement as both speech and action that are enacted forms of knowledge and understanding. We maintain that such dynamically embodied signifying acts in symbolically rich spaces are the dialogical, intersubjective means by which persons, social institutions, and cultural knowledge are socially constructed, historically transmitted, and revised, and so are constitutive of culture and self.