1Even today, most ancient historians and classical archaeologists, when investigating the origins and nature of the Greek city-state or Polis, are diverted from a clear perspective of the phenomenon by the abundant historical and archaeological data available for ancient Athens and Attica. Yet at the same time, the same scholars generally feel it necessary to remind us how abnormal and exceptional the Athenian Classical Polis is –probably from as early as the precocious union of Attica during the late Dark Ages.

2Although, therefore, I find my colleague Ian Morris’s radical reanalysis of Athenian society during the Dark Ages (Morris 1987) very convincing, I shall be deliberately provocative in arguing that the subtitle of his remarkable book Burial and Society, namely The Rise of the Greek City-State, is misleading, since the book’s focus on Athens deprives the reader of the appropriate viewpoint from which to comprehend the true nature of the typical Greek Polis.

3Throughout the 1980s and 1990s there have been many occasions when it has seemed that English-speaking scholars have been setting the pace in Classical Archaeology when it comes to fundamentally new and stimulating texts (one thinks of studies by Morris, Whitley, Osborne, Alcock). Yet these scholars are in reality operating within a narrow discourse focussing on the pathfinding research publications of their supervisor Anthony Snodgrass; it is a well-known phenomenon in Cambridge intellectual history that such fruitful but introspective schools can find themselves out-of-touch with developments in other scholarly traditions. One such, neglected, tradition is that of German historical geography of the ancient world, or “antike Landeskunde”, and it is from this source that I shall draw fundamental inspiration for the following analysis.

4Let us begin with my friend Eberhard Ruschenbusch, who in a series of penetrating studies (e. g. Ruschenbusch 1985, 1991) has synthesized what we can say about the typical characteristics of the 700-800 known city-states of the Classical Aegean world. He has shown that: 80 % of poleis studied have maximal territories of 100 km² and citizen males around 800 in number, whilst 69 % of poleis have only on average 400 adult male citizens-bringing total city-state populations (excluding slaves and resident aliens) to something of the order of 1600-3200 people. Now a “chora” or territory of 100 km² maximal is equivalent to a radial territory of 5-6 km-and with this geographical and demographic scale before us we can focus realistically on what Ruschenbusch and others have termed the “ Normalpolis”.

5The implications of such a small unit of space and population for a typical Polis were already brought out for us much earlier, in the works of another neglected scholar in the German antike Landeskunde tradition – Ernst Kirsten, most decisively in his major monograph of 1956 Die Griechische Polis (Kirsten 1956) with its significant subtitle als historisch-geographisches Problem des Mittelmeerraumes. The stimulus to Kirsten’s researches on the Polis go even further back within the same tradition-to a question innocently raised by his teacher Alfred Philippson of Griechische Landschaften fame: “Why were there so many poleis in Thessaly? ”

6What indeed was it about this observation that gave rise to Kirsten’s massive and brilliant monograph?

7Now a tradition still purveyed today in ancient history textbooks informs us that Greece is made up of little pockets of plain, surrounded by sea and mountains and cut off from each other, hence determining the formation of miniature city-states. Thessaly-two linked, massive plains of fertile land, gives no cause for such processes-yet by Classical times it was crowded with such poleis (Auda et al. 1991).

8Kirsten’s definitive solution to the problem is the touchstone for subsequent Landeskunde scholar- shipship. Ignoring the smokescreen of legalistic definitions of the polis-state, he sees the phenomenon as a geographer: the typical Greek city-state is in reality a village – a large village-but essentially a village, with its integral agricultural territory, and the Normalpolis is in origin a Dorfstadt/Dorfstaat. Kirsten set aside for separate investigation the rare and abnormal development of much larger city-states such as Athens, which he compared to Medieval towns and called Stadtstadt/Stadtstaat and Megalopolis. I shall return to the latter category later.

9I cannot more clearly show you the clarity of the Dorfstadt insight than through the example of modern Kastro village in Boeotia, Central Greece, which –neatly overlies the typical small city of Kopai and hints at a polis-village continuity even into the modern period, since its medieval occupants were an isolated pocket of Greeks amidst massive Albanian colonisation. However, since its medieval name –Topolia –is Slav, there should have been a strong influx of non-Greeks in early medieval times! Likewise little Chorsiai polis in S. W. Boeotia, studied by my friend John Fossey (Fossey/Morin 1987), with its compact small territory, has been replaced at no great distance by another surviving Greek community at the Byzantine to modern village of Chostiyani/Chostia.

