Selected Death Tolls for Wars, Massacres and Atrocities Before the 20th Century

Analysing statistics in Lawrence Keeley's War Before Civilization: the Myth of the Peaceful Savage (1996):

• Table 6.2 lists the Percentage of Deaths Due to Warfare. Of the 8 primitive societies that survived long enough to be analyzed by modern demographics, the median indicates that some 15.4% of all primitives, male and female alike, died by warfare. Of the 14 prehistoric cultures excavated and analyzed by archaeologists, the median indicates that about 14.8% of all prehistorics, male and female alike, died by warfare.
• Combining these into a sample group of 22 gives us a median of about 15.1%. The middle one-third of this combined sample runs from 12% to 16%. In practical terms, this means that for every 1,000,000 people who lived outside of a literate state, some 120,000 to 160,000 would eventually be killed in war. [For comparison, my calculation is that for every million people who lived in the 20th Century, some 37,000 died by violence.]
• Table 6.1 lists Annual Warfare Death Rates. The median for the 25 pre-state societies listed is 0.45%. The middle one-third runs from 0.3% to 0.6%. This indicates that if a region had population of, say, 1,000,000 typical primitives, 3,000 to 6,000 of them would be killed in war each year. That comes to about 450,000 (±150,000) per million per century, which nicely fits the 120-160,000 killed per generation in Table 6.2 if we assume some 3 or 4 generations per century.

According to Carl Haub, some 1138 million prehistoric humans were born between 50,000 and 8,000 BCE. According to the above analysis of Table 6.2, 170 million of them would have died in war.

Or instead, we could apply the analysis of Table 6.1. Assuming an average prehistoric population of 3 million, this indicates 4500 KIA each year worldwide, or 189 million for all the years between 50,000 and 8,000 BCE.

For later history, let's assume that from 8000 BCE to 1500 CE, the world's primitive population stayed about the same -- at 5 million. (After all, population growth would be confined to technologically innovative societies -- you know, civilization.) With 0.45% of them dying in war every year, this adds up to another 214 million killed in primitive war.

In total, all this indicates some 400 million people who died in primitive war before the primitives were wiped out or absorbed by civilization. Two caveats, however:

1. This does NOT displace WW2 as the worse thing that people have ever done to each other, because it's not a "thing". It's a category. The proper comparison is not "primitive war" vs. "World War II"; it's "primitve war" vs. "state-level war".
2. Don't you go tacking these 400M onto any kind of "death by government" category, because Keeley (and other anthropologists) specifically describe the societies in this category as "non-state".

In American Holocaust (1992), David Stannard estimates that some 30 to 60 million Africans died being enslaved. He claims a 50% mortality rate among new slaves while being gathered and stored in Africa, a 10% mortality among the survivors while crossing the ocean, and another 50% mortality rate in the first "seasoning" phase of slave labor. Overall, he estimates a 75-80% mortality rate in transit.

In Slavery A World History, Milton Meltzer estimates that 10 million slaves arrived in the Americas. This would be the residue after 12.5% of those shipped out from Africa died on the ocean, 4-5% died while waiting in harbor, and 33% died during the first year of seasoning.

In "The Atlantic Slave Trade and the Holocaust" (Is the Holocaust Unique, A. Greebaum, ed., 1996), Seymour Drescher estimates that 21M were enslaved, 1700-1850, of which 7M remained in slavery inside Africa. 4M died "as a direct result of enslavement". Of the 12M shipped to America, 15%, or 2M more, died in the Middle Passage and seasoning year.

Jan Rogozinski, A Brief History of the Caribbean (1994): "[A]s many as eight million Africans may have died in order to bring four million slaves to the Caribbean islands."

In The Slave Trade, Hugh Thomas estimates that 13M left African ports, and 11,328,000 arrived. Here are a few other numbers from Thomas:

• No year-by-year stats, but by piecing together scattered decade stats, I figure that 5M slaves were shipped in the 18th Century.
• Shipboard mortality among slaves:
  • Mercado in 1569 estimated an average shipboard mortality of 20%
  • Brazilian historians: 15-20% in 16th C; 10% in 19th C.
  • English trade:
    • 1680s: 24%
    • early 18th C: 10%
    • 1780s: 5.65%
  • Hugh Thomas: 9% reasonable est. for 18th C.
  • 19th C
    • Cliffe: 35%
    • House of Commons: 9.1%
    • Thomson: 9%
    • Hotham: 5%

In the chapter on African population in the Atlas of World Population History (1978), Colin McEvedy estimates that 9.5 million African slaves were imported into the Americas between 1500 and 1880. He also suggests a 15% mortality rate on the ocean.

