Joe Blythe, Rod Gardner, Ilana Mushin, Lesley Stirling (2018) Research on Language and Social Interaction, 51(2), 145-170. https://doi.org/10.1080/08351813.2018.1449441
The first CIARA publication is essentially a pilot study for the Turn-taking and Action Sequences sub-project. In the ‘Tools’ paper we take up one of the central claims from a well-known body of ethnographic literature relating to Aboriginal ‘styles’ of conversation, namely the ‘broadcast’ nature of Aboriginal interaction (Walsh 1991; 1994; 1997). The claim is essentially that within Aboriginal conversation, for the most part, talk is not directed to specific individuals. As a consequence, Aboriginal people are not particularly compelled to answer questions. They may opt-in or opt-out, as they see fit, because ‘control’ lies with the hearer. This claim poses a challenge to a ‘universalist’ model of turn-taking (e.g., Sacks et al 1974), specifically to the ‘rule’ that states when a current speaker selects a next speaker to speak (for instance, by asking them a question) then that selected person is obligated to speak (by answering the question, or by at least providing a reason for not answering).
Thus, we compared across different conversations conducted in four Aboriginal languages (Gija, Jaru, Murrinhpatha and Garrwa), each belonging to a different language family. We examined questions directed to toward specific individuals to see how these questions are packaged, and whether the selected person responded as expected. If they didn’t respond on the initial attempt to engage them, we then looked at what was involved in getting them to respond with a second attempt (e.g., by re-asking the question).
Although certainly not an exhaustive list, we identify a number of tools of engagement – features of turn design that increase the likelihood of securing a response. Address terms like names or nicknames, pointing, eye gaze and voice projection in the direction of the addressee; these are among the features of turn design that are normally involved in securing a response. Attaining mutual eyegaze (between the questioner and the would-be responder) is particularly effective for eliciting next turns from selected next speakers.
In fact, the physical arrangement of participants in space – whether in circular, semi-circular, or linear arrangements – can be a determining factor in whether the selected speaker actually speaks when selected.
When participants are able to return each other’s gaze without difficulty, would-be responders can see that they are being selected to speak, and thus generally respond accordingly. A failure to respond when selected usually results in a pursuit of a response (with a second attempt at asking the question). In these second attempts, speakers invariably gaze at (and project their voices) toward their targeted addressees, as they attempt to secure mutual eyegaze. In this respect, there is nothing unusual about the interactions we examined that would set these conversations apart from others described elsewhere in the world. For this reason, we see no grounds for proposing a culturally specific ‘style’ of turn-taking, as implicated by the ‘broadcast’ model.
However, we also found a few unanswerable questions – either questions that only the current speaker would be able to answer, or rhetorical questions that state a position or make a point, rather than elicit actual information. Interestingly, these questions were produced with gaze directed away from all other present participants. They are visibly produced for ‘nobody’ to answer.
We surmise that ‘non-selecting’ questions of this type may lie behind the ‘broadcast’ characterisation of Aboriginal conversation from the earlier ethnographic accounts, where questions don’t demand responses because they aren’t directed toward any specific individuals.
With the Tools paper our objective is not to debunk the broadcast model, but rather to take the impressionistic observations of an astute ethnographer as the basis for an empirical investigation into how next-speaker selection is enacted. We suspect the Tools of Engagement we observe to be operating in our corpora of Aboriginal language conversations to also apply to conversational turn-taking more broadly. Thus, our principle objective is to elucidate engagement of next speakers in any conversation – regardless of the language or cultu