CHAPTER 2

Hinduism and identity

1. INTRODUCTION

Who lives in South Asia? Of the present population of the region, between a quarter and a third are Muslims. They are, of course, very unevenly distributed: while Pakistan is as much as 97 percent Muslim and Bangladesh 83 percent, the census of 2001 showed the Muslims of India as only 13.4 percent.1 Most of the rest of the population can be described as in some sense Hindu, with the obvious implication that it is mainly to be found in India. In this chapter it is the Hindus, not the Muslims, that are at the center of our attention, and specifically the Hindus of India. But we will not be able to pursue this interest very far without encountering the Muslims of the subcontinent, and together with those of the world at large they will also provide us with a series of contrasts in both the pre-modern and modern settings. In fact a contrast already lurks in my presentation of the current demographic situation. The figures for the proportion of Muslims in the various countries may not be entirely accurate, but they are notably precise, and this reflects the fact that there is not much doubt about who is and is not a Muslim. By contrast, the vagueness of my remarks about the Hindu population has to do with the fact that it is often an open question who is or is not a Hindu, and the ramifications of this fact will occupy us extensively in the course of this chapter.

None of this says anything about ethnicity, where the regional variation is much greater. In the words of a fourteenth-century Sanskrit inscription from Āndhra, the subcontinent is marked by “diversity in language and customs” (bhāṣā-samācāra-bhidā) and by division into many distinct “lands” (deśas).² If we leave aside customs to concentrate on language, we can add that since ancient times India has been divided into two major linguistic zones: the larger north where Indo-Aryan languages prevail and the smaller south where the Dravidian languages hold sway. The major language of the north, and by far the most widespread in India today, is Hindi, spoken by over 40 percent of the Indian population; in its various forms it extends across the northern plains to form the “Hindi belt,” but without reaching the sea on either side. No other language is spoken by as much as 10 percent of Indians; thus Tamil, the Dravidian language spoken by the Tamils in the south east, is spoken by a little less than 6 percent.3 Just how far such linguistic differences are markers of ethnic differences is a question we can leave open; that they are so to at least some extent is clear.

2. THE WEAKNESS OF PRE-MODERN ETHNIC AND HINDU IDENTITY

As in the Islamic case, we begin by looking at Indian ethnic and religious identities in pre-modern times, that is to say, before the colonial period.

As far as ethnic identity is concerned, we can best begin with an exception. Among the peoples speaking Dravidian languages, it is the Tamils who preserve the most ancient literary heritage; it includes an epic that may date from the fifth century and was the object of a medieval commentarial tradition.⁴ This epic provides a vivid articulation of an ethnic pride directed against the Aryans of the north, and there is even a passing mention, not pursued, of the idea of “imposing Tamil rule over the entire world.”⁵ But the quarrel with the north is very clearly ethnic, not civilizational. Thus the Aryans and their kings are not demonized, just soundly thrashed by the Tamil king and his brave Tamil warriors, whose valor the northerners had been foolish enough to doubt;⁶ those who “had hurled insults at the Tamil kings, now bit the dust” and the army of the “northern Āryas” was routed.⁷ Likewise it is northern power, not northern culture that is rejected: on his triumphal expedition to the north, the Tamil king promises his support to “all those in the northern country who honor the Vedas.”8 We thus find ourselves in a familiar setting, quite comparable to the ethnic rivalries that existed to a greater or lesser extent among both Muslim and Christian populations in pre-modern times.

This epic, however, is highly exceptional. It is not typical of Tamil literature, and it seems to be without parallel in the vernacular literatures of the other peoples of traditional India, whether in the south or the north. As the author of a recent study of the emergence of vernacular literatures in medieval India puts it, nowhere in the pre-modern vernacular texts that modern scholars have read to date is it “possible to point to a discourse that links language, identity, and polity; in other words, nowhere does ethnicity … find even faint expression.”⁹ He goes on to draw an explicit contrast with medieval Europe: “narratives of ethnicity and histories of ethnic origins of the sort that obsessed late-medieval Europe did not exist in any form in South Asia before the modern period.”¹⁰ The contrast with the medieval Islamic world would be less drastic but still real. Why traditional Indian ethnic identities should have been so weak is a good question. Just as in the Islamic case, we have to do with a religious tradition associated with a core ethnicity, in this case that of the Aryans; as we saw, in the Tamil epic the Aryans appear as the ethnic group against whom the Tamils do battle. Yet the resulting tensions in the Indian case manifest themselves as social rather than ethnic—a point that will occupy us at some length in this chapter. As this may suggest, a large part of the explanation for the weakness of Indian ethnic identities is very likely to be found in the strength of the caste system;¹¹ but fortunately we need not pursue the question further.

We turn then to religious identity, which will require a more extensive treatment than its ethnic counterpart. But to anticipate, our main finding will be that Hindu identity too is in significant ways weak. Let us start with words; they are not everything, but they can sometimes be very revealing.

The terms “Hindu” and “Hinduism” are a case in point. Unlike “Muslim” and “Islam,” they are foreign, not native words. The people we call “Hindus” did not begin to apply the term to themselves until they learned the usage from others, specifically from Muslims.12 Thus we find “Hindu” as a self-appelation in vernacular texts composed by Hindus of the sixteenth to eighteenth centuries in Rajasthan,¹³ Bengal,¹⁴ and Maharashtra;¹⁵ and these texts also refer occasionally to Hindudharm, “Hindu religion,” or as we would say Hinduism.¹⁶ Further south, a Hindu ruler had adopted the title “Sultan among Hindu kings” (hindurāya suratrāṇa) as early as 1352; his successors were still using the title in 1551.¹⁷ The wholesale adoption of the terms “Hindu” and “Hinduism” in English as written and spoken under British rule was thus to be expected and need not point to dependence on Muslim informants. That the usage was foreign might not have meant much if the imported terms had merely taken the place of native ones serving the same function; but in fact this was not the case, for the foreign terms represented foreign concepts. Such borrowings can, of course, become fully domesticated, given time and the right conditions; it would be foolish to infer from etymology that there can be no authentic Hindu identity in India today. Likewise alien etymology has not prevented the Hindu nationalists from coining the term Hindutva, combining the borrowed term with an ancient Sanskrit suffix to convey the sense of “Hindu-ness,”18 or Hindu identity. As might be expected, the Hindu nationalists are not nationalist enough to renounce the loanword in which their identity is vested.

It is nevertheless worth noting here that the foreign origin of the term “Hindu” was to give rise to some discomfort in modern India. One response was to deny this origin: V. D. Sāvarkar (d. 1966), in a classic text of Hindu nationalism, felt obliged to argue on specious philological grounds that the term was not a loanword at all.¹⁹ A more radical response was to reject the term altogether: in 1873 Dayānand Sarasvatī (d. 1883), a radical religious reformer whose name will recur in this study, told his audience in Calcutta that they should discard the label “Hindu” imposed on them by foreigners, and instead call themselves “Aryans.”²⁰ Likewise in 1877 he inspired the establishment in Lahore of the reformist Ārya Samāj, the “Aryan Association”²¹—and not, as one might have expected, the “Hindu Samāj.” For unlike “Hindu,” “Aryan” is incontrovertibly native, an authentic ethnic and religious term inherited from high antiquity. Yet as early as 1902, Lajpat Rāi (Rāy) (d. 1928), a member of the Ārya Samāj of twenty years standing, was seeking to reinstate the use of the term “Hindu” rather than “Aryan,”²² and in 1911 the members of the Samāj declared themselves as Hindus to the census takers in the Punjab.²³ If this was the attitude of the Ārya Samāj, the chances of “Aryan” being adopted more widely to articulate a Hindu identity in modern times were slim. What, then, was wrong with the term?

One point that is relevant here is precisely the antiquity of the term “Aryan.” It was no longer in current use as an ethnic category, and other things being equal we would expect that in a matter as intimate as identity people would prefer continuity to revival. But the record does show that when other things are not equal, revived identities can stick. The people we know as Greeks mostly called themselves “Romans” for over a thousand years, yet they abandoned this usage on entering the modern world and instead took to calling themselves by the ancient term “Hellenes.”24 There must have been a time when this neologism (or rather paleologism) struck Greek speakers as formal and affected, but today it is as authentic a part of the language as the number system. If the Greeks could successfully adopt a revived identity on emerging from centuries of foreign rule, then why not the Hindus? Part of the answer is that, as we will see in a later chapter, Dayānand’s revival of Aryan identity was part and parcel of a peculiar fundamentalist doctrine that never had more than a marginal impact on modern India at large;²⁵ the return to antiquity was far more central to the cultural evolution of modern Greece than it was to that of modern India.

A more significant disadvantage of the term “Aryan” was its linkage to a theme that will figure prominently in this study: the caste (jāti) system.²⁶ In its classical articulation the backbone of traditional Indian society consisted of four classes (varṇas). Of these, the top three were “twice-born” (dvija), namely the Brahmins (typically priests), the Kṣatriyas (typically rulers or soldiers), and the Vaiśyas (typically merchants); by contrast, the Śūdras, who made up the lowest class of the four, were servile. The social system as it actually evolved was far more complex and at the same time varied from region to region. The number of castes was far greater than the number of classes, with each caste being assigned (with or without contestation) to an appropriate class; and beneath the Śūdras emerged the “Untouchables,” a large population divided into castes of their own. The total number of castes in India today is reckoned to be a few thousand. The relevant point here is that, according to the standard doctrine of the four classes, only the twice-born were Aryans; thus an ancient legal source uses the term anārya, “non-Aryan,” in the sense of Śūdra.27 This meant that much of the population—in all likelihood most of it at most times and places—was traditionally precluded from assuming an Aryan identity.²⁸

Like the foreign origin of the term “Hindu,” this caste-bound character of the term “Aryan” was to be a cause of some embarrassment in modern India. It would not have bothered Śraddhā Rām (d. 1881), a conservative paṇḍit of the Punjab who felt comfortable speaking of Hinduism as the “religion of the twice-born” (dvijadharma).²⁹ But his contemporary Dayānand seems to have thought very differently. The Aryans, he said, had already divided themselves into the four classes in the Tibetan homeland of mankind and then migrated to northern India, where there was no pre-Aryan population;³⁰ that implies that Śūdras are as Aryan as the twice-born.³¹ However, this more inclusive conception does not seem to have carried much conviction. Thus among the twentieth-century Yādavs, a Śūdra caste significantly influenced by the Ārya Samāj, the claim to be Aryan was closely linked with pretensions to Kṣatriya status.³² As a rallying point for the Hindu masses, “Aryan” was an unpromising identity.

Yet even the borrowed term “Hindu” could lose its apparent inclusiveness when adopted in the native social context: though it was readily used to embrace Śūdras, there was a tendency to exclude Untouchables from its scope. Until it became politically desirable for Hindus to maximize the size of their population, this exclusion was a problem for British census officials. As one of them noted in 1881, “Many of the more bigoted high caste Hindoos employed as census enumerators or supervisors objected to record such low persons as of the Hindoo religion.”³³ Moreover, the Untouchables themselves have shown a tendency to reserve the term “Hindu” for the castes above them.³⁴ In 1962 an Untouchable politician belonging to the Camār caste, which had taken to calling itself Jātav, campaigned in the ʿAlīgaṛh district of Uttar Pradesh with the slogan “Jātavs and Muslims are brothers, where do the Hindu people (Hindu qaum) come from?”35 The Camārs, then, were no more Hindus than the Muslims. As we will see, some of them, including the politician in question, had recently converted to Buddhism;³⁶ but Camār use of the phrase “Hindu people” to refer to higher-caste Hindus went back at least to the 1920s.³⁷

The corollary of this lack of a traditional native concept doing the work of the borrowed term “Hindu” is the absence of a clear-cut concept of the non-Hindu. In place of a dichotomy we find something more graded. Roughly speaking, the Brahmins are on the inside track, followed by the Kṣatriyas, followed by the Vaiśyas, followed by the non-Aryan Śūdras, followed by the Untouchable and tribal populations, followed perhaps by the barbarians (mlecchas), who in later times typically came from outside India altogether. Again, just where are we to draw the line, if a single line is to be drawn? Or more importantly, where are they to draw it? There is nothing here to compare with the relative sharpness of the Muslim partition of humanity into believers and unbelievers. Of course, Muslims argue about who exactly is an unbeliever and make distinctions within the category of non-Muslims; but Hindus lack such a category altogether. In the light of this it hardly seems out of place that the Hindus should owe to the Muslims their very concept of the Hindu.

In short, this glance at the traditional vocabulary available for the construction of a Hindu political identity has revealed two problematic features. One was the choice between a foreign term in current use and an ancient term in need of revival. The other was the lack of a sharp binary distinction between those who belong and those who do not. And thanks to their entanglement with the caste system, these features were linked to substantive issues that affected a significant part of the Indian population.

A rather different problem was the lack of a well-defined set of beliefs and rituals constituting the core of Hinduism. In this respect Hinduism is very different from Islam, with its two-part confession of faith and its five pillars. The ascetic Nārāyaṇa Guru (d. 1928), the leader of a low-caste religious movement in Kerala, was not without justification in taking the extreme nominalist view that “there is no such religion as Hinduism,” just a plethora of distinct religions.38 Likewise a Western scholar has argued cogently that, if we were to use our terms consistently, we would have to choose between referring to Hinduism as a plurality of religions and referring to Judaism, Christianity, and Islam as sects of a single religion.³⁹ Bāḷ Gangādhar Ṭiḷak (d. 1920), a Citpāvan Brahmin from Maharashtra and an early Indian nationalist, thus had a point when he averred in 1895 that for the Hindus of even one province to unite to worship the same god for ten days in every year was an event of no mean significance (he was referring to the Gaṇapati festival in Maharashtra).⁴⁰ Diversity of beliefs and rituals may not be an absolute bar to solidarity, but it certainly does not help—as Ṭiḷak was keenly aware.

3. ASPECTS OF HINDU COHERENCE

Yet there are countervailing aspects of the Hindu tradition. They are not common to all forms of Hinduism, but they are sufficiently widespread to give it a certain coherence. Three examples of such themes, and their continuity into modern times, should be enough to convey this; each of the themes introduced here will reengage us later in this study.⁴¹

A first example is the widespread, though by no means universal, recognition accorded to the Vedas, the texts whose transmission is central to the role of the Brahmins. To give just one example of the standing of the Vedas in pre-modern times, a purple passage on Muslim depredations in a southern text of the later fourteenth century laments that “the chant of the Vedas” has deserted the villages and that “the Veda is forgotten.”⁴² Yet, if we look at the attitudes of the devotional cults that have been so prominent in Hinduism for the last thousand years and more, we see great variation. At one extreme, the Tamil adherents of mainstream Śaivism regarded the Vedas as part of Śiva’s revelation.43 At the other extreme, sects like the Vīraśaivas or Lingāyats—a Śaiva sect in the region of Hyderabad and Mysore founded in the twelfth century—tended to reject the Vedas outright;⁴⁴ while in northern India Kabīr, a low-caste Vaiṣṇava or Ṣūfī who may have lived in the early fifteenth century and was a devotee of Rām, declared the Veda and Koran equally worthless.⁴⁵ The case of Kabīr is telling: if we consider him as a Ṣūfī, his attitude to the Koran seems very much an Indian one and sharply at variance with Ṣūfī views elsewhere in the Islamic world. Altogether, the Vedas do not have the same core status in Hinduism as the Koran does in Islam. And yet the fact that the Vedas were the obvious target for anyone in the business of rejecting ancient textual authority is in itself a testimony to their overall recognition.

This pattern continued into modern times. Thus the early liturgy of the Brāhma Samāj, a reformist movement founded by Rām Mohan Roy (Rāy, d. 1833) in Bengal in 1828, included the chanting of the Vedas, at first in a side room to which only Brahmins were admitted, but later for the whole congregation;⁴⁶ yet this was emphatically not a religiously conservative sect. Śraddhā Rām in the Punjab held uncompromisingly that anyone who “thinks that the Vedas, and Dharm Shastras are the products only of the Brahmans” cannot be “a true Hindu”⁴⁷ (the śāstras or dharmaśāstras are the Hindu law books). Vivekānanda (d. 1902), who did more than anyone else to render elite Hinduism a prestige export to the West, rather tendentiously told an Indian audience that the only point on which “all our sects” perhaps agree is that “we all believe in the scriptures,” namely the Vedas⁴⁸ (though he had views of his own regarding the extent of Vedic authority).⁴⁹ A more recent and less cosmopolitan figure, a Śankarācārya of the Śṛngerī monastery (an office we will come to shortly),⁵⁰ gave this advice: “Accept the teaching of the Vedas. Then one can worship any particular deity.”51 But again, while most people recognized the high standing of these venerable texts, others backhandedly acknowledged their prominence in the Hindu tradition by making a point of trashing them. Thus Jotirāo Phule (d. 1890), who developed an innovative low-caste ideology in Maharashtra, attributed the origin of the Vedas to Brahmā, here depicted as the cunning leader of the Aryan invaders; he “collected together some magical incantations and false fables that he knew off by heart.”⁵²

A second example of a feature that is widespread but not universal in Hinduism is revulsion at the killing of cows. This is a recurrent theme in friction between Hindus and Muslims.⁵³ What may be the earliest reference to Muslims in an Indian text correctly ascribes to them the view that there is no sin in eating animals such as cattle.⁵⁴ A late twelfth-century poet in the north gives a fanciful explanation of the ugly physical features of a Muslim ambassador in terms of “the vast number of cows he had slain.”⁵⁵ A southern poetess describing Muslim maraudings in the second half of the fourteenth century speaks of a river “flowing red with the blood of slaughtered cows.”⁵⁶ A Marāṭhī ballad that may date from the seventeenth century tells of a Muslim general who desecrated a Hindu idol and built a mosque in its place: “After the mosque was built,” the ballad continues, “a cow was slaughtered.”⁵⁷ Another early ballad describes a particularly obnoxious Muslim—a Rājpūt convert and a voracious cow-eater—who went so far as to order the sacrifice of a pregnant cow.⁵⁸ Muslim sources complement this picture. For example, Shaykh Aḥmad Sirhindī (d. 1624) saw the sacrifice of the cow as “one of the most important rites” of Islam in India,59 precisely because of its offensiveness to Hindus⁶⁰—though wise or weak Muslim rulers would from time to time forbid the practice for just that reason.⁶¹ This old theme gained new vigor with the development of mass politics in modern India. An early example is Dayānand’s cow protection campaign of 1881.⁶² Another dates from the time of the annual Muslim sacrifice in 1917, when “snowball” letters were circulating in the district of Shāhābād (now in Bihar).⁶³ One of them insisted on the need for “all of us Hindus to unite so that Mother Cow (gau mātā) may be protected.”⁶⁴ Another urged that “if you are a Hindu, you must liberate the cow” and spelled out the corollary: “That is why, you are instructed that wherever you catch a Musulman, you must kill him and loot his village.”⁶⁵

Here again, we are dealing with a feature of Hinduism that is widespread rather than universal. As an orthodox Tamil Brahmin declared in the mid-twentieth century, veneration of the cow could not be regarded as an essential feature of Hinduism; he pointed out that the “horror of cow-killing grew with time.”⁶⁶ Indeed the whole tradition is deeply conflicted. Thus the Ṛg Veda has the god Indra remark that “they cook for me fifteen plus twenty oxen” and refers to the slaughter of a cow for a marriage.⁶⁷ Yet in over forty passages the Vedas refer to the cow with an epithet meaning “not to be slain” (aghnyā).⁶⁸ Since Vedic times the tradition has shifted against the cow killers, but not all the way. Thus, juristic literature of the last millennium and a half generally disapproves of cow killing in the present age;⁶⁹ but the best-known commentator on the law book of Manu states that cow killing is proper on certain ritual occasions.70 Even Dayānand’s views on the cow were more utilitarian than spiritual: “One cow can feed 100,000 people, but its meat can feed scarcely eighty.”⁷¹

A third and final example is the cult of the god Rām, and in particular of his birthplace in Ayodhyā. This cult has, of course, become the subject of considerable obfuscation thanks to the destruction of the Bābrī Masjid, the mosque built for the Mughal emperor Bābur (ruled 1526–1530) in 1528–1529 and alleged to occupy the exact site of the birthplace of Rām. But for our purposes the outlines of the history of the cult are clear enough. It was an old one in the Ayodhyā region but not prominent until around the twelfth century.⁷² The relevant vernacular version of the Rāmāyaṇa, the epic devoted to the life of Rām, was that composed by the poet Tulsīdās (d. 1623). It was in Avadhī, the form of Hindi prevalent in the region around Ayodhyā,⁷³ and the poet began its composition when he was in Ayodhyā on the occasion of Rām’s birthday in 1574.⁷⁴ This was very much a living tradition: a peasant revolt that broke out in Awadh in 1920 was imbued with it, and the leading spokesman of the rebels was an itinerant reciter of the Tulsīdās Rāmāyaṇa.⁷⁵ There was likewise nothing inauthentic about the belief that Rām was born in Ayodhyā, where he came down to earth as a man “for the sake of Brāmans and cows and gods and saints.”⁷⁶ It is a good question where, if anywhere, Ayodhyā was located in ancient times, but the medieval Ayodhyā was undoubtedly where it is now,⁷⁷ and pilgrims have been visiting Rām’s birthplace there for centuries. A traditional pilgrim guide to Ayodhyā exists in several recensions, approximately datable between the thirteenth and the seventeenth centuries.⁷⁸ All describe Rām’s birthplace (the Janmabhūmi or Janmasthān).79 Thus, the oldest version assures the pilgrim that a man who has seen the birthplace will not have to undergo a further rebirth and will attain the same merit as someone who makes an offering of a thousand red cows a day.⁸⁰ What we do not know is the age of the tradition that locates the birthplace precisely where the mosque was built for Bābur⁸¹ and whether a Hindu temple marking the spot was in fact destroyed to build it. In the second half of the nineteenth century the Hindus were making their offerings on a platform outside the mosque.⁸² But in 1949 Hindu idols appeared inside the mosque,⁸³ and thereafter the microgeography of the conflict was unambiguous.

