Here’s a question for you: why were there only 300 Spartans at Thermopylae? According to popular history, this was supposed to be an all-or-nothing battle for the future of the Classical Greek World, a full-effort attempt to stop an immense Persian army from wiping out the city-states of the entire peninsula (a questionable assertion, but that’s a different discussion). Other, much smaller city-states sent more troops. Thespiae sent 700, Phocis sent 1000. Why did Sparta, a much larger Polis with tens of thousands of residents, only send 300 to fight a battle which would have supposedly demanded their maximum effort?
The answers to that question serve as an example of the factors which limit the size of any army, real or fictional – factors which any worldbuilder should really pay attention to if they want to have the sizes of their fighting forces realistically match the societies that field them.
So, back to our original question, let’s answer the second part of it first: how did other, much smaller city-states contribute more hoplites than Sparta, a major power of the region? This comes down a great deal to how the Classical Greek military system worked.
In short, Classical Greek City-States mostly fought wars with each other, which meant that military campaigns tended to be very short by our standards, and that armies didn’t have to travel very far. In addition, it meant that any decisive campaign was likely to end with one army besieging and potentially sacking the home city of another – which meant that war was short-term, short-range, and very high stakes. This led to a military system where individual city-states wanted to field as many useful troops as possible for a short amount of time.
Basically, how this worked was that every citizen in a city-state (citizen in this case meaning a male adult landowner) was required to maintain arms and armour for military service. In a time of war, those citizens would be mobilised as soldiers. In theory, a citizen’s rights were contingent on their ability and willingness to be called out to the battlefield in a season of war, but because these wars rarely lasted longer than a season, it meant that the citizen-soldier himself was not required to be a professional soldier. It was enough for him to own a shield, a spear, and a helmet, and be able to respond to the relatively simple commands needed to point a Hoplite phalanx at the enemy phalanx and send it charging forward.
This meant that relatively small city-states (with populations of a few thousand) could send much larger proportions of their population into battle simply because they were able to mobilise anyone who could afford their own weapons. When this system was transplanted to the much more populous and expansive Roman Republic, it led to the sort of gigantic (50 000+ soldiers) armies we see regularly raised during the Punic Wars. Of course, this system also had its disadvantages. Part-time soldiers can’t stay at war forever, and their nature as non-professionals means that preparing for war isn’t their main occupation. If a campaign requires an army of citizen-conscripts to remain in the field for years, or needs it to fight an army of professionals, it’ll usually lose. This is why the Romans eventually switched to an army of long-service professional volunteers – as did the United States, when its previous system of armed citizen militias proved woefully inadequate against far more capable indigenous fighters during conflicts like Little Turtle’s War in the 1790s.
It should be noted that professional armies tend to be smaller than part-time ones, simply because while part-time soldiers go back to their day jobs when the fighting’s over, professional soldiers don’t have a separate civilian job where they produce things for the local economy. While trained soldiers are an invaluable human resource in a time of war, they don’t contribute anywhere near as much to economic growth or standard of living in a time of peace as scientists, farmers, artisans, or poets. As a result, professional armies tend to rely on quality over quantity, and societies which have professional armies but haven’t been under existential threat for a long time tend to keep those armies far smaller than what they’re capable of. Even the United States – current poster-child for military overspending – only spends about half to a fifth of what it did as a proportion of GDP during the Cold War when armed conflict with a peer rival (the Soviet Union, which spent an even larger proportion of its GDP on its military) seemed imminent.
Of course, there were limits to the Classical Greek system as well. It only applied to the citizens of a city-state, not all its residents. Women, slaves, and immigrants were generally excluded from this (the reputation of Classical Greece as the brithplace of democracy is vastly overstated, in my opinion), and this applied in more or less every Classical Greek city-state. However, Sparta in particular possessed a much more restrictive citizenship than even most other city-states, leading to a situation where its larger population did not necessarily correspond to a proportionally larger body of citizen-soldiers – Spartiates – as other, smaller city-states. Part of this was because Sparta had a much more regimented citizen-society than most city-states, with citizens more or less living under military discipline and education from birth and with a far higher proportion of children “washing out” before reaching adulthood.
