The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways.

Bird, Douglas W. ; Bliege Bird, Rebecca L.

'The scientific principle produces abstractions, the anti-scientific principle destroys them on the grounds that such and such has not been considered. The scientific principle establishes strict concepts, the anti-scientific principle makes them ambiguous on the pretext of thus revealing their true variety.'

ALEXANDER ZINOVIEV (in Boyer 1995: 175)

Anthropologists will always struggle to understand diversity - in fact, this should be our ultimate purpose. But how do we begin? Postmodernist critique has left us more willing to recognize the political and historical context of both ethnographers and the people they study, while at the same time cultivating the anti-science notion that such historical and political contingencies make our observations, at best, indelibly infected with our own biases, and at worst, useless as data. Obviously, rejecting the role of observational data-gathering in anthropology will spell the death of the discipline and transform us into 21st-century versions of armchair Victorian speculators. If we accept the role of ethnographic observation in the science of doing anthropology, we are then faced with methodological dilemmas about how we make observations, what kinds of observations we make, and how we use and interpret those observations to serve the goal of understanding human diversity. The explicit aim of both Kent's and Kelly's volumes is to provide a better understanding of behavioural differences within and between populations of contemporary foragers. In our view, the degree to which this goal is served is a function of how the authors utilize ethnographic description. Where questions are clear and descriptions are aimed at testing the predictions of hypotheses to explain variation, the authors provide an important contribution to our understanding of contemporary hunter-gatherer societies. We will first focus on the different approaches illustrated in Susan Kent's volume and then turn our attention to Robert Kelly's monograph.

One of Kent's goals in compiling Cultural diversity among twentieth-century foragers is to document variability within and between African foraging groups. She and the contributing authors accomplish that goal splendidly, providing us with detailed descriptions of variability in forager lifeways which give greater depth to our general ideas about forager subsistence and socio-political organization. Given that Africa has been the geographic centre for many of the most influential ethnographic studies of foragers and debates about their significance, we could be forgiven for accepting Kent's descriptive aim and chalk up the volume as another nicely detailed empirical contribution to the literature.

In her preface, however, Kent makes the case that detailing the diversity of cultural behaviours among groups of Basarwa, Hadza, Okiek and west African foragers is the key to understanding the cause and significance of variability in their lifeways (pp. 3-6). She suggests that because all aspects of behavioural variability ultimately condition each other, specifics can only be understood in relation to the whole and have no particular meaning beyond their context. Description thus serves as both a source of interpretation and as a cautionary tale for those who might nominate any single group as 'typical' of hunter-gatherers. While Kent claims that this is not 'a call to return to the Boasian school of providing detailed descriptive information devoid of theory and interpretation' (p. 4), she indicates that observations of the range of variability in forager culture are the sources for behavioural explanations. Cultural diversity, for Kent, is the result of cognitive and cultural flexibility and mobility. Foragers are inherently more diverse than others because they are more flexible and mobile. She argues that recent approaches to understanding and explaining the variability in subsistence strategies among foragers have contributed to the homogenization of all foragers by limiting their focus toward 'economics or material culture' and labelling all other aspects of culture as epiphenomenal. She hopes that this book will help to redress the issue (pp. 6-9).

Kent's primary purpose in compiling the volume, then, is to detail (and interpret from all possible theoretical perspectives) differences in African foragers as a means of isolating similarities (such as flexibility and mobility) which can in turn serve as a basis for constructing general models to better understand all foragers, past or present (p. 2). The description will serve as its own explanation because the specifics are meaningless outside their context. For us, this is problematic.

While in principle Kent's goals seem reasonable, if our aim is to understand the 'why's of human behaviour, we have serious reservations about the actual utility of her approach. Why should aspects of behaviour that foragers have in common be more important for general models than their differences? Those contributors to the volume that seek explanation in terms of the variability they describe may contribute to a descriptive database of human behaviour, but their explanations are post-hoc and must be extracted from a mire of empirical complexity. For example, when Barnard & Widlok conclude their comparison of diversity in socio-spatial organization between the Hai//om and Nharo with the statement that the causes of diversity 'can be found in a complex of environmental, economic, political and cultural factors' (p. 106), they provide us with no hypotheses whose predictions might sort out which of these environmental, economic, political and cultural factors constrains and influences which aspect of behaviour. Likewise, Silberbauer concludes that while our task is ultimately to explain variability (p. 47), endogenous factors and external influences shape culturally specific learning among Kalahari Basarwa groups in such complex ways that appropriate generalizations can only be generated with more accurate and precise descriptions of variability, with special consideration of both emic and etic assumptions about the logic of behaviour.

