23 April 2024: We are investigating problems with ecommerce on Cambridge Core. We apologise for any inconvenience and hope to have full functionality restored shortly.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 10 May 2016
The following joint article is a departure from standard studies, in that historical research is put side-by-side with numismatic evidence. It reflects the growing awareness of the underlying concepts of steppe society that significantly shaped the formation and endurance of the Mongol Empire. With new analysis, it is apparent that the society was clear about these concepts and expressed them in very public pronouncements. They are most evident in the early period of the empire; during the formation of the state by Chinggis Khan and his first two successors, Ögödei (r. 1229–41) and Güyük (r.1246-48). However, the cataclysmic civil war in the middle of the thirteenth century between the Ögödeyids and Toluids removed direct acknowledgment of such a social ethos. Indeed, after 1250 khans strongly focused on pragmatic issues and relied less on philosophical theories of legitimacy, at least Mongolian ones. By contrast, the first three rulers were keenly aware of the theory of the state and the way society functioned within it. They developed this ethos into a fairly cohesive form that provided moral strength to a nascent regime. The evidence for this development emerges from the study of two particular words, tengri, “Heaven”, and especially töre, “grand principle”. Töre in this usage was the equivalent of the ‘binding and unbinding’ and the sunna of the messenger, Muḥammad, in medieval Islamic societies and of democracy in modern times. Tengri and töre are culturally defined theories closely related to the Aristotelian sense of positive law. In all cases, reality required various approaches to them at a given period of time. As a result, the concept of töre had existed before the empire and continues to this day, always implying the correct order of good governance.
1 Note that while in the present paper the authors follow Chinese transcriptions and reconstructed Middle Mongolian within passages quoted directly from the relevant primary sources, the Turkicised spelling Chingīz Khān for the founder's name is well attested on his coins.
2 The SH text used here is in Ligeti, Louis (ed.), Histoire Secrète des Mongols (Budapest, 1971)Google Scholar. See also the translation by de Rachewiltz, Igor, The Secret History of the Mongols, A Mongolian Epic Chronicle of the Thirteenth Century, 3 vols (Leiden, 2004–2013)Google Scholar. I would like to thank Eugene Anderson, Timothy May and Dagmar Schäfer for commenting on earlier drafts of this paper. Professor Schäfer suggested the framework for the study.
3 This implies that the qan too is subject to a proper ordering of society and is conscious of his role.
4 On the Mongol Empire, see Buell, Paul D., Historical Dictionary of the Mongol World Empire (Lanham, MD, 2003)Google Scholar.
5 An il, also el, was a pacified population and the term was used with reference to China's provinces by Rashīd al-Dīn. See Paul D. Buell, “Tribe, Qan and Ulus in early Mongol China: Some prolegomena to Yüan history”, doctoral dissertation, University of Washington, 1977.
6 See, as an introduction, Buell, Paul D., “Qubilai and the Indian Ocean: A new era?”, in Babones, Salvatore and Chase-Dunn, Christopher (eds.), Handbook of World-Systems Analysis (London, 2012), pp. 42–43,Google Scholar Ciolcîltan, Virgil, The Mongols and the Black Sea Trade in the Thirteenth and Fourteenth Centuries, translated by Samuel Willcocks (Leiden, 2012)CrossRefGoogle Scholar.
7 On the cultural exchanges, see Allsen, Thomas T., Culture and Conquest in Mongol Eurasia (Cambridge, 2001)CrossRefGoogle Scholar; and Paul D. Buell, “How Genghis Khan changed the world,” digital at http://www.mongolianculture.com/
8 See also the discussion in De Rachewiltz, Secret History, pp. xxv-cxiii.
9 After it was written, the SH underwent considerable editing, including the insertion of additional, sometimes anachronistic material, and also factual and terminological “corrections”. For example, Chinggis Khan was never a qaghan, pronounced in Middle Mongolian as qahan or qa'an, a Turkish term meaning emperor. It was later widely used to designate the Mongol ruler of China, among others, and because of the later prestige of the title, the qan of the text of the SH has been dutifully corrected in many places.
10 Nomadic patterns of movement among the early Mongols were generally what the Turkic peoples call yailaq-qishlaq, a regular alternation between summer pastures, yailaq, at higher altitude, and winter pastures, qishlaq, at lower altitude. Sometimes, as obviously occurred here, this meant movement along a river or stream. A rarer variant to such movement was a circular movement over quite large territories to maximize access to limited pasturage without much altitude change. See Buell, Historical Dictionary, pp. 1-2.
11 Tabooing is at work here. Yellow means yellow, but also white, one of the colours of Tengri, and a dog is here a stand-in for wolf, a Mongolian tribal totem.
