Revisiting the Archimedean structuralist

(an invited lecture, delivered at Asiatic Society, Kolkata, on Lévi-Strauss’s 100th birthday)

As we prepare, in this diversely anti-discriminatory period of human history, to reconsider our uses of structural analysis in the human and social sciences, we are obliged to revisit the moment of encounter – the moment at which structural linguistics, transmitted by Roman Jakobson, gave rise to a structural anthropology enterprise coterminous with Claude Lévi-Strauss. In particular, we are compelled to ask if this scholar from France, a country whose moral and mental exertions have traditionally been focused on freedom, helps us to identify a certain Archimedean point at which structural investigators take their stand and from which their free inquiry can attain a truly neutral view of human structures.

The range and depth of Lévi-Strauss’s writings make it necessary to choose with care the issues around which a focused discussion is possible. In the present exercise, we may usefully begin by considering a question he directed at his linguist colleagues at a 1952 Conference of Anthropologists and Linguists at Bloomington, a question that seems not to have been taken up by linguists.

Lévi-Strauss begins by invoking Kroeber: “As a matter of fact, Kroeber a long time ago noticed that no kinship systems are so completely different from each other as the Indo-European on the one hand and the Chinese on the other” (1968: 78-9).

He then goes on to formulate his question as follows: “If we try to interpret this picture, what do we find? We find that in the Indo-European case we have a very simple structure (marriage rules), but that the elements (social organization) which must be arranged in this structure are numerous and complicated, whereas in the Sino-Tibetan case the opposite prevails. We have a very complicated structure (marriage rules), with two different sets of rules, and the elements (social organization) are few. And to the separation between the structure and the elements correspond, on the level of terminology – which is a linguistic level – antithetic features as to the framework (subjective versus objective) and to the terms themselves (numerous versus few). Now it seems to me that if we formulate the situation in these terms, it is at least possible to start a useful discussion with the linguists. While I was making this chart, I could not but remember what R. Jakobson said at yesterday’s session about the structure of the Indo-European language: a great discrepancy between form and substance, a great many irregularities in relation to the rules, and considerable freedom regarding the choice of means to express the same idea. Are not all of these traits similar to those we have singled out with respect to social structure?” (1968: 79).

In the run-up to this formulation, Lévi-Strauss offers a tabulation (1968: 77) to clarify certain contrasts between the two cultural systems:

	Indo-European Area	Sino-Tibetan Area
Marriage Rules	Circular systems, either resulting directly from explicit rules or indirectly from the fact that the choice of a mate is left to probability	Circular systems, present in juxtaposition with systems of symmetrical exchange
Social Organization	Numerous social units, with a complex structure (extended family type)	Few social units, with a simple structure (clan or lineage type)
Kinship System	(1)               subjective (2)               few terms	(1)               objective (2)               numerous terms

In the present exercise, our goal is to take Lévi-Strauss’s long neglected question as a point of departure for a discussion that must begin at the methodological level. Few scholars have devoted as much energy as Lévi-Strauss has to showing that what we can learn from the recorded history of settled civilizations that have been able to preserve considerable archival material contrasts with, and also exhibits certain curious complementarities with, what we learn from the panoramic data base – spatially broad though temporally shallow – of the non-literate cultures ethnographically described over the last couple of centuries. In view of this emphasis, it is striking that Lévi-Strauss should have brought a combination of historically and ethnographically obtained input to bear on this question that he expects linguist colleagues to help anthropologists to articulate more rigorously and answer.

It is also striking that, in the run-up to this question, Lévi-Strauss should have voiced an unusual type of dissatisfaction with the language-culture correlations proposed by Whorf. He writes: “Whorf has tried to establish a correlation between certain linguistic structures and certain cultural structures. Why is it that the approach is unsatisfactory? It is, it seems to me, because the linguistic level as he considers it is the result of a rather sophisticated analysis – he is not at all trying to correlate an empirical impression of the language, but, rather, the result of true linguistic work (I don’t know if this linguistic work is satisfactory from the point of view of the linguists, I’m just assuming it for the sake of argument) – what he is trying to correlate with this linguistic structure is a crude, superficial, empirical view of the culture itself. So he is really trying to correlate things which belong to entirely different levels” (1968: 73).