10Living every summer for almost two decades, as I have done, in rural Boeotia, where traditional villages are regularly spaced about the landscape in comparable scale to Ruschenbusch’s territories, and with populations likewise of 1-3000 people, also brings home the “village” reality of Normalpolis life. In 1879 even the two “towns” of the Boeotian province had only 5 000 souls each, making them only marginally urban rather than “very fat villages” with special functions (Sauerwein 1991). The fact, however, that over 100 years ago there were far more villages in the Boeotian countryside than today, and of smaller size and territory, stimulates us to search likewise for a dynamic development in early historic times underlying the mature form of Kirsten and Ruschenbusch’s Dorfstadt/Normalpolis.

11Boeotia offers a useful database for this “origin of the polis” analysis (Bintliff 1994), as a by-product of the Boeotia Project’s continuing archaeological investigation into its regional settlement history (a Project sponsored by the British Academy and directed by myself and Anthony Snodgrass). Our intensive surface survey both in S. W. and Ν. E. Boeotia has shown that despite discovery of several hundred settlements of all periods, settlement activity in the Dark Ages of Protogeometric and Geometric times is very confined, and indeed predominantly focussed on locations later to become named Classical towns and villages, with occasional small sites clustered in their immediate vicinity. This confirms the less well-researched picture for the entire province of Boeotia (fig. 1), where recorded PG and G settlements and cemeteries are almost entirely locations later to become Classical towns and villages. These widely-spaced settlement foci appear, not least from our own survey results, to be hamlets or villages with extensive individual territories (Bintliff 1994).

12By final Archaic and Classical times these “pioneer” hamlet/villages have further multiplied, through a process of internal colonisation creating new settlements in the interstices of the original Dark Age network; the subsequent reduction in average distance between nucleated settlements results in an average territory radius of modular 2-3 km or some half-hour’s walk (fig. 2). However over the same period political differentiation has separated the modular nuclear foci into three classes: of “polis” –large and small-and dependent villages or komai. The territory of the poleis, which now incorporates that of their dependencies (komai alone for small poleis, komai and small poleis for large poleis), is larger, ranging from 5 to 10 or even 15 km radius with the variable power of individual city-states (fig. 3). Nonetheless-and this is the critical element-underneath the territorial radius of the larger and smaller poleis the fundamental settlement and territory unit remains that of the “village” territory originally specific to each polis and that of each of its satellite villages, whilst citizenship at both levels of the hierarchy rests upon landownership in the relevant community lands or chora.

13I think it most helpful to see the modular infill of essentially large and small farming villages in Late Geometric to Archaic Boeotia as a human ecological process, influenced by Strange Attractors. I find it highly appropriate to introduce Complexity Theory into my argument at this point (Lewin 1993). This is an interdisciplinary body of ideas in the Biological and Physical Sciences which focusses on the tendency for complex systems with strong random oscillatory tendencies to converge on recurrent, semi-stable configurations or structures (Strange Attractors) through processes such as adaptive advantage, functional resonance, etc.

14The first “strange attractor” for my analysis is hypothesized to be-(empirical data are rare if suggestive) –a preference for farmers to concentrate activity within a one-hour radius of settlement (or some 5 km radius of level terrain) (cf. Bailey/ Davidson 1983). A second “attractor” is the well-attested anthropological mechanism of village fission when a community experiences social stress due to increasing size (Forge 1972) –producing internal and external colonisation across the landscape. This attractor has been given a putative biological foundation by the physical anthropologist Robin Dunbar (Dunbar 1992), having previously been highlighted by social anthropologists such as Forge, and includes the claim that there is a limit on “face-to-face” societies lacking strong stratification to a community size of 150-200 members or less. (The theory has also been used to good effect in Morris 1987, for rather different purposes).

15These two initial “strange attractors” can help us to postdict the modular colonisation of the Boeotian landscape. Let us take our initial, large-territory, Dark Age farming communities, and assume that the majority are in a political sense weakly-differentiated village/hamlet societies. Once these had risen in population above the hypothesized threshold for face-to-face social cohesion, they would split and cause colonisation of intervening land. Given Kent Flannery’s important discovery (Flannery 1976) in early farming Meso-america that this stage of fission long precedes land exhaustion around pioneer hamlets, we might expect also for Boeotia that social factors dominated over economic, and that land use was still at a small percentage of available cultivable land when village fission occurred. Indeed the archaeological and historical evidence, together with the later development of intense land use, confirms the model for Central Greece. In such circumstances offshoot hamlets carved out territory from the outer parts of their parent’s “parish” without threat to parental subsistence; if one sees this from the perspective of spatial geometry, a pioneer “maximal” territory of 5 km radius, after two phases of settlement fission and associated halving of territory on each occasion, would be reduced to a territory of 2-3 km radius. I would argue that further territorial subdivision, in dry-farming landscapes and with sizeable villages, would be a dangerous strategy, removing any economic cushion against resource fluctuations; this predicts a cross-cultural regularity –a third “Strange Attractor” –of stabilisation of mature dry-farming settlement systems into networks of some half-hour radius (fig. 2), except under conditions of overpopulation. Indeed settlement systems with much smaller territories may be symptomatic of overpopulation.