12 See Buell, Historical Dictionary, p. 5. The fullest treatment of the Mongolian social and kinship system is D. Gongor, Khalkh Tovchoon, 2 vols, BNMAU Shinhlekh Ukaany Akademiyn Tüükhiyn Khüreelen (Ulaanbaatar, 1970-1978).
13 On the qan’s title see Igor de Rachwiltz, “The title Činggis Qan/Qaγan reexamined”, in W. Heissig and K. Sagaster (eds.), Gedanke und Wirkung. Festchrift zum 90. Geburtstag von Nikolaus Poppe (Wiesbaden, 1989), pp. 281-298.
14 Or fallow doe.
15 This term occurs here and in female names. No one is quite sure what it means.
16 A jarliq was an imperial pronouncement pertaining to the total ulus of the Mongols, the Yeke Mongol Ulus, or “Great Mongol Ulus”.
17 The discussion here is largely drawn from Paul D. Buell, “Some aspects of the origin and development of the religious institutions of the early Yuan period”, unpublished master's thesis, University of Washington, 1968.
18 Töre also occurs elsewhere in the SH. Yeke töre, for example, is also found in 220 with much the same meaning. Alone töre in 121 and 178 occurs in a general sense as circumstances, conditions, perhaps an older, a less specialised meaning. In 263, this meaning of töre occurs as a binome in yosun töre, “preferred ways”, in this case meaning the “preferred ways” of the cities of the Islamic world and China, i.e., the proper ways for cities. For a complete, historically-driven discussion of the word see Humphrey, Caroline and Hürelbaatar, A., “The term töre in Mongolian history”, in Sneath, David (ed.), Imperial Statecraft: Political Forms and Techniques of Governnance in Inner Asia, 6th-20th Centuries (Bellingham, WA, 2006), pp. 263–292 Google Scholar.
19 Patrimony is ulus, the empire as the joint property of the imperial clan at one level and the holdings of individuals at another. See Buell, “Tribe, Qan and Ulus”.
20 Cook is something of a misnomer. Bawurci comes from bawur, liver, and means dispenser of delicacies.
21 See “Some royal Mongol ladies: Alaqa-beki, *Ergene-Qatun and others”, World History Connected, 2010, digital at http://worldhistoryconnected.press.illinois.edu/7.1/buell.html. On the Öng’üt see now Christopher P. Atwood, “Historiography and the transformation of ethnic identity in the Mongol Empire: The Öng’üt case”, Asian Ethnicity, digital at http://dx.doi.org/10.1080/14631369.2014.939333
22 On this campaign see Buell, Paul D., “Early Mongol expansion in Western Siberia and Turkestan (1207-1219): A reconstruction”, Central Asiatic Journal 36 (1992), pp. 1–32 Google Scholar, and idem, “Sübötei-ba'atur”, in Igor de Rachewiltz, Chan Hok-lam, Hsiao Ch'i-ch'ing and Peter W. Geier (eds.), In the Service of the Khan. Eminent Personalities of the Early Mongol-Yuan Period (1200-1300) (Wiesbaden, 1993), pp. 13-26.
23 See Buell, “Tribe, Qan and Ulus”.
24 See Buell, Paul D., “Chinqai (1169-1252): Architect of Mongolian Empire”, in Kaplan, Edward H. and Whisenhunt, Donald W. (eds.), Opuscula Altaica, Essays Presented in Honor of Henry Schwarz (Bellingham, WA, 1994), pp. 168–186 Google Scholar.
25 Discussion in Buell, “Tribe, Qan and Ulus”.
26 This implies that the qan himself was not going to eat until his retainers are satisfied.
27 That is, positions in the qan’s ger. Such positioning was also expressed when regular palaces were built.
28 Buell, “Chinqai”, In the Service of the Khan, pp. 95-111.
29 Ibid ., p. 103.
30 Such an inscription, using Turkic khan and the Middle Mongolian form qa'an, qaγan in Classical Mongolian, was largely for the consumption of the world outside Mongolia, which was used to Turkic titles. Qaγan was not used as a formal title by the Mongols until later although it already appears on some of the coins of Cinggis-qan.
31 For an example, see the Oriental Coin Database (zeno.ru), no. 5002.
32 Although there was a regular style, there were variations. The Shahāda coinage seems to have actually begun with billon as represented with the Timurdh coin of 628/1231: Petrov, Pavel N., Ocherki po numizmatike mongol´skikh gosudarstv XIII-XIV vekov (Nizhny Novgorod, 2003), p. 109 Google Scholar, no. 68. However, the main group was electrum. The whole series was short-lived but geographically extensive: Kolbas, Judith, The Mongols in Iran: Chingiz Khān to Uljaytu, 1220-1309 (London, 2006), pp. 92–96 Google Scholar. In this paper, the mints do not have diacritic marks that would represent Arabic spelling.