The problem with such approaches is that any description is inherently synthetic, subject to tremendous bias, and never complete. Silberbauer suggests that the full range of cultural variability among the G/wi may only be manifest over a period of two or more generations (p. 30). We might suggest that two generations is enough time for the production of a tremendous amount of new variability. If the explanation of diversity can be found only in the articulated pattern of the whole, what pieces (if any) of a cultural puzzle are most important for composing a picture, and what aspects of cultural variability should we begin describing? Likewise, whose empirical picture of the world should we utilize if all versions of history or synchronic variability are not (or are) equally valid? While Silberbauer suggests that the task of anthropology does not end with descriptive typology, he would have us believe that more and better description is critical for a complete understanding of diversity, which will in turn establish a foundation for generalization and comparison: '. . . it would be an empty or misleading ethnography which recorded only what the anthropologist considered to be significant differences . . .' (p. 29). Is it possible, or even useful, to attempt a more complete description by including variability that the ethnographer feels is insignificant? Clearly outlining specific questions and deductive methods seems to be the only way to decide what observations are relevant. Whether the questions and methods are relevant, interesting or appropriate is another matter.

Lewis Binford (1967) pointed out three decades ago that archaeological descriptions do not speak for themselves in explaining variability. After reading this volume we suggest that studies of contemporary foragers face the same dilemma: increased precision in illustrating diversity, new correlations between social factors and aspects of subsistence behaviour, a better understanding of the dialectical nature of ethnography, recognition of the importance of emic perspectives - all these may serve as important sources of cautionary tales, but they are not sources in and of themselves for explanations of why behaviour varies. Why not admit that our descriptions are always inherently incomplete and that we always impose on data by assigning to them values we think are significant, and turn our focus toward utilizing details of diversity as sources to falsify hypothetical predictions? This is a fundamentally different approach from the inductive endeavour of seeking 'appropriate' abstractions (emic, etic or otherwise) for comparison and generalization, and it is, we argue, more sensitive to the post-modernist critique that bias infects everything we observe. The scientific approach minimizes bias by sacrificing holism for analytical precision.

An approach like Kent's that appeals to all-inclusive theoretical orientations fails to generate any testable hypotheses about why behaviour might vary between groups. Analyses of diversity best begin with simplifying (and explicit) theoretical assumptions and move toward explanation by focusing on predicted relationship between one or two particular economic, political, cultural, historical or environmental factors and one or two aspects of behavioural variability. We can draw upon a theoretical foundation of behaviour (lacking in Kent's 'holistic description' approach) for hypotheses to clarify how we might expect these external factors to influence behavioural outcomes, and test our predictions against observations. This is not to say that predicted relationships are more important than others in affecting patterns of behaviour. Obviously, by isolating aspects of behavioural variation and extracting them from their contexts for examination, we can only hope to explain a fraction of the whole complex of cultural traits at any one time. But by proposing many different explanatory hypotheses to account for variability in many different aspects of behaviour, we may be able to piece together a more coherent explanatory picture of the whole. Reading Kent's volume only offered us more encouragement that it is this 'reductionist' approach (as Silberbauer terms it), cumulatively applied and grounded in some sort of theory of behaviour, which offers genuine utility for deductively explaining diversity.

These two approaches, testing explanations vs holistic description, are well represented in the volume and differences in the utility of each can be illustrated by contrasting Kent's contribution (pp. 125-56) with the chapter provided by Blurton Jones, Hawkes & O'Connell (pp. 159-87). Kent demonstrates a finding with regards to hunting variability among the Kutse: that consumption returns per hunter are conditioned by post-acquisition sharing. Skilful Kutse hunters give more than they acquire, while less skilful hunters receive more than they acquire. She then suggests that this is a product of the inherent flexibility (resulting from mobility) required to maintain an 'atmosphere of egalitarianism', the essential ethos of Basarwa culture (p. 155). We are left with questions of infinite regress: if mobility conditions flexibility which requires egalitarianism that in turn promotes sharing to even out hunting variability, then why one type of mobility as opposed to another? Noting that hunting variability is conditioned by sharing does not explain the pattern, it explains it away. Predicting this specific relationship with a hypothesis would be another matter.

How can we begin to sort out what factors are important in conditioning variability when we know that not all are of equal importance? As opposed to Kent, Blurton Jones et al. suggest we cannot begin by continuing to seek more correlations to interpret. The utility of their approach (regardless of whether or not one agrees with the tenets of their theoretical framework) is self-evident: by clearly delineating hypotheses with competing predictions we can evaluate how well each can account for behavioural variability, in this case differences between the !Kung and Hadza in children's foraging, fertility, parental investment, and hunting strategies. A similar approach by Hewlett (pp. 215-44) to understand variability between west African hunting practices is also productive. As Bird-David notes (p.302), strict hypothesis testing may trade-off 'ethnographic flesh' for clarity, yet the sacrifice can be of tremendous heuristic value.

This brings us to Robert Kelly's ambitious monograph, The foraging spectrum. Kelly provides a useful resource for students and researchers who are looking for syntheses of conceptual approaches to hunter-gatherer studies, a framework of theory to explain variability, and empirical diversity in the life-ways of foragers. While the data he uses to assess his and other's ideas about factors that condition diversity are limited to addressing specific questions, they are drawn from an astonishingly wide range of studies. In attempting to 'review some of what anthropology has learned about variability among ethnographically known hunter-gatherers' (p. 6) '. . . and [provide] some idea of what might account for it' (p.7), Kelly's goal is not dissimilar from Susan Kent's. However, Kelly's limited analyses of specific questions and his focus on hypothesis testing greatly enhances the overall clarity of the work and sets a precedence for future syntheses of foraging studies.