16I have deliberately avoided telling you how some of these modular village-territory units of Late Geometric-Archaic Boeotia emerged as Classical poleis as opposed to others destined to become satellite komai. This is for the very good reason that I consider them all at a certain stage to be protopoleis or potential poleis of Dorfstaat type. However, I also suspect that there may have been from the first, a small number of exceptional, larger communities with strong political differentiation outside of this face-to-face community fission process. It is a combination of historic, unique factors in the histoire événementielle of each proto-polis and converging “strange attractors” (some not yet mentioned-such as inherent territorial fertility) that will weight the scales for or against any one of our modular villages becoming a “polis”.

17The key to understanding this stage of village to polis transformation can be found both in the village ethnography of traditional, post-Medieval Europe, and in the insights of historians studying the development of early Medieval villages in Western Europe. Both research communities have, independently, presented similar models for transformation phenomena in village development. It will also help us to introduce yet another Strange Attractor unknown to these research communities: the prehistorian Martin Wobst’s argument (Wobst 1974, 1976) that healthy human populations require marriage/mating networks of 5-600 people to be viable in the long-term. I shall briefly attempt, using these insights, to reconstruct the main lines of village development in Medieval and post-Medieval Western Europe of relevance to us.

18Let us begin by taking the simplest form of village colonisation, such as may have been common in post-Roman Dark Age Europe: small, weakly socially-differentiated villages/hamlets of less than 200 people undergo regular fission due to social tension, and colonise interstitial land, with a resultant subdivision of original large territories. The “strange attractor” of face-to-face social relations has been operative (consider the eleventh century English Domesday Book village databank for an excellent confirmation1), but our modular villages, being too small for an endogamous marriage network, are necessarily being drawn through Wobst’s attractor into cross-community intermarriage, to satisfy prohibitions on close-kin marriage. Thus adjacent locations within the former outer lands of pioneer villages are ideal for daughter foundations, since your new neighbours are kin, but after a while not so close-kin, yet close enough in common background and in space for easy socializing. Land is still plentiful since village fission has occurred long before full exploitation of the cultivable landscape.

19Let us now envisage the effects of what happens, demonstrably, throughout much of Western Europe across the early Medieval centuries of the later 1st millennium AD: a continuing rise of population. Our new modular, smaller territories, often some half-hour in radius, become ever more fully exploited. Now the gravity-force of out-marriage becomes an economic disadvantage; in return for participation in a district mating-pool, your community loses land and stock through dowries, receiving distant plots and grazing-rights in other communities’territory.

20The resolution is simple but dramatic: further increases in village populations, achieved by agricultural intensification, allow the community to rise above the Wobstian threshold figure of 5-600 people and therefore to be largely endogamous; and the leaking of its life-blood, its territorial resources, shrinks to a trickle.

21In a fascinating anthropological analysis Susan Tax Freeman (Freeman 1968, 1970) has shown how traditional Early Modern village Spain was dominated by large villages which had succeeded in this goal of territorial integrity and dominant endogamy-the “pueblo” village-town; alongside these idealized communities there existed smaller villages that openly acknowledged the disadvantages of their failure. Information on rural endogamy/exogamy in Early Medieval Europe is less easily extractable in comparison to recent village Europe, but by the High Middle Ages when average village territories in lowland England and France, for example, were 1-3 km radius, and populations were frequently of appropriate size for a high level of endogamy, we might reasonably argue for processes comparable to those typical for Early Modern Spain and Italy with high levels of endogamy.

