Readings on population and army size?

[Marja]

Also, Scheidel alludes to Ottoman census figures and occasionally other 19th-century census figures. Can anyone recommend good sources for these for comparison?

for clarification, I'm looking for 19th-century census data for the Mediterranean, to compare with estimated 4th-to-4th century populations in the same area.

[Eleatic Guest]

Ottoman censuses are used for the carrying capacity approach. There are two paragraphs about this method in this review of Hansen's Shotgun Method (begin from "The fourth method H. examines involves applying the concept of the "carrying capacity"..."): http://bmcr.brynmawr.edu/2007/2007-04-58.html

[Sean Manning]

Also, Scheidel alludes to Ottoman census figures and occasionally other 19th-century census figures. Can anyone recommend good sources for these for comparison?

The heuristic "ancient populations can have been no larger, healthier, or better fed than 19th century ones" is pretty much disproved I think. Mogens Hansen does a good job in his book, but see also Geoffrey Kron's article on heights or Ian Morris' work on house sizes and consumption. This is one of the traditional buttresses of the "low count" which has been undermined recently.

[Marja]

But that's not why I'm looking for 19th century census data. I'm looking for them to understand Scheidel's allusions to them, and because the 1st century and 4th century estimates I have [Beloch and Russell are the only ones with region by region overviews] are all over them place regarding both the absolute and relative populations of Aegyptus, Syria, parts of Asia minor, etc. 19th century estimates don't have to be upper limits to be reasonable clues to the relative populations.

[Sean Manning]

That is a good point Marja; my apologies. I have not been feeling level-headed lately and should hold my tongue. I think that Beloch refers to some censi in Die Bevölkerung der greicische-romische Welt, and Hansen's overview of the history of the “carrying capacity” approach should have footnotes to detailed studies. Unfortunately for your purposes, Beloch focused on the Mediterranean and skims over points north and east, so I'm not sure if he has much on the Germans.

Colin McEvedy and Richard Jones, Atlas of World Population History (Penguin, 1978) has some handy figures from early censi but I would ignore their estimates of ancient populations. They have curious conclusions such as that modern Greece had three times the population of modern Iraq in 400 BCE ... very frustrating when you are trying to estimate the population of the Achaemenid empire.

[Eleatic Guest]

Colin McEvedy and Richard Jones, Atlas of World Population History (Penguin, 1978) has some handy figures from early censi but I would ignore their estimates of ancient populations. They have curious conclusions...

Theirs is the lowest count I know of, 44 million for the entire Roman Empire, but since they are also about the only ones who provide a breakdown by province, their figures are still used. Recently, Maddison has used them for calculating comparative wealth, with surprising results: Income and population distribution: Europe, Asia, Africa

Maddison's Contours of World Economy, 2007, pp. 32ff. offers a fairly detailed, yet concise discussion of Roman demography related to economic performance.

[Sean Manning]

Colin McEvedy and Richard Jones, Atlas of World Population History (Penguin, 1978) has some handy figures from early censi but I would ignore their estimates of ancient populations. They have curious conclusions...

Theirs is the lowest count I know of, 44 million for the entire Roman Empire, but since they are also about the only ones who provide a breakdown by province, their figures are still used. Recently, Maddison has used them for calculating comparative wealth, with surprising results: Income and population distribution: Europe, Asia, Africa

Maddison's Contours of World Economy, 2007, pp. 32ff. offers a fairly detailed, yet concise discussion of Roman demography related to economic performance.

Yes, one reason that makes this so difficult is that estimates of population are tied into the question whether the Roman empire was an advanced preindustrial economy like Song China or 18th century England, or poor but showy due to greedy elites and a habit of working in stone and pottery. That is a hard debate too, but I note that Scheidel accepted Ian Morris' conclusions about the Iron Age economic boom in Greece in the Cambridge Economic History of the Greco-Roman World. But Morris' conclusions about the rate of growth depend on the current chronology of the Early Iron Age being right, and it is rather shaky at present. So its all about interlocking systems of assumptions not all of which people spell out ...

I can't give much credence to any study which works on the basis of ancient figures in McEvedy and Jones. Their estimate for the Greek population of Anatolia in 400 BCE is ten times smaller than Mogens Hansen suggests, and its not based on much more than Pompey's boasts and the implicit assumption that population should rise exponentially from prehistory to 1800 ...

[Marja]

A lot to think about.

Would you consider the Hansen book a must-read? I picked up the Bowman and Wilson book [Quantifying the Roman economy] and the Manning and Morris book [The Ancient Economy: Evidence and Models] some time back, but unfortunately, don't have the econ background to follow most of the arguments in either volume. It's made me skittish about ordering either economics or demographics books.

I generally tend to divide the Roman world into several agricultural zones:

1. Egypt, shaped by the Nile River

2. Mesopotamia

3. The Mediterranean Coasts, including peninsular Italia and Achaea

4. The Anatolian Highlands

5. The Atlantic Fringe

6. Central Europe, including cisalpine Gallia

7. The Steppes

I know the focus on agriculture is most often associated with the carrying-capacity approach. I'm not sure about that approach, but I'm sure it makes sense to try to keep population estimates within each zone broadly proportionate to each other, while population estimates between zones can be disproportionate to each other,

In Egypt, the population went from about 2½ million to 10 million in the 19th century. I get the first impression that other factors overwhelm any 'carrying capacity,' inherent stability or inherent growth trends there. I don't see how to decide between a low count of about 4 million and a high count of about 8 million, or dramatic population changes within the period. But are there any reasons a low count in Egypt should contradict a high count in Italia or vice-versa?