Rummel estimates a total death toll of 17,267,000 African slaves (1451-1870)

• Among slaves going to Orient: 2,400,000 dead
• Among slaves staying in Africa: 1,200,000 dead
• Among slaves going to New World: 13,667,000 dead

Fredric Wertham claims that 150,000,000 Africans died of the slave trade.

My Estimate:

Looking at all the scholarship on the subject, it looks like, at the very least, 35% of those enslaved in Africa died before they were ever put to work in America. On the other hand, at least 20% of them survived. Between these extreme possibilites (35-80%), the most likely mortality rate is 62%.

In terms of absolute numbers, the lowest possible (and only barely possible at that) death toll we can put on the trans-Atlantic slave trade is 6 million. If we assume the absolute worst, a death toll as high as 60 million is at the very edge of possibility; however, the likeliest number of deaths would fall somewhere from 15 to 20 million.

If 5 million slaves were shipped in the 18th Century (the busiest century, see Hugh Thomas, above), then the 18th Century death toll could be around 8.1 million. (=5/11*17.8)

Keep in mind that these numbers only count the dead among the first generation of slaves brought from Africa. Subsequent generations would contribute additional premature or unnatural deaths.

Ronald Segal, in Islam's Black Slaves, estimates the total number of African slaves shipped to the Muslim world at 11.5M-14M. This breaks down as follows:

• From 650-1600 CE
  • Citing Ralph Austen:
    • Trans-Saharan: 4,820,000
    • Red Sea: 1.6M
    • East Africa: 0.8M
    • TOTAL: 7.22M shipped
  • Citing Paul Lovejoy: 3.5-10.0M shipped
• 17th Century
  • Sahara: 0.7M
  • Red Sea: 0.1M
  • East Africa: 0.1M
  • TOTAL: 900,000 shipped
• 18th C
  • Sahara: 0.7M
  • Red Sea: 0.2M
  • East Africa: 0.4M
  • TOTAL: 1,300,000 shipped
• 19th C
  • Sahara: 1.2M
  • Red Sea: 0.45M
  • East Africa: 0.442M
  • TOTAL: 2,092,000 shipped
• TOTAL: 11,512,000 shipped

Segal also mentions estimates by Raymond Mauvy:

• 7th C: 0.1M
• 8th C: 0.2M
• 9th C: 0.4M
• 10th-13th Cs: 2.0M
• 14th C: 1.0M
• 15th-19th Cs: 10.0M
• First half 20th C.: 300,000
• TOTAL: 14M shipped

What was the mortality rate among these slaves? Here are a few estimates in Segal:

• Wylde: Each eunuch in Cairo represented 200 dead Sudanese.
• Hourst, 19th C: each sale represented a loss of ten in the original population, including raids.
• Livingstone: 1 living = 10 dead.
• British Govt Rpt: For every 10 slaves reaching 19C Cairo, 50 died on the way.
• Nachtigal: on one large [typical?] Saharan caravan, 3 or 4 died for every survivor.
• UK Consul in Zanzibar: 1:1 ratio
• Mahadi: 20% d. in Saharan trade
• Lovejoy, citing Martin: 9% overall in 19th C. East Africa. (Segal: safe estimate)
• [MEDIAN of these estimates: 3 to 5 deaths for every 1 live import]

Death Toll

How many people died in all the slave harvesting by Moslems over the centuries? I hesitate to estimate, but I think we can safely assume that at least 3 people died for every 2 living slaves delivered (similar to the death rate in the Atlantic trade), which comes to about 19M deaths. Keep in mind that the data is so spotty and the margin of error so wide that we can't honestly or definitively accuse either the Christian or Moslem slave trade of being worse than the other.

