A Radical Empiricist Theory of Speaking: Linguistic Meaning Without Conventions

Abstract

I propose an account of speaking that draws on radical empiricist philosophy, which conceives of the environment as a network of relations. The argument takes the following form: (a) affordances are relations that exist between an animal and structure in its environment; (b) these affordance relations exist even when the animal is not attending to them; (c) these relations are themselves public and can in principle be perceived by an observer; (d) in the case of language-using humans, these relations can be directly acted upon by other speakers; and (e) one tool for acting upon another person's web of relations is linguistic structure. This description scheme offers a way to incorporate linguistic meaning into the framework of ecological realism while avoiding any notion of conventional meaning. Speech is conceived not as the production of messages to be decoded but as action controlled with reference to relational properties of the environment; the function of verbal actions is to influence the behavior of others in ways that are adaptive to the speaker's purposes. It is argued that the shape of the system of relations must result from the personal learning history of individual language users and therefore that early word learning is the appropriate place to begin an ecological analysis of speaking.

Acknowledgments

I am indebted to Carol Fowler, Eric Charles, and an anonymous reviewer for their invaluable comments on an earlier draft of this article.

Notes

¹ One reason for this difficulty may be that Gibson's (1979/1986), theory of affordances, to the extent that he was able to develop it, maintained a crucial ambiguity about the status of the social. Costall (1995) asserts that Gibson was attempting to maintain two conflicting ideas: on the one hand, that meaning is a relational property arising from the mutuality of animal and environment (and here it is acknowledged that the environment encompasses other actors), but on the other hand, that in order to avoid slipping into cultural relativism it is necessary to assert that the asocial physical structure of the environment is primary. This ambiguity seemingly threatens to relegate the social to the status of secondary property.

² It must be acknowledged that the term “convention” is used in a great variety of ways. For instance, Millikan (2003) used the term to denote patterns of activity that span multiple individuals; on this scheme a complete exchange can be said to constitute an instance of one single convention, for example, the phrase “Can you pass me the bread” followed by the complementary action of a second person handing a loaf to the first. I confine myself to attacking only the notion of convention as a relation between a pattern in sound and an event in the world.

³ Clearly it is true that a sound wave makes contact with a surface inside the listener's ear, but this is to describe the phenomenon at a different level of analysis: at the biological level rather than the psychological level. See Michaels and Carello (1981, p. 103) on “grains of analysis.”

⁴ Is there not a contradiction here? If the claim is that the function of a verbal action is to direct a listener's attention (implying that speaking is about indirectly affecting the behavior of others), how can this be compatible with the claim that speaking is action directed at knowledge relations (implying that speaking is a means of directly acting on the speaker's environment)? My working assumption is that these claims are compatible. I here call upon Tinbergen's (1963) distinction, in ethology, between a trait's evolutionary function and the mechanism by which that trait operates. It is presumably functional, or adaptive, for members of a species to be able to reliably direct one another's attention to threats in their shared environment (note, also, that attending is itself a form of behavior). In order to achieve this, an individual may rely on the mechanism of producing an action that is controlled with reference to a knowledge relation standing between a conspecific and the threatening object. In that case, the function of speaking is to direct attention, whereas the mechanism by which speaking is achieved is through action directed at knowledge relations.

⁵ Granted, there may be some verbal utterances for which this description does not seem to apply. The word “hello,” for instance, does not obviously connect the addressee to any structure except for the speaker. The word “hello” may simply be a device for orienting the addressee toward the speaker.

⁶ A related approach is Thompson's (1997) natural design perspective. Thompson conceived of design as an association between two arrays—an array of structures (verbal actions, say) and an array of uses (outcomes that the speaker is trying to achieve). In the case of communication, the speaker is said to be attempting to cause the addressee to enact a further design (e.g., in the bread-passing example, to cause the addressee to enact the action of bread passing).

⁷ There is, of course, a vast empirical literature on the nature of speech in early infancy and its development (e.g., Reed, 1995). I am not concerned here with the question of in what precise order a child's skills develop; I restrict myself to the more theoretical question of what kinds of explanatory structure we should be looking for if we wish to pursue a genuinely nonrepresentational account of speaking.

⁸ Here I acknowledge another account of speaking that also emphasizes the importance of relations: relational frame theory (Hayes, Barnes-Holmes, & Roche, 2001; Tonneau, 2004). This is an intellectual descendent of B. F. Skinner's behavior analysis approach to verbal behavior (Skinner, 1957). An important difference between this account and the one presently being outlined is that the relational frame theory does not conceive of relations as things in the environment that can be directly perceived but as things that a person does: “People frame events relationally in the moment as an active process that is a function of their extensive learning history and stimulation in the present environment. ‘Storage’ of these frames as structures is not implied and not required” (Blackledge, 2003, p. 429).

⁹ The proposal here is not novel. Vygotsky (1978) outlined the same process, describing the child's language learning, and particularly learning to use private speech, as a process of “internalizing” a behavior first learned in interaction with others. E. B. Holt also proposed something along these lines in his radical empiricist concept of the “recession of the stimulus” (e.g., Holt, 1915, p. 75). Holt suggested that as an animal learns to respond appropriately to structure at increasingly higher levels of organization, its behavior is less and less constrained by the immediate contingencies of the surroundings.