Introduction

This article is occasioned by two years of fieldwork among Aboriginal people of Central Australia to 'assess the impact of commercial TV on remote Aboriginal communities'. The concern motivating the award of the fellowship to undertake research was promoted by the Australian Government's plans to launch a direct broadcasting satellite (AUSSAT) at the end of 1985 which would make television, but possibly telephony and other services, available to isolated communities throughout the vast interior of the continent. For those Aboriginal people who had maintained a continuous cultural tradition for over 50,000 years, isolation from the mass media has been a factor in their resistance to acculturation. What might the effects of a daily television schedule be?

46

In fact, the question is asked before the fact. I intentionally chose to work with a community far distant from any television receiver. The problem was to determine if a useful course of action could be developed here, rather than simply to observe the possible deterioration of a society whose choices had already been made for them. The project has combined development, policy research and a study of the traditional oral communications system. In this article, I report not my findings, but on methodological, even epistemological questions that have arisen in the nearly completed three years of study: how can I come to know and answer the urgent questions that face these people?

'What do Aboriginal people want from the satellite?'

Background

'Remote' is defined by Telecom (the Australian Government telephone service) as being beyond a certain number of kilometres from any automatic telephone exchange. Remote in Aboriginal Australia means essentially being unable to make many or any transportation/ communication trade-offs. Yuendumu settlement, 300km up the mostly dirt Tanami highway northwest of Alice Springs, is such a community. Although it is perhaps the fifth largest concentration of people in the Northern Territory, a radio-telephone is the only 'public' telecommunications link to the supply centres and capital cities for the 1,000 residents. It is minimally reliable for about an hour, five mornings a week. If the electricity is on and the post mistress is in town, we can expect to get out three to five calls a day. In-calls cannot expect to connect to their intended receiver unless

47

that person happens to be near one of the 20 CB units mostly in European venues. Yet European service workers comprise less than ten per cent of this community which is otherwise Aboriginal in ownership and intent. Peter Toyne and I have described this situation in a longer paper which suggests the staggering costs of maintaining such a community where supplies and services for nearly one thou sand people cannot be co-ordinated reliably except by face-to-face communication (Michaels and Toyne, 1983). It might be expected, therefore, that automatic telephone services would be the top tele communications priority identified by all surveys. That it is not will be explained below.

Yuendumu appears to be precisely the sort of community for which AUSSAT was designed. It is typical, if larger, than most remote settlements in that its economic marginality is enforced by its great distance from any telecommunication trunk lines and its social marginality is enforced by its isolation from the electronic mass media. In principle, as in economic fact, the great advantage of satellites is their ability to access such locations as easily and economically as any other, rendering the costs of information equivalent to more densely populated and more geographically accessible areas.

Yuendumu is also typical of such unserviced settlements in that it is predominantly Aboriginal in population; it seems hardly an accident that the majority of people in Australia without automatic tele phone or access to at least one television channel are Aboriginal, al though they comprise something less than two per cent of the population. In arguing for the benefits of satellite in terms of value to remote regions, the Australian government presumably would have computed these figures, and made special provision for assessing Aboriginal opinions, as well as needs. This was not entirely the case.

What research was performed initially was mostly limited to studies commissioned by Telecom. The public, published, versions of these studies seem to prove rather conclusively that for 200 years, Aboriginal Australians have wanted nothing so much as a Digital Radio Concentrator System. (IMG, 1980: Bliss and Wild, 1984). Likewise, casual surveys (the only ones) made by the Northern Territory Government (Ridge, 1983, P.C.) indicated that 'top end' communities wanted TV more than anything, even telephones. It does not seem incidental that, in one sense or another, these are precisely the services that the investigators were 'selling'. In the last few months, by contrast, there has been a sudden host of short-term inquiries, investigations and research projects aimed directly or indirectly at assessing Aboriginal needs and projecting impacts of new media on remote Australia. Coming so late in the game, and tying up independent

48

Aboriginal initiatives, some communities and organisations began to wonder if these studies may not, in fact, be more of a tactic to stall planning until it was too late. But here we will attend only to the question of what kinds of studies might produce useful answers, and which are most likely simply to reify the preconceptions of the researcher.

