One question that arises more than almost any other in truly nerdy discussions about A Song of Ice and Fire relates to the population of the Seven Kingdoms and how many people live in each region, in the lands beyond the Wall and in the cities. That’s quite a big discussion and there’s been lots of conclusions drawn over the years.
How to determine the size of the population? Elio Garcia of westeros.org has an excellent video which outlines several methods of doing so. The first involves extrapolating from the military figures we are given. These figures vary due to George R.R. Martin having characters using guesswork, estimates and sometimes (as in the case of Dorne) misinformation in their figures. But certain trends have emerged throughout the books that have allowed some fairly accurate military figures to be gained and thus a civilian population to be extrapolated. Another method involves mapping the square mileage of Westeros (in a similar manner to what we have already done) and then comparing that to the population of real-life corresponding areas (Scandinavia for the North, France for the Reach, Spain for Dorne etc) in the medieval period, so see what results you’d get, adjusting for factors like the latitude and the long winters.
For the military technique, we need to know the armies of the Seven Kingdoms. Using the books as a base, Garcia extrapolated as follows:
- 15,000 for the Crownlands.
- 35-40,000 for the North.
- 35-40,000 for the Riverlands.
- 35-40,000 for the Vale of Arryn.
- 50-55,000 for the Westerlands.
- 15,000 for the Iron Islands.
- 20,000-25,000 for the Stormlands.
- 25-30,000 for Dorne.
- 120,000 for the Reach.
This totals 380,000 troops. Given unknown variables (like Dorne’s unknown true strength and the extra potential forces available on Skagos or sellswords), this can be raised to 400,000 for the entirety of the Seven Kingdoms.
Various percentages have been given for the ability of a medieval population to support a large force under arms for a prolonged period of time. The figure varies from around 0.5% of the population up to almost 2%, although that seems rather high (10%, the figure reached in WWII, was only possible through mass industrialisation and mechanisation on a colossal scale so the figure is certainly far, far less than that). A commonly-cited figure is 1% which is what Garcia goes with, and seems reasonable.
Using a combination of these techniques, Garcia estimates a population of 40 million people, give or take 5 million. For comparison’s sake, the population of Europe was about 25 million in 800 AD, 56 million in 1000 and 78 million in 1300. This means that Westeros was considerably more lightly populated than Europe of the corresponding time period, possibly as low as half the population density. Given the harshness of the winters, this may not be too surprising.
Using the same technique we can work out the approximate populations of the constituent regions of the Seven Kingdoms.
- 1.5 million for the Crownlands.
- 4 million for the North.
- 4 million for the Riverlands.
- 4 million for the Vale of Arryn.
- 5.5 million for the Westerlands.
- 1.5 million for the Iron Islands.
- 2.5 million for the Stormlands
- 3 million for Dorne
- 12 million for the Reach.
If the population of the Reach seems disproportionately huge, it’s worth remembering that in 1300 the population of France was approximately 17 million compared to 3 million for England, which makes it kind of incredible that England was able to maintain any possessions at all on the continent given the overwhelming French numerical superiority.
The next thing to look at is urbanisation. The Seven Kingdoms, at first glance, look extremely under-urbanised. There are only five cities in a region the size of Europe. However, I would argue that this is down more to the Westerosi custom of only counting “cities” as extremely large settlements, whilst in the real medieval period a city could have only a few thousand inhabitants (in fact my home town, with a population of just under 200,000, is not counted as a city by UK standards even though it would be in many countries). This can be seen in the Free Cities, where the Rhoyne river towns Volon Therys, Selhorys and Valysar are not counted as cities despite being larger than any settlement in the Seven Kingdoms (including, allegedly, King’s Landing and Oldtown).
The “cities” of Westeros are counted as follows:
- King’s Landing (c. 400,000)
- Oldtown (c. 350,000)
- Lannisport (c. 200,000)
- Gulltown (c. 50,000)
- White Harbor (c. 30,000)
The population figures are worked out as follows: King’s Landing has over 500,000 people in and around it in A Storm of Swords when Oberyn Martell arrives and meets Tyrion Lannister, but this includes parts of the Lannister and Tyrell armies and a flood of refugees from the Riverlands. Ordinarily I’d say that there’s no way that Tyrion could know how many transients were in the area, but Littlefinger had introduced a tax on refugees for this exact purpose. As a result, King’s Landing’s population is swollen significantly above its norms. In addition, the city-as-mapped appears too small to support 500,000 people inside it, especially as unlike in many real cities there was no major urbanisation outside the walls. To be frank, the city is probably too small for 400,000 people to live comfortably either (at least not without the place catching fire every five minutes) but we can go with that figure. Oldtown is then cited by GRRM as being slightly smaller, a “big drop” to Lannisport and then a further big drop to Gulltown and White Harbor, with White Harbor being the smallest city.