At the same time the Rām cult extended far beyond Ayodhyā and its environs. Its presence was not uniform: one scholar describes the god’s popularity today as concentrated in the Hindi-speaking north,⁸⁴ and this pattern doubtless antedates the twentieth century. Yet versions of the Rāmāyaṇa existed in most of the major vernacular languages of India, and the story also found its way into the literatures of the Jains and Buddhists.⁸⁵ Significantly, the legend of Rām was already being invoked in the context of resistance against Muslims in pre-modern times; the Rāmāyaṇa, unlike the Mahābhārata, has the advantage of demonizing the enemy.⁸⁶ Thus a court poet of the Hindu ruler Śivājī (ruled 1674–1680) dubs his patron’s Muslim enemy a “second Rāvaṇa” (Rāvaṇa being the leading villain of the epic), and puts these words into Śivājī’s mouth:

The Muslims are demons incarnate

Arisen to flood the earth with their own religion.

Therefore, I will destroy these demons

Who have taken the form of Muslims

And I will spread the way of dharma fearlessly!⁸⁷

In the Tamil country an indigenist revaluation going back to the late nineteenth century turned the tables and denounced Rām as an Aryan villain;88 later the Tamil separatist Periyār (d. 1973) was the author of a notoriously blasphemous pamphlet about the Rāmāyaṇa.⁸⁹ But this was modern innovation.

To these themes that give Hinduism a certain coherence we should add that the tradition possesses a remarkably sharp territorial sense. The Hindus may owe to the Muslims the idea of the Hindu, but their traditional idea of India is very much their own.⁹⁰

Let us start with a set of Vedāntist monasteries (maṭhas) whose foundation was attributed to the Vedāntist sage Śankara. Like so many of the major religious figures of the Indian past, Śankara is resistant to precise dating; current scholarship tends to place him around 700.⁹¹ We know him above all as the author of a set of works that made him the classic exponent of a particular tradition of religious philosophy, Advaita Vedānta. What concerns us here, however, is the action he took, or was later believed to have taken, to institutionalize it. According to the medieval biographical tradition, much of it already legendary in character, he traveled all over India, founding a number of major monasteries (maṭhāmnāyas) where his teachings were preserved and propagated. There was general agreement that one of them was located at Śṛngerī, in what is now Karnataka, and that others were placed at Dvārkā on the coast of Gujarat, at Purī on the coast of Orissa, and at Badrīnāth in the Himalayas; the status of a further such institution at Kāñcī near Madras has been more contentious.⁹² These monasteries are spread out over India at large distances from each other, with no sign of regional clustering; all remain within the borders of India at the present day. Each one of these major monasteries was—and is—headed by a Śankarācārya. The individual monasteries typically preserve lists of their successive heads; this tradition seems to date from the later middle ages and cannot take us much further back with any confidence.⁹³ But there is a geographical vision here, and while we cannot say for sure that it goes back to Śankara himself, it is clear that it had been realized on the ground by late medieval times if not before.

This vision invites comparison with the geographical terms enshrined in the tradition. Āryāvarta, the land within which Aryans may dwell, is defined by Manu as extending from the Himalayas in the north to the Vindhyas of central India in the south and from the sea in the west to the sea in the east.94 In this connection Badrīnāth, Dvārkā, and Purī have a clear geographical significance as marking the northern, western, and eastern limits of Āryāvarta. Only in the south do the monasteries depart from this pattern: they are located too far south to mark the southern border of Āryāvarta but too far north to fit the later concept of Bhāratavarṣa, which extends all the way south to Cape Comorin (Kanyā Kumārī).⁹⁵

These monasteries represent a Śaiva vision of religious geography, but a comparable if less dramatic sense of India as a sacred space emerges from the record of Vaiṣṇava pilgrimages. Vallabhācārya, a Vaiṣṇava sect-founder who died—or ascended to heaven—in 1531, went on pilgrimages that had a geographical reach comparable to Śankara’s travels; the locations of significant events in his life are commemorated by eighty-four widely dispersed shrines.⁹⁶ A Vaiṣṇava saint in the late eighteenth century visited Badrīnāth, Purī, Kāñcī, Rāmeśvaram, and Dvārkā;⁹⁷ Rāmeśvaram is located on an island between India and Ceylon. The practices of Vaiṣṇava ascetics in the twentieth century are also relevant here—and likely to be old. The Rāmānandī ascetics traditionally sent new recruits to circumambulate India, visiting Badrīnāth, Purī, Rāmeśvaram, and Dvārkā.⁹⁸ One Rāmānandī was told by his guru to go to the source of the Ganges, fill a copper pot with its sacred water, and pour it out at Rāmeśvaram—an act of devotion considered extremely auspicious; he also visited Badrīnāth, Purī, and Dvārkā.⁹⁹

Śaivas and Vaiṣṇavas thus shared a Hindu idea of India as a domain of religious activity distinct from the rest of the world. As we have seen, this idea was linked to the old notion of Āryāvarta, later extended as Bhāratavarṣa to include the south. Though these conceptions may have meant little or nothing to most of the inhabitants of the peninsula, they were not confined to the scholastic culture of the religious elite. We find both terms used in the context of military hostilities against Muslims.100 A Maharashtrian chronicle of the late eighteenth century, written in the Marāṭhī vernacular of the region, devotes a passage to sketching this idea of India as a domain of religious actions (karmabhūmī). This domain is encircled by the sea on the east, west, and south and bounded on the north by the Himalayas. It is characterized by abundant temples, study of the Vedas and śāstras, places of pilgrimage that bring instant rewards, abodes of gods, families of Brahmins, and the like. Our author concludes his sketch with the remark that “this auspicious land, where the good karma ripens, is limited to fifty-six regions (deśas).”¹⁰¹ Hinduism was thus endowed with a conception of its territory that was both clear and, once the south had been incorporated, stable. Islam too has a clear sense of its territory, the “abode of Islam” (dār al-Islām), but the borders of this abode are not intended to be stable; it is only the original core territory of Islam, the Arabian peninsula (arḍ al-ʿArab, Jazīrat al-ʿArab), or still more narrowly the Ḥijāz, that remains fixed.¹⁰²

The role of pilgrimage in maintaining the territorial cohesion of India was not limited to marking its borders. Long-distance pilgrimages involving large numbers of people could do something to integrate a religious community in another way, by enabling its members to make contact with each other and disseminate ideas. The most impressive example of this anywhere in the world is, of course, the Muslim Ḥajj, a pilgrimage that brings together people from all parts of the Islamic world at the same place and time each lunar year.¹⁰³ Hinduism has nothing quite like this, any more than do Buddhism and Christianity. But it does have a set of large-scale pilgrimages each of which takes place every twelve years.¹⁰⁴ Of these the most important is the Kumbh Melā at Hardvār, near Badrīnāth, in the region where the Ganges descends from the Himalayas onto the plains. Though not ancient, this pilgrimage seems to be old, and Hardvār is already mentioned as a place of pilgrimage in a medieval Purāṇic text.105 We lack evidence of the geographical provenance of the pilgrims before modern times, but data from 1968 show pilgrims coming to Badrīnāth from all over India though, as might be expected, some regions were poorly represented and pilgrims tended to associate with fellow pilgrims from their own region.¹⁰⁶ It is significant that the Kumbh Melā figures prominently in the life of Dayānand, despite the fact that he did not believe in it. He was present three times: in 1855, when he was too taken up with yoga to participate; in 1867, when he distributed thousands of copies of a reformist pamphlet to little effect, railing against such evils as idolatry and sacred rivers; and in 1879, when he distributed a manifesto but came down with acute dysentery (ideas are not the only entities that can spread at mass pilgrimages).¹⁰⁷ For comparison, in 1968 the Indian authorities were making similar use of the pilgrimage for their own purposes, setting up a temporary family-planning exhibition with a clinic.¹⁰⁸

To sum up, we have seen that both ethnic and religious identity were considerably weaker in the traditional Hindu world than in the traditional Islamic world. The weakness of the Hindu identity in particular comes across in the problematic character of the terms associated with it, the ambiguity as to who is in and who is out, and the lack of a common core of beliefs and rituals. But there were countervailing elements that gave the religion a degree of coherence, notably the widespread recognition of the Vedas, the hostility to cow killing, and the veneration of Rām; to these we can add the existence of a territorial idea of India reinforced by monasteries and pilgrimages. In all this two things are worth highlighting. One is the role of caste in the fraying of Hindu identity as one moves down the social pyramid (and perhaps also in the weakness of ethnic identity). The other is the part played by Muslims not just in giving the Hindus a name for themselves but also in serving as their historic enemy for the last thousand years.

4. MODERN ETHNIC AND HINDU IDENTITY

We now turn to the role played by these identities in the context of modern politics, that is to say, in forms of politics heavily influenced by Western institutions and values. One of these values is, of course, nationalism, and its reception in the region might lead us to expect ethnic identities to acquire a higher political profile than religious identities. But this has not in fact been the case. Religious identity has continued to play a very prominent role in making the subcontinent what it is. It lay behind the Muslim separatism that led to the partition of India in 1947 and has since been central to the politics of Indian Kashmir. For some years it inspired a militant Sikh separatism in the Indian Punjab. It has also given us the Hindu nationalist movement, in association with no small amount of communal violence. Ethnic identity can compete with this record only in two contexts, neither of which is central to our concerns. One is the northeastern fringe of India, where separatist insurrection has been rampant among populations that were to a large extent outside the traditional Hindu world. The other is Pakistan, where the hegemony of the Punjabis was a major factor behind the secession of Bangladesh and has given rise to violent tensions in the rest of the country. But within India since independence, outside the northeast, the role of ethnic identities has been much quieter. That this should be so was by no means obvious a priori. The early years of independence saw strong pressure for the creation of linguistic states within the federal structure, and the rulers of India gave way to this pressure with great reluctance, fearing that it would lead to the disintegration of the country. But since the major linguistic states were established in the decade 1956–1966,109 the problem has more or less gone away. This undoubtedly reflects the weakness of India’s ethnic identities in pre-modern times. Today even the Tamils, the only one of these linguistic groups to take their particularism to the point of separatism, seem satisfied. Thus, there is little more that needs to be said here about ethnic identity, though we will return to it briefly in connection with its role as an obstacle to Hindu nationalism, and it will resurface in the course of our discussion of caste.¹¹⁰

Our main concern from this point on is with Hindu identity. In the politics of modern India as a whole this identity is very much the property of the Hindu nationalist movement, a close-knit set of political organizations; and since the names of these organizations will recur in what follows, it may be useful to list them here. The Rāṣṭriya Svayamsevak Sangh (RSS) is the oldest member of the “family” of Hindu nationalist organizations (the Sangh Parivār), and its continuing core—or at least its conscience. It was founded in the city of Nāgpūr in 1925.111 In 1951 the movement established a political party, the Jana Sangh;¹¹² its heir is the Bhāratīya Janatā Party (BJP), formed in 1980.¹¹³ Meanwhile in 1964 the movement had also created the Viśva Hindu Pariṣad (VHP), or World Hindu Council as it is known in English, a body established for the defense of Hinduism.¹¹⁴ The other member of the family we should note is Sevā Bhārtī, an agency for social work set up in 1979.¹¹⁵ Also worth mentioning here is the Shiv Sena, the violent populist movement that dominates the politics of Bombay; though it is entirely independent of the Hindu nationalist family and has a quite distinct local origin, it has come to share some of the attitudes of the Hindu nationalists. The Shiv Sena apart, this ramification of institutions represents a division of labor within a movement that has so far proved remarkably free of schism; the contrast with Islamist movements is noteworthy. There has nevertheless been persisting tension within the family, and it relates to the terms on which the movement is to participate in the wider field of Indian politics.¹¹⁶ On one side are the RSS activists, who tend to remain loyal to the basic values of the movement through thick and thin; on the other side are the BJP politicians, whose pragmatism reflects their awareness of the compromises they have to make if the party is to enjoy the fruits of power through electoral competition.¹¹⁷ These remarks are intended only as a preliminary orientation; we will take up the question of the broader nature of Hindu nationalism toward the end of this chapter.118 In the meantime I will trace the political fortunes of Hindu identity on three fronts: ethnic, social, and religious—though in practice they are hard to separate.

As indicated, the ethnic front will not occupy us long. It is a well-known fact that Hindu nationalism finds much more support in the north of the country than it does in the south.¹¹⁹ One reason for this is historical: the Hindu nationalists originally wanted to see the adoption of Hindi as the national language.¹²⁰ This was a recipe for alienating the Dravidian south, now divided into four linguistic states, each with its own language. However, the Hindu nationalists were wise enough to abandon their Hindi chauvinism,¹²¹ so it hardly explains the fact that their support has continued to trail in the south. As we will soon see, this may have as much to do with caste as ethnicity—though there are contexts in which the two become blurred.

This brings us to the social front, which needs more elaborate treatment. Our concern here is with the caste system. Traditionally its central feature was a set of exclusions that grew ever more drastic as one descended the caste hierarchy: no intermarriage, no interdining, and the like. Since we have already encountered the Vedas and will meet them again, let us use them to illustrate the symbolic aspect of these exclusions. In traditional Hindu society, the Brahmins can teach and hear the Vedas; the other twice-born castes can hear but not teach them; the Śūdras can do neither, still less the Untouchables.¹²² Thus an ancient source prescribes that if a Śūdra listens to the Veda with a view to memorizing it, he gets molten lead in his ears, while if he utters it, his tongue may be cut off, and if he has mastered it, he should be hacked to pieces.123 This denial of access to the text also meant exclusion from Vedic ritual. Śūdras could not use Vedic mantras—Vedic verses possessing religious efficacy—in their worship.¹²⁴ In this connection the “mother of the Veda” (Vedamātā)¹²⁵—we might be tempted to say the mother of all mantras—is the Gāyatrī mantra. This Vedic verse is an invocation of the sun god: “Let us think on the lovely splendour of the god Savitṛ, that he may inspire our minds.”¹²⁶ It was to be taught to the twice-born boy following his initiation.¹²⁷ Thereafter multiple repetitions of the Gāyatrī were to play a key role in the ritual prayer (sandhyā) performed three times a day.¹²⁸ Over the course of a lifetime, one should repeat the verse twenty-four hundred thousand times, if need be getting others to help out; we hear of a ceremony held in the early twentieth century at which seventy Brahmins each repeated the Gāyatrī four thousand times a day for thirteen days.¹²⁹ Like the Vedas at large, the verse is closely associated with Brahmins; there is even a view in the ancient sources that it is for Brahmins alone, to the exclusion of the other twice-born classes.¹³⁰ In any case the Śūdras were excluded—and so by implication were the Untouchables.

This symbolic example doubtless makes caste sound like an archaism that would not survive the onset of modernity. In fact, as we will see, caste has often been fostered as much as eroded by modern conditions. What then are its implications for identity? Here two things seem obvious a priori. The first is that such a system was likely to give rise to enormous resentment in its lower reaches, particularly under modern conditions. The second is that the resentful would be strongly tempted to opt out of the Hindu hierarchy. Examples of both effects are not hard to find.

We have already met Jotirāo Phule, who in the mid-nineteenth century launched a non-Brahmin movement in the region of Poona in Maharashtra.¹³¹ His social program was designed to bring education—and hence power—to the non-Brahmins, his primary constituency being the Marāṭhās and associated castes, traditionally classed as Śūdras.132 The identity he offered them involved a dramatic reshaping of their history.¹³³ They were, he told them, the native population of the region. Originally they were Kṣatriyas who lived in peace and happiness (he supplied them with a new and appropriate etymology of the term “Kṣatriya” that associated it with the word for “field,” kṣetra, rather than with warfare). This golden age continued until disaster struck in the shape of the Aryan invasions; he owed this notion of the Aryans as alien intruders—as originally they almost certainly were—to European scholarship.¹³⁴ These invasions, he explained, brought Brahmin domination, and with it the reduction of the indigenous population to the status of Śūdras and worse. Eventually, he taught, the Creator in his mercy sent first the Muslims and then the English to liberate the natives from the Aryan yoke; the Muslims showed themselves to be utterly corrupt, but the English fortunately did not.¹³⁵ Meanwhile the Brahmins had elaborated their false religion as part of their conspiracy to subjugate and exploit the native population; we have already encountered Phule’s unflattering account of the compilation of the Vedas.¹³⁶ Two things are particularly striking about his vision. The first is his transposition of caste difference into ethnic difference. The second is his rejection, not just of the Brahmins but of the core of the religious tradition associated with them, from the Vedas to Śankara.¹³⁷ We hear less of what he wanted to see in its place, though he made much of Khaṇḍobā, a prominent deity of the castes to whom he was making his main appeal.¹³⁸

Phule’s exclusion of the Brahmins from the community was echoed by other dissidents over the following century. By far the most conspicuous instance of such notions was the indigenist movement that was to lead to Tamil separatism.¹³⁹ The Tamils, of course, were particularly well placed to play the indigenist card: unlike the speakers of Marāṭhī, they were one of the four major southern peoples set off from the northerners by the fact that they spoke and wrote Dravidian languages, and of these peoples they possessed by far the most ancient and distinctive literary heritage. Yet even in the north, caste divisions could be construed as ethnic divisions by those determined to do so. Thus Svāmī Achūtānand (d. 1933), an Untouchable leader in United Provinces and a Camār, in 1917 set out his views about the “Ādi-Hindus” or proto-Hindus, a category in which he included Śūdras, Untouchables, and Tribals—in contrast to the twice-born Brahmins, whom he considered “as foreign to India as are the British” (indeed he clearly considered the British an improvement on the Brahmins).140 The Ādi-Hindu religion was, he said, “the oldest religion,” a religion of hundreds of saints, one in which all are equal before God, in contrast to the ways of the Brahmins; “we do not need any other religion.”¹⁴¹

To return to Maharashtra, what Phule did for the lower castes in the nineteenth century, Bhim Rāo Āmbeḍkar (d. 1956) did for the smaller but more downtrodden Untouchable population in the twentieth and with a much greater impact on people of the same status in India at large.¹⁴² It was he who coined the term “Dalits,”¹⁴³ which has come to be used more or less with the sense of “Untouchables with attitude.” Again, he rewrote history to recover an allegedly lost identity for his constituency. This was primarily his own Mahār caste, the largest Untouchable caste in Maharashtra; its members traditionally provided their villages with a variety of services, including the removal of dead animals.¹⁴⁴ In this case, however, the lost identity was a religious, not an ethnic one: Dalits, Āmbeḍkar explained, were former Buddhists who had been reduced to their current abject condition through the medieval resurgence of Hinduism.¹⁴⁵ Considerable numbers of Untouchables in Maharashtra, mainly members of the Mahār caste, accordingly opted out of Hindu society by “returning” to Buddhism in a ceremony orchestrated by Āmbeḍkar in the year of his death.146 Thus Buddhism, a religion doctrinally indifferent to caste, was adopted as a badge of opposition to it.

