14

Food and Dietary Rules

abhakṣya, abhojya

Patrick Olivelle

It has long being noted that food—whether vegetable or animal—apart from being a biological necessity, has deep social, cultural, and religious meanings. Anthropologists have long noted that among the wide variety of edible animals available to a given social group, only a small fraction is actually consumed. Prohibited foods, whether by injunction or by custom, far outnumber permitted ones in most societies. The French anthropologist Claude Lévi-Strauss (1963 : 89) famously said that animals are “good to think.” The pioneering studies of the British anthropologists, Edmund Leach (1964 ) and Mary Douglas (1966 , 200) have illuminated the ways in which food prohibitions in various societies and religions are related to their social organizations and cosmological categories. The Sri Lankan anthropologist Stanley Tambiah (1969 : 423–4) recapitulates their conclusions:

Lévi-Strauss (1966 : 104) has formulated a view…that “there is an empirical connection between marriage rules and eating prohibitions.”…The theory [of Mary Douglas] argues that dietary prohibitions make sense in relation to a systematic ordering of ideas (a classification system) as exemplified for example by the abominations of Leviticus. Leach, partly building on Douglas,…demonstrates for the English and the Kachin the correspondence between three scales of social distance from Ego, pertaining respectively to marriage and sex relationships, spatial categories, and edibility of animals.

Thus, the dietary rules of the Dharmaśāstras discussed in this chapter are significant for the study of many aspects of ancient Indian society, including caste and marriage. Within the confines of this chapter, however, I will outline the Dharmaśāstric discussions of food and dietary rules, referring the reader to my earlier studies for analyses of their sociological significance. 1

Although rules about what can and cannot be eaten pertain to both vegetables and animals, it is to animals that dietary rules mostly refer. Therefore, I will take up rules on eating animals first. These rules often follow animal classifications based on where they live, what they eat, and how their teeth and feet are structured.

Following the division of space into land, water, and air, we have already in the Vedic texts a threefold classification of land animals, fish, and birds. Within the first and the third of these, there is a further distinction between village or domestic animals (grāmya and paśu ) and the wild (āraṇyaka ) often with the generic designation of mṛga . There is a twofold classification of land animals and birds based on their diet: the herbivores and the carnivores. The most significant classification with regard to animal prohibitions pertains to the structure of their teeth and feet. In terms of teeth, there is a twofold classification of land animals into those with incisor teeth in both jaws (ubhayotodat ) and those with incisors only in a single (lower) jaw (anyatodat or ekatodat ). We find this classification already in the Puruṣa hymn of the Ṛg Veda (10.90.10): “From it [the sacrifice] were born horses, as also whatsoever have incisor teeth in both jaws. Cows were born from it. From it were born goats and sheep.” 2 Even though the text does not mention the class with incisors in the both jaws, the anyatodat , it is clearly implicit; we have here a division of animals into two classes—those with two rows of incisor teeth headed by the horse and those with a single row of incisors, whose major representatives are cows, goats, and sheep, the most common farm animals. Again, although left unstated, the vast numbers of carnivorous animals with paws rather than hooves also belong to the category of animals with two rows of incisors. In terms of their feet, animals have either paws with five nails (pañcanakha ) or hooves (śapha, khura ). The latter is further subdivided into those with two (cloven) hooves (dviśapha, dvikhura ) and those with a single unparted hoof (ekaśapha, ekakhura ). Some of these categories overlap. For example, animals with incisors in both jaws fall under the categories of those with five nails (pañcanakha ) and those with single hooves (ekaśapha ). According to their foot structure, birds are divided into those with webbed feet (jālapāda ) and those without, the latter being further divided into those that eat by scratching the earth with their feet (viṣkira or ground feeders) and those that eat by pecking (pratuda ). 3

These classifications, especially in the case of land animals, play a central role in dietary prescriptions. Among land animals, three classes are prohibited: animals with five nails (pañcanakha ), 4 those with single or uncloven hooves (ekaśapha ), and those with incisor teeth on both jaws (ubhayatodat ). Among birds, those with webbed feet are generally prohibited, while those of the scratching and pecking varieties are generally permitted. Among both animals and birds, those that are carnivorous (kravyād ) are forbidden. In terms of animals falling within these categories, the single-hooved are few: horse, donkey, and mule. The ideal meat comes from the cloven-footed group: cattle, goats, and sheep, among the domesticated animals, and deer and the antelope family among the wild.

