1. “Kanun Daimer Nathi,” Chittagong District Collectorate Record Room, No. 64, bundle 78, case no. 5027; No. 7, bundle 62, case no. 4005.

2. Fātiḥa-yi imāmīn-i aiyām-i ‘āshūrā. Ibid., No. 251, bundle 55, case no. 3568; No. 127, bundle 63, case no. 4049; No. 232, bundle 30, case no. 1935.

3. For example, in 1749 the Mughal government granted Shaikh Jan Allah of Satkania Thana 25.6 acres of jungle so that he might perform fātiḥa-yi darvīshī at a shrine entrusted to him. Ibid., No. 196, bundle 19, case no. 1093.

4. Muṣallīān-i fātiḥa, or fātiḥa-khwānī, as in ibid., No. 31, bundle 77, case no. 4971; No. 40, bundle 47, case no. 3054. “Qur’an readers” were titled Qur’ān-khwānī, or tilāwat, as in No. 65, bundle 73, case no. 4677; No. 72, bundle 63, case no. 4100; No. 113, bundle 35, case no. 2296.

5. Melford E. Spiro, “Religion: Problems of Definition and Explanation,” in Anthropological Approaches to the Study of Religion, ed. Michael Banton (London: Tavistock Publications, 1969), 96.

6. Bhattacharyya, “Tiger-Cult and Its Literature,” 49–56; Tarafdar, Husain Shahi Bengal, 17–18, 164–66, 233–35; P. K. Maity, Historical Studies in the Cult of the Goddess Manasā (Calcutta: Punthi Pustak, 1966), 182. See esp. Asim Roy, The Islamic Syncretistic Tradition in Bengal (Princeton: Princeton University Press, 1983).

7. Bhattacharyya, “Tiger-Cult and Its Literature,” 49–50.

8. In parts of Bengal this sort of easy inclusion has persisted down to modern times. In 1956 it was noted that fishermen in the West Bengal Sundarban forests, before putting their nets to water, typically performed pūjā to the forest goddess Bon Bibi (literally, “forest goddess”) as a ritual intended to protect them from harm. For this purpose a small thatched and bamboo hut was constructed, in which was placed a clay image of Bon Bibi seated on a tiger. Flanking her on her right was an image of Daksin Ray, depicted as a strong, stout man standing with a sword. Behind him stood a bearded Muslim faqīr known as Ajmal, and in front of Daksin Ray lay the body and severed head of a young boy. Although the names and functions of the figures in the 1956 account differ from those of the seventeenth-century poem, the elements of both accounts—a tiger deity, a soldier, and a superhuman agent identified with Islam—have remained constant over the centuries, distinct from one another but included within a single religious cosmology. H. L. Sarkar, “Note on the Worship of the Deity Bon Bibi in the Sundarbans,” Journal of the Asiatic Society of Bengal 22, no. 2 (1956): 211–12.

9. Dinesh Chandra Sen, who collected and published this sort of oral folk literature in the early twentieth century, dated this particular ballad to the fourteenth century. But a recent study of the collector’s pioneering work concludes that “D. C. Sen, obviously influenced by his patriotic feelings, dated the ballads rather too far into the past; unable to give other reasons than rather vague ‘impressions of antiquity’ or social aspects which, as known, can be very misleading in such cases. There are, no doubt, old or even ancient portions, but the whole treatment of the stories, their ideology and psychological approach, point rather to more recent times.” Dusan Zbavitel, Bengali Folk-Ballads from Mymensingh and the Problem of Their Authenticity (Calcutta: University of Calcutta, 1963), 14.

10. Dinesh Chandra Sen, ed. and trans., Eastern Bengal Ballads, 2: 283–84. The text is to be found in Prācīn Pūrba Baṅga Gītikā, ed. Ksitish Maulik (Calcutta: Mukherjee Publishers, 1972), 4: 308–10.

11. Dinesh Chandra Sen, ed. and trans., Eastern Bengal Ballads, 1: 219–20; text, Prācīn Pūrba Baṅga Gītikā, ed. Maulik, 3: 33.

12. J. D.Y. Peel, “Syncretism and Religious Change,” Comparative Studies in Society and History 10 (1967–68): 124–25.

13. Melford Spiro, “Religion and the Irrational,” in Symposium on New Approaches to the Study of Religion, ed. June Helms (Seattle: American Ethnological Society, 1964), 110.

