Multiparty storytelling in Umpila and Kuuku Ya’u

ABSTRACT

This paper analyzes the functions of interactional devices used by co-tellers in multiparty stories in Umpila and Kuuku Ya’u, two closely related dialects of a Paman language of Cape York Peninsula, Australia. Within the Umpila and Kuuku Ya’u speech community there is a strong cultural preference for multiparty storytelling – a preference which has been noted in a number of Aboriginal Australian contexts. This paper seeks to understand the mechanisms through which co-tellers in these multiparty narratives contribute to the story. It first discusses co-teller roles, and distinguishes three key narrator roles and orders of conduct associated with each. The analysis then focuses on the use of questions and evaluative comments used by one type of co-teller, supporting narrators. The following discussion demonstrates that questions and evaluative comments go beyond immediate functions of seeking information or spontaneous expressive reactions. They help to fulfil expectations on supporting narrators to engage actively in the talk. It is additionally shown that these devices have functions in highlighting key aspects of the story and developing stance in intricate ways that complement the main line of the storytelling. The analysis demonstrates the close coordination of co-tellers in constructing a story; piece-by-piece they collaboratively describe and evaluate the story events as a group, prioritizing a local and situated shared telling over other potential story goals like performance and progressivity of a plot. The analysis of Umpila and Kuuku Ya’u storytelling contributes to the field of interaction and narrative studies by furthering our understanding of the organization of storytelling in different cultures and languages contexts.

KEYWORDS:

• Multiparty storytelling
• questions
• evaluative comments
• narrator roles
• epistemics

1. Introduction

This paper will analyze the functions of interactional devices used by co-tellers in multiparty stories in Umpila and Kuuku Ya’u, two closely related dialects of a Paman language of Cape York Peninsula, Australia. Umpila and Kuuku Ya’u storytellers describe the way they tell most narratives as we all talk one time. This expression characterizes a highly interactive mode of multiparty storytelling, where two or more narrators take turns or even co-produce turns to jointly tell a story.

Discussion of the interactional element of storytelling has often been sidelined in narrative research. In the Aboriginal Australian context, whether in text collections or narrative analysis, the focus has typically been on decontextualized ‘literary style’ storytelling events (Austin, 1997; Beckett & Hercus, 2009; Berndt, 1985; Berndt & Berndt, 1989; Clunies Ross, 1986; Dixon, 1991; Klapproth, 2004; Napaljarri & Cataldi, 1994; Róheim, 1988). This skewing is in part a result of the research interests of the field, and in part, the high cultural currency of traditional creation stories (often referred to as dreamtime stories in Australian English) and the formalities associated with their production. However, the growing interest in linguistics in interactionally oriented approaches has influenced the treatment of storytelling in the Australian context. There has been an increased focus on stories of personal experience and an increased acknowledgment of stories as interactionally constructed, emergent and dependent on a complex array of factors in the social setting (Blythe, 2011; Carew, 2016; Davidson, 2018; Hill, 2018; Mushin, 2016). As part of this work, multiparty narration has been posited as one of 10 characterizing features of Aboriginal Australian storytelling by Walsh (2016). Walsh’s observations are drawn from work on Murrinh-patha (Walsh, 1990) and a smattering of accounts that point to a cultural preference for multiparty interaction in specific language settings (McGregor on Gooniyandi and Kimberley languages more widely (1988, 2004), Haviland on Guugu Yimidhirr (1991), Black on Koko-Bera (2010)). The present study builds upon these observations by undertaking a close analysis of two interactive devices, questions and evaluative comments, in a corpus of multiparty stories in Umpila/Kuuku Ya’u. As this paper will show, the analysis of these two devices delineates co-teller roles, each contributing in circumscribed ways to collectively produce a successful narrative. Multiparty storytelling is not interactionally unusual in other linguistic and cultural settings, but cross-linguistic research has typically found that this type of storytelling is associated with contexts where there is shared epistemic access to events being narrated (Georgakopoulou, 2007; Lerner, 1992); for example a husband and wife co-telling a shared experience as detailed in many of Lerner’s English examples (1992) or best friends as discussed in Georgakopoulou’s Greek examples (2007). Umpila/Kuuku Ya’u storytelling is somewhat different from these accounts, in that multiparty delivery is the default style in nearly all contexts,¹ even though in some instances, many of the participants have only secondary knowledge of the events, or even no knowledge at all.²

Storytellers prod and question, and elaborate on and exclaim about each other’s narration, building up the story incrementally across turns. The lack of epistemic access does not in any way preclude participation in the unfolding narration in this way as it usually does in an English conversational storytelling context where normal turn taking is suspended and unknowing tellers refrain from substantive contributions (Sacks, 1984; Schegloff, 1982; Stivers, 2008). A preference for collaborative storytelling overrides any issues posed by the lack of epistemic access, with a group of two or more narrators working collectively to produce output that tells a single narrative. The way in which collaborative telling is interactionally prioritized suggests that these stories are “less geared to narrative as performance and more to narrative as a social forum for piecing together an evaluative perspective on an incident” (Ochs & Capps, 2001, p. 36). In fact, if an Umpila/Kuuku Ya’u co-teller does not participate enough, regardless of whether they know any details of the story beforehand, they are open to sanctions and demands to participate from the other interactants, and if these are not met the whole telling will often be derailed (Hill, 2018, pp. 41–42). The preference for collaborative telling raises the question of the ways in which unknowing co-tellers contribute to the storytelling in meaningful ways. This paper seeks to answer this question through the examination of the form and distribution of questions and evaluative comments. These two familiar interactional actions will be shown to have functions in highlighting key aspects of the story and developing stance in intricate ways that complement the main line of the storytelling. The paper will also show how they exert considerable influence on the local trajectory of the storytelling and generate zones of high collaborative input often featuring choral co-production. The analysis will show that the default practice to collaborate is achieved through delineated roles by different co-tellers each contributing in circumscribed ways associated with this role. How these are maintained, and how the contributions of all develop the story in concert, is one of the central findings of this paper.

The analysis presented in this study is based on the detailed examination of recordings of 20 multiparty narratives. The paper focuses on a close discussion of examples from two stories, Waiting for a Ride narrative and King Fred narrative, which illustrate patterns found in the broader collection (see §2). The rest of this paper is structured as follows. In §2, I will present the corpus of narratives used in this study. In §3, I will provide some background information on the narrator roles and will distinguish three participant roles (a primary narrator; a supporting narrator; and audience role) on the basis of linguistic and behavioural correlates. In §4, I will analyze the distribution and form of questions produced by supporting narrators. This analysis will show that questions have functions beyond obtaining information with wider interactional purview in highlighting key aspects of the stories and providing collaborative tools for co-telling. In §5, I will move onto looking at the evaluative comments produced by supporting narrators. The examination of these with regard to narrator roles and distribution will show how closely supporting narrator evaluations (assessments in Conversation Analysis terms) work in tandem with the primary narrator’s development of the main line of narration. These evaluative comments are important in displaying stance and interpretative responses to the narrative events and have similar highlighting functions to questions. Finally, in §6, I will draw together the discussion of these two interactional devices with general conclusions about how they assist in satisfying the cultural preference for collaborative storytelling for Umpila/Kuuku Ya’u narrators.