There is another category of animal and bird that is forbidden, and that is the “village” (grāmya ). This category is distinct from the “domesticated” because farm animals called paśu, namely, cattle, goats, and sheep, constitute the three paradigmatic permitted animals. The category grāmya appear to include animals and birds whose habitat is the village, such as the village fowl and pig, which are forbidden, even though the former falling under the “scratching” (viṣkira ) category and the latter under both the cloven hoof (dviśapha ) and the single row of incisor teeth (ekatodat ) should have been permitted. In fact, their wild (araṇya ) counterparts, the jungle fowl and wild hog, are indeed permitted. The grāmya animals and birds are those that have chosen to live in the village, whereas paśu are domesticated animals that people rear in farms.

Other broad categories of land animals and birds prohibited are the solitary (ekacara ), the unknown (ajñāta ), and those that are not specified (anirdiṣṭa ), probably referring to animals and birds not listed specifically as permitted. At least in the case of birds, we have in addition the prohibition of nocturnal birds and in some sources birds with red feet or beaks.

Even though, as we saw, some categories of birds are prohibited, it is more common for sources simply to list the birds that are forbidden. Thus, we have the following given by Vasiṣṭha (VaDh 14.48):

Among birds, the following are forbidden: birds that agitate, scratch with their feet or are web-footed; Kalaviṅka sparrow, Plava heron, Haṃsa goose, Cakravāka goose, Bhāsa vulture, crow, Pārāvata dove, Kurara osprey, and Sāraṅga cuckoo; white dove, Krauñca crane, Krakara partridge, Kaṅka heron, vulture, falcon, Baka egret, and Balāka ibis; Madgu cormorant, Ṭiṭṭibha sandpiper, Māndhāla flying fox, and nocturnal birds; woodpecker, sparrow, Railātakā bird, Hārīta pigeon, Khañjarīṭa wagtail, village cock, parrot, and Śārika starling; Kokila cuckoo and carnivorous birds; and those living in villages.

The problems inherent in bird classifications, however, are exemplified in the way later authorities, such as Vasiṣṭha, Manu, and Yājñavalkya, handle the two major categories of birds: those that scatter (viṣkira ) and those that peck (pratuda ). All of the more ancient sources declare that these birds are permitted, whereas these later authors state just the opposite. 5 The reason for this reversal is unclear, but it may well have to do with a misunderstanding or a fresh interpretation of these classes. At least some commentators, for example, take the “peckers” to be carnivores, such as hawks and woodpeckers.

When it comes to fish, our sources, unlike the Hebrew authors of the Deuteronomy and Leviticus, totally give up on broad classifications, resorting instead to listing individual fish. Here also we detect an interesting change from the older to the younger sources. The older sources appear to assume that all fish, except those explicitly forbidden, may be eaten. The only broad category of forbidden fish in them is the rather vague one of misshapen or grotesque fish (vikṛta ). Manu and later writers, on the other hand, do not appear to favor the eating of fish. Indeed, in them, the process is reversed; all fish except those listed are forbidden. Indeed, Manu (5.15) says, “A man who eats the meat of some animal is called ‘eater of that animal’s meat’, whereas a fish-eater is an ‘eater of every animal’s meat’. Therefore, he should avoid fish.” 6 Yet, possibly bowing to tradition, even Manu (5.16) permits certain kinds of fish, especially in ancestral rituals: “The Pāṭhīna and the Rohita fish may be eaten when they are used in an offering to gods or ancestors; Rājīva, Siṃhatuṇḍa, and Saśalka fish may be eaten at any time.” One possible reason, besides the geographical locations of these authors, for this shift in attitude with regard to fish may have been the view that fish are by nature carnivores, as exemplified in well-known maxim of the fish, the bigger fish eating the smaller fish.