The other reason was that Sparta – to an even greater extent than most other Classical Greek city-states – was a city of slaveowners.
When people say that Sparta sent 300 soldiers to the Battle of Thermopylae, they’re only telling half of the truth. In reality, Sparta sent 300 Spartiates to Thermoplyae, and they were accompanied by three times as many slaves and other non-citizens acting as servants and support personnel. Sparta as a whole actually had an even greater proportion of slaves – Helots – to citizens – Spartiates – especially as time went on and the Spartiate class continued to dwindle. While the Spartiate class was able to devote all their time to military training, politics, and cracking one-liners, it was the Helot class which worked the fields, and kept the city running – usually while living under brutal discipline and in the kind of horrible conditions you normally find in 19th century chattel slavery. Naturally, this meant that the Helots ran away, disobeyed orders, and rebelled – quite often.
This of course, posed a problem to the Spartiate Class – they could either relent and give the Helots the same modicum of rights (like, for example, the right not to be hunted for sport by the Spartiate class) that other city-states offered, or they could simply repress the Helots harder. Guess which one they picked?
That, in turn, answers the other half of the question. Sparta only sent 300 Spartiates to Thermopylae because they needed the rest at home keeping the Helots in line. If they’d sent more, the Spartiates would risk what their latter day equivalents in the American South called “servile insurrection”, and while the Persians could be negotiated with (indeed, the Spartans did negotiate with them throughout the time period) the Helots – whose grievances were personal rather than geopolitical – were far less likely to acquiesce, especially when they knew that the only position the Spartiate class would accept was a return of the Helots to the most abject slavery imaginable. As a result, the Spartiates prioritised keeping their own slaves in line over this supposedly existential (but not really) threat, and distributed their military resources accordingly.
So what does this tell us about sizing armies in general? First of all, it tells us that the total number of troops a given society can put into the field is limited to a relatively small proportion of the society as a whole: even at maximum effort, a part-time citizen army like that of most Greek City-States represented maybe a tenth of its total population, with the rest being invalidated by gender, age, class, or political status. With a professional army in a society which has not known existential threat for a long time, this proportion is even smaller. The US Armed Forces has something like 1 active service member per 250 citizens. The Canadian Armed Forces maintain an even smaller proportion – more like 1 to 500.
Likewise, the full force of these militaries won’t be concentrated into a single big field army. Generally speaking, a large proportion – if not most – of a military’s human resources will be committed elsewhere: garrisoning fortifications, maintaining weapons and equipment, serving as police or internal security, or recruiting and training new units. As a result, an army which might boast hundreds of thousands of troops in total might only be able to deploy tens of thousands of troops in a single field army. In addition, a huge part of that army will be made up of servants, camp followers, and rear-echelon troops like pay clerks, engineers, farriers, staff officers, and other personnel who aren’t expected to actually go into battle. As a result, the number of troops which actually go into combat is a fraction of a fraction of the total mobilised armed forces – which is in turn a relatively small fraction of a society’s population – and these fractions tend to get smaller and smaller as war becomes more complex, requiring more civilian support on the home front (in terms of things like industrial production) while also requiring more specialists and technicians to allow for a field army to fight more effectively.
And none of this is getting into logistical restrictions. Just like the political entities we covered in last month’s article, armies are limited by technology: namely the ability to bring up all the food and supplies needed to keep that army fed and in fighting shape. Supplies – especially supplies pulled by animals who also inconveniently require constant supplies of food and water – are a major limit to how big armies can get in the field in more ways than one.
But that’s something for next time. This article has already gone long enough, and I’m going to have to split this into a two-parter.
Paul Wang 