There has of course never been any dearth of critics of linguistic relativity research in general, and of Whorf’s hypotheses about Hopi and Standard Average European in particular. But most critics accept Whorf’s question as a valid point of departure. They take the position that he fails to provide adequate evidence for his answers.

But Lévi-Strauss’s argument was directed against the form of Whorf’s question itself. Lévi-Strauss maintained that Whorf was trying to compare the structured results of linguistic analysis with impressionistic ideas about cultural difference. By proposing a new question at the linguistics-anthropology interface that focuses on a contrast – or a set of contrasts – between the Chinese and Indo-European systems, Lévi-Strauss was hoping to initiate a methodologically rigorous continuation of the interdisciplinary conversation. As he envisaged it, such a continuation would juxtapose rigorous products of structural analysis in linguistics and in anthropology so that we could ascertain the extent of correlation between linguistic patterns and cultural patterns: “I would say that between culture and language there cannot be no relations at all, and there cannot be 100 per cent correlation either. Both situations are impossible to conceive. If there were no relations at all, that would lead us to assume that the human mind is a kind of jumble – that there is no connection at all between what the mind is doing on one level and what the mind is doing on another level. But, on the other hand, if the correlation were 100 per cent, then certainly we should know about it and we should not be here to discuss whether it exists or not. So the conclusion which seems to me the most likely is that some kind of correlation exists between certain things on certain levels, and our main task is to determine what these things are and what these levels are. This can be done only through a close cooperation between linguists and anthropologists.” (1968: 79).

Can we in fact inherit and resume such a project today?

Does the Chinese and Indo-European material that Lévi-Strauss focuses on help us to bring the old but now flagging conversation between linguistics and anthropology back to life?

In linguistics, a great deal has changed. Generative grammar characterizes the species-specific and species-uniform linguistic ability of humans in terms of a biological endowment that precedes and grounds the specifications of this or that particular language spoken in this or that speech community. It is left to the Saussure-arbitrary residual specifications of particular languages – including lexical and functional vocabulary, inflections, and certain intricate consequences of inflections – to make sense of linguistic diversity, while the biologically unitary foundation articulates these particularities with the unity of human language.

Formal linguistics, so reconfigured, cannot inherit the Jakobsonian descriptions that Lévi-Strauss was coming from. Linguists today find it important to remove an unclarity in the formulation of Lévi-Strauss’s exploratory question. He had left tentatively open the scope of his inquiry. It is possible to interpret his formulations at the time-depth of reconstructed Proto-Indo-European or Proto-Chinese, or over the spatial breadth of Standard Average Indo-European and Standard Average Chinese today, on the tacit assumption that historical inheritance patterns tend to preserve the most pervasive features of a linguistic or cultural configuration. However, for a present-day generative grammarian, there just is no Standard Average Indo-European, not even as a rough and ready working notion. The typological range represented in the Indo-European language family today is vast. Generalizations concerning rigid or free constituent order or the overt marking of nominal definiteness or the omissibility of contextually evident object pronouns cannot be sustained over the entire area from Icelandic to Asamiya. Can we then retrieve any viable question at all from Lévi-Strauss’s text that today’s generative linguists can hope to formulate in dialogue with anthropologists?

Given the unavailability of a consensual linguistic palaeontology for Proto-Indo-European in contemporary thinking, and the unavailability of a “Standard Average Indo-European” notion, one is left with the option of considering, say, the Slavic subfamily of languages and cultures that Jakobson may have had in mind when he made the remark that led to Lévi-Strauss’s formulation. One would then seek to compare Slavic and, say, just the “Chinese language” (called “Mandarin” in one region of the literature). It is possible to envisage a synchronic comparative exercise at that level with some optimism about its fruitfulness.