22But Freeman also underlines the importance of socio-political changes in the village as its resources come under pressure from internal population growth. Her emphasis on the “corporate community” is echoed independently by medieval economic historians such as Harold Fox (Fox 1981, 1992) who has focussed on the problems arising from continual population increase within the dense network of early Medieval rural communities across the West European lowlands. Arable, and more particularly pasturage for farm and ploughstock became in critical short supply. Although by around 1100 AD the absolute number and spatial network of villages was largely in place, the result in many regions of colonising villages trying to stabilise at face-to-face level, between 1000 AD and 1250 AD most European villages continued to grow internally, doubling or tripling population from their original size. This may have relieved land pressure in one respect, through minimizing of out-marriages and maximizing resource negentropy, but it greatly increased the stress on internal land use. The solution found was the one that is also a characteristic of large traditional villages in Early Modern Europe: the creation of a “corporate community”2.

23This was a unified body, usually consisting of the wealthier adult male landowning heads of households, which assumed complete control over the entire village territory, reassigning land where necessary and generally assuming management responsibility over the resources of the village. A remarkable specific achievement of these Medieval West European village councils appears to have been a decision, which once taken spread rapidly throughout lowland West Europe, to confront pressure on the village arable and grazing territory through a complete root-and-stock restructuring of land use. The haphazard accumulation of individual family plots and pasturage was abolished and a new scheme was pushed through in which 2-3 giant fields were laid out and used in strict rotation by all villagers for similar annual land usage.

24Students of the ancient Greek Polis will I am sure have begun to recognize from far off some key features, perhaps the key features of Classical city-state life, in these Medieval and post-Medieval West European village developmental trajectories: the identity between closure of territorial integrity and a strong political community with its roots in land-holding within the territory; the necessity of endogamy for almost all polis citizens, since non polis-born residents cannot own land and are therefore de facto excluded from citizenship, landholding being the principal qualification for full membership of the political community. As an intriguing footnote to the fashion in which polis-exclusivety extended itself into all essential areas of community life, foreigners {i. e. non-citizens) in Greek poleis were only allowed to participate in polis cults through the medium of a citizen intermediary or “proxenos”; this even persisted in “interstate” sanctuaries such as Delphi, where citizens of Delphi had to act as continual intermediaries for pilgrims (Sourvinou-Inwood 1990).

25If you like, the Greek polis, the Spanish pueblo, the High Medieval West European village, have “squared the circle”: most of them will have begun life as “face-to-face”, fissioning and colonising hamlets gradually filling the landscape in modular patterns, pulled towards the strange attractors of social and ergonomie space. Colonisation of neighbouring sectors of landscape through territorial geometric subdivision was also in part a response to the pull of mating network requirements. Subsequent continuing expansion of population and intensification of land use placed growing stress on the territorial integrity of each community’s resource space, a tension which was resolved both naturally, through increasing closure of an endogamous marriage pattern, and artificially, through the establishment of a powerful council of the chief landholders with the power to intervene in questions of land use within the community’s territory. The central achievement of these recurrent settlement forms is that the mating community circulates, within the “chora”, its resources-human and environmental.

26In essence we have created the Dorfstadt-Dorfstaat, and if our analysis is correct we cannot conceivably view the Greek Polis as a unique phenomenon, just because it has special features-such as its own army, or claims complete political autonomy, and carves absolute laws on its walls: one could argue that the Medieval Three-field system is just as remarkable a community achievement.

27Yet I perhaps deliberately overstate the case, to prevent a chauvinistic blinkering amongst Classicists. Nonetheless, there are fundamental parallels in the processes I have described, and they are largely the product of the same “strange attractors”. What is it indeed that prevents the Spanish pueblo or the High Medieval village of say, Northern France from “U. D. I. ”? It is the fact that they have developed under the restrictive political superstructure of the Iberian kingdoms and lordships, of the French feudal kings, barons and knights-albeit with extensive local powers vested in the corporate communities of the village councils, especially in the realm of territorial management. What I would claim is bold enough: that in the absence of such prior authority, or in cases of weak expression of higher powers, the same strange attractors will begin to produce similar results to Greek polis formation. Symptoms can be seen in post-Medieval Italian hill villages waging war on each other (Tak 1990), or more recently terming themselves “villages by day, towns by night”. For a full realization of “polis” potential, we might turn to the rise of the genuinely autonomous Italian city-states of the Middle Ages –arising predictably in a regional power-vaccuum within North-Central Italy; or for a much earlier era, we could focus on the multiple small city-states of the Bronze Age in the East Mediterranean.

28Ruschenbusch’s statistical survey and examples I have cited from Boeotia make it clear that the vast majority of Dorfstädte or “village-poleis” remained in scale as large villages. Some such as Chorsiai or Kopai survived as single settlement, 2-3 km radius territory cities of “proto-polis” type, at the lower end of, or even below, Ruschenbusch’s average population figures, throughout Antiquity. But a typical 5-6 km radius territory polis, at the upper end of his average size scale (towards 3000 or more people) will generally have reached these spatial and demographic dimensions by enlarging its chora through absorption of one or more neighbouring “proto-poleis”.

