The fieldwork upon which this paper is based was carried out from Jan. 1970 through June 1971, under a Foreign Area Fellowship and while I was a Danforth Foundation Graduate Fellow. I am genuinely grateful for this support.

The approach taken here is strongly influenced by the recent writings of McKim Marriott, Ronald Inden, and Ralph Nicholas, cited below. The work of David Schneider (n. 8 below) and Stephen Barnett (“The Structural Position of a South Indian Caste,” Ph.D. dissertation, Univ. of Chicago, 1970; “The Process of Withdrawal in a South Indian Caste,” in Entrepreneurship and Modernization of Occupational Cultures in South Asia, Milton Singer, ed. [Durham, N. C: Duke Univ. Program in Comparative Studies on South Asia, 1973], pp. 179–204; “Urban Is As Urban Does: Two Incidents on One Street in Madras City, South India,” Urban Anthropology, II, 2 [1973], pp. 129–60) has been similarly influential. This paper has also benefited from a careful reading by McKim Marriott, for which I am thankful.

¹ “Status Evaluation in Hindu Caste System,” Journal of the Royal Anthropological Institute, XXXIV (1954), pp. 45–65Google Scholar.

² “Interactional and Attributional Theories of-Caste Ranking,” Man in India, XXXIX (1959), pp. 92–107Google Scholar; “Caste Ranking and Food Transactions: A Matrix Analysis,” in Structure and Change in Indian Society, Singer, Milton & Cohn, Bernard S. (eds.), Chicago: Aldine, 1968Google Scholar.

³ “The Conception of Kingship in Ancient India,” Contributions to Indian Sociology [hereafter CIS], 6 (1962), pp. 48–77; “A Fundamental Problem in the Sociology of Caste,” CIS, 9 (1966), pp. 17–32; Homo Hierarchies: The Caste System and Its Implications [hereafter HH], Chicago: Univ. of Chicago Press, 1970; French ed., 1967.

⁴ For a critique of attributional theories, see Marriott 1959 (n. 2 above); of interactional theories, see Dumont 1970 (n. 3 above), and Stephen A. Tyler (India: An Anthropological Perspective, Pacific Palisades, Calif.: Goodyear, 1973); of the purity-plus-power argument, see the review symposium on HH, organized by T.N. Madan (“On the Nature of Caste in India,” CIS, n.s. 5 [1971], pp. 1–81), also Nur Yalman (“De Tocqueville in India: An Essay on the Caste System” [review article on the French ed. of HH], Man, n.. IV, 1 [1969], pp. 123–31), Marriott (review of the French ed. of HH], American Anthropologist, LXXI [1969], pp. 1166–75), R.S. Khare (“Encompassing and Encompassed: A Deductive Theory of Caste System” [review article on HH], Journal of Asian Studies, XXX, 4 [1971], pp. 859–68), and Edmund Leach (“Hierarchical Man: Louis Dumont and His Critics,” South Asian Review, IV, 3 [1971], pp. 233–37).

⁵ Marriage and Rank in Bengali Culture: A History of Caste and Clan in Middle Period Bengal, Berkeley: Univ. of California Press, 1975.

⁶ “A Cultural Analysis of Bengali Kinship,” in Prelude to Crisis: Bengal and Bengal Studies in 1970, Peter J. Bertocci (ed.), East Lansing: Michigan State Univ. Asian Studies Center, Occasional Paper no. 18, South Asian Series, 1972; Kinship in Bengali Culture, Chicago: Univ. of Chicago Press, forthcoming.

⁷ “Towards an Ethnosociology of Hindu Caste Systems,” IXth Int'l. Congress of Anthropological and Ethnological Sciences, 1973, n. 2206; “Caste Systems,” Encyclopaedia Britannica, 1974.

⁸ Schneider, David, American Kinship: A Cultural Account, Englewood Cliffs, N.J.: Prentice-Hall, 1968Google Scholar.