Kelly begins the book with two chapters on the role of hunter-gatherer studies in anthropology and the paradigms that have structured anthropological thought about their importance. While brief, his outline of the history of hunter-gatherer studies sets an appropriate background for his discussion of contemporary approaches. In chapter 2 he details theory in behavioural ecology which, for the most part, sets up his treatment of subsistence, mobility, reproduction, sex differences, social organization and prehistory throughout the rest of the book. While much of Kelly's discussion of behavioural ecology is critical for the rest of his book, in some ways he undercuts the real advantage of the theory (and theory in general). At times, Kelly seems to confuse theory with the hypotheses it generates, and in so doing may misidentify the strengths and overplay the limitations (the same limitations that have been dealt with time and again over the last 15 years) of behavioural ecology.

For example, while he is correct in stating that behaviour is assumed to be linked to fitness, it is not the case that behavioural ecologists do (or can) test this assumption. This leads Kelly to surmise that 'behavioural ecologists are interested in the mean fitness of some class of organisms' (p. 52), and that they attempt to measure the fitness consequences of particular variants of behaviour by using proxy measures of fitness. This would be a fruitless task: inclusive fitness is a measure of lifetime reproductive success and genetic contribution to future generations, the cumulative product of innumerable trade-offs and decision-making strategies about food, sex, social interaction, parental investment - anything which impinges upon inclusive fitness. Our goal as behavioural ecologists is not to measure the fitness components of particular behavioural strategies, but to test hypotheses about how those strategies, as opposed to others, are maintained by selection. That the cumulative effect of behaviour over the lifetime of an individual is designed to maximize inclusive fitness is the assumption provided by evolutionary theory, an assumption we cannot test if we explicitly adopt the theory as the framework from which we derive testable hypotheses. We have the feeling that sometimes, failure to recognize this undercuts the strength of 'isolating different factors and idealizing away from reality . . .' (Boyer 1995: 175) for doing science. Kelly's discussion of culture (pp. 58-64) seems to indicate that he does not always appreciate the power of well established theoretical assumptions as a source for falsifiable and competing hypotheses. In some ways this leaves some dissatisfaction, especially when Kelly discusses issues like food sharing and the sexual division of labour.

Depending on the extent to which the reader is concerned with how well behavioural ecology, as opposed to other frameworks, can generate hypotheses to explain human variability, this may (or may not) be a relatively minor quibble. Despite this, we enjoyed Kelly's book immensely. His book is for the most part a lucid example of the advantage of explicit appeal to theory as a guide for questions, hypotheses and data, and as such it should be an ideal teaching tool to introduce students of anthropology to science. We agree with Kelly wholeheartedly when, in his concluding chapter on how we might understand past variability in hunter-gatherers, he writes, '. . . the construction of methods to make inferences from archaeological remains is inextricably linked to an understanding of the variability in behaviour. One cannot reconstruct the past without simultaneously trying to explain the past' (p. 343). This is especially important in issues where more accurate reconstructions require us to explain why behaviour in the past may have varied and should be of equal importance in dealing with contemporary diversity. As Kelly notes, the only way effectively to deal with this is by utilizing hypotheses that are independent of the empirics at hand.

The utility of answers to questions about why human behaviour varies depends on how well we can avoid explaining away variation. It may often be tempting to explain away behavioural diversity as the result of historical and proximate cultural processes. This, however, begs the question of interest: why is behaviour patterned in one way as opposed to another? Our understanding of variability will suffer if the answer to this question can only appeal to historical processes. Evolution is of course a historically contingent process: current variability is constrained by the trajectory of past change. Many phenotypic features exist not for functional reasons, but as a result of unique, sometimes stochastic, histories. As a result, any adaptation is seriously constrained by phylogeny (Williams 1992: 72-6). This does not mean, however, that we cannot understand the maintenance of variability in functional terms (as opposed to proximate causation, see Krebs & Davies 1993: 4-8), it just means that we won't be able to understand all variability this way. The only way to know, however, is to test our ideas against observations. An illustration of evolutionary history or a description of diversity, no matter how precise, will not explain variability. This was what distinguished Darwin from Linnaeus: Darwin realized that no matter how well we describe different historical trends or synchronic diversity, we cannot explain differences with increased precision of classification. Kent's volume and Kelly's book are commendable in that they provide description that assists in demolishing our imposed typologies of hunter-gatherers. But, as many of these authors note, tearing down taxonomies will not explain diversity any better than building them.

References

BINFORD, L.R. 1967. Smudge pits and hide smoking: the use of analogy in archaeological reasoning, American Antiquity 32: 1-12.

BOYER, P. 1995. Ceteris paribus (all else being equal), in J. Brockman & K. Matson (ed.), How things are: 169-75. New York (NY): William Morrow.

KREBS, J.R. & N.B. DAVIES. 1993. An introduction to behavioural ecology. Oxford: Blackwell.

WILLIAMS, G.C. 1992. Natural selection: domains, levels, and challenges. Oxford: Oxford University Press.