Collections: Hoplite Wars: Part IVa, The Status of Hoplites

This is the last part of our four-part series (I, II, IIIa, IIIb, Intermission) on the debates surrounding ancient Greek hoplites and the phalanx formation in which they fought. We’ve spent the last two entries in this series looking at warfare quite narrowly through the lens of tactics: hoplite spacing, depth, fighting style, and so on. I’ve argued for what I regard as a ‘blended’ model that sits somewhere between orthodoxy and heterodoxy: no ‘shoving’ othismos, but the hoplite phalanx is a shield wall, a formation with mostly regular spacing that is intended for shock and functions as a shock-focused shield wall formation likely from a relatively early date.

This week, we’re going to now ‘zoom out’ a bit and ask what implications the hoplite debate has for our broader understanding of Greek society, particularly polis Greek society. Hoplites, as warriors, were generally found in the Greek poleis but of course not all Greeks lived in poleis and areas of Greece without poleis largely lacked hoplites as well. In particular, our understanding of the place that hoplites have in polis society has a bunch of downstream implications in terms of social structure, the prevalence of slavery and even the question of how many Greeks there are in the first place.

I ended up having to split this into two parts for time, so this week we’re going to focus on the social status of hoplites, as well as some of the broader implications, particularly demographic ones, of a change in our understanding of how rich hoplites were. Then next week we’re going to close the series out by looking at hoplite ‘discipline,’ training and experience.

As always, if you like what you are reading, please share it as I rely on word-of-mouth to find readers! And if you really like it, you can support this project over at Patreon; I don’t promise not to use the money to buy a full hoplite panoply, but I also don’t not promise to do that. And if you want updates whenever a new post appears, you can click below for email updates or follow me on Twitter and Bluesky for updates when posts go live and my general musings; I have largely shifted over to Bluesky (I maintain some de minimis presence on Twitter), given that it has become a much better place for historical discussion than Twitter.

Orthodox Yeoman Hoplites

The key question we are asking here is fundamentally “how broad is the hoplite class?” That is, of course, a very important question, but as we’ll see, also a fiendishly tricky one. It is also a question where it can be unclear sometimes where scholars actually are which can render the debates confusing: heterodox scholars write articles and chapters against something called the ‘myth of the middle-class hoplite‘¹ but it isn’t always clear exactly what the bounds of the model they’re arguing against is, in part because orthodox scholars are not generally proposing hard numbers for the size of the hoplite class.\

Post-Publication Edit: We’ve already had some confusion in the comments so I want to leave a clarifying edit here. We’re about to dive into a lot of questions about the percentage of people in the hoplite class. But all of the scholars involve calculate those figures on a different basis – in particular does the denominator include women? children? slaves? the elderly? I try to homogenize those estimates here as best I can, often aiming for a ‘percentage of free households‘ (so the enslaved excluded) or ‘percentage of adult males’ (so women and children excluded, but slaves included) in a given status type. But I am afraid you will have to keep track fairly closely of exactly what percentage of what we’re calculating (and of course it is entirely possible I have simply made a math error somewhere, although I have tried to be careful).

By way of example, I want to take Victor Davis Hansen out to the woodshed on this point – because his half of this specific disconnect was brought up in the comments early in this series – in terms of the difference between how he sometimes imagines in words the size and social composition of the hoplite class and then how it looks when he uses numbers. In The Other Greeks, VDH’s preference for describing the hoplite polis of the late Archaic is ‘broad-based’ a term he uses for it about three dozen times, including on when he talks about the “broad base of hoplite yeomanry” and how “when middling farmers were in control of a Greek polis government it was broad-based: it was representative of the economic interest of most of the citizenry” and when he references “the yeomanry […] who had built the polis and created broad-based agrarian governments.”² These references are, in my digital copy, all within 3 pages of each other. They certainly give the impression of a middling, yeoman-hoplite class that dominated the typical polis. And indeed, in his more pop-focused works, like the deeply flawed Carnage and Culture (2001) he posits Greece as the origin point for a western tradition that includes “equality among the middling classes” tied to the hoplite tradition, which certainly seems to suggest that Hanson thinks we should understand the hoplite class as broad, covering even relatively poor farmers, and with a great degree of internal equality.

But then flash forward three whole pages and we’re calculating the size of that ‘broad-based’ class and we get a line like, “the full-citizen hoplites […] composed about twenty percent of the total adult resident population of Boeotia.”³ And pulling out just that second quote, someone might express confusion when I say that the heterodox argue that the hoplite class is small and exclusive, a rejection of the ‘middle class’ yeoman-hoplite of the orthodox school, because look there is VDH himself saying they’re only 20%! But equally, one may question the fairness of describing such a rate of enfranchisement as ‘broad-based!’