29Thus in S. W. Boeotia, Askra village, already by Hesiod’s time, at the dawn of the Archaic era, appears to be a satellite of 5 km-distant Thespiae. Plausibly at an earlier stage Askra may have been autonomous, and if some ancient sources can be accepted, during the Archaic era hostilities led to a Thespian expulsion of the Askraeans and a resettlement of the kome/village (Snodgrass 1990: a possible fortification of Archaic-early Classical style might relate to this episode).

30Boeotian history, indeed, from the start seems characterised by “polis prédation”, whereby over time those proto-poleis which emerge as true poleis do so either by survival in remote corners of the province, or more normally by increasing their regional power through incorporation of neighbours into their territorial states. The Normalpolis, therefore, at the medium to upper end of its scale, although it can be a single large settlement with a maximal farming catchment of 5-6 km radius, is much more likely to represent a larger community together with subordinate proto-poleispoleis and their lands. The pretensions and/or fears of all proto-poleis during the Archaic era of polis prédation seem to be reflected by the widespread practice of community fortification during this period for all communities beyond the hamlet.

31The organizers of this Colloquium have admirably posed questions of importance to my analysis, one of which I shall pause to consider: is the dynamic of State development linear-or even “Neo-Evolutionary linear”?

32In espousing Chaos-Complexity theory I am arguing for a certain timelessness, a certain semi-deterministic set of processes “waiting to act”; yet there are parameters with chronological meaning which affect the likelihood and scope of operation of the relevant strange attractors which I have sought to identify in Greek settlement history and elsewhere.

33The predominance of fixed, nucleated settlement is to be associated with settled village farming; the ability to produce sufficient food to enable a village to become demographically stable in virtual Wobstian endogamy is a feature especially to be. associated with additional factors-such as Mediterranean polyculture (relevant as early as the Bronze Age in the Levant and Southern Greece), or effective iron technology (relevant to early historic Greece and Italy), or social modifications such as the Medieval Three-field management system. Power vaccua, or weak overarching power-structures, however, once regional state systems are in existence in a particular part of the world, can occur at more arbitrary times, and hence within the agro-techno constraints already noted the corporate community and the Dorfstaat can appear in many different localities, cultures and eras-without prediction, but open to postdiction (to borrow a noteworthy concept of Stephen Jay Gould’s [Gould 1986]).

34To return to the dynamics of polis-formation in ancient Greece: whether or not one is tempted to extend a biological analogy to erect an hypothetical strange attractor of territorial prédation, it is an empirical fact that, progressively, most of the 2-3 km radius proto-poleis are absorbed over time into the territories of larger poleis in the course of the Archaic centuries (fig. 4) if not in final Geometric times. Thus in Boeotia, most proto-poleis lose their independence and there remain only 14-15 truly, or semi-autonomous poleis for the province in mature Classical times. These survivors are nonetheless regularly at odds with one another, and are also all threatened by the giant community of Thebes. Dominating local politics through most of the C5th BC, Thebes succeeds to total control over Boeotia and then to hegemony over Southern Greece during the early C4th BC; subsequently the other cities manage to reassert themselves, and most continue as political entities, with diminishing powers, till the end of Antiquity-when they, often already shrunken to proto-polis size again, become transformed into Medieval villages.

35I have yet to say something about “Princes” or indeed about social hierarchy. The time has come! I shall do so through the intermediary of further intriguing details in the Boeotian polis development sequence.

36From the earliest clear archaeological evidence in our survey areas of S. W. and Ν. E. Boeotia, the “pioneer” Geometric villages show quite disproportionate sizes (Bintliff 1994): Askra is a small, tightly –focussed hamlet and will become a large kome/village under Thespian control by Archaic times (perhaps after its polis pretensions are thwarted); Hyettos has a reasonably extensive area, perhaps in the form of several close hamlets, in Geometric-early Archaic times, and will become a middle-range autonomous polis; Haliartos appears to be yet larger in its formative phases, but a single extensive community –it will grow into a middle-to-large polis; finally Thespiae has a number of hamlet foci over a wide area and will become one of the larger poleis of Boeotia. It seems as if their future Classical status is already anticipated in the size of these villages in the final Dark Age to early Archaic era. That hypothesis receives confirmation if we turn to consider the case of Thebes: it has very widely-scattered multiple hamlet foci in Proto-Geometric and Geometric times, that seem to point to an exceptionally large community, appropriate in scale to its remarkable pre-eminence in size throughout the Classical era.