As with Phule’s ideas, subsequent decades have shown that Āmbeḍkar’s interest in Buddhism was not just a passing idiosyncrasy. A few months after he died—foully murdered, it was said, by Congress “giving him poisonous injections through a Brahman”¹⁴⁷—the energetic Camār politician active in the ʿAlīgaṛh district converted to Buddhism and went on to infuriate the upper castes by organizing a huge conversion meeting at which, he claimed, a hundred thousand Camārs adopted Buddhism, with Hindu temples becoming Buddhist ones.¹⁴⁸ Kānśī Rām (d. 2006), who seems likewise to have been a Camār¹⁴⁹ and had learned about Āmbeḍkar’s ideas from a Mahār Buddhist, was not in favor of conversion to Buddhism unless all Dalits were to convert at once.¹⁵⁰ But Māyawatī, a Camār woman with unlimited attitude, whom he had recruited to the cause, made use of Buddhist symbolism during her brief spells of power in Uttar Pradesh in 1995 and 1997, naming a district for the Buddha’s mother and referring to Āmbeḍkar as a Bodhisattva in a dedicatory inscription;¹⁵¹ during her third spell as chief minister in 2002–2003, she threatened to follow Āmbeḍkar and convert to Buddhism together with Kānśī Rām and several hundred thousand members of the Bahujan Samāj Party,¹⁵² and in 2006 she gave him a Buddhist funeral.¹⁵³ In Āmbeḍkar’s thinking the rupture between the Untouchables and the higher castes was religious, not ethnic; indeed it is a significant feature of the Buddhist option that Buddhism is incontrovertibly Indian—one reason he did no more than flirt with the Muslim option.¹⁵⁴ But we also find cases, several of them antedating Āmbeḍkar, where indigenist and Buddhist notions came together;155 thus, one indigenist from the Āndhra region celebrated the Buddha’s birthday in 1913.¹⁵⁶

Not every religious option open to those at the bottom of the pile could be presented as Indian. With the advent of Muslim and later Christian rule, it became relatively easy to leave the Indian religious tradition altogether, adopting instead a faith of foreign origin that was at least in principle caste blind: if you were not admitted to the Hindu temple, you could turn to the mosque or the church. The result was a hemorrhage of lower-caste and Untouchable Hindus through conversion,¹⁵⁷ and this hemorrhage did not stop when Muslim and Christian rule eventually came to an end. In 1981 Indian public opinion was badly shaken when some thousand Untouchables converted to Islam in a village in the Tamil south under the auspices of the leaders of the Muslim League; this was followed by thousands of conversions in the region. These events engendered reports of a grand Muslim conspiracy using Arab oil money to entice large numbers of Hindus to turn Muslim, and a majority of Indians—especially in the northern cities—wanted government action to end such conversions.¹⁵⁸

There was, of course, a potentially more harmonious alternative to the processes we have been looking at. The excluded could be included. Such a change might be initiated either from below or from above. Initiation from below tends to take the form of “Sanskritization.”¹⁵⁹ Here the caste hierarchy appears as a ladder, and an aspiring caste may seek to climb it—rather than walk away from it—by adopting the manners and customs of the higher castes. Vivekānanda was strongly in favor of this course: “I tell you men who belong to the lower castes, the only way to raise your condition is to study Sanskrit, and this fighting and writing and frothing against the higher castes is in vain”; “use all your energies in acquiring the culture which the Brâhman has, and the thing is done.”¹⁶⁰ We have already met one typical example of a caste seeking to climb the ladder: the claims of the Yādavs to Kṣatriya status.161 In taking to calling themselves Yādavs rather than Ahīrs, they were identifying themselves with an ancient tribe and a medieval dynasty.¹⁶² They were helped by the fact that they were traditionally herdsmen, which linked them to the sacred cow.¹⁶³ An earlier example of the same phenomenon would be the contentious claims of the Marāṭhā elite to Kṣatriya status in the mid-nineteenth century, at a time when the end of the rule of the Brahmin Peśvās and the indifference of the new British overlords to matters of caste discipline had made it harder to keep people in their places.¹⁶⁴ These claims were endorsed by the outcome of a formal debate held in 1830,¹⁶⁵ though as late as 1865 it was still being argued that no Marāṭhā could claim to be anything but a Śūdra.¹⁶⁶ Those whose attempts to climb the ladder were blocked from above could sometimes get results by threatening to walk away from it. Lower-caste Hindus seeking to pry open the temples of Kerala backed up their demonstrations with an implicit threat to convert to other religions if their demands were not met; the Maharaja of Travancore stepped in and opened the temples of his principality in 1936.¹⁶⁷ Yet such attempts to climb the social ladder were at the same time endorsements of the ladder itself.¹⁶⁸ One of the most outspoken protagonists on the Marāṭhā side in nineteenth-century Maharashtra was just as vehement in denouncing people lower down on the ladder who aspired to enter his own caste.¹⁶⁹

There were also pressures for inclusion coming from above. Modern politics, particularly electoral politics, put a premium on large constituencies. Hence the tendency of the excluded to opt out of Hindu society could not leave high-caste Hindus indifferent, and considerable efforts were made by Hindu reformers to eliminate or at least mitigate the traditional exclusions of Hinduism. Yet in real life concessions were not made easily or quickly. We may return here to the case of the Gāyatrī mantra, denounced by Phule as an incantation that the Aryan invaders had strict instructions not to reveal to the natives.170 It long remained a contentious issue in Maharashtra whether Śivājī had been permitted to recite it at his coronation;¹⁷¹ he might have been a celebrated champion of Hinduism, but he was no more than a Śūdra with recently concocted claims to Kṣatriya descent. Dayānand himself initially took for granted the exclusion of Śūdras from the Veda,¹⁷² but later he passionately supported their right to study it—had not God made it to shine for all?¹⁷³ In a part of the Punjab where his arrival was greeted by Hindu mobs hurling stones and bricks, he had the audacity to discuss the Gāyatrī in front of a mixed audience of Hindus and Muslims;¹⁷⁴ and if he thought it in order for Muslims to hear it, he can hardly have withheld it from Śūdras. But it was not all plain sailing. In 1924 the Hindu Mahāsabhā, an offshoot of Congress that did not want Untouchables converting to Islam, recommended a number of reforms to give them a better deal, including allowing them entry to temples; but it was still not prepared to countenance teaching them the Vedas.¹⁷⁵

The effort of reform from above that most concerns us is that of the Hindu nationalists. One of the problems of the early RSS in its Maharashtrian homeland was that it was not broadly representative of the local population, since the core of its membership was Brahmin.¹⁷⁶ Though the number of low-caste members in the movement has increased since then, they are still far from numerous and have tended not to rise very high in the organization.¹⁷⁷ This does not mean that the RSS represented a narrow projection of the interests of the upper castes—the Hindu nationalists were as aware as anyone of the danger of losing the Hindu masses to conversion,¹⁷⁸ and they were inspired by a nationalist vision of Hindu society as “one single homogenous family.”179 Statements made by the RSS and its leaders rejected the traditional Hindu class (varṇa) system and called for the abolition of caste, including Untouchability;¹⁸⁰ as far as Sāvarkar was concerned, the more everyone intermarried the better it would be for the Hindu people.¹⁸¹ This was rhetoric, but it was not mere rhetoric. Already in the early years of the movement, we hear of meals eaten in common with low-caste Hindus and of high-caste members living side by side with Untouchables.¹⁸² In the same spirit the VHP, when organizing the massive pilgrimages of 1983, included Untouchables among those assigned the task of carrying holy water.¹⁸³

The process of assimilation was nevertheless an asymmetric one. The culture of the movement was “Sanskritized,” a culture of Brahmins;¹⁸⁴ one Untouchable member accordingly made a point of learning Sanskrit and displaying his knowledge of it.¹⁸⁵ At a month-long Hindu nationalist course held annually in a village near Śṛngerī, Vedic coaching would be given to boys from “all sects and castes,” including a number of Untouchables; it was noted that “some of the best Vedic chanters turned out from such camps are from the backward sections.”¹⁸⁶ The Vedic theme recurs in the activities of Sevā Bhārtī, the RSS organization established in 1979 and devoted to something the Hindu nationalists have been good at, namely social work; thus Sevā Bhārtī runs schools for Untouchables in the slums of Āgrā.¹⁸⁷ As in all the schools operated by the organization, the day begins with the pupils—who here come disproportionately from the lowest of the Untouchable castes—reciting the Gāyatrī.188 Untouchables are likewise encouraged to participate in Hindu festivals from which they have previously been excluded.¹⁸⁹ All this is a program of uplift, intended to help “our underprivileged brothers” to “gain self-confidence.”¹⁹⁰ In an earlier epoch the message would have been expressed more bluntly, as in an address by the secretary of the Mangalore Depressed Classes Mission to an Untouchable audience in 1908: “If higher caste people ever beat you, that is because you become abusive when drunk. Behave properly, be humble and polite but at the same time try to improve your condition.”¹⁹¹ In our day the tone is somewhat subtler: “Social justice can be rendered to the weaker sections of society only when the entire society is imbued with the spirit of oneness and internal harmony,” as a leading member of the RSS admonished in 1993. “It is only with the goodwill and cooperation of the entire society that they can get the necessary opportunities to raise themselves up.”¹⁹² This patronizing tone is one reason why members of the largest Untouchable caste of Āgrā, the Camārs, have no use for Sevā Bhārtī, which they regard as condescending.¹⁹³ In the same way, as one observer puts it, this RSS strategy has found “few takers” on its home ground in Maharashtra, where the prevailing culture is antithetical to social hierarchy.¹⁹⁴ Even within the Hindu nationalist movement lower-caste resentment can occasionally be heard. In the 1990s Umā Bhārtī, whose powers of religious oratory already gave her a certain importance in the BJP of Madhya Pradesh despite her lower-caste origin, had not achieved full acceptance by the party chiefs there, and complained that the lower-caste cadres of the party “have an upper caste mentality.”¹⁹⁵

To complete this picture of the way in which caste politics pose an obstacle to the marketing of Hindu identity by the Hindu nationalists, we need to extend it to include the contentious matter of reservations. This institution is the prime example of the fact that caste politics are not just about symbols and sentiments; they relate to the real interests of the parties in very concrete ways. Reservations are quotas imposed by the state in favor of the less advantaged castes. They invite comparison with what Americans call affirmative action, though such policies have given rise to far more dissension in India than in the United States. Whereas white students in America mount legal challenges to affirmative action, their high-caste counterparts in India may burn buses, derail trains, sack government buildings, and immolate themselves.196

The institution of reservations has a history going back over a century. Because the Brahmins were traditionally the educated elite of Indian society, both British and native rulers tended to give them a disproportionate share of the administrative positions open to Indians. The question then arose whether this tendency should be countered by reserving some proportion of appointments (or of other good things dispensed by the state) for members of less privileged castes. An early move in this direction was made in 1902 by the Marāṭhā ruler of Kolhāpūr in what is now Maharashtra. He was on bad terms with his Brahmins, partly as a result of a dispute over the terms on which they were prepared to accept his claims to Kṣatriya status, and to spite them he decided to reserve half the future vacancies in his administration for non-Brahmins, primarily fellow-Marāṭhās.¹⁹⁷ Elsewhere reservations first made headway in the south,¹⁹⁸ but in due course they were extended all over India.¹⁹⁹ At independence the system was inherited by the new Indian rulers from the outgoing British.

Over time, it came to be marked by two significant asymmetries. The first was that, thanks to continuing initiatives taken by state governments in the south after independence, it was much broader and more effective there than in the north.²⁰⁰ As a result, the politics of the south had undergone a sea change by 1980, whereas in the north the higher castes remained dominant, with the positions set aside for the disadvantaged often left unfilled and their leaders regularly co-opted by their social superiors.²⁰¹ The second asymmetry was within the north. Here reservations in favor of Untouchables were an established feature of the system; for this purpose Untouchables had come to be known in official parlance as the “Scheduled Castes,” a term brought into use by the British in 1935.202 But little had been done in the north about reservations in favor of the castes in the middle of the pile, which came to be known in the same parlance as “Other Backward Classes” (OBCs) thanks to Nehru’s use of this phrase in a resolution of 1946.²⁰³ Two commissions were set up to report on the OBC question at the national level, one in 1953 and the other—the Maṇḍal Commission—in 1978.²⁰⁴ They differed in a significant respect: the first was led by a Brahmin, the second by a Yādav²⁰⁵—the Yādavs being a leading example of the OBC category. Yet nothing came of their reports until 1990, when prime minister V. P. Singh announced the implementation of the Maṇḍal report.²⁰⁶ This time it was for real, and the resulting “Maṇḍalization” played a major role in the “silent revolution” that, for all its incompleteness and unevenness, has upended the caste hierarchies of large parts of the north over the last generation.²⁰⁷

The implementation of the Maṇḍal report provides a good example of the problematic character of reservations for the Hindu nationalists. On the one hand they could not afford to alienate the party faithful, who stemmed disproportionately from the higher castes, but on the other hand they needed lower-caste votes to get elected. In this context the increased political consciousness of the OBCs was a crucial force working against the BJP.²⁰⁸ While the ideologically zealous RSS felt free to attack the Maṇḍal Report head-on,²⁰⁹ the vote-conscious BJP was inevitably under pressure to come to terms with the new social order (or disorder). Its attempts to do so have been a vital, if rather indeterminate, chapter in the history of the party, with considerable variation over space and time; the party organization in Maharashtra, for example, stood out for its clear-cut support for the Maṇḍal Report.210 To a significant extent the party is resigned to “Maṇḍalizing” itself,²¹¹ though the results have been mixed and more in evidence at the bottom of the party than at the top.²¹² Despite the elective affinity of Hindutva for the higher castes, and their historically dominant role in the movement, it is not inconceivable that the caste character of the party could eventually undergo more drastic change. The Shiv Sena, after all, represents a politically successful appropriation of Hindu nationalism by a party with no pretensions to high-caste status. But whether, or how fast, the BJP can evolve in such a direction remains to be seen.

One thing that this and other stories about reservations tell us is that the robustness of caste politics in India is not just a matter of social inertia. The fact that the state, even while disapproving of caste as a relic of an earlier epoch, was prepared to make it the basis of a set of policies shaping the distribution of good things to its subjects gave the institution a political salience it would not otherwise have enjoyed. There is a marked contrast here with the societies of Pakistan and Sri Lanka, where public policy does not engage with caste, partly at least because the dominant religious values are either hostile or indifferent to it.²¹³ The result is that in these countries caste, though socially real, is far less politically salient;²¹⁴ instead the conflict over who gets what is more ethnic and regional. But even in India the policies of the state were not formulated at the level of actual caste communities. Instead, they spoke of “Scheduled Castes” and “Other Backward Classes.” Neither sounds like an identity worth dying for. As noted, the Scheduled Castes are the Untouchables, itself a term of convenience that has no equivalent in the vernacular languages of India;²¹⁵ there is no Untouchable community, just a large number of castes for which it is useful to have this generic label. The same goes for the OBCs. Since the redistributive policies of the state do give these large-scale groupings of castes some interests in common, they can form political coalitions while the sense of common interest lasts. There was accordingly a significant coalition of OBCs in the early 1990s thanks to the implementation of the Maṇḍal report, but it faded in the later years of the decade.216

This is where social inertia—the survival of traditional caste identities—is crucial.²¹⁷ Embedded within the categories created by the state are real caste communities, and the relative solidarity of a limited number of large communities is the bedrock of caste politics.²¹⁸ We have already encountered two instances of such castes, and it is worth reviewing them here.

The first is the Camārs, the Untouchable caste of the Hindi belt that gave the world Svāmī Achūtānand, formed an alliance with Muslims, played with Buddhism, and had no use for Hindu nationalist social work. They are the major Untouchable caste of the region, indeed the largest in India; traditionally they had been leatherworkers, though in practice most of them are agricultural laborers.²¹⁹ Already in the first half of the nineteenth century the Camārs of central India had given rise to the Satnāmī movement, which articulated their resentment against the upper castes, partly by encouraging them to walk away from the ladder (since all men are equal), and partly by urging them to climb it (as by wearing the sacred thread of the twice-born).²²⁰ In India today the Bahujan Samāj Party, founded by Kānśī Rām and named after a phrase of Phule’s,²²¹ is a Camār party; other Untouchable castes tend to vote for different parties, including the BJP.²²² When Māyawatī wanted to highlight the turmoil that would ensue were she assassinated, her threat was that the members of the Camār community would “create a havoc”;223 in this emotive context she did not speak of Untouchables, Scheduled Castes, or even Dalits.

The second example is the Yādavs, the caste that came under the influence of the Ārya Samāj, aspired to Kṣatriya status, and supplied the politician who gave his name to the Maṇḍal Commission. Though traditionally herdsmen, they are in fact mostly cultivators, and form one of the largest castes of Uttar Pradesh and Bihar.²²⁴ Unlike the Camārs, they have no claim not to be Hindus, but they are uninterested in the politics of Hindu identity.²²⁵ Two Yādav politicians were prominent in state politics in the 1990s and beyond. One was Mulāyam Singh Yādav leading his Samājvādī (i.e., Socialist) Party in Uttar Pradesh. The other was the leader of the Janatā Dal in Bihar, who made a point of horrifying the elite by speaking in the local Bhojpurī dialect, or in English with a thick Bhojpurī accent; as so often in lower-caste politics, it is essential to be seen to “diss” the twice-born.²²⁶ But the gratifications that go with this political prominence are not just symbolic. In general, the benefits of the “silent revolution” went mainly to a small number of castes, rather than to the OBCs as a whole;²²⁷ and it is no surprise that the Yādavs are the prime example of this.²²⁸ They were accordingly resented by other OBC castes,²²⁹ who might show their dislike by voting for the BJP.²³⁰ Altogether, the Camārs and Yādavs constitute large lumps in the electorate that the BJP and its Hindu nationalist ideology has had no chance of digesting.

It is not that all castes matter in politics; probably most do not, either because they are too small and localized or because they have too little sense of community. But there are enough large castes enjoying the rooted loyalties of their members to have a serious impact on the political environment. For such castes old-fashioned caste consciousness has been far more of a given than any newfangled leftist class consciousness; the class identities so laboriously constructed by twentieth-century Communists look jerry-built by comparison. The past is not necessarily the future, but it does have inertia on its side.

In this context of vivid caste loyalties, at once inherited from the past and stimulated by modern politics, it was electorally disastrous for the BJP to be perceived as a party of the twice-born. As a prominent member of the party in Maharashtra lamented, “We are branded as a Brahmin party”; he saw this as a “major obstacle” in the way of the party’s success.231 All in all, the politics of caste in modern India have shown a potential to erode the clientelistic hierarchies that had previously maintained the dominance of the higher castes,²³² and have immunized large sectors of the electorate to the appeal of Hindu nationalism. This is a politics of attitude but also a politics of interest. “Do not take the candy of reservations and divide yourselves into castes,” implored Ritambarā, a Hindu nationalist female ascetic (sādhvī), giving a speech in Hyderabad in 1991;²³³ but to those who never had it before, candy tastes uncommonly good. Caste is the Achilles heel of Hindu identity.

What is so remarkable in comparative terms is that our long discussion of the travails of Hindu identity on the social front had no parallel in the Islamic case, where our attention was concentrated instead on ethnic tensions. Many Muslim countries contain non-Muslim or sectarian minorities, such as Christians and Shīʿites. But we encountered no populations that were excluded from the Muslim community, or felt themselves not to belong to it, for reasons related to social stratification. Thus Pakistan, like India, inherited the institution of caste at independence, and despite a hostile religious environment it remains a fact of life; and yet those at the bottom of the pile do not respond by rejecting a Muslim identity, nor are they immune to the blandishments of Islamists.²³⁴ Islamists can face many problems in mobilizing Muslim populations, but a tendency for the underprivileged to opt out of Islam is not one of them. A scholar who has paid attention to the social appeal of Islamism across the Islamic world stresses the “two-pronged” character of the movement from the outset: it was able to mobilize the devout middle class and the slum dwellers alike.235 Its problem was rather to hold together two constituencies with such divergent interests—witness the recent split in the Islamist vote in Egypt between the Muslim Brothers and the Salafīs and the analogous division in Tunisia. All this is in a way so obvious that it hardly needs saying. But it helps to dramatize a very basic contrast between Muslim and Hindu identity: socially speaking, Muslim identity is potentially integrative, whereas Hindu identity is almost inevitably divisive.

