The Neolithic dualist scheme

– The monumental twin steles of Göbekli Tepe are one in a long series of isomorphic compositions in Neolithic symbolism. Seemingly tracing back to the Palaeolithic, symmetry likely played a fundamental role for prehistoric societies. Ethnographers showed how hunter-gatherer ideology (mythology, totemism, etc. ) is often structured around a dualistic worldview (male/female; sum-mer/winter etc. ) taking root in the kinship system through a division of the community into exogamic subgroups. It is this dualism that is argued to be embodied in the


Introduction
The Neolithic site of Göbekli Tepe in Northern Mesopotamia has raised much interest, but the most relevant questions have mostly remained unanswered. One reason is that the field of prehistory alone is rather ill-adjusted to properly address matters of ideology. The ethnology of comparable extant societies offers an alternative means of exploration (Forest 1992.28-31;Yakar 2005.111-112), as it can reveal the concepts conveyed through the symbolism.
The present discussion concentrates on the most imposing feature of the Southwest Asian Neolithic symbolic world, that is, the monumental pairs of twin steles standing in the centre of the PPNA stone enclosures (A to H) of Göbekli Tepe III (Fig. 1). However, these must be conceived just as one particular case among the numerous contemporary isomorphic (architectural, geometric and iconographic) representations identified throughout (Fig. 2) the Neolithic period (Peters et al. 2005.31-32;Stordeur 2003): symmetric clay poles, parallel lines painted on floor, geometric figures on walls, antithetic or converging animals, twin figurines, couple of human skulls, symmetrical partition of communal buildings or of entire sites. A non-exhaustive list of such items

Problematic and methodology
The fact that the Neolithic symbolic material possesses such a meaning is made explicit by the planned and recurrent arrangement in which its various elements are invariably found. The symmetrical socalled 'pillars' systematically hold a central position in the communal buildings of a large number of spatially and chronologically separated sites in Northern Mesopotamia, the Levant and all the way to Late Neolithic Central Anatolia 1 1 . Moreover, this element often appears in association with the same set of figurative elements of strong symbolic connotation is presented elsewhere (Bodet 2021;forthcoming). The current paper concentrates on the likely meaning enclosed in these isomorphic symbols, or rather, on the ideology and the social structure they reflect.
Ethnoarchaeological analogy presents various problems (David, Kramer 2001.51-54), but the symbol in question here appears remarkably central to both archaeological and ethnographic societies. If the analogy proves appropriate enough, the analysis may somehow make the archaeological data 'speak' (Gould 1978.250), thereby unlocking some of the meaning enclosed in the symbolism (Wilson 2020.6). (Henderson 1964.154), in particular predators, snakes, birds (probably with a psychopomp function), bucrania, anthropomorphic statues or smaller side-steles. The ordering and redundancy of these symbolic compositions are doubtlessly not arbitrary and must correspond to a predefined logical system, conveying a particular message left to be deciphered (Stordeur 2003.32;Testart 1987a.171).

Fig. 1. Ground plan of Göbekli Tepe. GT_Gesamtplan_2014 (central area truncated), by Klaus Schmidt and Jens Notroff. © Deutsches Archäologisches Institut, Göbekli Tepe Projekt.
This message carried by such structures, together with the fact that nothing indicates that they supported anything (Jeunesse 2020), is, in passing, the reason why the word 'stele' is preferred here to that of 'pillar' generally used in the literature. In the same trend of thought, the term 'temple' is ill-fated to designate these enclosures, as these ceremonies probably do not imply a 'cult of deities' (Testart 2006a), which only arises when required for the ideological unification of large urban congregations during the much later Obeid/ Uruk horizons (Forest 1996a).
Though rarely or too briefly (Voigt 2002.254) mentioned (Roger Matthews 2003.37 is a significant exception), the decipherment of this symbolic message was successfully initiated nearly three decades ago by Jean-Daniel Forest (1993), providing a solid foundation on which further elaboration ought now to proceed.
Always placed in the centre, the twin composition is suspected to symbolize the highest sphere of Neolithic ideology. It is a simple symbol with which researchers are doomed to start with in order to patiently reconstruct the meaning it may hold in the social structure (Durkheim 1937.42-45). One precondition is not to underestimate the capacity of early communities to express abstract themes through corresponding symbols.
A symbol is a signifier standing for a signified. Repeated over and over again in places dedicated to communal matters, the signified in question must indeed be very significant for the community. This element can only be described, for now, as a symbol of symmetry, but it makes sense within a system of thought (ideology) deriving from a correspond-ing social context (Yakar 2005.111). Disconnected from this context, the symbol loses its meaning. This implies that in order to decipher this symmetrical symbol, the social context first ought to be reconstructed, at least in broad strokes.
This social coherence is an indispensable basis to start with, but without an intermediary reference, without a Rosetta Stone infusing the structure with meaning, the decipherment will be left to hollow speculations (Schmandt-Besserat 2013.xxv). It happens that, for the present concern, a reference exists. The latter is not a similar iconographic element but an abstract concept, and the correspondence appears too striking and the analogy too compatible not to be considered.

Dualism as a prehistoric principle
This investigation was originally inspired by Alain Testart's (1985) meticulous analysis of what he calls the 'primeval communist' societies. These are a pristine form of hunter-gatherers, prior to the advent of bows and arrows. This long Palaeolithic dawn of humanity appears, though with much caution, accessible through the abundant ethnographic documentation of Australian Aboriginal societies (Testart 1988.12). Any analogy led through the spectrum of a narrow technological comparison is doomed to failure, and absolutely no cultural comparison is attempted here; the social structure, because of its theoretical and universal nature, is the only element considered. 'Universal' is here to be understood in the sense that every human community necessarily has an economy, a kinship system, rules, customs and an ideology: these are the main structures that concern us here.