37We can therefore suggest that the “dice are loaded” when it comes to proto-polis interactions. Even though Askra will achieve (with a putative 1300 or so inhabitants and a discrete 2-3 km radius fertile territory), potential polis integrity by Classical times if not earlier-already by the commencement of the Archaic era, two centuries before, we have Hesiod to inform us that power lies with the corrupt chiefs or basileis in “the city” of Thespiae and not with the occupants of his “village”. So the basileis – the local aristocrats, are present, at least in a dominant, newly-emerged, larger Dorfstadt. This leads rather naturally onto the next very significant question: what then, does this make Hesiod, politically?

38Social theory, and now biological theory, argue that communities which rise above the face-to-face mode of social integration require either, or both, horizontal or vertical stratification. Some subdivision of the community into separate quarters (barrios) or kin groups, or the emergence of a group of powerful individuals (Big Men), even the stabilisation of a class of powerful versus a class of subordinate families-these are characteristic phenomena associated with the social infrastructure of large communities.

39We can surely see in the archaeological evidence for multiple foci of burial and occupation in the medium-to-large emergent village-poleis of Dark Age to early Archaic times, such as at Thespiae, Hyettos, and Thebes, as at Athens, Argos, and equally in the comparable phase of city emergence of Etruscan cities-the direct reflection of distinct social groups predicted by these general theories. As well as this evidence from which to infer a horizontal infrastructure, we can reintroduce Ian Morris’s interpretation of Dark Age Athenian cemeteries (Morris 1987), which has won unqualified support in claiming that a vertically-stratified society dominated life throughout the main part of the Attic Dark Ages: not only are cemeteries revealing high-status and lower-status families, but it is suggested that consistently only 1/3 to 1/2 of the adults of the community are receiving formal burial (the Agathoi), concealing a hidden lowest class (the Kakoi) 3.3. We can therefore recognize a three-level social hierarchy.

40This persistent inequality, of course, puts into helpful context the otherwise remarkable Lefkandi elite (house and?) tomb from early in the Greek Dark Age4, and the clear literary evidence from the final Geometric era at Askra/Thespiae for three levels of status in Hesiod’s society (with upper class basileis, yeoman farmers such as Hesiod with dependents, and free/unfree dependents together with the poorer villagers). James Whitley (Whitley 1991) has argued that the dominant families comprising the uppermost stratum of the upper class (Agathoi) were unstable Big Men, citing the role of hero-cults, archaising art, and false genealogies in early historic Greece as evidence for an absence of genuine aristocratic stability; social mobility between the upper and lower divisions of the Agathoi is therefore envisaged.

41I return to the question: Who then is Hesiod, in social terms? Most commentators place him firmly in the class of wealthier farmer, chafing at the power of noble families in Thespiae. The Hesiodic “yeoman farmer” class is recognized to be a critical force in the rise of the polis, because over the Archaic period, (despite continuing debate over the precise details), it is generally agreed that a fundamentally new military and political factor established itself in all poleis-that of the heavily-armed citizen footsoldier army in an organized formation or phalanx.

42This vital defensive and offensive arm of the polis is essentially composed of “middle-class” farmers with surplus wealth sufficient to maintain the heavy armour and weaponry of their “hoplite” class. Aristocrats, although still the primary source of an accompanying cavalry, are in time not ashamed to stand anonymously in the hoplite formation. The military superiority of the new citizen army is such that it rapidly displaces other traditional forms of land warfare in the polis landscapes, and it has rightly been stressed that political power-sharing between the hoplite class and the aristocracy was both advantageous to polis security and a pragmatic necessity. The archaeological shift from weapon burial in tombs to its cessation in favour of dedication of the same items in “public” polis sanctuaries, (a strong contrast between the Late Geometric and the succeeding Archaic era), has also been seen as marking the military cohesion of the new polis-state, as it broadens political participation to the yeoman/hoplite class. Anthony Snodgrass has recently suggested (Snodgrass 1991) that the Polis is generally emerging in Southern Greece between the late C8th and the early C7th BC, with the hoplite phalanx in widespread use as an a posteriori proof of polis existence between 700-650 BC.

43Hesiod should then be seen as a member of a class which, it would seem, after his time, will use its growing economic prosperity to equip itself with appropriate military gear and obtain power-sharing in the polis-state, as the hoplite class. At the same time he is clearly in a distinct social stratum from a class lower than his which includes hired-labour and slaves.