14. Igor Kopytoff, “Classifications of Religious Movements: Analytical and Synthetic,” in Symposium on New Approaches to the Study of Religion, ed. June Helms (Seattle: American Ethnological Society, 1964), 78–79.

15. Ibid., 80.

16. D. C. Sircar, “Veraval Inscription of Chaulukya-Vaghela Arujuna, 1264 A.D.,” Epigraphia Indica 24 (1961–62): 141, 150.

17. From the thirteenth century on, Muslim Turks in Bengal built mosques with dedicatory Arabic inscriptions similar to that of the Veraval mosque. Unfortunately, though, no Bengali mosque has yet been found with a companion Sanskrit or Bengali text that might tell us how non-Muslims perceived such mosques and the deity worshiped there. It is only in premodern Bengali literature that we begin to see identifications of the type illustrated in the Veraval bilingual inscription.

18. Haji Muhammad, Nūr Jamāl (Dhaka University Library MS. No. 374, sl. 260), fol. 6 mc. Cited in Asim Roy, “Islam in the Environment of Medieval Bengal” (Ph.D. diss., Australian National University, 1970), 194.

19. Saiyid Murtaza, Yoga-Qalandar (Dhaka University Library MS. No. 547, sl. 394), fol. 1a. Cited in Roy, “Islam in the Environment,” 299.

20. Saiyid Sultan, Nabī-Baṃśa, ed. Ahmed Sharif (Dhaka: Bangla Academy, 1978), 1: 1ff.

21. ‘Ali Raja, Jñāna-sāgara (Dhaka University Library MS. No. 146b, sl. 9), fols. 109, 215, 216. Cited in Roy, “Islam in the Environment,” 199, 202.

22. Actually, these literati were building on a long tradition of Persianization. As a result of prolonged contact between Persianized Turks and Bengalis since the early thirteenth century, many Persian words and phrases were absorbed into the Bengali language. These became so thoroughly indigenized that Bengalis today are often unaware of their foreign origin. In 1966 Shaikh Ghulam Maqsud Hilali published a lexicon of over nine thousand Bengali words and expressions of Persian origin. See Shaikh Ghulam Maqsud Hilali, Perso-Arabic Elements in Bengali, ed. Muhammad Enamul Haq (Dacca: Central Board for Development of Bengali, 1967).

23. See Mannan, Emergence and Development, 86–102; Roy, “Islam in the Environment,” 315–21.

24. Asim Roy, “The Social Factors in the Making of Bengali Islam,” South Asia 3 (August 1973): 29.

25. Mannan, Emergence and Development, 99. Mannan’s translation.

26. Saiyid Sultan, Nabī-Baṃśa, 1: 115.

27. As we saw in Chapter 5, non-Bengali sources attest to the emergence by Saiyid Sultan’s time of a rural Muslim community in the area in which that poet lived. In 1567 Cesare Federici found Muslims inhabiting Sondwip, a large island that is actually visible from Sitakund, in northern Chittagong District, where Saiyid Sultan lived and wrote some twenty years later.

28. ‘Abd al-Nabi, Vijay-Hamza, quoted in Muhammad Enamul Huq, MuslimBāṅglā Sāhitya, 2d ed. (Dacca: Pakistan Publications, 1965), 214–15. Translated by Rafiuddin Ahmed, “Conflict and Contradictions in Bengali Islam: Problems of Change and Adjustment,” in Sharī‘at and Ambiguity in South Asian Islam, ed. Katherine P. Ewing (Berkeley and Los Angeles: Universitiy of California Press, 1988), 122.

29. Cited in Ahmed Sharif, Saiyid Sultān: Tār Granthābalī o tār Jūg (Dhaka: Bangla Academy, 1972), 203.

30. Jack Goody, “Religion, Social Change, and the Sociology of Conversion,” in Changing Social Structure in Ghana, ed. Goody (London: International African Institute, 1975), 103.

31. Acts 17:22–23.

32. “The Lo Dagaa [people of West Africa] did not initially think of the acceptance of Christianity as conversion, because the introduction of a new cult does not involve a displacement of other gods,” Jack Goody notes (“Religion, Social Change, and the Sociology of Conversion,” 103).

33. Communication from Tony K. Stewart, July 1990. Although study of the Satya Pir cult has thus far been rudimentary, Stewart has begun a careful investigation with a view to translating representative samples of Satya Pir literature and to tracing the cult’s evolution over time.