Fig. 2. Distribution of Neolithic sites with identified dualist symbolism (see Bodet (2021 forthcoming)).
Australian Aborigines are the only living mirror of Pre-Mesolithic-type of societies. A coherent and allencompassing theoretical reconstitution of their social structure was achieved by Testart, as stable and lengthy as the Palaeolithic period itself, as it is known from archaeology at least. But the ultimate reason why Aborigines are of interest to us here is that, together with Neolithic societies, they clearly appear to hold dualism as a keystone of their ideological construction and that only a living society can reveal its meaning.
A word of warning, however. This is not simply about making "connection between two entirely different societies on the basis that they use symmetrical symbols in their ideology", as one reviewer of an earlier version of this paper suspected. There could be a slight chance that isomorphism may relate to something utterly different in the two societies although as dualism and exogamy are extremely widespread (near universal) in the ethnographic record, there is a much higher chance that the Neolithic isomorphism is a delayed expression of the Palaeolithic dualism. Moreover, the modern understanding of dualism, "two irreducible, heterogeneous principles" (Britannica.com) must be here understood as being thoroughly complementary in their opposition, which is what we will try to report.
Ethnographers have been struck by the extreme attention given to kinship by all traditional societies, and in particular to one critical point (Spencer, Gillen 1899;Frazer 1910;Howitt 1905): such societies are always divided into several subgroups, at least two, exchanging sexual mates every generation for the sake of procreation. This 'artificial' social division is at the root of exogamy ('marrying outside'), an absolutely fundamental principle from which later social and kinship systems evolved (Freud 2010(Freud [1913. 39-53, 255-256;Lévi-Strauss 1967.80-97;Testart 1985;1988). One direct consequence of this law is that, as Robert S. Walker et al. (2011.1) say, "arranged marriages [necessarily among relatives] are inferred to go back at least to first modern human migrations out of Africa".
The archaeological data offers monumental evidence to support the idea that this arrangement perpetuated at least until the Neolithic.
It is certainly biased and erroneous to designate a society by what it does not have (bows and arrows). What Palaeolithic-type societies do have, and even more so than later hunter-gatherers, are relations of production entirely based on and shaped by an elaborate kinship system. In these small communities turned inward, 'elementary' to use Claude Lévi-Strauss' (1967) term or 'universal' to use Alan Barnard's (19782020.50-53), everyone is somehow related to everyone else, and this kinship or marital relation dictates the modalities of their social interaction in every way. The way individuals respect, joke, avoid, command/obey, punish, teach/learn, give/receive, conduct ceremonials (initiation and funeral rites), and, most importantly, the way they marry, are thus prescribed primarily by the subgroup the people in interaction belong to (Malinowski 1926;Radcliffe-Brown 1952.90-104;Woodburn 1982;Ghasarian 1996.152-159,185-197;Walker et al. 2011.2;Bird-David 2019.15-16).
"The fundamental feature in the organisation of (…) Australian tribes, is the division of the tribe into two exogamous inter-marrying groups. These two divisions may become further broken up, but even when more than two are now present we can still recognise their former existence" (Spencer, Gillen 1899.55). Even though odd numbers may also be found (as a result of historically induced disparities), 'primeval communist' societies are often separated into parallel subgroups or phratries: eight subsections, four sections, or, for the most genuine case, two moieties (halves) (Barnard 2020.52). Dualism is thus generally considered to be the most original and purest form of this form of social organization (Cook 2003.65;Freud 2010Freud [2013. 50-51;Testart 1978.15-22;1985.478-479). But whatever the number of subgroups, this plurality is necessarily reduced in conceptual terms to the number two, because it is the number par excellence that embodies the concept of 'differentiation' (Girard 1972.87-92), making (equal) exchange possible.
Dualism is much more than a marital arrangement. Organically articulated to the economic system, it reflects on the symbolic sphere: totemism, mythology, rituals, etc. (Testart 1985.451-489). It finds in nature an obvious mode of expression, through fixed oppositions such as day (sun) and night (moon), winter (cold, wet) and summer (hot, dry), and, more particularly, males and females, the interdependence of which is naturally indispensable for the perpetuation of the cycle of life and death, oppositions themselves seen through their own interdependence.
This fecund sexual opposition was suspected by André Leroi-Gourhan (1964.108), among others (Testart 2006b.26), for Upper Palaeolithic societies, and by James Mellaart (1967.48, note 27-28), Ian Hodder or Forest (Matthews 2003.46) for Neolithic ones. Dualism is also the principle behind the famous Yin (female/earth/moon/water) and Yang (male/sky/ sun/fire) of the Chinese tradition (Granet 1929. 225). A similar symbolic partition of fundamental opposite elements is still present today, for example, in the arrangement of the Berber house in northern Africa (Bourdieu 1980). The philosopher Volkert Haas likewise refers to the concept of separation of the 'undifferentiated' cosmos in primeval times into two sets of opposite but mutually interdependent elements, in particular above-heavenmale and below-earth-female (Becker et al. 2012. 30). This widespread differentiation is personified in the antithetic heroes, twins or brothers/sisters in many founding myths all over the world (Girard 1972.247-248).
Among Australian Aborigines, this binary interdependence becomes a ubiquitous principle encompassing inorganic elements like mountains, water holes, stars or meteorological events (storms, rainbows). The entire world is thus systematically divided into separated but interdependent halves, a reflection of the society itself, as ideology generally does (Testart 1985.467-489). Dualism thus does not appear as a cultural but as a structural element deeply wired in the constitution of (all?) early human societies.
Barnard (pers. com.) tells me that this dualist division of the society "is true for Aboriginal Australia, but not necessarily for hunter-gatherers in general". This is a crucial point because it shows the chronological and structural evolution from 'primeval communists', for which Australian Aborigines are the sole ethnographic representatives, towards 'later' hunter-gatherers like the !Kung San, in which relations of production seem to have been altered by a certain spur of individualization (see below ;Testart 1985.56-60;1987b). This evolution would explain the distinction between the 'socio-centric system' and the 'ego-centric system' made by Alan Barnard (1978.77), as well as, the full "correlation between the system of kin categorization as a whole and the rules of marriage" (ibid. 75) that characterizes the Australians but is not found among the San. All this tends to show how 'primeval communism' could represent the genuine social background, characterized by "a lack of ambiguity of categorization" (ibid.) and from which later developments are likely to have derived.
Among these later developments there is the Neolithic period. Right in the centre of the a priori mysterious symbolic repertoire on display at Göbekli Tepe, there is a pair of huge parallel stone slabs standing majestically, seemingly conveying an abstract statement (Becker et al. 2012.14). They appear as nothing but a material representation of this universal dualist scheme. This is the hypothesis that will presently be explored by trying to understand what this dualism is really about.
It goes without saying that Australian Aborigines have absolutely nothing to do with Göbekli Tepe, it is just that they seem to share a similar social structure, thus opening the door to a possible analogy, which now needs to be questioned. A major obstacle first ought to be removed: if the aspect and centrality of the 'primeval communist' principle and the Mesopotamian Neolithic symbol present a striking similarity, these societies must be somehow structurally compatible for this analogy to function.

The analogy
Ideology is a central social organ in close interaction with the relations of production (Giddens 1971. 42), which implies that, for the analogy to be acceptable, the mode of production of the societies in question ought to be comparable 2 2 . From this point of view, primeval hunter-gatherers appear starkly different from Neolithic proto-farmers (Willcox, Stordeur 2012.112;Asouti, Fuller 2013.308). However, a sociological rule needs to be considered here: if technical and economic changes can diffuse rapidly, their repercussions for ideology (and, subsequently, for symbolism) are always very much delayed. This fact has been well attested by anthropologists studying the appearance of agriculture. "People can hold on to ideologies (mode of thought) reflecting foraging for generations, even when their systems of production have undergone transition"; "relations of production among proto-agriculturalists (…) tend to retain the structures of a huntergatherer habitus" (Barnard 2007.8,14, quoted by Asouti, Fuller 2013).
An ideological structure should not be seen here as a conscious and planned construction, but, indeed, somehow like Pierre Bourdieu's (1980) habitus, that is, continuously shaped by an everlasting accumulation of practice and experience. The ideology of the earliest farmers is thus likely to be largely inherited from a Palaeolithic background, built over hun-dreds of thousands of years: Forest (2006.126) thus states without hesitation that "in the case of the Pre-pottery Neolithic, the analogical basis to take into consideration is of course the 'primeval communism'". This implies that Neolithic people are likely to have possessed a cosmogony organized around some form of totemism, animism or shamanism (Lewis-Williams 2002.132;Bischoff 2002.237;Yakar 2005.112).
From there, it is difficult to support the idea that Neolithic people started to cultivate grains because they would have begun (why?) to conceive differently (how?) their relation to deities (are there any?). This is where Jacques Cauvin's (1997) famous model is problematic, and why the chain of causal effects may benefit if reversed (Testart 1998.27).