Now on the one hand VDH’s argument in this passage is about the relative inclusivity of ‘moderate’ oligarchies (the ‘broad-based’ ones) as compared to radical Greek democracies and so the question of the relative breadth of the hoplite class itself is not particularly his concern. But I think he’s also hiding the ball here in key ways: Boeotia is a tricky test case – unusual and famous for both its significant cavalry (drawn from an unusually wealthy aristocracy) and light infantry manpower (drawn from an unusually impoverished peasantry). VDH notes the low property qualifications for citizenship in Boeotia but does not stop to consider if that might be connected not to the hoplites, but to the unusually large numbers of Boeotian light infantry.

Moreover, there is a lack of clarity when presenting these percentages as to exactly what is being included. VDH’s 20% figure is 20% of the total “adult resident population,” rather than – as we might expect – a percentage of the adult male population or frequently the free adult male population. So he is actually asserting something like almost 45% (really probably 43 or 44%) of free households serve as hoplites (once we adjust for women and the elderly), which, as we’ll see, I think is pretty doubtful.⁴ For the sake of keeping comparisons here ‘clean,’ I am going to try to be really clear on what is a percentage of what, because as we’ll see there is in fact, a real difference between the orthodox assumption of a hoplite class of 40-50% of free households and the heterodox assumption that is closer to 25% of free households.

So when I say that heterodox scholars generally argue for a smaller, economically elite hoplite class while orthodox scholars generally assume a larger ‘yeoman’ hoplite class, it can be tricky to pin down what that means, particularly on the orthodox side. We need apples-to-apples number comparisons to get a sense of where these folks differ.

And I think the place to actually start with this is Karl Julius Beloch (1854-1929); stick with me, I promise this will make sense in a second. Beloch’s Die Bevölkerung der griechisch-römischen Welt (1886, “The Population of the Greco-Roman World”) is the starting point for all of the debates of Greek and Roman demography, the first really significant, systematic effort to estimate the population of the entire classical world in a rigorous way. Now if you recall your historiography from our first part, you will quickly realize that as a German writing in the 1880s, Beloch was bound to have drawn his assumptions about Greek society and the social role of the hoplite class from those early Prussian and German scholars who serve as the foundation for the orthodox school. They were, after all, writing at the same time and in the same language as he was. Equally useful (for us) Beloch’s basic range of estimates for Greece remain more-or-less the accepted starting point for the problem, which is to say that a lot of current historians of ancient Greece when they think about the population of the Greek poleis are still ‘thinking with Beloch’ (typically mediated by Corvisier and Suder, La population de l’Antitiquité classique (2000)).

So analyzing Beloch’s approach – and because he is estimating population, he is forced to use numbers – can give us a sense of the society that the ‘orthodox’ vision of hoplites imagined at its inception and which it largely still imagines when it thinks in terms of raw population numbers. And that can help us lock down what we’re actually arguing about.

In very brief, Beloch had a problem to solve in estimating the population of Greece. Whereas in Roman Italy, he had census data to interpret, we have no equivalent in Greece (ancient reports of population in Greece are rare and almost invariably unreliable). So instead he adopts the method of estimating from maximum military deployments, the one number we reliably get from ancient sources. Doing so, of course, requires squaring away some key questions: what percentage of adult males might be called up for these armies? Our sources often give us only figures for hoplites, so this question really becomes, ‘what percentage of adult males served as hoplites?’ And then following on that, what percentage of people were female, children, elderly or non-free?

Beloch answers those questions as follows: he assumes that roughly half of all free households are in the hoplite class, so he can compute the free adult male population by multiplying hoplite deployments by two, that he can compute the free population by multiplying the adult male population by three, and that the non-free population is around 25% of the total (significantly concentrated in Sparta and Athens), including both slaves and serfs. You can see the logic in these assumptions but as I am going to argue all of these assumptions are wrong, some more wrong than others. We’ll come back to this, but I think Beloch’s key stumbling block (apart from just badly underestimating the number of children in a pre-modern population – he should be multiplying his adult males by four, not three) is that he largely assumes that the Greek poleis look more or less like the Roman Republic except that the Romans recruit a bit further down their socio-economic ladder. And that’s…not right, though you could see how someone working in the 1880s might jump to that expedient when the differences in Greek and Roman social structure were less clear.

