Could the Persian Empire logistically support an army several million strong?

[Erpingham]

I of course bow to your personal knowledge of squatter camps and I see the sense in your comment about them being organised - I can't see how they would work otherwise.  I still have my doubts that they are a good model for an ancient army though.

[Patrick Waterson]

(Students of that conflict will recall the communist Chinese managed to move massive armies around in mountainous areas and keep them supplied, albeit not by Achaemenid methods!)

I'm no expert on the Korean War but I'm not sure this is really compatable.  The Chinese deployed a smaller army over a much wider front and weren't tied to naval supply.

As I pointed out, they did not use Achaemenid methods.  The essence of my point is that we tend to be too 'road-bound' in our military thinking; the PRC put 400,000 men into terrain where MacArthur did not think they would be able to deploy 100,000.

I am a bit baffled by the approach to modern analogy of the "source first" group.  Maurice has been rejected as he is too early 20th century British but Lawrence of Arabia has been cited with approval.  18th century practice is right out but the PVA in 1950 is OK.  Modern US and Pakistan army experience with mule transport is disregarded as too humane (as opposed to being based on operational requirements).

Then let me explain. 

Lawrence is not trying to perform logistical calculations for Herodotus but for troops under his care.  We can therefore take his figures as applicable to his situation; whether they also apply to 5th century BC Achaemenids is a moot point: their importance is to show that Maurice's staff figures are not universal and in particular seem overstated in a non-European Near Eastern context.

The communist Chinese are invoked not to advocate an inapplicable supply system but to demonstrate that we English-speakers tend to be too wedded to communications routes when considering what an army could or could not do, and this despite some very recent examples of our getting it wrong because other cultures do things differently.

Modern US and Pakistani mule transport might not be the best guide to how mules were loaded.

This site contains the following intriguing snippet:
General George Crook, in the late 1870's, preferred to ride his mule "Apache", which he considered much superior to the horse, and he continually stressed the importance of having healthy pack mules under his command. He believed that the success of any campaign, to a great extent, depended upon them. General Crook's mules easily carried twice the load the Army manual stipulated because he allowed only the best equipment to be used on the best mules - - and each pack saddle was tailored to fit each mule. Crook's troops always had the ammunition they needed because his mule trains never failed.

We do not know what carrying harnesses, panniers etc. the Achamenids used, but we do know they came from a long tradition of mule use.  I therefore suggest they had learned to optimise the use of their mules, notably by using equipment fitted to the individual animal, and hence their loads - which, as explained earlier, were not constant but regularly diminishing - could easily be double the sustained loads carried by more modern mules such as those cited by our Mr Clipsom.

How the mighty have fallen.  We've gone from one of the most highly organised armies in pre-modern history to a migrating squatter camp of gigantic proportions.  Perhaps time for some reflection on the difference? I don't think the highly ordered supply and march routine and the mobile squatter camp are easy bedfellows.

I know it is hard to resist a drift into sarcasm, but you must resist.  

I'm also wedded in numbers games to what I was taught in physics at school - does your answer look right when compared with other things you know?  So, in this case, if all other armies in Europe for which we have reliable figures in the pre-modern era are under 600,000 strong, is there a good case why this one should buck the trend?

Yes, and a very good one.  Although previous Achamenid armies, or at least those for which we have source numbers, did not exceed 700,000 men, this one represented a full mobilisation of the Empire, or as full as could be managed without leaving everything defenceless, and was the first and last time this was done for the purpose of expanding the Empire.  There would be later mass mobilisations for defensive purposes, threatened Achaemenid monarchs massing armies recorded as being around the 1 million mark in Mesopotamia (Artaxerxes II in 401 BC; Darius III in 331 BC) but never again for a conquest or reconquest.  This was a unique event, and like all unique events, does not fit comfortably into a standard pattern.

[Patrick Waterson]

I of course bow to your personal knowledge of squatter camps and I see the sense in your comment about them being organised - I can't see how they would work otherwise.  I still have my doubts that they are a good model for an ancient army though.

They may or may not be, but they do show the density achievable for a sustained human settlement*, thus indicating a transient military camp could easily be more densely inhabited.

*In particular, one without multi-storey constructions, which would considerably distort the picture.

[Prufrock]

The point is that the army does not march through any narrow gaps that oblige it to contract to a column 8 men wide. It marches in a column or columns several hundred yards wide with the column(s) about 20 km long or so, which causes as much inconvenience as the columns of a regular army.