There is a catch, though. China and the Slavic-speaking societies have been through a socialist trajectory. Their fiction retains the historical signature of the subtleties of kinship in and before the nineteenth century and thus fiction readers in these literate societies are aware of the kinship systems at the level of verbal comprehension to this day. But we need to work out, methodologically, what it means to take such literary historical awareness on board when we are trying to make sense of ethnographic realities. Where is the authentic present-day culture of these societies to be located in such an exercise? Are we supposed to bracket out the written thread of the past?

In order to ask this question less abstractly, we may usefully look at the way Lévi-Strauss himself modulates the more natural and the more cultural phenomena within culture in his analytical work. Here is an extended quote from Lévi-Strauss 1973 that may help us to get our bearings:

“The myths about the origin of cooking relate to a physiology of the marriage relationship, the harmonious functioning of which is symbolized by the practice of the art of cooking, whereas, on the acoustic and cosmological levels, charivari and eclipses refer to a social and cosmic pathology which reverses, in another register, the meaning of the message conveyed by the introduction of cooking. Symmetrically, the myths about the circumstances surrounding cooking elaborate a pathology of the marriage relationship, the embryo of which is symbolically concealed in culinary and meteorological physiology: this is because, just as the marriage relationship is constantly threatened ‘along its frontiers’ – on the side of nature by the physical attractiveness of the seducer, and on the side of culture by the risk of intrigues between affines living under the same roof – cooking, too, because of the availability of honey or the discovery of the use of tobacco, is in danger of moving entirely in the direction of nature or of culture, although hypothetically it should represent their union.

“The pathological condition of cooking is not simply linked with the objective presence of certain types of food. It is also a function of the alternation of the seasons which, by being characterized, as they are, sometimes by abundance and sometimes by shortages, allow culture to assert itself, or force mankind to move temporarily closer to the state of nature. Consequently, whereas, in one case, culinary physiology is reversed so as to become a cosmic pathology, in the other case, culinary pathology looks for its origin and objective basis in cosmic physiology, that seasonal periodicity which is distinguished by its regularity and is part of the given nature of things, unlike eclipses which, at least according to the native way of looking at them, are non-periodic accidents.

“It would have been impossible to unravel the complexities of this problem if we had broached it simultaneously at all levels; in other words, if, like someone deciphering a text on the basis of a multilingual inscription, we had not understood that the myths transmit the same message with the help of several codes, the clues of which are culinary – that is techno-economic – acoustic, sociological and cosmological. However, these codes are not strictly equivalent to each other, and the myths do not put them all on the same footing. The operational value of one of them is superior to that of the others, since the acoustic code provides a common language into which the messages of the techno-economic, sociological and cosmological codes can be translated. I showed in The Raw and the Cooked that cooking implies silence and anti-cooking din, and that this was the case for all the forms that might be assumed by the contrast between a mediated relationship and a non-mediated relationship, independently of the conjunctive or disjunctive character of the latter. The analyses contained in the present book confirm this finding. While the myths about the origin of cooking establish a simple contrast between silence and noise, those concerned with the substances surrounding cooking go more deeply into the contrast and analyze it by distinguishing between several of its modalities. This being so, we are no longer dealing with din, pure and simple, but with contrasts within the category of noise, such as those between continuous and discontinuous noise, modulated or non-modulated noise, and linguistic or non-linguistic behaviour. As the myths gradually widen and specify the category of cooking, which was originally defined in terms of presence or absence, they widen or specify the fundamental contrast between silence and noise, and introduce between these two opposite poles a series of intermediary concepts which mark out a frontier that I have done no more than reconnoitre, taking great care not to cross it in either direction, so as not to find myself venturing into two foreign fields: the philosophy of language and musical organology” (1973: 470-72).