34. Dinesh Chandra Sen, The Folk Literature of Bengal (1920; reprint, Delhi: B. R. Publishing Corporation, 1985), 99–102.

35. Thus the modern folklorist D. C. Sen interpreted the Satya Pir cult as a hybrid cult synthesized by “Hindus” and “Muhammadans.” “When two communities mixed so closely, and were so greatly influenced by one another,” he wrote, “the result was that a common god was called into existence, worshiped by the Hindus and Muhammadans alike. His name was formed by compounding an Arabic [sic] word with a Sanskrit word. He was called Satya Pir.” Dinesh Chandra Sen, History of Bengali Language and Literature (Calcutta: University of Calcutta, 1954), 677.

36. Ibid., 677–78. Sen, Folk Literature, 100–101. Sarat Chandra Mitra, “On the Worship of the Deity Satyanarayana in Northern India,” Journal of the Anthropological Society of Bombay 9, no. 7 (1919): 768–77.

37. Pyrard, Voyage, 2: 336.

38. See n. 35 above.

39. Zbavitel, Bengali Folk-Ballads, 133.

40. Haji Shariat Allah was described on his tombstone as a “defender of religion against the menaces of the Shi‘ahs and the disbelievers and against all misguidance, valiant fighter for righteousness against all falsehood and vanity, deliverer of Islam (which) was covered by darkness like the sun enveloped in clouds.” Muin-ud-Din Ahmad Khan, History of the Fara‘idi Movement in Bengal (1818–1906) (Karachi: Pakistan Historical Society, 1965), p. xxxiii.

41. Cited in ibid., 8.

42. Wise, “Muhammadans of Eastern Bengal,” 61.

43. Risley Collection, “Reports on the Religious and Social Divisions amongst the Mahomedans of Bengal” (India Office Library, London, Eur. MSS. No. E 295), 9: 417.

44. Webster, Eastern Bengal and Assam District Gazetteers: Noakhali, 39.

45. A. K. Nazmul Karim, Changing Society in India and Pakistan: A Study in Social Change and Social Stratification (Dacca: Oxford University Press, 1956), 132.

46. For example, Muslim men were urged to grow long beards, women were discouraged from wearing saris or using henna to stain their feet and nails, and all were urged to eat grasshoppers on the grounds that Arabs ate locusts. See Wise, “The Muhammadans,” 56; The Cambridge History of Islam, ed. P. M. Holt et al. (Cambridge: Cambridge University Press, 1970), 2: 77.

47. H. H. Risley, Tribes and Castes of Bengal (reprint, Calcutta: Firma Mukho-padhyay, 1981), 1: xxx.

48. For a differing view, see Bruce B. Lawrence, Defenders of God (New York: Harper & Row, 1989), in which the author argues for the “modernness” of Islamic reform movements.

49. “The decisive consideration was and remains: who is deemed to exert the stronger influence on the individual in his everyday life, the theoretically supreme god or the lower spirits and demons? The process of rationalization (ratio) favored the primacy of universal gods; and every consistent crystallization of a pantheon followed systematic rational principles to some degree, since it was always influenced by professional sacerdotal rationalism or by the rational striving for order on the part of secular individuals.” Max Weber, The Sociology of Religion, trans. Ephraim Fischoff (Boston: Beacon Press, 1964), 20, 22.

50. See J. T. Milik, “Inscriptions grecques et nabatéennes de Rawwafah,” appended to P. J. Parr, G. L. Harding, and J. E. Dayton, “Preliminary Survey in N. W. Arabia, 1968,” Bulletin of the Institute of Archaeology 10 (1971): 54–58.

51. The Ecclesiastical History of Sozomen, Comprising a History of the Church from A.D. 324 to A.D. 440, trans. Edward Walford (London: H. G. Bohn, 1855), 309–10.

52. Bishop Sebeos, Histoire d’Héraclius, trans. F. Macler (Paris: Imprimerie nationale, 1904), 94–96. Cited in Patricia Crone and Michael Cook, Hagarism: The Making of the Islamic World (Cambridge: Cambridge University Press, 1977), 6–7.

53. Qur’an 36:23, 43:86, 53:26.

54. Qur’an, 53:23.

55. Syed Sajjad Husain, ed. and trans., A Descriptive Catalogue of Bengali Manuscripts in Munshi Abdul Karim’s Collection, by Munshi Abdul Karim and Ahmad Sharif (Dacca: Asiatic Society of Pakistan, 1960), xxiv.