44Now here is a paradox: whilst Ian Morris convincingly documents that 1/3 to 1/2 of Dark Age society, at least in the larger settlements, may be taken to represent a generalized upper class (the Agathoi), with the remainder belonging to a lower class (the Kakoi), he fails to observe the resonance that these proportions evoke. The reason, once more, I suspect, is the Atheno-centric bias of his study, where the future development of the polis-state is towards an extreme experiment in democracy encompassing all of the Kakoi and the Agathoi.

45In the Other Greece, the Greece of the 7-800 “Normalpoleis”, however, the typical political constitution by Classical times was just as likely to be that of a very broad oligarchy for all or most of Antiquity. Boeotia is a good example. Citizenship came for all free-born of citizen-parents within the state-territory, but political power was restricted to those male family-heads with a minimum property-status-almost always limited to the hoplite class and above. Now as is well-known, and here the numerical resonance occurs, the hoplite-hippeis (noble) class together comprised, on average, 1/3 to 1/2 of the citizen population in a typical Greek polis (cf. Snodgrass 1991, 19).

46The inevitable conclusion has to be, that in the development of the Normalpolis of ancient Greece, the crucial division of an upper class and inferior class, typified by the oligarchic constitution of all Archaic Greek city-states, and remaining typical for at least half of all Classical city-states (Prof. Peter Rhodes, pers. comm.),-already existed in the same proportions within Dark Age communities, at least in larger settlements such as at Athens. The Dark Age Agathoi included what will later be formalised as the hoplite class. Put this way, the generalized emergence of the Polis during early Archaic times is not the great democratic extension of political status to the lower class or Kakoi that will be the ultimate achievement of Athens in the C5th BC, and more subtle interpretations are called for.

47Here a clue may lie in Snodgrass’s recent (Snodgrass 1993) reanalysis of the origins of hoplite warfare. It does seem likely that in the later Dark Age at least, the posturing antics of the aristocratic warrior class were accompanied, not just by a rabble of ineffective and poorly-armed peasants, but by a more significant military force of “middling-status” folk-even though this has not yet formed into a coherent formation or phalanx. Gradually, from Late Geometric times into the C6th BC these proto-hoplites as a class become more fully-armed, more formation-armed and more formation-active, until the polis army is effectively the citizen body as a solid phalanx.

48Adopting a gradualist position enables us to find more subtle shifts across the Dark Age to Archaic transition. Following Whitley’s “unstable” nobility model, let us assume that the Dark Age communities of larger size (those of say five hundred or more), were composed of leading families associated with a class of prosperous yeoman-farmers (the “proto-hoplites”) –both levels being permeable to two-way social mobility along Big Man principles, together with an equally, or more, numerous lower class. In warfare and economic rivalry the families of both upper-class levels may have witnessed cyclical fortune. We might mention in connection with this, Bill Cavanagh’s (Cavanagh 1991) comment on the “genealogical shallowness” suggested by the short life of Dark Age tombs at Knossos, which he interprets as showing limited kinship links and a low degree of continuity in lineages of powerful families. Set apart, however, from the two-level upper class, whose adult members are privileged by formal burial, is a lower class of lesser folk, the Kakoi, dependent on the Agathoi both socially and economically.

49We might hypothesize further, that although privileged in burial, and considered a useful fighting element in inter-community warfare, the “proto-hoplite” class has limited power or influence in “affairs of state” in the emergent Dorfstaat, even if far from being servile; it would comprise the fully autonomous and prosperous clients of the leading families, people whose allegiances could shift and be of some potency in the aristocratic politics of the pre-hoplite reform era.

50Smaller communities, which I have earlier suggested to have been the most likely source of landscape infilling by village-fission, may well have lacked members of the uppermost class of basileus status, and even its prosperous peasants of the lower Agathoi class could have been subject to two-way mobility with other free-born Kakoi families in their communities-probably Hesiod’s Askra was such a place.

51At different times in different places across the early Archaic era, the growing numbers of prosperous “middle-class” farmers and their increasing military effectiveness, and the “scissors” created by the need for greater territorial integrity associated with heightened endogamy, stimulated the creation of the corporate, power-sharing community, the broad-based oligarchic polis, in which the hoplite voice was truly powerful5. Let us recall that even Athens only became a true citizen-democracy with the Kleisthenic reforms at the very end of the Archaic era. Such a transformation in the structure of power, driven on by demographic and economic trends, refocusses our attention on the very real modifications in the levels of human activity across the Greek landscape between earlier Dark Age and final Geometric to early Archaic times, which Ian Morris’s revelation of the artificially-exaggerated rise in burial numbers across this transition has unintentionally served to obscure for the scholarly world.