33 At the death in 623/1226 of the Caliph al-Ẓāhir, Mongol coinage from Samarqand did not cite the new caliph. Only the title remained as a continuation of Islamic practice. See Kolbas, The Mongols in Iran, p. 65.
34 In Chinese sources, the population from the older census (or censuses) in China is called jiu hu 舊戶, “old families”, and newly-enumerated population xin hu 新戶, “new families”. Population belonging to the great warlords and princes was not counted: Otagi Matsuo 愛宕松男, “Mōkojin seikenji shita no kanchi ni okeru hanseiki no mondai 蒙古人政權治下の漢地に於ける版籍の問題, Asiatic Studies in Honour of Torū Haneda, 383-429; Buell, “Tribe, Qan and Ulus”; and the overview in Allsen, Thomas T., “The rise of the Mongolian empire and Mongolian rule in north China”, The Cambridge History of China, VI, Alien regimes and border states, 907-1368 (Cambridge, 1994), pp. 321–413 Google Scholar (at pp. 373-379). Note that no Mongol census was a complete canvass but focused on new population elements first available for distribution after a conquest.
35 Kolbas, The Mongols in Iran, p. 229.
36 In 632/1234, after the first quriltai, taxes were simplified in the extreme so that it took about two years after the second one to develop the full policy: Kolbas, The Mongols in Iran, pp. 96-97.
37 For Pūlād, Oriental Coin Database (zeno.ru), no. 4789; for Imil, no. 47899.
38 Kolbas, The Mongols in Iran, pp. 108-111.
39 Buell, “Chinqai”, p. 100.
40 Herat had been in ruins from 1222 and began to be re-established in 634/1236. Tolui had taken 1,000 weavers to Besh Baligh, where they had made fabric for the court. Finally, a noble was allowed to lead a small group of 100 families back to the city. In 636/1239, 200 more families joined them; and remnants from the countryside started to swell the population. A census was taken, and by 637/1240 there were 6,900 people. The city continued to grow. In another case, Körgüz greatly encouraged the rebuilding of devastated Ṭūs: Howorth, Henry H., History of the Mongols, from the Ninth to the Nineteenth Century, III, The Mongols of Persia (London, 1888), pp. 40–41 Google Scholar.
41 Later in 651/1253, Rubruck noted a thriving Muslim section of the town with shops and an open-air market. He does not, however, indicate their origin. See Jackson, Peter, translation and edition, The Mission of Friar William of Rubruck, The Hakluyt Society, no. 173 (London, 1990), p. 211 Google Scholar.
42 Heidemann, Stefan, Kelzenberg, Hendrik, Erdenebat, Ulanbayar and Pohl, Ernst, “On ‘The first documentary evidence for Qara Qorum for the year 635/1237-8’”, Zeitschrift für Archäologie aussereuropäischer Kulturen 1 (2005), pp. 93–102 Google Scholar. Also presented by Whaley, Mark A., “ The first documentary evidence for Qara Qorum from the year 635/1237-8”, Acta Mongolica 11 (2011), pp. 113–128 Google Scholar. The key word töre on the coin was first read for the authors by Timothy May in 2009 and later confirmed by Paul Buell. Stefan Heidemann, however, in the article as well as in personal communication, continues to consider the script illegible.
43 Pavel Petrov and Vladimir Belyaev have discovered and analyzed Ögödei's tamgha on other coin series: P. N. Petrov and V. A. Beliaev, “K voprosu o personalizatsii tamg na monetakh Chagataiskogo ulusa”, Trudy Mezhdunarodnykh Numismaticheskikh Konferentsii: Monety i denezhnoe obrashchenie v mongol´skikh gosudarstvakh XIII-XV vekov (2001), pp. 79-95.
44 One coin is not enough to determine the position of the issue in the 24 carats that constituted the mithqāl system; however, because of its weight and size, the coin is about a third of the presumed complete mithqāl weight, making it a rather light coin for the period.
45 See the progression of this symbol on coinage from Kurram to Herat during the period that Ögödei was in command of Zābulistān: Kolbas, The Mongols in Iran, pp. 91-95.