Greeks are not Romans and the Greek polis is not the Roman Republic.⁵

Nevertheless those assumptions suggest a vision, a mental model of the social structure of the typical Greek polis: wealthy citizens of the hoplite class make up roughly half of the free households (he explicitly defends a 47/53% breakdown between hoplite and sub-hoplite), while the landless citizen poor make up the other half. Beloch assumes an enslaved population of c. 1m (against a free population of c. 3m), so a society that is roughly 25% enslaved, so we might properly say he imagines a society that is roughly 37.5% hoplite class (or richer), 37.5% poorer households and 25% enslaved households. And returning to a moment to VDH’s The Other Greeks (1995), that’s his model too: if 20% of adults (not just adult males) were citizen-hoplites in Boeotia, then something like 43% of (free) households were hoplite households (remember to adjust not just for women, but also for the elderly),⁶ which is roughly Beloch’s figure. It is a touch lower, but remember that VDH is computing for Boeotia, a part of Greece where we expect a modestly larger lower class.

What does it mean for a society if the hoplite class represents approximately 40% of households (including non-free households)?

Well, this suggests first that the hoplite class is perhaps the largest or second-largest demographic group, behind only free poor citizens. It also assumes that nearly all of the propertied households – that is, the farmers who own their own farms – both served as hoplites and were members of the hoplite class.⁷ In particular, this imagines the ‘typical’ member of the hoplite class (this distinction between hoplites and the hoplite class will matter in a moment) as a middling farmer whose farm was likely small enough that he had to work it himself (not having enough land to live off rents or enslaved labor), essentially a modest peasant. Moreover the assumption here is that this broad hoplite ‘middle class’ dominates the demography of the polis, with very few leisured elites above them and a similar number of free poor (rather than a much larger number) below them.

And I want to note here again there is an implicit – only rarely explicit (Beloch makes the comparison directly) – effort to reason from the social model we see in the Roman Republic, where the assidui (the class liable for taxes and military service) as a group basically did include nearly all farmers with any kind of property and ‘farmers with any kind of property’ really does seem to have included the overwhelming majority of the population. There’s an effort to see Greek ‘civic militarism’ through the same frame, with the polis a community made up of small freeholding farmers banding together.⁸ I think scholarship has not always grappled clearly enough with the ways in which Rome is not like an overgrown polis, but in fact quite different. One of those differences is that the assidui is a much larger class of people than anything in a polis, encompassing something like 70% of all adult males (free and non-free) and perhaps as much as 90% of all free households. That is an enormous difference jumping even from 37.5% to 70%. What that figure suggests is both that Roman military participation reached much more robustly into the lower classes but also that (and we’ll come back to this in a moment) land ownership was probably more widespread among the Roman peasantry than their Greek equivalents.

In short part of what makes the Roman Republic different is not just where they draw the census lines, but the underlying structure of the countryside is meaningfully different and that has very significant impacts on the structure of Roman society.⁹ Taken on its own evidence, it sure looks like the organization of land in the Greek countryside was meaningfully less equal¹⁰ and included meaningfully more slaves than the Italian countryside, with significant implications for how we understand the social position of hoplites. And that brings us to the heterodox objections and thus…

Divisions Among Hoplites

The response to the ‘yeoman hoplite’ model of hoplite orthodoxy has been Hans van Wees’ assault on the ‘myth of the middle-class hoplite.’¹¹

What van Wees does is look specifically at Athens, because unlike anywhere else in the Greek world, we have the complete ‘schedule’ of wealth classes in Athens, denominated in agricultural production. He’s able to reason from that to likely estate size for each of the classes and from there, given the size of Attica (the territory of Athens) and the supposed citizen population (estimates from 40,000 to 60,000) the total size of each wealth class in terms of households and land ownership, in order to very roughly sketch the outlines of what wealth and social class in Attica might have looked like. Our sources offer little sense that they thought Athenian class structure was ever unusual or remarkable beyond the fact that Athens was very big (in contrast to Sparta, which is treated as quite strange), so the idea here is that insights in Athenian class divisions help us understand class divisions in other poleis as well.

What he is working with are the wealth classes defined by the reforms of Solon, which we haven’t really discussed in depth but these are reported by Plutarch (Solon 16) and seem to have been the genuine property classifications for Athenian citizens, which I’ve laid out in the chart below. Wealth was defined by the amount of grain (measured in medimnoi, a dry measure unit of 51.84 liters), but for non-farmers (craftsmen and such) you qualified to the class equal to your income (so if you got paid the equivalent of 250 medimnoi of grain to be a blacksmith, you were of the zeugitai, though one imagines fairly few non-landowners qualify for reasons swiftly to become clear).