52It is surely not coincidental, that in the moderate oligarchy which typified the Archaic polis, and remained as common as a wider democracy in the Classical polis, a citizen population of average 1 600 people (some 69 % of Ruschenbusch’s database) would imply a ruling hoplite-hippeis class of some 134-200 adult males-ideal for a face-to-face politics.

53Finally some comments on the Abnormal Polis (Bintliff 1994; Kirsten 1956). In Boeotia, Thebes may have retained a large population and an elaborate social structure throughout the Dark Ages, and will sustain this lead in size through inter-com-munity aggression through the final Geometric, Archaic and Classical periods, until achieving regional hegemony in practice during the C5th BC and in official name during the early C4th BC. By its apogee its walled area may have been 350ha, one-and-a-half times the walled area of C5th Athens (c. 211 ha), and one may suggest a population in Thebes itself of some 30-40,000 people. It is truly on the scale of a major medieval city and exists in another class altogether from all other poleis in its province. As a Primate City its role has nothing to do with the Dorfstadt, and its economy and politics are on a far more grandeous level. It is, to use Ernst Kirsten's useful phrase – a Megalopolis, playing to a very different set of rules from its neighbouring poleis.

54Nonetheless from the Dark Ages onwards it has had to fight its way across the landscape from a confined territory, absorbing or expelling adjacent populations in a process of relentless expansion. Its territory from the first appears to be larger than average, and this hints at a continued subordination of satellite hamlets from Bronze Age times, or a very early incorporation of its nearest neighbours. Yet the behaviour of the Theban Megalopolis goes well beyond the typical centre-plus-satellite system which larger Normalpoleis grow into: Thebes is known to have forcibly resettled some villages within its walls and expelled other competing settlements entirely from the region. This sharper version of predatory territorial expansion is reminiscent of Syracusan “ethnic cleansing” on Sicily (a Megalopolis of comparable size) (Berger 1991), and to my mind deserves further investigation, arguably from a socio-biological perspective.

55Moving across the Boeotian frontier south, it is intriguing to re-examine the case of the rise of the Athenian Polis “from the bottom-up”, since the Attic province is almost identical in size to the whole of Boeotia. Ian Morris's helpful synthesis allows us to suspect a similar process of landscape infilling to that seen in Boeotia during Dark Age to Archaic times. We would also follow Osborne (Osborne 1985) and Whitehead (Whitehead 1986) in concluding that the resultant dense network of Attic settlements is essentially a village network. In their view the Kleisthenic constitutional reforms of 507-8 take as given the fact that the people of Attica live in some 139 rural villages or urban suburbs (demes), from which already existing units, a centralized political system for a unified single-city, Athenian state can be forged using such natural “building blocks” as constituencies.

56A geographical analysis of the locatable demes (Bintliff 1994) suggests strongly that the rural districts of Attica have village territories of on average 2-3 km radius, just as in Boeotia, whereas the urban and “rural-suburban” (astu) demes are far more territorially-packed with an average 1-2 km radius parish6 (fig. 5). From the councillor quotas for 507-8 BC, which are believed to represent each village proportionate to population, we can also see that there are a series of very large “villages”, or rather small towns, in Attica-Athens does not stand alone above a sea of small, rural hamlets7.

57Let us recall that Athens is unusual in its precocious unification of a large province based on a single city, and that because this happened before recorded history-perhaps in the vigorous C8th BC-we find it difficult to give proper credence to what then have to be classified as semi-legendary tales of events preceding unification. In particular we need to take more seriously those repeated stories of open rivalry between the nascent power of Athens and that of several alternative and autonomous centres of Dark Age power in Attica, out of which Athens alone emerged victorious. In our comparative perspective from Boeotia and the Normalpolis world of the Aegean, such a scenario seems all too reasonable-with a similar putative early structure of numerous proto-polis villages (the demes), some already unusually large and associated with high-status burials (and some of these by 507-8 BC are appropriately of the size of a medium-to-large polis in other landscapes). The “demes” of Athens are not merely the reflection of the state or the “Demos”-as Robin Osborne seems to suggest: the village-deme is the basic building-block of the “polis” in “normal Greece” and will also have played a surprisingly important part in the formation of the Athenian state8.