Methodologies

The difficulty which must be admitted is that Aboriginal culture was not designed to serve the interests of either researchers or the government. From a European perspective, the languages (and the cognitive systems assumed to underlie them) are damned difficult, based on distinctions and categories unfamiliar to us. And the political and social structures have proved especially resistant to translation into terms of Western law and political representation. So much of the European approach to unfamiliar peoples can be glossed as, 'Take me to your leader'. If the indigenous political structure is not hierarchical and does not identify general purpose representatives, the usual alternative seems to be to attempt to establish such structures and designate leaders. While this point may seem essentially an administrative problem, it is also a methodological issue.

The statistical protocols most familiar in communications research assume a normal distribution of characteristics across a population. An individual's response to a survey or interview question will be assumed in some (corrected, statistical) sense to 'speak for' the population as a whole. In this sense, most sampling models in communications studies assume a (democratic, post-industrial) social model as well. 'Critical' international communications researchers (eg., Katz, Schiller, Tunstall, McAnany, Nordenstreng, Jouet) dub such statistical methods as characteristic of 'administrative research', and claim such research has administration rather than scholarship as its objective: to turn societies into these forms. This imposition of the normal curve on inappropriate social systems would be a criticism applied especially well to communications research in Aboriginal society. Here the demography of opinions no less than of persons or groups is intricate in ways that are exemplified by land rights cases where attempts to determine 'tribes and tribal boundaries' have revealed a situation extraordinarily difficult to translate into common law (Maddock, 1983). The 'boundary' of an Aboriginal population may be impossible to establish and thus the generalisability of any individual datum can be utterly problematic.

49

This can be illustrated easily by examples from questionnaire surveys administered by Europeans to traditional Aboriginal people. Typical demographics require the respondent to be the head of a household, or in a stated relationship to such a 'head'. But what constitutes the 'head' of an 'eight-subsection kinship system grouping with patrilineal emphasis but matrilineal rights through mother's brother', as described for the Warlpiri. What indeed is a 'household' in a mobile, foraging society in which residence may express itself in dozens of ways depending on demands of ceremonial life, age grade, rainfall, season and other variables. Anthropologists do not have easy answers to such questions after a century of Aboriginal fieldwork. But social psychologists have too often ignored the question, settling for anyone who offers him/herself as 'head of household' (Noble, 1984) .

There have been a number of studies proposed recently by academics who wanted to translate their very real concern about the plight of Aborigines facing the introduction of television into useful research. They could not understand why they met such extreme resistance from Aboriginal organisations and were not funded. In one case, a researcher took his case to the press, claiming an important scientific possibility had been irrevocably lost. The facts bear some examination.

That proposal made the obvious choice of protocol as a pre/post test design, where the 'test' is the onset of transmission. The elicitation and measurement procedures were to be self-reported evaluations elicited through questionnaire or interview. To achieve statistical power, several hundred questions were proposed for several hundred remote Aboriginal respondents. These would then, presumably, be collapsed into categories of response capable of testing particular propositions and producing 'findings'.

If I could ask three such questions of three respondents after my three years of fieldwork and could feel secure that I understood what was said to me, I would consider it an extraordinary achievement. Self report has generally proved to be unreliable in American communications research. In cross-cultural settings and bilingual inter views, the problem is exacerbated.

Generally, subordinated classes do a better job figuring out the language and rules of the dominating class than the other way around. The stakes are of course much higher for the subordinated. Aborigines strike me as profoundly interested in understanding (as distinct from becoming) Europeans, an issue which has a lot to do, incidentally,

50

with the interest in imported media. Researchers, and other alien presences in remote settlements are tolerated, I suspect, partly be cause of what can be learned from such people. But Aboriginal organisations and communities who wish to manage their own affairs now demand that research priorities reflect their needs for information and can provide a basis for development projects. They also demand a respect of Aboriginal people and culture which they find lacking in some kinds of research.

Anthropology is familiar with such situations, and provides a methodological contrast to these research procedures. Anthropological models allow us to test propositions by evaluating the exceptions, to describe a field in terms of its margins, to evidence rules by observing their violation. Naturalistic observation and by participation which seeks to identify, describe and operate within categories and issues emergent in the cultural setting (rather than imposed by experimental and measurement requirements of a rigid empiricism) have much to offer cross-cultural communications research. Let me illustrate the contrast by example.

