From Ethno-Science to Science, or ‘What the Indigenous Knowledge Debate Tells Us about How Scientists Deﬁne Their Project’

ROY ELLEN∗

ABSTRACT This paper begins by examining the response of the organised scientiﬁc community to the claims of the indigenous knowledge lobby, and with some observations on the dichotomy between science and traditional technical knowledge. It reiterates the view that the potency of the distinction arises from a fusion of the general human cognitive impulse to simplify the processes by which we understand the world, reinforced by the socially-driven need of science to maintain an effective boundary around the practices which scientists engage in. The paper goes on to argue that the existence of these two epistemological meta categories obscures the presence of different ways of securing predictive knowledge of the material world, each of which is characterised by a distinctive conﬁguration of cognitive and technical features, and which in several ways cut across the usual dualism between science and traditional knowledge. The argument is illustrated using examples from the history of biology and the ethnography of ethnobiological knowledge. It engages critically with insights drawn from cognitive psychology, the philosophy and sociology of science, and cognitive anthropology, as well as with scientists’ own descriptions of what distinguishes the mental operations in which they engage.

1. The history of ‘indigenous knowledge’

The surge of interest in indigenous knowledge during the eighties and nineties of the twentieth century produced, in its turn, a counter offensive from some scientists concerned at the kinds of claims that proponents of indigenous knowledge were making, an attempt by others to reconcile

∗University of Kent at Canterbury.

c© Koninklijke Brill NV, Leiden, 2004

Journal of Cognition and Culture 4.3

traditional knowledge and science, and to distinguish both from pseudo science.1 Many otherwise pragmatic individuals were wary of endorsing the utility of ‘indigenous’ knowledge because this seemed to question their own credentials as scientists and professionals. Similarly, some scientists’ detected an embedded anti-scientiﬁc tradition, historically rooted in certain formulations of cultural relativism, and most recently evident in its post modernist guise. The more extreme versions of this position, it is often said, undermine the possibility of objective science, and go against our common sense experience that enough science ‘works’ sufﬁciently well to allow us to rely on the predictions upon which medicine and engineering depend. This overcharged interchange has been dubbed the ‘science wars’ (Anderson 2000). In retrospect, both sides were making some outrageous claims, and for the most part there is now a large degree of consensus. I do not wish to revisit this ground here, merely to register its existence. Nor do I wish to return to the terminological squabbles which have confused the issue, or to provide further endorsement of the the value of indigenous knowledge. The important thing is to note the terms, conditions and timing of the debate. For over much the same period other scientists have been embroiled in a parallel and equally acrimonious debate with cultural theorists, sociologists and historians of science, intent, as it was seen, in undermining the ‘objectivity’ and value-neutrality of science. Some, it is asserted, sort to rewrite the great scientiﬁc breakthroughs in ways which the authors would claim improved on older, ideological, revisionist and selective ‘presentist’ histories of science, histories which had airbrushed out the inconvenient non-scientiﬁc context and presented us with representations of the minds and achievements of great scientists which suited the present day high priests and guardians of scientiﬁc method. At a distance, the potency of the sterile dichotamies being drawn here arise from a fusion of a general human cognitive impulse to simplify the processes by which we understand the world (reinforced by the socially-driven need of science to maintain an effective boundary (Nader 1996: xii-xiv, 3-4) around the practices which scientists engage in), and of the West’s mission to preserve its cultural pre eminence. I wish to suggest that the opposition of science and ‘indigenous knowledge’ (which in the minds of the more extreme protagonists becomes 1See e.g. the editorials (Anon. 1999, 1999a) and correspondence (Diamond 1986; Johannes 1987) in Nature. See also ICSU 2002.

ne between science and superstition) is yet another manifestation of what anthropologists have called ‘the great cognitive divide.’ Even though the
pposition between primitive and civilised thought has fallen under the weight of the evidence, and its surrogates (most recently that between literacy and orality) have similarly been shown to be problematic (Frake 1983), the apparent need to divide the world into just two ways of thinking persists.
- The existence of two epistemological meta-categories – let us call
them for the time being science and indigenous knowledge – unfortunately
bscures the evident presence of many different ways of securing predictive knowledge of the material world, each of which is characterised by a distinctive conﬁguration of cognitive and technical features, and which in several dimensions cut across the usual dualism between science and indigenous knowledge. In this paper I try to look at science as a cognitive anthropologist might look at a local ethno-science. My examples will be largely, though not exclusively, drawn from natural history and biology, on the grounds that this is my own area of specialist knowledge, and I shall attempt to relate this to the historical emergence of global, transmodern, ‘scientiﬁc’ biology, with its formative phase in Western cultural history between 1600 and 1900. I shall ﬁrst examine theories of science, in relation to its actual practice; consider the proposals of those who see science as different only in degree from wider cognitive practices, to varying extents constrained by neurophysiology, either distinctively human or shared with
ther higher organisms. This de-construction of the uniqueness of science inevitably means that it becomes ever more difﬁcult to distinguish and label the practices in which those we call scientists engage. I shall examine the meta-category of science in relation to the historical generation of its antonym (indigenous, local, traditional, intuitive, ordinary, knowledge), and in the light of all the deconstruction try to characterise the combination
f cognitive, technical and social practices which help us maintain modern global science as a social and cultural entity distinct from the many cross- cutting ethno-sciences.

2. Diversity in biological knowledge traditions

2.1. Indigenous knowledge

Indigenous or traditional knowledge emerged as a generic category quite late, during the second half of the twentieth century (Ellen and Harris 2000). In the seventeenth and eighteenth centuries European natural ists and medical practitioners readily assimilated knowledge from newly colonised or contacted people, and it was only really with the rise of the ‘development industry’ in the twentieth century that such knowledge, hav ing been at ﬁrst mute and then actively rejected, was subsequently redis covered and celebrated. From about the mid nineteen-sixties the tendency to marginalise local biological knowledges had begun to be put into re verse, prompted by a combination of romantic idealism and pragmatism. Attempts were made to make it more like science, to reduce it to a codi ﬁed, packageable commodity, secure and easily transferable from one place to another. And, once ethnobiological knowledge had been drawn within the orbit of modern science and its origins forgotten, it became difﬁcult to know where to place the boundary between the two. Indeed, changing the boundaries was often sufﬁcient to redeﬁne something as science, as what deﬁnes it as such was to a considerable extent determined by who prac tised it, and in what institutional context the practices took place. We can see a parallel here between the scientist and the artist under high mod ernism and post-modernism, where it is less the substantive materiality of the creation which is crucial than the context and who does the creating.

Another way of highlighting the absurdity of the category ‘indigenous knowledge’ or ‘traditional knowledge’ as applied to biology, is to deﬁne it as what is left once ‘biological science’ has been subtracted. Looked at this way, ‘indigenous knowledge’ is just like the multitude of manifestations of ‘the other,’ what is left when we take ourselves out. As a single all-encompassing notion it is nonsense, as terminologies, deﬁnitions and cognate concepts vary throughout their geographical, local-global and various historic and disciplinary refractions. There are many indigenous biological knowledges, each accessing the real world to various degrees of imperfection and subjectiveness. These diverse biological knowledges, for well over 60,000 years, have constituted the main body of our adaptive knowledge as a species. Some are part of specialised scholarly textual traditions (which I shall return to shortly), but those outside such traditions have been represented as sharing a number of broad common characteristics (Ellen and Harris 2000: 4-5). They are rooted in particular places and sets of experiences; are generated by people living in those places; are mostly orally-transmitted or transmitted through imitation and demonstration; are a consequence of practical engagement in everyday life constantly reinforced by experience and error; are the product of generations of intelligent reasoning. They are empirical in character, orality to some extent constraining the kind of organisation necessary for the development of abstract theoretical knowledge. The redundancy which they embody aids retention and reinforces ideas; they are ﬂuid and the outcome of continuous negotiation, constantly changing, being produced as well as reproduced, discovered as well as lost; though often represented as static. They are characteristically shared to a much greater degree than global biological science, but are still socially clustered within a population, by gender and age, for example, and preserved through distribution in the memories of different individuals. Specialists may exist by virtue of experience, but also by virtue of ritual or political authority. Although knowledge may be the prerogative of particular individuals and may achieve a degree of coherence in rituals and other symbolic constructs, it does not exist in its totality in any one place or individual, devolved not in individuals at all, but in the practices and interactions in which people themselves engage. Such knowledges are characteristically situated within broader cultural traditions, so that separating the technical from the non-technical, the rational from the non-rational, is problematic.

2.2. Scholarly ways of knowing

By comparison, the great ‘scholarly ways of knowing,’ or what Dunbar (1995a: 36) rather derogatorily calls ‘cookbook science,’ come midway be tween these essentially local folk knowledges and biological science. They combine knowledge dependent on an agreed shared authority with that of the personal authority of a practitioner. They are often grounded in writ ten texts, and resemble the European scholarly traditions. Galenic, Tibetan, Chinese and Ayurvedic traditions of medicine differ from each other, but each have a notion of scholarship in common (see e.g. Bates 1995). Where the scholarly and local folk traditions merge is unclear, and as in the Eu ropean case there is historical evidence to suggest, for example, that the great Asian herbalist systems have been systematically absorbing and then replacing local folk knowledge. We see here something very reminiscent of the codifying and simplifying processes which accompanied the incorpora tion of European folk knowledge into the early modern scholarly traditions. These ‘folk’ traditions have themselves become highly codiﬁed. During me diaeval and early modern Europe, proto-scientiﬁc knowledge of plants and animals superseded folk-knowledge by classiﬁcation, analysis, comparison, dissemination (usually through books and formal learning) and thus gener alisation. The process was not sudden: for a long time common experience, oral tradition, personal experience and learned authority contributed to the ‘received wisdom’ upon which organised specialist knowledge, particularly medical knowledge, depended; and delineating the boundaries between un codiﬁed folk knowledge, professionally restricted organised knowledge, and proper scientiﬁc knowledge is not always easy. Some have argued that the phylogenetic taxonomies of contemporary post-Linnean biology are based on a European folk template (Ellen 1979; Atran 1990) and some have gone even further by claiming that the European folk scheme and that which emerged in nineteenth century professional natural history are no more than variants on a single cognitive arrangement to which all humans are predisposed through natural selection (Atran 1998; Boster 1996).