What would it be like if Islam did provide doctrinal support for a caste system? The question is not entirely hypothetical. Even in the Arab core of the Islamic world there exist, or existed until recently, a few Muslim societies with social groups that could aptly be described as castes. Two cases in point are Upper Egypt and Ḥaḍramawt (not to mention Yemen at large). In each case society is or was divided into three broad castes: an aristocracy of descendants of the Prophet (Ashrāf or Sāda); a tribal population disproportionately endowed with the means of violence (referred to in Upper Egypt as Arabs);²³⁶ and an underclass of peasants, artisans, servants, or the like.²³⁷ The descendants of the Prophet enjoy a certain prestige in many Muslim societies, but only in exceptional cases like these have they constituted a religious aristocracy comparable to the Brahmins. In such societies it is obviously the members of the underclass that have the least stake in the system and are most likely to feel themselves to be the victims of discrimination; notably, they are precluded from marrying women of higher castes.²³⁸ On the Indian analogy one might accordingly expect that they would be tempted to convert to other religions. In fact, of course, nothing of the kind has happened. This is hardly puzzling in the case of Upper Egypt, where the process of labor migration that in recent decades has opened the underclass to external influences took its members to the solidly Muslim—and Muslim-ruled—societies of the Persian Gulf.239 It is perhaps a little more arresting in the Ḥaḍramī case, where much of the corresponding migration took place earlier and was directed to regions under European rule whose populations were less uniformly Muslim.²⁴⁰ Yet in both cases the discontent of the underclass was expressed in properly Islamic terms, by appealing to the wider Islamic tradition against the socially inegalitarian forms of it that had hitherto prevailed locally.²⁴¹ Thus in each case the socially disadvantaged rejected the aristocratic pretensions of the descendants of the Prophet; but in each case they did this by invoking the message of the Prophet himself—something that they could do because the mainstream tradition is not a socially exclusive one.²⁴² Thus Phule had no counterpart in Upper Egypt or Ḥaḍramawt, and Bin Laden—whose family is reputed to stem from the Ḥaḍramī underclass—was destined to become an Islamist rather than an apostate. One reason for this contrast is that the religious discrimination to which members of the Muslim underclass were exposed was less marked than that experienced by their Indian counterparts: they were not excluded from mosques, and in Upper Egypt they were in fact conspicuous for their attendance.²⁴³ The other reason is, again, that Muslim identity and Hindu identity are different. The result is that the defection to other religions so common in the Hindu case is not just unparalleled but also unnecessary and indeed unthinkable in the Islamic case.

In a traditional agrarian society, where the majority of the population was effectively outside the political community, an identity that did not reach significantly beyond the elite was not a serious problem. Perhaps in a modern society under strongly authoritarian rule it might still be viable, though probably not indefinitely; under such conditions the twice-born might function as some kind of local equivalent of the Communist Party in contemporary China. But given the robust electoral politics of India since independence, and the disruption of traditional hierarchies, the old dominance of the upper castes was unsustainable—and has not been sustained. Marketing a Hindu political identity under these conditions has not been easy.

5. RALLYING HINDUS AGAINST MUSLIMS

Given all this, one might begin to wonder how Hindu identity could ever have stood a chance in the politics of modern India. Yet the political movement formed around it eventually achieved sufficient electoral success to serve more than once as the core of a governing coalition. There are no doubt many reasons for this success but only one that is so central both to the rise of the BJP and to our concerns in this study as to merit discussion at some length. This is the repeated efforts of the Hindu nationalists, with varying success, to rally Hindus against Muslims.

Before we come to these efforts, a glance at the Shiv Sena may be instructive. This organization is well acquainted with the importance of choosing one’s enemy wisely. When it was formed in 1966, the first enemy it chose for itself was the immigrant south Indian population of Bombay; but the southerners showed themselves to be bad enemies by learning to speak fluent Marāṭhī and putting up busts of the Marāṭhā hero Śivājī in their restaurants.244 The movement was then in some trouble until it discovered the Muslims in the 1980s, as the RSS had done long before, and borrowed the RSS rhetoric of Hindutva; however shallow and opportunistic its commitment to Hindutva may appear to members of the RSS,²⁴⁵ from then on Shiv Sena did not look back.²⁴⁶ To make do for now with a trivial example, on one occasion when its leader was concerned to make political noise, he demanded that two women in a film who develop a lesbian relationship be given Muslim names.²⁴⁷ Without question the change of tack has paid political dividends, and the moral is that Muslims make better enemies than southerners.

There are several reasons why Muslims make good enemies in the Indian context. One is that Muslims, whether they like it or not, are historically identified with the invaders who did most to destroy Hindu culture. Of course, not all the Muslim elite in India espoused this program assiduously, and many did not pursue it at all or indeed did the opposite. But there was enough rhetoric of destruction, and enough actual destruction, to lend support to a deep sense of Hindu grievance. For example, we have seen how Muslim cow killing had given rise to a centuries-old tradition of insult and injury;248 and whether or not the Muslims demolished a temple at the site of the Rām’s birthplace, they had unquestionably destroyed many others, for example, that of Somnāth.²⁴⁹ As to rhetoric, Sirhindī, who, as we saw, laid such emphasis on sacrificing cows to offend Hindus, tells us that the coexistence of Islam and unbelief is unthinkable and that one of them can only get ahead at the expense of the other. Within this zero-sum relationship, “to honour the one amounts to insulting the other.” The honor of Islam thus lies in insulting unbelievers, and anyone who respects them dishonors the Muslims.²⁵⁰ In mining the resulting vein of Hindu resentment, the Hindu nationalists were not particularly scrupulous; it would, for example, be hard to cite chapter and verse for the seventy-seven battles that according to the VHP had been fought by Hindus against Muslims at Ayodhyā to prevent the destruction of the temple and completion of the mosque.²⁵¹ But even if their scholarly standards had been higher, the Hindu nationalists would not have needed to be very selective.

A second reason is that since Partition the Muslims of India, again whether they like it or not, have been associated with Pakistan, the Muslim separatist state that broke the unity of undivided India, treats India with studied hostility, and has a track record of supporting its insurgent movements. The loyalty of India’s Muslims is accordingly easy to call in question. As a Shiv Sena politician put it in 1992: “India and Pakistan play a cricket game. The Indian team loses and Pakistan wins. Fire crackers go off in Bhendi Bazaar. Bhendi Bazaar [is a] Muslim area. That means what? Their loyalty is not for India but for Pakistan.”252 At the same time Partition encourages the sense that if India’s unity has been broken and the massive Muslim populations of the western Punjab and eastern Bengal have seceded, then at least what remains should be Hindustan, the land of the Hindus. As the ascetic Ritambarā puts it, “The Muslims got their Pakistan”; “Hindustan is ours!”²⁵³ The Shiv Sena politician followed his account of the cricket match by telling his interviewer that “this country is Hindustan. It is Hindustan of the Hindus. What is wrong with calling it Hindustan? Like Pakistan, which is a Muslim country. This is a country which belongs to Hindus. Communal riots will increase in the future.”²⁵⁴ From this perspective secularism appears as “a euphemism for the policy of Muslim appeasement.”²⁵⁵ This does not leave much room for Muslims in India, and according to one slogan shouted by demonstrators, Muslims belong either in Pakistan or in the cemetery (qabristān).²⁵⁶

A third reason is the existence of a sufficient number of Muslims with enough attitude to provide an inviting target—Muslims who do not behave in the manner of the Shiv Sena’s ill-chosen southern enemies, and who accordingly, in a favorite phrase of the Hindu nationalists, need to be taught a lesson.²⁵⁷ In 1987 the Muslim countermovement that developed during the struggle to defend the Bābrī Masjid obliged the Hindu nationalists by calling for a boycott of India’s Republic Day ceremonies.²⁵⁸ A couple of years earlier the “Shāh Bāno affair” of 1985, a legal fracas to which we will come in a later chapter, may have involved the largest mobilization of Indian Muslims since the partition of the subcontinent.²⁵⁹ External factors also helped, notably the global Islamic revival and the wealth accruing to Indian Muslims thanks to labor migration to the Persian Gulf.²⁶⁰ Even among the Tamil Muslims, a population traditionally at ease with the wider culture of their homeland,261 two recent songs include the tactless claim to have “ruled India for eight hundred years.”²⁶²

By way of illustration, one observer has given us a fine portrait of a Muslim with attitude in the southern city of Hyderabad, where the Muslim community is unusually large.²⁶³ This Akbar, a wrestler from a family of wrestlers, is the proud and violent champion of his community. Muslims, he tells his interviewer, are tougher than Hindus; in Hyderabad, at least, more Hindus than Muslims get killed in riots. Likewise Muslims are united, whereas Hindus are divided. Moreover Muslims have larger families than Hindus, and lower-caste Hindus convert to Islam. So the future, he declares, is on the side of the Muslims: “There won’t be many Hindus left.”²⁶⁴ Interestingly, the same observer also gives us long extracts from the speech given in Hyderabad in 1991 by the Hindu nationalist ascetic Ritambarā, who echoes the same themes from the other side.²⁶⁵ Hindus are tolerant—maybe too tolerant; the devotees of Gandhi are wimps, our current rulers are hermaphrodites. Muslims breed like mosquitoes and flies and have no use for family planning. There is a conspiracy to make Hindus a minority in their own country.²⁶⁶ The enemy in this rhetoric is not the Muslims as such: “Wherever I go, I say ‘Muslims, live and prosper among us. Live like milk and sugar.’ ” The enemy is the Muslim with attitude: “Is it our fault if he seems bent upon being a lemon in the milk?”²⁶⁷ In all this there is no disagreement between Akbar and Ritambarā over the facts. It is just that Akbar’s dream is Ritambarā’s nightmare—and both are projections of communal hatred.

Of course many Muslims are not militant, even in the face of strong provocation. A few years after the massive anti-Muslim riots in Bombay in the winter of 1992–1993, a Muslim businessman remarked: “The moment the Shiv Sena is out of power, Muslims are in danger. So we have to just try and keep them in power.”268 At the same time Muslim leaders adopted a policy of developing closer relations with the police, since, as one of them put it, “Now I realize that working with the police is the way to prevent another riot here”; leading police officials were invited to Muslim festivals, and small constructions were erected at street corners to provide shade for police constables.²⁶⁹ Muslims in Karnataka did not appreciate it when the Congress Party leadership sought their votes by laying on broadcasts from Bangalore in Urdu, thereby provoking anti-Muslim rioting in 1994.²⁷⁰ In 1979 Muslims in an industrial center in Bihar agreed to allow a Hindu procession to pass through the Muslim quarters, thereby temporarily upsetting the plans of the RSS—though in the event the organization succeeded in starting its riot all the same.²⁷¹ Muslim leaders in Bhopāl undertook to organize and pay for the repair of a Hindu temple damaged in the riots following the demolition of the mosque at Ayodhyā.²⁷² And one Muslim cleric argued against any attempt to secure the restoration of the Bābrī Masjid itself on the ground that this would only exacerbate hostility against Muslims.²⁷³ But from the Hindu nationalist point of view that still leaves enough Muslims asking to be taught a lesson.

A final reason why the Muslims made good enemies was that they were, in a way, just the kind of community the Hindu nationalists wanted to create. The Hindus, or so a Hindu nationalist lamented in 1923, are divided into numerous “water-tight compartments” that rarely think it worth their while to come to each other’s aid; the Muslims, by contrast, form “one organic community,” any part of which feels an injury done to another part.²⁷⁴ Muslim identity thus provided an image of what Hindu identity ought to be.

The Muslims were accordingly the obvious domestic enemy of the Hindu nationalists. They did not belong in the country. As M. S. Goḷwalkar (d. 1973), leader of the RSS from 1940, put it in a book of 1939, “The Hindus alone are the Nation,” while Muslims, “if not actually antinational are at least outside the body of the Nation.”275 They were welcome to “merge themselves in the national race and adopt its culture,” but if they did not wish to avail themselves of this offer, their only option was “to live at its mercy so long as the national race may allow them to do so and to quit the country at the sweet will of the national race.”²⁷⁶ Since then the tone of the Hindu nationalists has mellowed somewhat.²⁷⁷ Vajpayee in 1961 observed that the Indian Muslims were the descendants of Hindus and that “by changing religion one does not change one’s nationality or culture”;²⁷⁸ we will see more of such thinking later in this chapter.²⁷⁹ But a strong vein of hostility to Muslims has remained an endemic feature of Hindu nationalism. This brings us to a notorious aspect of urban politics in India: the communal riot.

These riots are typically the “orchestration of spontaneity.”²⁸⁰ Spontaneity matters, since even the most accomplished arsonist is doomed to fail if he has nothing inflammable to set a light to. Hence a robust heritage of communal mistrust, fear, and hatred greatly increases the chances of a violent outbreak; it would be a thankless task to try to foment a riot against Episcopalians in Princeton, New Jersey. In recent Western academic writing on Indian politics, as might be expected, there has been a tendency to downplay this role of the communal heritage.²⁸¹ Such a tendency is also widely shared, as an acute observer has pointed out, by “men of goodwill” in riot-prone towns and cities of India.²⁸² But it proves hard to adhere consistently and convincingly to such a position,²⁸³ particularly in the light of evidence that the phenomenon of communal riots antedates the coming of modern politics to India.284 What is, of course, true is that the communal heritage is not the same everywhere. Thus there has been a marked contrast between the north and south of India, with more hatred in the north.²⁸⁵ In December 1992 a poll showed 53 percent approving the demolition of the Bābrī Masjid in the north, but only 17 percent in the south.²⁸⁶ In the following year BJP representatives in the southern states were demanding that the national leadership talk about issues other than Ayodhyā because a communalist message “would not sell in the South.”²⁸⁷

What concerns us here, however, is the orchestration.²⁸⁸ It can be a fairly simple matter to start a communal riot. Hindu religious processions are a very convenient way to trigger them,²⁸⁹ as the founder of the RSS knew well from events that took place in Nāgpūr in 1923.²⁹⁰ For this purpose a procession needs to take a route through a Muslim quarter,²⁹¹ preferably timing its movements so that it passes by a major mosque during the Friday prayer (if there is a risk of arriving at the mosque too early, the procession can always slow down).²⁹² Such a procession is, of course, accompanied by musicians. In the good old days, when Hindus and Muslims participated in each others’ processions, the musicians would have been Muslims and would respectfully have fallen silent while passing the mosque;293 but these days they are Hindus and make a point of maintaining the decibel level.²⁹⁴ The Muslims can be expected to respond angrily to this provocation. They stream out of the mosque, or they may already be praying in the street because the mosque is overflowing with worshippers; they attack the procession, and a riot has begun. Processions were already being used, albeit on a more modest scale, to mobilize Hindus in the late nineteenth century. But there are, of course, other routes to the same result. In Bombay a ritual known as the Mahā Ārtī was developed into a stationary challenge to Muslim congregational prayers;²⁹⁵ since no procession was involved, the problem of timing was eased. Nationally, the agitation over the birthplace of Rām in Ayodhyā was associated with massive processions all over India, and at the same time it elevated the hated Bābrī Masjid and the projected temple of Rām into a source of synchronized communal tension for the entire country. When the news of the destruction of the mosque spread in December 1992, some of the ensuing riots were started by Hindu victory processions, while others arose from reactions to the demolition on the part of angry Muslims.²⁹⁶

The political advantage to be gained by Hindu politicians from a successful communal riot is clear enough.²⁹⁷ The point is to get people to identify first and foremost as Hindus and vote accordingly—for the BJP or, in the case of riots in Bombay, for the Shiv Sena. There is no doubt that Hindu nationalist politicians believe that communal riots can get out the Hindu vote for them, and students of Indian politics have noted many instances where the outcome suggests that they were correct.298 Thus in Gujarat, where there is a strong history of Hindu-Muslim riots, those of the early 1990s helped to bring the BJP to power for the first time in 1995.²⁹⁹ However, the government soon fell, owing to caste divisions within the party.³⁰⁰ Fresh riots in 2002 then gave the BJP a much stronger grip on power in the state.³⁰¹ So, under the right conditions the communal riot is a winning strategy. But, of course, conditions are not always right;³⁰² if they were, the BJP would by now have engrossed enough of the Indian vote to be able to rule the country without the need to accommodate difficult and demanding coalition partners.³⁰³ Nevertheless, conditions have been favorable in enough places for enough of the time to make the orchestration of communal riots a familiar implement in the Hindu nationalist toolbox.³⁰⁴

In their efforts to bring about the triumph of a Hindu identity, the Hindu nationalists fight on two fronts, the caste front and the Muslim front. Their underlying strategy could be summed up as recouping on the Muslim front their losses on the caste front. In this context it is worth noting that the agitation over the Bābrī Masjid, the most aggressive anti-Muslim campaign ever launched by the Hindu nationalists, was at the same time an attempt to change the subject by diverting attention from the implementation of the Maṇḍal report.305 The strategy can work well, as when Untouchables join in attacks on Muslims, as they did during the Bombay riots of 1992–1993, despite their antipathy toward the Shiv Sena.³⁰⁶ But it can also work badly, especially when the caste enemies of the Hindu nationalists join their Muslim enemies.³⁰⁷ At the Untouchable level we have already met a slogan proclaiming a Camār-Muslim alignment against the upper castes: “Jātavs and Muslims are brothers, where do the Hindu people come from?”³⁰⁸ Ironically, this invites comparison with Gandhi’s nightmare of 1932 during his fast-unto-death against a separate electorate for Untouchables: “ ‘Untouchable’ hooligans will make common cause with Muslim hooligans and kill caste-Hindus.”³⁰⁹ The same alignment was reasserted when the Camār politician Māyawatī extended reservations to low-caste Muslims in Uttar Pradesh in 1995 and reserved 8 percent of police officers’ posts for them; Muslims also appreciated her refusal to allow the VHP to target a mosque said to occupy the birthplace of Kṛṣṇa in Mathurā.310 At the OBC level Caraṇ Singh (d. 1987), the leading proponent of “peasant (kisān) politics,”³¹¹ in 1979 demanded the expulsion of the Hindu nationalists from the Janatā Party (an amalgamation of opposition parties formed in 1977) on the ground that they would cost it support among Muslims.³¹² In 1994, following the Bombay riots, the Samājvādī Party moved into the city through alliances with prominent Muslim businessmen; its leader, Mulāyam Singh Yādav, generously promised a large crowd in Bombay that he would guarantee the safety of Muslims in India.³¹³ He had some standing to do this, thanks to his role during the agitation over Ayodhyā: it was under his government in Uttar Pradesh in 1990 that large numbers of Hindu nationalists were arrested and a smaller number of them shot, earning him the nickname “Mullā Yādav.”³¹⁴ Allying with the group that Hindu nationalists saw as the internal enemy cannot have cost these politicians the support of their own constituencies or they would not have done it.³¹⁵

The importance of both fronts is indicated by the fact that each involves what by the standards of a Western democracy is a high level of violence. As we have seen, the characteristic form taken by communal violence is the riot; its counterpart in caste politics is the atrocity.³¹⁶ Caste atrocities are counted in thousands. Thus, in Maharashtra in 1994 the number of cases filed was over four thousand, involving rape, murder, and arson;³¹⁷ in Uttar Pradesh in 1995 it reached almost fifteen thousand.³¹⁸ In Bihar caste militias have been involved.³¹⁹ Usually it is the upper castes who take it out on the lower castes, but the contrary is not unheard of. After the assassination of Gandhi in 1948, there were lynchings of Brahmins in Maharashtra and riots that rendered thousands of them homeless.³²⁰ It would be of some interest to know how the number of fatalities in the sudden cloudbursts of communal rioting compares with those attributable to the steady drizzle of caste atrocities. And yet all this violence should not obscure the contribution of communal and caste tensions to a key structural fact about Indian politics: over time, as one observer has put it, the plurality of identities in play “prevents conflict from being concentrated along one particular fault line in Indian society”321—including the fault line along which the Hindu nationalists would most like to concentrate it.

Another testimony to the significance of the two fronts is the power of the emotional symbols used to rally the contenders, whether in “just and urgent defence” or to give “wanton offence” to their enemies.³²² In communal politics no symbolic issue was more potent than the scheme to demolish the mosque at Ayodhyā and erect a Hindu temple in its stead—though as an Untouchable villager in Uttar Pradesh told a reporter in 1991, “We can’t even enter the temple in the village, so what does the Ram Mandir in Ayodhya mean to us?”³²³ Of course, there were more overt ways to hit back at Rām: Māyawatī was delighted to convene a meeting in Lucknow in 1995 to honor the legacy of Periyār, the Tamil separatist author of the blasphemous pamphlet denouncing Rām as a villain³²⁴—a man who had also championed the conversion of Untouchables to Islam and the formation of Pakistan.³²⁵ He was to have had a huge statue in the new Revolution Square in Lucknow alongside Phule and Āmbeḍkar.³²⁶ The nearest equivalent to Ayodhyā in caste politics might be putting up and pulling down statues of Āmbeḍkar;³²⁷ Māyawatī erected fifteen thousand of them while in office in 1997.³²⁸ The parallels between caste alienation and communal alienation can go quite far. Āmbeḍkar had ideas of a Dalit homeland, though he did not take his dream as far as Jinnah was to take his.³²⁹ Likewise the slogans shouted by demonstrators may give Muslims the choice of Pakistan or the cemetery,330 or they may call for the Brahmin bastards (Brāhmaṇ sālā) to leave the land.³³¹ It all depends.