Name	Wealth Requirement	Notional Military role	Percentage of Population Following van Wees (2001)
Pentakosiomedimnoi (“500 Bushel Men”)	500 medimnoi or more	Leaders, Officers, Generals	1.7-2.5%
Hippeis (‘horsemen’)	400 medimnoi	Cavalry	1.7-2.5%
Zeugitai (‘yoked ones’)	200 medimnoi (possibly reduced later to 150 medimnoi)	Hoplites	5.6-25%
Thetes (‘serfs’)	Less than 200 medimnoi	Too poor to serve (later rowers in the navy)	90-70%

Now traditionally, the zeugitai were regarded as the ‘hoplite class’ and that is sometimes supposed to be the source of their name (they were ‘yoked together’ standing in position in the phalanx), but what van Wees is working out is that although the zeugitai are supposed to be the core of the citizen polity (the thetes have limited political participation) there simply cannot be that many of them because the minimum farm necessary to produce 200 medimnoi of grain is going to be around 7.5 ha¹² or roughly 18 acres which is – by peasant standards – an enormous farm, well into ‘rich peasant’ territory. It is, in fact, roughly enough farm for the owner to not do much or any farming but instead subsist entirely off of either rents or the labor of enslaved workers.¹³

In short, the zeugitai aren’t ‘working class’ ‘yeoman farmers’ at all, but leisure-class elites – mostly landlords, not farmers – albeit poorer than the hippeis and pentakosiomedimnoi even further above them. And that actually makes a great deal of sense: one of the ideas that pops up in Greek political philosophy – albeit in tension with another we’ll get to in a moment – is the idea that the ideal hoplite is a leisured elite and that the ideal polis would be governed exclusively by the leisured hoplites.¹⁴ Indeed, when a bunch of Greek-speakers (mostly Macedonians) find themselves suddenly in possession of vast kingdoms, this is exactly the model they try to build their military on (before getting utterly rolled by the Romans because this is actually a bad way to build a society). And of course Sparta’s citizen body, the spartiates, replicate this model as well. Often when we see elements in a Greek polis try to create an oligarchy, what they are intending to do is reduce political participation back to roughly this class – the few thousand richest households – which is not all the hoplites, but merely the richest ones.

Of course with such large farms there can’t be all that many zeugitai and indeed there don’t seem to have been. In van Wees’ model, the zeugitai-and-up classes never supply even half of the number of hoplites we see Athens deploy; they only barely crawl over half if we assume the property qualification was (as it probably was) reduced at some point to just 150 medimnoi. Instead, under most conditions the majority of hoplites are thetes, pulled from the wealthiest stratum of that class (van Wees figures these fellows probably have farms in the range of ~3 ha or so, so c. 7.5 acres). Those thetes make up the majority of hoplites on the field but do not enjoy the political privileges of the ‘hoplite class.’ And pushing against the ‘polis-of-rentier-elites’ model, we often also find Greek sources remarking that these fellows, “wiry and sunburnt” (Plato Republic 556cd, trans. van Wees), make the best soldiers because they’re more physically fit and more inured to hardship – because unlike the wealthy hoplites they actually have to work.

What the transition to the Athenian democracy meant was the full enfranchisement of this large class of thetes, both the fellows who could afford to fight as hoplites (but previously didn’t have the rights of them) and the poorer citizen thetes.

And of course this isn’t only Athens. The only other polis whose complete social system we can see with any clarity, of course, is Sparta and when we look there, what do we find? A system where political participation is limited to the rentier-elite class (the Spartiates), where there is another class of poorer hoplites – the perioikoi, who fight as hoplites – who are entirely blocked from political participation. It appears to be the same kind of dividing line, with the difference being that the spartiates had become so dominant as to deny the perioikoi even citizenship in the polity and to physically segregate themselves (the perioikoi lived in their own communities, mostly on the marginal land). It is suggestive that this sort of divide between the wealthy ‘hoplite class’ that enjoyed distinct political privileges and other ‘working-class’ hoplites who did not (and yet even far more poor farmers who could not afford to fight as hoplites) was common in the polis.