• Exodus 32: 3,000 Israelites killed by Moses for worshipping the golden calf.
• Numbers 31: After killing all men, boys and married women among the Midianites, 32,000 virgins remain as booty for the Israelites. (If unmarried girls are a quarter of the population, then 96,000 people were killed.)
• Joshua:
  • Joshua 8: 12,000 men and women, all the people of Ai, killed.
  • Joshua 10: Joshua completely destroys Gibeon ("larger than Ai"), Makeddah, Libnah, Lachish, Eglon, Hebron, Debir. "He left no survivors."
  • Joshua 11: Hazor destroyed. [Paul Johnson, A History of the Jews (1987), estimates the population of Hazor at ?> 50,000]
  • TOTAL: if Ai is average, 12,000 x 9 = 108,000 killed.
• Judges 1: 10,000 Canaanites k. at Battle of Bezek. Jerusalem and Zephath destroyed.
• Judges 3: ca. 10,000 Moabites k. at Jordan River.
• Judges 8: 120,000 Midianite soldiers k. by Gideon
• Judges 20: Benjamin attacked by other tribes. 25,000 killed.
• 1 Samuel 4: 4,000 Isrealites killed at 1st Battle of Ebenezer/Aphek. 30,000 Isr. k. at 2nd battle.
• David:
  • 2 Samuel 8: 22,000 Arameans of Damascus and 18,000 Edomites killed in 2 battles.
  • 2 Samuel 10: 40,000 Aramean footsoldiers and 7,000 charioteers killed at Helam.
  • 2 Samuel 18: 20,000 Israelites under Absalom killed at Ephraim.
• 1 Kings 20: 100,000 Arameans killed by Israelites at Battle of Aphek. Another 27,000 killed by collapsing wall.
• 2 Chron 13: Judah beat Israel and inflicted 500,000 casualties.
• 2 Chron 25: Amaziah, king of Judah, k. 10,000 from Seir in battle and executed 10,000 POWs. Discharged Judean soldiers pillaged and killed 3,000.
• 2 Chron 28: Pekah, king of Israel, slew 120,000 Judeans
• TOTAL: That comes to about 1,283,000 mass killings specifically enumerated in the Bible. The battle of 2_Chron_13 is so much larger than all the others that we probably should doubt it.

[FAQ: "How reliable are these numbers?"]

• David Barrett, Todd Johnson, Justin Long
  • World Christian Encyclopedia (2001): This book is the standard reference work for religious statistics of all kinds, and both Britannica and the World Almanac cite from it. It has a single page [http://gem-werc.org/gd/gd16.pdf] estimating the number of martyrs since the origin of each religion:
    • Muslim martyrs: 80M
    • Christian martyrs: 70M
      • 20th Century: 45.4M
      • At the hands of...
        • Atheists: 31,689,000
        • Muslims 9,121,000
        • Ethnoreligionists: 7,469,000
        • Christians: 5,538,000
        • Quasi-religionists 2,712,000
        • Mahayana Buddhists: 1,651,000
        • Hindus: 676,000
        • Zoroastrians: 384,000
    • Hindu martyrs: 20M
    • Buddhist martyrs: 10M
    • Jewish: 9M
    • Ethnoreligious: 6M
    • Sikh: 2M
    • Baha'i: 1M
    • Other religious martyrs: 5M
  • The book defines martyrs as
    1. Believers in Christ (U.O.N.)
    2. Who have lost their lives
    3. Prematurely
    4. In situations of witness
    5. As a result of human hostility.
  • World Christian Trends by Barrett et al: Apparently this book gives more details about martyrdom statistics, but so far, I haven't found it in a library (even our local seminary), and I can't afford to buy it myself. (But see http://gem-werc.org/gd/listings.pdf) Here are some of the major episodes that I've been able to scrounge from websites and articles. Some of my sources have been somewhat ambiguous in their wording, so I honestly don't know whether they mean to say that the Amerindians and Baghdadians, for example, are meant as Christian martyrs or Muslim or Shamanist, or whether the various subclasses who died at Soviet or Nazi hands are already included in the largest number or meant to be added to it.
    • Martyrdom at Soviet hands:
      • 1921-50: 15M Christians in prison camps
      • 1950-80: 5M Christians in prison camps
      • Orthodox: 14.5M k. by Stalin, 2.7M of them martyrs (1929-37)
      • Roman Catholics (1925): 1.2M martyrs
    • By Mongols
      • 1214: Genghiz Khan massacres 6M Christians: 4M martyrs
      • 1214: Diocese of Herat sacked by Genghiz Khan: 1M
      • 1258: massacre in Baghdad by Hulaku Khan: 1.1M
      • 1358: Tamerlane destroys 15-million-strong Nestorians: 4M martyrs
    • Conquistadors kill 15 million Amerindians: 2M martyrs (1560)
    • Christians executed by Nazis in death camps: 1M
      • Nazis exterminate 0.5M Gypsies
    • Khmer Rouge slaughter 2M (1975)
    • Kurds massacre 20,000 Nestorians (1843)
    • Massacre of 40,000 Vietnamese Catholics (1970)
• The Ottawa Citizen (20 Dec. 1998)
  • 15M Christians d. in Soviet prison camps because of their faith.
  • citing Paul Marshall, Their Blood Cries Out
    • ca. 400,000 Chinese Christians died during the Cultural Revolution.
    • ca. 100,000 Christians k by Idi Amin (out of 300,000 total k.)
    • Up to 500,000 Rwandan Christians died as witnesses to their faith (out of 700,000 total k.) (1994)
• The Ottawa Citizen (6 Feb. 1993)
  • Citing D. Barret, Our Globe and How to Reach It
    • 40M Christians martyred throughout history.
    • ca. 24M by secular governments and atheists
    • ca. 8M k. by other Christians