6. WHAT ARE THE HINDU NATIONALISTS?

So far in this chapter we have encountered the Hindu nationalists repeatedly without asking what they are. Calling them “Hindu nationalists,” of course, implies that they are a species of nationalist and thus comparable to nationalists in the Islamic world and elsewhere; religion, in other words, is secondary. But they are also sometimes seen as the counterpart of the Islamists, which is presumably what those who refer to them as “Hindu fundamentalists” have in mind.³³² Which is correct? A full answer to this question cannot be given at this point because it turns on the relationship of the movement to Hindu values, as opposed to Hindu identity. But certain points are already worth making here.

At first sight the role of Hindu religious figures in and around the movement might suggest that we are dealing with a religious rather than a nationalist phenomenon. Of these the most august are the Śankarācāryas, the heads of the major monasteries that act as custodians of the teachings of Śankara. In the twentieth century they came to form something of an all-Indian religious leadership, though by no means a well-coordinated one. It helped that they were above sectarian squabbles: one Śankarācārya of Dvārkā held the balance by installing 1331 phallic symbols (lingas) of Śiva on the one hand and 1220 black stones (śālagrāmas) with images representing Viṣṇu on the other.³³³ It also helped that their origins were not always local. One Śankarācārya of Purī was from Gujarat, one Śankarācārya of Dvārkā was from Karnataka.³³⁴ They were often good speakers in several languages and traveled widely throughout India; one Śankarācārya of Kāñcī did 1500 miles on foot.³³⁵ At the same time modern communications helped to put them in closer contact with each other. Admittedly, one result of this was a vocal dispute over the claim of the Kāñcī monastery to be an authentic foundation of Śankara’s; Śṛngerī and Kāñcī have been at loggerheads over this since about the middle of the nineteenth century, and the other Śankarācāryas have been drawn into the conflict.336 But increased contact has also meant increased cooperation, as meetings of the Śankarācāryas became more frequent in the second half of the twentieth century.³³⁷ Moreover, these leaders have come to terms with at least some aspects of the modern world. The Śankarācārya of Śṛngerī in the first half of the twentieth century was against modern education and Westernization,³³⁸ but the Śankarācārya of Dvārkā in 1960 established an Indological research institute affilated with a major university.³³⁹

The Śankarācāryas were thus well placed to advance the Hindu nationalist cause, and to an extent they did so—but only to an extent. In 1966 the Śankarācārya of Purī fasted for seventy-three days in the course of a massive agitation for cow protection mounted by the Hindu nationalists—and was jailed for his pains.³⁴⁰ The Śankarācāryas of Purī were in fact precociously political,³⁴¹ and of the major Śankarācāryas they were perhaps the least unreliable as allies of the Hindu nationalists.³⁴² But even they could be loose cannons: in 1969 the same Śankarācārya of Purī angered the Maharashtrian Buddhists by asserting on a visit to Poona that untouchability was an integral part of Hinduism.³⁴³ The general unreliability of the Śankarācāryas was particularly apparent in 1993, when four of them held a meeting at Śṛngerī and agreed on a compromise proposal to resolve the issue of the Bābrī Masjid, thereby undermining the position of the VHP.³⁴⁴ Admittedly the relationship had its moments: in an emotive scene played out in 1990, a minor local Śankarācārya joined the VHP at a mass meeting in Bhopāl and handed out medals to volunteers (karsevaks) returning from the agitation in Ayodhyā.345 But this was untypical of a complicated and mutually rather unsatisfactory relationship.

Less prestigious religious figures—ascetics (sādhus) and others—have also played their part and in much larger numbers.³⁴⁶ They too have not been very dependable allies³⁴⁷ nor altogether welcome when they have displayed too direct an interest in getting into politics.³⁴⁸ Their support can nevertheless extend the reach of the movement on the ground and perhaps help to confer a measure of religious legitimacy on it—though one should not overestimate this latter effect, since Indian public opinion takes almost as dim a view of religious leaders as it does of politicians.³⁴⁹ In any event, it is noteworthy that these saffron-clad religious figures, like the Śankarācāryas, have mostly remained on the periphery of the movement. Umā Bhārtī, who owes part of her prestige to her familiarity with the Rāmāyaṇa of Tulsīdās,³⁵⁰ wears saffron, and she is not the only BJP politician to do so.³⁵¹ But the activists of the RSS and even the VHP wear white, not saffron.³⁵² In short, the role played by religious figures is not evidence that the fundamental character of the movement is religious; the men in saffron are not the counterparts of the Iranian clergy.

At the same time the presence of non-Hindus in the movement provides positive evidence that it is not religious in a way that would make it comparable to Islamism. Thus Jains play a minor but persistent role.³⁵³ Of the twelve volunteers who received medals at the mass meeting staged by the VHP in Bhopāl in 1990, two were Jains.354 We find Jains holding office not just in the RSS³⁵⁵ but also in the VHP.³⁵⁶ The BJP nominates Jain candidates,³⁵⁷ rich Jains may contribute to its funds,³⁵⁸ and Jain holy men may support it.³⁵⁹ There were Jains among those carrying holy water during the “pilgrimage” (yātrā) organized by the VHP in 1983,³⁶⁰ and Jain Studios played a role in broadcasting Hindu nationalist propaganda.³⁶¹ Just as noteworthy is the lack of any reference to tension in this regard—even hidden tension of the kind that has affected the participation of Christian Arabs in Arab nationalist movements.³⁶² Jainism, after all, is a religion that broke with the Hindu mainstream some twenty-five centuries ago, rejecting the Vedas and the rituals associated with them. Indeed Dayānand, for whom such things mattered, saw the Jains as alien intruders who could be blamed for corrupting Indian religion by introducing idolatry³⁶³—and this despite the fact that he came from a region in which Hindus and Jains lived in a peculiarly intimate symbiosis.³⁶⁴ But the Hindu nationalists do not think in this way. Sāvarkar, for example, appealed to the example of the Jains to prove his point that “a man can be as truly a Hindu as any without believing even in the Vedas as an independent religious authority,”³⁶⁵ and he listed Mahāvīra, the founder of Jainism, as a paradigmatic Hindu.³⁶⁶ It is as if the Muslim Brothers were to see no problem in recruiting the local Christians, Jews, and Samaritans. What this suggests, of course, is that for the Hindu nationalists—in marked contrast to the Muslim Brothers—religion is a culture rather than a creed; for despite its credal incompatibility with Hinduism, Jainism is unmistakably a product of the same Indian cultural milieu. Even Dayānand must have shared this attitude at some level: a couple of years before his death he was investing members of the higher castes in Rajasthan with the sacred thread, and half of them were Jains.367

The underlying attitude of the Hindu nationalists to Buddhists and Sikhs seems no different, though surface features tend to overlay it. In the Buddhist case the survival of the religion down the centuries was associated with regions outside India; when Untouchables reintroduced it in the twentieth century, this marked a deliberate break with the very community that the Hindu nationalists stand for. In other words, the trouble with Buddhism is not that it split off from Hinduism almost as long ago as Jainism but rather that its modern Indian reincarnation took place in a divisive caste context. Thus the Shiv Sena sides with those Untouchables who are “proud to be Hindu,” in contrast to the Buddhists among them.³⁶⁸ The Buddha is nevertheless a revered figure in the Hindu nationalist pantheon. In his book of 1939, Goḷwalkar referred to him as “the Great Master,” and spoke of “the spiritual awakening under Lord Buddha.”³⁶⁹ A later Hindu nationalist author tells us that Lord Buddha came at a time when “addiction to religious practices and blind adherence to tradition brought society to a state of stagnation”; he accordingly “battled with the distortions that had crept into Hindu religion.”³⁷⁰ The RSS “morning meditation” (prātaḥsmaraṇam) invokes him in a long list of Indian worthies.³⁷¹ Ritambarā, the Hindu nationalist ascetic, begins a flight of oratory by hailing the Buddha among others, though we hear no more of him once she shifts her saintly attention to the Muslims.³⁷² And it was the Dalai Lama who opened the conference held by the VHP in Allāhābād in 1979,³⁷³ for all that he is neither an Indian nor a Hindu. One Hindu nationalist author describes how the distinguished visitor was met at the railway station with Vedic chanting and later reciprocated by declaring that Buddhism was “truly a product of this soil.”³⁷⁴

In the Sikh case the complicating circumstance was again historical rather than doctrinal. Unlike common or garden Hindu sects, the Sikhs had acquired a military base among the caste of peasant warriors known as the Jāṭs, and in the late eighteenth century they had used it to establish a state. Hence no doubt their tendency—already apparent in the late nineteenth century—to see themselves as a non-Hindu community,375 culminating in the separatist terrorism that afflicted the Punjab in the 1980s. Dayānand, of course, insisted on doctrine: he regarded the Sikhs as in effect idolaters,³⁷⁶ and Ārya polemics against Sikhism soon put an end to the early presence of Sikhs in the Ārya Samāj.³⁷⁷ But again, the Hindu nationalists do not seem to have been bothered by considerations of creed. For Sāvarkar, “Sikhs are Hindus in the sense of our definition of Hindutva and not in any religious sense whatever.”³⁷⁸ Instead, they were accorded an iconic status thanks to their historical record of fighting Muslims. Goḷwalkar had nothing but praise for “that band of unconquerable Hindu heroes, the Sikhs, headed by their immortal Gurus.”³⁷⁹ Hence the Hindu nationalist reaction to Sikh terrorism in the Punjab—which included the slaughter of considerable numbers of RSS activists and the ethnic cleansing of Hindus—was marked more by sorrow than by anger.³⁸⁰ Sikh devotees of the movement were sent to reason with the terrorists,³⁸¹ and the RSS took pride in the role played by its members in sheltering Sikhs during the disturbances that followed the assassination of Indira Gandhi in 1984.³⁸² In earlier, happier days there had been Sikh representatives present at the founding congress of the VHP held in 1964.³⁸³ Even in 1983 Sikhs shared with Jains the task of carrying holy water in a pilgrimage organized by the VHP.³⁸⁴ In short, the Hindu nationalists share the wider indifference of Indian society to credal differences between culturally Indian religions.

If credal differences do not matter, it becomes an interesting question whether the movement can in principle accommodate Muslims. This is a delicate issue. There are in fact Muslims in the movement, even in the core membership, though as one might expect they are not numerous. One Muslim, Sikander Bakht, has risen high in the BJP, though not through the RSS,385 and the party sometimes nominates Muslim candidates in elections.³⁸⁶ Of the twelve volunteers recognized in Bhopāl in 1990, one was a Muslim—and received the warmest applause.³⁸⁷ The author of a good-news book on the activities of the RSS published in 1988 remarked with satisfaction that instances of Muslim boys and youths attending meetings at RSS branches were “though not very common, not rare,” and gave a couple of examples. One was an apparently troubled Muslim youth who joined the movement while his father was absent in Dubai; when his father heard this alarming news, his wife was able to reassure him that, thanks to the care of the local RSS leader, “Aziz has improved a lot.”³⁸⁸ But as might be expected, many in the movement were not keen to include Muslims. When an elderly Gandhian idealist invited the RSS to open itself to Muslims and Christians in a speech of 1977, the response was tepid.³⁸⁹ In 1982 there was disapproval in the ranks of the RSS when BJP politicians sought to cultivate Muslims by participating in their prayer rituals.³⁹⁰ And in 1993 local activists sabotaged the electoral campaign of a Muslim candidate.³⁹¹ But we are concerned here mainly with principle, and at this level too there has been a significant divergence of views.

One view, which we have already met in Goḷwalkar’s book of 1939, categorically excluded Muslims from the nation and a fortiori from the movement.³⁹² The general question Goḷwalkar was addressing was “what is to be the fate of all those, who, today, happen to live upon the land, though not belonging to the Hindu Race, Religion and culture?”—a category in which the Muslims clearly predominated. His answer was that such people “can have no place in the national life, unless they abandon their differences, adopt the religion, culture and language of the Nation and completely merge themselves in the National Race.”³⁹³ Sāvarkar had already considered the question in his book of 1923. His first line of argument was that the exclusion of Muslims was not a matter of “any religious beliefs that we as a race may hold in common” but was rather due to the fact that they do not share “our Hindu culture.”394 But what if they did share Hindu culture? Here he went on to introduce a specifically religious criterion: a Hindu was someone for whom the country was not just a fatherland but also a holy land (puṇyabhūmi). Thus the reason why even Muslims sharing a fatherland and most of their culture with Hindus could not belong was that their holy land was “far off in Arabia”—“their love is divided.”³⁹⁵ They were, of course, more than welcome to return to the Hindu fold.³⁹⁶

Against this background Sāvarkar’s polite nod to the idea that Hindus, Muslims, and the rest might come to feel “as Indians first and every other thing afterwards” seems unconvincing.³⁹⁷ But such a view has certainly existed in the movement—the view, that is to say, that there can be such a thing as a good Muslim. A good Muslim shares Hindu culture and reveres the heroes of the Hindu epics as national symbols, just as Sikander Bakht does.³⁹⁸ One high official of the RSS in 1993 held up the Javanese Muslims as an appropriate model: “Islam did not prevent Indonesian Muslims to proclaim Mahabharat and Ramayan as their cultural epics and Ram and Krishna as their forefathers.”³⁹⁹ A VHP convention organized in Hardvār in 1988 is said to have demanded a little more: if Indians regard Rām and Kṛṣṇa as “national heroes,” and touch India’s sacred ground three times in prostration while saying, “Hail, Mother India!,” then they may belong to “any religion”⁴⁰⁰—a view that invites comparison with the attitude of imperial Rome toward its Christian minority. In 2011 a Hindu nationalist politician was happy to include in the nation Muslims who “acknowledge with pride that though they may be Muslims, their ancestors were Hindus.”⁴⁰¹ A similar benevolence toward good Muslims is found among the Shiv Sena.402 One follower of the movement strongly approves of the Muslims of the Konkaṇ as “Maharashtrians in their heart” who “speak Marathi” and “even dress like Hindus”; he contrasts them with other Muslim groups who “keep to themselves” so that one never knows “where their loyalties lie.”⁴⁰³ In this way the Shiv Sena too is opposed only to “antinational” Muslims; indeed one of its Muslim adherents has been minister of labor,⁴⁰⁴ a position that one might expect to confer considerable patronage resources. Of course, many, perhaps most, Muslims are not by Hindu nationalist standards good Muslims, but that is not the point.

In short, despite the hard line taken by the patriarchs of the movement, if someone has the right cultural loyalties, then religion as such is not of overwhelming importance to the Hindu nationalists. It is in this sense that the BJP’s rhetoric of “unity in diversity” has in principle a potential to include even Muslims.⁴⁰⁵ Indeed what finer icon of cultural diversity could there be than the projected Sanskrit translation of the Koran, announced by the Minorities Cell of the BJP in 1995 as a way of building “a bridge to the Muslim masses”?⁴⁰⁶ (Presumably it will not include the verses that Umā Bhārtī assures us instruct Muslims to skin idolaters alive and “stuff them in animal skins.”)⁴⁰⁷ And even in his categorical excision of Muslims from the nation, Sāvarkar introduced only the bare minimum of religious belief needed to exclude them. In all this what we see is the logic of cultural nationalism, not the logic of religious fanaticism.

So far we have looked at the role played in the movement by religious figures on the one hand and non-Hindus on the other. But these phenomena, though instructive, are somewhat peripheral, and we should now turn to the movement at large. Its members are certainly not alienated from religion. From the start the RSS had significant features in common with Hindu sects,408 and it possessed a ritual calendar mixing traditional Hindu festivals with nationalist elements relating to the figure of Śivājī.⁴⁰⁹ Yet, as one commentator has aptly observed, the literature of the RSS has remarkably little to say about religious beliefs and does not waste much time on deploying proof-texts or asserting the authority of a scriptural canon; moreover he reports that “in private conversations its members frequently emphasize that they are not religious.”⁴¹⁰ This was a movement that in its early days owed much to European nationalism, Italian and German,⁴¹¹ while the Sunday parade of its original devotees was marked by the use of English martial music and drill commands.⁴¹² Perhaps the symbol that best captures what these devotees were about is the Śivājī cult: their day began with a salute to the Bhagvā Dhvaj,⁴¹³ the saffron flag particularly associated with Śivājī, an emblem that “brings before our eyes the living image of our ancient, sacred and integrated national life in all its pristine purity.”⁴¹⁴ The members of Shiv Sena likewise sport the saffron flag.⁴¹⁵ As their leader explains the “Shiv” in “Shiv Sena”: “In politics it’s Shivaji, in religion it’s Shiva.”⁴¹⁶ And indeed they honor Śiva by equipping themselves with the trident (triśūl), his traditional weapon.⁴¹⁷ Yet despite such elements of religiosity, it would be hard to mistake either the RSS or the Shiv Sena for a religious movement.

If there is a specifically religious organization among the Hindu nationalists, it is the VHP.⁴¹⁸ It has indeed concerned itself with properly religious questions, and as we will see it has sought to reshape Hinduism in significant ways, both by organizing it and by seeking to identify elements of the religion that everyone can agree on.⁴¹⁹ At this point we need remark only that the religiosity of the VHP does not seem to be out of line with that of the Hindu nationalist movement in general. A telling example is the initiative of the VHP in summoning assemblies to deliberate on the state of Hinduism in 1966 and 1979.420 Both times it held them at Allāhābād in Uttar Pradesh.⁴²¹ There was an explicit historical reference here.⁴²² Allāhābād is the ancient Prayāg, a Hindu holy place at the confluence of the Ganges and the Jumna, and it was there that Harṣa, the last native ruler of an imperial state in northern India, held his quinquennial assemblies in the first half of the seventh century.⁴²³ But quinquennial assemblies were a Buddhist, not a Hindu tradition, for all that Brahmins were also present at Harṣa’s convocations.⁴²⁴ As this indicates, the VHP’s project is primarily a nationalist one;⁴²⁵ it is in line with the tradition of the early nationalist Ṭiḷak, who wanted “to consolidate all the different sects into a mighty Hindu nation”—not a sublime Hindu religion.⁴²⁶

One final point of contrast between Hindu nationalists and Islamists concerns the transnational dimension of the identity they champion—or the lack of it. For Islamists the worldwide Muslim community is a primary focus of loyalty irrespective of political boundaries. Hindus lack a wider community on this scale; in contrasting the situation of the Muslims and Hindus of India, an early member of the Ārya Samāj observed that “Mohammadans have Constantinople behind their back,” whereas the Hindus “are circumscribed within the four walls of Hindustan and have no outside assistance to influence the attitude of their rulers.”⁴²⁷ There are, however, significant Hindu populations outside India in several countries of South Asia, namely Pakistan, Nepal, Bangladesh, and Sri Lanka, and there is a considerable Hindu diaspora in other parts of the world. We could thus imagine a Hindu movement aiming to unite Hindus wherever they may be found, stirring up the Hindu populations of the smaller South Asian countries and feeding on tensions between the Hindu diaspora and its host societies. But such a movement does not exist. The VHP is very interested in the Hindu diaspora, particularly the part of it that resides in the First World—and for good reason: it needs to raise money.428 But we do not hear persistent complaints about Hindu nationalists making political trouble among the Hindu populations of Nepal, Sri Lanka, Fiji, or Surinam, let alone Bali; Nepal in particular, during the decades when it was the world’s only surviving Hindu monarchy, seems to have functioned more as an icon and a ceremonial stage than as a place to engage in serious political work.⁴²⁹ In any case, the political horizons of the Hindu nationalists do not reach beyond the boundaries of India as defined by Śankara’s monasteries.⁴³⁰ And even in the case of the territories lost to India through Partition, there has been a notable lack of irredentism—hardly what one would expect, given the strength of the movement’s original commitment to an undivided India.⁴³¹ In sum, the aspiration behind the Hindu nationalist invocation of religion is to redefine the nation within the current borders of India, not to transcend it; this is religion in the service of nationalism, not religion sweeping nationalism away.

How then should we characterize the conception of Hindu identity prevalent among the Hindu nationalists? Obviously they are not secular nationalists in the style of Nehru, but neither are they true believers in the manner of Dayānand. One way to describe them, and an adequate one, is to call them “cultural nationalists,” meaning nationalists for whom the nation is made up of the bearers of a particular cultural tradition.⁴³² It was not just wishful thinking when an RSS activist from an Untouchable background averred that there was “no caste and no religion” in the RSS, the only religion being Mother India⁴³³—where Mother India (Bhārat Mātā) is a modern nationalist addition to the Hindu pantheon.⁴³⁴ This was also how the patriarchs of the movement saw things. Thus Sāvarkar, who had a seminal influence on the RSS, minimized the religious component of Hindutva in his book of 1923, putting most of his emphasis on a shared fatherland, blood, and culture;435 “Hinduism,” he wrote, “is only a derivative, a fraction, a part of Hindutva.”⁴³⁶ He did not eliminate religion altogether: as we have seen, he made recognition of India as a holy land a litmus test, and in that respect Hinduism had a key role to play in his idea of Hindutva.⁴³⁷ But he was consistent in emptying the notion of any credal content: “Some of us are monists, some, pantheists; some theists and some atheists. But monotheists or atheists—we are all Hindus and own a common blood.”⁴³⁸ It is hard to imagine an Islamist saying anything like this. Goḷwalkar in 1939 sounded more like an Islamist. “With us,” he wrote, “every action in life, individual, social or political, is a command of Religion.”⁴³⁹ But unlike an Islamist, he had no problem with the “variety” comprehended within the “organic whole” of his religion;⁴⁴⁰ and most tellingly, the reason he gave for not giving up “religion in our National life” was that “it would mean that we have turned faithless to our Race-Spirit.”⁴⁴¹ Like Sāvarkar, he showed no interest in the question whether traditional Hindu beliefs are true.