The Palaeolithic 'middle range'
The idea that the dualist ideology could be present in Neolithic symbolism as a continuum of a much older tradition would certainly gain some weight if found directly in the Palaeolithic period proper. As mentioned above, Leroi-Gourhan (1964) pointed out such reciprocal dichotomy in the Franco-Cantabrian cave paintings, Lascaux in particular. Horned animals (placed on protruding parts of the cave wall) are supposed to represent a male abstraction (horns being an obvious phallomorphic symbol, see Hansen 2017), while hornless animals (horses, in concave spaces) a female one, their interaction leading to fertility. These images are moreover painted in the very depth of caves, an obvious symbol of Mother Earth's vagina and womb, where not only humans but all organic forms come to life. "The earth would have been considered the source of all life's elements" writes Yak Yakar (2005.111-112), specifying that such "communal fertility-related rituals may have originated in the period before farming became the principal subsistence economy". Dimitrij Mleku∫ Vrhovnik (2021.3) further says that a cave is "a womb and a tomb at the same time", which would fit well with the above-mentioned idea that life is conceived as taking place in an eternal cycle where death is its inevitable opposite (Gibson 2009(Gibson [2010). Jean-Loïc Le Quellec (2015.259-260) comes to the same conclusion after his comprehensive studies of ancient myths from around the world, where humans and animals emerge from a hole in the ground, making the underground at the same time the place where life originates and where the deceased return. The cave paintings thus seem to have put into action symbolically the "structuring principle (of) vitality (fecundity, life-force)" (Ver-hoeven 2002.244), astonishingly resembling the 'tao' of the Chinese (Granet 1929.293).
But if all forms of life are concerned, it is certainly the community that is primarily envisaged by the principle of fertility.
The dualist principle seems to have been known by the Neanderthals as well (Fig. 3): the symbolic composition found at the bottom of the Bruniquel cave (Jaubert et al. 2016) represents two piles of stalagmites (another phallomorphic symbol) in one circle (a shape often connected with maternal womb and fertility, see Haland 2017.166), making the composition strikingly similar to the circle surrounding the twin steles of Göbekli Tepe III (Figs. 2, 3). Since absolutely no cultural connection can be established between these cultures, such similarity can, here again, only make sense if dualism and exogamy are understood as extremely widespread principles among early humans (Freud 2010).
Because the animal species painted in the Palaeolithic caves are represented as isolated groups, Testart (1985.276-290;2012.254-267) sees them as totems. His analysis is based on the identification, mostly by James George Frazer (1910) and Alfred William Howitt (1905) (who were also Sigmund Freud's main anthropological sources when he wrote his famous Totem and Taboo), of a strong correlation between totemism and exogamy. Totems are natural species (animals, sometimes plants) representing a specific social subgroup, as if the natural world, classified into species, was called in to naturally classify the community among separate groups. Both Leroi-Gourhan and Testart thus consider that Palaeolithic societies likely knew some form of 'classification', in Lewis Henry Morgan's (1871) use of the term (see also Radcliffe-Brown 1972.98-103;Bloch 1983.8-13), so as to ensure the practice of exogamy. And we saw that every classification ultimately and theoretically resumes as a two-fold division, that is, dualism.
In the absence of writing, how are people to express what matters most to them, that is, the (male) society, the (female) engendering principle and exogamy, if not by using elements with readily identifiable characteristics, such as, respectively, the horn, the circle and isomorphic forms?
It is remarkable how early anthropologists from all corners of the world like Morgan in North America, Marce Granet in China, Spencer and Gillen in Australia, Marshall Sahlins in Oceania, Claude Meillassoux in Africa or Marcel Mauss and Bronisław Malinowski in Melanesia recognized related practices of inter-clan exogamy, cross-cousin marriage or 'classificatory' structures to describe the internal organization of pre-state societies, an organization so different from their own Western 'complex' type of kinship (where marriage is practiced with the outside world, as opposed to closed-in and 'elementary', systems to use Lévi-Strauss terminology). This all-encompassing dualist classification appears as the principle according to which early human society coped with the distribution of sexual mates in order to ensure its own perpetuation. There is thus nothing surprising in finding it all the way to the Neolithic, before the Agricultural Domestic system altered it profoundly (infra).
There seems to linger in Western thought an ethnocentric reflex to consider pre-state societies as unfamiliar with elaborate forms of conceptualization (Asouzu 2007.192). The ethnography of hunter-gatherer societies largely suggests the contrary (Barnard 2020). It is much beyond the scope of this paper to develop the Palaeolithic symbolic world, but it was essential to show that the Neolithic dualist system is a natural offspring of a much more ancient and complex ideological background (Verhoeven 2001.84). We are now ready to investigate more precisely what this 'dualist scheme' is all about.