That's not what it means at all. It means that they didn't use the Tempe pass. It does not mean that they had a mountain superhighway over which they could march several million men in columns 200 men wide with neat spacing between ranks while also accommodating cavalry horses, pack animals and the supply train.

I'm trying to form an idea of just how difficult the ground north of the pass is. Here's an image of a Drakensberg hike I did, from Giant's Castle hotel (red dot at bottom) to Bannerman's hut (red dot at top). The distance is 7km and took about 4 hours. It's what hiking books would term 'difficult'. Tiring but not really exhausting and not dangerous.



Here, on the same scale, is the hilly ground north of the river valley. Distance from the flat plain below to the summit ridge is about 4km; distance from the summit ridge to the plain beyond is about 3km. It's less steep than the Drakensberg climb. I'd say quite manageable for men and animals in a number of columns that avoid the steep bits. The traversable area looks to be about 1,5km wide.

There is a fair old difference between a fellow going on a hike with a backpack for fun and 3,500,000 + fellows going to war and relying on cavalry mounts, animals, and baggage trains for their survival now and over the coming weeks. If we use other ancient hillwalkers as examples, Xenophon and the 10,000 valued their roads, and while delegated portions of the 10,000 could go over the rough portion of a hill if they absolutely needed to, it was out of military necessity, and not part of everyday movement. It took its toll when you look at the rest days required, and these guys were both fit and motivated.

And the Achaemenids had a thing for roads - they had the royal ones, of course - so they knew how to make them. But their royal roads measured 7m across, not the 200m plus that has been asserted applied to armies of this size. What evidence is there for these massively wide columns?

Is there any positive evidence for column widths of the size here maintained?

[Dangun]

With this in mind, 1000 people per hectare is workable for a permanent settlement.

If we compare Mongkok in Hong Kong to Kibera, Kenya it is interesting that the application of technology and capital doesn't do a lot to maximum population density - it rises from about 90,000/km2 in Kibera to 130,000/km2 in Mongkok. But I live on the 44th floor in HK which was unlikely an option for a Persian camp.

For the perfect s%t-storm of technology and poverty, consider Kowloon's Walled City, also in Hong Kong, which peaked out a population density of 1.3million per km2 - about 0.8sqm per person! Again, not relevant, but interesting.

But something that IS relevant, is that Kibera, or Kowloon Walled City are only 1 to 10% the size of our posited Persian army. So NONE OF THEM have to grapple with the same walking-to-edge-of-camp problems that the posited Persian camp would. In other words, the evidence suggests that as scale increases density must decrease - there are no slums in the world as big as this Persian camp.

What you want to be true is really that unprecedented.

[Justin Swanton]

There is a fair old difference between a fellow going on a hike with a backpack for fun and 3,500,000 + fellows going to war and relying on cavalry mounts, animals, and baggage trains for their survival now and over the coming weeks. If we use other ancient hillwalkers as examples, Xenophon and the 10,000 valued their roads, and while delegated portions of the 10,000 could go over the rough portion of a hill if they absolutely needed to, it was out of military necessity, and not part of everyday movement. It took its toll when you look at the rest days required, and these guys were both fit and motivated.

Sure. The kind of difficult patch I gave above was exceptional: most of the time the army was walking on flat ground. It went the hilly route only because there was no other option.

Hiking and military marching are comparable IMHO with two important differences: a hiker like myself is not really fit when he goes on a hike, whereas the Persian infantry, who came from far more active backgrounds, had been marching for weeks and were as fit as fiddles. On the other hand, a hiker uses trails which wind up a slope, substantially easier than going straight up over virgin ground. I suspect the two more-or-less cancelled out, so that a Persian infantryman tramping over hilly ground didn't enjoy himself but wasn't clapped out at the end.

And the Achaemenids had a thing for roads - they had the royal ones, of course - so they knew how to make them. But their royal roads measured 7m across, not the 200m plus that has been asserted applied to armies of this size. What evidence is there for these massively wide columns?

Is there any positive evidence for column widths of the size here maintained?

That's the point of this thread. It's a feasibility study of the ability of the Persian empire to maintain and move such huge armies in the field. If there was conclusive evidence the Persian army did march that wide then we wouldn't be having this discussion.   If 3,4 million men are to move from one campsite to another in a single day they have no choice but to march in a column several hundred men wide - not in neat parade-groups files and ranks of course, just wide. They can't do it any other way. There are indications the army did indeed march this way: the Thracians who don't cultivate the ground the Persians have trod, the need for the army to clear a forest in Macedonia during their passage (why clear a forest if they can just use the regular roads?).