What then was the point of adopting a cultural nationalism in the Indian context? Here the key contrast is with the secular nationalism of Congress, not with Islamism. The Indian nationalist movement that developed under the rule of the British and came to power on their departure was a secular one. This feature had its origin in the nationalist package imported from Europe, but in the Indian environment it proved adaptive to the extent that it made possible an appeal to all Indians irrespective of their religious affiliations: “We are Indians first,” went a statement of the message dating from 1905 “and Hindus, Mahomedans, Parsees or Christians afterwards.”⁴⁴² But unlike the secular nationalisms of Europe, this inclusive Indian nationalism was not grounded in a strong sense of ethnic community; the subcontinent was marked by an ethnic diversity more reminiscent of Europe as a whole than of any single European country. Secular nationalism in the Indian context was thus characterized by a certain emotional threadbareness, largely based as it was on a combination of the geographical unity of the subcontinent and the shared exposure of its inhabitants to British rule.443 “The Congress,” as it said about itself in 1886, “is a community of temporal interests and not of spiritual convictions.”⁴⁴⁴ This pragmatic conception of the nation led Goḷwalkar to complain that Congress was turning India into one big hotel: “Indeed they have made our country a veritable serai!”⁴⁴⁵ The eventual departure of the British may have been a triumph for the secular nationalists, but it could only exacerbate the problem of threadbareness.

An alternative was to cultivate a nationalism that foregrounded a religious heritage as a national culture. The pull of such a cultural nationalism was already evident in the last decades of British rule and not just among the Hindu nationalists and their precursors; perhaps more significant at the time was the presence of such attitudes within Congress itself. Though relatively inhibited at the center, they were widespread at the local level. Thus, a study of the politics of the United Provinces in the last decades of British rule shows how persistently Congress politicians made use of the Holī festival, the Rāmāyaṇa, and cow protection for the purpose of mobilization.⁴⁴⁶ It would have been politically irrational for them to forego the use of these resources, but even with the best intentions (and intentions were not always the best), such a repertoire tended to suggest that Muslims were not part of the nation. The contrast between the central and local manifestations of Congress helps to explain how the members of the same party could appear as self-hating Hindus to the Hindu nationalists and as Hindu chauvinists to many Muslims. Independence and its aftermath then benefited the Hindu nationalist cause in two practical ways. First, the secession of the two main Muslim-majority regions of the subcontinent—what we can call the hard partition of India—considerably reduced the need for a nationalist doctrine that sidelined religion. As we have seen, once the Muslims had their Pakistan, India came to look a lot more like Hindustan, thereby increasing the plausibility of Goḷwalkar’s insistence in 1939 that “in Hindusthan, the land of the Hindus, lives and should live the Hindu Nation.”447 Second, the restructuring of India as a federation of linguistic states—the soft partition—averted the danger of political disintegration under the impact of competing ethnic nationalisms, though at the cost of dimming the prospects of Hindi as an indigenous national language.⁴⁴⁸ Thus, Hindu nationalism figured as the only credible alternative to secular nationalism if India was not to dispense with nationalism altogether. It is this recycling of a religious heritage as cultural nationalism that has won for the Hindu nationalists their current role in Indian politics. On the one hand it has enabled them to achieve a centrality that eluded true believers like Dayānand, and on the other hand it has allowed them to take full advantage of the inhibitions of Congress in relation to the Hindu heritage. What it has not given them is dominance of the political system: too many Indians have reason to feel more comfortable with Indian nationalism that with Hindu nationalism.

Before we leave South Asia, it is worth pausing to identify the counterparts of the Hindu nationalists across the border in Pakistan. It should be clear from what precedes that they are not the Islamists. They are, in fact, the politicians of the Muslim League, those who brought Pakistan into existence and ruled it off and on thereafter.⁴⁴⁹ Like the Hindu nationalists, the members of this elite were not particularly religious, but they were strongly identified with the cultural and political fortunes of a religiously demarcated community. At first they also resembled the Hindu nationalists in another significant respect: exacerbating the tensions between Muslims and Hindus with cries that Islam was in danger could serve to divert the Muslim vote from their competitors.⁴⁵⁰ There was even a minor Muslim League analog of Ayodhyā at Sukkur in Sindh. Here a domed building, the Manzilgāh, had been claimed by the local Muslims to be a mosque since 1920 (the Hindus had no claim to it themselves but wanted the Muslims kept out). In 1939 the cause was taken up by the Muslim League in order to undermine the position of the communally mixed government then in power in Sindh. The agitation involved Muslim volunteers congregating in Sukkur, the seizure of the building by Muslim demonstrators despite a police presence, and several days of communal rioting following the eventual expulsion of the demonstrators; the Hindus appear to have started this outbreak by jeering at the Muslims when they were expelled, but the loss of life was mainly on the Hindu side.451 After independence, however, such strategies tended to lose their effectiveness in Pakistan.⁴⁵² One reason for the divergence between the two countries may have been demographic, namely the absence of a politically significant Hindu minority in West Pakistan; by 2000, there were fewer than three million Hindus in Pakistan, though many more in Bangladesh. Another reason was the weakening of electoral democracy. Unlike India, Pakistan was dominated by an ethnic core that was unwilling to accommodate other ethnic groups through an Indian-style federal system. Moreover, the bottom line of the Punjabi hegemony was the army, which unlike the Indian armed forces, has intervened three times in politics to establish long-lasting military regimes. Only during the Emergency of 1975–1977 did the politics of India look at all like those of Pakistan, and even then the reaction of the Indian politicians was to anticipate the resumption of politics as usual. For both these reasons, playing the old communal card came to be far less advantageous in Pakistan than in India.

This was not, however, the only way in which the cultural nationalists of Pakistan found themselves in a very different political environment from their Indian counterparts. For one thing, the secular, territorial conception of nationalism that was dominant in Nehru’s India could have no viable equivalent in Pakistan. There had, of course, been Indian Muslims who thought like Nehru, such as Sir Sayyid Aḥmad Khān (d. 1898), for whom “the words Hindu and Mahomedan are only meant for religious distinction,” so that all inhabitants of the country should be regarded as “belonging to one and the same nation.”⁴⁵³ Even Jinnah could sound like this on occasion: in a speech of 1947 to the Constituent Assembly of Pakistan, he looked forward to a time when, in a political as opposed to a religious sense, Hindus would cease to be Hindus and Muslims to be Muslims; both alike would be citizens of the state.454 But this kind of thinking did not make much sense in Pakistan, a country created for a community whose “bedrock and sheet-anchor” was Islam.⁴⁵⁵ For another thing, the rulers of Pakistan had to share their country with Islamists, for whom the political claims of Islam went far beyond the creation of a mere Muslim state, in the sense of a national state for the benefit of Indian Muslims; as we have seen, the Islamists and their Islamic state had no Hindu counterparts.⁴⁵⁶ In short, Jinnah and his successors differed from the Hindu nationalists in having no Nehru to their left and Mawdūdī to their right.

In each country we can accordingly think of the cultural nationalists as clinging to a slippery slope—but with the slopes facing in opposite directions. To hold power, Hindu nationalist politicians are obliged to do deals with people of more secularist inclinations than their own.⁴⁵⁷ Hence such projects as the construction of the temple of Rām in Ayodhyā get shelved, as happened in the coalition building of 1998.⁴⁵⁸ To quote a comment made at that point by the leading academic expert on the Hindu nationalists: “In some ways, the BJP appeared to have adopted the role of the post-independence Congress Party by providing a relatively neutral pivot around which other groups could arrange themselves”⁴⁵⁹—an outcome that, taken on its own, would place the BJP well on the way to the bottom of the slope. Pakistani politicians, by contrast, are obliged to co-opt or placate people more religious than themselves, resulting in a persistent drift toward Islamization. One result of this has been increasing deference to Islamic law, another the rising level of sectarian conflict among Muslims. Whereas in 1953 the politicians held the fort in the face of religious agitation against the Aḥmadīs, they were unable or unwilling to do so in 1973–1974;460 and since the late 1970s, for reasons partly domestic and partly foreign, sectarian conflict has extended to relations between Sunnīs and Shīʿites.⁴⁶¹ One might contrast this with the absence of any revival of traditional tensions between Śaivas and Vaiṣṇavas in the politics of India.⁴⁶² No commentator on modern Indian politics ever emphasizes the importance of not alienating the Śaiva vote or speculates about which of the two sects might be more likely to form an alliance with the Muslims against its rival.⁴⁶³ All in all, it is hard to say which slope is more slippery, and in neither country have the cultural nationalists yet reached the bottom. We tend to have a greater awareness of the Pakistani case, where the results look malign, than of the Indian case, where the results, if we notice them, look benign. The net effect is that in each country the cultural nationalists are in recurrent danger of being outflanked, but on different sides. The findings of a survey carried out in 2006 shed an interesting light on this. In Pakistan 87 percent of Muslims identified as Muslims first, rather than citizens of their country; in India only 10 percent of Hindus identified in this way.⁴⁶⁴

7. CONCLUSION

In the course of the millennia of its history, India had acquired a certain diffuse unity. It was not an ethnic unity, though its ethnic divisions were at least relatively shallow. Nor was it a religious unity: there was no sharp sense of who was and was not a Hindu, and indeed the very idea of a Hindu was of foreign origin. Nevertheless, a number of religious themes—such as the veneration of the Vedas and the Rām cult—were widespread in the subcontinent, and there were elements of the religious tradition that staked out a clear territorial definition of the country. Yet this was far less than would have been needed to make India a plausible European-style nation. Here the Hindu nationalists offered a way to bridge the plausibility gap in the form of a cultural nationalism that highlighted a religious heritage. In this respect they had something to offer that Congress with its commitment to secularism could not provide—or at least not overtly. The major problem the Hindu nationalists faced in marketing their conception of Hindu political identity was the way the caste system worked against them. The tactless manner in which the Hindu tradition articulated its vision of the subjection and exclusion of the lower castes was in itself an incitement to its victims to defect under modern conditions, to walk away from the caste ladder rather than try to climb it. Such defection could be articulated in either ethnic or religious terms. Moreover, the increasingly extensive system of reservations exacerbated the situation: the more the underprivileged castes stood to benefit from horizontal solidarity, the less they were attracted to vertical ties with upper-caste patrons. This meant that the lowest castes—the Untouchables—had little reason to revisit their traditional sense of exclusion from Hindu society, and those immediately above them—the OBCs—lacked motivation to behave as political Hindus and so forego the candy of reservations. It was thus an ambitious program for the Hindu nationalists to seek to create a single homogeneous family. The solution to their caste problem, insofar as there was one, was to rally Hindus against their historic Muslim enemy, orchestrating the spontaneity of communal tensions to win elections. Sometimes they have used this strategy with success, but at other times they have seen their caste enemies—be they Camārs or Yādavs—ally with the Muslims against them. All this meant letting religion flood into politics. But at least in the field of identity—which is all that we have looked at so far—this did not mean the development of a Hindu emulation of Islamism. As we have seen, the Hindu nationalists had no problem with welcoming into their movement non-Hindus with the right cultural credentials, be they Jains, Buddhists, or Sikhs—perhaps even good Muslims, though this was rare and contested. Had the Hindu nationalists been Muslim politicians in Pakistan, they would have had just as little trouble including Shīʿites and Aḥmadīs. In other words, they were the counterparts of the Muslim League, not of the Islamists, and unlike the Islamists, they adhered to the broad global consensus that modern political identity has to be national. Their problem was that large numbers of Indians—not just Muslims—still found their particular brand of nationalism unappealing.

___________

1 See 94n266.

² So the Vilasa Grant of around 1330 (Venkataramanayya and Somasekhara Sarma, “Vilasa grant,” 241, 260 lines 10–11; for the dating, see 256). The passage is cited in Chattopadhyaya, Representing the Other, 58.

³ Data from the census of 2001, http://censusindia.gov.in/Census_Data_2001/Census_Data_Online/Language/Statement6.htm.

⁴ Parthasarathy, Cilappatikāram; see Zvelebil, Tamil literature, 114. The poem survives in at least twenty-five manuscripts.

⁵ Parthasarathy, Cilappatikāram, 225.

⁶ Ibid., 225, 233–36, 238, 253–54, 258.

⁷ Ibid., 207, 236.

⁸ Ibid., 237.

⁹ Pollock, Language of gods, 475.

¹⁰ Ibid., 476, and more generally 468–77, of the section on “European particularism and Indian difference.” He notes the contrast between the way things were seen in India and the epigram of a tenth-century Christian poet that “a language makes a people” (474)—an epigram that, as we have seen, would definitely have meant something to Ibn Taymiyya (see above, 15).

¹¹ Cf. ibid., 475–76: “No doubt social forces such as caste endogamy … worked against integrative ethnicity.”

¹² For evidence that this process was already well advanced by the time the British arrived in India, see Lorenzen, “Who invented Hinduism?,” 648–53.

¹³ The word appears several times in various spellings in the hagiographical poem Kabīr parcaī of the Rāmānandī ascetic Anantadās (fl. around 1588). See Lorenzen, Kabir legends, 166 (7.2 = 107), 167 (7.5 = 107), 170 (7.14 = 109), 171 (7.15 = 109), 205 (13.3 = 125). For occurrences of the term in the earliest manuscript (dating from 1636), see 230 (80), 231 (85), 233 (92), 251 (182). See also Callewaert, Hagiographies of Anantadās, 73 (7.1), 74 (7.4), 76 (7.14), 93 (13.3). For the date, affiliation, and location of Anantadās, see 1, 31–32; for the manuscript of 1636, see 23. Cf. Vaudeville, Weaver named Kabir, 152, 154, 227, 233, 240, for the term “Hindu” in translations of the poetry of Kabīr himself. See also Lorenzen, “Who invented Hinduism?,” 649–51, for the use of the term “Hindu” by both poets.

¹⁴ O’Connell, “Word ‘Hindu’ ,” 340–44, cited in Jaffrelot, Hindu nationalist movement, 4.

¹⁵ Wagle, “Hindu-Muslim interactions,” 139–41 (quoting from a sixteenth-century Hindu text from Maharashtra describing a Hindu-Muslim debate, for which see also Lorenzen, “Who invented Hinduism?,” 648–49); Wagle and Kulkarni, Vallabha’s Paraśarāma Caritra, 2.49 (33 = 155–56), 3.10 (60 = 40), 4.34 (46 = 73), 4.45 (47 = 74–75; this work dates from the late eighteenth century, see 2). Cf. also Raeside, Decade of Panipat, 39, 53, 72, 95, 98 (translating a chronicle probably dating from the late eighteenth century, see xi), 125 (translating a letter of 1757); Gordon, “Maratha patronage,” 334 (“Hindu” in a Marāṭhā documentary context, 1760).

¹⁶ O’Connell, “The word ‘Hindu,’ ” 342b, 344a; Wagle and Kulkarni, Vallabha’s Paraśarāma Caritra, 3.26 (41 = 63); also Laine, Shivaji, 51; and cf. Raeside, Decade of Panipat, 53, 72, 98.

¹⁷ Wagoner, “Sultan among Hindu Kings,” 861–63 (drawn to my attention by Sanjay Subrahmanyam).

¹⁸ “Had not linguistic usage stood in our way then ‘Hinduness’ would have certainly been a better word than Hinduism as a near parallel to Hindutva” (Savarkar, Hindutva, 4; this book was first published in 1923).

¹⁹ Ibid., Hindutva, 6–10, and cf. 70–71.

²⁰ Jordens, Dayānanda Sarasvatī, 90. For a Maharashtrian parallel from the same period, see Dalmia, Nationalization of Hindu traditions, 35n17.

²¹ K. W. Jones, Arya Dharm, 37–38.

²² Ibid., 251; for a short biography, see 332–33.

²³ Jaffrelot, Hindu nationalist movement, 18.

²⁴ For these competing identities see Herzfeld, Ours once more, 124.

²⁵ See 404–9.

²⁶ The nature and history of caste in India have been the subject of much research and controversy. For a survey of both, see S. Bayly, Caste, society and politics. Like Bayly, I take it for granted that caste is real (see, for example, 6, 345, 366), and that it is ancient (see, for example, 13, 25, 366). She also argues throughout the book that in recent centuries caste has become much more salient and widespread in Indian society, and she dates the beginning of this change to the late pre-colonial period (see, for example, 25, 196, 368–69). Though this view may be correct, and is to an extent historically plausible, the book does not present evidence that would unambiguously support it.

²⁷ Kane, History of Dharmaśāstra, 2:35.

²⁸ Contrast the situation among a religious community that preserved a comparably ancient ethnic identity: all Jews are Israelites.

²⁹ K. W. Jones, “Two sanātan dharma leaders,” 231.

³⁰ Jordens, Dayānanda Sarasvatī, 254.

³¹ For a second-generation member of the Ārya Samāj who likewise considered Śūdras to be Aryans, see Jaffrelot, Hindu nationalism, 52.

³² Jaffrelot, India’s silent revolution, 187–89, 192–96.

³³ Mendelsohn and Vicziany, Untouchables, 27–28.

³⁴ U. Sharma, Caste, 51, citing Searle-Chatterjee, “Caste, religion,” 161; also Searle-Chatterjee, “Urban ‘untouchables,’ ” 17 (the fieldwork relates to Benares, see 14).

³⁵ Jaffrelot, India’s silent revolution, 109 (for the Camārs as Jātavs, see 207–8, 457).

³⁶ See 77.

³⁷ See J. Leslie, Authority and meaning, 56, for an instance from 1927.

³⁸ K. Sreenivasan, Sree Narayana Guru, 133, quoted in A. Sharma, Modern Hindu thought, 226–27.

³⁹ Stietencron, “Hinduism,” 40–41, 46–47.

⁴⁰ Jaffrelot, “Politics of processions,” 63n2.

⁴¹ For the Vedas, see 400–2; for cow-killing, see 296, 421–22; for the Rām cult, see 422–24.

⁴² Nilakanta Sastri, Pāṇḍyan kingdom, 242–43, whence Chattopadhyaya, Representing the Other, 57.

⁴³ Hudson, “Arumuga Navalar,” 28 (on the Śaiva Siddhānta of the thirteenth to eighteenth centuries).

⁴⁴ Renou, Le destin du Véda, 3–4, 6–8 (with further sectarian examples); contrast Rao, Śiva’s warriors, 7 (on the thirteenth-century Vīraśaiva Somanātha, who rejects the Brahmins but not the Vedas).

⁴⁵ Vaudeville, Weaver named Kabir, 154, and cf. 149, 151, 216, 221, 234, 246, 262 (but 228 is kinder).

⁴⁶ Damen, Crisis and religious renewal, 25, 33.

⁴⁷ K. W. Jones, Arya Dharm, 27.

⁴⁸ Rambachan, Limits of scripture, 57.

⁴⁹ See 401.

⁵⁰ See 67.

⁵¹ Cenkner, Tradition of teachers, 136. This Śankarācārya, I think, held the office from 1954 to 1989 (see ibid., 122, and Jaffrelot, Hindu nationalist movement, 470).

⁵² O’Hanlon, Caste, conflict, 145; and cf. his repeated references to the spurious scriptures of the Brahmins (Phule, Collected works, 2: 11, 31, 32, 77, 83).

⁵³ For references to the cow as a point of friction between Hindus and Muslims in pre-modern times further to those that follow, see Haider, “Holi riot,” 131, 133, 143n37.

⁵⁴ Kuijp, “Earliest Indian reference,” 200 (drawn to my attention by Shahab Ahmed).

⁵⁵ Pollock, “Rāmāyaṇa and political imagination,” 276–77 (with further examples from the same poem); for the dating, see 274.

⁵⁶ Nilakanta Sastri, Pāṇḍyan kingdom, 242, whence Chattopadhyaya, Representing the Other, 57. The Vilasa grant of around 1330 identifies the three characteristic vices of the invaders as drinking wine, eating the meat of cows (gopiśita), and slaying the twice-born (Venkataramanayya and Somasekhara Sarma, “Vilasa grant,” 241, 261 line 36, cited in Chattopadhyaya, Representing the Other, 59).