46 Igor de Rachewiltz (Secret History, III, p. 133, n. 41) considers that the word is not töre but yörı, the second person imperative of Turkic yör, “to walk, move”. He takes this as imitating the tong 通, “pass through”, and xing 行, “move”, on Chinese coins. He cites an Ögödeyid silver issue inscribed on one side as da chao tongbao 大朝通寶, “currency of the great court”. The authors of this article suspect that the purpose of that Ögödeyid joint-language issue was to tax the Chinese population, but the issue did not combine two numismatic traditions on the same side. Therefore, a direct Chinese instruction was probably not transferred to this Islamic-style coinage. Rather, this one Uighur word and the tamgha were the only items on the excavated coin not influenced by external factors. Moreover, the later Güyük coins, which were much more prolific, do not have any counterparts with Chinese characters. After further examination of several more coins, this author re-confirms that the first letter is ‘t’. Its initial form in Uighur is a small circle with a short upward stroke on its left side. There is no beginning base line. After that letter, in the word under discussion, comes the much larger and oblong open form for ö that is like an Arabic ص/ṣad so that the ‘t’-circle is often overshadowed by it. Depending on the calligraphy, the first letter looks like an open English ‘o’ (Oriental Coin Database [zeno. ru], no. 121161) or ‘c’, but more often a filled-in ‘o’ that melts into the short vertical stroke (silver coin from the British Museum, no 1967.0112.2, 3.19 grams). The open circle is visible, for example, on the reverse of the gold coin owned by Terry Lee. Some examples, however, have an unfinished circle pushed into the short line, such as on the reverse of the British Museum coin. Moreover, since many coins start the letter on a base line, two other forms appear: one has an anvil shape (Oriental Coin Database [zeno. ru], no 121210, with illegible reverse); another makes a single upward sweep to the top left of the letter. This last shape creates a solid right triangle (Oriental Coin Database [zeno. ru], no 121171, with the mint Qara Qorum at the top) that matches the initial ‘y’ that Igor de Rachewiltz has read. Vladimir Nastich has also studied many of these coins. Of particular interest is his Table 2 on page 20 of his report, giving the calligraphy for this word from coins of the Utrār hoard. Of the eleven examples, not one quite matches another one, which is further visual proof of the difficulty in deciphering the word. In spite of such a detailed study, Dr Nastich declined to provide a confident reading of the word (reconfirmed by private communication). However, the largest and best circle of all existing specimens appears on the 635/1237 Qara Qorum coin itself. Therefore, the authors consider that this wider survey combined with the 635/1237 coin bears out the reading of töre. This argument and some of the previous discussion with illustrations are also covered by the authors in “Töre and Ulūsh: Discovering the emergence of a Mongol state ethos through coinage”, in Golden Horde Numismatics (Kazan) 3 (2013), forthcoming (p. 7).
47 See the discussion above.
48 For example, Chinqai instructed Yelü Chucai in the script: Buell, “Chinqai”, p. 102.
49 “. . .the Mongols’ ideology. . .of special good fortune, or royal charisma,. . .(was) a concept that passed to the Turkic nomads through the agency of the Sogdians. . . .It was a central element in the quest for legitimacy. In this respect,. . . .the Mongols closely followed their Turkic predecessors”: Allsen, Thomas T., “A note on Mongol imperial ideology”, in Rybatzki, V. et al. (eds.), The Early Mongols: Language, Culture and History: Studies in Honor of Igor de Rachewiltz (Bloomington, 2009), pp. 1–8 Google Scholar (at p. 1).
50 Buell, “Chinqai”, p. 106.
51 For an example with the mint of Qara Qorum at the top and töre starting with a base line, see Oriental Coin Database (zeno.ru), no. 121161.
52 Note the discussion in Heidemann, “On the first documentary evidence”, p. 95, of the Utrār hoard and other excavation finds at Qara Qorum.
53 Some people, including Heidemann, consider that the mint is referred to in the main inscription. However, the mint is rarely alone in the middle of the main field; and the preceding similar types indicate that the main inscription did not indicate the ruler's ordo as the mint site.
54 K. M. Baipakov and V. Paul D. Buell, Charite Universitäts Medizin and Max Planck Institute, Berlin, and Judith Kolbas, Central Asian Numismatic Institute, Cambridge. Nastich, N, “Klad serebrianykh veshchei i monet XIII v. iz Otrara”, in Kazakhstan v epokhu feodalizma (Problemy etnopoliticheskoi istorii) (Alma-Ata, 1981), pp. 20–62 Google Scholar.
55 Terry Lee, Canberra, private collection; also Oriental Coin Database (zeno.ru), no. 90114. Because of the shahāda, it seems that the gold type was struck further west, perhaps closer to Samarqand than to Qara Qorum. There is no indication that the type was issued under Güyük, but it does try to follow his common reverse inscription.
56 Kolbas, The Mongols in Iran, pp. 124-128.
57 Janjava, Giorgi and Paghava, Irakli, “Ahar: A new mint issuing Ulugh Mangyl Ulus Bek type coins”, Journal of the Oriental Numismatic Society 215 (2013), pp. 22 Google Scholar-23.
58 Kolbas, The Mongols in Iran, p. 134.
59 The Pulad (6)33/1235-6 silver coin does not state Allāh but instead has Ögödei's tamgha at the bottom. Above it is the Mongolian phrase (in awkward Arabic script) kuchigdur Tengri (“By the power of Heaven”): Oriental Coin Database (zeno.ru), no. 47899.