Let me emphasize that this is just a feasibility study. Nobody is being asked to believe the Persian army was in fact several million strong. One can argue feasibility and disbelieve Herodotus' numbers for other reasons.

[Justin Swanton]

With this in mind, 1000 people per hectare is workable for a permanent settlement.

If we compare Mongkok in Hong Kong to Kibera, Kenya it is interesting that the application of technology and capital doesn't do a lot to maximum population density - it rises from about 90,000/km2 in Kibera to 130,000/km2 in Mongkok. But I live on the 44th floor in HK which was unlikely an option for a Persian camp.

For the perfect s%t-storm of technology and poverty, consider Kowloon's Walled City, also in Hong Kong, which peaked out a population density of 1.3million per km2 - about 0.8sqm per person! Again, not relevant, but interesting.

But something that IS relevant, is that Kibera, or Kowloon Walled City are only 1 to 10% the size of our posited Persian army. So NONE OF THEM have to grapple with the same walking-to-edge-of-camp problems that the posited Persian camp would. In other words, the evidence suggests that as scale increases density must decrease - there are no slums in the world as big as this Persian camp.

What you want to be true is really that unprecedented.

There are squatter camps that are comparable - that one example I gave from Nairobi numbering 800 000 to 1 000 000, jammed far closer together than I would conceive for the Persian camp, in a permanent settlement, and it still manages to function more-or-less.

[Jim Webster]

2) Young assumes no fodder anywhere along the route, despite the Persians having marched 'at the best season of the year' (Herodotus VII.50).

The best season for marching is not the best season for grass growth
We know they were marching in the dry season. So grass was not growing. It might have been 'ripening' but that merely means its feed value was dropping. Indeed you would start to see the phenomena that is known as 'burning off' where grass effectively 'disappears' because the the heat and dryness.
So once the fodder had gone, it had gone, it would not be replaced.
So perhaps Young might have a better idea about what he's talking about that somebody who has never had to farm grass or forage?

[Justin Swanton]

2) Young assumes no fodder anywhere along the route, despite the Persians having marched 'at the best season of the year' (Herodotus VII.50).

The best season for marching is not the best season for grass growth
We know they were marching in the dry season. So grass was not growing. It might have been 'ripening' but that merely means its feed value was dropping. Indeed you would start to see the phenomena that is known as 'burning off' where grass effectively 'disappears' because the the heat and dryness.
So once the fodder had gone, it had gone, it would not be replaced.
So perhaps Young might have a better idea about what he's talking about that somebody who has never had to farm grass or forage?

The harvest in Greece is in June. The army is marching from April to, say, September. Does that allow for decent grazing for the animals?

[Jim Webster]

There are squatter camps that are comparable - that one example I gave from Nairobi numbering 800 000 to 1 000 000, jammed far closer together than I would conceive for the Persian camp, in a permanent settlement, and it still manages to function more-or-less.

Squatter camps are not comparable to ancient military camps.
Firstly each unit has to have room to form up as a military formation without having to take down its tents  (should they have any) because the unit might have to leave the camp ready to fight.
There have to be lanes through the camp wide enough and clear enough to let the equivalent of the military police move at speed to break up trouble
These same lanes will be used to let the formed up units march out of the camp without having to march through the camping areas of other troops

I think we take the Roman camps as an absolute minimum for the amount of size needed

[Justin Swanton]

There are squatter camps that are comparable - that one example I gave from Nairobi numbering 800 000 to 1 000 000, jammed far closer together than I would conceive for the Persian camp, in a permanent settlement, and it still manages to function more-or-less.

Squatter camps are not comparable to ancient military camps.
Firstly each unit has to have room to form up as a military formation without having to take down its tents  (should they have any) because the unit might have to leave the camp ready to fight.
There have to be lanes through the camp wide enough and clear enough to let the equivalent of the military police move at speed to break up trouble
These same lanes will be used to let the formed up units march out of the camp without having to march through the camping areas of other troops

I think we take the Roman camps as an absolute minimum for the amount of size needed

Roman camps are estimated at an upper limit of 1186 men per hectare. If we make the Persian camp 1000 men per hectare would that be reasonable?