2. The stories

The stories discussed in this paper are those of a small group of elderly Umpila and Kuuku Ya’u women, respectfully and affectionately referred to as the old girls within their community. The women are notable as the remaining fluent speakers of Umpila and Kuuku Ya’u, two closely related dialects of a Paman language spoken on the eastern coast of Cape York Peninsula, Australia. Umpila and Kuuku Ya’u are the key southern and northern coastal dialects respectively from a dialect complex that had six or more varieties prior to non-Indigenous colonization. In contemporary community-based life, Umpila and Kuuku Ya’u have a composite form among the small speech community of remaining speakers. They reside near Lloyd Bay in Lockhart River Aboriginal Community where the vernacular is Lockhart River Creole, an English-lexified creole (Mittag, 2016).

The analysis presented in this paper comes from detailed examination of video recordings of 20 multiparty narratives totalling around 3 h talk time. The stories in this collection are told collaboratively by between two and six co-tellers. Table 1 provides a short title, an overview with a brief description of the story, and the date recorded. These stories recount a wide array of new and old events of personal experience, i.e. events that at least one of the narrators participated in themselves. They retell events from a recent weekend’s camping trip and are often filled with misadventures of broken-down cars or humorous hunting escapades. They also retell events of long past-times and wider historical-social relevance, which the narrators themselves participated in or witnessed, such as the flu epidemic in the Old Mission, and army related activities in the region during World War II.

Table 1. Narratives in the corpus.

Title	Topic	Date recorded
Buthen Buthen	Stranded near miners camp at Buthen Buthen	14/03/07
King Fred	King Fred discovers a sea mine	23/03/07
Old Woman Wuthathi	Fishing trip with a Wuthathi elder	08/02/05
Land Handover	Trip to Coen for KULLA land trust agreement	04/09/08
Charlie	Pet emu Charlie chases Islander men visiting the Lockhart Mission	23/06/08
Spirit Relative	Encounter with ghost of an ancestor	12/08/07
That Sick	Flu epidemic in Mission times	21/05/05
We Go Country	Journey to Umpila country	18/08/08
My Country	Journey to Massey Creek	17/08/08
Kangaroo	Encounter with a large kangaroo	10/06/08
Preparation for Dance	Preparation for the Laura Dance Festival	05/07/07
Give Fright	Two boys play a trick on an old woman	30/06/08
Suzie and Cilla	Suzie and Cilla’s encounter with a sorcerer	04/09/08
Escaped Prisoners	Discovery of escaped prisoners living in the bush	02/04/07
Waiting for a Ride	Travelers stranded overnight near the old Lockhart River Airstrip	27/03/07
Stranded	Stranded at Pascoe River Mouth	25/07/11
Perth	Trip to Perth for an art exhibition opening	12/07/11
WW2	Encounter with American army planes in WW2	05/04/04
Hospital	Emergency trip to hospital in Cairns	17/07/10
Canberra	Trip to Canberra with Athol Chase	19/07/11

The stories were recorded as part of several major language and cultural documentation projects carried out by the author in collaboration with the speakers and local community organizations between 2004 and 2011.³ Given that Umpila/Kuuku Ya’u is not the everyday language of this community, the recording of these stories was an organized language documentation activity. The documentary sessions often took place on a veranda at the Lockhart River Arts and Culture Centre where the speakers would gather to share the news of the day, drink tea and tell stories. They were often joined by family members who were co-tellers of, and audience to, the stories. In these sessions, the stories themselves were recorded as spontaneously and naturally as the language situation allowed. The narrators self-selected the topics of narration and self-organized levels of participation and speaker roles. The narratives were not formal performative events resulting in seamless blocks of talk by a narrator: they were sometimes fragmented and fleeting, and always interactionally emergent and crucially dependent on receptive co-narrators and a receptive audience (Hill, 2018, pp. 67–77).

For the purposes of illustration, the paper focuses on a close discussion of examples from two stories, Waiting for a Ride narrative (told by five co-narrators) and King Fred narrative (told by two co-narrators). These two stories were selected as representing clear and indicative exemplars of the function of questions and evaluative comments produced by supporting narrators noted in the broader collection. A close discussion of two narratives allows the article to explore the relationship between sequences or spaces within the story and between the co-tellers as a story develops.

3. Narrator roles

Storytellers in Umpila/Kuuku Ya’u multiparty narratives have different types and degrees of participation through the ongoing talk. The analysis presented in this paper will distinguish between three participant roles: a primary narrator; a supporting narrator; and audience role (for similar co-teller distinctions see Lerner’s “story consociate” (1992) and Schegloff’s “active recipiency” (1982)).⁴ These roles can be distinguished on the basis of linguistic and behavioural correlates. Briefly discussed here are three main correlates: production of new narrative content; the right to select speakers; the right to question or correct. This section provides a brief sketch of participant roles required for the coming analysis; for a more detailed discussion, the reader is referred to Hill (2018, pp. 39–59).

As the name suggests, primary narrators typically have primary knowledge of the events being discussed and contribute the bulk of new narrative content describing these events. As such, they produce more turns and hold the interactional floor for longer during these turns. Co-tellers can transfer tellership during a narrative so that the participant with primary knowledge of a particular section of the story inhabits the primary narrator role for that section (“rendering one’s own part” as in Lerner (1992, pp. 265–266)).⁵ Supporting narrators and other recipients (i.e. audience members as termed here, though this obscures some complexity) fix their eye-gaze on the primary narrator(s) for much of the interaction. This fits with research on recipient eye-gaze patterns, where it has been noted that it is normatively required to gaze at speakers in extended tellings (Rossano, 2012; Rossano et al., 2009). To illustrate, consider (1) below. This a short excerpt from a narrative in which RP recounts her experiences encountering a wapa ‘a supernatural being’ to five other participants (referred to here with initials, NB, DG, CS and KH and along with the author/linguist). RP is the primary narrator; she takes the lead in narrating the events underway and introduces the people and describes the actions of the people in the story. In (1), RP produces a burst of new narration where she describes how a nephew nicknamed Marbot drives them to the beach and the food cooking that happens there: ‘Marbot carries us’ (line 1), ‘those two had a car too’ (line 3), ‘over there she buried the food there’ (line 5). The other four participants direct their gaze towards RP. They also speak, but they produce notably less linguistic output, and their contributions are derivative upon RP’s, as in lines 9–14 where NB and DG confirm and elaborate on the damper being buried which had already been described by RP in line 5. NB and DG’s repeats and elaborations in this example are typical of supporting narrators. They assist in the delivery and provide ongoing active reception of the story.

Table

(1)	1	RP	Marbot	nganan	kali-na
			marbot	1plexcACC^6	take-NF
			Marbot carries us
	2		(1.0)

Table

	3	RP	pula	pa’amu all be get a car too them too
			3plNOM	two
			those two (Anita and husband) had a car too
	4		(1.6)

Table

	5	RP	ngku-lu	mayi	atha-na	ngulu	Anita-lu	mukana
			dem.dist2-DM	food	bury-NF	3sgNOM	anita-ERG	Big
			over there she buried food there (implies damper buried in ashes), Anita and big mob there
	6		(0.5)
	7	NB	hm
	8		(1.1)

Table

	9	NB	pulka-nguna	atha-na =
			hot.ashes-LOC	bury-NF
			(she) buried (the food) in the hot ashes
	10	RP	=pulka-nguna	atha-na
			hot.ashes-LOC	bury-NF
			(she) buried (the food) in the hot ashes
	11		(.)
	12	DG	malngkan-nguna
			beach-LOC
			at the beach
	13		(.)