2.3. Proto-scientﬁic and scientﬁic knowledge

A similar process in terms of the epidemiology of ideas was at work in the way non-European folk and scholarly traditions were absorbed into science, and simultaneously employed – at least in the early modern period – to legitimate it. I have, for example, discussed elsewhere (Ellen and Harris 2000: 8-10) the work and inﬂuence of the Dutch East India Company naturalist George Rumphius, while Grove (1996) has provided us with the parallel case of Garcia da Orta in relation to Hendrik van Rheede. Their work resulted in the publication of scientiﬁc accounts of new species and revisions of taxonomies which, ironically, depended upon a set of diagnostic and classiﬁcatory practices which though represented as ‘Western science,’ had been derived from earlier codiﬁcations of indigenous knowledge. Rumphius and other scholars of their time also sought to confer authority on their work and writing by appealing to the wisdom of local peoples. Linnaeus, for example, self-consciously drew on the traditions of the Lapps in order to promote his scientiﬁc work, to the extent that he had his portrait painted wearing Lapp costume, even though – and revealingly – the details of the clothing are ethnographically questionable (Figure 1).2 Seeing Linnaeus dressed as a Lapp in early eighteenth century Sweden conveyed much the same impression and authority as the proverbial photograph of the anthropologist’s tent in the context of ethnographic writing (Clifford 1986: 1-2).

Rumphius and Linnaeus make an interesting comparison because their lives and work fall around the boundary between what Foucault (1970: 127-8) has called the natural history and biology epistemes, or what we might call a Foucouldian moment or transition, when traditional knowledge or scholastic knowledge becomes science. Rumphius had no pretensions to be a ‘scientist’ and no understanding of what this might have meant in contemporary terms; and otherwise performed successfully as a Company historian and naturalist. Linnaeus, by contrast, and certainly the later Linnaeus, was more self-consciously operating within a scientiﬁc paradigm, controlling a network of acolytes who contributed to his wider comparative project. But on either side of the epistemic divide those who would call themselves scientists often had no problem in underpinning their scientiﬁc observation with ideas and perceptions which today we would consign to the realm of the non-scientiﬁc: Rumphius, like his inﬁnitely more famous contemporary, Isaac Newton (e.g. Dobbs 1975), believed in magical ‘ancient wisdom,’ astrology, the spontaneous generation of living matter, species transformation and the idea of the homunculus (Figure 3). This latter notion, whereby the discovery of the spermatozoan through microscopy is combined with an imaginative folk embryology which places the pre-formed human actually within the head of the spermatozoan, 2Linnaeus exaggerated his dependency on Lapp knowledge and included various sleight of-hand textual and illustrative devices in his Flora Lapponica to increase his scientiﬁc authority, he assembled a Saami costume for his grand tour of Europe and embroidered his travel tales. For Koerner (1999: 64) his ‘Lappland invention’ passed beyond ‘self aggrandizement’ and ‘became governed by a central fantasy: the coloniser masquerading as the colonised.’ ‘As for Linnaean lists generally, minerals were followed by plants, animals, and local technologies and ethnography. On a conceptual level, and within the master plot of import substitution that governed his natural history, Linnaeus had four Lapland strategies: to harvest natural resources; to support dairy and grain farming; and to introduce exotic cultivars’ (Koerner 1999: 77).

Figure . 1

Mezzotint of Linnaeus in Lapp costume; Dunkerton, from a painting

by M. Hoffman. Reproduced by permission of the Linnean Society, London.

Figure . 2

Figure . 3

Spermatozoa as depicted in different seventeenth century images: a, b

ﬁnds an interesting echo in the anthropomorphic faces which Rumphius’s plagiarist and follower, Francois Valentyn, placed on the heads of nymphs of the arthropod Irona renardi observed under a magnifying glass (Figure 4).

During the nineteenth and twentieth centuries local knowledge was increasingly tapped and codiﬁed. Such practices became so routinised that once absorbed into ‘scientiﬁc’ solutions local biological knowledge disappeared from view, separated from its symbolic underpinnings, and insufﬁciently ‘real’ to merit any certain legal status or protection in the same ways which gave value and ownership to western scholarly knowledge and expertise. Even when the knowledge was clearly being utilised it was often redescribed in ways which eliminated any credit to those who had brought it to the attention of science in the ﬁrst place. With the reiﬁcation of the modernist project of science-driven progress during the middle part of the twentieth century, the traditional became widely accepted as irrelevant, and indigenous knowledge was used much more selectively, being virtually abandoned for several decades in certain ﬁelds, such as pharmacology. Under ‘high modernism,’ science becomes the superior a priori standard by which all truths, including those derived from traditional teachings are measured, validated and valued (Pfaffenberger 1992; Turnbull 2000). Thus, the boundaries between science, scholarly knowledge and folk knowledge, as these terms apply

Figure . 4

Early illustrations of the crustacea Squilla mantis and the nymphs of

to biological phenomena, are constantly shifting, and the distinctions themselves are not always helpful. All knowledges are anchored in their own particular socioeconomic milieu; all are indigenous to a particular context (Agrawal 1995: 5).

3. Theories of science

The ways in which science has been hitherto theorised are broadly of two kinds. The ﬁrst, what we might call traditional philosophy of science, is characterized by idealistic or programmatic theories (which take their cue from assertions about how good science should be conducted). The second kind of theory varies somewhat, but includes cognitive theories, emphasising the empirical character of scientiﬁc thought (either as a species of everyday cognition or as a special kind of thought); theories which focus on the empirical analyses of laboratory practice (which emphasise what scientists do, say and how they interact in working relations); and cultural theories, either those which scientists construct for themselves about their conduct, or those derived from outside observers. These latter emphasise science as a combination of shared norms and values, cognitive and technical practices. The data which sustain all of these are of four kinds: historical studies of prominent scientists, ethnographic studies of the conduct of contemporary science, ethnographic studies of indigenous knowledge, and psychological studies of science as a kind of thought.

Programmatic theories which emphasise ideal and formal methodology (‘laws,’ falsiﬁcation and objectivity) have become increasingly difﬁcult to defend and largely abandoned. Although some philosophers of science and scientists themselves do so, most who seriously reﬂect upon the subject rarely seek to defend them in their pristine form (Oatley 1996: 131). The archetypically austere Popperian account still has surprising currency in some quarters, but is now much critiqued in the world of philosophy of science itself. Such a view is often linked to the idea of science as the ‘real’ or ‘absolute’ truth, as a repository of ‘facts,’ ‘out there,’ waiting to be discovered by great individual minds rather than being the outcome of interactions in intersubjective space (ibid., 126, 133-4). Rather, the new consensus is that ‘philosophers have grossly overestimated the way in which science is actually done’ (Dunbar 1995a: 78), and that ‘we get trapped by the ideals of science when we insist on an exclusive role for well formed computation, veriﬁability, and truth conditionality’ (Bruner 1996: 100). The real ‘multi-layered’ world of science, it is observed, involves contradictions in actual practice and radically different kinds of theory, where some ‘operate programatically and others (are) more concerned with the details.’ Whereas the programme may function after the fashion of a Kuhnian paradigm (a framework theory), subsidiary hypotheses are formulated and tested according to neo-Popperian criteria (Dunbar 1995a: 23). In a sense, theoretical and empirical observations represent two parallel worlds with different internal rules, which feed into and reinforce one another, a view which is essentially that of Lakatos and of Carl Hempel’s hypothetico-deductive model (ibid., 25). This accords with the notion that science is a methodological prescription rather than a particular body of theory, a method for ﬁnding out about the world (ibid., 32). Nevertheless, most scientists continue to deﬁne their project as the search for truth, accurate data about the world, and in terms of theories and hypotheses that model this search (e.g. Wolpert 1992; see Oatley 1996: 133-4), and for this reason tend to imagine that their thinking is special, despite the fact that cognitive psychologists using these very methods ﬁnd it difﬁcult to locate its specialness. For Bruner (1996: 103) there can be no one model of cognition or of the mind, and we should ‘avoid theories of meaning making tied exclusively to the needs of science’ for which no reductionist theory will ever do proper justice. Indeed, simple, formal, deﬁnitions of what science is are always going to be problematic because they end up excluding practices and kinds of knowledge which are, in common-sense terms, integral to how science works. Science is not consistently ‘rational, objective and produced according to the canons of scientiﬁc method,’ but is rather ‘messy, contingent, unplanned and arational,’ a polythetic practice and conditional ‘assemblage’ of local knowledges (Turnbull 2000: 6, 14).