Milk is the most restrictive of all food items. The milk of most animals—women and other five-nailed animals, single-hooved animals such as mares, and even most animals with parted hooves, such as sheep and deer—is explicitly forbidden. The only permitted milk is that of cows, buffaloes, and goats. Even in their case, milk during the first ten days after giving birth, or of one whose calf has died is forbidden.

Prohibitions with regard to vegetables also involve individual plants rather than categories, and the lists of such prohibited plants differ from source to source. In general, mushrooms, onions, garlic, and leeks are forbidden, as are growths on tree stumps, red sap of trees, and anything that grows on filth or excrement. 7 Food prohibitions, however, are much more focused on animals than on vegetables.

The rules we have encountered thus far deal with what kinds of animals, birds, fish, milk, and vegetables a person may or may not eat. These refer to sources of food rather than to prepared food, about which there are also rules as to their suitability. Dharmaśāstras use two pairs of terms for these two kinds of rules regulating food: bhakṣya /abhakṣya (permitted and forbidden food) and bojya /abhojya (fit and unfit food). 8 The first pair refers to food sources that we have described above. Abhakṣya (“forbidden food”) refers to items of food, both animals and vegetables that are completely forbidden; they cannot be eaten except under the direst circumstances, such as when someone is dying of hunger. Generally, lists of forbidden food contain food sources rather than cooked food served at a meal. Thus, carnivorous animals, village pigs and fowls, web-footed birds, garlic, and the red resins of trees are all abhakṣya . Foods other than the abhakṣya would constitute permitted food (bhakṣya ). The term abhojya (“unfit food”), on the other hand, refers to food that is normally permitted but owing to some supervening circumstances, has become unfit to be eaten. Lists of unfit food items contain not food sources but food that is actually served at a meal. Thus, food contaminated by hair or insects, food touched by an impure man or woman, and food that has turned sour or stale are all abhojya . This term takes on a secondary meaning referring not directly to food but to “a person whose food is unfit to be eaten”; 9 for example, a person during the period of impurity following a death (āśauca ), a physician, or a hunter. Conversely, bhojya in the compound bhojyānna is a person from whom one may accept food. The Gautama Dharmasūtra (17.6–17.7), for example, states: “A man who looks after his animals or plows his fields, a friend of the family, his barber, and his personal servant—these are people whose food he may eat (bhojyānnāḥ ), as also a merchant who is not an artisan.”

When anthropologists working in the field describe food prohibitions and how they relate to caste divisions—that is, from whom can you receive food or water—they are dealing with abhojya . Given the importance of food transactions in modern Indian society, the abhojya kind of food regulation becomes significant. I will not address here issues relating to abhojya , however, because in the world of the dharma literature, there was little by way of restriction on food transactions between various social groups. Indeed, it was taken as a matter of course that a Brahman could eat at the house of even a Śūdra, entertain people of all classes, and have Śūdra cooks in his household.

The category of abhojya relates to the deeply transactional nature of food: food is a social good that is intended to be shared within and across various social boundaries. The sociologist R. S. Khare (1976a) has focused on the transactional nature of food and identified several “food cycles” in Indian culture, the most important of which are the socio-ritual and the cosmological (Olivelle 1995b ). Food, in the form of meat, vegetable products, and milk and milk products, such as ghee, constitute the major ritual offerings both in the Vedic sacrifices and in the later domestic rituals and temple offerings. The cosmology where gods provide rain for grain and plants to grow and humans provide food for the gods is the basic principle of do ut des at a cosmic level. This is extended in the śrāddha ancestral offerings, where the deceased ancestors are dependent on food and drink offered by their living relatives. At the social level, there are the food transactions among individuals exemplified in the rituals of hospitality toward guests and the offering of food to beggars and religious mendicants.

These food transactions are codified in the five great sacrifices (mahāyajña ) that are closely associated with the householder. 10 Except for the sacrifice to the Vedic seers consisting of reciting the Veda, all these involve the offering of food: to gods, to ancestors, to human beings, and to all beings (bali offerings to bhūtas ). The householder thus stands at the center of these cycles of food transactions. Vasiṣṭha (8.14–8.16) calls the householder the best of the four āśramas , because “as all living beings live dependent on their mothers, so all mendicants live dependent on the householder.”

