Implicit and Explicit Linguistics

Abstract

In this paper the affiliation of Lévi-Strauss with the functional structuralism of Trubetzkoy and Jakobson is defined as “explicit linguistics” and the consonance with the works of Franz Boas is also thrown into relief. As “implicit linguistics” is defined his accurate attention to lexical facts, in the case of questions of nomenclature and in particular of the terminology of kinship. As this regards, Lévi-Strauss perfectly captures not only the differences between linguistic systems and anthropological systems, but also their respective integration. In this way he comes to surmount the model itself of the functional structuralism of a certain linguistics which is too predictive. Finally, in homage to the great scholar, two examples are proposed of the integration of linguistics “without adjectives” and anthropology “without adjectives” in relation to the etymology of gr. anthropos, seen from the point of view of a specific sociolinguistic hierarchy and within the frame of the Indo-Mediterranea area, and of the Indo-European name of the woman of marriageable age (gr. gyné) seen in the perspective of the innovative conditions of exogamous marriage.

Lévi-Strauss’ discovery of linguistics – more specifically, the Prague School phonology of Nikolai Trubetskoy, which he encountered in America through Roman Jakobson – represents a sort of milestone in the arc of his scholarly oeuvre. Even if his career subsequently evolved autonomously along very particular lines, with the extraordinary results we know today, there can be no denying the contribution of linguistics to his thought. I will use the term ‘explicit linguistics’ to refer to that part of Lévi-Strauss’ oeuvre in which, beginning with The Elementary Structures of Kinship and later culminating in Structural Anthropology, he acknowledged the debt his anthropology owed to structural linguistics (albeit with some disclaimers). In contrast, I will use the term ‘implicit linguistics’ to refer to the work in which, addressing problems of nomenclature, particularly that of kinship (Lévi-Strauss, 1969: XXIII–XXIV), he paid close attention to lexical data without however deploying it as a form of cultural proof, giving it if anything the status of eloquent but ultimately inconclusive evidence. I will also use ‘implicit linguistics’ (to me the more interesting aspect of the illustrious scholar's long-standing interest in matters of language) to indicate all those places in his work where linguistic data (particularly lexical data, with its inevitable etymological implications) might open potential new avenues of research for anthropology more generally.

In terms of ‘explicit’ (or as one might say in this case, ‘programmatic’) linguistics, Lévi-Strauss insisted on the importance of the structural method for anthropology in the very book whose title eponymously described its epistemological dimension. An ‘implicit’ comparison with linguistics can be seen in the clear distinction he drew between the ‘comparative nature of the sociological method’ (Lévi-Strauss, 1963: 1), in which the anthropologist is positioned as a comparative sociologist, and the ‘documentary and functional character of the historical method’ (Lévi-Strauss, 1963: 1), where ‘historical method’ is understood to mean that which focuses on events and not that of the longue durée represented by the Annales school (the direction in which the most sophisticated historical linguistics turned in the second half of the twentieth century). The descriptive exasperation of ethnography when confronted with ‘human groups considered as individual entities’ (Lévi-Strauss, 1963: 2) was quickly overcome and subsumed through the epistemological stance of social and cultural anthropology, in which human institutions revealed themselves to be specific ‘systems of representations’ (Lévi-Strauss, 1963: 3–4; this was not a casual choice of words; Lévi-Strauss used this phrase repeatedly). This is the path that would lead him to his ‘American’ encounter with the ‘hardest’ aspect of imported European structuralism, the functional and teleological phonology of Jakobson.

But first Lévi-Strauss paid homage to one of the great figures of anthropology and linguistics, Franz Boas (Lévi-Strauss, 1963: 4–6), who was always attentive to the problem of ethnological transformations, recognizable ‘with indirect methods’ as in comparative philology – the British and, later, American term given to the Indo-European method of comparative reconstruction. Lévi-Strauss also approached structural linguistics (with the consequent suspension of the historical dimension) when he observed that ‘[h]istory organizes its data in relation to conscious expressions of social life, while anthropology proceeds by examining its unconscious foundations’ (Lévi-Strauss, 1963: 18). Now we come to linguistics and its merits, and particularly ‘the high value of [the] linguistic method’ (Lévi-Strauss, 1963: 19) for ethnological research. Here, Lévi-Strauss quotes Boas directly, giving us the chance to see how Boas had ‘helped’ him arrive at this critical discovery:

The great advantage that linguistics offers in this respect is the fact that, on the whole, the categories which are formed always remain unconscious, and that for this reason the processes which lead to their formation can be followed without misleading and disturbing factors of secondary explanations, which are so common in ethnology, so much so that they generally obscure the real history of the development of ideas entirely. (Lévi-Strauss, 1963: 19–20; cf. Boas, 1911: 70–71)

The reference to the unconscious dimension of the linguistic (and in particular the phonological) competence would become the epistemological foundation of the structural method in Lévi-Strauss’ anthropology.¹

Comparative linguistics, which must be understood here in a strictly synchronic sense and in terms of ‘explicit linguistics’, draws its legitimacy from ‘something more than a mere fragmentation’ – namely, a real analysis. From words the linguist extracts the phonetic reality of the phoneme; and from the phoneme he extracts the logical reality of distinctive features’ (Lévi-Strauss, 1963: 20). It thus positions itself as a ‘hard’ epistemological model for every human science that hopes to transcend the limits of subjective perception. From this point of view, Lévi-Strauss seeks to reaffirm in a subtle yet decisive manner that in the process of understanding of the human sciences more linguistico demonstrata, ‘[t]he transition from conscious to unconscious is associated with progression from the specific toward the general’ (Lévi-Strauss, 1963: 20–21). In the first edition of Structural Anthropology, he stated plainly that ‘[l]inguistics occupies a special place among the social sciences, to whose ranks it unquestionably belongs’ (Lévi-Strauss, 1963: 31). From the perspective of explicit linguistics, synchronic and structural, this is put forward as a methodological perspective. But in the lens of the other, ‘implicit’ linguistics, there is another important observation:

The linguist provides the anthropologist with etymologies which permit him to establish between certain kinship terms relationships that were not immediately apparent. The anthropologist, on the other hand, can bring to the attention of the linguist customs, prescriptions, and prohibitions that help him to understand the persistence of certain features of language or the instability of terms or groups of terms. (Lévi-Strauss, 1963: 32)

This is valid, as we shall see in greater detail, for the terminology of kinship and its evolution, at times explicable only through the rapprochement of anthropology and linguistics.

The moment has come for declarative emphasis. In the words of Lévi-Strauss and ‘his’ explicit linguistics: ‘But, after all, anthropology and sociology were looking to linguistics only for insights; nothing foretold a revelation’ (Lévi-Strauss, 1963: 33).² And then: ‘The advent of structural linguistics completely changed this situation […]. Structural linguistics will certainly play the same renovating role with respect to the social sciences that nuclear physics, for example, has played for the physical sciences’ (Lévi-Strauss, 1963: 33). This path leads Lévi-Strauss to some interesting observations, such as when he recalls Trubetskoy's polemic against the ‘atomism’ of the old historical linguistics, and accuses a certain type of kinship ethnography of getting tangled up in usages and terminologies understood as either consequences or vestiges, falling into a ‘chaos of discontinuity’ (Lévi-Strauss, 1963: 35), while in contrast structural linguistics teaches one to view things according to synchronic systems and diachronic teleology. Nevertheless, he rightly warns against making any superficial parallels between kinship terms (operating thus on the plane of implicit linguistics which, being circumstantial and diachronic, cannot escape a value-based perspective!) and phonemic systems, declaring: ‘In our own kinship system, for instance, the term father has positive connotations with respect to sex, relative age, and generation; but it has a zero value on the dimension of collaterality, and it cannot express an affinal relationship’ (Lévi-Strauss, 1963: 35). For our purposes it is enough to note that in the case of archaic Latin, which perpetuated a specific Indo-European heritage, the primary sense of pater implies a positive primary connotation linked to sexually marked power; a woman who enters the sphere of the pater familias establishes through marriage ‘an acquired kinship relationship’, albeit one amended by the eloquent syntagma filiae loco. An expression like bonus pater familias has a triple anthropological connotation because it refers to a person who is ‘good’ (bonus), in the sense of being capable, is endowed with full authority (pater) over all subalterns (wife, children, servants, animals, etc.), who in fact constitute the familia (understood as a collective of famuli, or servants). The linguistic syntagma becomes an effective diagram of a specific anthropological order, and its proper translation should be ‘capable master of the house’, not ‘good family father’, which is only the result of anthropological evolution over the longue durée.

