War Before Civilization
Page 26
Whatever their personal biases and favored theories, archaeologists basically and ultimately want to know what happened in the past. The physical circumstantial evidence already available repeatedly attests that what transpired before the evolution of civilized states was often unpleasantly bellicose. It also demonstrates that, as with the Native American accounts of the Battle of the Little Bighorn, we cannot summarily dismiss the ethnographic reports that give the same message. As Thoreau said, when he suspected his milkman of watering the milk, “Some circumstantial evidence is very strong, as when you find a trout in the milk.” This book has been an extended exercise in finding the trout in the milk.
APPENDIX
Tables
Table 2.1 Political Organization Versus Frequency of Wars
Table 2.2 Subsistence Economy Versus Frequency of Warfare
Table 2.3 Political Integration Versus Frequency of Warfare
Table 2.4 Frequency of Offensive Raids and of Defense Against Raids Among Western Indians
Table 2.5 Frequency and Duration of Warfare by Nation-States, 1800–1945
Table 2.6 Combat Unit Sizes and Social Unit Populations
Table 3.1 Association Between Weapons and Armor
Table 3.2 Prestate Fortifications
Table 4.1 Casualties from Formal Battles
Table 6.1 Annual Warfare Death Rates
Table 6.2 Percentage of Deaths Due to Warfare
Table 7.1 Territorial Gains and Losses from Warfare per Generation
Table 7.2 Population Density and Width of Buffer Zones
Table 8.1 Motives and Causes of Wars in Nonstate Societies
Table 8.2 Causes of Warfare in New Guinea
Table 8.3 Population Density versus Frequency of Warfare
Table 9.1 Distribution of Arrow Wounds at febel Sahaba
NOTES
Chapter 1
In order not to clutter the text with footnotes, the references for each paragraph have been consolidated into the footnotes attached to the first or the final sentence of each paragraph.
1. See Divale 1973: 3–9; Ferguson 1988: 114–21.
2. This original spelling is used by several anthropologists as a shorthand reference to Hobbes’s vision of small-scale societies and to characterize some ethnographic situations in which violence of all kinds was extremely common.
3. This is, of course, a libel, since Hobbes “concluded” no such thing. It is interesting that the neo-Rousseauian, Brian Ferguson, repeated this misrepresentation in 1990 but neglected to acknowledge Rousseau’s precedence or even to mention his existence!
4. Ryan 1981: 49–57.
5. Sumner 1911 versus Malinowski 1941.
6. Divale 1973: xvii.
7. Herdt 1987: 47–48.
8. Keegan 1976: 36–46.
9. Divale 1973: xxii.
10. For example, the anthropology graduate student whose master’s thesis was part of the project, Harry Hoijer, later co-authored the most widely used anthropology textbook of the 1950s and 1960s (Beals and Hoijer 1965). Thus anthropologists did not need to consult Wright’s massive book to be influenced by it.
11. Wright 1942 [1964]: 7.
12. I can find nothing in his writings or in anything written about him to indicate that he ever experienced combat. He certainly would have seen much of its ugly aftermath in liberated Belgium. In any case, Primitive War was written before World War II, which was his only chance to see combat.
13. Wright 1942: 62, 69, 74–76; Turney-High 1949: 141–68; 1981: 26, 36–40. Regarding the sportive or entertainment motive among primitives, Wright offers no documentation for his statements. Turney-High’s arguments and examples on this point are rather strange: war stories are “the most entertaining stories, and in order to spin yarns there must be wars”; California Indians knew they were “athletic humbugs” but would not admit it (!); and so on. No one reading his works can doubt that Turney-High thought war was fun—an easier attitude for a rear-echelon M.P. to maintain than for a front-line “grunt.”
14. Wright 1942: 80–85; Turney-High 1949: 21–137.
15. Turney-High 1949: 85, 87.
16. Turney-High 1981: 34.
17. Turney-High 1981: 34.
18. Various places in Turney-High 1949: 25–137; summary in 1981: 35–44, 56, 58.
19. Keegan 1976: 22–23.
20. Turney-High 1981: 69.
21. Wright 1942: 85–88; Turney-High 1981: 38.
22. Wright 1942: Appendix XII, 569–70. In this appendix, Wright listed annual war death rates for four tribal societies, three of which were from three to ten times higher than ninteenth-century France’s war death rate (the highest civilized rate known to him in 1941).
23. Wright 1942: 242–48; 1964: 59–62. Ennumeration indicated that civilized battles have been becoming less deadly over the last four centuries. Thus to save his hypothesis, Wright had to include all deaths “indirectly related” to war, dismiss the figures from seventeenth-century Britain and Germany, and include some highly estimated “indices” created by sociologists.