Some interesting information regarding the prestige of various foods comes from the kinds of dishes offered to ancestors during a śrāddha ceremony. Different foods provide nourishment for the ancestors for varying lengths of time, the more prestigious (or nutritious?), the longer they can survive on it. The earliest record comes from Āpastamba (2.16.23–2.17.3):

The materials used in this rite are sesame and beans, rice and barley, water, roots, and fruits. When the food is made greasy, however, the gratification it gives the ancestors is more ample and lasts longer, as also when one gives righteously (dharma ) acquired wealth to a worthy person. With cow’s meat their gratification lasts for a year, and even longer than that with buffalo meat. This rule makes clear that the meat of domestic and wild animals is fit to be offered. With the meat of a rhinoceros offered on a rhinoceros skin, their gratification lasts an unlimited time, as also with the flesh of the Śatabali fish and the Vārdhrāṇasa crane.

The list of food at a śrāddha becomes longer and more complex in later texts. The long list with corresponding lengths of time of Manu exemplifies this trend:

I will explain exhaustively the types of sacrificial food that are efficacious for a long time and those that are efficacious in perpetuity, when they are offered to the ancestors according to rule. By offering sesame seeds, rice, barley, beans, water, roots, and fruits according to rule, ancestors of men rejoice for one month; by offering fish, for two months; by offering the meat of the common deer, for three months; by offering sheep meat, for four months; by offering here the meat of birds, for five months; by offering goat meat, for six months; by offering the meat of the spotted deer, for seven months; by offering the meat of the Eṇa antelope, for eight months; by offering the meat of the Ruru deer, for nine months; by offering boar or buffalo meat, they are satisfied for ten months; by offering rabbit or turtle meat, for eleven months; and by offering beef, milk, or milk-rice, for one year. The satisfaction from the meat of a Vārdhrīṇasa hornbill lasts for twelve years. The Kālaśāka herb, Mahāśalka crustacean, the meat of the rhinoceros and the red goat, and honey, as well as every type of sage’s food 11 are efficacious in perpetuity.

The lists of food at a śrāddha, as well as the lists of permitted and forbidden animals discussed earlier, brings up the topic of meat eating in ancient India, a topic that has drawn attention and controversy in contemporary India. The history of vegetarianism and noninjury to living beings (ahiṃsā ) is complex, and I do not have the space to discuss it here. 12 The repugnance toward the killing of animals for food is clearly old as already noted in Aśoka’s inscriptions from the middle of the third century bce . The early Dharmasūtras present the eating of meat as a common and noncontroversial practice, even though ahiṃsā or not killing living beings is given among the virtues that individuals should cultivate. Yet this virtue is not linked to refraining from killing animals for food or to vegetarianism. The first hints that a new ethic was emerging come from the latest of the Dharmasūtras, that of Vasiṣṭha (VaDh 4.5–4.8), who argues that killing for certain purposes is not only permitted but is really not killing at all:

The treatise of Manu states: “An animal may be killed only on the occasion of paying homage to ancestors, gods, or guests.”

When offering the honey mixture, at a sacrifice, and during rites for ancestors and gods—only on these occasions, Manu has declared, should an animal be killed.

Without killing a living creature you can never obtain meat; and killing living creatures does not get you to heaven. Killing an animal at a sacrifice, therefore, is not a killing.

He should, moreover, cook a big ox or a big male goat for a Brahman, for a Kṣatriya, or for a visitor. In this way they show him hospitality (cf. ŚB 3.4.1.2).

It appears likely that this is an argument against an unstated opponent of killing animals even for sacrifices or hospitality. That killing an animal, even a cow, for an important guest is evident in the term goghna (cow-killer) used as an epithet for a guest. The only time abstention from meat is mentioned in these early texts is within the context of certain vows and sacred times: Vedic students, wandering ascetics, people getting ready for an ancestral rite, teachers preparing themselves for the annual course of study, and widows. 13

