Similarly, in the instance HOOK, from Section Three above, we see George performing, in "overlap" with Monica's talk, a nonvocalized, embodied action turn-unit; and we observe that, like Mary, Monica halts in mid-turn, alters her own turn direction, and immediately orients to George's performance, displaying not only agreement with his nonvocalized claim but also interpreting/articulating that claim in her spoken utterance ("yeah nOw you are frEE").

(Click on PLAY button for HOOK)

4 George: but (0.2) NO:W [I'm ((EA--(0.2)------>))

5 Monica: [but i f w e : sa:y-

6 (0.1) yeah nOw you are FREE

Our primary interest is focused on the remarkably non-problematic nature of turn-taking in these instances. The participants display smoothly unfolding, untroubled interaction. Their observable behavior in these instances supports the argument in Proposition One: that embodied actions -- and particularly nonvocalized embodied actions -- comprise a fifth domain of turn construction-unit types available to co-present communicators.

However, it is important to observe that turn overlap of this nature (one participant is speaking and another produces a nonvocalized embodied action-turn at the same time) is a new emphasis for turn-taking among speaking/hearing participants. In recent work by McIlvenny (1995) on how the deaf make use of visual space and the body in constituting and managing their interactions, he builds on past findings that for deaf interactants, "turn transitions are managed in particular and novel ways in and through the lived-in visual-space modality" (147). MvIlveny found that:

The visual-spatial modality shapes the social organization of

interaction and of talk-in-interaction specifically in that it must

be accomplished in situ with hands, eyes, faces and bodies,

but not with ears. (132)

For example, among deaf interactants, overlap is NOT entailed in simultaneous signing because those signers may or may not be being attended to visually by recipients; and individual recipients may experience an instance of overlap when others co-present do not see it simply because their visual attention is directed elsewhere in that moment. In her 1995 study of the interactive structures of encounters among deaf individuals, Madeline Maxwell has noted that: "Signing rather than speaking is central to interactions in these events. Vision rather than audition is a key influence on interaction structures" (2). (Footnote *1).

What is most germane to our discussion is that the unproblematic recurrence of embodied action-turns produced by and oriented to by speaking/hearing co-participants demonstrate that they, too, inhabit McIlveny's "visual-space modality" and rely on its resources. The data show that physical/visual communicative resources are used and relied upon by co-present interactants; those who can see most assuredly are counted on to see in particular moments of interaction that have been produced for that recipient's visual appreciation. And so we observe that turn overlap -- where embodied action-turns are involved -- is a phenomenon handled differently that speech overlap and is differently problematic: producers must "show" for a "seeing" recipient and therefore these turns are inherently collaborative in nature.

In the DROP-CATCH instance from Chapter One, the child produces a coordinated and dynamic request-configuration with her body, and her father's rapid response provides evidence for the closeness of his visual monitoring and attending to her interactive turn. In the examples presented in this chapter, we have observed how closely co-interactants are attending to one another and how they produce turns that display evidence of that close attention. Providing and relying upon such close aural and visual attentiveness is part of the work interactants must do in order to analyze the unfolding structures of turns, to recognize places of possible completion, and to accomplish smooth turn-transition. In their description of the turn-taking system, Sacks et al made specific reference to the fact that gaze direction could be used to select a next speaker. What this entails, of course, is that the recipient is already attending visually to the current speaker such that he or she might be "selectable-as-next" by a turn-holder using gaze-direction as his/her selection device.

The interactive space between co-participants is what Kendon has called "the transaction space" (1990, 211), and this is the domain of McIlveny's "visual-spatial modality." Producers of embodied action-turns make their physical performances visually available to their co-participants and in so doing, provide evidence for them (and for analysts) of their actions' interactional-relevance and recipient-design.

Therefore, as McIlveny and as Maxwell have demonstrated, interactive structures -- such as overlap -- get collaboratively worked out in different ways among deaf interactants. Here, in another instance with an embodied action-turn, we see non-deaf interactants effecting smooth turn-taking based on their shared reliance on the visual-spatial domain. Co-presence makes it available to them, and their use of it to employ embodied action-turns specifically requires the physical performance by the producer of the turn and visual attention of the recipient. The closeness of this visual attending and monitoring is demonstrated in the following instances, HAREM, involving a nonvocalized second assessment, and SEWING, involving a word search. They are treated at length.