Africa's historical population

[Lightwell]

Whenever the topic of Africa's historical population is approached, I always see two methods for determining the population, and two camps. It's odd that the historical community seems split like this, and I wanted to see how the people on this forum would weigh in.

The first group believes Africa's population didn't exceed 50 million until at least 1500 A.D. They usually justify this by using back-projections from colonial population censuses. When the question of how the population could remain so low emerges, the answer given is that there isn't a suitable climate for mass production of calorie-dense goods.

The other group usually doesn't offer an estimate for the entire continent, but offers high enough estimates for particular regions of Africa, like the Kingdom of Benin, the Malian Empire, the Kingdom of Kongo, or the Kilwa sultanate, that the estimates can't peacefully coexist with those of the first group.

The reason this is a problem is because the second group tends to offer evidence directly from those regions, such as testimony from contemporaries, food production methods, documented village densities, and inferences from side texts. Evidence like this has placed me in the second group, and the wide-reaching explanations of the first just seem simplistic to me.

The fact remains that, not only is the second group split, but it also seems to be somewhat under-spoken when compared to the first group, as you have to do a bit of extra searching to find their answers.

When it comes to the population sizes of historical African cities, there are massive divides on how many people it could've held. Great Zimbabwe and Mbanza Kongo are estimated at having been capable of holding anywhere from 10,000 to 50,000 people, depending on who you ask.

Here's the question: In the years before 1500 A.D, do you believe Africa had a population in excess of 50,000,000 people? Yes or no? Why?

(I want to wait until others have provided their opinions before I bring in the evidence I've gathered.)

 

[elvain]

Interesting topic.

But before I answer I wonder why is it important to know whether Africa had 50 million population or not?

From my quite recent (non-professional, yet quite detailed) research of African medieval history I can only say that the method of the first group you mention is just plain wrong. We can't extrapolate early-colonial population censuses to pre-colonial periods, because demography wasn't linear, not even in long term... also with other things in mind, it puts me into the second group, if it matters.

In the Sahel for instance, there is quite satisfactory archaeological data, which proves that population of vast areas of medieval kingdom of Wagadu (AKA Ghana empire) were very populous between cca 800 and cca 1200, but the towns and villages were getting deserted from about 1100-1200 and after 1300 most of them completely ceased to exist. I don't remember the exact data, but as far as I remember, regions such as Mema or Awkar (the core of Wagadu around the capital Kumbi Saleh) had at least 5-10 cities with populations of 5.000-15.000 with the capital being twice as populous... and each of those big cities was surrounded by cluster of 3-5 smaller "service towns" with populations of 3-5 thousands and dense net of villages.

These data support information provided by medieval historians or travellers who wrote about Sahelian kingdoms such as Ghana/Wagadu, Kawkaw/Gao, Takrur or Kanem, which speak about powerful kingdoms in areas, which are until today a semi-desert or desert.
This corresponds with archaeological data from Inland Niger delta and the surroundings of Jenne-Jeno, from areas on the lower parts of the Niger (areas around Kukiya, Sirba valley etc.) in the same time period. In later centuries (13th -14th) the core of political power moved southwards and also archaeological data suggest, that situation of kingdoms such as Mali was similar in terms of population density. We also know that the Brong areas of today's Ghana had from 12th century settlements (although not as populous and dense as those described above) which later (15th century and later) declined and were shadows of their past.
And the list goes on - similar situation is in the Yorubaland (Old Oyo, Old Ifé), you already mentioned Benin, there's Igbo-Ukwu etc.

The problem of this "second group" being splitted might be caused by the lack of "newly compiled" continental histories - the ones I came across are mostly 30-50 years old, while all the researches we both described above happened during the last half-century, hence there are very few attempts to do continental-wide estimates.
Perhaps the lack of these continental-wide estimates is also because even the regional studies are often very carefull and their authors often avoid giving hard numbers for wider than researched regions, their estimates are in very wide range of estimates (i.e. the population of Wagadu can be estimated from 200.000 to 500.000 or even above million) and they emphasize that the data they provide can't be extrapolated to other regions in the neighbourhood (so we can't say that if Ghana/Wagadu had some 500K people, some of its neighbours roughly 100-200K, the other neighbours and countries with similar conditions could have this or that number of people).