We need to read this passage from Lévi-Strauss with our key question in mind. Recall that our main point in the present exercise is “to ask if this scholar from France, a country whose moral and mental exertions have traditionally been focused on freedom, helps us to identify a certain Archimedean point at which structural investigators take their stand and from which their free inquiry can attain a truly neutral view of human structures.” Thus, we are interested in the intersection between considerations of human freedom – in a world that wishes to uphold human rights and eliminate all forms of discrimination – and the application at the language-culture interface of procedures of structural analysis as these procedures were understood by Lévi-Strauss, the first major anthropologist to take on board the cognitive seriousness of the linguistic and cultural systems of the so-called primitive societies.

From this standpoint, what we are bound to find especially striking in the passage just quoted is Lévi-Strauss’s willingness to characterize the relations between the more natural and the more cultural entities in terms of the claim that “the acoustic code provides a common language into which the messages of the techno-economic, sociological and cosmological codes can be translated.” He goes so far as to say “that cooking implies silence and anti-cooking din”. It must be his explicit reticence to venture into the field of the philosophy of language that deters him from co-articulating this insight with a formal theory that situates speech, silence and writing in relation to the architecture of language as an interface between nature and culture.

Just as we must update our linguistics to be able to inherit Lévi-Strauss’s question at the language-culture interface, we are likewise obliged to augment our understanding of the cultural dimension beyond the explicit terminus that Lévi-Strauss himself had been able to reach in his written work. It seems reasonable to read Derrida (1974) as providing such an update, and as offering an affirmative answer to our question about whether a reconsideration of Chinese and Slavic should take the literature-nourished consciousness of present-day readers of fiction on board.

It is possible not to see at first glance just how making this move is helpful at the level of advancing our understanding of scholarly issues that interface with human freedom. Let us elaborate. On the face of it, Derrida (1974) simply offers a critique of some early remarks by Lévi-Strauss about the role of writing in culture. But a close reading shows that Derrida brings out and develops a deeper skepticism about the closure of semiotic codes that precisely Lévi-Strauss’s later work has articulated with other issues in the social sciences. To see this, it is best to move directly into a closer reading of Derrida’s theses about writing. The moves we are about to make are to be read as evidence in favour of our main proposal. We may need to reiterate the proposal: a serious comparison of linguistic and cultural indicators in the Chinese and the Slavic populations does have to take on board (“culturalistically”, and cannot afford to elide on “naturalistic” methodological grounds) the presence of fiction in the textual knowledge base that the literate adult public of these regions never leave home without.

The moves we shall now make constitute an interdisciplinary cross-fertilizing exercise. It is suggested here that certain core concerns of generative grammar and of one version of postmodernist theory – deconstructionism – can be brought to bear on each other with useful results. If the reasoning offered here is on the right track, it also has consequences for the investigation – in cognitive science and other relevant contexts – of just where language is located in human biology.

We begin the argument at generative grammar, whose classical insights have to do with the infinity of the syntactic system. This infinity is formally associated with embedding. (We prescind here from the issue of whether there may be an exception or two to the ubiquity of embedding in human languages.) For our purposes we need to distinguish obligatory and optional embedding – two ways in which a guest sentence may be “embedded” in a host or “matrix” sentence.

In obligatory embedding cases, the embedded clause could not possibly have stood alone as a full, unembedded sentence. For instance, “We hope to win” is a case of obligatory embedding, where “to win” – a construction built around the verb “win” and its silent subject (coreferential to “we”) and thus technically a clausal structure in certain approaches to syntax – could not have stood alone as a full sentence.

In the second, optional kind of embedding, we find an embedded sentence that could have functioned as a standalone full sentence instead of being embedded. Thus, in the example “We hope that we shall win”, what is embedded in the matrix “We hope (..)” is the sentence “that we shall win”. When one removes the word “that”, which is an embedder or “complementizer” device, the result “We shall win” can stand alone as a sentence.

Embedding can in principle be extended beyond just one matrix and one embedded clause. Both types of embedding permit extension. A series of five obligatory embeddings, for instance, yields (a) “John expects Mary to believe Bill to be quite likely to keep hoping to try eventually to win”. However, extended obligatory embedding rapidly lapses into obscurity. In contrast, one can preserve naturalness over long stretches of optional embedding.