Much of my research consists of teaching Aboriginal people how to use media technology to recreate and communicate their own world. The technical procedure is simple; it takes about a half an hour to show anybody how a VHS camera/recorder outfit works. Any further instruction is limited to response to direct questions about how to solve particular problems that arise in production. I had imagined that in this way, we might create an audio-visual lingua franca in which we could communicate about the medium under consideration. I might be able to compare Aboriginal production, products and viewing with their European equivalents and develop a model for describing the differences of need and usage on which to base a solution to the problems which face us.

The discovery of what Aboriginal people want from the media emerge only out of the dialogue that takes place around the production, post-production, viewing and reviewing of these tapes as they insert themselves into the community's life, or fail to. And these inter actions lead tangentially into related areas of Warlpiri life which impinge on the question of communication, graphic production and interpretation, and traditional law. Thus, I spend evenings talking with the younger people about Hollywood videos they are watching in the camps. I spend weeks painting canvasses with senior men of the designs associated with the 'dreamings' for their country, and take them on trips to these places as well. I spend days with the video makers poring over a single tape sequence in the editing room, translating the words and discovering what the event boundaries are

51

which allow for an edit to be made. And while my command of the

language is poor, I have worked with expert linguists, both Aboriginal and European, to come to some understanding of its unfamiliar grammatical structures and lexical categories. In all of this, I can barely claim to understand but a small percentage of what is told me. But, I can document in some detail the circumstances surrounding our discourse and advance some reasons why I was told particular things- These are experiences which have led me to be highly sceptical of large samples and unexamined cross-cultural dialogue.

It is nearly two years since we began this collaboration at Yuendumu. And it is difficult to claim that this lingua franca has been established. Mostly, I discover that Warlpiri Aborigines have very different interests in media than those described for people elsewhere. Production styles differ in ways that are best explained by reference to traditional social organization, as is viewing (Michaels and Kelly, 1984). The video-taped product itself can only be understood by reference to these contexts. Aboriginal people that I work with now express their interests in media by their activities and their plans. These findings would not be captured by survey research, or individual opinions, as they are socially emergent facts, expressed not as individual observations but by collective community action.

Of all the research questions I formulated to bring with me into the field, the most provocative one turns out to be, 'What conditions are necessary for Aboriginal people to develop and operate their own communications facilities?' In pursuing this question, I am continually reminded of Herbert Schiller's observation that TV is not the sole or initial agent of imperialism or cultural homogenisation. It is always introduced after a host of other social and economic impositions on traditional and third-world cultures designed explicitly to bring such enclaves into a dependent relationship to the world market and the dominant society. The imposition of European forms on to Aboriginal language and custom does not, therefore, require electronic media as a cause. Indeed, by focussing too narrowly on media, or television, we risk impoverishing a broader social criticism necessary to explain the contemporary Aboriginal experience. Had films been easily transported here, or theatre, or had literacy been easily accomplished, these mass media would have carried the ideological messages which are of such concern. In fact, the experience of European life has already come to Yuendumu, albeit in skewed neo-colonial forms. The effects on the population in terms of deteriorated health, cultural and social disorganisation and material impoverishment are clearly visible. And these effects were results of previous policy decisions, it should be recalled. So research will inform the current government's choices of telecommunications services.

52

What will in fact happen at Yuendumu, or throughout Aboriginal Australia, when the satellite is launched is a political question, which regards the media, but does not depend on it. Research which prematurely identifies television as an isolated variable, risks obscuring the fact that television can be many things, and carry many contents. Which contents, through what social and technological systems, will be a political and economic choice of government. And the 'effects' of the medium will be the consequences of these choices, which instruct us as much about the policy of the state as the conditions of any medium. If development and the research underpinning it are to have any more favourable consequences, we require far more precision in our philosophy, if not our methods. It does not serve to run such people through our research mill any more than we can humanely shower them with the residue of our mass media. If we single out indigenous cultures for special concern, as the Australian Government has in this case, a re-evaluation of our research tools may also be required.

Examples

Three brief examples should serve to illustrate the above remarks:

Case No 1

Several telecommunications options had been discussed during a survey of current services at Yuendumu. Before presenting the finding to appropriate agencies in government, it was decided to discuss them first with the Yuendumu community council. The findings revealed that the economy of the community was suffering inordinately from the lack of telephone services. Minister Duffy had recently announced that television and radio, but not telephones, would be made available through AUSSAT. In a meeting with the council, I described various communications services, and the councillors reviewed telephony at great length, arguing that it would enhance their traditional status, while television was seen as weakening it. I was familiar with previous surveys which had rank-ordered these options, and that some had ranked television as a first priority. Also, I had been criticized by urban Aborigines for advancing communications over 'necessities' such as health and utilities. I asked the council (mostly older, more traditional men) which they wanted more, Toyotas or telephones.