56. Gordon Darnell Newby, The Making of the Last Prophet: A Reconstruction of the Earliest Biography of Muhammad (Columbia: University of South Carolina Press, 1989), 7, 3.

57. Cornell Fleischer made similar observations respecting the literary careers of the historians Ibn Khaldun (d. 1406) and Mustafa ‘Ali (d. 1600), both of whom were greatly influenced by the frontier environments in which they lived and wrote. See Fleischer, “Royal Authority, Dynastic Cyclism, and ‘Ibn Khaldunism’ in Sixteenth-Century Ottoman Letters,” Journal of Asian and African Studies 18, nos. 3–4 (1983): 216.

58. Richard Bulliet, Conversion to Islam in the Medieval Period: An Essay in Quantitative History (Cambridge, Mass., Harvard University Press, 1979), 44.

59. Newby, Making of the Last Prophet, 10. According to the early Muslim scholar al-Bukhari, in Muhammad’s own day “Jews used to read the Torah in Hebrew and interpret [Arabic fassara] it to the people of Islam in Arabic.” Quoted in ibid., 12.

60. See ibid., 8–14.

61. Although the identification of the Arabic nabī with the Sanskrit avatār was not unique to Saiyid Sultan, it is very rarely met with in the history of the Islamic encounter with Indian civilization. The earliest incidence of such an identification appears on coins minted by Sultan Mahmud of Ghazni in Lahore in 1027 and 1028 (A.H. 418 and 419). On these coins the sultan inscribed a Sanskrit translation of the Islamic confession of faith in which Muhammad was called the avatār of God, translated as Avyaktam, “the Unmanifested.” This bold experiment at translating the most fundamental Islamic ideas into the Indian religious universe is unique in the history of Indian coinage, however, and it was never tried again. See D. C. Sircar, Studies in Indian Coins (Delhi: Motilal Banarsidass, 1968), 19. Isma‘ili traditions dating perhaps to the fifteenth century also made use of the avatār notion, identifying the Prophet Muhammad’s son-in-law ‘Ali with the tenth of Vishnu’s avatārs. See Azim Nanji, The Nizari Isma‘ili Tradition in the Indo-Pakistan Subcontinent (Delmar, N. Y.: Caravan Books, 1978), 110–20.

62. Saiyid Sultan, Nabī-Baṃśa, 1: 2.

63. Ibid., 3, 4, 6.

64. Ibid., 24–25.

65. Ibid., 11–12.

66. Ibid., 38.

67. Ibid., 25.

68. Ibid., 181.

69. Ibid., 242–43.

70. Ibid., 244.

71. Ibid., 252.

72. Ibid., 132–33.

73. This fundamental point is repeatedly proclaimed in the Qur’an itself—e.g., “And We have sent down on thee the Book making clear everything, and as a guidance and a mercy, and as good tidings to those who surrender.” The Koran Interpreted, trans. Arthur J. Arberry (New York: Macmillan, 1970), Sura 16:91.

74. Peel, “Syncretism,” 139–40. See also Jack Goody, The Logic of Writing and the Organization of Society (Cambridge: Cambridge University Press, 1986), 1–44.

75. In the case of Qur’an schools (madrasas), of course, the government’s support of literate Islam was explicit. In one 1732 sanad, for example, 390 acres of forest in Sylhet were granted in the expectation that the land, once cleared and cultivated, would support the students of a Qur’an school (kharj-i ṭālibān-i madrasa). “Register of Sanads,” Sylhet District Collectorate Record Room, 20, no. 932.

76. Losty, Art of the Book, 10–12, 113.

77. Rockhill, “Notes on the Relations,” 440. See also S. A.K. Ghori and A. Rahman, “Paper Technology in Medieval India,” Indian Journal of the History of Science 2 (November 1966): 136. For a history of the diffusion and technology of papermaking in the premodern Muslim world generally, see Johannes Pedersen, The Arabic Book, trans. Geoffrey French, ed. Robert Hillenbrand (Princeton: Princeton University Press, 1984), 59–67.

78. Mukundaram, Kavikaṅkaṇa Caṇḍī, 346.

79. Muhammad Khan, Maqtul Husain, Dhaka Museum MS. No. 2826, Acc. No. 6634.

80. For example, Shaikh Muttalib, Kifayat al-Mussalin, Dhaka Museum MS. No. 2825, composed in 1856.

81. As recently as 1967, a judge on the Supreme Court of Pakistan suggested that, in the interests of fostering unity between Pakistan’s eastern and western wings, Bengali, like Urdu, be rendered in the Arabic script. But this was resisted by the Bengalis themselves, who several years later wrested their independence from Pakistan. See Golam Morshed, “Cultural Identity of Former East Pakistan and Conflicting State Policies,” Journal of the Institute of Bangladesh Studies 3 (1978): 69.