Table

	14	NB	°atha-na	[yeah°
			bury-NF	yeah
			buried yeah
	15	RP	[malngkan	miintha
			beach	good
			nice beach/sand
	16		(0.3)
	17	CS	hm
	(Hill 2007–2010: LCCH-V04Sep08_Suzie and Cilla)

Primary narrators prompt for confirmation and support from supporting narrators, typically selecting them as recipients of these utterances using eye-gaze and body orientation. Word searches are interesting in that they show that the speaker thinks that another participant could recognize the object of the search, which also provides a way for a co-teller to enter the storytelling (Lerner, 1992, pp. 255–258). Example (2) shows an instance of a word search produced by RP a little further along in the narrative from (1). In line 5, RP struggles to think of the Umpila/Kuuku Ya’u language form, instead using pipi, a generic term for all small shell molluscs in Lockhart River Creole. The word search is produced with tell-tale rising intonation and lengthening; at the same time RP directs her gaze towards NB and CS (and not to the three other participants present: DG, KH and the author). NB responds in line 7 by producing the missing form nyakun.

Table

(2)	1	RP	ngulu	nga’a-lu
			3sgNOM	dem.dist1-DM
			that one
	2		(1.0)
	3	RP	api-na
			pick.up-NF
			picked up
	4		(.)
	5	RP	pipi:::::
			pipi (shells)?
	6		(0.4)
	7	NB	minya	nyakun =
			meat	mollusc
			nyakun [species of mollusc]
	8	RP	=yaw	[ nyakun
			yes	mollusc
	(Hill 2007–2010: LCHVDP04Sep08_Suzie and Cilla)

In the reciprocal role to the primary narrator, questions or corrective actions regarding aspects of the story are directed towards the primary narrator by supporting narrators. These actions are the focus of §4. These two reciprocal selections by primary and supporting narrators respectively reveal which participants are deemed able to assist in filling in gaps in the narration and who have the rights to query the unfolding narration. In contrast, audience members have a passive role, not questioning nor being questioned. RP’s daughter (MI) is present at the telling of the story from which (1) and (2) are drawn. She inhabits an audience role as distinct from the supporting narrators. She displays recipiency through eye-gaze directed at RP and the various supporting narrators and through the production of more minimal feedback like confirmatives, nods, laughter (see (8)–(10) in §5 for illustration of audience participant feedback). This is a frequent pattern in participant dynamics: MI is of a younger generation and with less traditional language proficiency, which are noted as recurrently associated with an audience role, though not always. In other contexts these factors would be less relevant if there were cultural motivations (e.g. relationship to land, totemic or family connections) prioritizing her input into a story.

While a good portion of supporting narrators’ contributions are derivative from the primary narrators telling (as above) this is not to suggest that supporting narrators do not at times direct the flow of the storytelling or make meaningful additions to the story. They do, and can do so within multiple perspectives in the story. Unlike primary narrators or audience members, they traverse between the narrated event perspective, in which the story events are recounted, and a meta-linguistic perspective, external to the story-world. In the external perspective, they prod at and prompt primary narrators with questions (§4). They produce evaluative comments and exclamations which draw conclusions from the other speakers’ narration (§5). These orders of conduct associated with supporting narrators will be explored in the remainder of the paper.

4. Questions

This section will analyze the functional workload of questions and describe the influence they exert on the flow of storytelling in multiparty stories in Umpila/Kuuku Ya’u. Questions are a crucial tool in multiparty narration because they deal with the fundamentally cooperative nature of this task: establishing and maintaining mutual understanding and coordination of knowledge states across the co-narrators with varying knowledge of the events being narrated. Given the epistemic skewing between participants in multiparty narratives, questions are produced largely by supporting narrators requesting additional information about the story from a knowing primary narrator, e.g. ‘where was that?’; ‘what did they do now?’; ‘how much did they find?’. However, the function of questions also goes beyond mere informational imperatives in many instances. This is most immediately apparent where questions are posed that the questioner already knows the answer to. For instance in one case, following the introduction of a protagonist in a story a supporting narrator asks of the primary narrator wantuka kalmanana ‘where is he from?’, despite less than a minute prior herself saying ngulu kamanana yiipaymunu, Night Island ‘he came from the South, from Night Island’ (Hill 2007–2010: LCCH-V23Mar07_King Fred). Asking a question that you know the answer to does not break the interactional preferences in these multiparty narratives.⁷ Such utterances are not uncommon nor treated as problematic by the narrators in these stories. They are enthusiastically taken up and responded to, often at length. In this section, I will show that the use of questions by supporting narrators functions as a way to highlight key aspects of the story (when the answer to the question is known or not). I will support this argument by looking at the form of questions and their uptake by co-tellers. I will show that post-question spaces involve a shift to a different participation framework in which supporting narrators routinely, in choral style co-participation, work over details of the story.

This section will consider several excerpts from a comedic story about the misadventure of a group of travellers stranded overnight near the old Lockhart River Airstrip (see Waiting for a Ride in Table 1). The narrative was jointly told by five speakers. The events in the story were experienced by NB, who is the primary narrator, while the other four participants (CS, DG, RP and LP) do not know the story prior to this telling and take a supporting role. Questions are frequently used by these co-tellers to participate in the interaction, representing a quarter (20/85, 24%) of all their contributions (see Table 2 which gives an overview of supporting narrator contributions vis-à-vis utterance type).

Table 2. Supporting narrator contributions to Waiting for a Ride.

Clause type	Question	Declarative	Evaluative	Continuer
Frequency	20	43	6	16
Total %	24%	51%	6%	19%

In (3) CS asks the primary narrator NB, ‘hm what did those two say, the two Islanders?’ (line 3). In (4) RP asks NB, ‘how are those two? are they not frightened?’ (line 4). These examples have been chosen as they display the typical form and functions of questions in Umpila/Kuuku Ya’u multiparty stories. The content word form of the questions and lexical design of the question help to draw out and highlight the main topic of the story. At the point in the story in which (3) and (4) occur, the scenario of a group of travellers stranded has been well established. The characters have been told that they have to stay overnight away from the airstrip facilities, and the two Islanders have been singled out from the group as a key focus of the story. With the aid of CS and RP’s questions in (3) and (4), it becomes apparent that this will be a lawalawa ‘lair’ or ‘joke’ story (Hill, 2018, p. 66). This is a genre of comedic contact story about interactions with outsiders, often Torres Strait Islander people, as in the case of the Waiting for a Ride narrative. Typically, these outsiders through lack of knowledge and experience in the Cape York environments find themselves afoul of some type of misadventure. The questions in (3) and (4) both ask for more information about the Islanders’ reactions to the situation, and they do so in ways which demonstrate an understanding of centrality of these to the primary narrator’s account. In (3), CS’s question ‘what did those two say, the two Islanders?’, hones the story in on the Islanders’ specific response. It is produced in response to a detailed description by the primary narrator of the travellers’ camp arrangements for the night. A key thing to note about CS’s question is the use of overt person references, namely pula ‘they’ and a lexical NP headed by the ethnonym thathimalu ‘islander’. In Umpila/Kuuku Ya’u narration minimal and ellipsed person references are the default (Hill, 2018, pp. 260–268). The use of thathimalu ‘Islander’ in this context – where something more semantically minimal would have sufficed in securing reference – highlights the most thematically relevant aspect of the identity of these referents. It highlights their outsider status, which is crucially relevant to their reactions in the unfolding scenario. Use of ethnonyms over other reference forms are closely associated with lawalawa ‘lair’ or ‘joke’ stories (Hill, 2018, pp. 194–198). This person reference choice breaks the social world into ethnic groups (e.g. Islander, Aboriginal person, white person, Chinese person) and so relates thematically to the insider/outsider thematic of these stories. Twenty plus turns later RP asks ‘how are those two? are they not frightened?’ (line 4). This once again draws the storytelling towards the Islanders’ response, specifically their emotional state as contrasted with the local Umpila and Kuuku Ya’u peoples.