The 'primeval communist' social structure and its persistence in the Early Neolithic
The relations of production According to the analysis that Testart (1985) proposed of at least certain genuine (i.e. matrilineal) Australian tribes (in the southeast, especially), the hunter is not supposed to eat the prey he has killed, but to give it away to the community. "For example, in south-west Victoria, the hunter is said to receive nothing, and his brothers are treated in the same way (Howitt 1904.765)" (Testart 1987b.296). This is the basic opposition this author makes between 'primeval communists' and later hunters who usually distribute their prey according to a pattern which "leaves no doubt about the sharer's close kinship ties" (Bird- David 2019.17-19). Indeed, "possession of a kill in a hunting society confers not the right to its consumption but the privilege of performing its distribution" (Ingold 1980.158, citing Dowling 1968. Who, then, appropriates the prey among 'primeval communists'? It is often individuals belonging to the social group opposite to that of the hunter (Testart 1988.10). According to Morgan's (1871) 'classificatory system', these groups are defined by filiation and generation, thus grouping all siblings in the same class (Radcliffe-Brown 1952). Because of exogamy, the opposite moiety is the one where the hunter finds his spouse. His prey may then go to his spouse's parents. "Among the Ngatatjara, the parents-in-law take first and the brothers last. (…) Among the Maljangaba of New South Wales, the tribe is divided into matrilineal moieties and a man gets very little meat from his maternal kin because they belong to the same kinship group as he does. He receives much more from his father, since he is not maternal kin (Beckett I967. 459)" (Testart 1987b.296). There may be as many rules as there are societies, but it is significant that if the exchange of meat proceeds according to the kinship system among hunter-gatherers, it is in particular the non-producers who generally appropriate the product among 'primeval communists'. Generally speaking, the producer is never the consumer and the consumer never the producer (Testart 1987b. 294). Because the rule applies to every hunter, the latter eventually always gets his share, and, if the production is denied to the producer, it is, in the end, to the benefit of the society as a whole. Reciprocity as a rule of traditional economies is also well known in the ethnography of Melanesia (Malinowski 1926. 33;Mauss 1924) and elsewhere (Barnard 2020. 31), but it is in Australia that this form of exchange appears the most equalitarian. Comprehensive and equal internal cooperation has thus been identified as the dominant mode of production (Testart 1985. 115,169).
The universality of this rule can be questioned for the early stages of humanity, but the general paradigm seems to be that sharing follows the division traced by the kinship pattern in the social body (Speth 2010.xiiiv). This could help shed light on specific archaeological traces. For example, food exchange can be inferred from such data in the PPNA sites of the Northern Levant, where harvested grains are assumed to have been stored in communal buildings (Stordeur 2000.3;2003.20;, more or less symmetrically divided into two equal parts along the axis of the building (Stordeur et al. 2001.32-33, Fig. 5/1). Though this is nothing but the author's speculation, it could be that this symmetrical division of the village granary was made according to the kinship division within the community for matters of exchange (each subgroup producing and storing for the other). This division goes much beyond economic matters. At Faynan, in the southern Levant, the communal buildings are characterized by "general symmetry to the structure along an axis formed by a deep trough" (Mithen et al. 2011.354). Sometimes, such complementary division has been identified at the level of the entire site: "the small settlement at Qermez Dere had been laid out in two contrasting halves that performed complementary functions. Part way through its life, the village was re-formed, but once again in two complementary halves" (Watkins 2006.16). Leaving aside the case of Asıklı, which has a street dividing the village (Özbasaran et al. 2012. 140) but in less clearly symmetrical parts, Hodder (2012.304) identified such an arrangement at Çatalhöyük, with "a large dip or trough across the middle, dividing it into two hills. The mound does seem to have developed in two halves (north and south). (…) In addition we have found some differences in the genetic make-up of the humans buried in the two halves". Finally, in the late Neolithic (but in fact, contemporary with the local emergence of the farming system), Ulf-Dietrich Schoop (2005.49), concerning the "lines of parallel houses facing one another" in Hacılar and several other western Anatolian Late Neolithic sites, writes that "this brings to mind the social organization known in the ethnographic record as the 'moiety system', in which a community views itself consisting as two competitive 3 3 halves. I do not wish to elabo-rate on this, for at the present state of investigation it would only be grasping at straws" 4 4 . This precaution appears academically wise, but the straws grasped here are arguably nothing but the very root of prehistoric ideology.
A slow change can be detected in the following PPNB period at Çayönü, with granaries attached to every house from the Grill Building phase onward indicating "that economic emphasis may have been shifted from community to family based production and consumption" (Yakar 2003.442). This reorganization indeed seems to reflect a slow trend towards an economic (and marital) autonomy of lineages, following a "segmentation and separation of balanced components arranged in relation to each other" (Hodder 2020.49-51), and possibly leading in late Neolithic Çatalhöyük to "the House as a historical and genealogical social unit" (Kuijt 2018. 584; infra) based on a line of ancestors. This late Neolithic fission into autonomous families apparently emphasizes, by contrast, the closed-in reciprocal pattern that was arguably still strong in the earliest Neolithic.
This economic evolution also seems supported by the genetics of wild food. The PPNA plant material, though anthropologically managed, is not morphologically domestic yet. The subsequent physical domestication, eventually including animals, implies that the originally loose farming mode of production is, relatively speaking, gaining in intensity during the PPNB (Zeder 2011.230;Willcox, Stordeur 2012.112;Asouti, Fuller 2013.329). This trend goes well with the idea of weak communal production (PPNA) gradually intensifying (PPNB) towards the specific interests of each lineage (PN), in particular for the constitution of bride-prices (herd animals), suspected elsewhere (Bodet 2019b) to have begun in these latter periods. This morphological evolution of resources emphasizes, by contrast again, the economically loose, reciprocal and equal form of food exchange expected to characterize the earliest Neolithic groups.

The relations of reproduction
Economic reciprocity appears to reflect marital patterns. Among hunter-gatherers "marriage prescriptions commonly involve real or classificatory crosscousins and (…) exchange between two kin lineages" (Walker et al. 2011.4). The parents of 'cross-cousins' are the children of a brother and sister; because the social affiliation comes from either the father or the mother, cross-cousins necessarily belong to a different 'class' and are expected to marry. This is just one straightforward example; there are many possible types of marital alliance, with many more subgroups but the founding exogamic principle remains everywhere the same: the hunter does not 'consume' his sister but the sister of a hunter from the opposite moiety (hence the famous 'exchange of sisters').
Close-kin mating is often thought to have been prohibited to lessen biological complications coming from consanguine mating, but there is no genetic difference between a cross-(prescribed mate) and a parallel-cousin (proscribed). The key point is that healthy mate circulation in the long term implies not only separate groups (at least two), but reciprocity among them. It is obvious that this social division is not prompted by emulation or rivalry but, quite on the contrary, by the welfare of the entire society. "A social benefit results from an exogamic marriage (…), the law of exogamy is omnipresent (…). It is the archetype of all other manifestations at the basis of reciprocity, it provides the fundamental and immutable rule ensuring the existence of the group as group" (Lévi-Strauss 1967.551). Exchange and the social classification permitting it thus appear again ultimately as modes of (re)production. The tight infrastructural correspondence between the kinship system and the economy seems to strengthen the fundamental role played by reciprocity and dualism in the ideology (Lévi-Strauss 1967.48-170;Bloch 1983.9-10).
As for the reflection of these principles in archaeology, the internal subdivision of the PPNA communities suspected above may be continuing in PPNB Çayönü, where the two large buildings just north of the Plaza could house the elder(s) of each moiety. Another contemporary hint is found in Nevalı Çori with "two groups of houses with different orientation (…) that could have belonged to two groups of families with different lineage" (Yakar 2003.443), two groups expectedly related through permanent intermarriages. Whatever the case, all hunter-gatherer societies seem to know one form or another of kinship classification (Ghasarian 1996.31;Walker et al. 2011.1): there seems to be no viable reason not to expect a similar system of reciprocal mate exchange among Neolithic communities. And this could be what the twin steles state out-loud.