82. Dinesh Chandra Sen, History of Bengali Language and Literature, 674.

83. Zbavitel, Bengali Literature, 161–62.

84. Karim, Social History of the Muslims in Bengal, 171. Karim’s translation. Cf. Viyaya Gupta, Padma Purāṇa (Calcutta: University of Calcutta, 1962), 140.

85. For an excellent summary of how the belief in bhūt permeates contemporary Bengali Muslim life, especially among women in rural society, see Thérèse Blanchet, Meanings and Rituals of Birth in Rural Bangladesh (Dhaka: University Press, 1984), 50–63.

86. Toward the end of the sixteenth century, Mukundaram wrote that mullās also performed life-cycle rituals for Muslims and collected from their clients regular payments for their services, paralleling in this respect the functions of Brahmans in Hindu society: “The Mullās perform the ceremony of the nikā [wedding] and get a reward of four annas and bless the couple by reading the Kalimah. He [the mullā] takes a sharp knife, kills the fowl, and gets a reward of ten gaṇḍās of cowri. For butchering a she-goat, the mullā gets six buṛis of cowri, and also the head of the animal killed.” Mukundaram, Kavikaṅkaṇa Caṇḍī, 344. Karim, Social History of the Muslims in Bengal, 172. Karim’s translation.

87. The report read: “A circle is described round the affected person with a stick, charms being recited at the same time; a verse from the Koran is read over a quantity of mustard oil, and a wick smeared with the oil thus charmed is burnt and the smoke out of it is thrust into the nostrils of the affected person. To the bhut this is unbearable and so it speaks through the possessed. The sorcerer (Ojha) then asks the bhut his name and whereabouts and how he came to have possession of the person, and he gives replies. The sorcerer then makes the bhut promise in the name of Suleman Badshah, the sovereign of bhut, that it would never more enter the body and then, when he is satisfied, he cuts the magic circle and the bhut goes away.” Bhattacharjee, “Folkcustom and Folklore,” 19–20. Relevant in this regard are Thérèse Blanchet’s remarks on bhūt as a fundamental category in Bengali folk belief: “Bhut are said to originate from the north, in the mountains. Bhut used to be masters of the land, before, when there was no civilization. But then came the ‘great religions.’ These had powerful gods who could dominate the bhut. Their learned men, or priests, could pronounce the words that keep bhut at bay. Indeed both Sanskrit and Arabic incantations (montro) are effective against bhut. Their written words have the same effect (i.e. tabij). Civilized ways did not penetrate everywhere however and bhut, even after the advent of the ‘great religions’ still played tricks on men, but even more on women. Bhut are believed to be numerous in the jungle or where habitations are sparse (on the chor). They are few in towns.” Blanchet, Meanings and Rituals, 54.

88. In 1986 it was reported that an ojhā had healed a mentally deranged Chakma woman by “using all the abusive words to the evil spirit who was believed to have possessed this village woman. His whole body began to shake and tremble. He then recited some incantations taken from the religious books of the Muslims and the Hindu community. He was using names of Allah, Hari, Muhammad, Krishna, Kalima, Fatima, and deities in Kamakhya, a place in Assam noted for its magic, and made a circle around the woman. He then made mild strokes on the body and poured hot mustard oil into the nostrils of the patient. The attendant of the Ojha was beating a tin can. The Ojha then burnt incense and asked the patient to bend over it and to inhale the smoke.” When the ceremony was concluded and the woman fully recovered, the ojhā asked the patient’s guardians to make a sacramental sacrifice at the mazār, or shrine, of a certain pīr. Anwarul Karim, “Shamanism in Bangladesh,” Bangladesh Observer, February 7, 1986. The last detail is significant, as it shows that the healing ceremony was specifically linked to the agency of a local Muslim saint, thereby associating Islam in the villagers’ minds with physical or psychic health.

89. Buchanan, Geographical, Statistical, and Historical Description, 92.

90. Bhattacharjee, “Folkcustom and Folklore,” 139. Emphasis mine.

91. The Glorious Koran: A Bilingual Edition with English Translation, Introduction and Notes, trans. Marmaduke Pickthall (London: George Allen & Unwin, 1976), Sura 4:34.

