There’s a thread about the Rape of Nanjing, and when I read about it on Wiki, I noticed something I’ve seen many times before about events of the 20th century - a huge possible range of deaths.
In the specific example of Nanjing, the estimates of deaths are between 40,000 and 300,000. How could we possibly not have a better idea of how many people died? I understand that not all the bodies could be accounted for, but there have been censuses for a very long time, so shouldn’t it be possible to know approximately how many people were in Nanjing at the time, and how many were left afterwards?
Likewise the earlier Armenian genocide in the second decade of the 20th century where estimated deaths range from 800,000 to 1,500,000. Similarly the much, much later Cambodian genocide in the 1970s could’ve killed 1,500,000 or up to 3,000,000.
And even in the 1990s we don’t know if 500,000 or 1,000,000 Rwandans were killed.
Even if it’s a matter of many people fleeing just before or after the atrocity occurred, they had to go somewhere, and many wouldn’t be in the position to go very far. It would seem noteworthy if a hundred thousand or more people suddenly showed up elsewhere.
100,000 people can scatter well enough to not be all that disruptive to the societies at their destination, especially in the pre-modern era. Keeping track of number would be especially challenging in an area where people are semi-nomadic to begin with.
The Japanese military kept secret or destroyed most of the records concerning the massacre shortly after their surrender. Even if censuses had been held, many people would have fled the city before, during, and after the massacre. The last thing the Japanese would have been interested in would have been holding a census right after the massacre.
Never mind the 20th century, it still happens in the 21st. Take the Boxing Day tsunami in 2004. Estimates vary wildly. In Thailand, there were many, many Burmese illegals working as laborers and such. They were never counted, and no one is sure exactly how many there were but certainly in the thousands
And for massacres specifically, they’re usually of people that those in power don’t care much about. You might run a census to get a count of the Right Sort of People, but why would you bother counting Those People?
In the United States, great care is taken to accurately research and account for the number of people living here every 10 years. Perhaps spending into the billions to do it. Yet somehow between 10 and 50 million illegal aliens live here, the true number of which we do not really know (or publicly do not want to admit to IMO). These are relatively stable circumstances. This would not be the case in times of turmoil.
Besides accurate record keeping of genocide only really occurred once.
Basically, the people committing the atrocities have reason not to document them, and the people being subjected to the atrocities are not in a position to gather data until long afterward.
AFAIK, places in war zones and third world countries did not do intensive censuses. The number of people in a big city or scattering of villages was always an estimate. A census of Cambodia - if there was one - was probably not done accurately since the French occupation before WWII? A census would be difficult to do while there’s a guerilla liberation war going on. Nobody knows for sure how many people lived in Rwanda before the massacres. As mentioned, a lot of people are in some places illegally and don’t want to be counted by the local government.
I have a very interesting book, Atlas of the Holocaust, that tries to guesstimate the deaths from Hitler’s holocaust and the source of the populations that died. Between the random destruction of records when possible as the Allies approached, and the general destruction of records by Allied bombing, it’s difficult even there to know the tally, although with the surviving pre-war population records they can make a more accurate estimate.
Then again, for propaganda purposes, each side will claim varying numbers depending on their agenda.
I’ve mentioned several times my cousin who registered thousands of Venezuelan citizens who had not previously been registered because they thought they didn’t have the right to register. I don’t know what’s more confusing, parsing the grammar of a sentence like that or the mindsets behind it, but the thing is, there are many places which don’t have anything near a trustworthy census.
If you don’t know how many people you had before, and you don’t know how many you have after, you can’t get anything approaching an accurate differential.
A very common reason is that in those pre-internet days the atrocity would kill them and also destroy all evidence of their existence. Like at Hiroshima and Nagasaki obviously, but also during natural disasters, genocides etc. If the public records office at Nanking got burnt to a crisp, there is not going to be too much information about lots of recently dead people.
How often do people conduct a full and accurate census in the middle of a civil war? Now imagine there’s also a foreign invasion, and the area is occupied by the foreign invaders who are still fighting. And since there’s a war, lots of people scatter as refugees, while others are worried about being found by the wrong side of the civil war. Do you really think you’re going to get an accurate count of who was in the city when there are multiple competing authorities, people with good reason to hide from the authorities, massive destruction of infrastructure, and one of the authorities attacking the general population?
And in Cambodia specifically, it’s even worse, because the genocide was specifically targeted at intellectuals, including all the folks who would be conducting censuses or estimating populations in other ways.
Besides the data problems others have mentioned, in at least the first two of those cases there’s also the issue of political motivation in raising or lowering estimates. By and large there isn’t a political debate among those trying to count the victims of Pol Pot or in Rwanda (maybe more questionable in the second case, the successor govt was ostensibly ‘national unity’ and anti-ethnic division but practically speaking represented the victim group). But there’s white hot political debate in the first two cases.
I’m not proposing a concept of ‘your truth and my truth can be different’ here, not for something like this. There is an answer in each case. It might be more complicated than it first appears. For example Nanking, how many victims were uniformed Chinese military prisoners (some of the more horrendous executions documented involved them), military men caught trying to escape in civilian clothes, military age men simply assumed to be in the second category, clear non-combatants. At the lower range of plausible estimates the first three categories might be the bulk of the victims, and arguably that might change somewhat the view of the incident, not make it OK, but might reasonably be believed relevant to a full understanding. And the answer complicated or simple might be difficult to put in a narrow range with complete objectivity. But lack of such objectivity is also a factor in some of those cases.
A related question which begs to be asked here is, what is the goal of making that accurate a tally?
I’m concerned, because I have too many times, witnessed people who wanted to IGNORE a new atrocity, or “downgrade it” politically, based on the numbers of people involved not being AS HIGH as some previous atrocity.
In particular, many modern Holocaust Deniers, base their denials on the total number of people killed as being UNDER 4 million, as opposed to the more commonly cited “over 6 million.” They want everyone to “go home, nothing to see here,” on the grounds that there MIGHT have been a slight over-count of carnage.
igor: People in the bulk are dumb when it comes to politics. It’s a game we just barely know how to play. We’re better at it than chimps are, but not much.
There are two kinds of events, no matter the type: The absolute worst there ever was, and all the other background events.
In the US there was Hurricane Katrina. Then a bunch of dumb ol’ garden variety hurricanes.
In Russia / SU there was The Great Patriotic War. Then all those other minor skirmishes.
When it comes to genocides mainstream public opinion holds that there was the Nazi Holocaust. Then all those other unfortunate incidents.
Anyone who can alter the pecking order so the Nazis don’t come first will have done a huge service in vastly and disproportionately minimizing the perception of their evil. That’s what the Holocaust minimizers are doing. Raw denialism won’t work (yet; give it a century or two) but minimizing can chip away on the Nazi’s lead in the Tournament of Greatest Historical Evil.
Second place is waay less visible than first place. It’s really as simple as that.
Quick, name me some *silver *medal winners from the recent Olympics.
Based on my research on my ancestors who were in the Civil War, two of whom were POWs, relatively good for that time period. Units had daily roll calls. When my ancestors were captured, that they were MIA was documented within a day or two. Of course there were MIAs whose fate was never determined, and there were plenty of unidentified dead, but the general scope of casualties was reasonably well documented.
A factor which enters in military ‘casualties’ is whether the speaker means, as strictly speaking they should, killed wounded and missing/prisoners or just mean those killed. Discussions of atrocities might have imprecision or debate about the cause and motive for deaths but don’t typically get confused about victims other than the dead.
Some of the apparent imprecision in counts of military deaths also arises from inclusion or not of non-battle (eg. disease) or after-battle deaths, as in ‘died of wounds’. Joshua Chamberlain the famous regimental commander of the 20th Maine at Gettysburg ‘died of his wounds’ (suffered in a later Civil War battle) in 1914.
Civil War records can be a bit hit and miss. As Colibri said, unit roll calls tended to be fairly accurate. A lot of those records got lost or damaged, though. Rolls would be collected and sent up the chain of command, and sometimes those records would be sent by mule and would be completely ruined by rain, or an enemy attack would wipe out the unit where the records were and all paperwork would be lost.
If a man was taken prisoner or killed, what happened to him often wasn’t recorded correctly, or sometimes wasn’t even recorded at all. His name might just mysteriously stop showing up on roll calls with no explanation for what happened. Sometimes men who were taken prisoner then died from their wounds a few days later, and their deaths weren’t recorded. Men who were taken prisoner or just went MIA were often listed as being killed. Battles were big and chaotic, and sometimes you just didn’t really know what happened to some of the men.
We have pretty decent enlistment records for the Union side of things. Confederate enlistment records are almost nonexistent. A lot of Confederate records were also lost in the years after the war, as things got a bit chaotic in the South due to the economic and social upheaval that was going on. Union records tend to be a bit more intact, generally speaking.
A lot of prison records on both sides were also lost after the war. Folks doing genealogy sometimes only know that their ancestors died in a particular prison because of the records of the pensions that their family received after the war, and not from prison records as those were lost.
Again, speaking generally, Union casualty estimates tend to be more accurate. Confederate casualty estimates involve a bit more guesswork due to a larger number of missing or lost records.
It’s also important to note that WWII was the first war where more soldiers died from battle than from disease. WWI was close to 50/50 with disease being the slightly higher number. In the Civil War, for every 3 soldiers killed in battle, 5 died from disease.
As already noted, some casualty estimates don’t differentiate between the various causes of death.