⁵⁷ Laine, “Śivājī as epic hero,” 6–7.

⁵⁸ Laine, Shivaji, 29, and cf. 24, 49, 59.

⁵⁹ Friedmann, Shaykh Aḥmad Sirhindī, 33 (min ajall shaʿāʾir al-Islām fī ʾl-Hind).

⁶⁰ See Rizvi, Muslim revivalist movements, 249, and cf. 309.

⁶¹ Sirhindī was lamenting that the tyrannical ruler of his day had done so. For another instance, see Laine, Shivaji, 61.

⁶² See Jordens, Dayānanda Sarasvatī, 220–23, and cf. 290 for the anti-Muslim animus of the campaign.

⁶³ Pandey, Construction of communalism, 262–66; for the background, see 167–74. The sacrifice is that of the ʿĪd al-Aḍḥā, on which any of a variety of animals may be sacrificed; but it is known in north India as Baqr-Id, sc. the festival at which cows (Arabic baqar) are sacrificed.

⁶⁴ Ibid., 264 no. 5.

⁶⁵ Ibid., 265 no. 6.

⁶⁶ Rangaswami Aiyangar, Hindu view of life, 19.

⁶⁷ D. N. Jha, Holy cow, 29, 33; and see Kane, History of Dharmaśāstra, 2: 772.

⁶⁸ D. N. Jha, Holy cow, 37; and see Kane, History of Dharmaśāstra, 2: 772.

⁶⁹ D. N. Jha, Holy cow, 113–15, 119.

⁷⁰ Ibid., 93, citing the commentary of Medhātithi, for which see G. Jha, Manusmṛti, 5: 39, 40, 52.

⁷¹ Jordens, Dayānanda Sarasvatī, 125, and see 219, 220–21. His cow-protection campaign thus smacks of political opportunism.

⁷² Bakker, Ayodhyā, 1: 66 (summarizing a detailed discussion of the rise of the Rām cult).

⁷³ Masica, Indo-Aryan languages, 57, 422.

⁷⁴ Bakker, Ayodhyā, 1: 124; Hill, Holy lake, x.

⁷⁵ Lutgendorf, “Interpreting Rāmrāj, 269f, based on Pandey, “Peasant revolt,” 258–61. Pandey’s authorities include Nehru (Autobiography, 53).

⁷⁶ Hill, Holy lake, 87–89 (Tulsīdās here refers to Ayodhyā with the vernacular form “Avadh”). The belief is ancient (Vālmīki, Rāmāyaṇa, 159, and cf. 134).

⁷⁷ Bakker, Ayodhyā, 1: 12.

⁷⁸ Bakker gives a summary presentation of his datings of the versions of the Ayodhyāmāhātmya in a table (ibid., 1: 153). For an English rendering of the guide, see Rám Náráyan, “Ayodhyá Máhátmya.”

⁷⁹ Baker, Ayodhyā, 2: 143, 149.

⁸⁰ Ibid., 1: 128.

⁸¹ Cf. ibid., 2: 146.

⁸² Ibid., 2: 147–48.

⁸³ Jaffrelot, Hindu nationalist movement, 92–96.

⁸⁴ Fuller, Camphor flame, 270. Fuller links the popularity of Rām to two characteristics of the region: the role of the Tulsīdās Rāmāyaṇa as a favorite devotional text and the observance of the Rām Līlā festival (see 121–23).

⁸⁵ For a survey of these versions, see Brockington, Righteous Rāma, 260–62, 266–86.

⁸⁶ Pollock, “Rāmāyaṇa and political imagination” (see especially 283, 287); more summarily Lutgendorf, “Interpreting Rāmrāj,” 258–59.

⁸⁷ Laine, “Śivājī as epic hero,” 10–11. See also Laine, Shivaji, 23.

⁸⁸ Irschick, Politics and social conflict, 283–84, 294, 339.

⁸⁹ On Periyār (as his admirers called him) and his pamphlet, see Richman, “Ramasami’s reading.”

⁹⁰ Khilnani, Idea of India, 154–57.

⁹¹ Thus Pande, Life of Śaṅkarācārya, 52 (arguing for the second half of the seventh century).

⁹² Ibid., 338, 357, 363, and cf. 337 on his travels.

⁹³ Ibid., 43, 360–61.

⁹⁴ Manu, Laws, 19 (2.21–22). Other early texts have different views, particularly on the east-west axis (Kane, History of Dharmaśāstra, 2: 13–15).

⁹⁵ Kane, History of Dharmaśāstra, 2: 17.

⁹⁶ Barz, Bhakti sect, 27–28.

⁹⁷ R. B. Williams, New face of Hinduism, 10 (for the chronology, see 9, 11).

⁹⁸ Gross, “Hindu asceticism,” 172.

⁹⁹ Ibid., 604–8 (and see also 188, cited in Jaffrelot, Hindu nationalist movement, 361n97).

¹⁰⁰ Pollock, “Rāmāyaṇa and political imagination,” 278 (Āryāvarta); Venkataramanayya and Somasekhara Sarma, “Vilasa grant,” 241, 260 line 10, cited in Chattopadhyaya, Representing the Other, 58 (Bhāratavarṣa).

¹⁰¹ Wagle and Kulkarni, Vallabha’s Paraśarāma Caritra, 2.12–14 (24 = 152). For the term karmabhūmi, see Kane, History of Dharmaśāstra, 2: 17; for the fifty-six regions, see 3: 136.

¹⁰² For traditions requiring that only Muslims live in the core area, see Friedmann, Tolerance and coercion, 85–86, 90–93.

¹⁰³ Note that Muslim pilgrims patrol the sacred core of the Islamic world, but not its borders.

¹⁰⁴ Cf. Bhardwaj, Hindu places of pilgrimage, 117–19, 212.

¹⁰⁵ Ibid., 62, map (Haridvāra, from the Garuḍa Purāṇa).

¹⁰⁶ Ibid., 125–27, and see 119, 222 (on Hardvār).

¹⁰⁷ Jordens, Dayānanda Sarasvatī, 23–24, 46–47, 185, 207–8.

¹⁰⁸ Bhardwaj, Hindu places of pilgrimage, 215 (and see the photographic section following 96).

¹⁰⁹ For this process see Brass, Politics of India, 169–74; R. Guha, India after Gandhi, chap. 9. Contrast the lack of such accommodation of ethnic groups in Pakistan.

¹¹⁰ See 73, 75–76. In a very Indian fashion caste played a hidden but significant role in the formation of the linguistic states (see Jaffrelot, “From Indian territory,” 201–2).

¹¹¹ Jaffrelot, Hindu nationalist movement, 33.

¹¹² Ibid., 118–19.

¹¹³ Ibid., 315.

¹¹⁴ Ibid., 193.

¹¹⁵ Ibid., 530.

¹¹⁶ See, for example, ibid., 10, 222, 230, 315, 368.

¹¹⁷ The pragmatism has also been in evidence in foreign policy. On a visit to Lahore in 1999, Vajpayee visited one of Pakistan’s key symbolic sites, the Mīnār-i Pākistān, where the “Lahore Resolution” of 1940 was passed by the Muslim League (Jaffrelot, History of Pakistan, 118; and cf. Jaffrelot, Hindu nationalist movement, 285, for his visit to Islamabad in 1978). In Karachi in 2005, Advani put in an appearance at Jinnah’s mausoleum, where he praised Jinnah in terms that caused great distress to the RSS (Iyer, “Jinnah favoured a secular Pakistan”).

¹¹⁸ See 103–20.

¹¹⁹ Jaffrelot, Hindu nationalist movement, 439–41. Of the four southern states, Karnataka is the only one in which the BJP has made serious headway; in the general election of 2004, it won 18 of its 138 seats in this state and received 35 percent of the vote there. This is a significant modification of the party’s geographical base and may be one reason for its diminishing interest in issues such as Ayodhyā (Jaffrelot, “BJP,” 240–41, and the table, 242; in the general election of 2009, the party won 19 of the 28 seats in Karnataka and received 42 percent of the vote, while winning no seats in the other three states; see the Wikipedia article “Indian general election, 2009,”). In this connection it may be worth noting the weakness of political bonding around Kannaḍa, the language of the state (see Manor, “Language, religion,” especially 171–75).

¹²⁰ Jaffrelot, Hindu nationalist movement, 160.

¹²¹ Ibid., 439.

¹²² For the exclusion of the Śūdras, see Kane, History of Dharmaśāstra, 2: 154–55. There is an isolated view implying the contrary (36, 156–57).

¹²³ Ibid., 2: 155.

¹²⁴ Ibid., 2: 158, 159.

¹²⁵ Ibid., 2: 303, 304.

¹²⁶ This is Basham’s translation of Ṛg Veda 3.62.10 (Basham, Wonder that was India, 162, with the original).

¹²⁷ Kane, History of Dharmaśāstra, 2: 300.

¹²⁸ Stevenson, Rites of the Twice-born, 210–25, especially 218–19, 221–23.

¹²⁹ Ibid., 348–50. For a recitation of the Gāyatrī by 125 Brahmins in 1890, see K. W. Jones, Socio-religious reform movements, 79.

¹³⁰ Kane, History of Dharmaśāstra, 2: 302–3.

¹³¹ O’Hanlon, Caste, conflict (a monographic study); Jaffrelot, India’s silent revolution, 153–57. Jaffrelot notes that Phule found inspiration in the emancipation of the black slave population of the United States (153).

¹³² O’Hanlon, Caste, conflict, 129, 140.

¹³³ Ibid., 136, 137, 141–42; cf. 72–73.

¹³⁴ As might be expected, the idea that the Aryans were invaders has not found favor with the Hindu nationalists (Jaffrelot, Hindu nationalism, 278, and cf. 366).

¹³⁵ Phule, Collected works, 2: 10–11, quoted in A. Sharma, Modern Hindu thought, 145; see also O’Hanlon, Caste, conflict, 129.

¹³⁶ See 63.

¹³⁷ For the latter, see O’Hanlon, Caste, conflict, 204.

¹³⁸ Ibid., 152–53.

¹³⁹ Jaffrelot, India’s silent revolution, 168–72, 244–45.

¹⁴⁰ Ibid., 200–205. For such ideas see also 157 (an Untouchable from Nāgpūr in 1909), 223 (a dissenting member of the first Backward Classes Commission, that appointed in 1953), 390–91, 394 (Kānśī Rām, who founded the Bahujan Samāj Party in 1984); J. Leslie, Authority and meaning, 55–57 (the Ādi-dharm movement in the Punjab in the later 1920s); Schaller, “Sanskritization, 111–13, and Juergensmeyer, Religion as social vision, 45–46, 112, 261 (also on the Ādi-dharm movement).

¹⁴¹ Jaffrelot, India’s silent revolution, 203–4.

¹⁴² Ibid., 19–23.

¹⁴³ Ibid., 22.

¹⁴⁴ Gokhale, From concessions to confrontation, 29–34.

¹⁴⁵ Jaffrelot, India’s silent revolution, 22–23.

¹⁴⁶ Gokhale, From concessions to confrontation, 180–81. The Mahārs were seriously considering a conversion as early as 1935 (164), but in those days Sikhism seemed a more likely choice (169–70). For Buddhist activity in India in the decades prior to the Mahār conversion, see Coward, “Revival of Buddhism,” 277–84.

¹⁴⁷ For this belief see Duncan, “Levels,” 287, citing an election pamphlet from ʿAlīgaṛh.

¹⁴⁸ Jaffrelot, India’s silent revolution, 109. For Camār Buddhism, see also 457.

¹⁴⁹ Mendelsohn and Vicziany, Untouchables, 219; Jaffrelot, India’s silent revolution, 388.

¹⁵⁰ Jaffrelot, India’s silent revolution, 390, 423.

¹⁵¹ Ibid., 414–15, 419n78. For Āmbeḍkar as a Bodhisattva, see Gokhale, From concessions to confrontation, 126, 182, 186.

¹⁵² “Maya threatens to embrace Buddhism.”

¹⁵³ See, for example, “Mayawati justifies lighting Kanshi Ram’s funeral pyre.” I owe this information to Christophe Jaffrelot.

¹⁵⁴ See Gokhale, From concessions to confrontation, 168–70.

¹⁵⁵ Jaffrelot, India’s silent revolution, 168, 170, 203–4, 208.

¹⁵⁶ See Omvedt, Dalits, 121–24.

¹⁵⁷ See, for example, K. W. Jones, Arya Dharm, 5, 10–12, 144, and, for a conspicuous example of conversion to Christianity among an Untouchable caste of the far south, K. W. Jones, Socio-religious reform movements, 156–60.

¹⁵⁸ Jaffrelot, Hindu nationalist movement, 340–42.

¹⁵⁹ The term was coined by the anthropologist M. N. Srinivas in 1947 to label a process by which a caste stakes a claim to a higher position in the hierarchy by adopting “the customs and way of life of a higher caste” (Srinivas, “Note on Sanskritization,” 200, 202; for the date, see further his article “Cohesive role of Sanskritization,” 235n1).

¹⁶⁰ Vivekānanda, Complete works, 3: 291, 298 (partly quoted in Dixit, “Political and social dimensions,” 304–5; this article is thoroughly hostile to Vivekānanda but rich in quotations). Vivekānanda himself was a Kāyasth and thus a member of a literate caste of ambiguous status; he claimed vociferously to be a Kṣatriya rather than a Śūdra (305–6; Vivekānanda, Complete works, 3: 211; and cf. Kane, History of Dharmaśāstra, 2: 75–77).

¹⁶¹ See 59.

¹⁶² For these, see Basham, Wonder that was India, 41, 75, 134, 304.

¹⁶³ Jaffrelot, India’s silent revolution, 187–88.

¹⁶⁴ For these claims and their ramifications, see O’Hanlon, Caste, conflict, 24–27, 32–45.

¹⁶⁵ Ibid., 33–34.

¹⁶⁶ Ibid., 41–42.

¹⁶⁷ Deliège, Untouchables of India, 156–57. Cf. also Jaffrelot, Hindu nationalist movement, 21.

¹⁶⁸ Compare Āmbeḍkar’s analysis of “graded inequality” (Jaffrelot, India’s silent revolution, 20–21).

¹⁶⁹ O’Hanlon, Caste, conflict, 38–39.

¹⁷⁰ Ibid., 147.

¹⁷¹ Ibid., 20n12, 147n19.

¹⁷² Jordens, Dayānanda Sarasvatī, 96, 104 (with a qualification).

¹⁷³ Ibid., 262. He included even Untouchables (ati-śūdras), not to mention girls.

¹⁷⁴ K. W. Jones, Arya Dharm, 40; cf. also Jordens, Dayānanda Sarasvatī, 51, 52.

¹⁷⁵ Jaffrelot, India’s silent revolution, 16.

¹⁷⁶ Jaffrelot, Hindu nationalist movement, 45–47.

¹⁷⁷ Ibid., 49.

¹⁷⁸ See, for example, ibid., 358–59, on the reconversion efforts of the VHP in 1981–1982.

¹⁷⁹ Ibid., 349.

¹⁸⁰ Ibid,, 61, 235n19, 349, and cf. 347 (on the VHP). For a rousing statement by Sāvarkar to the same effect, see Lederle, Philosophical trends, 290; and cf. Savarkar, Hindutva, 39.

¹⁸¹ Savarkar, Hindutva, 138–39.

¹⁸² Jaffrelot, Hindu nationalist movement, 45.

¹⁸³ Ibid., 361, and cf. 403.

¹⁸⁴ Ibid., 47.

¹⁸⁵ Ibid., 49; Jaffrelot, India’s silent revolution, 477; and cf. Jaffrelot, Hindu nationalist movement, 509, on low-caste members of the BJP.

¹⁸⁶ Seshadri, RSS, 125, in a section headed “Vedas for All” (“backward” is a euphemism for “low-caste”). The book is a compilation of good news sent in by members of the RSS and edited by Seshadri.

¹⁸⁷ For the work of Sevā Bhārtī in the slums of Āgrā, see Jaffrelot, India’s silent revolution, 456–62; for Sevā Bhārtī in general, see 454–56. The Hindu nationalists likewise went to work among the unemployed rural immigrants of Bhopāl (Jaffrelot, Hindu nationalist movement, 511), and responded to natural disasters (ibid., 284n11; Hansen, “BJP,” 144). At the time of the partition of India, they did much to aid Hindu refugees from West Pakistan, especially in Delhi (Jaffrelot, Hindu nationalist movement, 75–76); this city, with its large refugee population, thus became a Hindu nationalist stronghold (ibid., 226n170).

¹⁸⁸ Jaffrelot, India’s silent revolution, 458; and see Jaffrelot, “Hindu nationalism,” 208, on a village school in Madhya Pradesh.

¹⁸⁹ Jaffrelot, India’s silent revolution, 459.

¹⁹⁰ Ibid., 455.

¹⁹¹ S. Bayly, Caste, society and politics, 184.

¹⁹² Jaffrelot, India’s silent revolution, 454 and n4; Jaffrelot, Hindu nationalist movement, 535 and n4.

¹⁹³ Jaffrelot, India’s silent revolution, 461–62.

¹⁹⁴ Hansen, “BJP,” 124–25.

¹⁹⁵ Jaffrelot, India’s silent revolution, 477, 478; Jaffrelot, Hindu nationalist movement, 388, 509. She is a Lodhī, and thus stems from the “Other Backward Classes” (OBCs), not from the Untouchables.

¹⁹⁶ Jaffrelot, India’s silent revolution, 316, 326, 347.

¹⁹⁷ Cashman, Myth of the Lokamanya, 115–17.

¹⁹⁸ Jaffrelot, India’s silent revolution, 172–75, 238–41.

¹⁹⁹ Ibid., 185.

²⁰⁰ Ibid., 237–46, 324–25.

²⁰¹ Ibid., 90–92, 102.

²⁰² Ibid., 214; Galanter, Competing equalities, 34n46, and cf. 130.

²⁰³ For the background of this term, see Jaffrelot, India’s silent revolution, 214–15; Galanter, Competing equalities, 159.

²⁰⁴ Jaffrelot, India’s silent revolution, 221–29, 320–24.

²⁰⁵ Ibid., 221, 268.

²⁰⁶ Ibid., 338.

²⁰⁷ For a summary statement, see ibid., 9–10. For the unevenness of this transformation, see, for example, 362–63 on the contrast between Bihar and Rajasthan.

²⁰⁸ Jaffrelot, Hindu nationalist movement, 493.

²⁰⁹ Ibid., 414–15; Jaffrelot, India’s silent revolution, 453.

²¹⁰ Hansen, “BJP,” 134.

²¹¹ Jaffrelot, India’s silent revolution, 426.

²¹² Ibid., 463, 473–74, 478–80, 484–90, 490–91.

²¹³ Compare also the role of Marxism as a caste-blind ideology in West Bengal under Communist rule (ibid., 255), for all that the Party has tended to be dominated by the kind of Communists that Āmbeḍkar referred to as “a bunch of Brahman boys” (Mendelsohn and Vicziany, Untouchables, 211).

²¹⁴ For the social reality of caste in Pakistan, see Jaffrelot, History of Pakistan, chap. 9; for the contrast with India, “a country where the castes are willing to speak their names,” see ibid., 216.

²¹⁵ Deliège, Untouchables of India, 12.

²¹⁶ Jaffrelot, India’s silent revolution, 349, 363, 384.

²¹⁷ The extent of this survival is evident from the “matrimonials” section of the Sunday Times of New Delhi for December 23, 2012. Many advertisements for brides and grooms appear under the relevant caste names; though some announce that caste is no bar, they are a minority. Muslims play the game alongside Hindus: one “Ansari boy” generously states that subcaste is no bar, another that his preference is for a girl from the same caste. I owe my copy of this item and my understanding of its significance to Bernard Haykel.

²¹⁸ Jaffrelot, India’s silent revolution, 384 (on the OBC castes). Compare the role of the Mahārs as the solid core of Āmbeḍkar’s movement.

²¹⁹ Mukerji, Chamars of Uttar Pradesh, 19, 21, 26; Mendelsohn and Vicziany, Untouchables, 48, 219.

²²⁰ K. W. Jones, Socio-religious reform movements, 128–31.

²²¹ Jaffrelot, India’s silent revolution, 153.

²²² Ibid., 424, 462.

²²³ Ibid., 424n94.

²²⁴ Ibid., 187–88.

²²⁵ Note, however, the emergence of Kṛṣṇa as the patron deity of the Yādavs (Fuller, Camphor flame, 275–76); the claim that “Krishna was a Yadav” could thus facilitate a convergence with Hindu nationalism in the event that the Yādavs’ conception of their caste interests were to change.