⁹ Strictly speaking, the ethnographic base for generalizing from Hindu villagers of the Narayanghar kingdom to Hindus throughout West Bengal State is limited. The generalizations are ventured nevertheless because, as Geertz, Clifford, in The Inter-pretation of Cultures (New York: Basic Books, 1973), p. 22Google Scholar, has pointed out, “The locus of [an- thropological] study is not the object of study. Anthropologists don't study villages (tribes, towns, neighborhoods… ); they study in villages.” Their data is extremely specific and circumstantial, but the importance of tha t data, at least potentially, is not so much where it came from, as how far it can be taken by analysis and interpretation. This said, it must be added that frequent checks on the information gathered within villages of the Narayanghar kingdom were made with other Bengalis elsewhere; thus, my generalizations are not blindly founded. And, as will be noted elsewhere, at least some of these generalizations also find support outside of West Bengal.

¹⁰ For a description of the major and minor worlds—their placement in the universe, the physical features of each, and the life forms that occupy them—see the Srimad Bhagavata Mahapurana [hereafter SBM], V, 16–26.

¹¹ The sanskritic terms may be more familiar to non-Bengalists: jangama, “moving beings,” ineluding the whole animal world; sthāvara, “stationary beings,” including the whole plant world; and khanija, “pit-born,” or those beings taken from the earth.

¹² Creation is not here considered the origination of something new, something different and unlike Brahma. Rather, in creation, all that already exists is made manifest and apprehensible as various emanations or evolutes of Brahma. For a detailed description of the creation process and the order in which all substances and categories of the phenomenal world were made manifest, see SBM, II, 5; also III, 26. A more abbreviated account is also given in the Manu Dharmasastra [hereafter MD], I.

¹³ This point is made in SBM, XI, 25:29 (C. L. Goswami, trans., Gorakhpur: Motilal Jahan, 1971) in the following passage, which summarizes a rather lengthy description of the operation of the three gun: (In this way) substance (food as classified in verse 28 above), place of abode (as mentioned in verse 25), fruit (in the shape of joy referred to in verse 29), time (as indicated in verses 13 to 16), knowledge (referred to in verse 24), action (classified in verse 23), the agent (as mentioned in verse 26), faith (referred to in verse 27), state of consciousness (as told in verse 20), form (state of existence as classified in verse 21), and final destiny (as referred to in verse 22)—everything is as a matter of fact constituted of the three Gunas alone.

¹⁴ For a further description of the three gun, see SBM XI, 25; also the Bhagavad Gita, XIV: 15–21, in which sattva, rajas, and tamas are equated with goodness, passion, and darkness, or essence, energy, and inertia, respectively. Bühler's 1886 translation of MD (The Laws of Manu, Delhi: Motilal Banarsidass, 1964), unfortunately, does not speak of the three gun as such, preferring to translate these terms as the “three qualities.” Sherry Ortner's paper “Sherpa Purity” (American Anthropologist, LXXV, 1 [1973], pp. 49–63) may be read for a description of the Tibetan equivalents to the three gun), which are analyzed very differently.

¹⁶ In reading the paradigm, it may be asked how rajogun can activate sattvaguri and tamogun when it is itself suppressed. It is not that two different conceptions of “activity” are involved here. Rather, Bengalis hold that, like Brahma, all life forms are ever animate—only more or less so, according to their likeness to Brahma. Thus, to say that rajogun is suppressed is to describe a relative, not an absolute, disposition. Even when relatively suppressed, rajogun is still active and can still activate the other gun.

¹⁶ Goswami trans, (n. 13 above). For an analogous listing of the “effects” of the three gun working in various combinations, see SBM, XI, 25:8–10, 12–17, 22–23, 25–26.

¹⁷ “A similar account is also found in SBM, II, 5:34–36.

¹⁸ A description of the capacities of each of the four varna is given at various points in MD; see especially X, 74–100.

¹⁹ Bengali villagers are generally uncertain as to why their own life experiences should vary from the account given in respected texts. Bengali scholars generally argue that the Kshatriya and Vaisya varna in Bengal have not been eliminated forever; it is only that in this Kali Yuga they remain unnoticed or degraded to the rank of Sudra for failure to observe their appropriate behavioral codes (see Dutt, N. K., Origin and Growth of Caste in India [Calcutta: Firma K. L. Mukhopadhyay, 1969], vol. II, pp. 78–90)Google Scholar.

²⁰ Both the terms rank and standing affirm the premise of inequality, and designate relationships of relative highness and lowness. But as used here, rank is the unmarked term while standing is applied restrictively to those units that cannot be ordered in comprehensive rankings. Pocock, David (Kanbi and Patidar, Oxford: Clarendon Press, 1972)Google Scholar makes a similar distinction between status and standing, where status is used to refer to the ranked relationship between castes and standing is used to refer to the ranked relationship between persons, families, or descent groupings within a caste.

