Question and Answer Sequences in Garrwa Talk

Abstract

For question–answer sequences in Australian Aboriginal talk, it has been claimed that answers are not necessarily a required response. This would contrast with findings reported in recent cross-linguistic work on such sequences. In a corpus of 62 question sequences from conversations in two Garrwa communities on the west side of the Gulf of Carpentaria in northern Australia, 34 questions were answered, and a further 12 dealt with the question in some other way. Sixteen received no response in the proximally subsequent talk. Whilst most of these questions were answered, the offset time between question and response was long compared to previous studies. There was also a higher rate of non-answers and non-responses. For some cases of non-responses, contingent factors easily explained the lack, but in a few the reasons were not so apparent. It is argued that a significant factor in the relatively long silence between question and answer, and the relatively high rate of non-answers or non-responses, is that the parties in the talk spend much of the time in ‘continuing states of incipient talk’, rather than in tightly focused and temporally bound conversation, which may help to account for the apparent relaxation of gap minimization and response mobilization.

Keywords:

• Aboriginal Conversation
• Conversation Analysis
• Questions and Answers
• Conversational Style

Notes

¹This is not to say that there is a denial in CA work that cultural differences exist, but that these differences play out through variations in practice of these fundamentals. For example, it would appear that in some cultures compliments are generally rejected by compliment recipients, whilst in others they are more likely to be downgraded or accepted. However, in all cases the recipients are dealing with the compliment, and it would be very rare (and subject to special contextual considerations) for a compliment to be responded to by, for example, an insult or a greeting.

²This contrasts strongly with the Stivers et al. (2009) study of question–answer sequences in 10 languages, in which polar questions were found to be the commonest type in nine of the languages.

³Of the eight polar questions, seven are in Kriol, and none of these has polar question syntax (i.e. auxiliary verb inversion); and six are declarative questions with rising terminal intonation (the seventh is a terminally falling declarative). The eighth example is a rare case of the use of ‘Kuna’ (Ilana Mushin, pers. comm.), which is a polar question particle. Interestingly, this utterance has falling terminal pitch, similar to most English WH-question utterances. It is also striking how few polar questions there are in the dataset, even though I undertook a second careful search for examples on the suggestion of one of the reviewers.

⁴Comparisons with Stivers et al. (2009) are constrained by that study being limited to polar questions. It is nevertheless striking that on several of the parameters, these Garrwa conversations lay outside the range of results for those 10 languages: offset times for responses were longer than for any of those languages, the proportion of questions not receiving an answer was higher, and the proportion of questions receiving no response at all was higher (with the exception of Lao, which was similar).

⁵All names are pseudonyms.

⁶In the interests of readability of the transcripts, notation of non-verbal activity has been kept to a minimum, as the transcripts are already fairly unwieldy with glossing and translations. The downside is that the reader is required to do more work to locate reference to non-verbal activity in the discussion.