Table

(3)	1	NB	ngampula	got-im	yuma	mukan	pal’ana-na =
			1plincNOM	got-him	fire	big	prepare.fire-NF
			“we have to make a big fire”
	2	NB	=haaahuhhhh=
			haaahuhhhh

Table

	3	CS	hm	pula	wantant	pa’a-	inga-na	pa’amu	thathimalu
			hm	3plNOM	IGNOR	two-	say-NF	two	islanders
			hm what did those two say, the two Islanders?
	4		(0.2)
	5	NB	apa
			INTJ
			“oh dear!”
	(Hill 2007–2010: LCCH-V27Mar07_Waiting for a Ride)

Table

(4)	1	NB	yuma-na	mukana	pal’ana-na
			fire-?	big	prepare.fire-NF
			we made a big fire
	2		(.)
	3	CS	hmmm =
	4	RP	=pula	pa’amu	wantantu	ngampa	wini-na
			3plNOM	two	IGNOR	NEG	fright-NF
			how are those two, are they not frightened?
	5		(0.6)

Table

	6	NB	pula	nga’a-l	pa’amu
			3plNOM	dem.dist1-DM	two
			those two
	7		(0.4)

Table

	8	NB	aa	ngay	inchi-nya-lana	ngampa	wini-mpu
			Ah	1sgNOM	tell-NF-3plACC	NEG	fright-IMP.PL

Table

			ngampula	wuna-tha
			1plincNOM	sleep-FUT
			I speak to them, “don’t be frightened, we will sleep”
	(Hill 2007–2010: LCHVDP27Mar07_Waiting for a Ride)

As in (3) and (4), questions by supporting narrators are overwhelmingly WH-questions. In this format the questioner contributes some information about what they already know, and the question is targeted to seek out information about one element of the proposition.⁸ The information included is often based on content from prior talk, either extrapolated or in other cases explicitly part of the prior talk (once again highlighting functions beyond an information imperative mentioned above). For example, in (4) where RP asks, ‘how are those two? are they not frightened?’ (line 4). At this point the frightened state of the Islanders is an established and ongoing topic of discussion in this part of the story. RP herself had even commented winina ‘(they) are frightened’ 20 turns earlier. The WH-form is key here. Its displays the comprehension of the supporting narrator, and functions to highlight topical information (like the emotional state of the couple in (4)) by incorporating that content into the question. As in the case of the two examples we have looked at, questions function to highlight meaningful information, ensuring the important points are emphasized and ‘driven home’ to the audience.

The uptake of questions routinely marks a shift to another format of storytelling, where multiple co-narrators join in and work over the details of the story. The question by CS in (3), shortly followed by RP’s in (4), marks the start of a lengthy and highly interactive sequence which largely halts the narration of subsequent events in the story for no less than 60 turns – only a small part of this section of the story is shown in excerpts in this paper. The length of this is more drawn-out than usual, but the type of narrative ‘work’ carried out by the co-narrators is typical of such post-question spaces. These sequences continue the work started in the questions themselves and function to emphasize and embellish important information in the story. Example (5) shows more of the sequence following the question in (3):

Table

(5)	1	CS	hm	pula	wantant pa’a-	inga-na	pa’amu	thathimalu
			hm	3plNOM	IGNOR two-	say-NF	two	islanders
			hm what did those two say, the two Islanders?
	2		(0.2)

Table

	3	NB	apa
			INTJ
			“oh dear!”
	4		(1.0)
	5	RP	ngulu	wini-na
			3sgNOM	frighten-NF
			he was frightened
	6		(.)

Table

	7	NB	nganan-	ngampulan	[ngaani	ngachi-ka
			1plexcACC	1plincACC	IGNOR	find-FUT
			“something will find us lot – us all!”

Table

	8	LP	[nga’a-l-ki	xxxxx
			dem.dist1-DM-DUB	?
			those/that one might
	9		(.)

Table

	10	DG	chilpu	kalma[-na	lo
			old.man	come-NF	GEN
			the old man came
	11	LP	[kalma-na	Masic
			come-NF	Masic
			came from Masic (island)
	12	RP	nyii =
			yes
			yes
	13	DG	=chilpu	kalma-lu	[chilpu-lu
			old.man	come-?	old.man-ERG
			the old man came, the old man
	14	RP	[nyii
			yes
			yes
	15		(0.4)

Table

	16	DG	nga’a-l	chilpu	yuma	nga’a
			dem.dist1-DM	old.man	fire	dem.dist1
			that old man at the fire there
(Hill 2007–2010: LCHVDP27Mar07_Waiting for a Ride)

In the first few turns of (5), NB responds to the question about the Islanders’ reaction with a representation of the Islanders’ speech: “oh dear!” (line 3) and “something will find us lot – us all!” (line 7). Following this, the co-tellers repeat and embellish content already established. RP adds about the Islander husband, ‘he was frightened’ (line 5). DG and LP reiterates that the Islander couple are from Masic Island in the Torres Strait (lines 10–11). This further highlights the participants’ Islander origin and experience of the world which is crucial for understand their reactions in the story. DG elaborates on NB’s previous description of fire building, by describing the Islander old man standing by the fire built for him (line 16). Example (5) shows how the co-tellers use this interactional space in the storytelling to add small details about the current scenario and reiterate the emotional state of the Islanders.

To show this pattern further, another excerpt from this sequence is shown in (6). CS reiterates ‘that one is frightened’ (line 1). This is confirmed by NB (line 3). RP repeats twice that the Islander man sits up till daybreak (line 2, line 5). This contrasts with NB’s description of herself and his wife sleeping (line 8). CS adds further detail by specifying their proximity to the fire (line 10). The building up of the situation by redoing and enhancing details helps to construct key background to the Islanders’ reaction to the situation they find themselves in.

Table

(6)	1	CS	nga’a-l	wini-na =
			dem.dist1-DM	fright-NF
			that one is frightened

Table

	2	RP	=pa’anama	paacha[na
			sit.down	daylight
			sit down till daylight

Table

	3	NB	[wini	nyii
			fright	yes
			frightened yes
	4		(.)
	5	RP	pa’anama	[paachana
			sit.down	daylight
			sit down till daylight

Table

	6	DG	[oh	nga’a-lu	xxxxx
			oh	dem.dist1-DM	?
			oh that one
	7		(1.0)
	8	NB	wayimu	wuna-na ngayu	<nga’a-l >
			woman	camp-NF 1sgNOM	dem.dist1-DM
			the woman and I sleep
	9		(1.0)
	10	CS	yuma	yinchu
			fire	close
			close to the fire
	(Hill 2007–2010: LCHVDP27Mar07_Waiting for a Ride)

The post-question spaces are characterized by a certain style of co-production, which is clearly shown in (5) and (6). Use of transitions and pausing between utterances is an important rhetorical device in Umpila/Kuuku Ya’u narrative. Long dramatic pauses (of several seconds) versus rapid fire narration (latched or in overlap) are used for different effects and are associated with different contexts in the narration (see Hill (2016), McGregor (2004) and Muecke (1982) for similar points on storytelling in Australian Aboriginal contexts). In post-question spaces, more than one narrator speaking simultaneously is routine, resulting in choral-style interaction. They are high pace sections of the storytelling which feature overlapping between turns (see (5) lines 7–8, 10–11, 13–14; (6) lines 2–3, 5–6), or micro pauses between turns (see (5) lines 6, 9; (6) line 3) and latching of turns (see (6) lines 1–2; (5) lines 12–13).⁹ Choral co-production in English conversational settings has been understood as an affiliation device where a co-participant associates themselves and displays agreement with what is being said in another turn (Lerner, 1987).¹⁰ This function shows strong similarities to that observed in these stories. The way that the multiple co-narrators construct the narrative in concert with each other through these spaces simultaneously functions to display their mutual engagement in, and mutual understanding of meaningful pieces of information in the emergent story.