That leaves the notion of a truly ‘broad-based’ hoplite-class that runs a ‘broad-based’ agrarian polis government that consisted of ‘middle-class’ ‘yeoman’ hoplites largely in tatters. Instead, what you may normally have is a legally defined ‘hoplite class’ that is just the richest 10-20% of the free citizen population, a distinct ‘poor hoplite’ class that might be around 20% and then a free citizen underclass of 60-70% that cannot fight as hoplites and also have very limited political participation, even though many of them do own some small amount of land.

Once again, if you’ll forgive me, that looks nothing like the Middle Roman Republic, where the capite censi (aka the proletarii) – men too poor to serve – probably amounted to only around 10% of the population and the light infantry contingent of a Roman army (where the poorest men who could serve would go) was just 25%.¹⁵ So whereas the free ‘Roman’ underclass of landless or very poor is at most perhaps 35% of (free) households,¹⁶ the equivalent class at Athens at least (and perhaps in Greece more broadly) is 60% of (free) households. Accounting for the enslaved population makes this gap wider, because it certainly seems like the percentage of the enslaved population in Greece was somewhat higher than Roman Italy. It is suddenly less of a marvel that Rome could produce military mobilizations that staggered the Greek world. Greeks are not Romans.

This is a set of conclusions that naturally has significant implications for how we understand the polis, particularly non-democratic poleis. Older scholarship often assumes that a ‘broad’ Greek oligarchy meant rule by the landholding class, but if you look at the number of enfranchised citizens, it is clear that ‘broad’ oligarchies were much narrower than this: not ‘farmer’s republics’ (as VDH supposes) but rather ‘landlord‘s republics.’¹⁷ That is quite a different sort of state! And understanding broad oligarchies in this way suddenly restores the explanatory power of what demokratia was in Greek thought: it isn’t just about enfranchising the urban poor (a class that must have been vanishingly small in outside of very large cities like Athens) but about enfranchising the small farmer, a class that would have been quite large in any polis for reasons we’ve discussed with peasants.

I think there’s also a less directly important but even more profound implication here:

Wait, How Many Greeks Are There?

The attentive reader may be thinking, “wait, but Beloch’s population estimates assume that the hoplite contingent of any Greek polis represent half of its military aged (20-60) free adult males, but you’re saying that number might be much lower, perhaps just 30 or 40%?”

I actually haven’t seen any scholars directly draw this connection, so I am going to do so here. Hell, I’ve already seen this blog cited quite a few times in peer-reviewed scholarship so why not.

If it isn’t already clear, I think when it comes to the size of the hoplite class, van Wees is correct and that thought interlocks with another thought that has slowly crept into my mind and at last become lodged as my working assumption: we have significantly under-counted the number of Greeks. Or, more correctly, everyone except Mogens Herman Hansen has significantly under-counted the number of Greeks. So good job to Mogens Herman Hansen, everyone else, see me after class.

Now these days the standard demographic reference for the population of Greece is not Beloch (1886), it is Corvisier and Suder, La population de l’Antitiquité classique (2000). Unlike Beloch, they do not reason from military deployments, instead they reason from estimated population density. Now I want to be clear, they are reasoning from estimated rural population density, which is not the same as reasoning from built-up urban area¹⁸ The thing is, we can’t independently confirm rural population density from archaeology (unlike urban area estimates) so this method is entirely hostage to its assumptions. So the fact that Corvisier and Suder’s estimates fall neatly almost exactly on Beloch’s estimate (a free population of c. 3m in mainland Greece) might suggest they tweaked their assumptions to get that result. And on some level, it is a circular process, because Beloch checks his own military-based estimates with population density calculations in order to try to show that he is producing reasonable numbers. So if you accept Beloch’s density estimates at the beginning, you are going to end up back-computing Beloch’s military estimates at the end, moving through the same process in reverse order.

But you can see how we have begun to trouble the foundations of Beloch’s numbers in a few ways. First off, we’ve already noted that his multiplier to get from military aged males to total population (multiply by three) is too low (it needs to be four). Beloch didn’t have the advantage of modern model life tables or the ability to see so clearly that mortality in his own day was changing rapidly and had been doing so for a while. Adjusting for that alone has to bring the free population up to support the military numbers, to around 4m instead of 3m (so we have effectively already broken Corvisier and Suder (2000)). Then there is the question of the prevalence of the enslaved; Beloch figures 25% (1m total), but estimates certainly run higher. Bresson, L’économie de la Grèce des cités (2007/8) figures perhaps 40-50% and 30% is also a common estimate, though we are here, in practice, largely guessing. Even keeping the 25% figure Beloch uses, which we now have to acknowledge may be on the low side, we have to raise the number of enslaved to reflect the larger free population: 1.33m instead of 1m, for a new total of 5.33m instead of Beloch’s original 4m.