Despite such an emerging consensus, there is rearguard resistance to the abandonment of programmatic theories. This is in part fuelled by prevailing popular conceptions of science as narrow, compartmentalising, technical, esoteric, accessible and relevant only to specially-trained experts. But resistance is also encouraged by the perceived excesses of the cultural constructionists (the likes of Kuhn, Bloor and Latour) who, it it asserted, in sist on trying to show that science is ‘nothing more than,’ those great minds such as Darwin, Boyle or Hooke, or their lesser contemporary counterparts who beavered away in laboratories, formulating their problems in a way which ‘merely’ reﬂects the society around them. Some scientists, trapped by assumptions that all social science commentaries are tarred with the same simplistic post-modern brush, feel justiﬁed in resorting to the equally fatuous assumption that if sociologists make such superﬁcial claims it is because they are ‘unable to grasp the scientiﬁc arguments’ (Dunbar 1995a: 157). The problem with the sociology and cultural history of science is not so much that it claims that scientists are creatures of their particular social and cultural circumstances, but that it does not go beyond this to examine the similar cross-cultural cognitive practices within which science might be embedded. It seems to me that the sociological ‘strong theory’ is a wholly illusory challenge to the credentials of science as a legitimate and powerful tool for understanding the world, and is now, anyway, so obvi ously true that it is no longer interesting. What is much more interesting is that despite its social determination, science is so obviously effective. With Anderson (2000), I take the view that even if we could demonstrate that science is one hundred percent socially constructed, generally-speaking, it is still an accurate representation of the world. To say that something is socially constructed is not to say that it is inaccurate, only that people learn what they do (their ‘social constructions’) from each other. I shall return to this idea later, but in the following section provide the evidence for the apparently opposite view: that ‘science-like thinking’ is not merely a human universal, but perhaps even derives its universality from cognitive imperatives which are a propensity of all higher organisms.

4. Science-like cognition as a human universal

The growing recognition of, and respect for, traditional non-western forms of knowledge of the environment led during the seventies and eighties of the last century to claims that it was a kind of science. This view gained support in the context of the many failures of non-participatory, apparently ‘science-driven,’ top-down development (Ellen and Harris 2000: 19) and the consequent increasing utilisation of discoveries originating in local knowledge systems. A revival in the recognition of the historical inputs of such systems into the growth of Western science, for example Indian innovations such as the decimal system, the concept of zero, and procedures in trigonometry and algebra, two millennia before Newton and

Liebnitz, was also important in this respect. Partly due to the force of such claims, some scientists have themselves adopted the view that empirical science is a human capacity characteristic of all societies (Dunbar 1995a: 58). Indeed, although local ecological and social circumstances provide different constraints and opportunities, what is very clear is the remarkable extent to which cultures converge on the same answers and solutions. Among the most widely quoted anthropological evidence for this is the claim that classiﬁcatory knowledge of the kind which Brent Berlin calls general purpose – reﬂects universal cognitive regularities. But there are also other kinds of evidence, such as the effective demonstration that farmers facing similar problems in widely separated parts of the world have developed basically the same dynamic understandings of the relationship between elements in an agroecological system, for example in relation to fodder crops and nitrogen ﬁxation (Walker, Thorne, Sinclair, Thapa, Wood and Subba 1999; c.f. Iskandar and Ellen 2000). Other patterns of convergence in general perception of biological properties are observable in relation to medicinal plants: the widely-recognised ‘hot-cold’ frameworks, and patterns in the selection of taxonomically unrelated plants on the basis of chemical similarities, biases towards certain plant families displaying useful kinds of bioactivity, and a recognition that the properties of plants which make them toxic are the same as those which make them medically potent (Johns 1990; Moerman, Pemberton, Keifer and Berlin 1999). Such claims to the universality of particular practices and their out comes are interpreted by some as evidence that all peoples share a basic way of apprehending the natural world grounded in a common evolu tionary history. The notion of a shared infrastructure of perception and cognition with respect to biological knowledge has been termed by its pro ponents ‘natural history intelligence’ and is linked most prominently to modular theories of the mind. The diagnostic features of such intelligence, or such a module, generally include a shared concept of basic natural kind (a species-like concept) reﬂecting a view of the biological world as a series of discontinuous entities; an ability to recognise and respond to things as living matter (and more speciﬁcally an ‘algorithm for animacy,’ or theory of ‘aliveness,’ a set of core principles which guide identiﬁcation and reasoning about animals arising from the recognition that they pos sess an internal source of movement); a capacity to intuit certain kinds of behaviour based on expectations arising in part from common experiences derived from phylogenetic similarities or observations of human behaviour, and strategies for classifying biological diversity (Atran 1990, 1998; Boster 1996; Carey 1996; Ellen 2003; Gelman 1990; Keil 1994; Mithen 1996: 52-4).

Because none of this is accessible other than through its local cul tural versions, distinguishing what are shared human universals from what are simply culturally widespread is problematic.3 But whatever the degree and character of epigenetic embedding, strong evidence for the cogni tive universality of science-like processes is found in the ability of people everywhere to generalise the principles of plant and animal biology on the basis of knowledge of individual organisms. In terms of understanding human cultural adaptation to different environments, knowledge of such general principles of biology may be more important than breadth of for mal knowledge or depth of substantive knowledge of individual organisms, or indeed of any universality in the taxonomic ordering of biodiversity. What is central here is the ability to transfer general lessons learned with respect to one organism to another organism. To some extent this may encourage us to generalise a module for ‘natural history intelligence,’ with its predisposition to recognise common aspects in the functioning of living 3There is dispute as to whether folk biology is a non-theoretical innate core module, a congenitally innate module, an intuitive theory, or a framework theory constructed during the ﬁrst decade of life (Carey 1996: 192; Keil 1992, 1994). For Carey (ibid., 193), for example, the position espoused by Atran (and Sperber) that folk biology is a non-theoretical innate core module is problematic ‘because features of folk biology which emerge in early childhood are not domain speciﬁc, and because those features which are domain speciﬁc are probably neither innate nor theory-neutral, because they fail to confront (the) full implications of perception, and because they face the ‘problem of theory laden attribution.’ While Atran wants taxonomy and essentialism to be both biological and pretheoretical (ibid., 196), for Carey they are ‘plausibly innate but not speciﬁc to folk biological classiﬁcation, and true only as properties of the adult system.’ In support of this she cites child study data from the US and Nigeria (ibid., 199) which suggest that only by age 10 does general reasoning become incorporated as intuitive biology, thus supporting Atran’s claim that folk biological understanding is part of universal folk taxonomy but conﬂicting with his vision that this is part of an innate cognitive model. She concludes that there is no good evidence to see folk biology as an innate module like language, number, reasoning about objects, or reasoning about persons, but that it is rather a ‘framework theory’ (ibid., 189).

things, but much substantive knowledge of individual types of animal de rives from analogical reasoning with respect to human bodily functioning. Thus, knowledge of human anatomy mutually reinforces knowledge of an imal anatomy. Knowledge of the human body is, therefore, partly based on knowledge of animal bodies acquired in hunting, food preparation and livestock keeping; while understanding of animal physiology, pathology – and even psychology – derives from modiﬁed human experience.

But if science-like cognition is a pan-human universal, and if this universality is accompanied by evidence for common underlying cognitive operations under ontogenetic control, then the question must arise as to whether such cognition is uniquely human at all. Some would claim that it is not, and that the methods of low-level empirical science are practised by all advanced non-human animals, or in an even stronger formulation that ‘Western science is a product of a highly formalised version of something very basic to life’ (Dunbar 1995a: 58; see also 33, 57). Two processes underpin an organism’s ability to learn about the world: classiﬁcation and causal inference, on the basis of which are built up logical chains of cause effect sequences which permit leaps of inference (ibid., 58). Johnson-Laird (1982), for example, claims that storing knowledge as causal hypotheses (or models) is efﬁcient because humans do not have sufﬁcient memory to make the right responses by induction alone. The difﬁculty for anthropologists here has been in identifying cultural and cognitive traits of sufﬁcient discreteness to be accepted as unitary memes in the ﬁrst place, and the ways in which the human mind unhelpfully interferes with the conventional forces of selection by re-forming such units, linking them together in novel ways and attributing to them new (and sometimes contradictory) linkages and meanings (Aunger 2000).

5. Science as a specialised form of folk knowledge

We can see from both the preceeding brief history of the category ‘indigenous knowledge,’ and from the equally brief review of theories of science, that where the dividing line between scientﬁic biology and other kinds of biological knowledge lies is by no means obvious, and that both science and ethnoscience have common and interconnected origins. Indeed, ‘the whole of science’ as Einstein once famously remarked, ‘is nothing more than a reﬁnement of everyday thinking,’ an assertion which cognitive psychologists such as Deanna Kuhn have sought to substantiate, even citing hypothesis-testing in children’s verbal exchanges and the way they learn word endings as evidence in support of this (Kuhn 1996: 261-2, 275).

How ways of thinking (cognitive practices) develop and become avail able to individual humans can be instructively compared with how we develop ways of speaking. We are genetically provisioned with a range – if not an inﬁnite variety then a very wide variety – of cognitive options, just as we are provided with a wide congenital range in terms of our ability to generate and recognise phonemes through the audio-vocal tract (Lieber man 1984). But these options become more limited and less ﬂexible as we are habituated and socialised into local cultural practices – such as occu pational or social role playing. Within this range of cognitive activity, the most obvious distinction is not necessarily between scientiﬁc thinking and the rest (Kuhn 1996: 276), but rather ‘between thinking that is more versus less skilled, with skilled thinking deﬁned in its essence as thinking that re ﬂects on itself and is applied under the individual’s conscious control.’ For most of the time widely shared and deeply-embedded cultural responses which do not involve too much skill and reﬂexivity serve us well. Only when habitual behaviour and rules of thumb fail do we resort to conscious and reﬂexive science-like activity. In order to understand any differences there might be between science as a speciﬁc kind of reiﬁed knowledge making process, and science as an underlying proclivity, it is useful to list the characteristics of both which are often claimed to distinguish the ﬁrst from the second.