Anyway, looking at the data I have studied and attempting to make a un-educated rough guess for Sahelian population, I would put it no higher than some 3-5 millions and certainly below 10 millions for the whole Sahel in the 11th century (in 9th or 10th centuries the population could have been slightly higher - wildly guessing by roughly one million, but still IMHO below 10 in total). Since this is by historians and archaeologists described as "the optimal zone", I would wildly guess that in Sub-Sahelian areas of West Africa , which I have studied much less than the Sahel, the populations in 13th century would not be higher than the data of the peaking Sahel... which gives us still roughly 5-15 millions in the entire West Africa...

If the treshold of 50 millions is of any importance, then considering West Africa as quite large and one of the more advanced regions of medieval Africa, I would wildly guess that the medieval (11th - 14th century) Sub-Saharan Africa probably would not top this number and would stay below 50M, although I wouldn't argue it could reach slightly below... so if we add North Africa and contemporary Maghreb's 2-4 millions and Egypt's 3-6 millions, the whole Africa could get above those 50 millions. But this is considering the higher numbers, I myself would prefer to be bit more conservative... and in this regard I would say that medieval Africa probably did not reach the population of 50 millions. I however have to admit I know virtually nothing about the civilizations of Kingdom of Kongo or Great Zimbabwe.... or anything south of line Igbo-Ukwu - Wadai - Darfur - Kaffa - Ajuraan.

 

[Grandpa Maur]

I was under impression that 50m figure is calculated not projecting back from colonial times, but taking what we know about extent and type of agriculture and calculating population density from it. Granted, it requires knowledge about area, but not nearly as extensive as other methods. And it circumvents the problem of what exactly the size of some city means for the population of an area.

 

[elvain]

I was under impression that 50m figure is calculated not projecting back from colonial times, but taking what we know about extent and type of agriculture and calculating population density from it. Granted, it requires knowledge about area, but not nearly as extensive as other methods. And it circumvents the problem of what exactly the size of some city means for the population of an area.

And do we know what extent of agricultural land and methods of agriculture does it take into account?

For instance it is well known fact that even if the same technologies were used, the output largely differed and that climatic conditions hugely affected agricultural productivity - which means that same regions were able to supply much larger populations even within relatively short periods of time with no change in technologies (decline of productivity in late 13th - early 14th century France or decline of productivity of the Sahel around year 1100 etc.)

 

[Grandpa Maur]

Eh, no, the approach is used because our knowledge is very limited, and it's an approximation. It's the best method, not good method. Historical climate and material culture are still more reliable than projecting back censuses, or, even worse, extrapolating from limited literary sources. But you don't seem to disagree with that, based on your long post?

 

[elvain]

Thanks.
And yes, you're right, I don't disagree. I'm just interested in learning more about origins of that number. (I still haven't received an answer why does exactly this number matter)

And when wondering about details of the figure, I'm just interested to know for instance which things were taken into account and which haven't... and how rough the approximation is.
The reason for these questions is that for instance before 1100 the area usable for agriculture in the Sahel was certainly far larger than any time later. And hence taking or not taking this into account could largely influence the final value of the approximation as you described it. In other words, how rough the approximation is.

Is there any source about how was the approximation done and how was the number 50m acheived? I guess it is some study, right?

 

[Grandpa Maur]

I don't *know* why the number matters, but i can guess - if the number is high in 1500, then we it kind of implies the slave trade was worse, demographically, than if it's low. The reasoning being is that if the population didn't grow much (compared to other continents, for example), we could perhaps infer that it's because of severe impact of slavery. And vice versa.

That's my guess. Of course, the reasoning is kind of... not exactly bulletproof. But these are political implications and they don't have to be so strict.

Same as with Americas pre-1492 population. Which is even worse case, tbh, with estimates varying so wildly... (in Africa we at least don't have to contend with societal collapse, guesswork about diseases, eradication of natives and lack of even oral history)

I can't answer your questions exhaustively. But this is not a study per se - it's higher level work, analysis of work of more specialized academics by more general ones. But it's often very crude because there is, or at least wasn't, much legwork done. So, when estimating pre, say, 1800 population of central Africa it's basically guesswork on some Portuguese and Kongolese, Luango etc archeology or accounts or oral history to draw conclusions about the level of economy. And when it comes to previous times it's even worse and what's left is "okay, so Bantu migrated here in 1st millenium and brought agriculture and metalworking, displaced hunting-gathering pygmes, so let's assume average pop density rose from 0.1 per sq km to 1 per sq km during that millenium based on what we know of similar events elsewhere. And then let's draw a straight line between the two points. That's afaict how the number was achieved, and no, it's not pretty

Oh, and towns/cities are very unreliable indicator of population in these times, sadly. So are literary sources of people impressed by amount of population dying in wars, famines, or impressed by great cities. For example, most Europeans were concerned about population decline and decline and depopulation well into XVIII century, when Europe population growth has greatly accelerated. This only changed in second half of the century when it became known the situation is very different. But that's already a period of strenghtening of state apparatus.

 

[Geriander]

IOh, and towns/cities are very unreliable indicator of population in these times, .

Indeed, if that method was used for western Europe you could conclude that the population dropped by more than 90% after the fall of the Western Roman Empire.

 

[elvain]

Oh, and towns/cities are very unreliable indicator of population in these times, sadly. So are literary sources of people impressed by amount of population dying in wars, famines, or impressed by great cities. For example, most Europeans were concerned about population decline and decline and depopulation well into XVIII century, when Europe population growth has greatly accelerated. This only changed in second half of the century when it became known the situation is very different. But that's already a period of strenghtening of state apparatus.

Indeed, if that method was used for western Europe you could conclude that the population dropped by more than 90% after the fall of the Western Roman Empire.

You're right, especially if the data about cities would be used as sole source.
Which in the context I mentioned it, isn't the case, of course.

We can't take the disappearance of cities alone as a proof of population decline. I meant it the other way around. The existence of towns as specialized centers of crafts (as proved by archaeology), which means not "agriculturalist" towns, with proofs of a net of dense agricultural settlements supporting them is however an indication of relatively high population density and specialization. But in areas, which are mainly desert or very sparsely populated for at least the latest 500+ years.

I probably wasn't very clear in my previous posts, but the studies which described the downfall of those urban populations also presented numerous other data, which support depopulation. One (but not the only) is also gradual decline and disappearance of vast majority of the agricultural settlements in the same area. Together with climatologic data indicating that the areas we talk about were the worst hit by very remarkable aridization of the area this all is indeed a clear indication of population decline.

So while judging only from disappearance of urban settlements in post-Roman Europe one could wrongly deduce a population decline by some 90%. But looking at other available data, one would realize that while there was some decline, the biggest difference was a change of settlement structure, with increased number of different forms of settlements.
OTOH on the edges of the Sahel we can see that it was not just cities, but also all other minor settlements either disappearing or largely declining... which suggest a population decline far bigger than what we can read from sources about post-Roman Europe.
And climatic data give us clear explanation of what could have caused it:

 

[Grandpa Maur]

Thank you for your response. I pondered a bit how to reply, but since there is nothing i disagree with all i could come up is thanking you for nice conversation

Hm. Okay, maybe about the cities. I agree - but there is another caveat, in other direction. There are cases where there are no specialized cities but the population is relatively dense. It still should show in archeological record as some kind of dense rural settlement/manorial economy, so it's small caveat.

I am completely ignorant about the specific Sahel situation you talk about, so it's quite interesting for me. In general arid environments seem to be much more prone to changes in climate, so it fits the pattern. On a side note, are you aware of factors that influenced the climate changes? To be precise, i don't mean underlying causes like sun activity but more how did that happen - like, pressure high changed position and thus rainfall patterns changed or something like that.