37. Scheidel 2014, esp. 21–23. For Orbis, see Scheidel and Meeks 2014.
38. “Economic” refers to the movement of goods, “military” to (fast or slow) troop movements. Steps of expansion: Scheidel 2014: 13–14, with 17, figs. 1–2.
39. For the localized nature of Roman fiscal extraction under the mature empire, see Scheidel 2015a: 248, 251.
CHAPTER FOUR
1. For the case for counterfactuals, see the final section of the introduction.
2. Demandt 1999: 75 invokes the counterfactual of an early “Etruscization” of Rome and consequent lack of Latin support as a potential obstacle to imperial state formation.
3. On the Akkadian imperial project, see Foster 2016: 80–83; on its fall, 22–25. For Neo-Assyrian state-building, see Bedford 2009.
4. Briant 2002 is the most comprehensive account of Achaemenid history.
5. On the war from a Persian perspective, see Briant 2002: 146–61, 525–42; Cawkwell 2005. Several historians have explored the counterfactual of Persian victory in Greece in 480 BCE. Strauss 2006: 99–106 argues for a scenario of a failed Persian invasion of southern Italy followed by revolt and liberation in mainland Greece. Demandt 2011: 83 similarly considers a later revolt the most plausible outcome. Hanson 2006: 71–72 downplays Persian problems of holding on to various peripheries. For Persia’s persistent problems with Egypt, which was lost or only partly controlled by the former for approximately 83 of the 194 years from 525 to 332 BCE, or 43 percent of the time, see Ruzicka 2012: 23 (c. 522–c. 518 BCE), 27–29 (c. 487–484 BCE), 29–33 (462–454 BCE), 35–198 (c. 405–342 BCE), 199–205 (338–335 BCE). Persia’s rule also frayed in Cyprus (380s, c. 348–344 BCE), Asia Minor (366–360 BCE), and Phoenicia (c. 348–345 BCE), and the extent of its control in the Indus Valley remains obscure (Briant 2002: 754–57). See Hansen 2006 for Greek manpower, and Ober 2015: 21–44 on city-state ecology and esp. 71–122 on economic development. Ameling 1993: 15–65 tracks Carthaginian state formation.
6. Athens’s ambitious (if unsuccessful) intervention in Egypt in the 450s BCE in support of a local revolt highlights the scope of its designs on Achaemenid possessions (Ruzicka 2012: 30–32); it was accompanied by operations in Cyprus and on the eastern Mediterranean coast (Biondi 2016). In the 420s BCE, Athenian comedy could at least joke about the possibility of a campaign against Carthage: Aristophanes, Knights 1300–1304. The massive Sicilian expedition in 415 BCE, which in a counterfactual world might have succeeded instead of ending in disaster, would have opened the door to involvement in Italy.
7. Contra I. Morris 2009. The substance of Runciman’s 1990 argument—that Greek city-states were too inclusive and democratic to form viable larger states—has not been refuted: Athenian direct democracy in particular would have militated against the formation of a Gellnerian ruling class that transcended individual communities (Gellner 1983: 9, fig. 1, pictured in chapter 2, figure 2.6 [a] of this volume). For criticism of Morris’s thesis, see Raaflaub 2016: 120–25; Scheidel in press-a.
8. A large army of Macedonian and Asian, mostly Iranian troops, had been assembled at the time, and he had appropriated large quantities of gold and silver from the Achaemenids and other sources (Holt 2016).
9. Diodorus 18.4.4. For contrasting positions regarding its reliability, see the contributions by Fritz Hampl and Fritz Schachermeyr in Griffith 1966: 308–34. Bosworth 1988: 185–211 defends the veracity of this account; see esp. 190–202 for evidence in support of massive war preparations in the Mediterranean.
10. Bosworth 1988: 83–93 analyzes the ancient tradition.
11. Demandt 2011: 86, however, considers Macedonian expansion beyond the Adriatic unlikely: the western regions would have been left to Romans, Celts, and Germans. But see also Demandt 1999: 75–76 for Alexander’s chance of preventing the Roman empire.
12. Toynbee 1969: 441–86, at 467–70. Cf. Ober 1999: 46–47 against Toynbee’s anachronistic model of an enlightened super-state: given Alexander’s earlier conduct, further campaigning would most likely have resulted in more carnage and predation.
13. Livy 9.17–19. Oakley 2005: 184–99 provides painstaking commentary.
14. Seleucus: Grainger 1990: 76–94. Billows 1990: 158–59 argues that Antigonus did not aim to be Alexander’s successor, but compare 183 for the view that an Antigonid victory in 301 BCE would have paved the way for imperial reunification. See also Grainger 1990: 215–17 for powerful constraints on Seleucid expansion.
15. Fronda 2010: 288–300 discusses several counterfactual scenarios, noting Hannibal’s inability to win over enough allies to overcome Rome’s manpower advantage (50–52, 299).
16. Demandt 1999: 76 and 2011: 94 also deems Hannibal’s efforts too late and Carthaginian manpower insufficient.
17. I discussed the alternative counterfactual of a durable Macedonian-Persian empire from Alexander’s rule onward earlier in this chapter. It would also be possible to envision an intermediate scenario of more rapid reunification, in time to intervene against Rome on Carthage’s side, but this would effectively be largely indistinguishable from the first one.
18. Partly based on Fischer-Bovet 2014: 76 and Murray 2012. I reckon with 90,000 soldiers each for the Ptolemies and Seleucids, 40,000 for Macedon, and 15,000 each for the two major Greek leagues, and with 300 warships for the Ptolemies, 100 for the Seleucids, and 100+ for the Aegean powers.
19. Mouritsen 1998 is the best analysis of the relationship between Romans and allies.
20. Irrelevant: unless we assume that Roman expansion into the European periphery depended on material resources from the east, which is a valid concern.
21. Much the same was true of the short-lived civil war of 69 CE, at a time when the legions still heavily relied on recruitment in Italy and Spain.
22. Enduring fragmentation (and just conceivably wholesale failure) first became an option again in the 260s CE but was averted by peripheral military rulership, a development I discuss in forthcoming work.
23. For documentation, see chapters 1 and 8 in this volume.
CHAPTER FIVE
1. Unless noted, all dates in this chapter are CE.
2. Demandt 1984: 695.
3. Quotes: Gibbon 1781: 631.
4. Cross-cultural approaches include Toynbee 1934–1961; Tainter 1988; Yoffee and Cowgill 1988; Randall Collins 1995: 1554–67; Motyl 2001; R. Li 2002; Turchin 2003, 2006; Turchin and Nefedov 2009; S. A. Johnson 2017; Middleton 2017. For commentary, see briefly Scheidel 2013b: 38–40; Middleton 2017: 29–46.
5. Paradigmatically, out of a vast literature, contrast Heather 2006 (external forces) and Goldsworthy 2009 (internal problems). On environment, see Harper 2017.
6. Demandt 1984: 597–600 offers an elegant objection to traditional blends of internal and external causes that seek flaws in entities that fail, and prioritizes external factors (588–97). But this is a problem separate from that of distinguishing internal from external factors.
7. On the mechanisms of migration and settlement, see most lucidly Halsall 2007: 417–54; and cf. also Heather 2009: 335–59; Sarris 2011: 33–40. For the role of the Huns, see Kelly 2009. For the fiscal interdependence of the empire’s regions, see Hopkins 2002; Scheidel 2015a: 251–52.
8. For the use of counterfactuals, see the final section of the introduction in this volume.
9. See chapter 8 in this volume.
10. See Sarris 2011: 127–34 for the various frailties of the East Roman state.
11. Gold reserves had risen from 32 tons in 457 CE to 129 tons in 527 CE (Scheidel 2009c: 175).
12. Summaries in Haldon 1990: 21, 33–34. For the Balkans see now esp. Sarantis 2016.
13. Summaries by Haldon 1990: 20, 34–46; Sarris 2011: 146, 239–54.
14. Haldon 1990: 50–84 and now esp. 2016: 31–55. Picard 2018 describes the rise of caliphal naval power.
15. Harper 2017: 199–245 is the latest analysis.
16. Survival: Haldon 2016: 55–57; for an answer, see 282–94. Attrition: D
ecker 2016: 8, 22 (maps); Treadgold 1995: 42–75 (army); Haldon 2016: 27–29 (revenue). Archaeology: Decker 2016: 191–94, in an account that is scrupulously wary of declinist narratives of the Byzantine “Dark Ages” in the seventh through the ninth centuries.
17. See chapter 4 in this volume.
18. H. Kennedy 2007 is a standard narrative of the Arab conquests.
19. Size: Taagepera 1997: 496 (but cf. Scheidel in press-b). Population estimated from McEvedy and Jones 1978, adjusted for these authors’ undercount (see the technical note to chapter 1). Population shares: see chapter 1 in this volume.
20. See H. Kennedy 2007: 366–70 for a good survey, and also Sarris 2011: 272.
21. Quote: H. Kennedy 2007: 370. Early successes: see H. Kennedy 2001: 2 for another counterfactual.
22. See Stathakopoulos 2004 for a dossier and discussion of the various episodes. H. Kennedy 2007: 366–67 on plague-induced depopulation as a factor.
23. H. Kennedy 2001: 6; H. Kennedy 2007: 370–74; Sarris 2011: 273–74.
24. Quoted from H. Kennedy 2007: 1, in Sebastian Brock’s translation from the original Syriac.
25. H. Kennedy 2001: 4, 8–9, 18–19, 30–51 for elements of this summary.
26. Wickham 2009: 322–23, 325, 331–33. See Gordon 2001 for the rise of Turkic forces in the Samarra period, and more generally chapter 7 in this volume.
27. For divergent fiscal evolution under the Germanic successor regimes and the Arabs, see chapter 7 in this volume.
28. H. Kennedy 2001: 59–95 on military payments, esp. 59, 62–64, 74–78; H. Kennedy 2015: 393–94, 397; and also Haldon 2016: 75–77.
29. H. Kennedy 2001: 19–21, 76–78.
30. As noted by Wickham 2009: 334–35.
31. H. Kennedy 1995: 256–58.
32. Quote: Gibbon 1788a: 408–9.
33. Strauss 1999: 85–89.
34. H. Kennedy 1995: 259–69; Wickham 2009: 338–47.
35. H. Kennedy 1995: 251–54.
36. Parker and Tetlock 2006: 378–79 also stress the likelihood of fragmentation in the event of Arab success.
37. Kasten 1997 is an exhaustive study of the succession of Frankish rulers and the attendant partitions of power and territory. See esp. 9–14 for the origins of this practice.
38. For the mode of succession specifically in this period, see Kasten 1997: 136–98.
39. For summaries, see Nelson 1995a; Roger Collins 1999: 333–63, as well as Costambeys, Innes, and MacLean 2011: 379–427, on fragmentation after 843, and Wickham 2009: 427–52, on the successor states of Carolingian France, esp. 439–44 on West Francia; also Wickham 2016: 102, 105. See also Tabacco 1989: 144–81 for anarchy in post-Carolingian Italy.
40. Kasten 1997: 574 concludes that shifts to more primogenitural arrangements were generally limited to periods of political and material weakness.
41. See chapter 7 in this volume. I sidestep the debate over whether these developments ushered in “feudalism”: see Patzold 2012 for a recent survey of the controversy, esp. 14–43 on eighth- and ninth-century Francia. Cf. also Costambeys, Innes, and MacLean 2011: 317–19; Wickham 2016: 10. Reynolds 1994 remains the classic challenge to the application of the feudal model to the early Middle Ages. In the present context, this issue is not of critical importance.
42. Halsall 2003: 25–27, 31; Halsall 2007: 492–97.
43. Wickham 2005: 102–15 is the classic analysis.
44. Halsall 2003: 54–56, on the pre-Carolingian period.
45. Fouracre 1995: 86, 89, 93; Halsall 2003: 73–77.
46. Wickham 2001: 72 (quote); Nelson 1995b: 387, 394.
47. Roger Collins 1999: 295–98.
48. Airlie 1995: 448–49; Nelson 1995b: 394; Halsall 2003: 92.
49. Wickham 2001: 91; Halsall 2003: 93–95, 99–101.
50. Fouracre 1995: 108–9.
51. Successor states: Halsall 2007: 508–12; S. F. Johnson 2012: 50. Government: S. F. Johnson 2012: 48, 550. Charlemagne: Fouracre 1995: 107.
52. Halsall 2003: 119–33; Barbero 2004: 249–71, esp. 252, 256–59, 265–68.
53. McCormick, Dutton, and Mayewski 2007, esp. 875–89; Newfield 2013: 169. “Charlemagne was more than vigorous and smart: he was, with respect to volcanic aerosols and rapid climate change, a very lucky ruler” (McCormick, Dutton, and Mayewski 2007: 892).
54. See Moreland 2001: 404–6, 414–17 for an admirably sober account.
55. Most remarkable: thus Fouracre 1995: 86.
56. Wickham 2009: 430–32.
57. See C. Morris 1989: 113–21, 154–73 for the conflict over lay investiture.
58. E.g., Tabacco 1989: 182–236, esp. 220, and chapter 10 in this volume. Lombard League: Raccagni 2010. The Mediterranean disease environment also played a role: both Otto II and Otto III succumbed to malaria, which repeatedly ravaged invading armies.
59. P. Wilson 2016, esp. 9–11, 44–45, 343–44, 348–52, 356–61.
60. P. Wilson 2016: 365–77 (quote: 366).
61. France 1999: 77–106, esp. 85–87, 107, and 230–34 (quote: 232). See also Brauer and van Tuyll 2008 for the underlying economics: 45–79.
62. P. Wilson 2016: esp. 325–27, 332, 372, 389–90, 404.
63. P. Wilson 2016: 87, 319, 324. Analogy to sovereign rulers: Arnold 1991: 284.
64. Angelov and Herrin 2012: 167–71; P. Wilson 2016: 46–76.
65. Arnold 1991: 281 notes that “if the experience of other west European kingdoms, Angevin England, Hauteville Sicily, and Capetian France, is anything to go by, then the regal machinery equal to overturning the local rule of the princes would have had to include regular taxation throughout Germany, annual judicial assizes, and a capital residence gradually enforcing political centripetence.” None of these features was on the horizon (280–84).
66. P. Wilson 2016: 380 (quote), 385–87.
CHAPTER SIX
1. From the rich literature, see, e.g., May 2012; Man 2014. For the role of climate in the rise of the Mongols under Genghis, see Pederson et al. 2014 (unusual warm and wet conditions from 1211 to 1225), and cf. also Putnam et al. 2016 for the southward spread of grasslands more generally.
2. For the European campaigns, see Chambers 1979; Hartog 1996: 29–41; Jackson 2005: 58–86; Man 2014: 138–45; Sverdrup 2017: 293–327. Medieval reports and modern estimates regularly posit Mongol forces in excess of 100,000 men including auxiliaries (e.g., Hartog 1996: 29; Sinor 1999: 19); but see Sverdrup 2010: 114 and 2017: 297, who argues for 40,000 Mongols, an estimate that, however, neglects the potential for drawing in allies from the western Eurasian steppe.
3. E.g., Chambers 1979: 70–82. For more on Kievan Rus’, see chapter 8 in this volume.
4. For conditions in Europe, see the previous section, and briefly Chambers 1979: 87–90; Hartog 1996: 34–35; Jackson 2005: 9–10. Army sizes: France 1999: 13, 128–30.
5. Specifically for Central Europe, see Chambers 1979: 86–113; Jackson 2005: 63–74. Co-optation of Turkic warriors: Hartog 1996: 29, who gives a total of 120,000–140,000 men; but see note 2 above.
6. Chambers 1979: 110–11; Jackson 2005: 65–67; Man 2014: 143–44.
7. For discussion of divergent scholarly positions on this matter, see especially the surveys by Rogers 1996 and Pow 2012: 12–45; and cf. also more briefly Jackson 2005: 71–74. Raid: Jackson 2005: 73–74, who also considers other reasons. Sverdrup 2017: 325 notes the disposition of Batu’s forces in early 1242.
8. See Pow 2012: 34–41 for critical discussion of this position.
9. See Rogers 1996: 12–14 and Pow 2012: 41–45 for discussion. For a cold and wet spell in the winter of 1241–1242 that reduced pasture and mobility, see Büntgen and Di Cosmo 2016, rejected by Pinke et al. 2016. The signal is not strong, and does not explain why the Mongols did not return sooner.
10. Accepted by Strauss 1999: 92; Parker and Tetlock 2006: 380. Jackson 2005: 72 assigns some significance to this event, noting the conflict between Batu and Guyuk. Cf. also Mielants 2007: 159, who claims that the Mongols
could easily have overrun Europe.
11. See Rogers 1996: 8–9, 15–18; Pow 2012: 12–24 for discussion and criticism. Their reference to continuing operations in Anatolia and Syria is not compelling, for the reason given in the text.
12. Pow 2012: 46–77 is by far the best discussion of this issue in relation to Mongol expansion.
13. Jackson 2005: 69; Pow 2012: 59. Quote: Annales S. Pantaleonis in Jackson 2005: 64.
Escape From Rome Page 61