Totemism
Some form of totemism seems rather common among hunter-gatherer communities, though its absence among the San shows it is not universal (Testart 2006c.149;Barnard 2020.46). It is nevertheless thought by a number of specialists cited by Freud (2010.42, note 2), in particular Frazer (1910), to have been very widespread at an original stage which would correspond to 'primeval communism'. There is, again, no a priori reason to exclude its presence in the Neolithic, as strongly suggested by the twin stele arrangement. The totem consist of natural species, usually an animal, considered to be the ancestor of a clan or tribe. Individuals maintain a very specific relationship with it, being strictly forbidden to consume it, except once a year during the ritual known as 'Intichiuma', aimed at magically increasing the totemic resource for the opposite moiety to consume (Spencer, Gillen 1899.169). This is the symbolic projection of the rule of reciprocity reviewed for production (the hunter not eating his prey but hunting it for the other moiety to consume) and for reproduction (individuals not marrying within their own subgroup, but 'producing' children to be 'consumed' by the opposite one). By way of animal species, totemism can be conceived as a symbolic representation of the dualistic kinship system ongoing in the community (Freud 2010.203-204, 255-256).
Göbekli Tepe, sometimes seen as a place of interclanic reunions for very extended kinship groups (Schmidt 2001.52-53;Belfer-Cohen, Goring-Morris 2002;Peters, Schmidt 2004.210-212), would fit well as a place where Intichiuma-type ceremonies were taking place. In fact, if contemporary levels are to be found, the high number of large early Neolithic sites, like Karahantepe, recovered within a radius of about 20km all around Göbekli Tepe (see below), may have composed this population (Bodet 2019a).
For Hans Georg K. Gebel (pers. com.), "the ideology of the early Göbekli Tepe Culture represents a symbolically sustained system needed to serve the integration of growing group numbers (…). Mutual understanding and conflict management of groups not knowing each other were reached by commonly accepted strong and binding ideologies and conventions. One may speak of ideocratic territories mediated through the fixed image programs". A common ideological background indeed certainly played an important binding role among all these communities, and, looking at the homoge-nous symbolic program on the entire site, this binding element must have been related to an isomorphic, i.e. dualist conception, shared long before the construction of Göbekli Tepe.
The Neolithic iconography could fit neatly with the totemic analogy, by supposing, like the excavators of Göbekli Tepe and other specialists (Peters, Schmidt 2004.209-212;Kornienko 2018.17-18), that the animals carved on the steles represented totems. However, for Forest (2006.134) this is probably not the case, because these elements interfere with each other, being complementary or synonymous, so as to convey a message, while totemism simply classifies in purely equal terms. Totemism and message/law indeed stand at different levels in the ideological structure, but are not at all incompatible (Testart 1985.510). The numerous animals represented on the side steles could fall under Forest's warning (Bodet 2021), while the few ones on the central steles, absolutely alone, may be more in tune with a totemic classification. Though intrinsically related in form and signification, we will see that side and central steles may hold a different symbolic value.
The rich animal repertoire represented on the side steles is very similar to that deciphered by Forest (1993; at Çatalhöyük: bulls, predators, psychopomp birds, found to symbolize the society, death of the body and transportation of the soul, respectively. Göbekli also has many snakes, understood as lineages (Forest 2006). The same geometric elements, thought to symbolize the two moieties or the sexual mates they exchange, are also found on both sites (and many others): parallel lines, zigzags, triangles, and chevrons. The several side steles could then represent the subgroups the community is composed of, linked to each other by the circular stone wall, forming a large matrimonial self-sufficient unit. The non-totemic (message-delivering) symbolic animals carved on them would suggest the endless (feminine) cycle of life (snakes) and death (birds, predators) in which these groups were involved (Bodet 2021). On the other hand, if the central parallel steles stand for the dualist subparts of the (masculine) society (product of the feminine principle), the isolated animals carved on them may indeed be totemic. For example, the reflecting foxes on the twin steles of enclosure B (Peters, Schmidt 2004.184) would represent the two inseparable but distinct subparts of the same totemic clan (see Malinowski 1926 for compatible ethnographic examples).
A look at a tightly interwoven subject -mythologywill allow a more comprehensive understanding of the dualistic nature of totemism.

The mythology
The species included in the totemic partition have a correspondence in the mythology. The main characters of the Australian Aboriginal myths of the Dream Time (Testart 1978;1985.390-395), with a correspondence worldwide (Girard 1972), are generally divided in two types: the violator and the counter-violator. For anything to happen in the founding myths, the fundamental rule (exogamy) must be violated, which invariably entails a counter-violation. The widespread myth of the eagle (the hunter, the creator, the counter-violator) and the crow (the scavenger, the trickster, the violator) illustrates this point. The crow steals the fire from the eagle (violation), and, as he escapes, he drops the fire, allowing humans to capture it and cook food. The eagle avenges himself by causing a huge fire that threatens humans (counter-violation). Some variants of this myth are about stealing water (violation) and so permitting life but provoking floods when uncontrolled (counter-violation). Myths thus explain the origin of the society as an interaction between two opposite but interdependent poles (Testart 1978.95, 118-125;1985.384-387,432-444). Just like later religions, mythology aims in the end at securing the social order through ideology.
Forest (1993.17-21; 2006.134) reads the elements in the iconography of Çatalhöyük as principles conveying a message: life and death on a vertical line, two exogamic moieties on a horizontal one, all intersecting in a cross pattern to permit the existence of the community, thus recalling its fundamental rule (exogamy). This message is essentially the same as the one present in myths. It is open to question whether the animals in the iconography, strikingly similar in Göbekli Tepe and Çayönü despite a wide chronological and spatial gaps, directly represent mythical characters. But in the end it matters little, since they are likely to stand for the same opposition among interdependent elements (as an image, again, of the subgroups of the society). Just like in many myths of the primeval world, a dualistic opposition can be suspected with a certain degree of confidence among the antithetic heroes of Neolithic mythology, symbols of a fertile opposition.
René Girard (1972.88-95) presents a somewhat different interpretation of early myths and rituals, focusing on sacrifice as catharsis, expelling the tensions accumulated within the community on an innocent victim, but the ultimate goal remains to prevent the 'divided' community from the risk of becoming "undifferentiated". Jungian psychology also notes the case of twin snakes in mythology. "These are the famous Naga serpents of ancient India; and we find them in Greece as the entwined serpents on the end of the staff belonging to the god Hermes 5 5 . An early Grecian herm is a stone pillar (…). On one side, are the intertwined serpents (in the act of sexual union) and on the other an erect phallus: we can draw certain conclusions about the function of the herm as a symbol of fertility. (…) But Hermes is (also) Trickster (…) the leader of souls to and from the underworld" (Henderson 1964.155). The two snakes appear as the dualist lineages whose union alone can engender society.

The blood ideology
Based on ethnographic data, Chris Knight et al. (1995.89,93-97) have proposed a Palaeolithic "symbolically structured sexual division of labour" where, notably, the recurrent use of red ochre would be utilized in menarchal rituals to symbolize fertility. This interpretation could fit well with Testart's (1985.345-475) reconstitution of the primeval mind, according to which the female compensation for the masculine blood-soiled meat brought by the hunter is the feminine blood-soiled newborn child.
The widespread presence of red paint on the floor of special buildings all through the Neolithic of southwestern Asia (Gökce 2021) supports the idea that blood played a fundamental symbolic role. At Çayönü, actual traces of human (and animal) blood have been detected on a one-ton slab in the courtyard of the so-called 'Skull-building' (Özbek 2004.20). Given what is known about these societies, notably the classification according to generations, it seemed likely to Forest (1996b) that this blood was that of initiation rites. The blood of circumcision (symbolizing the first hunt?), equivalent to that of the (first) menstruations (Doyle 2005.280;Knight et al. 1995. 95), can be understood as a separation between two crucial statuses: not simply synchronically between male and female (or between their respective moieties), but diachronically between consumers (children) and (re)producers (adults).
Through its intimacy with both life and death, blood is thus considered by Testart to have played a synthetic role in the primeval communist ideology. A symbol of order and life when running in the closed system of the veins (=exogamic rule respected), blood represents chaos and death when running out of a disrupted vein system (=exogamic rule violated). Just like for the mythical figures (divided into violators and counter-violators), for the natural species in totemism (divided among social subgroups) and for society (divided into parallel moieties), blood, the one and same blood, is artificially divided and separated into distinct but mutually interdependent classes so as to promote their mutual interdependence and strengthen the unity of the whole.

Synthesis: the dualist scheme
Because the data mobilized here is not archaeological in nature, it is perhaps not superfluous at this point to synthesize what we have proposed. The kinship dual classification of the society is reflected in the economy and ideological structures, such as totemism, mythology and blood ideology. Beyond a mechanical Marxist view that would present the social superstructure as invariably determined by the economic infrastructure, kinship thus appears to dominate the primeval communist relations of production, making the reproductive infrastructure the very root of dualism. Yet, as Alan Barnard (1978.78-79) writes, "Australian systems differ from other universal systems in that Australian universality is not confined to kinship", it is "closely connected with totemism and with other aspects of cosmology" (which) "divide the universe -nature and culture alike-into named categories [which] represent a concept of world order in which kinship is only a part". It thus seems that what determines kinship and all other structures is, in fact, the exogamic principle which must be conceived as an overarching pattern imposed on the entire social fabric. Opposing sets of the natural world like male/female, sun/moon, winter/summer, water/fire, dry/wet or life/death are 'given' to humans, who use them as symbols to express and justify the only opposition on which they have a hold, the division of society itself into exogamic lineages or moieties. In other words, just as in later religious systems (Forest 1996a), the cosmos is mobilized to promote, through its own perpetual and fecund oppositions, the perpetuation of the fragile opposition between lineages or sub-clans in order for the society as a whole to reproduce. It is this all-encompassing exogamic division that is termed the 'dualist scheme' by Testart (1985.207-218, 477-515), and which, as we intend to show, was still very vivid in the Neolithic.

The Neolithic dualistic symbolism
Two parallel steles seem to be the symbol chosen by Neolithic people to represent the concept of exogamy/dualism. Given the pervasive twin steles present in a large number of sites (Bodet forthcoming), it appears that early Neolithic communities felt the urge, maybe more than Palaeolithic ones, to recall and impose this principle. Is it possible to be more precise as to the message conveyed and to the cause of the monumentality given at this precise moment to an immemorial principle?

The stele as a symbol of the lineage
In the communal building of Nevalı Çori (devoted to reunions, given the bench running at the base of the surrounding wall), Alexis McBride (2013.54) proposed that the anthropomorphic steles inserted in the bench are 'participants' along with the real humans seating there. This makes sense indeed, but the verticality of the stele must be meaningful. According to several researchers, like Forest (1993.7), Tatiana V. Kornienko (2018.17) or Christian Jeunesse (2020.54), the stele stands for a genealogical line of ancestors related over time, that is, a lineage, built up generation after generation. This interpretation is well supported by ethnographic observations: in the American North-West coast, the "totem poles, house posts, memorial posts (…) record the household's lineage" (Banning 2011.626). The endless continuity in time of the lineage could be the reason why certain Neolithic steles are reused at the same location phase after phase (Watkins 1996), while others are buried or ritually broken.
Seeing side steles as lineages implies that each circle could represent a larger social group, like a tribe or clan, divided into a number of subsections, an idea already alluded to, in one way or another, by several authors (Belfer-Cohen, Goring-Morris 2002;Yakar 2013.438;Hodder 2020.50). Beyond the representation of the cycle of life and death as suggested above, the surrounding wall of the Göbekli Tepe enclosures could bind the lineages in an endless circle of marital exchange, a stone materialization of a closed-in 'generalized-type' of kinship pattern 6 6 (Lévi-Strauss 1967), where lineage A gives a mate to lineage B, B to C… back to A (Bodet 2012). There would then be between four and twelve subsections for each tribe; interestingly, twelve is also the number of subgroups chosen by Freud (2010Freud ( [1915.51) to present a typical totemic society. The interconnected side steles would then stand, like a temenos wall, as a transition between the real world and the sphere of pure abstraction, which, we will now see, seems to take stage at the centre of the circle.

The reciprocal relationship
Central and side steles have the same monolithic structure and same morphology, the latter being simply smaller and less well executed. For the symbolic program in question, they are likely to be a related signifier standing for a related signified, but on a different scale. If side steles may represent the actual lineages making up the community, the central twin steles would then represent two parallel lineages, but on a purely conceptual level. They seem to stand for an abstract idea, an allegory of the exogamic rule, the active (feminine) principle of fertility. Parallel, isomorphic and face-to-face, the twin steles seem to express a bilateral relation of strong symmetry, where each subgroup is at the same time the donor and recipient of a (marital) transaction. Behind the exogamic rule (Forest 1996b.29), the composition seems to express the type of social relation that exogamy entails, one of pure reciprocity. As Lévi-Strauss (1967.97) puts it, the dualist principle is itself only a modality of the principle of reciprocity.
This emphasis on the relation itself appears hindered precisely by the fact that the twin steles do not generally enter in relation with each other, just like parallel lines painted on floors (also recalling lineages). The relation between the steles is suggested elsewhere: a low bench, a slight clay lip or a slab set on edge at Qermez Dere, Beidha, Çayönü (Skull Building), Musular and at late Göbekli Enclosure A (E PPNB) (Watkins 1996;Makarewicz, Finlayson 2018;Erim-Özdogan 2001.208;Özbasaran et al. 2012;Schmidt 2001.50 respectively). In the enclosures of Göbekli III, twin steles can be said to be connected by the ground, Mother Earth, of which we saw the importance for the concept of fertility. But there is more. The lack itself of any obvious connection between the parallel steles implies exactly the contrary: perpetually reflecting each other (as well as the -totemic? -decorations carved on them), one is nothing but the permanent counterpart of the other, each stele fundamentally dependent on the other to exist. The intrinsic relation between the two steles is conspicuous by its absence.
The composition seems to be bluntly saying: a moiety is the mirror of its counterpart, and only the two together, as equal partners in the (marital) exchange, can engender the society altogether, which, in turn, can only exist divided into equal subparts intrinsically bound to each other. But if twin steles represent an abstract sphere of symbolization with such majesty, it is exactly because this notion is not simply an allegory: reciprocity must be conceived as a law governing social conduct. In other words, the full message conveyed by the twin composition appears as such: 'marital reciprocity must be respected for the sake of the whole society'. We will later see why this antediluvian rule took such a 'monumental' urgency in the Neolithic.