[Jim Webster]

2) Young assumes no fodder anywhere along the route, despite the Persians having marched 'at the best season of the year' (Herodotus VII.50).

The best season for marching is not the best season for grass growth
We know they were marching in the dry season. So grass was not growing. It might have been 'ripening' but that merely means its feed value was dropping. Indeed you would start to see the phenomena that is known as 'burning off' where grass effectively 'disappears' because the the heat and dryness.
So once the fodder had gone, it had gone, it would not be replaced.
So perhaps Young might have a better idea about what he's talking about that somebody who has never had to farm grass or forage?

The harvest in Greece is in June. The army is marching from April to, say, September. Does that allow for decent grazing for the animals?

No
It's as simple as that. Study the agriculture of Greece, they didn't rely on Grazing animals because the vast amount of the country was not fit for grazing. (Hence the use of olive oil rather than butter)
They also used to move what grazing animals they did have up into the mountains because that is where the grazing was in summer.
If you want to move large numbers of horses and graze or collect fodder as you go, then the rule of thumb is to move them through areas where cattle graze all the year round.

[Jim Webster]

There are squatter camps that are comparable - that one example I gave from Nairobi numbering 800 000 to 1 000 000, jammed far closer together than I would conceive for the Persian camp, in a permanent settlement, and it still manages to function more-or-less.

Squatter camps are not comparable to ancient military camps.
Firstly each unit has to have room to form up as a military formation without having to take down its tents  (should they have any) because the unit might have to leave the camp ready to fight.
There have to be lanes through the camp wide enough and clear enough to let the equivalent of the military police move at speed to break up trouble
These same lanes will be used to let the formed up units march out of the camp without having to march through the camping areas of other troops

I think we take the Roman camps as an absolute minimum for the amount of size needed

Roman camps are estimated at an upper limit of 1186 men per hectare. If we make the Persian camp 1000 men per hectare would that be reasonable?

Given that we have no evidence whatsoever that the Persians had camps organised on anything like the efficiency of the Romans we are flattering them. Certainly the Greeks didn't didn't have anything that would pass as an organised camp that Romans would recognise, and they never commented on the Persians being remarkably efficient
But because we need a rule of thumb than if you take a thousand man per hectare, this gives you the absolute minimum camp size.
Given at times the army plus baggage seems to have been anywhere up to five or six million strong, the camp north of Thermopylae for example, could have quite a size

[Dangun]

There are squatter camps that are comparable - that one example I gave from Nairobi numbering 800 000 to 1 000 000, jammed far closer together than I would conceive for the Persian camp, in a permanent settlement, and it still manages to function more-or-less.

The present-day evidence suggests it is not comparable.

800,000 is a quarter of the size of your Persian camp.
The largest dense slums on the planet today are Dharavi, Kibera etc., nothing much more than a million people.

The slums that exist today which are larger, e.g. the largest - Orangi (2mn people) is much less dense.
This is probably telling us something.

Conceptually a slum/camp is limited by how far you can walk in a day to get food, fresh water and bring it back again, or dispose of your waste.
This limits the maximum size, as observed by... Maurice.

Herodotus' Persian camp is not just unprecedented at the time, but without precedent since.

[Justin Swanton]

There are squatter camps that are comparable - that one example I gave from Nairobi numbering 800 000 to 1 000 000, jammed far closer together than I would conceive for the Persian camp, in a permanent settlement, and it still manages to function more-or-less.

The present-day evidence suggests it is not comparable.

800,000 is a quarter of the size of your Persian camp.
The largest dense slums on the planet today are Dharavi, Kibera etc., nothing much more than a million people.

The slums that exist today which are larger, e.g. the largest - Orangi (2mn people) is much less dense.
This is probably telling us something.

Conceptually a slum/camp is limited by how far you can walk in a day to get food, fresh water and bring it back again, or dispose of your waste.
This limits the maximum size, as observed by... Maurice.

Herodotus' Persian camp is not just unprecedented at the time, but without precedent since.

The main differences between a squatter camp and a Persian military camp is that squatter camps don't all have to be in the same place and are not supported by the machinery of an imperial economy (they're hardly supported by anything). Xerxes' army had to be in one place and was supplied by food dumps and the entire Persian fleet. If 1 000 000 people congregate of themselves into a single area at 1000 people/ha or more, what is the great difficulty in 3,4 million people congregating into an area under the command of the Great King, an area moreover that they don't occupy for more than a few days at a time?