But then if the number of men who fight as hoplites is not, as Beloch supposes, roughly half of polis society, but closer to 40% or even less, then we would need to expand the population even further. If it is, say, 40% instead of 50%, suddenly instead of Beloch’s computation (very roughly) of 500,000 hoplites giving us 1,000,000 free adult men giving us 3,000,000 free persons, resulting in a total population of 4,000,000 including the enslaved, we have 500,000 hoplites implying 1,250,000 free adult men implying 5,000,000 free persons, to which we have to add something like 1,500,000 enslaved persons¹⁹ implying a total human population not of 3 or 4m but of c. 6,500,000.

And there’s a reason to think that might be right. The one truly novel effort at estimating the population of Greece in the last few decades (and/or century or so) was by Mogens Herman Hansen. Having spent quite some time on a large, multi-scholar project to document every known polis (resulting in M.H. Hansen and T.H. Nielsen, An Inventory of Archaic and Classical Poleis (2004)), M.H. Hansen decided to use that count as a basis to estimate population, assigning a rough estimate to the size of small, medium and large poleis – using the built-up urban area of poleis we knew relatively well – and then simply multiplying by all of the known poleis to exist at one point in time. The result, documented in M.H. Hansen, The shogtun method: the demography and ancient Greek city-state culture (2006), produced an estimate of 4-6m for mainland Greece and I think, to be frank, Hansen pulled his punch here. His method really produced the top figure in that range, a significantly higher figure that generally postulated for Greece.²⁰

My strong suspicion – which the evidence is insufficient to confirm definitively – is that van Wees is right about the relative size of the slice of men who fight as hoplites (distinct from the ‘hoplite class’) and that M.H. Hansen is correct about the population and that these two conclusions interlock with each other to imply a rather different Greece in terms of equality and social structure than we had thought.

Looping back around to what is my repeated complaint this week: we were often conditions to think about Greek agriculture, the Greek peasantry, the Greek countryside through the lens of the much better documented Roman Italian agriculture, peasantry and countryside. After all, it is for Italy, not Greece, that we have real census data, it is the Roman period, not the classical period, that gives us sustained production of agricultural treatises. We simply have a much better picture of Roman social structures and so it was natural for scholars trying to get to grips with a quite frankly alien economic system to work from the nearest system they knew. And that was fine when we were starting from nothing but I think it is a set of assumptions that have outlived their usefulness.

This isn’t the place for this argument in full (that’s in my book), but briefly, the structure of the Roman countryside – as we come to see it in the late third/early second century BC – did not form naturally. It was instead the product of policy, by that point, of a century’s worth of colonial settlements intentionally altering, terraforming, landholding patterns to maximize the amount of heavy infantry the land could support. It was also the product of a tax-and-soldier-pay regime (tributum and stipendium) that on the net channeled resources downward to enable poorer men to serve in that heavy infantry.²¹ Those mechanisms are not grinding away in mainland Greece (we can leave Greek colonial settlement aside for now, as it is happening outside of mainland Greece), so we have no reason to expect the structure of the countryside to look the same either.

In short the Romans are taking steps to ‘flatten out’ their infantry class (but not their aristocracy, of course), to a degree, which we do not see in Greece. Instead, where we get an ideology of economically equal citizenry, it is an ideology of equality within the leisured elite, an ‘equality of landlords’ not an equality of farmers. We should thus not expect wealth and land distribution to be as ‘flat’ in Greece as in Italy – and to be clear, wealth distribution in Italy was not very flat by any reasonable standard, there was enormous disparity between the prima classis (‘first class’) of infantry and the poorest Roman assidui. But it was probably flatter than in Greece within the infantry class (again, the Roman aristocracy is a separate question), something that seems confirmed given that the militarily active class in Roman Italy is so much larger and more heavily concentrated into the heavy infantry.²² Consequently, we ought not assume that we can casually estimate the total population of Greece from hoplite deployments, supposing that the Greeks like the Romans, expected nearly all free men to serve. Instead, the suggestion of our evidence was that in Greece, as in many pre-modern societies, military service (and thus political power) was often the preserve of an exclusive affluent class.

Implications

But returning to Greece, I would argue that accepting the heterodox position on the social status of hoplites has some substantial implications. First, it suggests that there was, in fact, a very real and substantial social division within the body of hoplites, between wealth hoplites who were of the ‘hoplite class’ as politically understood and poor hoplites who fought in the same way but only enjoyed a portion of the social status implied. That division suddenly makes sense of the emergence of demokratia in poleis that were more rural than Athens (which is all of them). The typical polis was thus not a ‘farmer’s republic’ but a landlord’s republic.