To summarize, questions are an important and frequently employed tool for supporting narrators to join in the narration. They are particularly powerful for unknowing narrators who can contribute meaningfully through extrapolated information or common ground. Through the use of content question formats supporting narrators are able to display comprehension of the story, sometimes while still asking for more information, other times while knowing the answer already. In addition, the proffering of this type of question (like has been shown for other displays of uncertainty (Goodwin, 1987)) reciprocally invites the other co-tellers to join-in the storytelling. Beyond the response to a question proffered, questions often initiate extended collaborative sequences, featuring choral co-production, which highlight and embellish important aspects of the unfolding story. In Umpila/Kuuku Ya’u narration, questions appear to invoke a shift to a different participation framework (Goodwin & Goodwin, 1986); a framework where supporting narrator talk dominates and where highly collaborative discussion is typical, if not interactionally required.

5. Evaluative comments

This section considers another interactional device employed by supporting narrators to contribute to Umpila/Kuuku Ya’u multiparty narration. Like co-tellers and recipients of many stories, supporting narrators often produce evaluative comments (assessments in Conversation Analysis terms) and express reactions to the story as it unfolds. For instance in (7), supporting narrator LP exclaims ‘that one, son, I cry out for poor son!’ immediately following a key and troubling revelation in the events being described by the primary narrator. This section will show that these comments have extra functional payload; much like questions discussed in §4, they highlight key aspects of the story and work in tandem with the main line of the narration to build up the story.

Table

(7)	1	LP	ngulu	(.)	maampa	nga’a-lu	ngayu
			3sgNOM		♀C	dem.dist1-DM	1sgNOM

Table

		ing-	ingka-n	maampa-laka
		cry	cry-NF	♀C-PATHOS
		that one, son, I cry out for poor son!
	(Hill 2007–2010: LCHVDP23Mar07_King Fred)

Supporting narrator evaluative comments are of three general types. They provide exclamations displaying surprise or dismay or other emotional reactions, like in (7) above (or as in, apa apa ngululaka ‘oh dear the poor thing!’), or explanations in response to events in the story world (yakay awuki ‘oh dear it must be a devil!’) or comments that directly evaluate characters and scenarios (kampinu wu’umana ‘the man was bad’). These evaluative comments are momentary asides to the main recounting of the story events, and as reactions to the narration itself, they are often deictically set in a perspective external to the storyworld being crafted by the interactants’ talk. They often feature interjections, evaluative morphology (see (7)), imperative tense, first-person reference forms and self-associated kin-terms (see (7)), and prosodic markedness such as broadening of pitch range. These features work to mark the utterance as outside the telling of the narrative events. Given that evaluative comments are often highly emotive and expressive, they appear to be produced reactively in a ‘spontaneous’ and less ‘controlled’ way than other aspects of the narration. On a closer look, however, they are better understood as interactionally co-constructed and interactionally managed by joint work between the primary and supporting narrators (Hochschild, 1983; Wilkinson & Kitzinger 2006). As will be shown in this section, the performative rather than spontaneous nature of these utterances becomes apparent through the examination of narrator roles, distributional patterns and the target recipient of the comment.

The discussion here will begin by looking at the distribution of evaluative comments across a single story, the King Fred narrative, which (7) above is from. This story is told by RP and LP, with RP in the primary narrator role, and with RP’s daughter and the author as audience. It is a story about the ‘king’ of Night Island (‘king’ as appointed by white colonial powers) and his discovery of a washed-up sea mine which he unwittingly attempts to open. The narrative ends with the ‘king’ running away while the explosive device is being defused by a group of white men. He hides in shame and fright as it explodes. LP has a supporting role in the narration and produces 10 evaluative comment-type utterances throughout the story. These comments are produced in response to key events in the story. This is presented in Table 3, which summarizes the narrative context and the supporting narrator comment in consecutive columns. Questions are also used in this narrative in the same pattern as in §4, but with less frequency. As represented in Table 3, RP reports that King Fred ordered his fellow Night Island people to go down to the beach and try to open the mine (lines 53–57). LP exclaims ‘poor old man! son, hey son!’ (line 58). RP reports that a group of young girls have carried down an axe to prise open the mine (lines 70–71) and LP cries out ‘oh no they carry the axe for something! (they) might hit it!’ (line 72). RP reports that a group of white men repairing a light buoy notice the events unfolding and are coming to intervene (lines 90–95) and LP exclaims ‘that one, son, I cry out for poor son!’ (line 96) and so on. This is the typical pattern. Evaluative comments tend to punctuate stories during key junctures and moments of high drama in the action in the story – they are particularly prevalent at moments when something dangerous is befalling the story participants. The evaluative comments make clear and highlight important information within the story at these moments. They support the action of the narrative being built up by the primary narrator. This pattern of high-drama and evaluation corresponds to findings in a wide range of studies: most notably Labov and Waletzky’s (1967) narrative macro-structure model where the evaluative component follows the narrative high point, but also Heritage (1984), Jefferson (1984, 1988), Kockelman (2003, p. 479), Ponsonnet (2014, p. 125) and Selting (1996) who all noted special interactional functions and distributional profiles of displays of emotion in storytelling and interaction.

Table 3. Distribution of supporting narrator evaluative comments in King Fred.

Turn no.	Preceding narration summary	Supporting narrator comment
53–58	King Fred orders his fellow Night Island people to go down to the beach and try to open the mine.	Poor old man! Son, hey my son!
70–72	The girls carry down an axe to help prise open the mine.	Oh no they carry the axe for something! (They) might hit it!
90–96	The white men at the lighthouse notice the events unfolding on the beach.	That one, son, I cry out for poor son!
121–123	King Fred hangs his name plate around his neck to identify himself as chief to the white men.	Chief hey they call him, son that belongs me, that one.
127–136	The white men explain that they have been hitting a mine and that the explosion could have killed them.	Could burn. Oh no! No. Hit it and it will explode! (explosion gestured)
139–143	King Fred runs away, abandoning the Night Island group he leads.	(i) Don’t run away with fright son! (ii) Son was lucky he didn’t hit the right place, if it had blown up he would have turned to nothing.
147–157	The white men instruct them to watch the mine being detonated. Description of King Fred running away is repeated.	(i) Oh dear, he was frightened, that one, my poor son! (ii) If son had hit that right spot the light would burnt them to nothing.
169–171	The smoke of the explosion rises up and spreads across the sky.	Oh no!