In the Steppes, there seems to be as much instability, probably due to the shifting balance between agriculture and pastoralism. This sometimes results in outward migration, culture collapse, etc.

[D B Campbell]

A lot to think about.

Is this the subject of your thesis, Marja?

I confess that this subject doesn't particularly interest me (yet); I looked at modern attempts to gauge Classical Spartan population, and realized that they were all simply guessing.

[Marja]

Oh, no, this is for a game!

I was getting tired of implausibly large 'barbarian' armies in ancients games, such as Caesar: Conquest of Gaul, and want to find better ways of estimating the size and composition of 'barbarian' armies and other poorly-documented armies.

[Sean Manning]

A lot to think about.

Would you consider the Hansen book a must-read?

The chapter where he talks about methods and historiography is very good (and much more measured than my youthful grumbles). However, his focus is estimating the total population of every Classical Greek polis and its hinterland based on the areas of walled towns. He divides his figures into six regions, but it wasn't even easy to get his figure for the Asian Greeks let alone other Anatolians. So I am not sure how helpful it would be for estimating the populations of the northern barbarians.

In Egypt, the population went from about 2½ million to 10 million in the 19th century. I get the first impression that other factors overwhelm any 'carrying capacity,' inherent stability or inherent growth trends there. I don't see how to decide between a low count of about 4 million and a high count of about 8 million, or dramatic population changes within the period. But are there any reasons a low count in Egypt should contradict a high count in Italia or vice-versa?

The only problem I can see is where a high count in one area and a low count in another give implausible ratios. I would be surprised to see someone who guessed 4 million in Egypt and 15 million in Italy under Augustus, because that would imply that Egypt lacked the population to be a significant center of power.

[Marja]

More on the army size side of things, I'm also interested in the levies the Roman Empire demanded from 'barbarian' peoples, especially in the later period.

So for instance, Ammianus book 26 chapter 10 claims that Procopius, under treaty, called for 3,000 Gothic soldiers. In book 31 chapter 4 he claims that Valens hoped to recruit soldiers from the refugees. In book 31 chapter 6 he may be referring to two such units, or they may have some other origin. Orosius book 7 claims that 10,000 Goths died at the Frigidus [doubtless an exaggeration]. Olympiodorus claims that 12,000 of Radagaisus' leading troops joined the Roman army.

I'm looking for some comparative list...?

[Marja]

As an addendum to #32117, I probably should have separate zones for the Atlas Highlands and central Hispania, and some way to account for the less densely populated Mediterranean islands.

Beloch estimates that there were about 6,000,0000 people in Syria (55 per km2), 6,000,000 people (44 per km2) in provincial Asia, and 500,000 people (33 per km2) in Cyrenaica. So this low-count estimate would actually fit right in with certain high-count estimates for other areas in the Mediterranean zone. Suppose 4,600,000 people in peninsular Greece (40 per km2 x 115,000 km2) or 6,400,000 people in peninsular Italia excluding Gallia Cisalpina (40 per km2 x 160,000 km2)... [strike]Certainly Beloch's low count for Greece seems way too low relative to the rest of the Mediterranean.[/strike] [Edit: Beloch counted a much larger portion of the Balkans, including most of Thrace, as part of peninsular Greece, so that his estimate was much less dense than his estimates for Asia or Italia.] I hope Hansen's work sorts out better figures for Greece, and maybe his count for Greece can shed light on Asia, Syria, and Italia.

Given how long the Punic Wars lasted, especially the 1st, I expect there would have been similar populations in the core areas on each side. So if 6,400,000 in Italia is about right [probably somewhat lower in the 3rd Century BCE] 5,000,000 to 7,000,000 in Africa makes sense, without needing to figure out the areas and densities for the coast and plains vs. the Atlas Highlands.

[Marja]

I'm still waiting for the Hansen book to arrive. I hope it sheds light on the population of Greece and Peninsular Italy. But that still leaves a lot of uncertainty surrounding the population of the European interior. I figure Cisalpine Gallia probably had a population density no greater than Hispania in the 3rd century B.C.E. but no less than Peninsular Italy by the 6th century C.E.

But that leaves the rest of the Central European Zone uncertain. If we extrapolate similar population densities throughout the Mediterranean Zone [excepting the highlands on its periphery] [implying that Greece and Peninsular Italy had the same densities as Syria and western Asia Minor, 40-55/km2] and extrapolate similar population densities throughout the Central European Zone [excluding the Alps, the densest forests, and the densest marshes], we get extremely high population totals for the last.

[Marja]

Soil maps are very inconsistent. Some show forest soils all around the Po valley, similar to central Europe and unlike the Mediterranean, while others don't. Some show forest soils in northwestern Iberia, while others don't. Soil maps could be useful for identifying different agricultural zones with different needs, such as heavier plows in heavier soils, but under the circumstances they aren't as useful as they could be.

Soil erosion and/or depletion is controversial but might contribute to an absolute decline in population in the Mediterranean in late antiquity. I can't see any reason why the western Mediterranean would be hit harder than the eastern Mediterranean though. Soil erosion and/or depletion is still compatible with the increase in the cultivated area in parts of the east in late antiquity.

I need a better understanding of the medieval population boom to understand the ancient population changes in central Europe, including cisalpine and transalpine Gallia.