²²⁶ Jaffrelot, India’s silent revolution, 379. Compare the comment of a contemporary on the Buddhist politician in ʿAlīgaṛh: “it is Maurya’s abuse of Brahmans and Thakurs that earned him the adoration of his caste” (Duncan, “Levels,” 274–75). The Ṭhākurs are Kṣatriyas.

²²⁷ Jaffrelot, India’s silent revolution, 352.

²²⁸ Ibid., 364, 377–78, 383–84.

²²⁹ Cf. ibid., 234–35 (an earlier case of such resentment).

²³⁰ Ibid., 376–77.

²³¹ Hansen, “BJP,” 126; cf. the derogatory identification of the party as the new Peśvās (130) and “the murderers of Gandhiji” (132–33). The RSS has been tactless enough to hold an annual meeting at the former Peśvā palace (Hansen, Saffron wave, 117).

²³² Jaffrelot, India’s silent revolution, 385.

²³³ Kakar, Colors of violence, 158.

²³⁴ If there is such immunity anywhere in the Islamic world, it is ethnic—as with Kurds and Kabyles.

²³⁵ Kepel, Jihad, 6. This is perhaps the central theme of the book.

²³⁶ For this usage, see Fandy, “Egypt’s Islamic Group,” 612–13. Here, as with “Aryan,” an ethnic term has been incorporated into a caste system. Contrast the unrestricted application of the term “Arab” to all who speak the language by ʿAbd al-Ghanī al-Nābulusī and Bakrī (see above, 13, 14n54).

²³⁷ For Upper Egypt, see Fandy, “Egypt’s Islamic Group,” 612–13; for Ḥaḍramawt, see Bujra, Politics of stratification, 13–15 (a summary statement expanded in the rest of the chapter); also Freitag, Indian Ocean migrants, 39–42. As might be expected, there is some academic mileage to be gained from stressing diversity within the stratificatory patterns of Ḥaḍramawt (see Camelin, “Reflections”).

²³⁸ Bujra, Politics of stratification, 95.

²³⁹ Fandy, “Egypt’s Islamic Group,” 617–18.

²⁴⁰ Bujra, Politics of stratification, 4–5, 77–81; Freitag, Indian Ocean migrants, 50–59.

²⁴¹ In the Egyptian case, the discontent came to a head in the late twentieth century and was articulated by radical Islamism (Fandy, “Egypt’s Islamic Group,” 611, 613–14, and cf. 615n29). In the Ḥaḍramī case the dissidence began early in the twentieth century and was articulated by an Islamic reformist movement known as Irshādism (Bujra, Politics of stratification, 94, 130–33; Freitag, Indian Ocean migrants, 245–58).

²⁴² For the Islamic terms of the argument in Upper Egypt, see Fandy, “Egypt’s Islamic Group,” 613–14; for the Ḥaḍramī case, see Bujra, Politics of stratification, 94, 131; Freitag, Indian Ocean migrants, 245–46; Mobini-Kesheh, “Islamic modernism,” 241.

²⁴³ Fandy, “Egypt’s Islamic Group,” 614.

²⁴⁴ Eckert, Charisma of direct action, 88–89.

²⁴⁵ Hansen, “BJP,” 131.

²⁴⁶ Eckert, Charisma of direct action, 90–91, 103–4.

²⁴⁷ Ibid., 257.

²⁴⁸ See 63–64.

²⁴⁹ For the modern resonances of this destruction, see R. Guha, India after Gandhi, 141–42. For a list of eighty historically attested Muslim temple desecrations in India over the period 1192–1760, see Eaton, “Temple desecration,” preceding 275. Eaton argues that, despite the religious rhetoric accompanying such desecrations, the motivation was political (see for example 260, 268).

²⁵⁰ Rizvi, Muslim revivalist movements, 248.

²⁵¹ Jaffrelot, Hindu nationalist movement, 402, 423.

²⁵² Interview quoted in Banerjee, Warriors in politics, 150; for more such denunciation of the disloyalty of the inhabitants of this Muslim quarter in Bombay, see Eckert, Charisma of direct action, 91. Cricket matches between India and Pakistan have been described as “war minus the shooting” for most viewers (R. Guha, India after Gandhi, 725).

²⁵³ Kakar, Colors of violence, 161, 165.

²⁵⁴ Banerjee, Warriors in politics, 150. The politician gives these views on the authority of his “political guru,” the leader of the Shiv Sena.

²⁵⁵ Graham, Hindu nationalism, 50.

²⁵⁶ Jaffrelot, Hindu nationalist movement, 459.

²⁵⁷ Jaffrelot, “Politics of processions,” 86; and cf. Eckert, Charisma of direct action, 119.

²⁵⁸ Brass, Politics of India, 247.

²⁵⁹ Jaffrelot, Hindu nationalist movement, 334–35. For this affair, see below, 288–89.

²⁶⁰ Cf. ibid., 338–40, 362n103. Hansen reports that in private a fair number of Muslims who had worked in the Persian Gulf would confide that their experiences with Arab employers had destroyed their faith in Muslim solidarity, but in India they could not say this in public (Saffron wave, 256n24).

²⁶¹ See Narayanan, “Religious vocabulary,” 74–76. The text to which the article is devoted is a seventeenth-century Tamil life of the Prophet Muḥammad whose title combines two literary terms, the Muslim sīra and the Hindu purāṇa (79).

²⁶² Ibid., 93–94.

²⁶³ Kakar, Colors of violence, 59–69.

²⁶⁴ Ibid., 62, 64.

²⁶⁵ Ibid., 153–66.

²⁶⁶ This Hindu demographic paranoia was a product of the decennial census instituted by the British in 1871. Thus in 1909 one Hindu commentator was warning that Hindus would disappear within 420 years (Jaffrelot, Hindu nationalist movement, 24). For a Hindu nationalist account of the continuing demographic peril in a publication of 1999, see Jaffrelot, Hindu nationalism, 244–50. The fear does not come out of thin air: the decennial censuses show the proportion of Muslims in the Indian population rising from 10.7 percent to 13.4 percent between 1961 and 2001 and the proportion of Hindus falling from 83.5 percent to 80.5 percent (Pew Forum, Spirit and power, 81).

²⁶⁷ Kakar, Colors of violence, 159–64.

²⁶⁸ Eckert, Charisma of direct action, 204.

²⁶⁹ Hansen, Wages of violence, 153–54.

²⁷⁰ Manor, “Southern discomfort,” 190–91.

²⁷¹ Jaffrelot, “Politics of processions,” 74–75.

²⁷² Jaffrelot, Hindu nationalist movement, 463n62.

²⁷³ Zaman, Ulama, 183–84 (on Mawlānā Waḥīd al-Dīn Khān).

²⁷⁴ Jaffrelot, Hindu nationalist movement, 20. Jaffrelot labels the effect seen here as “stigmatisation and emulation.”

²⁷⁵ Golwalkar, We or our nationhood, 53 n.

²⁷⁶ Ibid., 47 (another such passage is quoted in Jaffrelot, Hindu nationalist movement, 56). For Goḷwalkar, concessions to minorities courted the “disastrous fate of the unfortunate Czechoslovakia” (49–50).

²⁷⁷ For a possible softening in Goḷwalkar’s own later views, see Gold, “Organized Hinduisms,” 566.

²⁷⁸ Graham, Hindu nationalism, 96.

²⁷⁹ See 110–11.

²⁸⁰ Eckert, Charisma of direct action, 126; similarly Jaffrelot, “Politics of processions,” 84.

²⁸¹ For an outstanding study that follows this line of analysis, see Brass, Hindu-Muslim violence.

²⁸² Kakar, Colors of violence, 150–51. His comments on the whole debate seem eminently sensible (149–53).

²⁸³ Brass states that his analysis “relegates all spontaneity theories of the causes of riots to the realm of blame displacement” (Hindu-Muslim violence, 16). But he later observes that there is “little spontaneous about Hindu-Muslim riots” (377–78—he does not say “nothing spontaneous”); he speaks of “elements both of spontaneity and planning” that contribute to them, of “prejudices” that interact with purposive action (358), and of “latent hostilities” that are deliberately mobilized (378). Perhaps most strikingly, he remarks that “there exists in India a discourse of Hindu-Muslim communalism that has corrupted history, penetrated memory, and contributes in the present to the production and perpetuation of communal violence in the country” (34). The only question that leaves is how long this discourse has been in existence.

²⁸⁴ For a riot “between the communities of Hindus and Muslims” in Gujarat in 1714, as described in sources written by contemporaries, see Haider, “Holi riot,” 128–32. For one in Calcutta in 1789, see C. A. Bayly, “Pre-history of ‘communalism,’ ” 199.

²⁸⁵ For the case of Karnataka, see Manor, “Karnataka,” 322; Manor, “Southern discomfort,” 182. For an explanation of the Tamil case in terms of the early development of caste-based politics in the south, see Wilkinson, Votes and violence, 194–96. But for adverse trends in the south, see Jaffrelot, “Politics of processions,” 87.

²⁸⁶ Jaffrelot, Hindu nationalist movement, 473, and cf. 467. But see also 425–26, 476, on more irenic attitudes in the rural north; communal riots are an urban, not a rural phenomenon (Varshney, Ethnic conflict, 6, 95–97).

²⁸⁷ Manor, “Southern discomfort,” 189. There is also a marked contrast between the west (where Gujarat became a disaster area) and the east (where Orissa has never been one); see Varshney, Ethnic conflict, 97–101.

²⁸⁸ For orchestration at the micro-political level, note one observer’s account of the role of meetings of “strong men” representing different localities in deciding where “the wind (havā) is to be spread” in the course of a riot (Kakar, Colors of violence, 80).

²⁸⁹ Jaffrelot, “Politics of processions,” with numerous examples.

²⁹⁰ Jaffrelot, Hindu nationalist movement, 34; Jaffrelot, “Politics of processions,” 66.

²⁹¹ As in Poona in 1982 (Jaffrelot, “Politics of processions,” 76).

²⁹² As in Madras in 1995 (ibid., 87).

²⁹³ As in Allāhābād before 1924 (ibid., 68–69).

²⁹⁴ As already in Nāgpūr in the 1920s (ibid., 66).

²⁹⁵ Ibid., 86; Eckert, Charisma of direct action, 115–16.

²⁹⁶ Jaffrelot, Hindu nationalist movement, 458–63; for the case of Bombay, see also Eckert, Charisma of direct action, 114–17.

²⁹⁷ Muslim politicians, by contrast, do not stand to benefit. The fact that Muslims are a minority means that they are likely to be on the losing side in a communal riot and that solidifying the Muslim vote does not help to get Muslims elected. But Muslims do start riots. This was the case in Calcutta in 1996, where it was a Muslim procession that insisted on making its way through a Hindu quarter (Jaffrelot, “Politics of processions,” 88), and on some accounts in Gujarat in 2002 (for the Gujarat riots, see Brass, Hindu-Muslim violence, 387–91); for the disputed events that triggered the riots, see also Jaffrelot, “2002 pogrom,” 174–75.

²⁹⁸ For the link between riots and elections, see Jaffrelot, Hindu nationalist movement, 395, 448, 513; Jaffrelot, “Politics of processions,” 68–71, 73, 78–81; Shah, “BJP’s riddle,” 248, 249; Eckert, Charisma of direct action, 127–28; Brass, Hindu-Muslim violence, 33–34; Wilkinson, Votes and violence, 49–51. There is also a scatter of southern examples (Manor, “Southern discomfort,” 167, 172, 173, 174).

²⁹⁹ Shah, “BJP’s riddle,” 249.

³⁰⁰ Ibid., 261–65.

³⁰¹ Jaffrelot, “2002 pogrom,” 182.

³⁰² Two recent studies probe these conditions. One compares cities that are riot prone with cities that are not and concludes that intercommunal civic engagement is a major factor in reducing the probability of riots (Varshney, Ethnic conflict, 281). Another makes the case that the key factor is the electoral interests of state governments, since they control the police and thereby decide whether an incipient riot is to continue or not (Wilkinson, Votes and violence, 4, 16–17; for his explanation of the Gujarat riots of 2002 in these terms, see 7–8, 59–62, 154–60). Cf. the view of a senior police officer in 1995 that no communal riot can continue for more than twenty-four hours unless the state government wants it to (Jaffrelot, “Politics of processions,” 90).

³⁰³ The ambivalence of the BJP response at the national level to the Gujarat riots of 2002 is instructive. Prime minister Vajpayee denounced the riots on television, but the state government was not dismissed for its failure to maintain law and order; he visited Gujarat, but only after a considerable delay, and he took Umā Bhārtī with him; a firm, even-handed police chief with a track record for handling communal violence effectively was dispatched to Gujarat, where he was less than welcome to the state leader, but he was not given the force he asked for (Brass, Hindu-Muslim violence, 390–91). Presumably one factor here was the need to appease coalition partners; another may have been the emergence among the BJP’s constituencies of a middle class that wants to see India governed by a strong, disciplined, no-nonsense nationalist party—not a party of chaos and disorder (see Jaffrelot, Hindu nationalist movement, 432–35, 448, 449–50, 451, 458, 528, and Manor, “In part, a myth,” 66; and cf. Jaffrelot, “Politics of processions,” 86–87 on rural voters, and Manor, “Southern discomfort,” 189–90 and n73 on both middle-class and rural voters in Karnataka).

³⁰⁴ This is not to say that the technique was monopolized by the Hindu nationalists: for the role of Congress politicians in fomenting riots against Sikhs following the assassination of Indira Gandhi in 1984, see R. Guha, India after Gandhi, 565–66.

³⁰⁵ Jaffrelot, Hindu nationalist movement, 415–16.

³⁰⁶ Eckert, Charisma of direct action, 221, and cf. Hansen, “BJP,” 158–59; also Jaffrelot, Hindu nationalist movement, 448, and Brass, Hindu-Muslim violence, 184–85.

³⁰⁷ See, for example, Manor, “Karnataka,” 354 (the coalition of the disadvantaged put together by an enterprising politician in the 1970s); Jaffrelot, India’s silent revolution, 248 (Gujarat in the late 1970s), 391, 394, 402 (Kānśī Rām’s parties from the 1970s to the 1990s), 371–72 (a Yādav-led socialist party of the 1990s), 381–83 (a Yādav-led party in Bihar in the same period). The availability of allies gives the Muslims a leverage in Indian politics that they would not have outside the framework of electoral democracy. Thus the autonomy of the Muslim University in ʿAlīgaṛh was curtailed in 1951 but expanded in 1981, and their protests over the Shāh Bāno affair won them a law resolving the issue in their favor (for both examples, see Brass, Politics of India, 233–34).

³⁰⁸ Jaffrelot, India’s silent revolution, 109 (the slogan was used in the campaign of a Camār Buddhist politician in Uttar Pradesh in the context of an alliance with Muslims in 1962, see above, 59–60). Compare a document of the Ādi-dharm movement in the Punjab which opposes independence for India in the immediate future as prejudicial to “India’s weak minorities—especially Ad Dharmis and the Mohammedan people” (Juergensmeyer, Religion as social vision, 125).

³⁰⁹ Desai, Diary, 301; Jaffrelot, India’s silent revolution, 24. For an example from 1926 in which the Brahmin leader of the RSS was physically attacked by Muslim and low-caste Hindu activists at a public meeting, see Bacchetta, “Hindu nationalist women,” 128.

³¹⁰ Jaffrelot, India’s silent revolution, 416 and n66.

³¹¹ Ibid., 279–89.

³¹² Jaffrelot, Hindu nationalist movement, 308.

³¹³ Hansen, Wages of violence, 181, 182.

³¹⁴ Jaffrelot, Hindu nationalist movement, 420–22.

³¹⁵ Note also that Muslim votes are relatively cheap: Muslims are looking for security, not the candy of reservations (Wilkinson, Votes and violence, 17, 144–45).

³¹⁶ Mendelsohn and Vicziany, Untouchables, chap. 2.

³¹⁷ Eckert, Charisma of direct action, 218–19.

³¹⁸ Jaffrelot, India’s silent revolution, 411, table 11.10; and cf. 405, 445.

³¹⁹ Mendelsohn and Vicziany, Untouchables, 57.

³²⁰ Jaffrelot, Hindu nationalist movement, 86–87; Hansen, “BJP,” 122–23.

³²¹ Manor, “Regional parties,” 121 (with a contrast to Canada and Sri Lanka); see also 118–19, 122.

³²² The phrases are Sāvarkar’s (Hindutva, 140). The salience of such symbolic politics in India marks a strong contrast with China since the Cultural Revolution.

³²³ Fuller, Camphor flame, 261.

³²⁴ Jaffrelot, India’s silent revolution, 415, and see above, 67. Imagine a conference to celebrate the life and work of Salman Rushdie being held in Lahore.

³²⁵ See More, Muslim identity, 148–50, 153, 160. I owe my knowledge of this study to Mahmood Kooria.

³²⁶ Mendelsohn and Vicziany, Untouchables, 230.

³²⁷ Jaffrelot, India’s silent revolution, 412, 415, 419; cf. Trivedy, “Breaking the status quo,” 53–54 (on the vandalization of Āmbeḍkar statues by “upper caste goondas” in villages near Lucknow), and Eckert, Charisma of direct action, 220n15, 221 (on riots following the desecration of an Āmbeḍkar statue in Bombay).

³²⁸ Jaffrelot, India’s silent revolution, 419.

³²⁹ Ibid., 105 and n49. A Camār leader of the same period flirted with the idea of a homeland for Untouchables (Achūtistān, see J. Leslie, Authority and meaning, 60).

³³⁰ See 93.

³³¹ Jaffrelot, India’s silent revolution, 347.

³³² For this usage, see 399.

³³³ Cenkner, Tradition of teachers, 116, and see 143–44.

³³⁴ Ibid., 125.

³³⁵ Ibid., 103, 122, 123, 124, 125, 126, 130–31.

³³⁶ Pande, Life of Śaṅkarācārya, 35; Cenkner, Tradition of teachers, 114–15.

³³⁷ Cenkner, Tradition of teachers, 114–15; Jaffrelot, Hindu nationalist movement, 470.

³³⁸ Cenkner, Tradition of teachers, 104. There are two splendid photographs of this Śankarācārya in Glasenapp, Heilige Stätten Indiens, plates 126–27.

³³⁹ Cenkner, Tradition of teachers, 119.

³⁴⁰ Ibid., 125, 132; Jaffrelot, Hindu nationalist movement, 204–8.

³⁴¹ Lütt, “Śankarācārya of Puri,” 414–19; Cenkner, Tradition of teachers, 131–32. But see also Jaffrelot, Hindu nationalist movement, 23–24, on a minor Maharashtrian Śankarācārya.

³⁴² See Jaffrelot, Hindu nationalist movement, 198, 356–57, 362, 372, 413, 470, 488.

³⁴³ Cenkner, Tradition of teachers, 143; his predecessor had opposed Gandhi on this issue (132).

³⁴⁴ Jaffrelot, Hindu nationalist movement, 470–71.

³⁴⁵ Ibid., 423–24. Such volunteers are typically unemployed urban youths; their involvement in the movement gives them a measure of self-respect and an opportunity to model themselves on the “stereotypical heroes of Hindi popular cinema” (428f). The Shiv Sena has a similar appeal to the young and unemployed male who longs “to be the hero of a Bombay film in which violence leads to justice” (429; cf. Banerjee, Warriors in politics, 71–72).

³⁴⁶ See, for example, Jaffrelot, Hindu nationalist movement, 194–96, 198–200, 206–7, 338, 350–55, 362, 363, 373.

³⁴⁷ Ibid., 210, 348–49, 488.

³⁴⁸ Ibid., 479–80, 483–84.

³⁴⁹ Ibid., 480, reporting a poll conducted in five metropolitan cities in 1993.

³⁵⁰ Ibid., 388.

³⁵¹ Ibid., 479.

³⁵² Cf. ibid., 357–58. There is perhaps more spillover of color in the case of the Shiv Sena: members paint their buildings in saffron (Eckert, Charisma of direct action, 16, 211), and wear it on occasion (211).

³⁵³ I have not seen a study devoted to this theme. Presumably it is because relations between the Hindu nationalists and the Jains are not a problem for either side that nobody seems to be interested in the topic—which is a pity.

³⁵⁴ Jaffrelot, Hindu nationalist movement, 430.

³⁵⁵ Ibid., 140–41 (also mentioning a Jain active in the Ārya Samāj).

³⁵⁶ Ibid., 353, 354.

³⁵⁷ Ibid., 509, and cf. 38n114.

³⁵⁸ Laidlaw, Riches and renunciation, 104.

³⁵⁹ Shah, “BJP’s riddle ,” 258.

³⁶⁰ Jaffrelot, