As presented, Sneath’s “Tribe, Ethnos, Nation” and The Headless State, which it epitomizes, offer a salutary corrective that deconstructs the idea of “tribes” as preexisting and autochthonous kinship groups that in time evolved to form nations and states. But in replacing them with their functional equivalents of aristocratic “houses” in his intellectual schema, he may risk underestimating the roles that kinship beliefs had in shaping social formation and vice versa as well as, through limiting the debate to the elucidation of social entities or states, miss an opportunity for a more thoroughgoing comparative analysis of the ritual and discursive practices that contributed to and maintained various types of social bodies.
The notable penchant of Western academic writers to argue their positions ab originis (as they think) by invoking as authorities ancient Mediterranean authors or works about those societies appears even in discussions that bear principally on Eurasian pastoral nomads. Yet this particular debate seems to take as its point of departure not the extensive classical Greco-Roman ethnography of pastoral nomads but rather the more abstracted theoretical musings of someone such as Aristotle whose Politics offers a comprehensive set of claims regarding the social evolution of city-state society. According to Aristotle (Politics 3.1280b), the genos or clan constituted a primordial social unit formed from the agglomeration of individual households or oikiai; these households and clans in turn coalesced into civic community in a process of synoikismos as part of man’s pursuit of the good life that only a polis can offer. Such an evolutionary and accretive scheme of civic formation gained credence as a way to conceive of societal change and teleology, though perhaps more among modern political theorists than historians. Sneath therefore seems justified in devoting a portion of his analysis to the debunking of the idea that ancient Greek and Roman history offers a convincing case of tribes as natural kinship-based social units that evolved to form classical city-states. He adduces as support for his own view the observation that there has been a recent fundamental shift in classical studies whereby a previous scholarly view of clans and tribes as the building blocks of city-states, poleis and civitates, has been superseded by more nuanced reflections on the historical dynamics of state- and tribe-formation. Thus, whereas influential writers a century or so ago, such as Numa Denys Fustel de Coulanges, deployed an interpretation of archaic Greek and Roman society as the intellectual foundation to elaborate broader theories about the development of the ancient city, historians of antiquity in today’s world – and stretching back now more than one generation – tend to take the inverse intellectual position and rely increasingly upon anthropological and sociological studies of other societies to inform their own approaches to the question.
“Les notions de tribu, tribalisme et société tribale ne sont plus maniées qu’avec réticence par la plupart des anthropologues contemporains, dont certain même enclins à les rejeter complètement…” – thus writes Roussel in his refutation of “essentialist” interpretations of the genos, phratry, and phylon.^[1] Roussel’s study is highly relevant in that it systematically reexamines the extant source evidence in light of a sophisticated approach informed by a post-Durkheimian methodology (to which we shall return) and also in that it specifically offers a thorough refutation of the thesis of Lewis Morgan, whose work was fortuitously translated into French in 1971.^[2] But indeed Roussel might have been too modest regarding the intellectual contributions that ancient historians themselves made to the critical understanding of these matters, just as Sneath has underestimated how, for more than a century already, classical scholars have been expressing opinions sympathetic to his own position. In 1901, Szanto argued in his Die griechischen Phylen that the Dorian and Ionian Greek tribes should be regarded as artificial sociopolitical units created in the course of common large-scale undertakings such as conquest, land division (as in the case of the Dorian tribes), and overseas colonization (as in the case of the Ionian tribes), concluding

“…dass die dorischen Phylen so alt also die dorische Wanderung sind und daher bei allen Doriern vorkommen, dass die ionischen Phylen altattische sind und nur vereinzelt im ionischen Colonialgebiet adoptirt wurden.”^[3]

Thus no longer should the Dorian tribes in the Greek historical period, for example, be described as the direct descendants of invading Indo-European tribes that ended Bronze Age Mycenaean palace-cultures in the course of their cataclysmic migrations. Rather, historians now conceptualize the process by which such “tribes” came to be formed in more complex and nuanced ways, akin to how historians of early medieval Europe have revised the traditional narrative of “barbarian invasions” to one of longer-term cultural synthesis or ethnogenesis, a project ably promoted by Wolfram and Pohl in Vienna and made even more relevant to an Anglo-Saxon audience by Geary in his Before France and Germany.^[4]In such views, all references to the early tribes ought to be read as invocations of past sociopolitical entities that had in reality always shifted in composition and meaning during long centuries of change and adaptation.
Ancient Greek and Roman tribes are likewise no longer widely regarded by scholars as a priori kinship groups that predated nations or civic communities. The historical record is full of cases of the wholesale invention of new tribes (phylai) and systems of tribes as part of civic self-transformation, which in turn appears to demonstrate Sneath’s thesis. Division and alteration in the structure of tribes seem to have accompanied the changing nature of the ancient city along with the process of competition over resources and equity that accompanied it. At times, brand new tribes would even be created along a redistributive model with each having a defined share in territories in the city, along the coast, and in the inland area. Interestingly, such a process often focused on existing population centers rather than traced strictly spatial boundaries beyond the urbanized zones so as to maintain the fiction of tribes as kinship groups rather than residential wards. As the Greek polis developed, it was often the case that new tribes would be superimposed upon former ones so that, in the small Peloponnesian city-state of Phleious, for example, “…‘ancient’ tribes presumably carried out certain cult activities while the new tribes were the ones which participated in, and were subdivided for, political (and perhaps military) purpose.”^[5] The identity of tribes become especially legible during the (admittedly infrequent) process of naturalization of new citizens into a Greek city-state whereby incorporation into the Athenian body politic, for example, required the inductee to be enrolled successively into a deme, a tribe, and a phratry.^[6] The electoral reforms of Cleisthenes of Athens, commonly referred to as the creator of democracy in Athens, amounted to an overt form of interested gerrymandering that would make even a modern politician proud. So too are the changes attributed to his namesake Cleisthenes the tyrant of Sicyon (c. 600–570 BC) who altered his city’s tribes in a way that made allusions to a pre-political past that just as clearly speaks to the manipulation of his political present. Sicyon then were thought of as consisting of three dominant Dorian tribes that were privileged in their access to power and wealth in the city on account of their members’ claim to rights by conquest in a far distant past. Cleisthenes proceeded to take away these prerogatives and renamed the Hylleis, Pamphyli, Dymanatae, the Dorian tribes also thought of as the “children of Herakles,” respectively, as the Hyatai, Oneatai, and Choireatai (pig, ass, and swine people) while he gave so-called pre-Dorian tribes, including one thought to be that of his own, elevated names; thus the Aegialeis became the Archelaoi or the “Rulers of the People.” After one has discounted the truth value of the “old model” of the Dorian invasions, which allows this episode to be read as the reassertion of the power of the pre-Dorian Greeks – and hence the revolt of suppressed indigenous people against their oppressive colonizers – it virtually invites us to reread it as a case when a change in contemporary sociopolitical and economic relations and realties had to be made through the transformation of “ancient” identities because the latter had been asserted as the basis for privileged entitlements enjoyed by certain social groups over other ones.
The Greek and Roman tribes that we encounter in the historical record therefore indeed do appear as less than fully autochthonous. Instead they were often said to be the objects of the manipulation by named figures. Here we need to consider the Greek tendency to posit the agency of a founder (ktistês, oikistês) and/or lawgiver (nomothetês) in conferring distinction and identity on a social entity such as a polis. Indeed, one might even say that the abundant evidence in Greek and Roman accounts for the deliberate interested manipulation of social units such as tribes conformed to a mythologizing historical imagination that posited specific historical foundation moments and the agency of (often eponymous) heroic founders in the constitution of social bodies. What does such a realization mean for our own expectations regarding what we can learn from the available evidence? Here I would like to briefly digress to consider the evidentiary question. Indeed the issue pertains not only to the scantiness of the relevant evidence regarding the nature of tribes in early societies but also to its character. Sneath freely admits that in fact little is known about most early tribes (here he seems to be hedging his bet that there could be tribes before states) but at the same time asserts that this lack of evidence constitutes an argument against their being articulated kinship descent groups in any case. This amounts as much to an argumentum e silentio as Rincin’s proposal, which Sneath cites and criticizes in the article, that the Secret History of the Mongols does not show the pre-chinggisid (i.e., pre-imperial) Mongols as possessing an elaborate kinship structure because it had broken down even earlier. A similar dilemma faces ancient Mediterranean historians who are unable to travel back in time to interrogate their subjects or to observe the societies that they study, and have to rely virtually entirely for their primary source information a set of highly refracted literary and epigraphic records. What this means is that tribes as such become only visible or legible to us after the development of literacy and the development of a statal society (polis or civitas). In other words, neither the Greek nor Latin literary and documentary tradition offers contemporary witness to “pre-political” tribes even should such entities have ever existed; in corollary, they speak only to tribes as they existed in a “post-political” environment in which the latter appeared either as vestiges of an archaic past or as manipulated political subgroups, or both. Thus using the Greek and Roman evidence to argue for the nonexistence of tribes prior to the rise of the city-state is just as, or even more, methodologically suspect than arguing for their preexistence and role as primordial building blocks. A similar observation applies, I suggest, even to our knowledge regarding the Xiongnu and the Mongols, two other cases that form part of Sneath’s dossier. Neither Sima Qian’s Shiji nor the later Hanshu bears transparent witness to a time prior to when the Xiongnu had already become an imperial people,^[7] nor should we see in the Secret History of the Mongols anything other than a post-political/imperial Yuan dynasty document that aimed to construct an archaic pre-imperial tribal past for the Mongol imperial ethnos that was trying very hard to negotiate its present circumstances with particular versions of a stylized and legitimating chinggisid past. But these questions are sufficiently well known to require much treatment here.
One of Sneath’s useful insights is that if the concept of “nation” has been so thoroughly deconstructed of late, it is now time to do the same to that of the “tribe” (or the “house” for that matter). But should we not take the exercise even a little further? Indeed, why argue against the priority of one (nation/state vs. tribe or vice versa) simply by asserting the primacy and anteriority of the other? Why not regard them as dynamic, similar, and, in some cases, even mutually embedded processes, that is, as social and cultural dynamics rather than entities or states. Max Weber has maintained an important distinction between “political processes” as such and the state proper whereby the former are independent of as well as logically and temporally prior to the latter. In such a view, just as much as nation-formation, tribe-formation ought to be brought under scrutiny as the result of a political process but not necessarily one influenced by a preexisting state. I would therefore suggest that while tribes can and should be regarded as political units just as much as states were, a prior-existing state is not a sine qua non for the former. Rather, a broadly similar “political process” involving the active interruption and substitution of (claimed) preexisting hereditary and kinship lineages and redistribution of a population into mixed social groupings accounted for tribal formations as well as – especially when projected onto a large scale – the formation of nations or states. When this consideration is conjoined with my earlier observation regarding the evidentiary record (i.e., most of our knowledge of “tribes” came from a “post-political” era and/or from their sedentary neighbors), I find that I cannot fully accept the following assertion by Sneath: “Comprehensive kinship organization, where it appeared, seems to have been a product of the state, not a precursor to it, even in the Greek and Roman world that had inspired Morgan.”^[8] For if even the very basis of our knowledge about the tribes about which we wish to theorize is a “product of the state,” as broadly defined, Sneath’s claim above reveals itself as tautological at core.
Another quibble I have with Sneath’s otherwise very telling critique has to do with his broad generalizations regarding complex and nuanced thinkers. In his article, Durkheim is mentioned as a student of Fustel de Coulanges and a fellow traveler who bears the “imprint of Renan’s nationalism.”^[9] Yet Durkheim’s contribution to the present set of deliberations is much greater than is conveyed in these dismissive characterizations. Returning to the study of early tribes among ancient historians mentioned earlier, Szanto’s argument was almost immediately made known to and taken up by Durkheim who published a broadly sympathetic review of it in his journal L’Année Sociologique.^[10] Durkheim in fact used his dual review of Szanto’s work and a like-minded article by Hopzapfel on the origins of the Roman tribes both to highlight the move among certain scholars to reject the more essentialist and evolutionary notion of city-state formation (contra Theodor Mommsen, among others) as well as to add his own third perspective to the debate. For while Szanto and Hopzapfel forcefully contended that tribes among the early Greeks and Romans, respectively, were not antecedent to but actually a product of the archaic city-state, Durkheim suggested that this inversion of the traditional view conveyed but a half-truth only. For while he appreciated the Szanto’s and Hopzapfel’s arguments highlighting the inadequacy of an evolutionary paradigm, he identified two principal weaknesses: one is that their scheme provides no explanation for the trenchancy and universality of tribes as a form of sociocultural and religious organization and, moreover, seems to suppose that political creations conjured up ex nihilo without any reference to the social and religious imagination of the population in question could possibly have had the success that they seemed to enjoy. Thus Durkheim proposed that one needs to examine the form of social rationality that tribal group formation presupposed: “Ainsi la tribu doit être considérée comme un groupe naturel, en ce sens que c’est sous l’influence d’affinités naturelles qu’elle s’est formée primitivement; c’est la seule manière d’expliquer son extrême généralité, non seulement en Grèce mais en Italie.”^[11] He further contended that even if tribes were formed in reference to social principles that were believed to have been “natural,” they could still be shaped by a political process:

“Elles (i.e., tribes) sont naturelles en tant qu’elles reproduisent une forme d’organisation qui s’était produite naturellement. Mais la symmétrie de leurs subdivisions en phratries et genê démonstre évidemment l’intervention du législateur. Telles que nous les connaissons à travers les documents historiques, elles ne sont plus le produit d’une formation spontanée. Mais elles ne sont pas davantage des cadres de pure inventions, créés de toutes pièces par les hommes d’État.”^[12]

Central to Durkheim was of course the nexus between group beliefs and rituals. What remains to be added to the analysis then is to trace the symbolic actions, or ritual enactments, and the social beliefs that attended all these dynamic processes: such indeed was to become the project of Durkheim’s Les formes elementaires de la vie religieuse published in 1912.