These figures may seem okay, but it’s worth noting that in 1328 Paris had a population of around 250,000 and London about 80,000. So, slightly oddly, we have a situation where the Seven Kingdoms have a much lower population than the real medieval Europe but rather larger big cities.
The low urbanisation issue can be taken care of by counting the smaller towns, however. Martin has been much cagier about the number and populations of the walled towns of Westeros and has in fact been happy to introduce apparently quite big towns (Stony Sept, Weeping Town, Planky Town) out of nowhere when there had been no clue as to their existence in prior books. There seems to be quite a fair number of these “big towns” in the books, including: Duskendale, Barrowton, Lordsport, Sisterton, Saltpans, Lord Harroway’s Town, Maidenpool, Stony Sept, Kayce, Tumbleton, Hull (on Driftmark), the Weeping Town, the Planky Town, Starfish Harbor and Vintown. The populations of some of these towns, particularly Duskendale and Stony Sept, seems to comfortably be in the thousands.
This isn’t even addressing the issue of “castle towns”, that is the fact that some of the castles in the books seem to have quite large towns surrounding them. There’s the winter town of Winterfell, Ashford, Seagard, Sunspear and Rosby as the very least. I wouldn’t be surprised if the overwhelming majority of the castles of the Reach had reasonably-sized towns attached to them, and a reasonable number in the Westerlands, Riverlands, Crownlands and Vale either.
So there we have it. The population of Westeros would appear to be in the ballpark of 40 million people, maybe slightly less. Working out the population of Essos is altogether more difficult, but is something we may attempt further down the road.
Brett said:
1.5 million seems high for the Iron Islands, since they’re described as having poor soil for farming and heavily dependent on fishing. Given that they lean heavily on fishing and traditionally supported themselves with raiding as well, the 15,000 might represent a disproportionately high percentage of their free men (although that could be countered by the need for them to maintain forces back home to keep the thralls and salt wives under wraps).
LikeLiked by 1 person
SeanF said:
Interesting as always. I’d probably give the Crownlands a higher population, as I’d expect it to be densely populated.
LikeLike
Martin said:
Riverlands seems underestimated. This land is the second most fertile soil of seven kingdoms.
But, in general I agree with your numbers.
If total westeros population is -40 millions,
Reach> riverlands `> werterlands> North >vale > Dorne> stormlands> crowmlands> Iron Islands
11 Reach
6 Riverlands
5,5 westernlands
4,5 North
4,3 vale
3,5 dorne
3,3 Stormlands
1,7 crowlands
0,7 Iron islands
According to 2007 map release, Dorne seems to have 2/3 fertile lands. (not very)
LikeLiked by 1 person
rattaman said:
All strong men in Iron Islands are warriors; they don’t farm..
So the population of Iron Islands if the warriors are 15.000 is <100.000
LikeLike
werthead said:
That’s the perception of the Iron Islands, but it’s not remotely in touch with the reality of how population bases worked in the real medieval period. To support thousands of fighting men, you needed hundreds of thousands of non-fighting men, women, children etc working the land, fishing, defending the home and working in other areas.
George’s worldbuilding for the ironborn (and also the Dothraki) is not tremendously convincing on this front, it has to be said.
LikeLiked by 1 person
Pingback: #TPTWP Episode 13.1- Show Notes… – The Westerosi Companion
Simon said:
Since Westeros is more centralized than Europe during the Middle Ages, isn’t it logical that the cities should be larger?
LikeLike
Anna said:
I always thought Westeros was about the size of the United Kingdom, which in the Middle Ages had a population of 4-6 million, which is a tenth of the amount estimated above.
If it is however the size of Europe, then the number is too small as Europe had a population of 70-90 million at the time, which is more than double the estimated amount.
LikeLike
werthead said:
Yes, that is very much the dichotomy involved in the worldbuilding in the series.
LikeLike
Gerard P said:
GRRM has commented several times that Westeros is the size of South America.
LikeLike
bellerophon said:
Interesting. Oldtown supposedly lost a full half of its population to Grey Plague less than a single lifetime ago, I wonder how large it was at its maximum. It must have only been by that point that King’s Landing overtook it.
LikeLike
werthead said:
It should be noted that spamming screechy, incoherent and ill-informed comments in ALL CAPS that do nothing but betray an utter lack of even the most basic knowledge about medieval history is not a recommended way of getting your comment published.
LikeLike
JoséMST said:
I just want to ask, why were the Stormlands so low in population? Isn’t Dorne the least populated of the Seven Kingdoms? I would personally put Dorne at 2.5-3 million and the Stormlands with a million or a million and a half more
LikeLike
aldariontelcontar said:
There are actually several approaches which can be made to estimate population. I have made it here:
https://militaryfantasy.home.blog/2020/04/21/population-of-westeros/
But essentially, population of Westeros can be anywhere between 10 million and 100 million. But most realistic estimate would likely be 40 – 60 million.