How accurate are these numbers? Well, at first glance, I'm sure that they overstate the number of Christians in Central Asia before Genghis and Tamerlane, and I can't recall any event in recorded history that put 676,000 Christians at the mercy of Hindus. Nor can I find a massacre of Vietnamese Catholics in 1970. (1870, yes, but not 1970) And I'm not sure what they mean by "Quasi-religionists". And a million Bahai's? No way. But all in all, I'd say that the 20th Century numbers seem to be in the right order of magnitude (probably too high, but the right number of digits) if we accept their definition of martyr. That definition, however, can be debated.

The hard part of tallying martyrdoms is that not only do we have to figure out who, when and where, but also why. Did the Bolsheviks choose their victims because they were Christians, or dissidents, or middle class, or just in the way? According to Bruce Lincoln, Red Victory, "a commission appointed by [anti-Communist] General Denikin to look into Bolshevik atrocities indicated that more than five times as many teachers and professors, and more than seven times as many physicians, died at the hands of the Bolsheviks than did priests."

Most of these martyrologists seem to count any Christian (no matter how nominal) who died under persecution (no matter the reason). For example, many would count the Rwandan massacres as religious persecution because so many victims tried to take sanctuary in churches, even though both the victims and the killers were usually of the same religion. [see http://www.pbs.org/newshour/bb/religion/july-dec00/rwanda_8-31.html] A tighter definition obviously would yield smaller totals. Let's look at possible alternatives.

At one extreme, there is the notion that the universe is the scene of an eternal struggle between the wicked and the righteous, therefore all violent deaths among the righteous are martyrdoms. There's no arguing with this viewpoint, so let's just nod and grin and back away slowly. A calmer argument with the same result is that a person's belief system is such a central component of his character that every decision in life has religious roots, and therefore every conflict in society has a religious element to it.

At the other extreme we find the argument that, deep down, every high-minded ideology (such as religion) is really a mask for cold self-interest, and therefore no one really dies for their faith.

Between these all-or-nothing alternatives, there's plenty of room to debate just what is or isn't a martyrdom. While Barrett et al use a theological definition which focuses on the mental state of the victim ("in a situation of witness"), I would recommend a secular definition that focuses on the mental state of the killers ("for religious reasons"). After all, you can't die for your beliefs unless your beliefs are under attack.