This precise element is deemphasized or even wholly missing from Sneath’s current account of tribe, ethnos, or nation. The structurating role that public ritual practices played and how, in general, the mobilization/articulation of beliefs through practice constituted social groups and identities, and the study of religious and political ritual à la Durkheim and his followers in general, will have much to add to the present debates. The way early Greek and Roman tribes constituted themselves through the enactment of beliefs in ritual action presents excellent case studies opportunities. Indeed, many tribes may be regarded as religious or cultic associations at base while at the same time appearing just as readily as political entities. Old Greek and Roman tribes appeared above all as kinship groups that claimed monopoly over particular sets of “ancient rituals” that these practitioners asserted as ones that predated even the foundation of their city.^[13] Roussel comments that

“On notera tout d’abord que ce n’est pas parce qu’un group social apparaît à une certaine époque comme voué au service d’un culte très ancien, que l’origine de ce groupe doit nécessairement remonter très loin dans le passé.… Pourquoi devrait-on toujours rechercher l’origine des génè bien avant le temps de la Cité? Celle-ci, nous le savons, pouvait fort bien conférer le monopole d’un culte à un personnage jugé digne, pour services rendus à la collectivité, d’être ainsi honoré, lui et sa descendance.”^[14]

Early Greek tribes were devoted to the cult of heroes, and references to tribes among the Atthidographers or local Attic historians, the earliest writers of Athenian history, appear in their discussion of the Greek heroic cults. Civic processions were often associated with the cult of local heroes, who in some cases served as the eponymous founders of particular tribes within the city, and the presence of armed males in the procession accompanying paraphernalia of cult heightened the symbolic identification of the social group with the totemic symbols of its identity. Religion and cult continued to be important in defining tribes and tribal identities well into the city-state period so that, in one view, tribes only truly came alive as it were in the course of the performance of rituals such as public or civic processions. In the case of Athens, as elsewhere, the involvement of tribes in civic processions during religious festivals in Greek city-states allowed the civic community to become conscious of it own social identity in the best Durkheimian sense. Processions served as positive public rites that could not only articulate and bring into social consciousness the normative social identities of these groups but also allowed for the negotiation of their relations to other groups within the city through either placement or priority in the procession or other means of indicating differential status.^[15]
The central argument underlying Sneath’s study and his rejection of the evolutionary model of social formation that underpins it may therefore be said to be an ongoing intellectual project that dates back to more than one century. The evidentiary basis of the “Morganian thesis” – Sneath’s straw man – has also been examined long ago and found wanting, and indeed the scholarly communis opinio among historians does not in fact wholly support the preexistence of hereditary descent groups as the primordial basis of tribes (and later of states). But the debate should not be just about whether “tribes” (or “houses”) or nations came first. What Durkheim called for, and what still remains a worthwhile point to ponder today, was that we should not limit our debate only to questions about the relative priority of tribes versus states or nations but rather we must more deeply interrogate the symbolic actions and rationalities that underpinned the dynamic and necessarily dialectical processes that led to the constitution of every kind of social entity since only the latter constitutes what was in essence “the real.”