LikeLike
Robert Smith said:
The calculation of 1.5 million in population for the Iron Islands is not supported by the population density of Westeros as a whole. Another chapter reviews the population of each kingdom and provides the following: The Reach has 12 million in roughly 480,000 sq. miles; The Westerlands has 5.5million in 192,720 sq mi; The Crownlands has 1.5 million in 130 sq mi; The Iron Islands are projected at 1.5 million in 11,136 sq mi. Population density based on this is: The Reach, 25 per sq mi; The Westerlands, 28.5 per sq mi; The Crownlands, 11.5 per sq mi; The Iron Islands, 135 per sq mi.
The Iron Islands have few resources, other than the sea, by which it can feed its population. Yet the writer is postulating a population density of over five times that of The Reach and The Westerlands, two of the richest of the Kingdoms. This makes no sense. The 1% analysis is flawed in considering that in The Iron Islands their naval forces were not only used as a military force, but were also an economic force through piracy and raiding. It is likely that its population was closer to 300,000, certainly no greater than 500,000.
LikeLike
josmst said:
Considering as an example the population of England during the Middle Ages, the population of the Iron Island its not that flawed, estimating the population in 5 million in a land mass of 50345 gave us a density population of 99, lets say 100 people per square mile, you still argues the Iron Islands have to much people?
LikeLike
aspectemperor said:
Werthead estimates the size of the Iron Islands at 11,136 square miles, not 50,345. And with a population of 1.5 million as estimated by Werthead, not 5 million, we get a very unrealistic measure of population density of 135 per square mile. That is roughly equivalent to modern-day Nicaragua.
Given that the Iron Islands not one big island, and that the area of 11,136 sqmi is broken up into 7 main islands and 24 other smaller ones, and that the islands are rocky with infertile soil, AND that the islands lack any sizeable populations of working animals (horses, oxen, cows(which also means that they don’t have access to foodstuffs from those animals)), a population density of 136 people per sqmi is absurdly impossible. Even your estimate of 100 people per sqmi is extremely unrealistic for a medieval society. For example, England alone had a density of at most 25 people per sqmi in 1300.
A real-life comparison can be drawn between the Iron Islands and Iceland. Icleand, like the Iron Isles, has poor soil unsuitable for agriculture (although they do have sigificant numbers of livestock animals), and for many years relied on fishing to provide subsistence. The population of Iceland in 1100 CE was 60,000 and the area 40,000 sqmi. This gives a population density of just 1.5 people per sqmi. A far cry from 100.
Another useful comparison is Estonia. I arrived at this nation by observing the latitude of the Iron Isles (about 42-43) and I found that this correlates with the Japanese island of Hokkaido. It is fair to assume that the Iron Islands and Hokkaido have similar climates. I could not find any population estimates online for medieval Hokkaido, so I chose a country with the same Köppen climate classification (Dfb) as Hokkaido that I was able to find a medieval population estimate for. Estonia had a maximum population of 200,000 in 1200, within an area of 17,000 sqmi. This gives a population density of ~12 people per sqmi.
Do you understand now why the estimated population of the Iron Islands is much, much too high? The measurement based on Werthead’s finding is literally impossible for a small, infertile, medieval archipelago. This is not an dig at them, but rather at GRRM for creating a very unrealistic world, at least when it comes to population demographics.
LikeLike
Sean Morris said:
I think Gulltown population should be higher. You put the population of the Vale at 4 million. We know that the population of the Vale is concentrated in one general area in the vale proper. If Lannisport has 200k with a population less concentrated in one area, wouldn’t Gulltown be closer to 100k? Like 80k-120k. I am only saying this because since all of the 4m are mostly living in the same area you would think the only major port city would be packed with people.
LikeLike
conor said:
the Iron Islands can’t be both tiny and thinly peopled or they would have no people at all given the times they lost battles, so either they do have a million plus people and the Iron Isles are bigger than they look on the map (because they hate maesters for example) or they don’t have hundreds of ships and an army/navy of 15,000
Because the iron islands being more populous than the Reach is highly unlikely and if it were true you’d expect someone like Ned to have mentioned them having a million more people than expected.
LikeLike
thr2 said:
I think that using military sizes is misleading because they depended more on the finances of the state than they did the population or even the economy. Key to note, many armies were supplemented by resources. The wealth of the region would obviously play a role, but so would the state structure in each region( How much power the lord has, how loyal their subsidiaries are[certainly a gamble in all universes,especially in a civil war like this] and how much of a treasury they’ve accumulated). Medieval armies were not tied as much by population: Army sizes increased by a significant amount even as europe suffered the bubonic plague.
LikeLike
Daniel said:
Medieval armies weren’t tied to population because they lacked the ability to mobilize anything like as many people as their population would allow, there are exceptions such as Poland due to a combination of wealth and desparation or just being amazingly combative
on the flipside the Iron islands have to be over mobilizing to get even as any as they have raised.
LikeLike