5.1. The grounds for belief in science and folk science are the same. Both require suspension of belief and the free and vigorous expression of contrary views; not simply the mechanical application of existing knowledge. As Kuhn (1996: 275) has demonstrated, ‘in both scientiﬁc and everyday reasoning, people must be able to distance themselves from their own beliefs to a sufﬁcient degree to be able to evaluate them, as objects of cognition, in a framework of alternatives that compete with them and evidence that bears on them.’ Thus, Richards (1993: 62) notes that the range of skills and strategies employed by farmers often extends beyond simple applied knowledge into a ‘ﬂuid body of improvisations’ about inter species ecological complementarity relevant to immediate needs, rather than the outcome of a prior ﬁxed ‘stock of knowledge.’ An essential dimension of this kind of thing is ‘thought becoming aware of itself,’ which in developmental terms occurs very gradually rather than at some discrete point during maturation (Kuhn 1996: 277).

5.2. Differentiated concepts of causality. Causality is a universally under stood notion, but as Lloyd (1990) suggests, it everywhere varies between different domains. In its widest understood form it is based on mechanical or physical analogy.

5.3. Philosophers differ in their rational reconstruction of what constitutes a theory (Carey 1996: 190), and some cognitive anthropologists deny that or dinary cognition relies on theory-like representational structures, and assert that folk biology, for example, is a pretheoretical cognitive module,’ even among adults (Atran 1994; Sperber 1994). For Atran (1996), ‘the structures of ordinary conceptual domains may strongly constrain, and thereby ren der possible, the initial elaboration of corresponding scientiﬁc ﬁelds.’ But for most cognitive psychologists, and I suspect, ethnographers, theory is as much an instrument of ethno-scientiﬁc thinking about the natural world as it is of science. It may, however, make sense to distinguish framework theories (foundational theories) from speciﬁc theories. If Wellman (1990: 191) is correct when he argues that cognitive scientists who write about intuitive theories usually mean framework theories, then by this logic folk biology is a framework theory.

5.4. A common feature of ethnobiological knowledge is the way in which knowledge is structured in terms of networks of understanding, linking individual species together in living contexts and entire landscapes, in contrast to formal science in the West which historically early reiﬁed the species and species-centred approach to understanding. This kind of systemic knowledge differs from biological science in emphasising long term processes, including cyclical environmental change. However, it re-emerges in modern science in ecology, climatology and the earth sciences. 5.5. Suspension of belief makes possible acceptance that the world is constructed differently from conclusions based on initial perceptions or pre existing cultural authority. Some scientists have held such counter-intuitivity to be the main condition of thought practices we describe as ‘scientiﬁc’ (Wolpert 1992), though there is increasing evidence that this is not the case. Atran (1996: 234) notes, for example, that supernatural beliefs are just as counter-intuitive for those who think them true as those who think them false, an obvious biblical example of which is the miracle in which Jesus turns water into wine. The types of belief which we regard as intuitive depend in part on substantive domain, place and specialist training. For a physicist what is counter-intuitive is, in certain respects, not the same as for a non-physicist. For instance, the idea that ice freezes faster when hot water is added (the so-called Mpemba effect, is counter-intuitive for most of us (including most physicists), but for an ice cream maker or somebody who reﬁlls livestock troughs in freezing conditions it is not (Chaplin 2004).

5.6. Experimentation. All science and folk science requires that in some form or another its practitioners evaluate relevant data against a possible set of explanations, and embrace a methodology which requires an expla nation which best ﬁts the data. Evidence for the repeated and systematic evaluation of experimental situations has by now been well-documented in the local agroecological knowledge literature, with respect to plant breeding and pest control (e.g. Cleveland and Soleri 2002; Richards 1985). Moreover, the experimental method in Europe preceded seventeenth cen tury prototypes by over 1500 years, Aristotle in his Metereologica showing clear understanding of this knowledge generation technique (Lloyd 1991; in Dunbar 1995a: 40).

5.7. But it is not simply a matter of extending recognised characteristics of science to ethno-science, but also of the reverse. Thus, the emphasis given to mathematics and abstraction in modern science sometimes seems to deny the signiﬁcance of narrative. As Oatley (1996: 123-4) has pointed out, scientists do not restrict themselves to mathematics, diagrams, and experiments. They use narratives to spin stories but can switch from these to the ‘paradigmatic mode,’ such as an equation, when necessary. Similarly, while Bruner (1996: 100) acknowledges that while there just might be ‘branches of mathematics and of the physical and biological sciences that are so formally or propositionally entrenched as to permit deviations without the support of folk takes,’ this is rare, and recalls a claim by Niels Bohr that the inspiration for the Principle for Complementarity grew by analogical story-telling drawn from an autobiographical experience.

6. Nature, the etak of science

It is not just that the speciﬁc articular and underlying cognitive practices of science and ethno-science appear to be the same, but also the cultural ‘doxa’ in which they are embedded and through which their performance is measured. By doxa I mean here – following Bourdieu (1977) – that network or framework of shared assumptions, habitual practices and axioms which we take for granted, but which give meaning to individual acts and thoughts and which allow them to function as a system. Doxa, for Bourdieu (1990: 20) is ‘the coincidence of the objective structures and the internalised structures which provides the illusion of immediate understanding, characteristic of practical experience.’ I would include here also ‘beliefs’ in the sense of the kind of propositions described by Goodenough (1990), a propensity for which hominin brain evolution has permitted.

For science and folk science to work there needs to be a framework of assumptions about how the world is constructed and how human actors relate to that world. Such a framework often corresponds to what we conventionally call ‘nature,’ that is some kind of conscious broad brush deﬁnition of the objects under scrutiny in relation to the human mind and actions in the world. Just as knowledge of individual organisms is embedded in ecological knowledge of the relations between them, and the relationship of assemblages of living things is understood in a wider landscape and in functional contexts, so there is a link between all culturally varied biological knowledges and local constructions of that aspect of the world we call ‘nature.’ Of course, this does not mean that this framework is everywhere constructed in the same way, nor does it mean that such a framework needs to be universally ‘accurate,’ only sufﬁciently robust to serve as a shared point of investigative departure. We can draw a parallel here with other, more speciﬁc, framework theories in science, such as those of Newtonian physics, which while inconsistent with prevailing conceptions of matter, and in other ways ﬂawed, have nevertheless provided an essential practical basis for most engineering. Such frameworks are cognitively reminiscent of the Micronesian etak system (Akimichi 1996; Frake 1985; Gell 1985; Gladwin 1970; Oatley 1977), for long a familiar ethnographic case study in undergraduate cognitive anthropology courses, included – one assumes – to demonstrate the impossibility of typologising modes of thought.

The etak system, you will recall, combines detailed empirical observations of environmental data – tides, currents, animal movements, winds, star movements – with the imaginary existence of an island and other symbolic entities, to make calculations of distance and direction which achieve a remarkable degree of navigational accuracy in inter-island voyages. The Micronesian etak, like ‘nature,’ is a moving target, if you will a ﬁction, which nevertheless aids investigative focus, and only makes sense because it helps achieve the desired objectives.

But even if we acknowledge that nature is no more than a convenient ﬁction, it does appear that all conceptions of nature are themselves underpinned by pan-human conceptual universals: one being the notion of what is ‘natural’ (primordial, essence); a second a tendency to contrast ourselves as humans and individuals with those biological others which lie outside of and around us; and thirdly, a compulsion to recognise and classify natural kinds as things in ways which suggest that we are evolutionarily adapted to cognise the natural world in broadly the same way (Ellen 1996). Thus, human biological knowledge, in whatever tradition or particular social world it has developed, always and simultaneously informs and reﬂects adaptive behaviour through ﬂexible cultural learning constrained by a common human cognitive framework. The cognitive roots of ‘nature’ indicate clearly how important such constructions are to basic adaptive ‘scientiﬁc’ thought, in that without certain cultural interventions or stimuli the cognitive processes themselves are not activated.

What is notable about conceptions of nature, including and perhaps most obviously in the etak system, and what has for long puzzled anthro pologists, is the ambiguous relationship in all cultural repertoires between the empirically observable and the symbolically asserted. This is often por trayed as a tension between levels of understanding distinguished as symbolic and technical (mundane) knowledge (knowledge versus know-how, savoir versus savoir-faire, ontological knowledge versus practical knowledge), or slightly differently between knowledge of and knowledge for, a genus of distinctions which goes back beyond Durkheim and Mauss to earlier philo sophical dualisms, particularly mind/body theories. Attempts to make such distinctions are not completely separate from attempts to distinguish knowl edge from skills, and the difﬁculty with virtually all such qualitative con trasts is that they are never hard and fast. Those anthropologists (such as

Brent Berlin) who have accepted the legitimacy of distinguishing the tech nical from symbolic have, on the whole, been those who have undertaken specialised studies of kinds of ‘indigenous knowledge,’ the ethnosciences in the sense used above. By contrast, those anthropologists who have avoided analysing technical knowledge, or who reject the distinction between tech nical and symbolic in the ﬁrst place (such as Mary Douglas), have been more interested in symbolic and social knowledge per se. This convergence of cognitive and symbolic approaches (Colby, Fernandez and Kronenfeld 1980; Ohnuki-Tierney 1981) is consistent with the recognition that all hu man populations apprehend both the social in terms of the natural world and the natural in terms of the social world, making many claims of uni directional metaphoric extension problematic. The two are intrinsically complementary. The classiﬁcatory language we use for plants and animals is derived from the way we talk about genealogical relations, and we un derstand the functional dynamics of both organisms and ecological systems in terms of our experience of participating in social systems, where technol ogy provides numerous productive analogies: say, the heart as a pump, the blood vascular system as a thermostat or the brain as a computer. More generally people attribute meaning to parts of the natural world around them by investing them with human and spiritually anthropic qualities (an imism). Indeed, there are striking similarities across cultures (Atran 1996: 234) between natural causality and the supernatural, in terms of symbolic content (e.g. animated substances and monstrous beings) organised through the same core set of cognitive mechanisms.