For instance, the following sentence (b), despite its length, turns out to be easier to follow than example (a), of which (b) is a sort of translation: (b) “John expects that Mary will believe that it is quite likely that Bill has not yet lost the hope that eventually he will make an effort that will lead him to victory”. The extended optional embedding example (b) is noticeably clearer than the extended obligatory embedding example (a).

What is involved here? One possibly pertinent difference has to do with clausal perspective. Each embedded sentence in (b) seems to adopt a viewpoint and to construct a view. One is not dealing with a bare expectation, belief, likelihood, hope, effort, or victory. Instead, the listener is responding to facts visualized from a concrete standpoint each time – that “John expects”, that “Mary believes”, that “Bill hopes”, and even that “it is likely”, a standpoint whose coordinates we refrain from commenting on in this intervention.

While embedding as such is what technically underwrites infinity, the factor directly associated with the real indefinite stretchability of sentences is clearly optional embedding. We conjecture that multiple perspective – a major concomitant factor – also plays a role. The technical details of the theory of embedding built around these assumptions (Dasgupta 2008; Dasgupta 2009) are not germane to the present discussion. Nor is the possibility that invoking a perspective associated with the name of a person potentially conjures up that person’s face and may, in certain experimental settings, turn out to be associated with the human ability to record and recall faces, though such a line of inquiry is also of interest in the context of cognitive science.

Our project here is to show the mutual relevance of certain core concerns of generative grammar and of deconstructionism. Enough has been said for us to take the next step in our argument; we mention alternative paths we choose not to take only to make it clear that we have no difficulty with colleagues preferring to focus on other approaches to these and related topics.

We juxtapose this generative syntax argument with deconstructionism at a useful intermediate vantage point constructed by Bleich (1988), who proposes that literacy can be equated with the ability to perceive a situation from more than one perspective. Bleich, who offers his analysis as a user-friendly version (for education-focused users) of deconstructionism, builds his reasoning around the claim that literacy in this sense – the ability to differentiate individuals and to segment space – is a precondition for the child to begin to speak.

Generative theory’s basic story resonates with this analysis in ways that need to be highlighted. A human child, unlike even a very bright young ape, has the innate capacity for recursion (including “optional embedding” in the sense of the discussion above). Thus, it makes sense that, in the ontogenetic schedule of a human child’s development, the activation of this innate capacity by relevant experience (exposure to particular kinds of handling by caregivers – see Bleich’s account of what Anne Sullivan had to do to make Helen Keller begin to segment space and only then begin to acquire words) should precede the standard milestones of “language development”. But, unless more is said, a normal generative grammarian cannot readily go along with the thought that the capacity for recursion should meet literacy, of all things, at the vantage point that Bleich is calling “perspective”.

“Why is there supposed to be any special connection”, the normal generative grammarian is bound to ask, “between recursion and literacy? What an absurd idea! We know for a fact that millions of children, deprived of reading and writing, entirely confined to orality, move into the full use of sentence recursion every year. What does the Derrida-Bleich theory say about such children?”

By literacy the Derrida-Bleich account means what the generative idiom would term the human faculty of writing – not the manifest individual ability to read and write that this faculty gives rise to when coupled with appropriate social exposure to the actual use of a script system. The “illiterate” children that our reader is referring to exemplify what happens when that social exposure is unavailable: the faculty still supports recursion in the spoken competence, but it does not translate into any written competence in the absence of exposure to textual material.

The claim we are making here – a claim that should be easier to grasp now that the Derrida-Bleich proposal and the generative theory of recursion have been placed in dialogue at Bleich’s “multiple perspective” vantage point – is as follows (we call it Claim A for ease of reference):

Claim A: Recursion and the faculty of writing are closely interrelated properties of the specifically human language faculty, as distinct from the conceptual capacity that humans share with other higher mammals (which, in the domain of language, underlies the theta-theoretic structure of the clausal nucleus built around a verb).