²¹ See Inden and Nicholas 1972 (n. 6 above) for a more developed statement about the changing memberships of Bengali kul, and of the changing features of the women who pass between them; also Inden and Nicholas forthcoming (n. 6 above). David, Kenneth (“Until Marriage Do Us Part: A Cultural Account of Jaffna Tamil Categories for Kinsmen,” Man, n.s. VIII, 4 [1973], pp. 521–35)CrossRefGoogle Scholar makes a similar argument for the changing features of women among Jaffna Tamils.

²² Inden and Nicholas forthcoming (n. 6 above).

²³ Other life forms possess a bhōg deha, a body with which to sustain (bhōg) themselves only. In the instance of animals, plants, and objects, the possession of a bhōg deha means these life forms undergo transformations only through the direct intervention of the deities. In the instance of deities, it means that they are forever subject to ruling over the worlds of the universe, and especially over the affairs of persons on earth. Man, by contrast, is responsible for his own activities and the transformations experienced as a result of those activities.

²⁴ In SBM, these alternate sets of activities and the transformations resulting from each are referred to as the “three destinies” (III, 22:35). A description of the three destinies is also given in the Bhagavad Gita (XIV:I8) as follows: “Upward [rebirth as a god] is the path of those who abide in Goodness, in the middle [rebirth as a human] stand the men of Passion. Stuck in the modes of the vilest constituent the men of Darkness go below [rebirth as an animal].” (R.C. Zaehner, trans., London: Oxford Univ. Press, 1969)

²⁵ Note 1 above.

²⁶ HH (n. 3 above), pp. 46–61.

²⁷ Marriott and Inden 1973 (n. 7 above), p. 9.

²⁸ The discussion of marriage forms that follows reflects the insight of Inden's 1975 (n. 5 above) analysis of marriage and rank among Bengali Brahmans and Kayasthas, particularly his blending of Dharmasastra and present custom.

²⁹ “In all castes (varna) those (children) only which are begotten in the direct order on wedded wives, equal (in caste and married as) virgins, are to be considered as belonging to the same caste (as their fathers).” (MD, Bühler, trans, [n. 14 above], X:5)

³⁰ “Sons, begotten by twice-born men on wives of the next lower caste, they declare to be similar (to their father, but) blamed on account of the fault (inherent) in their mothers. Such is the eternal law concerning (children) born of wives one degree lower (than their husbands).” (Ibid., X:7–8)

³¹ “Twice-born men who, in their folly, wed wives of the low (Sudra) caste, soon degrade their families and their children to the state of Sudras.” (Ibid., III:14)

³² Inden 1975 (n. 5 above), pp. 60–82.

³³ Inden and Nicholas 1972 (n. 6 above), p. 5.

³⁴ The same is also true of Jaffna Tamils (David [n. 21 above], p. 522).

³⁵ For somewhat different listings of hot and cold foods in Tamilnadu and Mahdhya Pradesh, see Beck, Brenda (“Colour and Heat in South Indian Ritual,” Man, n.s. IV, 4 [1969], pp. 553–73)CrossRefGoogle Scholar and Babb, Lawrence A. (“Heat and Control in Chhattisgarhi Ritual,” Eastern Anthropologist, XXVI, 1 [1973]. PP. 11–28)Google Scholar.

³⁶ The following discussion of food exchanges reflects the work of Marriott 1968 (n. 2 above) and 1976 (“Hindu Transactions: Diversity Without Dualism,” in Transactions and Meaning: Directions in the Anthropology of Exchange and Symbolic Behavior, Bruce Kapferer, ed., [Philadelphia: Institute for the Study of Human Issues, 1976]), and Marriott and Inden 1973 and 1974 (n. 7 above).

³⁷ 1968 (n. 2 above).

³⁸ Marriott has suggested (personal communication, 1975) that jal khābār might refer to food that is handled much as water is handled, thus “water-like food.”

³⁹ The ranking of individuals is discussed in my “Rank and Rivalry in Rural West Bengal” (Ph.D. dissertation, the University of Chicago, 1975).

⁴⁰ Description and Comparison in Cultural Anthropology, Chicago: Aldine, 1970Google Scholar.