My questions were translated by the council president and the responses translated back to me (although I had some limited Warlpiri,

53

and most of the councillors had workable English). The proceedings were audio-taped.

The council president gave the reply that they wanted telephones most- This surprised me, as I knew that Toyotas were critical to the maintenance of outstations, hunting, and of ceremonies, and they were a very high priority among these same individuals, most of whom I had talked to about these things before. I was excitedly anticipating the publication of this finding.

Reviewing the tape with a Warlpiri translator the next day, I asked just what they had said at this point. My Warlpiri colleague said, 'They say they want a telephone and a Toyota'. But I pointed out that I had asked which one. 'Oh ... telephone. That's what they want'. And that's what I wrote. The next day, I was explaining this interesting development to David Nash, a dedicated linguist specialising in Warlpiri. David asked to hear the tape, and then he explained it to me.

Indeed, the men had said, 'We want both ... ' This was because Warlpiri, like other Aboriginal languages, does not have an 'either/or' comparative construction. It is very difficult to ask a rank-order or preferential question in Warlpiri. The fact that the men had combined both objects in a single reply, indicated an equivalence. That one object preceded the other might be construed as preference (which perhaps my original translator had done) but this was more dubious. What was reported to me was probably an attempt to match my expectations without engaging in too complicated a linguistic discussion.

Case No 2

In testimony before the Australian Broadcasting Tribunal regard ing issues which might arise in relation to new media delivery systems, held at the Kintore Aboriginal Homeland near the West Australian/Northern Territory border, old and young Warlpiri men spoke eloquently of their concerns (Yuendumu women could not attend for reasons related to local ceremonies). Elders spoke through interpreters, the younger interpreted for themselves. The other groups decided to evidence their language facility by speaking only their language and being translated by an interpreter. But this interpreter was female, and of a different language group; there were many things the men said that she would not interpret for traditional reasons.

54

Throughout the hearing, two concerns kept recurring: 'culture' and 'old people'. In English, these terms have a particular range of denotations and connotations. In fact, culture, per sc, is an English term, associated with the anthropologists who are familiar people to the Aborigines. But the Warlpiri term being used, jukurrpa, (and the . equivalents in the language used by the other groups present) is probably more closely equivalent to 'law'. There is a considerable difference in European thought between a violation of culture, and a violation of law. Likewise, 'old people' are not necessarily chronologically old; it would be more accurate to call them 'wise', or, 'senior'. The concern was not, as it came out in English, for the quaint styles and sensibilities of an aged, dying generation. Rather, the concern was for the central tribal law as maintained by senior authorities with wisdom and respect. It may be a difference of emphasis, but it seems to be more than a trivial one.

Case No 3

Early in my work, I became familiar with the film 'Two Laws' and the film-makers' rationale for using a wide-angle lens, and limiting themselves to group shots at some distance. This was partly to respect traditional avoidance laws and regulations which constrain direct eye contact in interpersonal communication. The wide angle technique translated this into a filmic convention and assured audience members that they would not be in the embarrassing position of look ing directly into the eyes of an inappropriate relative on screen.

In the first video-tapes shot locally by Warlpiri Aborigines, this same distancing convention was used. I was ready to see this as confirmation of an hypothesis about the medium for traditional practitioners. Only later, in accompanying people while shooting, did the curious dependency on the cigarette jack in the Toyota become apparent. Because the video-makers hadn't been adequately instructed in the use of the portable batteries, they remained 'chained' to the truck, at some distance from subjects and events. Once the battery was used, close-ups became common and my naive hypothesis had to be rejected (to be replaced by a more complicated one regarding particular prohibitions later).

Discussions of Cases

These three examples are excerpted from my field notes, and therefore are not necessarily the best examples of the problem in that

55

I do not usually seek the 'objective measures' of phenomena or statistical correlations which I am criticizing. Even so, it should be evident that in all three cases, the first two linguistic and the last semiotic, preconceptions got in the way of any valid understanding of events. Only in review, accorded by an on-going association with the community, were these flaws revealed, and corrected. The third example underscores a problem of hypothesis testing. An hypothesis is also a bias; where insufficient attention or scope is available to the investigator, false proofs may be readily available.

The diversity of opinions and perspectives characteristic of Aboriginal life means that a researcher can usually find at least some informants to support any given proposition. Worse, the reason why Aborigines, no less than anybody, offer a particular opinion, or co-operate at all with a researcher, may be complex and have nothing to do with the explicit objectives of research. Fieldworkers represent political and material resources. Individuals may wish to provide or withhold the kinds of answers perceived to be sought on the basis of considerations that may remain utterly masked to the researcher. Or, the re searcher may wish to disguise the situation from himself. A psychologist studying kinesic behaviour sought 'natural' actions, but needed to record them. He created a camera with an angled mirror which made it appear that he was shooting 90 degrees away from his real subject. I submit that this camera was only effective at fooling the psychologist; the Warlpiri had it figured out in no time (the camera has since been abandoned, and more sophisticated and sympathetic techniques developed by the psychologist).

With the notable exception of the contribution of Elspeth Young (who has been working in Aboriginal communities for some years) to the IMG reports (1980) nearly all the research on telecommunications in remote Aboriginal Australia has been based on rapid site surveys in which researchers spend at most a few days in a given community. Typically, they speak to a few individuals, designated as 'leaders', and, if this proves too much of an effort, to European administrators or 'experts' such as myself who might be around. There is little or no opportunity to cross-check information within a community, and so there is little chance to discover when mistakes of the sort described here, or other errors, are made. Thus the quality of data on which planning and policy are developed is highly suspect. In such a situation, the statistical collapsing of such data mostly serves to create enough distance between data and findings to assure such errors will be undetectable.

I realise that in discussing these cases, I have collapsed what many academics would distinguish as 'pure/theoretical' research with

56

'applied/policy' research. In the current context, in fact, in any research on a living population, I believe the distinction is false. My own work proceeds from what I take to be profound questions about the contrast between oral and electronic societies in order to inform a theory about what it means to be human, and to communicate that. But anything that I write, or publish, is apt to be used by policy makers or by technicians, or government, to affect the lives of the people about whom I write. Thus, though my emphasis may be different in different articles, I would argue that in Aboriginal Australia, at least, no research is 'pure', only unaccountable.

The only distinction between approaches worth noting is between what might be called an investigative vs. a marketing approach. In the first case, one wants to discover something, in the second, one wants to prove something. Ideally, 'scientific' research is associated with investigation‹an intellectual curiosity regarding a phenomenon, an issue or an event, or perhaps some relationship between these. The answer, even the question, may not be known beforehand and both question and answers evolve in relationship to field circumstances: what is knowable, what the participants in the research want, what communication is possible. But 'marketing' research has its question and/or answer determined: will these people buy tele phones, or, having decided that telephones will be available, can we assemble evidence to show that people really want them, or that it will be good for them to have this? Or, more typically, what will be the most efficient, cost-effective means of introduction?

The experiences reported here suggest that in cross-cultural re search, the importation of rigid 'methodologies', especially those where non-English speakers are forced to respond in English language/culture categories (either through elicitation or coding) to questions posed by the researcher, even with the most 'scientific' intentions, converts the procedure to marketing research. The statistical operations performed on these answers and the attention given to data manipulation rather than elicitation in the typical academic or policy presentations serve to mask the most fundamental criteria for validity: who asked what of whom, in what language, under what circumstances?

Summation and Conclusions

It would be useful to provide methodological solutions to the problems outlined in this article. And, of course, the present author modestly implies that 'anthropological approaches' are exempted from these problems. In fact, anthropology has earned much well

57

deserved criticism from its insensitivities; perhaps we have learned something from these. There is, admittedly, little chance that crucial social issues can often be approached by long-term anthropological fieldwork, or that they necessarily should be. Policy research usually requires more rapid answers to immediate questions, and protocols which can be more widely employed than the very personal methodologies of ethnographic fieldwork. Even so, if particular research approaches demonstrably result in certain types of error, those errors may be identified to some advantage.

Three categories of error have been identified here as the source of invalid of misleading results in cross-cultural communications re search. Without determining how they can be corrected efficiently in other research approaches, they are offered as caveats for future studies.

1. Grammar/Lexicon Errors: From the earliest research in non Western linguistics, it has been apparent that diverse languages have diverse lexical and grammatical systems which are associated with quite different cognitive systems and conceptual categories. Questions not merely asked, but also conceived, in English or other European languages may be conceptually unfamiliar and inappropriate to speakers of other languages. While most languages can be made to express any concept, so that universal translation is superficially possible, a respondent to the question may have great difficulty with some important element of it, even in translation. In the quite common case where indigenous speakers have some facility in the dominant language (eg., English, Spanish, French) the facility may be misleading. Creole and pidgin forms carry first language concepts into, eg., English. Even apparent fluency may mask persistent conceptual forms. Very careful work with linguists, indigenous and academic, may be required in constructing questions and evaluating re plies. Some questions may simply be unaskable, or prove unanswerable.

2. Demographic Errors: Survey/interview sampling and many or most statistical operations are based on certain social-structural assumptions about the organization of society. Concepts such as 'individual', 'family', 'community/population', (indeed, 'normal' and 'deviant', are largely unexamined social concepts. This can become quite clear when seeking the 'head' of an Aboriginal household, or finding the population of a community was 1,200 last week, but is 300 this week (and 150 of these people were not here the week before). The most important male in a child's life might be the mother's brother (quite typical) or someone even more 'distant' in terms of European kinship For female children, another relative may be most important.

58

Indeed, during childhood, different primary relationships and different residential patterns may obtain. Speaking rights are highly regulated in oral societies; only certain people have the right to speak of or for certain things. When identifying representatives for a sample, or collapsing individual replies to account for group behaviour, the appropriateness of the statistic to the observed social order is necessary. Impressive degrees of significance may in fact be invalid and misleading unless they are suitable measures for a given social system.

3. Interpersonal Error Data are always elicited in socially situated discourse and this situation is apt to be interpreted differently by per sons in the researcher role than the respondent role. It is especially helpful for researchers to develop some understanding of who the respondent thinks the researcher is, why they are asking questions and what the profit/outcome will be. Simple devices, such as monetary payment, do not answer these questions (unless one understands thoroughly the micro-economy of the community and the respondent's place in it). Particularly in linguistically or socially difficult or unequal settings, people being interviewed are likely to provide all sorts of answers for all sorts of reasons. I am most familiar with the answers offered when people are trying to be polite and provide what they think the researcher wants to hear. Conversely, the researcher may be seen as a conduit to send particular messages back to the church, the government, the school system, etc. The possibilities are legion. Without developing some model of the research interaction, the researcher is easily fooled, intentionally or unintentionally. With out describing this model in the report of the study, it becomes very difficult for other researchers to evaluate, reinterpret or test the results.

I believe some or all of these errors are contained in the bulk of the social research on which decisions have been made regarding Australian policy on Aboriginal telecommunications. In fact, these errors may be detectable in far more than this application. But here, I am particularly concerned because I can see the likely results of such re search in the very near future. They do not appear promising in terms of what I have come to understand to be Aboriginal needs and desires for new communications systems. The novelty in this case may be only that there is a government in Australia whose expressed intent is to serve Aboriginal interests, as opposed to ignoring them, as has been the case in too many other countries. And it is likely that a good deal of money will be spent on the matter. The government has sponsored a number of major research projects, but has been most comfortable with questions and answers framed in discrete,

59

quantitative terms, which is to be expected. Under these circumstances, I submit that research scholars might take particular responsibility for re-evaluating their tools and determining what sorts of approaches represent adequate scholarship.

Bliss, T. and Wild, S. (1984). 'Introduction of telephone services to remote Aboriginal and Islander communities', ANZAAS Workshop on the Introduction of New Media to Remote Aboriginal Communities, AIAS, Canberra.

IMG, (1980) A Study of Remote Area Telecommunications in the Northern Territory, (summarized in 'Remote Area Telecommunications Study: National Report', Telecom Australia).

Maddock, K. (1983) Your Land is our Land, University of Queensland Press, Brisbane.

Michaels, E., and Kelly, F. (1984) 'The social organisation of an Aboriginal video workplace', in Australian Aboriginal Studies, No. 4 (forthcoming).

Michaels, E. and Toyne, P. (1983) 'Out of touch: the costs of not having a telephone', in Working Papers in Aboriginal Telecommunications, Australian Institute of Aboriginal Studies, Canberra.

Noble, G. (1984) 'Influence of satellite television on the cultural values of Aborigines.' Application to the Australian Research Grants Scheme; preliminary findings presented at the Australian Communications Association Meetings, Perth, July, 1984.