Within the field of investigation that Claim A opens up, one important topic – highlighted by Derrida (1974) – is the nature of personal names. Derrida’s view was that names of individuals count as entries in a notional register that keeps tabs on all the members of the community. In other words, personal names presuppose the category of records and thus instantiate the faculty of writing – the term Derrida used was “archi-writing” – even in societies devoid of the practices of writing and reading.

Again, generative syntactic research resonates with such thinking.

Within the second generative revolution – the transition from the early rule-based grammars to current principle-driven and parameter-tuned accounts of syntax – syntacticians working on the basis of proposals made after 1980 have reached a consensus about the syntactic properties of names. In view of the articulation of the noun phrase into a substantive noun zone and a formal definiteness zone, inquiry has focused on the alternation between languages that use a definite determiner with proper nouns (names of countries in French – “la Chine” – and names of persons in Greek and colloquial German – “der Hans”) and languages that do not (English only has bare proper nouns like “China” or “Hans”, apart from complex structures like “the China that I was familiar with is no more”). These scholars agree that the analysis of this alternation must involve a process that indexically associates – to the point of identifying – the proper noun with the syntactic definiteness of the noun phrase.

To the extent that indexicality or deixis – characteristic of demonstratives such as “this” or “that” – is the salient content of the nominal construction’s formal definiteness zone, the idea that proper nouns are nouns uniquely associated with this indexical site amounts to the view that names count as pointing nouns – as the only nouns that point with reference to a field of indexically specified identities.

It is not the case that all names are associated with a “perspective” (in the sense of this term that drove the discussion of optionally embedded clauses). But personal or human names are. Syntax independently recognizes “human” as a distinctive feature with grammatical effects in the pronominal and inflectional systems. Thus, “human names” counts as a syntactic natural class. The fact that this class corresponds to indexed centres of perspective establishes an important affinity between the class of names (in particular, the subclass of human names) and the operation of embedding (specifically, the optional instances of this operation).

A closer look at these subcategories yields the result that “human” and “optional” have in common a direct connection with illocutionary freedom – the speaker’s freedom that goes into the making of the speech act itself. We may visualize the matter as follows: a human name designates someone in whom the speaker recognizes an image, an alter ego, of the self; an optional clause is one capable of standing alone as a full sentence, as a freely performed illocutionary act.

We may express this result as Claim B:

Claim B: (1) Optional embedding and human names are closely associated with the perspective-setting freedom of illocutionary centres. (2) The faculty of writing is identical to the capacity for multiple perspective. (3) Perspective-setting freedom shows up as the capacity for multiplicity of perspective; in other words subclaim (B-1) equals subclaim (B-2); human names correspond to the (“writable, recordable”) register of freedom.

We propose that Claim B, working within the theoretical possibilities implied by Claim A, establishes, via the notion of perspective, an exact correspondence between the generative syntax of proper nouns and the deconstructionist account of naming and writing.

This is a rigorous enough articulation to cast doubt on the widely held view that the deconstructionist account of language and writing is for some reason remote from the exact sciences or even from rationality itself. On the contrary, the foregoing considerations make it appropriate to reexamine the specific content of “human language” in the comparative and historical ethology of primates and humans.

At the ethological level, given the reasoning presented here, it should be possible for primatologists to design experiments that tap exactly what those specially trained chimpanzees and gorillas can do with personal names. Historically, the issue to look at has to do with the distinctly human surplus, in the use of names, that goes beyond primate naming capacity. If the perspectival centre approach to the theory of human names is on the right track, one needs to rearticulate the inquiry in terms of the multiplication of names. By this we mean not just the three-word structure of a name like Lev Davidovich Bronstein, but also the cross-level identification of formal Richard with informal Dick and with intimate Dickie.