The twin steles as a symbol of fertile regeneration
Let us first complete our reflection on the striking fact that the most central place of the entire composition at Göbekli Tepe (see also Jerf el-Ahmar) is the space left 'religiously' empty between the twin steles (Fig. 4). This vacant space must have been filled with meaning in the eyes of the audience. We saw that at Çatalhöyük two symbolic pillars are connected by the limbs of a feminine figure giving birth to a bull (which is a good enough reason to see this representation as a metaphor and not as a realistic scene). Again, according to Forest (see English summaries in Bodet 2012.7 and2021.149-151), this is not a woman, but the personification of the principle of regeneration, engendering the society. On the same line of thought, the space between the twin steles at Göbekli can be understood as a threshold to life (a symbolic vulva?) and death (a symbolic 'swallowing' mouth), because it is a representation of the vagina dentalia, the 'toothed vagina', a widespread mythological female principle of both regeneration and destruction among traditional societies (Forest 1993.22;Ross 2021).
This invisible principle of the regeneration of society and all life forms would then be put into action by the mutual interaction existing between the exogamic moieties represented by the two monoliths. This is a practical illustration of Trevor Watkins' (2006.21-22) statement that "architecture is a specially powerful mode of external symbolic storage". And these symbols convey a specific me-aning. The Neolithic twin steles seem to state: 'the eternal cycle of life and death can only be put into action by the principle of pure reciprocity'. Such prehistoric capacity of abstraction can only be a surprise to ethnocentric prejudices. McBride (2013.59) further suggests that the participants in a ritual or ceremony may have been asked to walk through the central steles at Göbekli Tepe. The idea deserves attention. This particular space, here putatively interpreted as the principle of regeneration of both life and death, would indeed be the ideal place to have adolescents pass through during their initiation ceremony, initiation being conceived in many societies as the death of the child and rebirth as an adult (Weiss 1966.72;Henderson 1964.120-121;Forest 1996b.28). According to Max Weber (1920), initiated aristocrats in China or India call themselves 'the twice-born'. The link between dualism, totemism and initiation is further supported by the fact that in order to become hunters, young Aborigines are systematically initiated by the opposite moiety. The same holds true for funerals and Intichiuma ceremonies (Testart 1987b.299).
Regeneration may also be the main principle displayed in the Franco-Cantabrian caves (also an ideal place for ceremonies of initiation). In that case, the cave paintings would 'magically' assist these principles by being represented in the 'womb of Mother Earth' (Henderson 1964.146-153), the latter also being a recurrent theme in Carl Jung's 'collective unconscious'. An analogy can also be made with the Turkish custom of Hıdırellez, still performed today, where the drawing of babies and cradles on the sand or earth in springtime (when nature comes back to life) is believed to enhance fertility 7 7 . The belief that the mere representation of symbols has a 'magical' active power is universal: it is the same when Christians hang a cross in their homes or Muslims a picture of the Mecca. This (and probably not a belief in a deity) would also explain why so many crude female figurines (representing the same allegory of fertility) were so common in domestic contexts in the Levant or Central Anatolia (Cauvin 1997.46-49), as well as many hand-size T-shaped figurines in the Urfa area (Hodder 2020). This same belief could finally explain why symbolic enclosures in sites like Göbekli Tepe, Nevalı Çori, Sefertepe, Karahantepe (Moetz-Çelik 2012.699) and as far away as Çayönü, Qermez Dere (Upper/Middle Tigris basin), Jerf el Ahmar (Middle Euphrates) and Beidha (Jordan), were carefully buried upon abandonment 8 8 , arguably so as to preserve the active principle of fertility they enclosed.

Epilogue: a likely cause for the monumentality
This investigation must now be placed in its socioeconomic context, thereby answering a last important question: why such monumentality, especially if, as discussed above, the displayed concept had been a basic one for tens of millennia?
We must come back to the idea that in spite of its 'universality' (Malinowski 1926;Lévi-Strauss 1967. 3-29,49), the division of society into subgroups is not founded in nature, it is a social product; this implies that nothing can physically guarantee the respect of exogamy, and that its importance must be permanently reinforced in the community, in particular to the newly initiated generations.
In hunter-gatherer societies marriage is not left to the free-will of individuals; it is codified by tradition and contracted among more or less closely related individuals, like cross-cousins (Lévi-Strauss 1967;Walker et al. 2011;Barnard 1978;Ghasarian 1996. 147-174;Bird-David 2011). The distribution of sexual mates is thus regulated so as to avoid a dangerous anarchy for the entire community. Once settled, communities naturally continued this immemorial tradition of 'prescriptive' (or pre-arranged) mating, every new generation being bound to stay within the village so as to comply with this systematic exchange. Coupled with the fact that settled life and farming naturally lead to a strong demographic growth (Bellwood 2005.61-64), the consequence of this alliance system is that this growth is largely local. This age-old inward-looking 'elementary-type' (whether 'restricted' or 'generalized') of marital alliance rule (i.e. among the subgroups of the tribe) is most likely the ultimate cause for the appearance of Late Neolithic mega-sites like Çatalhöyük, Halula, Ain Ghazal, Shu'eib or Basta (Forest 1993;Bodet 2019a; this probably goes also for Neolithic megasites elsewhere like the Trypillian sites). Thus, if at Göbekli the symbolic emphasis is monumentally placed on the community altogether, in the much later Çatalhöyük horizon the same concern gradually shifts towards the intimacy of the (autonomous) lineage itself, or "multiple single-family households", to use Kuijt's (2018.565,584) words.
Such evolution did not go without problems. In farming families the elder son traditionally inherits from his father not only the estate, land and animals, but also a decisional power over his younger brothers and sisters, especially in terms of alliances. This leads to a growing internal stress with younger individuals searching to withdraw from the domination of their elders by splitting from the group, a situation well described in comparable ethnographic cases (Sahlins 1961.324-327;Meillassoux 1991.51-52, 122-124). For Çatalhöyük, Forest (1996b.5) devised a similar incongruous situation, all the more so that farming allows for (and is much more efficient with) small producing units (nuclear families) spread over the landscape, each family/farm on a separate piece of land. The resolution of this inextricable situation had to wait for the abandonment of the prescribed 'elementary-type' alliance system (among related individuals). This is indeed what seems to have happened nearly everywhere by 6500/6000 BCE at the latest, as suggested by the (gradual, then total) desertion of all mega-sites, followed by the establishment of gradually smaller farmsteads spread around, wherever land allowed for farming and herding. Huge Çatalhöyük East thus gives way to relatively smaller Çatalhöyük West (though this site still remains rather large); relatively small Musular, founded towards the end of large Asıklı, may represent an earlier (Late Neolithic), because eastern, example of the same process of site segmentation. This trend will continue throughout the early Chalcolithic, reaching its apex with the Halaf culture (Forest 1996a. 27-35).
The expected tense social context just referred to is crucial to understand the monumentality of the steles of Göbekli Tepe. In charge of maintaining the moral conduct of the society, the elders are constantly recalling, imposing and teaching the new (initiated) generations about the old traditions, in particular the reciprocal exchange of mates. They would have been particularly anxious and careful to avoid any disruption in the smooth circulation of women, and therefore, to counter, with the help of symbolism, the splitting of younger people from the larger kin-group. Indeed, the elders of each social subgroup, responsible for giving mates to the other subgroup(s), could not fulfil their duties if these younger individuals had left. From their point of view, such a fission would invariably lead to an outbreak of the inextricable internal violence that is known to have scared these societies so much (Girard 1972).
The fear of the loss of reciprocal exchange and of the cohesion of the community is hypothesized to have pushed the elders to express the old rule of (exogamic) alliances with much force and prominence by ordering the erection of the monumental twin steles. It is in this sense that we propose to illustrate Thomas Zimmerman's (2020.14-15) intuition that the symbolic program of Göbekli Tepe reflects much more a Palaeolithic cultural collapse than the advent of a new one, the way Cauvin (1997.50-55) sees it. "It is doubtful that the supernatural worldorder envisaged by earlier hunter-gatherers would have been entirely altered by new spiritual concepts"; (the Neolithic) "repertories of symbols (…) seem to have their origins in earlier periods" (Yakar 2005.111).
"The establishment of such symbolic systems, or the externalisation and canonisation of symbols, is not the result of a cognitive process but rather the result of a basic need, the need to sustain a current life mode by coping with newly arising social and ideological challenges of fast growing social aggregates in the Upper Mesopotamian grasslands": this statement of Gebel (2013.40) applies very well to our views, provided that the "basic need" in question is first and foremost that of a sustained marital alliance system.
The monumental isomorphic steles understood as an enforcement of respect for the old reciprocal alliance rule, can thus be seen as a form of propaganda, erected in the face of the threat of being abandoned at a time when Agricultural Domestic lineages were gaining economic and marital autonomy. These 'monumental' fears were indeed justified as, in spite of all these efforts, elementary alliances will prove obsolete by the Chalcolithic. This necessarily implies that nuclear families (a married couple and children), breaking free from their larger family groups, proceeded to 'complex' types of marriages taking the form of 'contracts' (hence called 'alliances') among unrelated larger families, probably secured by material transactions like the bride-price (herd animals) (Bodet 2019b). In such a context, exogamy became reduced to the prohibition of incest (Ghasarian 1996;Forest 1996a) and dualism naturally lost its ground as a principle of alliance together with its ideological relevance. In Gebel's words (pers. com.), "dualist schemes may even become extinct in early productive environments when strong relational ordering principles help or suffice to organize lineages and the societies they are part of".