The primary narrator does not usually present an assessment or evaluation of the events from their own perspective. Evaluative categories are virtually restricted to two contexts, supporting narrator comments and constructions of represented speech.¹¹ Primary narrator contribution is a combination of clauses recounting the events of the story in simple past-tense inflected clauses and reported speech constructions representing the perspectives of persons in the story. Primary narrator’s stance is of course implicitly embedded in the construction and construal of the narrative rhetoric, but this is achieved through a carefully constructed veil of a seemingly objective presentation of narrative events. Supporting narrators take a complementary role by presenting assessments and interpretations of the events from their perspective in a highly expressive and often forceful way. As such, their comments are filled with evaluative devices, such as, the suffix -laka pathos ‘poor thing’ in (7) and (8); interjections like apa ‘oh dear’ in (7); self-associated kin-terms like maampa ‘son’¹² in (7), (8) and (9); and modals like -tha ‘should’ in (8). While overtly evaluative, supporting narrators’ comments are also affiliative with the primary narrator’s presentation of the events. Looking at the extract presented below in (8), LP recycles the primary narrator’s own reporting of the character’s speech “I am going, I am frightened” (line 5) in ‘oh dear, he was frightened, that one, my poor son!’ (line 8) – albeit with the addition of the emotional interjection apa ‘oh dear’, a surprise interjection associated with concern and bewilderment, and the evaluative suffix -laka added to the kin-term to express compassion.

Table

(8)	1	RP	ngu’ulana	inga-na	para	(.)	thungky	mukan	paa’i-cha
			2plACC	say-NF	white.person		black	big	stand-FUT
			the white people said to you lot, “the big black smoke will stand up”

Table

	2		(0.8)
	3	RP	nhanu	wantantu
			2sgNOM	IGNOR
			“how are you?” (they asked King Fred)
	4		(0.6)

Table

	5	RP	ngayu	waatha-ngka	ngay	wini-na
			1sgNOM	go-PRES.CONT	1sgNOM	frighten-NF
			“I’m going, I’m frightened”
	6		(.)

Table

	7	RP	way	ngampula	[na-
			hey	1plincNOM	?
			“hey we all ?-”

Table

	8	KH	[apa	ngulu	wini-na	nga’a-lu
			INTJ	3sgNOM	fright-NF	dem.dist1-DM
			[maampa-laka
			♀C-PATHOS
			oh dear, he was frightened, that one, my poor son!

Table

	9	KH	[huuuuhaaahh
	10		(.)
	11	KH	chilpu
			old man
			the old man
	12	RP	hm
	(Hill 2007–2010: LCHVDP23Mar07_King Fred)

Likewise in (9), the main narrator describes King Fred running away back to camp leaving his people behind: ‘he ran and ran upwards’ (line 1). This description is repeated in LP’s following admonishment to King Fred ‘again don’t run away with fright son!’. In line 7 LP comments on his lucky escape from injury, and in line 9 provides an alternative possible ending to the story which articulates the character’s close shave with death. These comments along with the admonishment expresses judgement of the character’s actions (eliciting laughter from other participants in lines 4–5), but they are also highly emotive, expressively delivered with sorrow and pity for this character. In this sense, LP’s comments are derivative of, and aligned with, the main line of narration being crafted by RP, the primary narrator. The distribution suggests that the source turns are designed by the primary narrator to build the suspense of the narration and that the supporting narrator’s reactive comments are the conventionalized and relevant response to turns of that nature. The clear division of narrator roles suggests order and organization, rather than spontaneous emotional output.

Table

(9)	1	RP	pintipinti-na	kani-ma
			run-NF	up-DIR
			(he) ran and ran upwards
	2		(.)

Table

	3	LP	kuyi	ngampa	pintipinti	[maampa	wini-nyu=
			again	NEG	run.IMP	♀C	fright-?
			again don’t run away with fright son!

Table

4

RP

[huhhahhaha

Table

5

KH

=huhhuhu

Table

	6	RP	ngulu	atha-nya=
			3sgNOM	cover-NF
			he (took) cover/buried himself

Table

	7	LP	=maampa	mini-ku	wiiyanku	tha’i-na
			♀C	good-GEN	another	hit-NF
			son had luck, (he) hit it in some way
	8		(.)

Table

	9	LP	aachi-na-tha	now	well	antha	ulmpay
			burn-NF-DUTY	now	well	before	nothing
			and it should have burnt well before then, but nothing happened.
	10		(.)

Table

	11	RP	inga-na	ngu’ula	waathi	ngungku	ngaachi	ngu’ulungku
			say-NF	2plNOM	go	dem.dist2	place	2plGEN
			ilpi-n	kukuku
			return-NF	dem
			(they) said “you lot go over there to your place (the mission) and then look back from there”
	(Hill 2007–2010: LCHVDP23Mar07_King Fred)

The ordered nature of evaluative comments is further evidenced by the use of supporting narrators’ comments in response to well-known and recycled narrative content (see Wilkinson and Kitzinger (2006) for the same point regarding surprise tokens in conversations). Supporting narrators exclaim and call out in surprise or dismay even if those events are already known to them, which again suggests these emotive cries have other discourse functions. In (10), King Fred is described as running away after learning of his foolish behaviour. This section of the story is repeated just 10 turns later. The narrator reiterates the interaction with the white men, but with some extra detail added. Even though LP hears this for the second time within a very short time period, she still responds with an equally emotionally laden comment, as shown in line 6 of (10) below where LP exclaims ‘oh dear, he was frightened, that one, my poor son!’.

Table

(10)	1	RP	nhanu	wantantu
			2sgNOM	IGNOR
			“how are you?” (they asked King Fred)
	2		(0.6)

Table

	3	RP	ngayu	waatha-ngka	ngay	wini-na
			1sgNOM	go-PRES.CONT	1sgNOM	frighten-NF
			“I’m going, I’m frightened”
	4		(.)

Table

	5	RP	way	ngampula	[na-
			hey	1plincNOM	?
			“hey we all ?-”

Table

	6	LP	[apa	ngulu	wini-na	nga’a-lu
			INTJ	3sgNOM	fright-NF	dem.dist1-DM
			[maampa-laka
			♀C-PATHOS
			oh dear, he was frightened, that one, my poor son!

Table

	7	KH	[huuuuhaaahh
	8		(.)
	9	KH	chilpu
			old man
			the old man
	10	RP	hm
	11		(.)

Table

	12	LP	apa	maampa	tha’i-na	il’a	aachi-nya	ulmpay
			oh	♀C	hit-NF	light	burn-NF	nothing
			oh if son had hit that right spot the light would burnt them to nothing

Table

	13	RP	ngu’ula
			2plNOM
			(the white men said) “you lot

Table

	14	RP	nhanu	ngi’i	tha’i-ka
			2sgNOM	dem.prox	hit-FUT
			if you will hit this”
	15		(1.3)

Table

	16	RP	ngu’ula	maka-ka
			2plNOM	die-FUT
			(the white men said)“you lot will die”
	(Hill 2007–2010: LCHVDP23Mar07_King Fred)

A final piece of evidence for the ordered and planned nature of supporting narrator evaluative comments lies in the way different classes of recipients deal with them. In (10), LP’s comment in line 6 elicits a response from the main recipient of the narrative, the primary narrator’s daughter (KH). She has remained quiet through much of the storytelling, making her laughter (line 7) and production of ‘the old man’ (line 9) in (10) notable. Similarly, see (9) where KH also joins in laughter (line 5) following LP’s comment. As in these two examples, evaluative comments are places where audience members overtly respond to the narration, providing confirmations, displaying comprehension or showing their own reactions. This suggests that a supporting narrator’s evaluative comments work in tandem with the primary narrator to ensure audience comprehension. At least some of the time they function to draw together information delivered in the main line of narration making its significance clear. For example, in (9) and (10) LP succinctly and overtly articulates the main point and tension built up through the first half of the narrative with these one-liners ‘and it should have burnt well before then, but nothing happened’ ((9) line 9) and ‘oh if son had hit that right spot the light would burnt them to nothing’ ((10) line 12). These comments draw together over 100 turns of narration that described King Fred and his fellow countrymen’s repeated efforts to open the sea-mine.