Lévi-Strauss then introduces a distinction to surpass the limits of pure lexical data:

Thus, along with what we propose to call the system of terminology (which, strictly speaking, constitutes the vocabulary system), there is another system, both psychological and social in nature, which we shall call the system of attitudes. Although it is true (as we have shown above) that the study of systems of terminology places us in a situation analogous, but opposite, to the situation in which we are dealing with phonemic systems, this difficulty is ‘inversed’, as it were, when we examine systems of attitudes. (Lévi-Strauss, 1963: 37)

This is a point of fundamental importance. It allows us to see immediately that Lévi-Strauss’ structuralism is in no way characterized by a methodological ‘fundamentalism’. To the contrary, it admits open and integrated systems, following a formula that I would not hesitate to call ‘perfect’: ‘The system of attitudes constitutes, rather, a dynamic integration of the system of terminology’ (Lévi-Strauss, 1963: 39).³ The subsequent elaboration of a specific theory of attitudes, regarding the maternal uncle and the avunculate, brilliantly defined as the ‘atom of kinship’ (Lévi-Strauss, 1963: 48),⁴ resounds as a clear confirmation of this major epistemological turn.

For my part I would like to underline the fact that the disappearance of the Latin avunculus (maternal uncle) in Italian, where it was replaced by the neo-Greek zio, and its tenacious and even exemplary persistence in French – to the degree that it was in turn loaned out to both English and German – should be properly studied in terms of the dynamic integration between systems of attitudes and systems of terminology, first in the Latin world, then in the neo-Latin and Germanic, according to Lévi-Strauss’ instructions. The quadripartite system in Latin that has avunculus (maternal uncle) matched with matertera (maternal aunt), across from patruus (paternal uncle) and amita (paternal aunt), comes across through the nucleus and the various derivational suffixes, a peculiar system of attitudes which anthropological data cannot explain, while the linguistic data might in this case provide an eloquent bit of evidence. I refer to the fact that avunculus appears, by virtue of its morphological profile, to mean a ‘little grandfather’, but while avus can be paternus or maternus, the word in question refers only to the brother of the mother, with a ‘widening’ provided by * -en-/-on- (which recurs in the Celtic languages) before the appearance of the evaluative suffix. As for matertera, the maternal aunt, it is impossible to speak of equilibrium in the system of terminology, in that the nucleus mater does not refer to the previous generation, and the suffix -tera does not possess, as in the prior case, a evaluative function, but if anything marks an opposition that is, however, not a direct contrast (it is a mother a latere, ‘on the side’ – the mother's sister). The avunculate appears to suggest a positive trait, although – following Lévi-Strauss’ illuminating instructions, which dissuade us from ‘matrilineal’ oversimplification – it is not prestigious or authoritative, since it lacks a reference to the father. The paternal uncle, or patruus, does carry a name that mitigates, by virtue of the suffix which refers purely to the nucleus, any implication of parental prestige; the paternal aunt, amita, does not even descend from this nucleus, but rather appears to be a vague derivative of amma, a word used by infants to refer to the maternal figure. To sum up, from the system of terminology comes a system of attitudes with probable inter-systemic references that are not at all clear: if I might use a metaphor, the Latin system of terminology regarding this particular form of kinship can be considered as a linguistic architecture sui generis, built out of four different ‘recycled materials’ from different backgrounds.

For Lévi-Strauss, in cases like this the anthropologist would turn for help to the linguist – but, I would add, it is not a given that he would receive it. In essence, the linguist is caught in the jaws of a trap: confined on one side by the arbitrary and conventional character of the sign, and on the other by the risk of treating shadows like solid objects when he scours texts not for words themselves but for their long silhouettes in the complex play of their communicative agility. I would emphasize the fact that Lévi-Strauss, beyond a few careful but not completely convincing declarations of incompetence, was fully aware of the long voyages and peculiar metamorphoses suffered by words, especially when they become the visible idiosyncratic peaks of specific psycho-linguistic perceptions. I find this autobiographical disquisition particularly delicious:

As for myself, who has spoken English exclusively during certain periods of my life without, however, becoming bilingual, fromage and cheese mean the same thing, but with different shadings. Fromage evokes a certain heaviness, an oily substance not prone to crumble, and a thick flavor. This term is especially suitable for denoting what (French) dairymen call pâtes grasses (high in butter-fat content), whereas cheese, which is lighter, fresher, a little sour, and which crumbles in the mouth (compare the shape of the mouth) reminds me immediately of the French fromage blanc (a variety of cottage cheese). The ‘archetypal cheese’, therefore, is not always the same for me, according to whether I am thinking in French or in English. (Lévi-Strauss, 1963: 93)

If it is true that in terms of implicit linguistics (with all due respect for Trubetskoy's phonology, and certainly without forgetting Jakobson's ‘grammar of poetry’!) a system of terminology and a system of attitudes must necessarily be integrated in a vision that I would call ‘pragma-linguistic’, this is a case in which that process unfolds perfectly. Lévi-Strauss, an adept enthusiast of functionalist structuralism, thus transcends his own epistemological choices – as do all people of genius – and at the same time provides proof of an extraordinary freshness thanks to a truly profound perception of the facts.

Finally, in his honor I would like to propose two cases of the entanglement of ‘words’ and ‘attitudes’ that might represent a modest example of how a linguistics without qualifiers could dialogue with an anthropology similarly stripped of qualifiers. I refer to the extremely instructive use of the term ánthropos in Homeric poetry,⁵ which shapes a very specific attitude toward the human that unsurprisingly recurs in the greatest languages and civilizations of the ancient Near East; I also refer to the Greek name for woman, gyné, and some of its unexpected and inexplicable re-emergences in the modern argot. In the first case I will limit myself to noting that, in Homer, ánthropos designates the lowest level of the human, anonymous and plural; situated on an intermediate level there is anér, one who has a name and status (civilian or military), and at the highest level there is phōs, the Homeric hero, whose name almost – with the same morphological derivation, it needs only a slight variation in accent – denotes the same meaning as the Greek word for ‘light’, phôs (i.e., the clarissimus vir of Latin tradition). Within this system of terminology ánthropos reveals its oldest and most eloquent sense through the atonic suffix -opos, which refers to the ‘exterior aspect’, and the tonic syllable anthr-, indicating ‘that which is dark (like carbon)’, following the taxonomy of the anonymous and the indifferentiated that functions as the opposite of the named and individualized as a luminous being. It might be surprising (though surely not too much) to note that among the Sumerians, as well, an indistinct, anonymous mass of people was called ‘the multitude of black-headed (or black-faced) people’ (uĝ₃ saĝ gi₆.g), echoed and confirmed by the Akkadian salmat qaqqadi(m), ‘the black heads’ (literally, ‘black headed’). An additional element of proof can be found in the form of the ancient Egyptian expression harér eresh, ‘black heads’, which hyperonomically denotes ‘men’ without any reference to skin color (in the Egyptian of the Old Kingdom, as well, hr.w designated undifferentiated ‘faces’, anonymous and indistinct masses of men).

The Indo-European form reconstructed from the name for ‘woman’⁶ – it remains unclear which is the truly prototypical reference, as we shall see – can be represented as *gwenā and also, with a significant variation on the final vowel, as *gwenə and *gwenī (it is important to point out, however, that the first and the third form provide a clear, substantive indication of the feminine gender with the final vowel, whereas in the second the indistinct final vowel admits the possibility of its timbric levelling on the tonic syllable [e]). Add to this the fact that some important Indo-European languages (Sanskrit, Greek) allow the construction of a form with a reduced apophonic grade (i.e., the Sanskrit gnā, or the Greek gyné, from *gwnā), and the question immediately arises whether this name belonged to a notion originally expressed by a verb, which remains to be identified. In any case, in ancient Greek the indication of gender through gyné (Iliad, 6, 390) quickly shifted into a reference to a woman united sexually to a man, not merely as a concubine (Iliad, 24, 497) but also as a legitimate(d) wife, and then in the post-Homeric era became ‘the usual name of the married woman’ (Chantraine, 1990, s.v.), in a decidedly much higher register (in Vedic India gnā also meant ‘goddess’, while in Cornish – an Indo-European language in the Celtic sub-family – ben-en again means ‘wife’). The forms with the reduced grade include the Boeotian bana (as documented by the poetess Corinna), in addition to the ancient Irish ban- in composite terms, while the most significant responses in the ‘middle’ grade of the verb root with the vocalization [e] are provided by the Gothic qino (nasal infix) and the ancient Slavic žena (not to mention the modern English Queen, which raises the ancient reference to gender to the level of royalty). But what is the verb that lies beneath this incredibly important and widespread Indo-European name for ‘woman’, or rather a woman as understood in a very specific context (see above), from the perspective of the motivation behind the name itself?

The Indo-European verb believed to be the best candidate to trace back to the apophonic (and thus deverbal!) name of ‘woman’ (*gwen-/gwn-) is that which expresses the idea of ‘movement’, both in terms of approach and arrival as well as distancing and detachment. This presents itself according to the radical variation *gwen- (assured by the Greek βαίνω and the Latin ueniō, but also by the Oscan kúmbened, which corresponds to the Latin perfect conuēnit), and especially according to the primary radical form *gwem- (confirmed through the data provided by the Gothic qiman and the old English cuman, among others, while the Lithuanian gemù [‘to be born’, or literally ‘come into the world’] appears rather interesting with its additional semantic implications). At this point the (secondary) variant *gwen- assumes a form dissimilar to that of the primary expressed by *gwem- (from ‘labiovelar[gw]-bilabial[m]’ to ‘labiovelar[gw]-dental [n]’), probably already on the Indo-European level with a perfect parallelism between the Greek βαίνω (see the Latin ueniō, or old English kúmbened) and gyné, the Boeotian bana (from the Indo-European *gwnā). From a semantic perspective this rather plausible idea fits into the prehistoric institutional context of exogamous marriage, in which the woman ‘arrives’ or ‘enters’ (i.e., *gwem/n-) into the clan of the husband, precisely within the scheme of universal prohibition of incest illustrated by Lévi-Strauss in his Elementary Structures of Kinship.

This most ancient name for ‘woman’ seems to have its prototypical semantic fulcrum in the notion of ‘young age’ and the reference to ‘mobility’. In this perspective it seems to re-emerge, via unrecognizable linguistic paths and pertinentizations, in certain forms of argot like the Italian underworld/street term guagnastra (with a very clear evaluative suffix) or the (loutish) Milanese guanguana (an similarly clear reduplication), both meaning ‘prostitute’, to which we might immediately add the Neapolitan form guanguana, with its connotation of ‘mistress’ (heard also in Sicily). Angelico Prati, from whom I have drawn this source material, locates the center of expansion of this term in Naples: ‘The literary evidence and the various derivatives of guagnastra demonstrate that their birthplace is Naples, and that the primary meaning is “young girl, shapely young girl”: guagnastra in 1632; guagnasta in 1729; guagnastrella (young girl) in 1633; there are even a few guagnastro (“lover”, or “husband”) circa 1783; and no bello guagnastrone (c. 1621), “a beautiful woman” […] has the same meaning as guagnastra from 1633.’ For Prati, ‘guagnastro, -a is guagnone “young boy, young man” (c. 1635) with another suffix’ (1978: 85), but – in my opinion – guagnone (today's guaglione with a different regressive) cannot be primary, if one considers the compact (and complex) documentation of the female-gendered form studied above.

I would like to add, in conclusion, a new data point, which might confirm the antiquity and autonomy of this lexical series: this is the French gouine, a term from the argot picked up by Victor Hugo (see, for instance, in the collection Châtiments the verses Que la vieille Thémis ne soit plus qu’une gouine/baisant Mandrin dans l’antre où Mongis baragouine) with its connotations, as we have seen, of ‘bad woman’ or ‘prostitute’. Naturally, the vocalization of this form is not the same as that of the other terms above, but – notwithstanding the apophonic variance of the presumed Indo-European etymon (see above) – this does not present an obstacle; rather, it ends up being a confirmation of the belonging of this term to the series, with the advantage of restoring to us a primary form of the full apophonic grade (without suffixes or reduplication).

The history of a word in the longue durée, in so many ways emblematic, adds a new and unanticipated piece of the puzzle (with an equally unexpected assist coming from the direction of France) from right in the heart of Naples: that is, from a part of the ancient Mediterranean world incredibly rich in every era of history, and in every era shot through with episodes (including linguistic events) still little or insufficiently understood. In this way, in my view, both anthropology and linguistics once again come together, thanks to the illustrious man, so dear to us all, that was Claude Lévi-Strauss.

Translated from the Italian by Richard R. Nybakken