Synthesis: Dualism as a Neolithic scheme
By shedding light on the archaeological data using ethnographic social structures, we have here attempted to review how the Neolithic revolution transformed the Palaeolithic society into an Agricultural Domestic one. We were greatly helped in this task by the Neolithic symbolism on which social changes were invariably projected. This structural evolution is synthesized in Table 1 and Figure 5.
We started our investigation on the premise that hunter-gatherer societies do not marry with outsiders and are, as a rule, divided into (at least two) subgroups (moieties or lineages) as a direct outcome of the universal rule of exogamy so as to secure the distribution of mates and reproduction of closed societies: the hundreds of early societies reviewed in The Elementary Structures of Kinship by Lévi-Strauss (1967) as well as general handbooks (Ghasarian 1996) or articles (Barnard 1971;Walker et al. 2011) on early kinship make this point clear. This seems to apply during much of the prehistoric period, as symbolic representations in the depths of Middle and Upper Palaeolithic caves seem to suggest. Throughout the Mesopotamian and Anatolian Neolithic, this tradition continues with an impressive series of isomorphic representations, in particular twin steles. The message seemingly conveyed by this dualist symbolism can be read as follows: 'only the reciprocal (marital) relationship ongoing between the moieties (lineages) composing the society can allow for the society to reproduce safely'. Beyond the kinship pattern, reciprocity encompasses all other as-pects of the hunter-gatherer social structure, first and foremost its economy. Testart has therein come to the conclusion that dualism generally represents a fundamental scheme determining the structural and ideological composition of pre-state communities. It is suggested here that this dualistic ideology finds its most phenomenal transcription in the monumental central twin steles of Göbekli Tepe, at a time when it was in danger of being supplanted by the advent of a whole new social and ideological order, the Agricultural Domestic System, founded on autonomous (unrelated) agricultural lineages organizing marital alliances (with bride-prices) freely among themselves. As agriculture diffuses towards the west, and is appropriated by local hunter-gatherers, the same emphasis on dualism appears, this time on a more modest scale but widespread in every domestic context, as in Çatalhöyük. This latter social system (lineages, autonomous in marital terms) was destined, a few millennia later, to aggregate hierarchically into city-states on the pattern of the status differentiation between elder males and women/youngsters in the agricultural family itself (Forest 1996a;Meillassoux 1991).
According to the dictionary, a scheme is "a large-scale systematic arrangement for attaining some particular object or putting a particular idea into effect" 9 9 . For hunter-gatherers, the 'arrangement' is the reciprocal partition, and the 'object put into effect' the regeneration of the community. The symbolic repertoire of Göbekli Tepe and of a large num-ber of sites up until Çatalhöyük could show that this scheme may apply all the way to incipient farmers.
It is now possible to clarify a theoretical problematic raised above, and state that there is no coincidence in recognizing dualism in the ideology of societies as geographically, chronologically and culturally distinct as the Neanderthals, Australian Aborigines, Magdalenians, Ancient Chinese, SW Asian Neolithic or early historical Mesopotamians. This convergence becomes structurally logical when the ideology of these societies is ultimately determined by 9 https://www.encyclopedia.com/science-and-technology/computers-and-electrical-engineering/computers-and-computing/scheme

Fig. 5. The Neolithic Dualist scheme seen as an evolution of social structures (synthesizing graphic). Dualist and closed-in ('elementary') Palaeolithic societies confronted with growing autonomous Neolithic farming lineages trigger an ideological conservative reaction (monumental: Göbekli III, omnipresent: Çatalhöyük), but finally evolve towards open ('complex') Chalcolithic Domestic lineages.
strictly equal relations of (re)production (compensating for low productive forces). The human constitution, physically weak but with a very high potential for intra-specific communication, is such that the survival of the species is mostly dependent on the process of exchange between individuals. This perpetual need to distribute food and especially mates within the subgroups of the community and across generations required the adoption of a form of social conduct based on pure reciprocity. Dualism appears as the ideological result (and not the cause exnihilo) of this chain of causal factors. This is the reason why dualism should not be seen as a mere cultural tradition which, among others, would have been miraculously preserved until the Neolithic. It can be presumed to have been 'socially selected' in the Darwinian sense of the expression, that is, unconsciously over countless generations, for having provided humanity with the highest, maybe the only, probability of survival.

Evaluation
It should be noted here that Testart is suspected of having somehow distorted the ethnographical facts, although this is certainly due to the goal he set himself: not that of describing specific communities, but instead uncovering the purely theoretical structure that binds them all. He was thus able to reconstruct a coherent social system where every structure (production, reproduction, ideology) is absolutely in tune with all the others. The best clue to support the coherence of this reconstitution is that everywhere societies reproduced successfully the hunting-gathering way of life throughout the entire Palaeolithic period (one to two million years?). We saw how this success was achieved through the total annihilation of individual interests to the benefit of the whole, and that dualism has been identified as the keystone of this remarkably stable social construction. But the ultimate illustration for the central position held by dualism in prehistory, and without which dualism would have never occurred to the author's mind as a way to enlighten the Neolithic ideology, is the set of central monumental twin steles of Göbekli Tepe.

Conclusion
The following quote on the Australian Aboriginal social and ideological structure seems appropriate to conclude this paper: "to affirm that appropriation is the fact of the community as a whole only, to affirm that the latter is an inseparable totality, it was first necessary to break it into two, into two parts each closely dependent on the other. Each part was conceived from the start as part of the whole" (Testart 1985.478). It is remarkable and fortunate that these lines were written about a decade before the excavation of Göbekli Tepe. Today, the twin steles stand as a monumental confirmation of Testart's (1988) audacious intuition that Australian Aborigines likely reflect an extended Palaeolithic ideological background.