It is in Manu that we get for the first time a debate on the ethics of eating meat. Manu is torn between the traditional Dharmaśāstric position on not just the permission to eat meat of certain animals but also the injunction to do so in certain ritual settings, especially in ancestral offerings. The angst of Manu as he tries to thread this needle is palpable. After the section on forbidden food, he proceeds to discuss the issues relating to eating meat, a discussion that almost feels like a debate with a traditional opponent (pūrvapakṣa ). At the outset, Manu proposes his thesis regarding occasions when eating meat is permitted: “He may eat meat when it is sacrificially consecrated, at the behest of Brahmans, when he is ritually commissioned according to rule, and when his life is at risk” (MDh 5.27). A putative opponent gives a more liberal view that eating meat is natural and perfectly all right:

Prajāpati created this whole world as food for lifebreath; all beings, the mobile and the immobile, are nourishment for lifebreath. The immobile are food for the mobile; the fangless for the fanged; the handless for the handed; and the timid for the brave. The eater is not defiled by eating living beings suitable for eating, even if he eats them day after day; for the creator himself fashioned both the eaters and the living beings suitable for eating. (MDh 5.28–5.30)

Manu retorts (MDh 5.31–5.44) that it is legitimate to eat meat only on ritual occasions, especially a Vedic sacrifice. Killing (hiṃsā ) on such occasions is truly non-killing (ahiṃsā ), an argument that is repeated in later literature.

Then he returns, like a man pulled in several directions at the same time, to waxing eloquent on the evils of killing animals for food (MDh 5.45–5.55). One can never obtain meat without causing hiṃsā to living beings. There is nothing worse than wanting to fatten his own flesh at the expense of someone else’s flesh; there is, indeed, no greater sin than this. “The man who authorizes, the man who butchers, the man who slaughters, the man who buys or sells, the man who cooks, the man who serves, and the man who eats—these are all killers” (MDh 5.51). Manu closes with this interesting etymology of the term māṃsa (flesh or meat): “Me he (māṃ + sa ) will eat in the next world, whose meat (māṃsa ) I eat in this world—this, the wise declare, is what gave the name to and discloses the true nature of māṃsa ” (MDh 5.55). Manu concludes this thorny topic with a somewhat conciliatory statement, acknowledging the legitimacy of eating meat but showing the moral superiority of abstaining from it: “There is no fault in eating meat, in drinking liquor, or in having sex; that is the natural activity of creatures. Abstaining from such activity, however, brings great rewards” (MDh 5.56).

1 See Olivelle 2002a , 2002b [reprinted in Olivelle 2005b].
2 The translations of Rigvedic passages are from Jamison and Brereton’s (2014 ) translation of the Rig Veda. See MDh 1.39–1.42 for a similar but more elaborate classification of fauna and flora within the context of a creation account. The ruminant animals, such as bovines and antelopes, have only a thick dental layer called “the dental pad” in the upper jaw.
3 For an extensive analysis of animal classification, see Zimmermann 1987 .
4 There is an exception made with regard to a group of five five-nailed animals (pañca pañcanakhāḥ ). The five animals are porcupine, hedgehog, hare, tortoise, and monitor lizard (śvāvidh, śalyaka, śaśa, kacchapa, godhā ), although there are some variants of the list. For a study of this category, see Jamison 1998 .
5 See VaDh14 .48; MDh 5.13; YDh 1.172–1.173.
6 yo yasya māṃsam aśnāti sa tanmāṃsāda ucyate | matsyādaḥ sarvamāṃśādas tasmān matsyān vivarjayet ||
7 See MDh 5.19; YDh 1.170, 175.
8 For a detailed discussion of these two terms in the Dharmaśāstras, see Olivelle 2002a .
9 Grammatically, abhojya as “unfit food” is a Karmadhāraya compound, whereas abhojya as “a person whose food is unift” is a Bahuvrīhi compound.
10 See MDh 3.67–3.121. The grammarian Patañjali also refers to the offering of the five great sacrifices as a central feature of a householder (gṛhastha ): on Pāṇini 4.1.33.
11 Sage’s food (munyanna ) refers to uncultivated food, especially nīvāra, a kind of uncultivated rice. See MDh 6.12–6.21.
12 For studies of this topic, see Alsdorf 1962; Schmidt 1968; Bodewitz 1999.
13 See ĀpDh 2.5.16, 2.18.1; GDh 2.13, 14.39, 16.3; BDh 1.21.18, 2.4.7, 3.7.7; VaDh 10.24.