24. Turney-High 1949: xiv-xv, 25.
25. Ferguson 1984a: 6.
26. For example, Harris 1979.
27. For example, Harris 1975; Ferguson 1984a. However, it was never claimed that casualties alone could be a method of population control.
28. For example, Harris 1984: 129; Ferguson 1990: 29.
29. Chagnon 1983 (first edition 1968).
30. For example, Chagnon 1983; Koch 1974; Hallpike 1973, 1977.
31. Hallpike 1973: 454. Another neo-Hobbesian, K.-F. Koch (1974: 159–75), accepts four of five possible explanations for warfare in highland New Guinea—every possibility except the economic one.
32. For example, Fagan 1989; Wenke 1988; Sharer and Ashmore 1987; Thomas 1988. A recent exception is Hayden 1993.
33. For example, Green and Perlman 1985; Rouse 1986; Gregg 1988; Bogucki 1988.
34. Fagan 1989: 311; Whittle 1985: 219–20. Whittle does mention that at least one camp appeared to have been attacked by archers.
35. Dixon 1988; Mercer 1988.
36. For example, palisades around Mississippian villages to keep out deer (?); the “peaceful Pueblos” of the American Southwest (see Wilcox and Haas 1991); the “peaceful” Maya.
37. This example is not completely hypothetical example since the extensive Ancient Mayan road systems now being documented in the Yucatan are being interpreted by many scholars as “ritual roads” (B. Hayden, personal communication).
38. Ferguson 1992a, 1992b. His colleague, Neil Whitehead (1990: 160), blames Hobbes directly, claiming that intruding Westerners brought with them Hobbes’s “ideology of war” (what that ideology is remains unclear since Hobbes never praised war or suggested how it should be conducted). Another proponent of prehistoric peace is Blick (1988).
39. Ferguson 1992a: 113. Except for a single clause in one sentence, Hobbes did not mention any “wild violence” by natives to support his case.
40. Gabriel and Metz 1991: 3, 19. In his latest book, the justly celebrated military historian John Keegan (1993) “buys” Turney-High “lock, stock and barrel,” probably because the latter’s book remains the only general anthropological synthesis on prestate warfare available to nonanthropologists.
41. Rochberg-Halton 1991: B6-B7.
42. Manchester 1980: 102.
Chapter 2
1. Otterbein 1989: 21, 143–44, 148.
2. Ross 1983: 179, 182–83.
3. The Cayapa were indeed peaceful since they had no traditional memory of warfare since mythological times (HSAI vol. 4, 1948: 282).
4. Jorgensen 1980: 503–6, 509–15, 613–14.
5. The Panamint, Battle Mountain, and Hukundika Shoshone; the Gosiute and the Kaibab Paiute of the Great Basin; the Wenatchi and Columbia Salish of central Washington.
6. Harris 1989: 288–89; Meggitt 1962: 38, 42, 246.
7. Knauft 1987; Lee 1979: 387–400; Harris 1989: 288; HNAI vol. 5, 1984: 340–41, 401�
��402, 409, 429, 440–41, 455; J. G. Taylor 1974: 92–92; HSAI vol. 1, 1946: 94–95. Knauft’s (1987) paper on violence in “simple societies” is extremely useful, and most of the homicide rates referred to here were taken from his Table 2. He also calculates that the Semai, the archetype of a nonviolent society, had a homicide rate three times that of the modern United States.
8. HSAI vol. 1, 1946:94–95.
9. Lee 1979: 399; Harris 1989: 288.
10. To equal the Gebusi annual homicide rate of 683 homicides per 100,000 (Knauft 1987: 464), the armed forces of the United States (with an average population of 200 million and homicide rate of 10) would have had to kill 1,350,000 people each year. In nine years, this would amount to 12 million deaths; the population of South Vietnam in 1965 was less than 14 million.
11. Knauft 1987: 463. My conservative calculation (i.e., excluding deaths from disease and starvation) of the annual homicide rate of Nazi Germany (1933 to 1945) yields a figure of approximately 2,000 per 100,000 (over three times that of the Gebusi), indicating that it qualifies as the most homicidal society ever recorded.
12. HNAI vol. 5, 1984: 577–79, 585.
13. Even if only one homicide occurred every fifty years in such a small population, their homicide rate would equal that of the United States.
14. For example, Tonkinson 1978: 32, 118, 123–28; Steward 1938: 83, 91, 140, 176, 179.
15. See also Ember 1978.
16. See Ember and Ember 1992: 248–49.
17. Dentan 1979:58–59. See Knauft (1987: 458) for the Semai homicide rate.
18. Dentan (1979: 2) suggests that the Semai (and, presumably, the related Semang) tradition of flight from violence is a consequence of countless defeats and slave raiding at the hands of the more numerous and aggressive Malays. In other words, the Semai can be characterized as defeated refugees.
19. Appendix, Tables 2.1–2.4; see also Ember and Ember 1990: 255.
20. Heider 1970: 107; Chagnon 1968: 141.
21. Hackett (ed.) 1989: 140, 170, 193.
22. Appendix, Table 2.5.
23. Pospisil 1963: 59–60; Edgerton 1988: 39, 107; Steward and Faron 1959: 190, 209, 223, 245; Grinnell 1923 (II): 44–47; HNAI vol. 8, 1978: 219, 260, 380, 547; HSAI vol. 3, 1948: 480; Vayda 1960: 41; Meggitt 1977: 98–99.
24. Chandler 1966: 1, 102, 1, 106, 1, 113–14; Perret 1989: 553; Gabriel and Metz 1991: 89.
25. For example, Dart 1957; Roper 1969.
26. (Australopithicines) Brain 1981; (Neanderthals) Klein 1989: 333–34; Vend 1991.
27. Vencl 1991; Klein 1989: 387; Jelinek 1991; Gambier and Sacchi 1991; Svoboda and Vlcek 1991; Wendorf and Schild 1986; Wendorf 1968; Greene and Armelagos 1972.
28. Wendorf 1968.
29. Vencl 1991; Frayer, in press; Price 1985. See also Appendix, Table 6.2.
30. For example, Courtin 1984; Keeley 1990.
31. Wahl and König’s (1987) exceptionally intelligent and thorough analysis of the Talheim mass grave deserves far greater notice from archaeologists than it has received.
32. O. Bar-Yosef, personal communication. (Incidentally, Bar-Yosef interprets the Early Neolithic “fortifications” at Jericho as being flood protection and a temple tower.)
33. For example, Milner et al. 1991 (eastern United States); Jurmain 1988 (California); Charters 1989 (Columbia Plateau); Wilcox and Haas 1991; Turner and Turner 1992 (American Southwest). For additional references, see Appendix, Table 6.2.
34. For example, Milner et al. 1991; Rohn 1975; Wilcox 1989; HNAI vol. 7, 1990: 348; MacDonald 1989.
Chapter 3
1. Ferguson 1984a: 26 (referring to Otterbein 1989).
2. Rogers 1970: 14; Oliver 1974: 382; Vayda 1960: 38–40; Carniero 1990: 194–95.
3. Meggitt 1977: 67–69.
4. Turney-High 1981: 34.
5. Koch 1974: 214.
6. Warner 1931; Utley 1984: 105; Robbins 1982:187; Meggitt 1977: 86–91; Glasse 1968: 92; Heider 1970; Ferrill 1985: 22.
7. Utley 1984:99–118.
8. Malone 1991: 22.
9. Meggitt 1977: 57.
10. Turney-High 1949: 26.
11. Turney-High 1981: 69. Despite the importance Turney-High accorded to this “law” of warfare, it has not been taught to officer trainees by the modern armed forces of the United States, Britain, or the former Soviet Union since World War II.
12. Political system versus military sophistication, r = .64; but military success versus military sophistication, r = .44 (Otterbein 1989: 74, 95).
13. Otterbein’s “primary mode of subsistence” and “sociopolitical complexity” codes combined explain 52 percent of the variability (r2 = .52) in his “military sophistication index,” whereas the frequency of war (the lowest number in columns 4–6) and the “military success” codes together explain only 17 percent (r2 = .17).
14. Gabriel and Metz 1991: 56–75.
15. Driver and Massey 1957: 357.
16. See Appendix, Table 3.9.
17. Meggitt 1977: 57–58.
18. The elaborate, finely finished prehistoric axes commonly found at sites in the Southwest and Great Plains of North America—regions where both wood and woodworking were rare—may represent similar cases, especially on the Plains, where warfare victims have tomahawk traumas on their skulls (Willey 1990: 118).
19. Gabriel and Metz 1991: 72; Malone 1991: 15–18.
20. Tonkinson 1978: 32.
21. Meggitt 1977: 57; Connolly 1989: 162.
22. Gabriel and Metz 1991: 75; Handy 1923: 133.
23. DuBois 1935: 125; HSAI vol. 1, 1946: 295, 297, 425, 428; Handy 1923; Heider 1970: 285; Bohannon and Bohannon 1953; L. Bohannon, personal communication; Fadiman 1982: 116; Steward and Faron 1959: 190, 244, 249, 321, 323, 357; Spier 1930: 193–94; Gibbon n.d.; Steward 1941: 338; Aginsky 1943: 456; Stewart 1941: 385; Stewart 1942: 268; HSAI vol. 4, 1948: 4; Mercer 1980: 142; HNAI vol. 15, 1978: 112.
24. Underhill 1989: 221.
25. Gabriel and Metz 1991: 89–91.
26. Weber 1992: 229.
27. Webb 1974: 254–55; Keeley 1993; Wahl and König 1987: 178–79; D. Frayer, personal communication.
28. For exceptions, see Steward and Faron 1959: 190, 221, 358; HSAI vol.3, 1948: 35; HSAI vol. 4, 1948: 489; Morren 1984: 195.
29. Marshall 1987: 248–49. A similar calculation from Keegan’s (1976: 234, 255) figures for the British bombardment at the Somme in 1916 gives a ratio of 250 shells fired for each German casualty inflicted (from all causes).
30. Connell 1984: 259.
31. Turney-High 1981: 42.
32. The basic information and references for this section can be found in the Appendix, Table 3.2.
33. Keeley 1992; Bamforth 1994; Champion et al. 1984: 213–15, 283; HNAI vol. 9, 1979: 65–66, 136, 433.
34. HNAI vol. 8, 1978: 238; Chatters 1989: 241.
35. See Haas and Creamer’s (1993) fine study of this phenomenon in northeastern Arizona.
Chapter 4
1. Heider 1970: 107; Vayda 1976: 18; Dozier 1967: 68; Otterbein 1967: 352; HNAI vol. 8, 1978: 130, 198, 251, 344, 454, 488, 513, 697.
2. Glasse 1968: 92.
3. Keegan 1976: 296, 309.
4. Grinnell 1923 (II): 28–38; Hoebel 1978: 75–77.
5. Mae Enga warriors who killed or seriously wounded several enemies in a single formal battle, thus determining the successful outcome of the combat, were permitted to assume a knotted cord as a mark of honor. Additional knots could be added if the feat was repeated (Meggitt 1977: 66–67).
6. Hanson 1989: 190. No one can read Hanson’s descriptions of a Greek hoplite battle and imagine it a game.
7. The Allies did parole the Sicilians among the Italian prisoners taken during the invasion of Sicily in World War II.
8. Keegan 1989: 390.
9. Edgerton 1988: 178–79; Morris 1965: 449.
10. For example, Manchester 1980: 225–26.
11. HNAI vol. 8, 1978:698.
12. For example, Grinnel 1923 (II): 45–47; Hoebel 19
78: 79.
13. Of course, the frequency of battle dramatically increased under Grant and Sherman in the summer of 1864. It is interesting to note that in the fierce battles for Atlanta in July and August 1864 the proportion of Sherman’s army that was killed in action (usually half of those counted as “killed and missing”) never exceeded 2 percent (Sherman 1886: 608–11).
14. Oliver 1974: 398; Vayda 1976: 25; Carneiro 1990: 199.
15. For example, Herdt 1987: 48–55; Morren 1984: 186.
16. Vayda 1976: 22–23; Turney-High 1949: 124; Robbins 1982: 185, 188; Meggitt 1977: 75–76, 110; HNAI vol. 6, 1981: 408; HNAI vol. 5, 1984: 477; Morren 1984: 188.
17. Chagnon 1968: 141; HNAI vol. 6, 1981: 287; Chagnon 1983: 170; Hogbin 1964: 59.
18. Cannon 1992; Kent 1980.
19. (Plateau) Chatters 1989; (Illinois) Milner et al. 1991; (British Columbia) HNAI vol. 7, 1990: 58; (California) Walker and Lambert 1989; Lambert and Walker 1991; Jurmain 1988; Hohol 1982; (Egypt) compiled from Wendorf 1968 and Anderson 1968.
20. Milner et al. (1991: 583) estimate that the Norris Farms no. 36 cemetery was used only “for a few decades” (thus, let us say thirty years). Prorating the 43 homicides over 30 years gives 1.43 homicides per annum. If the base population of the group using this cemetary was 100, then the homicide rate was 1,430 per 100,000, or 140 times the U.S. rate of 10 per 100,000. If the population was 200, a more reasonable village size (R. Hall, personal communication), then the rate was 717. This latter homicide rate is seventy times that of the United States in 1980, 150 times that of the United States in 1953, and 1,400 times that of Britain in 1959 (Knauft 1987: 464).