92. Marshall Hodgson, Venture of Islam, 1: 342. The wearing of the veil would have inhibited the sort of manual labor associated with the lower classes, in much the same way as did the practice of footbinding among upper-class women of China between the eleventh and twentieth centuries. The wearing of the veil also appears to find support in the Qur’an (Sura 24:31, 33:59).

93. Ibn Battuta: Travels in Asia and Africa, trans. H. A.R. Gibb (London: Routledge & Kegan Paul, 1929), 124, 146–47, 243–44, 321.

94. Ibid., 146–47.

95. Rockhill, “Notes on the Relations,” 443. Hou Hsien would have seen a good bit of the interior delta, as he traveled from Chittagong to Pandua via Sonargaon.

96. Abu’l-fazl, ā’īn-i Akbarī (Lucknow ed.), 2: 75, 76. English trans., 2: 132, 134.

97. Saiyid Sultan, Nabī-Baṃśa, 1: 107–8.

98. Ibid., 110.

99. Dinesh Chandra Sen, trans. and ed., Eastern Bengal Ballads, 1, pt. 1: 307–8.

100. Analysis of how the purdah system restrains women from participating in economic activities outside the house is found in Mahmuda Islam, “Social Norms and Institutions,” in Situation of Women in Bangladesh, ed. Mahmuda Islam et al. (Dacca: UNICEF, 1979), 225–64.

101. “Traditional village religious figures and specially esteemed preachers, who are brought to the village from time to time, always stress the religious obligation men have to keep their women inside.” Tahrunnessa A. Abdullah and Sondra A. Zeidenstein, Village Women of Bangladesh: Prospects for Change (New York: Pergamon Press, 1982), 56. For a study of the rigidity with which the purdah system is practiced by women from different socioeconomic classes, see Florence McCarthy, The Status and Condition of Rural Women in Bangladesh (Dacca: Ministry of Agriculture and Forests, 1978).

102. Using indices such as life expectancy, literacy rates, undernourishment, and mortality ratios, the economist Pranab Bardhan has noted that the status of females is higher in West Bengal, Assam, and Orissa than in Punjab and Rajasthan. Bardhan explained this difference not in terms of culture or religion, but in terms of the different economy of eastern India as opposed to northern and northwestern India. In the east (and also the south), he argued, “the predominant crop is paddy which—unlike wheat and other dry-region crops—tends to be relatively intensive in female labour. Transplantation of paddy is an exclusively female job in the paddy areas; besides, female labour plays a very important role in weeding, harvesting and threshing of paddy.” From this Bardhan concluded that “the social and economic value of the woman is likely to be related to the ecological conditions and production relations of a particular region.” It is true that in predominantly Hindu West Bengal women are active in transplanting, weeding, and harvesting, and that their social status is higher there than in northern or northwestern India. But in predominantly Muslim East Bengal, where the economic base is identical with that of the western delta, women do not generally take part in these operations. In other words, contrary to Bardhan’s hypothesis, the region’s economic base does not dictate the social value of women. On the other hand one important cultural variable, Islam, does correlate with relatively low female status in the Bengal delta, as it does, too, in northern and northwestern India. See Pranab K. Bardhan, “On Life and Death Questions,” Economic and Political Weekly, August 1974, 1302–4.

103. Lucien Bernot, Les Paysans arakanais du Pakistan oriental (Paris: Mouton, 1967), 249–64. As the author notes: “Après quelques jours passés dans les plaines des environs de Chittagong, en milieu bengali, où les femmes, surtout les musulmanes, sont invisibles, le voyageur européen est toujours surpris de rencontrer dès les premiers villages des collines, des groupes de femmes bavardes, interrompant les conversations de leurs époux, nullement effrayées par la venue de l’étranger, lui demandant où il va, quel est son nom, s’il parle marma, lui offrant spontanément un cigare ou des fruits, de l’eau fraîche ou du bétel.” Ibid., 16.

104. Buchanan, “Account of a Journey,” 20.

105. Ibid., 36.

106. Ibid., 62–63.

107. Ibid., 67.

108. Around the turn of the twentieth century a Chittagong beggar, in explaining why he had renounced his earlier conversion to Christianity, declared that “a local Krishna is better than a foreign Krishna.” Risley Collection, “Reports on the Religious and Social Divisions,” 9: 286–87.