This idea is further supported by looking at who the evaluative comments are addressed to. Supporting narrators’ evaluative comments are directed talk, typically co-occurring with shifts in body orientation and in eye-gaze selecting the target recipient – eye-gaze has been shown to have an important role in selecting addressees and soliciting responses (see Blythe et al., 2018; Goodwin, 1981; Lerner, 2003; Rossano, 2012). Sometimes the comments are addressed towards the primary narrator and sometimes to the audience. The uptake of the two are different given their role in the interaction. When directed towards the audience, they often elicit a recipient token, an acknowledgement of comprehension like a confirmative (yes or repetition) or a nod. Looking at (8) again, LP immediately before and through the delivery of her comment in line 6 looks up and addresses the comment to KH. KH displays comprehension by laughing (line 7) and saying chilpu ‘old man’ (line 9), and following this RP adds the backchannel hm (line 10). In contrast, supporting narrator comments directed to primary narrators are not usually acknowledged with overt confirmations. Instead, the confirmation is implicitly expressed through the primary narrator providing further detailed narration of the particular events under comment. In (9), for instance, following LP’s admonishment ‘don’t run away with fright son!’, RP adds ‘he took cover’ and then reports the white men’s account to the abandoned group of the dangers and what will happen when the detonation occurs. Similarly, in (10), LP explains ‘oh if son had hit that right spot the light would burnt them to nothing’ (line 12), after which RP recounts the white men saying to the group that if they hit a certain spot on the mine then they will die (lines 13–16). Thus, while not overtly acknowledged by the primary narrator, supporting narrator evaluative comments are consequential for the local trajectory of the story. In this sense, the evaluative comment encourages further narration of details relevant to the important sequence that they highlight. They have a continuer like function, encouraging the primary narrator to continue the current explication.

To conclude, supporting narrator evaluative comments work in tandem with the primary narrator’s development of the main line of narration. While supporting narrator evaluative comments have features of high emotion, they have been shown to have a performative nature, both in terms of planned distribution and conventionalized narrator role arrangements. Supporting narrator evaluative comments overtly and forcefully display stance and interpretative responses to the narrative events, but in ways which are aligned with, and derivative of, the primary narrator’s narration.

6. Discussion and conclusion

This paper has sought to understand the mechanisms through which co-tellers in multiparty narratives in Umpila/Kuuku Ya’u contribute to the story. This is an intriguing issue given the strong preference for co-telling stories in almost all contexts and many resulting instances where some of the co-tellers have no first-hand knowledge of the story or have never heard the story previously. The analysis in this paper distinguished three participant roles – a primary narrator, a supporting narrator and audience role – and orders of conduct associated with each. The analysis then focused on the use of everyday interactional mechanisms of questions and evaluative comments used by supporting narrators. Both these devices are readily employable with limited prior knowledge of the story but the analysis demonstrated that they are able to contribute significantly to the narration, often garnered through extrapolated information and common ground. The analysis also demonstrates that questions and evaluative comments go beyond immediate functions of seeking information or spontaneous expressive reactions. Collaborative goals were shown to be central in the use of both. Their distribution and form indicated that they are often performative. Questions and evaluative comments were shown to highlight key parts of the narrative and build on the action of the story that the primary narrator is producing. These functions ensure that the important themes and aspects of the story are emphasized and ‘driven home’ to the audience. In particular, evaluations are strategically distributed in order to add emotional charge and express stance at narrative highpoints. The primary narrator’s storytelling typically lacks evaluative and perspectivizing devices while the supporting narrator’s contributions tend to be a locus for this. The description of this division of labour between co-tellers underlines the complex collaborative nature of these storytelling events.

Both questions and evaluative comments were also shown to exert influence on the local trajectory of the story. Questions were most interesting in this regard. They were shown to act as a pivot which suspends the narration and shifts the participation framework into a setting in which the co-tellers work together to confirm multiple current aspects of the narration in choral style co-participation. The interactional and extended nature of these post-question spaces demonstrate that joint telling has interactional priority over rapid narrative progressivity in this environment in the story. This once again highlights the importance of intricate co-telling practices within these interactions between Umpila/Kuuku Ya’u narrators.

This paper has presented a detailed analysis of aspects of storytelling in one language and cultural setting. While this micro-analysis helps us understand the nature of co-teller coordination in the corpus of Umpila/Kuuku Ya’u stories, it also poses new questions. Underpinning the description in this paper is a key typological issue regarding the universality and cross-cultural/linguistic diversity in storytelling. How similar or different is the delivery of Umpila/Kuuku Ya’u stories to those told by people in other Australian languages, and further afield? The findings presented relate to a general claim, and one yet to be robustly empirically proven, that multiparty talk is a characterizing feature of Australian Aboriginal storytelling and is different from the participant organization in storytelling in an Anglo Australian setting (Walsh, 2016). This claim is hooked into a wider typological debate which contends that Australian Aboriginal interactional practices have culturally specific features (Blythe et al., 2018; Gardner & Mushin, 2015; Liberman, 1985; Walsh, 1991). The analysis in this paper supports the strong preference for multiparty delivery in one Australian Aboriginal setting. However, the degree to which this is similar or different from other traditions requires some qualification, and ultimately, more data and analysis. A number of researchers have shown in English and other Indo-European languages/cultures that in some storytelling contexts multiparty delivery is typical – between family and close friends – and that it forges and signals shared group membership (Georgakopoulou, 2007; Lerner, 1992; Ochs & Capps, 2001; Ochs et al. 1992). The Umpila/Kuuku Ya’u stories bear similarities to the analysis presented in this body of work: in the identification of supporting roles (Lerner, 1992); fluidity in moving between roles in the narration (though less noted in the Umpila/Kuuku Ya’u context) (Georgakopoulou, 2007); and the locally situated and emergent nature of the shared narration (Georgakopoulou, 2007; Ochs & Capps, 2001). A better understanding of the nature of these contexts through a comparative lens is required: how similar is the dynamic within the small and close-knit community of Umpila/Kuuku Ya’u narrators to contexts where multiparty narrative is also typical in Indo-European languages and cultures? As has been speculated in recent work, some of the differences in storytelling observed in smaller speech communities may be a condition of the intimate nature of these communities and the high common ground between interactants (Stirling & Green, 2016). To continue with similarities, the multiparty interaction in Umpila/Kuuku Ya’u is achieved through the use of everyday interactional devices like questions and evaluative comments often employed with cross-linguistically familiar functions. However, the default nature of this mode of storytelling in Umpila/Kuuku Ya’u along with the interactional efforts the co-tellers go to collaborate does appear to be somewhat different from conversational European storytelling practices, though perhaps just in frequency and degree. Recall the ways in which the progress of a Umpila/Kuuku Ya’u story is halted and multiple co-tellers jostle for the floor and work over the details of the story with choral co-productions across many turns. These intricate co-telling practices unfold in a way which does not derail the storytelling action but is treated by other interactants as expected and pro-social. This interactional preference to deliver stories collaboratively in almost all contexts (thematically and interactionally) and sanctionable actions if one does not (Hill, 2018, pp. 41–42), suggests a ratcheted-up or amplified version of what is presented in the discussion of co-told conversational narratives in European language contexts (for example in “assisted storytelling” contexts explored by Lerner (1992), “small stories” by Georgakopoulou (2007) and by Bamberg and Georgakopoulou (2008), and “living narratives” by Ochs and Capps (2001)). The roles appear more delineated, and the interactional tools employed somewhat more conventionalized in the Umpila/Kuuku Ya’u stories. This may be a consequence of the comparatively ratcheted-up preference for collaboration: a stronger or more broadly applied social requirement may generate more conventionalized tools for collaboration to allow interactants to meet this preference.