It's probably best to balance several criteria:

1. If the only difference between the two conflicting groups is religion, then my guess is that we've got a religious conflict. Serbs and Croats are basically the same thing except for religion. Ditto for Hindi and Urdu, so whenever they fight it out, the victims could be considered martyrs.
2. If there are multiple differences between the two sides in a conflict, then it's a lot trickier determining whether religion, race, economics or ethnicity is the prime motive. Considering that the Roma (Gypsies) differed from the average European in just about everything - including religion - should we really count their annihilation as a martyrdom?
3. If the persecutors confess to religious motives, we should at least consider the possibility that they're telling the truth. When the Crusaders declared their intent to free their holy places from the unbelievers, then we might want to take them at their word rather than finagling a way to blame economic forces for it. Ditto for Communist attempts to eradicate religion in their countries.
4. If converting to the persecutors' religion does nothing to save the victims, then the motivation is probably not religious. During the Holocaust, Jews were killed regardless of whether they had converted to Christianity, making this an ethnic rather than a religious persecution. On the other hand, many of the Armenians attacked by the Turks were forcibly converted to Islam rather than killed, making this more likely a religious persecution.
5. When the victims and perpetrators are of the same religion, can we call it a martyrdom? Many martyrologists would include Martin Luther King on their lists. Their reasoning? He was a Baptist minister killed for his beliefs -- but he was killed for his political beliefs, at the hands of a Methodist -- hardly an example of religious persecution.

Finally, notice how these martyrmetrics use an interesting double standard. Atheism and secularism are counted as religions when they're the persecutors, but they aren't considered religions when they're the victims. Does this mean that no one in the history of humanity has ever been killed for being less religious than his enemies? Just to name names - Anwar Sadat, Yitzhak Rabin, Socrates, Mohandas Gandhi, Hypatia of Alexandria, Malcolm X and the occupants of the World Trade Center may count as secular martyrs. They were all murdered for religious reasons, but not because of their religion.

Several alleged martyrdoms and religious fights are widely doubted by historians.

• Book of Esther, chapter 9 (ca. 475 BCE)
  • Persian King Ahasuerus (Xerxes) allows the Jews under Mordecai to kill 500 enemies in the palace and 75,000 in the provinces.
    • Jewish Encyclopedia: "Comparatively few modern scholars of note consider the narrative of Esther to rest on an historical foundation." [http://www.jewishencyclopedia.com/view.jsp?artid=483&letter=E#2]
• Herod's Slaughter of the Innocents (ca.4 BCE, although no ancient historian mentions it.)
  • Catholic Ency. [http://www.newadvent.org/cathen/07419a.htm]
    • Greek tradition: 14,000 boys k.
    • Syrian tradition: 64,000
    • Medieval tradition: 144,000
• St. Ursula (3rd, 4th or 5th Century CE)
  • Catholic Ency. [http://www.newadvent.org/cathen/15225d.htm]: 11,000 virgins martyred in Cologne.
  • Cecil Adams doubts the likelihood of 11,000 virgins in one place. [http://www.straightdope.com/columns/020315.html]
• Battle at Cumorah (385 CE)
  • According to Mormon traditon, 230,000 Nephites were killed in this battle against the Lamanites fought in upstate New York. (Mormon 6:10-15) Also, 2 million men (plus their women and children) were killed in an earlier battle between Shiz and Coriantumr on the hill Comnor. (Ether 14:1) There is no corroborating evidence for any of this. [http://www.bcmmin.org/bomarch.html or http://nowscape.com/mormon/zindler1.htm or most especially, Mark Twain's take on this, http://www.classicreader.com/read.php/sid.1/bookid.1407/sec.17/]