At the same time, this also substantially alters the assumptions about ‘yeoman hoplites’ who have to rush home to pull in their harvests and who are, in effect, ‘blue-collar warriors.’ Instead, the core of the hoplite army was a body – not a majority, but a significant minority – of leisured elites who had slaves or tenants doing most of their farming for them. What kept hoplite armies from campaigning year-round was as much poor logistics as yeoman economics (something clear in the fact that spartiates – by definition leisured elites – didn’t campaign year-round either).

Finally, if we extend this thinking into our demographic analysis, we have to accept a much larger population in Greece, with all of the expansion happening below the men who fought as hoplites (both the hoplite class and our poorer working-class hoplites). It suggests a remarkably less equal social structure in Greece – indeed, perhaps less equal than the structure in Roman Italy – which in turn significantly caveats the way we often understand the Greek polis as a citizen community relatively more egalitarian and free than the absolute monarchies which pervaded Egypt and the Near East.

And of course, for one last return to my pet complaint in this post, it should reinforce our sense that Greek are not Romans and that we cannot casually supply the habits, economics or social structures of one society to the other to fill in gaps in our evidence. In particular, the assumption that the Greeks and Romans essentially share a civic and military tradition is a thing that would need to be proved, not assumed.²³

1. e.g. ch4 of van Wees, Greek Warfare: Myths and Realities (2004) is, “Men of Bronze: the myth of the middle-class militia” and his 2001 chapter, “The Myth of the Middle-Class Army” in T. Bekker-Nielsen/L. Hannestad (eds.), War as a Cultural and Social Force: Essays on Warfare in Antiquity (2001).
2. VDH, The Other Greeks (1995), 205-7. I have a digital copy of this (it’s a kindle version) which gives page numbers, but I do not know if these match the page numbers in the print version.
3. op. cit., 210.
4. Quickly on the math: if 20% of the adult population are hoplites, then shifting our statistic to the adult male population is going to basically double that figure to 40%. Factoring in the small but meaningful number of men over 60 who no longer serve is going to inch it up further, close to 45%. So the suggestion here is that nearly half of all households can furnish a hoplite, which we’re going to see in a second is exactly what I think VDH and other orthodox scholars mean. But notice how tricky that statistic is: we could say that under VDH’s model, 45% of households serve as hoplites in Boeotia, but flip that around and we could equally say with the same model that less than ten percent of free Boeotians at any given time were eligible to serve as hoplites, accounting for the fact that half of this society will be female and nearly 55% of it will be underage, so ‘adult males’ only make up around 22.5% of the population, so 45% of 22.5% ends up being almost exactly 10%.
5. While we’re beating up on VDH, at the start of his chapter on the ideology of the polis, he quotes Pliny, Natural History 18.4 (18.4.18), “Anyone for whom seven acres are not enough is a dangerous citizen,” which just about broke me because of course first Greeks are not Romans and this is Pliny the Elder (d. 79 AD) quoting Manius Curius (d. 270 BCE at Rome) and also because it is a mistranslation: it is not seven acres but seven iugera, which is 4.36 acres (he’s also somewhat mangled the relative clause of characteristic here). Needless to say, Pliny the Elder cannot be an authority for the ideology of the Archaic Greek polis but also the fact that the ‘ideal’ Roman farm suggested here is significantly smaller than the figures we’re going to see in a moment for the Greeks is also a tell that the Romans maybe do not have the same assumptions about the wealth of their heavy infantry.
6. Which I may note VDH fails to do, but men above fighting age (sixty for a Greek polis, generally) would make up about 2.5 of the population (and women of that age another 3.5%) even under the dismal estimated mortality figures in antiquity.
7. I am struck by Hanson’s use of ‘yeoman’ here because he may be thinking of that social class in Colonial North America, unaware that in Britain there was in fact a significant class of freeholding farmers poorer than the yeomanry, called the ‘husbandry.’
8. Again, Hanson makes this badly flawed ‘Greece and Rome are basically the same’ sort of argument more or less explicitly in his later works, like Carnage and Culture (2001). His grasp on the Roman Republic is, frankly, quite poor.
9. The most obvious is the enormous role of patronage in Roman social structures; parallel social customs in Greece are much less common and weaker.