Acknowledgements

I am grateful to the Umpila and Kuuku Ya’u speakers who taught me about their language and shared their stories with me, and to the wider Lockhart River community who encouraged and supported this work. I thank the Lockhart River Aboriginal Shire Council and Lockhart Arts and Cultural Centre for invaluable support. My fieldwork was funded by the Endangered Languages Documentation Programme (grant MDP0133) and internal grant funding from Language and Cognition Group at The Max-Planck Institute for Psycholinguistics. I thank Jean-Christophe Verstraete, Nick Enfield and Stephen Levinson for many insightful comments on the analysis of these stories through my PhD candidature. Part of this material was presented in talks at The Max-Planck Institute for Psycholinguistics and The University of Leuven, and also at the Australian Languages Workshop at Marysville, Victoria in 2018. I thank members of the audience for useful comments on all these occasions. I am also grateful to Lesley Stirling and two anonymous reviewers for many helpful comments and suggestions on a previous version of this article.

Disclosure statement

No potential conflict of interest was reported by the author.

Data availability statement

The data that support the findings of this study are available through restricted access (community consent required) in the Endangered Languages Archive at http://elar.soas.ac.uk/deposit/0059, reference number paman-hill-0059.

Additional information

Funding

This work was supported by the Australian Research Council under Grant CE140100041; Endangered Languages Documentation Programme under Grant ELDP MDP0133; and Max-Planck-Gesellschaft under an internal fieldwork grant within the Language and Cognition group at the Max-Planck Institute for Psycholinguistics.

Notes on contributors

Clair Hill

Clair Hill is a Lecturer in Linguistics in the School of Humanities and Languages at The University of New South Wales and an Adjunct Fellow in The MARCS Institute for Brain, Behaviour and Development at Western Sydney University. She has a special interest and expertise in Australian languages, especially Paman languages (Umpila and Kuuku Ya’u) of north-eastern Cape York Peninsula. There are two central threads in her work: an exploration of the interaction between language, cognition and culture, and collaborating with communities to translate this research into useful language documentation and language revitalization products.

Notes

1 The main exception being before time narratives (well-known central Australian counterpart referred to as dreaming or dream-time stories) which deal with the activities of mythological beings and/or ancestral peoples and the formation of the natural world as it appears today. A detailed discussion of the proprietary rights that motivate a single-narrator mode of storytelling for these narratives can be found in Hill (2018, pp. 40–52).

2 Relatedly Mushin (2016) and Sansom (1980) showed differences in what epistemic access means for interactional norms for Garrwa and people of Darwin regions respectively.

3 These projects were: (i) Oral Histories and Stories of the Umpila and Kuuku Ya’u Peoples of Lockhart River Language and Culture Documentation in 2004, undertaken in collaboration with Lockhart River Land and Sea Management Centre and Lockhart River Arts and Cultural Centre, funded by Australian Institute of Aboriginal & Torres Strait Islander Studies; (ii) Cape York Peninsula Language Documentation Teams in 2005, undertaken in collaboration with the Lockhart River Arts and Cultural Centre funded by Maintenance of Indigenous Languages and Records Program; (iii) Five Paman Languages of Cape York Peninsula between 2007 and 2011, undertaken in collaboration with Lockhart River Aboriginal Shire Council, which supported PhD fieldtrips of the author and was funded by the Endangered Languages Documentation Program.

4 Further specialised participant roles could be teased apart to account for aspects of the narration, e.g. following Goffman (1979, 1981) and related commentaries, roles such as transmitter vs. author and addressed recipient vs. unaddressed recipient may have some bearing on these stories. However, for the purposes of this analysis, the proposed distinction between primary and supporting narrator roles has considerable analytical power and accounts for the orders of conduct observed through stories and through the whole story collection analyzed.

5 Though in some instances the conditioning factors behind the allocation of a primary narrator role are not explained by epistemic access and may be motivated by complex social and cultural factors relating to country and kinship networks/kin avoidance (Hill, 2018, pp. 56–57). Further work on such factors remains a topic for future research.

6 The abbreviations used in glossing follow those of the Leipzig Glossing Rules (https://www.eva.mpg.de/lingua/resources/glossing-rules.php) with the addition of these abbreviations, used in example glossing: ♀ female anchor in kinterm, C child, DIR directional, DM demonstrative marker, exc exclusive, IGNOR ignorative, inc inclusive, INT interjection, NF non-future, PRES present. These notation conventions are used to indicate the timing and other aspects of the talk: ::: colons indicate a prolongation in production; ° the degree sign indicates that the talk following it was markedly quiet; = contiguous utterances, an equal sign is placed either side of utterances with no time gap; [ overlapping utterances, left hand square brackets mark the start of the overlap; (0.0) intervals within and between talk are measured in seconds and indicated in brackets; (.) micro-pause, a pause less than (0.2).

7 There are, of course, many interactional contexts where people ask questions they know the answer to: in interviews, in courtroom settings, in class rooms (Atkinson & Drew, 1979; Clayman & Heritage 2002; Levinson 1988).

8 This relates to some other work on questions in Australian Aboriginal languages, like Eades (1993) who suggests there is a preference for question formats which exchange information rather than just seek information. Eades (1993, 2013) and Walsh (1997) suggest Aboriginal Australian people tend to avoid answering some direct information seeking questions, particularly those of a personal nature. Albeit not an everyday interaction, this study finds no interactional issues with speakers producing information seeking questions or responding to them in this storytelling context.

9 Timing between turns in the transcripts (and subset of the analyzed narrative collection) are accurately measured directly from the digital sound file. Following conventions established in Conversation Analysis silences of 0.2 s or more were measured, and silences of less than 0.2 s are indicated by this convention (.) and referred to as a micro-pause (Atkinson & Heritage, 1984; Jefferson, 1989).

10 Some activities in conversation have been described as being achieved chorally rather than serially; e.g. greetings and leave takings (Lerner, 2002) and congratulations (Goodwin & Goodwin, 1987).

11 Verstraete (2011) makes a similar point for reported speech in narratives in neighbouring language Umpithamu.

12 In contrast to other parts of the narration which typically use simple human classificatory terms for third person reference (e.g. chilpu ‘old man’, anthaya ‘girl’, kanga ‘initiate’, para ‘white man’), supporting narrator’s evaluative comments often feature self-associated kin-terms. In the King Fred narrative, MP repeatedly refers to the protagonist as maampa ‘(my) son’ as can be seen in the opening example (4) ‘that one, son, I cry out for poor son!’; across the 10 comment utterances in this narrative, there are nine uses of maampa ‘(my) son’ (see Table 3). Often the use of kin-terms in these contexts is strategic and helps to ground the speaker’s assessments and reactions via a demonstration of the genealogical connections to the said character (Hill, 2016, 2018).