These remarks, however, take us perilously close to the enterprise of formulating a Claim C. Such extensions of the account offered here would involve making distinctions between, say, mother talk (initiating the young into the availability of language) and the architecture of a public linguistic space that draws on the full potential for infinity across perspectival centres in “society”. Trying to go so far afield would not serve the purpose of interdisciplinary cross-fertilization. Let us terminate the theoretical core of the reasoning at this point and move on to its relevance for other vital concerns.

Given the backdrop many of us take for granted, the Derrida-Bleich theory connects most directly with the struggle for the rights of indigenous peoples and for the conservation of their languages and cultures. Pre-deconstructionist theories of language and society – however thoroughly purged of the view that there are primitive and advanced societies, that some communities are intrinsically more intelligent than others, that languages are born unequal – nonetheless retained some residue of the doctrine of a fundamental linguistic rupture between preliterate communities and their merely spoken languages (corresponding to the classical category of “savage” societies) and literate communities endowed not just with written languages but also with a state to give this writing a raison d’etre.

Since the emergence of deconstructionism, the demonstration we need – that the rise of the state and of writing did not represent a rupture in the intrinsic development of individual humans (from an incapacity to a capacity for writing) but an extrinsic historical fact – has been within our reach, and with it a coherent foundation for a politics that ceases to disenfranchise the specific forms of living and knowing that characterize the indigenous peoples.

Now that we have a demonstration that deconstructionism and generative syntax provide closely allied approaches to the nature of the language faculty, a cognitive science take on these issues can be fashioned, and can support a rigorously articulated relation between scientific research and programmes for social transformation. Furthermore, we are now equipped with properly updated tools on both sides of the language-culture interface to take up the sequel to Lévi-Strauss’s dialogue with Jakobson in the 1952 conference that served as our point of departure. The political importance of such a venture is impossible to overstate.

In a period marked by the urgency of ecological issues, given that we know that preserving the indigenous peoples’ richly endowed languages (which encode practical acquaintance with details of the ecosystem and how to husband its replenishable resources) is essential for the survival of our planet, the conclusion that cognitive inquiry can directly interface with ecological concerns is a welcome result of the deliberations presented here. This result converges with independently established priorities in the syntactic research traditions associated with the rise of cognitive science – broadly, with generative research in recent decades. Derivational parametric work and the closely related lexicalist alternatives to its transformational approach have drawn heavily on syntactic data from the languages of certain indigenous peoples. The conceptualization-focused paradigm of cognitive linguistics, in the process of fashioning its theoretical cardinal points, has drawn on the lexically encoded conceptual structures available in a slightly different set of languages spoken by indigenous peoples. These prioritizations indicate that linguists working at the time of the rise of cognitive science are aware of the specific importance of these languages. At a time like this, the proposal that we should return to some of Lévi-Strauss’s central questions – with whatever updating may prove necessary – is likely receive a more serious hearing than might have been possible in the fifties, or even the sixties.

References

Bleich, David. 1988. The Double Perspective: Language, Literacy, and Social Relations. Evanston: Northwestern University Press.

Dasgupta, Probal. 2008. Knowledge and language in classical Indian linguistics: some observations. In Suresh Raval, G.M. Mehta, Sitanshu Yashaschandra (eds.) Forms of Knowledge in India: Critical Revaluations. Delhi: Pencraft. 89-104.

Dasgupta, Probal. 2009. ‘Transparency and arbitrariness in natural language: some empirical issues.’ In Rajendra Singh (ed.) Annual Review of South Asian Languages and Linguistics 2008. Berlin: Mouton de Gruyter. 3-19.

Derrida, Jacques. Of Grammatology (tr Gayatri Chakravorty Spivak). Baltimore/ London: Johns Hopkins University Press.

Lévi-Strauss, Claude. 1968. Linguistics and anthropology. In his: Structural Anthropology. Tr Claire Jacobson and Brooke Grundfest Schoepf. Harmondsworth: Penguin. 67-80.

Lévi-Strauss, Claude. 1973. From Honey to Ashes. (Introduction to a Science of Mythology: 2.) Tr John and Doreen Weightman. New York: Harper and Row.