The symbolic merges, therefore, with the mundane, and both may be truly local knowledge. Local people themselves simultaneously embed their mundane knowledge in the symbolic, and their symbolic knowledge in the mundane, but are nevertheless often able to make inferences which imply the separation of one (or certain aspects of one) from the other when it matters. Recogition of this possibility is vital, since by emphasising the merging of local symbolic knowledge (e.g. world views) and technical knowledge, those who criticise the use of anything less than fully symbolically contextualised knowledge sometimes seem to deny the ability of people to discriminate different orders of inference or kinds of data at all, as well as the strong probability that people know that certain kinds of solutions to problems are easier to identify than others. Indeed, one of the major philosophical contradictions in science as we know it, is that while its effective conduct requires some foundational theory, a set of assumptions and metaphors which are beyond empirical testing (a kind of doxa), it is also part of a cultural tradition which ideologically separates the symbolic from the technical in terms of the latter involving the application of self-conscious rationality. It is a tacit acceptance (or avoidance) of this contradiction through a pragmatic division between spheres of rationality which permits NASA scientists to be simultaneously evangelical southern Baptists, and creationists molecular biologists. But if we accept this, then we must also accept that the same kinds of distinction might apply to Trobriand gardeners when evaluating the reasons for crop failure. We know that the art of good story-telling and science (see above) is for the authors to provide the conditions under which the reader can suspend disbelief, so we might reasonably expect there to be conditions in any belief system where the conditions are presented which allow for the suspension – or even the instantiation – of belief.

7. Science as an instituted model If it is the case, as many argue, that insisting on seeing science as a special cognitive practice, or in denying the science in ethno-science or in the learning strategies of other species, is a result of confusing science as a general process of thinking with idealistic theories of particular individuals, then we still need to establish what it is which separates out what scien tists really do in cognitive terms from what the rest of us do. As Kuhn (1996: 279) notes, it is difﬁcult to deny that ‘non trivial differences exist between everyday and scientiﬁc thinking,’ and to say that science is no more than common sense in a special institutional setting is to come dangerously close to saying nothing at all. But if science, in all its many cultural mani festations, is only possible through reliance on a doxic infrastructure which is rarely fully articulated or understood, I wish also to suggest that we treat science as it has developed since the sixteenth century in Europe and the neo-Europes as a more speciﬁc kind of ‘cultural model.’ D’Andrade (1990: 809) deﬁnes a cultural model as ‘a cognitive schema that is intersubjectively shared by a cultural group,’ but as Shore (1996: 45) has noted this deﬁni tion fails to distinguish between ‘special purpose models’ and ‘foundational schemas.’ There is no doubt also that historically ‘Western’ biological sci ence emerged as a special purpose knowledge in a very speciﬁc context of use, with a speciﬁc division of labour, and like other local knowledges has simply reﬂected the interests of a particular local culture, that of scientists, who work as a community within social organisations and according to shared social norms. Thus, science conforms closely to Shore’s (ibid., p. 51) notion of an instituted model. For Shore (ibid., 52), models become socially institutionalised when they are ‘objectiﬁed as publicly available forms.’ However, having originated through a process of ‘intersubjective negotiation,’ modern science as an instituted model takes on a ‘second life’ as a series of mental representations or individual processes of ‘meaning construction.’ The practice of contemporary science provides some excel lent examples of personal mental models derived from an instituted model, and where conﬂict arises between the two there also arise possibilities for innovation. This feedback effect is easier to understand if we start from the ethnographic assumption that cultural models are intersubjective rep resentations created by individuals in a social environment, rather than from the base assumption of psychology, that mental models are subjec tive representations created by individuals. In looking at the growth of science the interesting question becomes, therefore, how external, publicly available instituted models are reconstructed as cognitive models, and how the relations between relatively conventional and relatively personal mental models work for any individual (ibid., pp. 48-50).

Some of the main factors which make science so effective as a social and technical practice are the mechanisms for the establishment, shaping, and maintenance of intersubjectivity, in other words for ‘making meaning’ (Bruner 1996: 94-5). In science, meaning is more secure than in other kinds of knowledge, and it is possible to more accurately establish that the receiver knows the message which the sender is transmitting. This is what Bruner has called ‘memory externalisation’ (ibid., p. 102, following Donald), through which process ideas and ‘facts’ become independent of those individuals who innovated them, and remain as part of the ‘exogram’ because they are formally necessary for the descriptive and explanatory system to work (ibid., p. 101). In terms of recent anthropological debates on the extent to which knowledge is shared, and of methodologies and modes of representation which conﬂate aggregate with individual knowledge, it may well be that in the exogram of certain scientiﬁc discourses we come the closest we ever can to that mythical ﬁgure, the ‘ideal’ or ‘omniscient speaker-hearer’ (see e.g. Ellen 1993: 43, 126-48).

For science to work there have to be mechanisms for securing intersub jectivity, for ensuring accurate memory externalisation, for distinguishing specialist knowledge from more generalised knowledge shared by a wider category of people, all of which in effect mean maintaining the boundary between scientists and non-scientists. In all of this, language use and liter acy have historically been crucial: the determination of specialist terminolo gies and notations (e.g. anatomical Latin and Greek letters) and linguistic registers, use of metaphor and abbreviations (from memes to quarks), of international rules of practice (say of taxonomic nomenclature), method ological protocols, specialist journals and other forms of publication. But at the same time, what is de facto legitimated as technical necessity is simul taneously symbolic and social. Thus, these same devices, representations and instantiations may also evoke secrecy, mystery and therefore power, maintaining boundaries and conferring legitimacy, authority and status; or, within science, may equally serve to divide. Dunbar (1995a: 139, 142) compares, for example, the way immunologists have managed to develop a separate professional language, while physicists are forced to borrow the language of everyday (and social) experience to talk about abstract rela tions and entities which are not perceivable in any way other than through mathematics.

Beyond the oral and written forms of language, the instituted model of science is maintained by conferring authority through social markers outside of the scientiﬁc practices themselves. The most striking of these are special places of work and presentation of work to the public (e.g. laboratories and museums as temples of learning), and places of training (universities): ‘real’ science, note Chambers and Richards (1995: xiii) is generated in laboratories, research stations and universities. But science also has its rituals, rituals of election to the Royal Society or British Acad emy, the wearing of distinctive garments, and in art. For example, there is a striking difference between two portraits of Linnaeus: a contemporary likeness commissioned by Linnaeus himself as a Lapp, which we have al ready encountered (Figure 1); and a painting of Linnaeus in Thornton’s Temple of Flora, published in 1807 (Figure 5). In the latter Linnaeus has been deiﬁed, consistent with the enlightenment idea of science as the new

Figure . 5

Illustration from R.J. Thornton, Temple of Flora, Volume II, 1807,

religion. Thus, science maintains its separateness and status, and therefore effectiveness, by self-consciously appropriating the sacred and demonstrat ing again the inescapable intermingling of symbolic and the technical. This role accorded to individually-named scientists in establishing science as an instituted model might appear to be at variance with the concept of sci ence as anonomysed intersubjectivity discussed above. However, we need really to draw a distinction between the process by which scientiﬁc knowl edge is created – in socially instituted instructional contexts and through biographised authors, a process which provides a necessary social legitima tion and authority and which provides social rewards to individual scientists – and its accumulated form as a body of shared knowledge.

The centrality of intersubjectivity and the maintenance of its stability through social means are therefore crucial also because they decenter the self in cognitive terms: science is a distributed process. Thus, while most researchers conceptualise thought as a process in the individual mind, it has been recognised since the work of Vygotsky (1930, 1978) and Harré (1983) that it may ﬁrst be social, and that only in later development can we think in certain ways as individuals. It has even been suggested that the social quality of science is reﬂected in an ‘inner dialogue’ of individual scientiﬁc minds. Thus ‘the social enters into thinking in the form of culturally elaborated methods of using inferences’ (Oatley 1996: 135). But it is not simply the transferal of social thinking to the individual mind which is involved, but in addition reliance on social interaction itself to enhance cognitive processes which individual minds are by themselves ill-equipped to undertake. Thus, as Oatley (ibid., p. 137) has pointed out, humans are not good at generating examples to disconﬁrm their own theories, and science would not work with individual knowers and doers. Empirical work in support of these generalisations can be found, for example, in the work of Kevin Dunbar (1995), who found that most productive conceptual advances in molecular biology occurred in laboratories where there were weekly meeting of researchers from different specialist backgrounds, who were able to engage in analogical reasoning between different parts of the same ‘region’ of understanding. If something was difﬁcult to understand, analogies drawn from another biochemical system, another subﬁeld, or some other organism were found to be helpful; results could be interpreted from different viewpoints, while conﬁrmation bias was counteracted by one person presenting results and others seeing different interpretations.

8. The emergent cognitive features of global instituted science

Global science, therefore, as we have come to understand it, is a social process, the product of self-reﬂective and metaconceptual thought between consenting adults requiring sophisticated instruction (Carey 1996: 206), and we can only sensibly talk about its distinctive features having ﬁrst deﬁned it as an instituted process.

Since so much attention in science is paid to boundary maintenance, to effecting the conditions of intersubjectivity, to communication and to social institutionalisation, this in itself suggests that its modes of cognition are less different qualitatively from what pertains in the wider social world than they are by degree. This observation follows from the well-worn anthropo logical dictum that differences between categories will be emphasised much more where these are not reinforced by multiple perceptual discontinuities (as exempliﬁed in Douglas 1973: 113-93). In turn, such accentuated fea tures are those most likely to be used in promoting self-consciouness in the performance of certain kinds of cognitive operation which are a necessary part of sharing. None in themselves point to a crucial cognitive difference in the way science and ethnoscience work, but all, polythetically, contribute to the way in which we explicitly or implicitly maintain the boundary be tween science and ordinary thinking, communicating and doing. In the context of global science, I think we can point to six general processes which in the degree to which they are evident serve to distinguish those we call scientists from ethnoscientists.

8.1. The lexicalisation of implicit knowledge

Science is a material practice, not just about thinking. Like other forms of knowledge, it is a combination of conscious representational knowledge (usually involving reﬂection: something we are aware of acquiring and using, and often do so purposely in order to solve various technical and social prob lems) and bodily knowledge (acquired and coded as part of doing and recog nising in particular practical contexts which require sensory and motor skills that are readily transmitted trans-generationally). In traditional soci eties (see, for example, Ellen 1999), much knowledge of the ﬁrst kind (such as we ﬁnd in plant and animal nomenclatures) is encoded in language, and is therefore lexical. Where this yields regularities in how people con ceive relationships between different living kinds, it becomes classiﬁcatory knowledge. However, much knowledge, particularly of natural processes is only partially lexically expressed. Where classiﬁcatory knowledge gener ates categories with no lexical markers, but where knowledge is manifestly evident though not necessarily systematically expressed in language, we might speak of substantive knowledge. Most knowledge of the biological world is substantive in this sense and classiﬁcations can be understood as codes to access and manipulate it. But modern science is also characterised by the degree to which practical procedures are lexicalised and formalised. Whereas in many folk contexts knowledge is not explicitly formulated into a set of rules, but rather acquired through mimicry, experience and infor mal apprenticeship; in science codiﬁcation and formal instruction is much more central to learning how to do things. Thus, we have dissection manuals, ﬁeld identiﬁcation protocols, laboratory practice handbooks, and so on. As we have seen, it is institutionalisation which makes this self-conscious in structional component effective, and to recognise its signiﬁcance is not to deny the crucial role which the transfer of tacit knowledge may have had in actual scientiﬁc discoveries.

8.2. The textualisation of lexical knowledge

The view associated with Goody (1968, 1977, 1986, 1987), and more recently with Olson (1996: 141), that literacy is the prime mover of qualitative changes in our mode of thinking about the world, and that science and certain kinds of logic were an historical by-product of literacy, must be substantially qualiﬁed in the light of other work. It is recognised that different traditions of literacy (say, Arabic versus English, versus Hanunóo) appear to have different implications (Foley 1997: 417-34), that orality and literacy are completely interdependent (Finnegan 1988), that analytic arguments have more to do with marketplace rhetoric than private literacy (Lloyd 1979), that being literate in many local languages has little effect on cognitive performance, while being schooled in English or some other lingua franca associated with scientiﬁc culture can have dramatic effects (Scribner and Cole 1981). Nevertheless, there is equally little doubt that while ethno-science is possible without literacy, science as it has developed in the various scholarly traditions, was institutionalised in the West and has coalesced globally, has become dependent on the ‘linear-sentential mode’ (Bloch 1991) for its forms of organisation, range of possible sustained cognition manipulations, and on the written word for storing and transmitting knowledge.

8.3. The formalisation of textual knowledge

Lexicalisation of data and the textualisation of utterance, as well as the diagramatic and similar devices that we embed in and interpret through text, permit the routine use of formalised protocols by a community of scientists; that is certain textual devices can be employed in a systematic way to better retreive and manipulate information, and to ensure conduct of procedures in the same way everywhere. A formal description of this kind may include everything from the deﬁnitive writing-up of a new species according to acceptable taxonomic conventions, the use of anatomical Latin in a medical diagnosis, a careful drawing of the reproductive organs of a plant, precision syllogism, a Euclidean proof, to the solving of a quadratic equation. Such formal conventions may exist independent of the rest of what we understand of science, but they can make all the difference between our superﬁcial recognition of something as science rather than ethno-science. Thus, while illiterate farmers may year-on year conduct experiments regarding the utility of different landrace seeds, proper recording on paper and analysis using abstract conventions permits a greater degree of experimental control and comparison. There has in the past been some resistance to the characterisation of science as merely careful perception and description, since it seems to deny the primacy of theory, and fudges the line of separation between, say, the artwork characteristic of the great scholarly traditions (say a Tibetan herbal) and that of European science since the time of Versalius or the microscopic images of Robert Hooke and Antony van Leeuwenhoek.

In some ways comparable to textual formalisation – and certainly involving the creation of texts such as labels and catalogues – are collections of specimens, whether in museums, laboratories, botanic gardens, herbaria or in living collections of animals. The accumulation of data – whether strictly textual or in some other form – typiﬁes both scholarly and scientiﬁc ways of knowing.

8.4. Simplﬁication and abstraction

If textualisation permits more effective formalisation, then formalisation in turn allows for simpliﬁcation and other forms of second-order manipula tion. Data simpliﬁcation is an essential part of handling large amounts of data, of comparison and therefore of inference. Simpliﬁcation may be ef fected through visual depiction (consider, for example, a medical drawing compared with a photograph, or a microphotograph of a histological spec imen stained to contrast its component parts), but in general the process is quintessentially associated with the application of mathematics, whether the grouping of like values in an equation, the ordering of statistical data through intervals, or some similar procedure. Carey (1996: 206) describes, for example, how physicists translate from physical to mathematical de scription, look for regularities in the simpler mathematical version, and translate back to deﬁne what become physical laws.

8.5. The application of material culture

Not only does the creation and dissemination of texts involve its own tech nology and material culture, but the doing of science from its earliest be ginnings as such in the sixteenth century has involved the manufacture and use of instruments. Instrumentation may enhance empirical observation, in the way the microscopy of van Leeuwenhoek and Hooke revolutionised bi ological depiction, providing prostheses to extend the capacity of available human sensory organs; or they may aid the process of simpliﬁcation, by – for example – permitting the recreation of organic, physical or behavioural processes in convenient laboratory spaces; or may simulate or model such processes by analogy. In this way some cognitive practices may become concretised in material form. In addition, the technology of science has become a salient icon of the enterprise itself (the images conjured up by the retort and the test tube), while museums of science when they are not exhibiting the material data of science (plants, animals and minerals for example), comprise collections of tools used in its conduct.

8.6. The routinisation of intense criticism and argumentative reasoning

Processes 1-5 by themselves might seem to suggest that what separates science from ethno-science is simply a number of mechanistic ways of eliciting, transforming and storing data, in which cognition is situated and distributed, and brain activity consequential and displaced rather than central. But, particularly idealistic, theories of science often put more emphasis on the quality of the reasoning involved. In sections 2-5 I have provided arguments which would seem to cast doubt on this dimension, but it is nevertheless undeniable that we see in the development of modern science the foregrounding of ‘a process of intensive criticism’ (Dunbar 1995a: 31), the elevation of uncertainty to a matter of principle, a kind of ‘argumentative reasoning’ involving the systematic and self-reﬂective coordination of explicit theory and evidence (Carey 1996: 205; Kuhn 1996: 264, 275) in which investigators are able to reﬂect on their own theories as objects of cognition sufﬁciently to recognise that they might be wrong, and a methodologically explicit recognition that evidence might disconﬁrm a theory. Science, in other words, provides a set of tools – perhaps the best set we have – for choosing between rival theories. Modern science also involves cognitive operations – such as induction – which most modern people ﬁnd intrinsically difﬁcult because the default form of folk explanation involves deduction: testing a pre-existing model of the world or current hypothesis against new data (Oatley 1996: 136). For Carey (1996: 137), following Pierce, the dislodgement of this predisposition requires the ability to recognise and understand the abstract processes of inference: namely abduction, deduction and induction.

* * * * *

Processes 1-6 exist not as intrinsic characteristics of how individual scientists work, but rather are distributed within common institutional practices, in social entitities, and in artifacts (such as books). Combined, they permit more effective sharing, codiﬁcation and transmission of information. They also allow for increasingly stronger inference, logical rigour of description, of testing and of experimentation (Dunbar 1995a: 102); for coordination of theory and evidence (Kuhn 1996: 264) and for inductive reasoning skills, especially those involving causal inferences in a multivariable framework (ibid., p. 275). But because science is detached from any one speciﬁc local context, these innovations have, most importantly, permitted the aggregation of knowledge from many localities and enabled trans-local generalisation and comparison through globalisation (Hunn 1993: 13-15). Thus, in a sense it was only when Linnaeus went beyond Lappland and applied his taxonomic criteria to ‘coconuts and cardamoms’ that arguably he was able to transcend what traditional scholarly knowledge systems had achieved (Koerner 1999). Lexicalisation, textualisation, formalisation and abstraction have, therefore, allowed scientiﬁc knowledge to escape its local roots, but trans-localism is not in itself unknown in local knowledge systems. Indeed, there is a uniquely human capacity for acquired biological knowledge to diffuse independently of what can be experienced in local habitats. Thus, people may have concepts for snakes, elephants, lions, even dragons, even though they have never seen one. Scientiﬁc biology is, in one sense, an extreme development of such an ‘intuitive biology,’ augmented by the possibilities offered by effective cultural transmission, since the capacity to generalise and hypothesise is grounded in the way science aggregates knowledge of species and ecologies beyond what a scientist might have local ﬁrst-hand experience of as a non-scientist. The corollary is that writing it down, the very thing which makes it more portable and permanent, also changes some of its fundamental properties, all of which reinforce dislocation. As several generations of development advisors have discovered, general comparative truths may often be at variance with the truth in a particular locality.

9. Conclusion

This paper is part of a special journal issue devoted to the cognitive anthropology of science, yet I am claiming that science can never be wholly deﬁned in cognitive terms. Science is not simply a way of thinking, but involves ways of doing (technical practices) and ways of representing (social practices), both to other scientists and to non-scientists. I also argue that although we need to distinguish between a general human aptitude for ‘science’ which is simultaneously pan-cultural and cognitive, and the socially constructed global science which has emerged since the seventeenth century, we must recognise the continuities between the two. Indeed, we will never understand modern global science, cognitively or otherwise, unless we look at it in the context of ordinary forms of reasoning and those specialist forms we call ethno-sciences. We need an integrated theory of science which incorporates them both, since continually re drawing the Rubicon will not work. Indeed, any approach which begins by reinventing a new version of the great cognitive divide, by opposing two meta-categories – science and traditional knowledge – is doomed to failure.

Because what counts as ‘indigenous knowledge’ is so protean and ex tensive, to claim that it is comparable or not comparable to science is misleading, a diversion from the real issue. It is not, in my view, par ticularly helpful to deﬁne indigenous knowledge as everything science is not, or in terms of tables of discrete opposing characteristics. While it may be useful as a beginning to characterise it as local, performative, or ganic, holistic, intuitive, socially constrained, practical, incremental, and egalitarian; in contrast to modern global science which is rationalistic, reductionist, theoretical, generalisable, objectively veriﬁable, abstract, non contextual, universal, based on cause-and-effect, and imperialistic; given the diversity of traditional knowledges, these sets of opposites do not take us very far, and exceptions and additions can always be found. The in digenous ethno-sciences can include propositional statements, experimental evidence, mathematical thinking, and literacy. And while science may chal lenge the credibility of ‘indigenous knowledge’ as science, the diversity of all that purports to be, or is by default treated as, indigenous knowledge makes any kind of generalisation hazardous. Professional scientists and others have coped with this diversity and ﬂuidity (as we have seen) by es sentializing, freezing, commodifying it, breaking it down into conveniently absorbable, observable and representational modules. But what we have to deal with as a result of such transformations and manipulations, even in some of the anthropological descriptions of indigenous knowledge, is something different, almost chimerical. I have – therefore – taken the view that the category of indigenous knowledge (which we may take to com prise the plethora of ethno-sciences) arises by default, as a counterpoint to what we call science, and is constructed largely on the basis of the kind of knowledge it is not. And as I have argued elsewhere, indigenous knowledge of a very important kind is ever present and constantly being reinvented to cope with new circumstances, existing as kind of buffer (Ellen and Har ris 2000), an intuitive, informal, knowledge which endures and inheres at every technical and social interface, and which enables technologies based on more formal abstract knowledge to work (c.f. Fischer 2004).

But it is not just the antithesis – indigenous knowledge, the ethno sciences – which defy placement in a single meta-category, but the thesis also. Science itself exhibits disunity as a theorised set of practices, and the vested interest which scientists understandably have in maintaining a clear boundary between science and non-science is thus ever challenging. Unhelpfully, methods and criteria which might assist in precisely stating where the boundary lies tend to change from time to time, and from subject to subject (Barnes, Bloor and Henry 1996: 140), and as these boundaries change so certain theories become more marginalised, until they slip over the edge into non-acceptability. So, what is it that makes evolutionary psychology science, and cold fusion, scientology or psycho-analysis not? But despite the polythetic character of science, the fuzziness of its boundaries and the complexity of its historic interconnections with other knowledge systems, and despite the formidable difﬁculties of representing the totality of relations between kinds of knowledge, this does not mean that it is therefore impossible to describe or compare the cognitive, representational and social contours of different knowledge systems.

Within anthropology too, strong views exist as to what is scientiﬁc and what is not, as the ongoing factionalism between post-modernism and ‘scientiﬁc’ anthropology attests (Kuznar 1997). Indeed, in anthropology the debate may be even more intense because it is fratricidal, and because what separates one side from the other is sometimes philosophical sentiment as much as technical practice. Those who have tried to make sense of this grey area, and show how the boundaries between science and non-science uncomfortably correlate with convenient ideological and social currents, are sometimes accused of being engaged in some kind of ‘pseudo-science’ themselves (Laudan 1981), or more straightforwardly of ‘relativism.’ While some approaches to understanding how science works as theory and practice do make it sound as no more than pernicious ideology, text, ‘rhetoric, persuasion and the pursuit of power’ (Wolpert 1992: 117), we are always dealing with degrees of truth and plausibility. Since science is dependent on metaphors and cultural constructs to determine shared knowledge, and is conducted by sentient persons with values and social differences in intersubjective space, it will always only be sufﬁcientyl precise, but for all that a remarkably reliable model of the world.

AGRAWAL, A.

REFERENCES

1995 Indigenous and scientiﬁc knowledge: some critical comments. Indigenous Knowledge and Development Monitor 3(3): 5. Elaborated as: (1995) Dismantling the divide between indigenous and scientiﬁc knowledge. Development and Change 26: 413 439.

AKIMICHI, T. 1996 “Image and reality at sea: ﬁsh and cognitive mapping in Carolinean navigational knowledge.” In R. Ellen & K. Fukui (Eds.), Redeﬁning nature: ecology, culture and domestication (pp. 493-514). Oxford: Berg.

ANDERSON, E.N.

2000 Maya knowledge and ‘science wars.’ Journal of Ethnobiology 20(2): 129-158. ANON.

1999 Caution: traditional knowledge: principles of merit need to be spelt out in distinguishing valuable knowledge from myth. Nature 401: 623 (14 October, issue 6754). 1999a ICSU seeks to classify ‘traditional knowledge.’ Nature 401: 631 (14 October,

issue 6754).

ATRAN, S.

1990 Cognitive foundations of natural history: toward an anthropology of science. Cambridge:

Cambridge University Press.

1994 “Core domains versus scientiﬁc theories.” In L. Hirschfeld & S. Gelman (Eds.), Mapping the mind: domain specﬁiciyt in cognition and culture (pp. 316-340). New York: Cambridge University Press. 1998 Folk biology and the anthropology of science: cognitive universals and cultural

particulars. Behavioural and Brain Sciences 21: 547-609.

1996 “Modes of thinking about living kinds: science, symbolism, and common sense.” In D.R. Olson & N. Torrance (Eds.), Modes of thought: explorations in culture and cognition (pp. 216-260). Cambridge: Cambridge University Press.

AUNGER, R. (ED.)

2000 Darwinizing culture: the status of memetics as a science. Oxford: Oxford University

Press.

BARNES, B., B. BLOOR & J. HENRY

1996 Scientﬁic knowledge: a sociological analysis. London: Athlone. BATES, D.

1995 Knowledge and the scholarly medical traditions. Cambridge: Cambridge University

Press.

BLOCH, M.

1991 Language, anthropology and cognitive science. Man 26(2): 183-198. BOSTER, J.

1996 “Human cognition as a product and agent of evolution.” In R.F. Ellen & K. Fukui (Eds.), Redeﬁning nature: culture, ecology and domestication (pp. 269-289). Oxford: Berg.

BOURDIEU, P.

1977 Outline of a theory of practice (Cambridge Studies in Social Anthropology 16) (trans.

Richard Nice). Cambridge: Cambridge University Press. 1990 The logic of practice. Cambridge: Polity.

BRUNER, J.

1996 “Frames for thinking: ways of making meaning.” In D.R. Olson & N. Torrance (Eds.), Modes of thought: explorations in culture and cognition (pp. 93-105). Cambridge: Cambridge University Press.

CAREY, S.

1996 “Cognitive domains as modes of thought.” In D.R. Olson & N. Torrance (Eds.), Modes of thought: explorations in culture and cognition (pp. 189-215). Cambridge: Cambridge University Press.

CHAMBERS, R. & P. RICHARDS

1995 “Preface.” In D.M. Warren, L.J. Slikkerveer & D. Brokensha (Eds.), The cultural dimension of development: indigenous knowledge systems (pp. xiii-xiv). London: Intermediate Technology Publications.

CHAPLIN, M.

2004 Water Structure and behaviour. CLEVELAND, D.A. & D. SOLERI (EDS.)

2002 Farmers, scientists and plant breeding: integrating knowledge and practice. Wallingford:

CABI Publishing. CLIFFORD, J.

1986 “Introduction: partial truths.” In J. Clifford & G.E. Marcus (Eds.), Writing culture: the poetics and politics of ethnography (pp. 1-26). Berkeley: University of California Press.

COLBY, B., J. FERNANDEZ & D. KRONENFELD

1980 Toward a convergence of cognitive and symbolic anthropology. American Ethnol

ogist 8: 422-450.

D’ANDRADE, R.G.

1980 “Cultural cognition.” In M.I. Posner (Ed.), Foundations of cognitive science (pp. 795

830). Cambridge, Mass.: MIT Press. DIAMOND, J.M.

1986 The environmentalist myth. Nature 324: 19-20. DOBBS, B.J.T.

1975 The foundations of Newton’s alchemy, or ‘The hunting of the greene lyon.’ Cambridge:

Cambridge University Press. DOUGLAS, M. (ED.).

1973 Rules and meanings: the anthropology of everyday knowledge. Harmondsworth: Penguin. DUNBAR, K.

1995 “How scientists really reason: scientiﬁc reasoning in real-world laboratories.” In R.J. Steinberg & J. Davidson (Eds.), Mechanisms of insight (pp. 365-395). Cambridge, Mass.: MIT Press.

DUNBAR, R.I.M.

1995a The trouble with science. London: Faber and Faber. ELLEN, R.F.

1979 “Introductory essay.” In R.F. Ellen & D.A. Reason (Eds.), Classﬁications in their

social context (pp. 1-32). London: Academic Press.

1993 The cultural relations of classﬁication: an analysis of Nuaulu animal categories from central

Seram. (Cambridge Studies in Social and Cultural Anthropology 91). Cambridge: Cambridge University Press. 1996 “Introduction.” In R.F. Ellen & K. Fukui (Eds.), Redeﬁning nature: culture, ecology

and domestication (pp. 1-36). Oxford: Berg.

1999 “Modes of subsistence and ethnobiological knowledge: between extraction and cultivation in southeast Asia.” In D.L. Medin & S. Atran (Eds.), Folkbiology (pp. 91-117). Cambridge, Mass.: MIT Press. 2003 “Variation and uniformity in the construction of biological knowledge across cultures.” In H. Selin (Ed.), Nature across cultures: views of nature and the environment in non-western cultures (pp. 47-74). Amsterdam: Kluwer.

ELLEN, R. & H. HARRIS

2000 “Introduction.” In R.F. Ellen, P. Parkes & A. Bicker (Eds.), Indigenous environ

mental knowledge and its transformations: critical anthropological perspectives (pp. 1-33).

Amsterdam: Harwood (Routledge). FINNEGAN, R.

1988 Literacy and orality. Oxford: Blackwell. FISCHER, M.D.

2004 “Powerful knowledge: applications in a cultural context.” In A. Bicker (Ed.),

Development and local knowledge: new approaches to issues in natural resource management, conservation and agriculture (pp. 19-30). London: Routledge.

FOLEY, W.A.

1997 Anthropological linguistics: an introduction. Oxford: Blackwell. FOUCAULT, M.

1970 The order of things: an archaeology of the human sciences. London: Tavistock. FRAKE, C.O.

1983 Did literacy cause the great cognitive divide? American Ethnologist 10: 368-371. 1985 Cognitive maps of time and tide among medieval seafarers. Man 20(2): 254-270.

GELL, A.

1985 How to read a map: remarks on the practical logic of navigation. Man 20(2):

271-286.

GELMAN, R.

1990 First principles organize attention to and learning about relevant data: number

and the animate-inanimate distinction as examples. Cognitive Science 14: 79-106. GLADWIN, T.

1970 East is a big bird: navigation and logic on Pulawat Atoll. Cambridge, Mass.: Harvard

University Press.

GOODENOUGH, W.H.

1990 Evolution of the human capacity for beliefs. American Anthropologist 92: 597-612.

GOODY, J.

1968 “Introduction.” In J. Goody (Ed.), Literacy in traditional societies (pp. 1-26). Cam

bridge: Cambridge University Press.

1977 The domestication of the savage mind. Cambridge: Cambridge University Press. 1986 The logic of writing and the organisation of society (Studies in Literacy, Family, Culture

and the State). Cambridge: Cambridge University Press.

1987 The interface between the written and the oral. Cambridge: Cambridge University Press. GROVE, R.

1996 Indigenous knowledge and the signiﬁcance of south-west India for Portuguese and Dutch constructions of tropical nature. Modern Asian Studies 30(1): 121-143. HACKING, I.

1999 The social construction of what? Cambridge: Harvard University Press. HARRÉ, R. 1983 Personal being: a theory for individual psychology. Oxford: Blackwell. HOOKE, R.

1961 [1665] Micrographia, or some physiological descriptions of minute bodies made by magnifying

glasses with observations and enquiries thereupon. New York: Dover.

HUNN, E.

1993 “What is traditional ecological knowledge?” In N. Williams & G. Baines

(Eds.), Traditional ecological knowledge: wisdom for sustainable development (pp. 13-15).

Canberra: Centre for Resource and Environmental Studies, ANU. ISKANDAR, J. & R. ELLEN

2000 The contribution of Paraserianthes (Albizia) falcataria to sustainable swidden man

agement practices among the Baduy of West Java. Human Ecology 28(1): 1-17. ICSU (INTERNATIONAL COUNCIL FOR SCIENCE)

2002 D. Nakashima and D. Elias (Eds.), Science, traditional knowledge and sustainable

development. ICSU Series on Science for Sustainable Development, No. 4. JOHANNES, R.E.

1987 Primitive myth. Nature 325: 478. JOHNS, T.

1990 With bitter herbs they shall eat. Tucson: Arizona University Press. JOHNSON-LAIRD, P.

1982 Mental models. Cambridge: Cambridge University Press. KEIL, F.C.

1992 “The origins of an autonomous biology.” In M.R. Gunnar & M. Maratsos (Eds.), Modularity and constraints in language and cognition. The Minnesota Symposium on Child Psychology 25: 103-137. 1994 “The birth and nurturance of concepts by domains: the origins of concepts of living things.” In L. Hirschfeld & S. Gelman (Eds.), Mapping the mind: domain specﬁiciyt in cognition and culture (pp. 234-254). New York: Cambridge University Press.

KOERNER, L.

1999 Linnaeus: nature and nation. Cambridge, Mass. & London: Harvard University

Press.

KUHN, D.

1996 “Is good thinking scientiﬁc thinking?” In D.R. Olson & N. Torrance (Eds.), Modes of thought: explorations in culture and cognition (pp. 261-281). Cambridge: Cambridge University Press.

KUZNAR, L.A.

1997 Reclaiming a scientﬁic anthropology. Altamira Press. LAUDAN, L.

1981 The pseudo-science of science? Philosophy of the Social Sciences 11: 173-198. LIEBERMAN, P.

1984 The biology and evolution of language. Cambridge, Mass.: Harvard University Press. LLOYD, G.E.R.

1979 Magic, reason and experience. Cambridge: Cambridge University Press. 1990 Demystifying mentalities. Cambridge: Cambridge University Press. 1991 Methods and problems in Greek science. Cambridge: Cambridge University Press.

MITHEN, S.

1996 The prehistory of the mind: a search for the origins of art, religion and science. London:

Thames and Hudson.

MOERMAN, D.E., R.W. PEMBERTON, D. KIEFER & B. BERLIN

1999 A comparative analysis of ﬁve medicinal ﬂoras. Journal of Ethnobiology 19(1): 49

67.

NADER, L. (ED.)

1996 Naked science: anthropological enquiry into boundaries, power and knowledge. Routledge:

New York.

OATLEY, K.G.

1977 “Inference, navigation and cognitive maps.” In P.N. Johnson-Laird & P.C. Wason (Eds.), Thinking: readings in cognitive science (pp. 537-547). Cambridge: Cambridge University Press. 1996 “Inference in narrative and science.” In D.R. Olson & N. Torrance (Eds.), Modes of thought: explorations in culture and cognition (pp. 123-140). Cambridge: Cambridge University Press.

OHNUKI-TIERNEY, E.

1981 Phases in human perception/cognition/symbolization processes: cognitive an

thropology and symbolic classiﬁcation. American Ethnologist 8(2): 451-467. OLSON, D.R.

1994 The world on paper: the conceptual and cognitive implications of writing and reading.

Cambridge: Cambridge University Press.

1996 “Literacy, consciousness of language, and modes of thought.” In D.R. Olson & N. Torrance (Eds.), Modes of thought: explorations in culture and cognition (pp. 141-151). Cambridge: Cambridge University Press.

PFAFFENBERGER, B.

1992 Social Anthropology of technology. Annual Review of Anthropology 21: 491-516. RICHARDS, P.

1985 Indigenous agricultural revolution: ecology and food-crop farming in West Africa. London:

Hutchinson.

1993 “Cultivation: knowledge or performance?” In M. Hobart (Ed.), An anthropological

critique of development (pp. 61-78). London: Routledge. SCRIBNER, S. & M. COLE

1981 The psychology of literacy. Cambridge, Mass: Harvard University Press. SHORE, B.

1996 Culture in mind: cognition, culture and the problem of meaning. Oxford: Oxford University

Press.

SEMALI, L. & J.L. KINCHELOE

1999 “What is indigenous knowledge and why should we study it?” In L. Semali & J.L. Kincheloe (Eds.), What is indigenous knowledge: voices from the Academy (pp. 3-57). New York and London: Falmer Press.

SILLITOE, P.

2002 “Globalizing indigenous knowledge.” In P. Sillitoe, A. Bicker & J. Pottier (Eds.), Participating in development: approaches to indigenous knowledge (pp. 108-138). (Association of Social Anthropologists, Monograph 39). London and New York: Routledge.

SINGER, C.

1959 A history of biology to about the year 1900: a general introduction to the study of living things.

London & New York: Abelard-Schuman. SPERBER, D.

1994 “The modularity of thought and the epidemiology of representations.” In L. Hirschfeld & S. Gelman (Eds.), Mapping the mind: domain specﬁiciyt in cognition and culture (pp. 39-67). New York: Cambridge University Press.

TURNBULL, D.

2000 Masons, tricksters and cartographers: comparative studies in the sociology of scientﬁic and

indigenous knowledge. Amsterdam: Harwood Academic. VYGOTSKY, L.S.

1978 [1930] “Tool and symbol in child development.” In M. Cole, V. John-Steiner, S. Scribner & E. Souberman (Eds.), Mind in society: the development of higher mental processes (pp. 19-30). Cambridge, Mass.: Harvard University Press. WALKER, D.H., P.J. THORNE, F.L. SINCLAIR, B. THAPA, C.D. WOLOD & D.B.

SUBBA

1999 A systems approach to comparing indigenous and scientiﬁc knowledge: consis tency and discriminatory power of indigenous and laboratory assessment of the nutritive value of tree fodder. Agricultural Systems 62: 87-103.

WELLMAN, H.M.

1990 The child’s theory of mind. Cambridge, Mass.: MIT Press. DE WIT, H.C.D.

1959 (Ed.). Rumphius memorial volume. Baarn: Hollandia N.V. WOLPERT, L.

1992 The unnatural nature of science. London: Faber and Faber.