• Alexander the Great (r. 336-325 BCE)
  • VD Hanson, Wars of the Ancient Greeks (1999)
    • Killed by Alexander the Great, considered the first war of anihilation in Greek/Western history.
      • "Very conservative figures suggest that in the space of just eight years Alexander the Great had slain well over 200,000 men in pitched battle alone, over 40,000 of them Greeks .... More Greeks in two engagements than had fallen in the entire history of pitched battle among city-states." [emphasis in original]
        • Granicus: 20,000 Persians and 15-18,000 Greek mercenaries in Persian service.
        • Issus: 50-100,000 Persians and 20,000 Greek mercenaries in Persian service.
        • Gaugamela: 50,000 Persians + a few thousand Greek mercs
        • 20,000 Indians at Hydaspes
      • Conservative estimate of a quarter million urban residents massacred, 334-324 BCE, incl:
        • Thebes: 6,000
        • Around Sindimana: 80,000
        • Sangala: 17,000
        • Tyre: 7-8,000 in streets + 2,000 crucified
        • Gaza: 10,000
  • Durant, Our Oriental Heritage (1935)
    • Killed fighting Alexander the Great
      • 20,000 Persians at Granicus
      • 110,000 Persians at Issus
      • 12,000 Indians in 326 BCE
  • Britannica, "Tyre": 10,000 inhabitants massacred in Tyre
  • Pitirim Sorokin: 14,750 Greeks k+w
• Graeco-Persian Wars, 499-448 BCE
  • VD Hanson: Carnage and Culture (2001): A quarter million Persian soldiers died total
    • Marathon: 6,400 Persian k.
    • Thermopylae: 10,000 Persians
    • Artemisium: A storm wrecked 200 Persian ships. "[N]early as many drowned as at Salamis."
    • Plataea: 50,000 Persians
    • Salamis: 40,000 Persians.
    • Retreat out of Greece: 100,000 Persians d.
  • Pitirim Sorokin: 57,000 Greeks k+w
• Trojan War
  • Dares (6th C. AD): 866,000 Greeks and 676,000 Trojans k. [cited (sceptically) in Hughes, Helen of Troy]
• Peloponesian War (431-404 BCE)
  • Pitirim Sorokin: 18,800 lost
  • Victor Davis Hanson, A War Like No Other:
    • Aggregate Athenian hoplite battle dead (p.146): 5,470
    • Plague of Athens (p.82): 70,000 to 80,000 residents of Attica
• Misc.
  • Pitirim Sorokin's estimates of Greek battle casualties, K+W (Only Greek armies. No Persians, Thracians, etc.(selected individual wars)):
    • Corinthian War (394-387 BCE): 34,000
    • Spartan-Theban War (379-362 BCE): 34,000
    • Holy War against the Phocians (355-341 BCE): 23,000
    • Diadochi Wars (323-251 BCE): 49,700
    • Spartan-Achaian War (227-221 BCE): 21,000
    • TOTAL: 303,460 Greeks lost on the battlefield from 500 to 146 BCE

[FAQ: "How reliable are these numbers?"]

Religious practices outlawed under Wm. Bentinck, r.1828-35

• Suttee (Sati) (from time immemorial to 1829) [make link]
  • Sakuntala Narasimhan, Sati: widow burning in India
    • 7941 widows burned alive in the Bengal Presidency, 1815-28
    • According to Rammohun Roy, almost ten times more incidents in Bengal than elsewhere
    • ANALYSIS: This indicates that there were some 8735 (=1.1 x 7941) satis in all of India in 14 years, or around 62,400 in a century. Naturally, this should be adjusted for population changes, shifting religious traditions, etc., but it does point to a century total numbered in the tens of thousands rather than the hundreds of thousands or mere thousands.
• Thuggee (Thagi) (13th C. to ca. 1838) 50 000 [make link]
  • The most common estimate is a million or so. It's even in the the title of one of the major books on the subject: James Sleeman, Thug; or, A Million Murders (London: Sampson, Low, Marston, 1926)
    • Colin Wilson, The mammoth book of true crime (p.42): over a million murders
    • 13 Oct. 2001 Toronto Star: 1M people strangled.
    • 6 Jan. 2001 Birmingham Post: Worst known serial killer in history was Indian Thug Behram, who strangled at least 931 victims in the Uttar Pradesh district of India 1790-1840
    • Michael Newton, Holy Homicide (1998)
      • British authorities est. 40,000 Thug-related homicides in 1812 alone.
      • 4,500 Thugs convicted, and 110 executed, 1830-48
    • Movie reviews for The Deceivers
      • 3 Oct. 1988 Toronto Star: 2M murders
      • 2 Sept. 1988 Bergen Record: 2M
    • Wm Saffire, "On Language", NYT 23 May 1999: ½M+
  • The most thorough study of the death toll is in Thug: the True Story of India's Murderous Cult by Mike Dash. From a 14 May 2005 Daily Telegraph (London) review: "James [Sleeman] calculated that Thugs had killed 50,000 people a year for up to 700 years, [but] Dash arrives at a more reasonable figure of perhaps 50,000 victims in total."