10. In particular, to clarify this point, the upper-end of the Roman aristocracy were probably meaningfully wealthier even before massive Roman expansion, than their Greek peers, so inequality at the top of the distribution was greater in Italy. Certainly the ability of those aristocrats to mobilize hundreds of clients is something we do not see often in Greek poleis and suggestive of the greater power Roman aristocrats wielded. But underneath those ‘big men,’ I think our evidence suggests that Italy had more middling and small farmers and fewer ‘very rich peasants’ than Greece, in part because the very action of Roman conquest and land distribution seems to have purposefully produced that effect to maximize the number of households liable for conscription.
11. As noted above, this is “Men of Bronze: the myth of the middle-class militia” in his monograph, Greek Warfare: Myths and Realities (2004) and his 2001 chapter, “The Myth of the Middle-Class Army” in T. Bekker-Nielsen/L. Hannestad (eds.), War as a Cultural and Social Force: Essays on Warfare in Antiquity (2001). The latter is more detailed to the point and mostly what I am covering here.
12. van Wees does his calculations in hectares.
13. That is, 200 medimnoi of grain is enough to feed the necessary enslaved work force and then still maintain the owner’s household in relative comfort. You’ll recall we did our peasant math in modii (8.73 liters) and our households needed around 200-400 modii of grain to survive and 400-800 to life in maximum comfort. The medimnos is a bigger unit (51.84 liters) so 200 medimnoi is 1,187 modii, more than enough to maintain one complete peasant household (of slaves) at subsistence and then another complete household (of leisured elites) in comfort.
14. This idea is discussed in Pritchard, D.M.  “Thetes, Hoplites, and the Athenian Imaginary,” in Ancient History in a Modern University, eds. T.W. Hillard, R.A. Kearsley, C.E.V. Nixon and A.M. Nobbs (1998) and also van Wees (2004), 55-7.
15. On this point, see N. Rosenstein, Rome at War: Farms, Families and Death in the Middle Republic (2004) which has an appendix directly addressing the size of the capite censi.
16. Keep in mind that a good number of the men serving as velites at any given time were doing so not because they were too poor to be hastati, but because they were too young: both the very young and the very poor get put out in front as skirmishers.
17. This point is made quite effectively in M. Simonton, Classical Greek Oligarchy: A Political History (2017).
18. As some works on the demography of Roman Italy have been able to do, e.g. de Ligt Peasants, Citizens and Soldiers (2012).
19. 1,666,666 to be precise in keeping the 25% ratio, but I want to keep my numbers here clean in part to be clear what wild estimates they are
20. MHH revisited these calculations and largely stood by the upper-end estimates (which I think is also indicative of how the upper-end estimate is the ‘real’ estimate) in M.H. Hanson, “An Update on the Shotgun Method” GRBS 48 (2008).
21. Because the tributum was collected as a tax proportional to wealth, but wages (stipendium) were paid out on a flat basis with all infantry receiving the same wage, so poorer infantrymen paid lower taxes when they weren’t serving but received the same wage back as their richer compatriots who paid heavier taxes. Thus poorer assidui end up as net recipients of tax money while rich assidui end up as net contributors. This argument on tributum relies very heavily on the excellent scholarship of James Tan on the nature and function of tributum.
22. Interestingly, the prima classis of Roman infantry, at around a quarter to a third of the total number of the pedites (very roughly estimated) approximates the slice of men in Greece serving as hoplites, which makes sense since our sources seem to think – there is some dispute as to if they think rightly – that the prima classis began as the part of Roman society that fought as hoplites. What is different is that the next set of fellows down the social ladder (the next four classes of pedites) in Roman Italy also fight mostly as armored heavy infantry (albeit with worse armor), whereas in Greece the ‘subhoplites’ fight as psiloi, ‘lights’ – skirmishers with very cheap equipment. That is, it turns out a big difference. if I am right above, mainland Greece actually had somewhat more total population than Roman Italy c. 212, but if you bolted all the poleis together you would still never match the Roman deployment of 185,000 men in that year (the largest combined polis army is at Plataea in 479, was probably around 80,000 men, with perhaps another 20,000 Greeks (or somewhat more) on the other side). It turns out spending a century terraforming Italy into an infantry-generation-machine generates a lot of infantry.
23. Once again, the idea that Rome was a simple extension of the Greek civic and military tradition is a foundational assumption (read: mistake) of VDH’s Carnage and Culture (2001), which uses this slight of hand to conjure a continuous ‘western’ military tradition (which somehow goes on holiday for most of the Middle Ages) for societies that have functionally no connection at all to the ancient Greek military tradition.