You, and you, and I belong to a network of constantly evolving institutions defined and differentiated on the basis of shared access to speech registers, linked repertoires of gestural, referential, and syntactic behavior. In enacting these linked repertoires—and this enacting includes passive comprehension as well as active production—we enact our membership in various speech communities. All of us, at every moment of our waking and dreaming lives, are participating in multiple gestural-referential-syntactic communities. But only certain kinds of behavior and certain kinds of artifacts count as language for the purpose of linguistics. One of the ways linguistics marks off its domain of inquiry is by specifying language as a phenomenon with particular qualities of extension in time and space.
Symbol-using behavior does of course, like everything else, have characteristic topologies, but it is by specifying a topology in advance that linguistics maps out a subset of symbol-using behavior as its own. This has consequences for linguistics, since certain phenomena that might usefully be considered part of the everyday unfolding of language become analytically invisible. It also has consequences for speech communities, in particular speech communities experiencing rapid shift to a language of wider communication. The common way of describing this phenomenon—in which, in the space of a generation or two, a language goes from being the main medium of communication in the community to being limited to a handful of domains, and in which children are no longer exposed to it in ongoing everyday interaction with fluent speakers and no longer become fluent speakers themselves—is to say that the language is endangered or going extinct. This usage, with its roots in an analogy from living things to languages, reflects a particular way of understanding how languages endure and change over time and space, that is, a particular way of understanding the topology of language.
This article shows how linguists have made certain assumptions about the topology of language and points to consequences these assumptions have had for the way documentary linguists have engaged speech communities under pressure to shift to a language of wider communication. My example comes from northwestern Australia in 1970. Topological qualities figure large in the social ontology of language, the set of assumptions shared among members of a community about the nature of language as a thing in the world. Linguistics has had a role in shaping these assumptions, not just in the social and cognitive sciences but across a broader array of domains, including law, government, and, more recently, the diffuse public of concerned citizens who avidly consume journalistic accounts of developments in evolutionary anthropology and the science of human cognition, not to say endangered language documentation. Endangered language reportage is never simply about a particular language but about how that language and its speakers fit into a topological schema constituted by assumptions about how populations of symbol-using creatures array themselves across the surface of the earth, how these populations grow, migrate, divide, and fuse over time, and how common participation in a speech community figures in the capacity of a population to represent itself, both internally and to outsiders, as an autonomous political community. When we make simplifying assumptions about the social topology of language, we also make simplifying assumptions about the topology of social identity. In doing so we risk both denying recognition of collective autonomy to communities that don’t have the shapes we’ve come to expect and foreclosing on possibilities for transition to some novel configuration of collective regard that could mitigate the very real pressures, language-related and otherwise, these communities face.
To say that language has topological qualities is to say that it has characteristic patterns of extension in time and space. Of course what we’re really saying is that language-enacting communities have characteristic patterns of extension in time and space, since language’s presence in the world, its ongoing manifest availability to sentient awareness, is contingent on the presence of a community of symbol-using beings who enact it, and enact it, and enact it. This is true of anything whose existence is a product of intentional behavior, but for many people language poses special problems in this regard because its sensible traces are ephemeral. Consider, for example, the comments of Bill Hillier, a theorist of urban form, on the relationship between language and built space. There exists, Hillier writes, “a class of artefacts which are no less dramatic in their impact on human life [than physical artifacts], but which are also puzzling in themselves precisely because they are not objects, but, on the contrary, seem to take a primarily abstract form. Language is the paradigm case. Language seems to exist in an objective sense, since it lies outside individuals and belongs to a community. But we cannot find language in any region of space-time. Language seems real, but it lacks location.” Or rather, Hillier continues, it is not that language and other abstract social artifacts do not manifest in space-time but that “these space-time appearances are not the artefact itself, only its momentary and fragmentary realisations” (Reference Hillier2007, 65).
The built environment too reflects the ongoing transitory social enactment of artifacts that exist strictly on the basis of convention, that is, artifacts whose presence in the world is given by the systematic articulation of referential (representational, meaning-generating) events to tangible things. Systems of referential convention (speech registers are one kind) afford the principle by which an accumulation of physical traces is configured into a language or a city. Cities, like languages, “are space-time manifestations of configurational ideas which also have an abstract form” (Hillier Reference Hillier2007, 68). The difference is that in the case of cities the tangible precipitate of the artifact-enacting process is a lot more durable, giving us a chance to inspect it at our leisure and form hypotheses about the relationships between the topologies of the social processes that make cities possible and those of the built artifacts that make them sensible.
Can we imagine something comparable for language? A remote-sensing apparatus that would afford us a synoptic feed of the ensemble of fleeting gestural artifacts and their haptic, sonic, visual, and graphic traces, ranging from touching your elbow to vocalizing to signing to the appearance of glyphs on the screen as I type, that make up the instantaneous activity of a speech community? If we had such an apparatus, how would we represent the data it generated in a way that made patterns manifest, how would we visualize it, how would we map it? Even in writing-saturated societies, the vast majority of language-enacting continues to play out on a temporal modulus dictated by the phenomenal experience of sensible co-presence: you and I, walking down the street, talking on the phone, trading text messages. Until rather recently, there was no way to monitor symbol-using activity on this modulus, let alone map it, for more than a handful of interactions (Berson Reference Berson2015).
This has not stopped anyone from forming hypotheses about language’s topological character. Linguists, among others, have had firm ideas about how language (the type-phenomenon, the behavioral faculty), languages (instance-artifacts, enactments of that behavioral faculty), and language-enacting communities unfold in time and space. We can imagine a number of ways of understanding language as a topological phenomenon. Elsewhere (Berson Reference Berson2012) I’ve discussed the hegemony of dendriform, as opposed to meshlike, topologies of language’s extension in time. Here I focus on language’s extension in space.
How people talk varies from place to place, and this variation unfolds in two more-or-less complementary patterns. On the one hand, the joint appearance of a particular bundle of symbol-using behaviors may be limited to an approximately convex region of the earth’s surface. In this case we say that the geographic topology of language consists in a tiling or tessellation of these convex regions. Depending on how we specify the bundle of behaviors we’ll get different tessellations for the same area, and tessellations mapped on the basis of one kind of behavior (vocal gestures, say) may not be congruent with those defined in different terms. Neighboring communities might use the same system of speech sounds but use them to form different words. The alternative to a tessellation is a reticulation: similarity in linguistic behavior may map not to a tiling of convex regions but to a network of intersecting lines. If a tessellation is like a map of political boundaries, a reticulation is like a route map.
Linguists have tended to assume that the topology of gross thresholds of linguistic difference—mutual nonintelligibility, among other things—takes the form of a tessellation, even if the spatial distribution of particular traits is reticular (Labov Reference Labov2007). Let’s look at how this has played out in the field.
Summer Institute of Linguistics
The history of linguistic ethnography is a history of what linguists and the institutions supporting them said and did. The people who spoke the languages in questions have, by and large, been represented neither in published accounts nor in the archived correspondence in which would-be ethnographers and government bureaucrats negotiated terms of access to Indigenous populations. From the onset of the colonization of Australia up to the present, with vanishingly few exceptions, linguists, evangelists, “native welfare” administrators, and policy makers have all had one thing in common: from the point of view of the people whose languages they aspired to catalogue, they were whitefellas (cf. Bashkow Reference Bashkow2006). Yet we do have a textual basis from which to form impressions as to what the informants, the speakers, made of the survey encounter: the field data itself.
Recordings and transcripts of direct speech elicitation represent one of the main sources from which to form impressions of what the speakers of Australian languages said in the course of encounters with one another and settlers. In contrast to the Americas, where reports of native oratory have long served a variety of rhetorical purposes in settler society, or Aotearoa–New Zealand, where Māori-authored grammars of Māori sprang up not long after colonization, in Australia, prior to the start of the land claims movement in the 1960s, the sole register in which Indigenous actors spoke for posterity, as far as the settler society was concerned, was that of the Dreaming: world-making stories of totemic ancestors who inhabit a time out of time.
In 1970 the Australian Institute of Aboriginal Studies (AIAS) commissioned the Summer Institute of Linguistics, Australian Aborigines Branch (SIL AAB), to conduct a survey of the languages spoken in the East Kimberley, a remote part of the northwest that had long been a problem area for Australian language classification. Transcripts of the conversations that yielded the 1970 “Surveys of Languages and Dialects of the North-East Kimberleys” show Indigenous Australians speaking in the present, as cattle station stockhands, as participants in the budding Indigenous revival movement at Kununurra, as women and men who grew up on this mission, worked on that station, have a father or brother on that reserve—above all, as people whose availability to answer linguists’ questions is constrained by the exigencies of wage labor. These transcripts also show Indigenous Australians acting not simply as passive sources of basic vocabulary but as guides, taking their linguist interlocutors in hand: correcting the linguists’ confusion of toponyms, ethnonyms, and glossonyms, setting them straight on which words change reference according to context and which words are names, telling them what they can and cannot truthfully write down about the language situation on this station, commanding them, in exasperation, “You look now,” while they explain the circulation of language and social identity through space and time.
Not all transcripts of elicitation offer the chance to hear the speakers as socially situated individuals rather than as linguistic type specimens. What makes the AIAS/SIL East Kimberley survey exceptional is that, even as the fieldworkers were compiling lists of basic vocabulary with the aim of using cognate densities to gauge degrees of relatedness, or at least of mutual intelligibility, they were also trying to figure out how many languages they were dealing with and where the speakers of these distinct languages had come from. The missions and the cattle economy had engendered a situation in which Indigenous people from all over the Kimberley and adjoining parts of the Northern Territory lived and worked together and shifted from place to place, compounding the difficulties Capell (Reference Capell1940) had encountered in the area in the 1930s. The only way to find out “which language” one was compiling a word list for was to ask the speaker. Speakers’ biographies and speakers’ metalinguistic expertise were the keys to reconstructing a “time before” the boundaries of intelligibility and descent had been blurred and erased by evangelism and wage labor. Asking the speakers was also the only way to find out where in the immense expanse (over 400,000 square kilometers) of the Kimberley one might find another potential informant who spoke the language named by a particular glossonym.
When SIL entered Australia, it was already recognized as the largest and most sophisticated linguistic fieldwork operation in the world (Dobrin Reference Dobrin2009). Into the 1950s SIL had concentrated its operations in Latin America, where the organization got its start, in the mid-1930s, translating the New Testament into the languages of highland Guatemala. Bible translation was, in fact, the point of SIL’s work and a strong factor in its operational efficiency, since the organization was never at a loss for idealistic young evangelical Christians ready to relocate to remote parts of the world for two years of total immersion in the local language, with financial support from their church communities back home. The vast majority of SIL’s fieldworkers came to the work with no training in linguistics—the Summer refers to an eight-week boot camp the organization ran in the Ozarks for new recruits. As a result, SIL had developed a program for teaching nonlinguists the art of acquiring a new language in the field without the benefit of dictionaries or intermediaries that was the envy of linguistics departments and colonial administrators around the world. When SIL expanded into the Pacific, it used its fieldwork bootcamp as a calling card, offering a modified version for Australian Department of Territories field patrols in exchange for help from the Australian government setting up a field station in the eastern highlands of New Guinea. This was in 1957. By the time SIL established its Darwin base of operations seven years later, its capacities were well known to those in the Australian government charged with the oversight of native peoples.Footnote 1
At this point in time, SIL’s priority was salvation, not salvage. Topological reconstruction was part of SIL’s work to the extent that understanding how neighboring and overlapping speech communities fit together, spatially and in terms of mutual intelligibility, would abet the planning of literacy development campaigns by guiding decisions as to which languages to develop writing systems and pedagogical materials for. You wanted to focus on languages that were not in danger of going out of use any time soon. You could not develop a literacy apparatus for every local language, but perhaps you could identify two or three that were stronger than the others, and induce speakers of the other languages to have their children taught to read and write in the stronger neighbor. SIL’s longer-term aim was to make the Gospel available to potential Christians in a language that was, if not the one they had grown up speaking, at least closer to their “heart language” (Handman Reference Handman2014), for example, closer in its strategies of information structure and rhetoric to what felt natural to those who would learn to read in it. Yet by the mid-1990s, SIL was on its way to becoming a model citizen among endangered language activists. The key to SIL’s reinvention was its atlas of dialect geography, Ethnologue.
Ethnologue originated in 1951 as a ten-page mimeographed circular. It was in 1971, when the editors started using a computerized database to generate the entries, that it began to assume the form it has today. These days, Ethnologue presents itself as a universal glossonymicon. When a new edition of Ethnologue comes out, it makes news. When journalists—and linguists—need to know “how many” languages there are in the world and what proportion will be gone in a hundred years, the numbers, ultimately, come from Ethnologue. This is partly a product of the fact that in a 1992 article widely considered to have set the agenda for the subsequent revival of documentary linguistics, linguist Michael Krauss relies on Ethnologue to make his case that the threat to global biodiversity was “relatively mild” by comparison with that to linguistic diversity. Krauss praised SIL for taking the initiative and demanded to know why, with the exception of SIL, linguists had been “so much quieter about” the impending extinction crisis than their colleagues in biology (Reference Krauss1992).
Where did the figures that populate Ethnologue come from? A look at SIL’s activities in northwestern Australia in 1970 sheds light on its field methods.
The Setting
Nineteen seventy was the tail end of what Indigenous oral histories refer to as the “station era,” a period extending from the 1880s up through the 1960s (Stanner Reference Stanner[1966] 1979; Berndt and Berndt Reference Berndt and Berndt1986). With the introduction of livestock to the dry rangelands of central and northwestern Australia, Indigenous Australians found themselves denied access to sources of water and food and struggled to maintain an adequate level of nutrition. Within the space of a generation, they were compelled to adapt to a new regime of provisioning, one based on flour and meat rations provided by missions and stock stations (ranches). By the 1930s the Indigenous community had become essential to cattle production and the ancillary activities—domestic work, road and airstrip construction, tracking fugitives for the police—it engendered. Conditions varied from station to station, but across the deserts of central and northwestern Australia Indigenous inhabitants were subject to living and working conditions that bordered on feudal obligation and chattel slavery. Stock workers were fed a diet of dry rations and beef. Their dependents often received nothing more than the bones and offal of animals that had been slaughtered to feed the station’s employees. In the Northern Territory the official minimum age for Indigenous stock workers was twelve, and the Aboriginals Ordinance fixed a schedule of cash wages, to be garnished for the upkeep of a worker’s dependents. These wages amounted to approximately twenty percent of those paid to white workers under the trade union–negotiated Cattle Station Industry Award. Indigenous women were often held in sexual thrall. In Western Australia, prior to a 1967 federal constitutional referendum that gave the Commonwealth government responsibility for making law with respect to Indigenous populations, there was no federal oversight of state Indigenous labor policy.
By 1970 the dry regions cattle industry was in crisis. Open-range pasturing combined with unrestricted herd growth and weak markets had produced unremediable soil erosion. In 1965, the North Australian Workers’ Union succeeded in getting the Commonwealth Conciliation and Arbitration Commission to conduct hearings on the extension of the Cattle Station Industry Award to black workers. At length the commission decreed that the Award Wage would apply to Aboriginal cattle workers from December 1968. Many cattle industry observers were of the opinion that coerced Indigenous labor was essential to the industry’s solvency. When, in 1968, the award decree came into effect, many station managers in the Northern Territory simply evicted Indigenous workers and their families from the stations, bringing the station era to a close. This was the context in which the 1970 East Kimberley survey unfolded.
Transcript
Much of the dialogue in the typed transcripts, from tapes recorded October 12–24, 1970, at cattle stations, missions, and Aboriginal reserves across the East Kimberley, reflects the fieldworkers’ concerns, first, to find enough reliable informants to scrape together word lists for the speech forms listed in Capell’s Linguistic Survey of Australia (Reference Capell1963) and, second, to connect languages spoken by an informant with places of origin.Footnote 2 The principal object of the interviews was to collect basic vocabulary (Berson Reference Berson2014a). To supplement test-list data, “informants were also engaged in discussions about languages and tribes of their area. This was also recorded, and transcription made where they seemed to contain information of interest” (Glasgow, Hocking, and Steiner Reference Glasgown.d.).
The transcripts of the October 1970 East Kimberley survey (Glasgow, Hocking, and Steiner Reference Glasgow, Hocking and Steiner1970) allow us to hear how the production of facts about what languages there are and about how these languages are arrayed in space involves the coordination of talk across multiple speech registers, each embodying an exemplary way of referring to the constituents of a particular ontology. Speakers and linguists brought with them markedly different ontologies of language, land inhabitance, and social identity. Distinct models of what language (or something roughly comparable to what the linguists intended by the word language) consists in demanded distinct patterns of phonological and syntactic expression, even when both speaker and linguist were speaking English, and to get their sense across, both parties, but especially speakers, relied on codeswitching.
In the conversation snippet reproduced below, the ethnographer simultaneously tries to elicit the speaker’s opinion on the identity or mutual intelligibility of the forms of speech named by distinct glossonyms and to locate those speech forms in space. Of necessity, the speaker’s personal history comes into play:
Consultant: I think the Forrest River tribe talk them Gunin language—that’s similar to Wunambal language.
Ethnographer: There’s another one I heard about—Gambera. Do you know if that Gambera language at Kalumburu?
Consultant: I never heard of it.
Ethnographer: Any Wunambal?
Consultant: Worora and Ngarinyin—you’ve heard of Ngarinyin language? Gibb River and Kurundji—a lot of them.
Ethnographer: Kurundji too eh? Is that Ngarinyin?
Consultant: Yeh.
Ethnographer: What do you think most people speak at Kalumburu … What language most of them speak?
Consultant: Wunambal language.
Ethnographer: Only a few Worora eh?
Consultant: Yeh, few Worora and Ngarinyin—few Ngarinyin—but mostly the Worora language comes from Mowanjum [Mission]—down Derby—well those people up there talk that language.
Ethnographer: Well I was there at Mowanjum six months ago or might be four months ago and I met a fellow called H.—you know H.? He came from Kalumburu Mission.
Consultant: He’s my father—I. you mean—I. from Kalumburu—mostly some people call him H.
Ethnographer: What’s his wife’s name?
Consultant: D.
Ethnographer: Yeh, I think that’s right—well they were there at Mowanjum and they said he talked Gambera and his wife she said she talked Gunin. Does that make sense to you?—that sound right?
Consultant: Well all depends on—we got different sort of tribes—like—the Worora and the Wunambal tribe—
Ethnographer: That Gambera same as Wunambal do you think eh?
Consultant: Yeh, I think so—but it’s a bit harder—you know—Gambera.
Ethnographer: So it’s a bit harder and Gunin do you think that’s the same as Gambera or is it different?
Consultant: I think it’s a bit different.
Ethnographer: Do you think that Gunin bit same as this Forrest River language?
Consultant: Yeh, I think so—Well Forrest River and Kalumburu they talk the same language—The people at Forrest River came from Kalumburu Mission.
Ethnographer: Same language eh? But might be words a little bit different do you think?
Consultant: Yeh. Some words a little bit different.
Ethnographer: Like they can talk to each other.
Consultant: Yeh they can talk to one another.
Ethnographer: Do you know any of that language? Wunambal eh?
Consultant: I don’t know how to talk Wunambal, but I can understand it you know.
Ethnographer: Yeh, now another one I heard—that’s Gwini—another name—do you know that name? Don’t eh? Long time ago. It’s probably called … I just wondered if you’d heard that name?—Gwini— Yeh, well I think that’s about all I want to ask you so thanks very much for your trouble.
(transcripts 4–5)
Here the speaker struggles to respond in the register in which the ethnographer has posed the questions. This is a register of linguistically differentiated tribes, each with its own fixed point of geographic origin. Other speakers, such as D. at Wyndham Reserve, adopt the idiom of whitefella linguistic ethnography more confidently:
Ethnographer: Can you tell me about Wumbulgaři—all the people here Wumbulgaři—or they when they say some Yeidji and some Wumbulgaři?
Consultant [D.]: Yeh they mixed people you know—some Wuladjangaři, some Wumbulgaři some what you call it now?
Ethnographer: Wunumbal? Footnote 3
Consultant: Yeh they talk different too.
Ethnographer: That’s different eh, Wunumbal?
Consultant: Two nations mixed—two tribes mixed.
Ethnographer: Wumbulgaři and Wunambal different?
Consultant: All the same, all the same—they understand one another.
Ethnographer: They understand one another a little bit eh?
Consultant: Yeh—no question, but not real well.
Ethnographer: Yeh—but different word here and there—some word different eh?
Consultant: Yeh that’s right.
Ethnographer: What about Gambera? There some people here Gambera? Gambera from Kalumburu Mission way—you know that—they talk Gambera language?
Consultant: Gambula language, Gambula—
Ethnographer: Gambera—
Consultant: All different tribe names—all different tribe … like that.
Ethnographer: Yeh.
Consultant: All the different names.
(transcript 10)
D. smoothly picks up the linguist’s frame of reference, matching, as best he can, his own locally situated knowledge to points of reference—proposed glossynyms—supplied by the ethnographer. The proposed glossonyms represent Capell’s distillation of two generations of ethnography, by degrees amateur and professional. Capell’s Survey was shaped not just by the heterogeneity of the source material but by the political context in which it was commissioned: the founding of the AIAS and, by extension, the institution of a government mandate to complete a census of historically attested, linguistically delimited Indigenous tribes before those tribes disintegrated (Berson Reference Berson2014b). That the cattle stations that provided the setting for many of the meetings between native speakers and linguists in October 1970 were understood to be one of the main sources of the pressures tending toward the disintegration of the region’s native tribes was one of the survey’s ironies.
D. seems unfazed by the profusion of unfamiliar names offered by the ethnographer, casually asking the linguist to repeat one: “some Wuladjangari, some Wumbulgari some what you call it now?—Wunumbal?—Yeh they talk different too.” Eventually, he reaches the limit of his capacity to characterize differences among named speech forms in the idiom proposed by the ethnographer. This is an idiom in which mutual nonintelligibility is a matter, first and foremost, of differences in basic vocabulary. The linguist prompts: “Yeh—but different word here and there—some word different eh?” D. assents but, offered another unfamiliar name, responds with, “All different tribe names—all different tribes … like that.”—that is, he refuses to commit as to what exactly marks the languages, tribes, and names in question as different.
This conversation took place at Wyndham Reserve on October 16, 1970. It was recorded on an early section on the tape numbered 17 in the ethnographers’ labeling (later A2183 in the Australian Institute of Aboriginal and Torres Strait Islander Studies [AIATSIS] media indexing system). In other instances, the register boundary between speaker and ethnographer is clearer. Tape 16 (AIATSIS A2182) features a conversation the same ethnographer had with J. at Fork Creek Reserve, again on October 16. After sounding J. out on Gambera and Wunambal, the linguist turns to the mysterious “Gingara,” provoking disagreement between James and an unidentified second speaker:
Ethnographer: And what about Gingara?
Consultant [J.]: Gingara you know King River?
Consultant 2: What King River?—that Wula—That’s country that one—King River.
Ethnographer: And what’s that Kalumburu?
Consultant: Same.
Ethnographer: Is that language or place?
Consultant 2: Language.
Consultant: No, that country they call Kalumburu.
Ethnographer: Kalumburu.
Consultant: That country belong River.
Ethnographer: And what they call the language there? The language that country—What language they talk?
Consultant: Worora.
Ethnographer: Worora.
Consultant: Yeh.
Ethnographer: And uh.
Consultant: I don’t know.
Ethnographer: What about—you know Wila Wila.
Consultant: Well some—same right round—Wila Worora and Gambera.
Ethnographer: Where there some man that speaks Wila. Where’ll find that man?
Consultant: Wila this here … (mandjaway)Footnote 4
Ethnographer: (Mandjaway)?
Consultant: Yes, (mandja)
Consultant 2: (Mandja) Where’s that?
Consultant: Over here longa Mission they call (mandja)
Ethnographer: Mowanjum?
Consultant: Mm.
Ethnographer: Oh yeh—and what about this mob at Kurundji?—What language they talk there?
Consultant: Wuladja, all the way.
Ethnographer: Oh yeh.
Consultant: You gettim here, you gettim there, all right through along Kurundji and Gibb River.
Ethnographer: And Gibb River too eh?
Consultant: Mount House (yirambu) all around everywhere.
Ethnographer: Yeh.
Consultant: Broome, Derby, all Wuladja.
Ethnographer 2: Big mob eh?
Consultant: Mm.
Ethnographer: You know this name Wembria?
Consultant: Belong this country or Kurundji?
Ethnographer: I don’t know? I’m asking you. You know that one?
Consultant: Wembria—language or country?
Ethnographer: I don’t know?
Consultant: Wembria—men or what?
Ethnographer: Men I think, or language?
Consultant: Wembria—might be country, I think.
(transcript 10)
Confronted by a demand to attach an unfamiliar language name to a hypothetical person or persons situated at some fixed point in space (“Where there some man that speaks Wila. Where’ll find that man?”), J. introduces a nomicon of his own, putting the ethnographer in the position of trying to match unfamiliar names to his own knowledge: “(Mandjaway)? … Mowanjum?” Asked to situate a place in linguistic topography, J. becomes expansive, depicting language as something organized in space not as a tessellation of discrete, abutting tracts, like lots on a surveyor’s map, but as a reticulation, with a particular speech form extending along a series of possible itineraries (“You gettim here, you gettim there, all right down … Broome, Derby, all Wuladja”). Finally, confronted with a bare indicium, a name without a country, he demurs. When the ethnographer is not even sure of his own ontology (“Wembria—language or country? … men or what?”), how can the speaker be expected to translate local knowledge into that ontology?
Elsewhere, scrambling for traction, the linguists cast aside any effort to meet local ways of understanding language halfway. In some cases the local expert responds easily, as in this exchange, which took place October 14 at Fork Creek Reserve, most likely involving a man identified as K.
Ethnographer: What’s this mob here now?
Consultant [K?]: This oh callim King River gabarindjʌ—King River gabarindjʌ gura wundjadun different—different mifella callim different country—half way—King River half way this way. —From right up Dunham come this way—this way—different way.
Ethnographer: What language?
Consultant: Wulaidja and Djerak.
Ethnographer: So that’s three languages—Wulaidja, Djerak and Yeidji. You got another language here?
Consultant: No we got none.
Ethnographer: Just three—all right.
Consultant: You can’t put-im-down another language.
(transcript 2)
Note how gracefully the speaker switches registers. When asked an open-ended question—“What’s this mob here now?”—he sketches a picture of a graduated network of affinity organized in space along a series of linked itineraries, with names for social collectivities, stretches of land, and ways of talking changing as one follows a path. His speech takes on a distinctive rhythm characterized by a pattern of alternating stress suggestive of walking: “different mifella callim different country—half way—King River half way this way.” Asked bluntly, “What language? … You got another language here?” he switches, instantly, to the clipped rhythm of the well-behaved atlas: “Wulaidja and Djerak. … No we got none.”
Not all speakers were so acquiescent. In this conversation, recorded at Kalumburu on October 22, 1970, the next-to-last day of the survey, A. resists prompts to make his knowledge more legible:
Ethnographer: Well you understand all these different kinds of Ngarinyin do you A.? And was this man talk the same kind of Ngarinyin as this man?
Consultant [A.]: Yeh, but (nebi) that one (wa’tad).
Ethnographer: Anybody else talk Ngarinyin in this country or—that well what I mean is your—these fellows Ngarinyin might be a little bit heavy or is it light [more or less difficult to understand]?
Consultant: You look now—that Ngarinyin goes like this see—there’s Gambera, Wunambal, Gunin—all that much see.
Ethnographer: Yeh.
Consultant: All come different, different and this Ngarinyin here—but we—we in this way you see—all that Ngarinyin here—inside here see—this is outside people Ngarinyin see—this is border of this Ngarinyin—all around language see—we have Ngarinyin, Wunambal, Gambera—all the same—We hear Gambera—that’s the edge of [the] boundary of all the Worora, Ngarinyin.
Ethnographer: Yeh.
Consultant: And then down to Gibb River language here see.
Ethnographer: Yeh.
Consultant: But then this lot the edge up to Ngarinyin.
Ethnographer: Yeh.
Consultant: All these edges can mostly pull their own language again see.
Ethnographer: Yeh.
Consultant: This is how this works.
Ethnographer: Well which way these two men.
Consultant: The end of this border.
Ethnographer: From the border of Ngarinyin.
Consultant: Yeh, yeh.
Ethnographer: From Gibb River Country and
Consultant: That’s right yeh,
Ethnographer: Gibb River Country your country or from this side Gibb River.
Consultant 2: Oh past through Kurundji side.
Ethnographer: Kurundji side.
Consultant: Ellenbray way.
Ethnographer: Ellenbray, I see.
Consultant: Look here now—there’s a Gibb River country and this ‘nother boundary here see—well that’s Kurundji country and then this is other place Ellenbray country—that’s how all the group run see.
Ethnographer: Yeh.
(transcripts 30–31)
Languages have boundaries, but these boundaries do not inhibit so much as facilitate communication, functioning in the constitution of a panregional discourse in a way that finds no idiomatic expression in the ethnographers’ register: “All these edges can mostly pull their own language again see.” What’s more, the topography of language does not necessarily align with the topography of country: a stretch of land associated with a particular set of cosmogonic narratives and a distinct array of resources for sustenance and ritual work may be inhabited or visited by speakers of multiple languages:
Ethnographer: B. now—you two from that same country as this man or …
Consultant [A.]: Yeh.
Consulant 2 [B.]: All those crowd now.
Ethnographer: Or you from different country?
Consultant 2: All that lot all one country.
Ethnographer: All one country—yeh—all right well—I think we get a lotta language from your country—we’ll just check here—Now how do you say? … (Elicitation commenced.)
(transcript 31)
In the end, the linguist can only fall back on word-list elicitation.
Elsewhere in the transcript, we see the linguist sounding out speakers, though not in any systematic way, on attitudes toward language within the community. Here again, as in this conversation between D. and his nephew and the ethnographer quoted here and there, which took place a day after the exchange with D. reproduced above, personal history comes to the fore. Note that parenthetical comments in the following transcript represent redactions, places where the person transcribing the tape skipped over a stretch of the conversation.
Ethnographer: When you were little boy where were you? At Forrest River Mission? When you were a little boy?
Consultant [D.]: Yeh, born there.
Ethnographer: Long time ago eh?
Consultant: Very early, long time ago.
Ethnographer: And did you work in the Mission?
Consultant: I work on Mission, yeh.
Ethnographer: Did you go away any other place to work?
Consultant: I went to a station one day.
Ethnographer: And what did you do there?
Consultant: On a stock job.
Ethnographer: What station that one?
Consultant: Rosewood Station.
Ethnographer: Rosewood, where’s that?
Consultant: Oh be a few run from Kununurra to Rosewood—couple morning’s run up long Kununurra town—You been to Kununurra?
Ethnographer: Yeh.
Consultant: Well you start long Kununurra at the breakfast—you get there smoko time—not far run.
Ethnographer: Oh yeh. (Discussion about stock work follows, then …)
Ethnographer: Many people now work on the stations—like— Wumbulgari people—
Consultant: The mission closed—
Ethnographer: Yeh.
Consultant: The cattle and the horse up there—a few cattle there—a few horses.
Ethnographer: What about on the stations any?
Consultant: Oh they wander, wander, wander—people still live there. They stop on still. They mob boys with myself—used to bring cattle overland to Wyndham [i.e., to the slaughterhouse at Wyndham].
Ethnographer: Oh yeh.
Consultant: And every evening the boys bring im and put him into the yard ready—make one of these— take cattle down to the race, down the jetty and the boys come alongside im (and so on) …
Ethnographer: See that old man down there? Who’s that old man?
Consultant: H.
Ethnographer: What language does he talk?
Consultant: Wumbulgaři language—same language as with us.
Ethnographer: Yeh. What about this little boy? What’s your name?
Consultant: R.
Ethnographer: Your son?
Consultant: He’s my brother’s son.—chap with artificial leg over there.
Ethnographer: Oh yeh.
Consultant: Lives in the green houses over there.
Ethnographer: With the artificial leg eh?
Consultant: Yeh, he’s my younger brother.
Ethnographer: R., you speak this Wumbulgari language too eh?
Consultant: He don’t understand language.
Ethnographer: He doesn’t understand?
Consultant: He talk English.
Ethnographer: Don’t you understand that language?
Consultant 2 [R.]: No.
Ethnographer: That’s too bad. The old people talk a language—you don’t understand what they’re saying eh?
Consultant: No.
Ethnographer: That’s too bad eh? They’d better teach you eh?
Consultant: Start teaching that to him to-day or to-morrow.
Ethnographer: What do you think if somebody come here and teach all the kids to write the Wumbulgari language—teach em to write it down—You think that good or not?
Consultant: Good idea.
(transcripts 11–12)
Five days later, the survey team had made its way to Kalumburu. Halfway through the first side of tape 29 we find this conversation with a speaker named P.:
Ethnographer: When you talk Gambera language many old people here speak that language or just a few Gambera people here?
Consultant [P.]: Oh a few old people down there camp.
Ethnographer: Oh yeh—what about young people—they learning that Gambera language or not?
Consultant: I don’t know—they don’t look like they learning any. That’s the thing we worry about. We are worried about these young people, I don’t know where they heading.
Ethnographer: Yeh.
Consultant: They want to try to go in modern ways to civilization—but they never get to that way yet.
(transcript 29)
Later the same day, the linguist pursued the matter further with M.:
Ethnographer: M.—the children here—are they still—they’re learning English in the school—but can they speak Gunin and these languages?
Consultant [M.]: Some of them—only very little tho’—some young ones they understand it—the old people talk—they answer them you see—for the old people they talk in language.
Ethnographer: They talk it to the old people?
Consultant: Yeh—they can understand their grandmothers and grandfathers—they have to talk in language—not to forget their own language see for them old people.
Ethnographer: Yeh—What about like these girls working in the sewing or in the kitchen.
Consultant: Yeh they understand too—they say (kabu, kedji) (kabu) means nothing (kedji) yes.
Ethnographer: Yeh.
Consultant: They say (paranga)—come here.
Ethnographer: So that when they talking to each other they can …
Consultant: No, they talk in English—the young ones you know,—only for the old people.
Ethnographer: Good, yeh. I see—well I’ll see if J.’s around.
(transcript 30)
The next day, October 23, with the time allotted for the survey drawing to a close, the ethnographer was back at Wyndham Reserve. On this second visit, he interviewed E. and N., older members of the community identified in the transcript as pensioners. Here, not for the first time, the linguist is confronted with the fact that even as young people are no longer learning local languages, older speakers are forgetting languages they did once speak. As with D. so with all the speakers interviewed in this survey: the speakers’ biographies are marked by periods spent at a number of missions and stock stations, living and working with people from across the Kimberley and the adjoining part of the Northern Territory. The languages one speaks best do not necessarily correspond to the languages associated with the countries one calls one’s own by virtue of totemic affiliation (Evans Reference Evans and Brenzinger2007). Nor are they necessarily the languages one learned as a child. Again, parenthetical comments are reproduced below as they appear in the transcript.
Ethnographer: Some Wuladja now or you bin forget that.
Consultant [E.]: Wuladja
Ethnographer: Yeh.
Consultant: No, I can talk Wuladja.
Ethnographer: You can talk Wunambal?
Consultant: Wunambal all right, I talk to you a few words—I
Ethnographer: What about Yeidji—Wumbulgari?
Consultant: Yeh, yeh I talk that.
Ethnographer: Wumbulgari.
Consultant: Yeh, Wumbulgari—that way Wumbulgari.
Ethnographer: Yeh, Where you come from yourself?
Consultant: Here—this my country.
Ethnographer: Wyndham?
Consultant: Wyndham.
Ethnographer: Were you at Forrest River?
Consultant: I was born there. (I don’t know if this refers to Wyndham or Forrest River.)
Ethnographer: You were born there?
Consultant: My mother they lose there, and body there too, for long time.
Ethnographer: Where—whereabouts.
Consultant: Pump you know where this pumping station. We callim pumping station.
Ethnographer: Pumping Station.
Consultant: Where water pumped.
Ethnographer: Yeh and you from this King River tribe eh?
Consultant: Yeh but I can’t—talk Wuladja.
Ethnographer: Yeh.
Consultant: I go way from there now when my little child I bin grow up long Mission Forrest River—
Ethnographer: Oh yeh was that King River people? King River tribe where they talk Wuladja?
Consultant: Wuladja—yes.
Ethnographer: Oh yeh
Consultant: Mixed (karεga) Yeh mixed—Wuladja, all that I talk Wuladja.
Ethnographer: (kařεga)
Consultant: King River people yeh.
Ethnographer: What’s this (kařεga)?
Consultant: (kaři)
Ethnographer: What’s that mean like—same like Wuladja?
(interruption) Your wife?
Consultant: Wife?
Ethnographer: What language was she when she was living?
Consultant: He belong whatchamicallem language—longa dis way.
Ethnographer: What was that one—what language?
Consultant 2 [N.]: He can’t talk now he forget all dat. Brother got all that language from this way. What they call people from that way—from this way?
Consultant: Miriwung.
Ethnographer: Miriwung.
Consultant: Like and you know
Ethnographer: Kununurra
Consultant: Yeh, yeh.
Consultant 2: (kareiyan)
Ethnographer: Yeh.
Consultant: He’s the sister now.
Ethnographer: Oh yeh.
Consultant: Long Kununurra—Wadi.
Ethnographer: Wadi.
Consultant: Mm.
Ethnographer: That’s your—I seen him—I got Wadi—I got on this tape recorder. I get im give me language.
Consultant: Oh yeh.
Ethnographer: He talk to me and tell me all the words.
Consultant: Yeh.
Consultant 2: But we no more gottim Wuladja language—belong dis way—King River—he talk Wuladja. Me—I belong this country. This my country—my mother born. Him bin loose there. I bin born there too, me.
Ethnographer: Mm.
Consultant: Well, I forget now. I go way, leave when me little time—me bin go back Mission.
Ethnographer: Yeh.
Consultant: Aruwadi side (aruwadi = south). Supposed to getim Wuladja, before. He forget all that now. No language belong this country.
Ethnographer: Where I find some men from King River Side.
Consultant: Eh?
Ethnographer: Where they now Fork Creek or?
Consultant: Finish.
Ethnographer: All finished eh?
Consultant: All finish.
Ethnographer: King River people.
Consultant: All die yep.
A few minutes later in the conversation, the linguist returns to Yeidji, which Earnest might have said he spoke (“What about Yeidji—Wumbulgari?” —“Yeh, yeh I talk that”).
Ethnographer: Yeh—well you can tell me some Yeidji talk now eh?
Consultant: Eh.
Ethnographer: Yeidji—you understand that or is it Wunambal?
Consultant: Wunambal—yeh I talk a little that Wunambal.
Ethnographer: I want to get somebody to tell me some Yeidji.
Consultant: R.
Ethnographer: R.
Consultant: Yeh.
Ethnographer: Maybe I’d better go see him eh?
Consultant: Yeh.
Consultant 2: He’s the proper Wuladja Number I. He belong that country.
Ethnographer: Oh yeh.
Consultant: (kular) (= west)
(transcripts 31–33, 34)
Of the language or languages from the country to the west, N. assures the linguist, “You get it properly once you get R. here.”
Later in the day, the linguist does manage to catch up with R. Their conversation is recorded on tape 31 (AIATSIS A2190) and was transcribed this way:
Ethnographer: They tell me you understand all about the people—different tribes and everything round here, like some people from Kurundji side, that’s Wuladjangari isn’t it?
Consultant [R.]: What’s that?
Ethnographer: Wuladjangari.
Consultant: Yeh.
Ethnographer: That from Kurundji way?
Consultant: Yeh.
Ethnographer: What about from King River?
Consultant: These all the same.
Ethnographer: Wuladjangari. And what about Forrest River?
Consultant: No, they Wunambal.
Ethnographer: Wunambal—
Consultant: Yeh.
Ethnographer: When you say Wunambal is that the same as Yeidji?—Yeidji, that same thing as Wunambal?
Consultant: What’s that?
Ethnographer: Yeidji.
Consultant: Yeidji—that’s Wunambal.
Ethnographer: What about Andidja?
Consultant: Same language isn’t it? Wunambal.
Ethnographer: And Andidja, and Gingara people—do you know what that is—
Consultant: No answer.
Ethnographer: Or is that the same King River people?
Consultant: King River, yeh.
Ethnographer: Yeh—when long time ago they—this Wuladja people they go to—Forrest River or—which?
Consultant: Wuladja people.
Ethnographer: Which?
Consultant: Wuladja yeh.
Ethnographer: Same as Forrest River eh? Yeh—what about King River people?
Consultant: King River people they walk up and down you know.
Ethnographer: They went in then came back?
Consultant: When they have big meeting they go to——some go Kurundji or some place.
Ethnographer: Yeh, I see—yeh—yeh. And what tribe do you belong to?
Consultant: (kular)
Ethnographer: (kular) and what Country’s that?
Consultant: Way down Gibb River.
Ethnographer: Gibb River.
Consultant: Yeh.
Ethnographer: Oh yeh and what language talk?
Consultant: Wuladjangari.
Ethnographer: Wuladjangari—same language?
Consultant: Yeh.
Ethnographer: Same language as Kurundji?
Consultant: Yeh.
Consultant 2: Old man and old woman out there.Footnote 5
Ethnographer: Well that’s all, I just wanted to ask you a little bit about the tribes here see—if you got time could tell me some Wuladjangari—but when you got time—You working now eh?—Going back to work?
Consultant: I working now.
Ethnographer: What time do you knock off tonight?
Consultant: Oh about 5.
Ethnographer: Can I come back and see you then?
Consultant: Yeh.
Ethnographer: Cause I gotta go back to Darwin to-morrow.
(transcripts 37–38)
At last we see and hear what might represent, for the linguist, an ideal informant. R.’s answers are unequivocal. He sticks to the categories introduced by the linguist’s questions: language, tribe, country. This is the speech of a busy man, someone who needs to get back to work. Like the linguist, R. is on a schedule.
As we read these transcripts in series, a narrative arc emerges in which the linguists’ quest for informants who will speak their language mirrors one of the central narrative tropes of contemporary archive-oriented endangered language documentation: the tracking down and authentication of an elusive Last Speaker. Just as the linguist comes to face to face with the individual who can reel off basic vocabulary in the western speech forms as they presumably were spoken prior to colonization, it is time to leave. Eight years later, in AIAS’s Revised Linguistic Field Manual for Australia, Sutton and Walsh would caution that, in “distinctively Aboriginal communities, where traditional life is often not very far in the past, brief ‘surveys’ which cannot result in the establishment of meaningful relations are to be discouraged” (Sutton and Walsh Reference Sutton and Walsh1979). The gap in the expectations ethnographer and speaker bring to the work of elicitation is partly a gap in life chances. The answer to the question of who can speak in a particular named code on demand is conditioned by mundane facts of upbringing, debility, and mortality that, as a rule, are more salient for speakers of languages under pressure than for linguists (Dobrin and Berson Reference Dobrin, Berson, Austin and Sallabank2011). But often, as here, the gap in expectations between speaker and linguist is also conditioned by differences in how the two participants in the encounter understand linguistic variation to unfold in space and time. A linguistics, or a state, that operates with fixed rubrics for locating that which, to return to Hillier, “lacks location” save in its recurring but ephemeral fragmentary realizations, risks mistaking the topologies inscribed in a particular set of artifacts—word lists, say—for those of the communities those artifacts are made to stand in for in synoptic accounts of communicative behavior.
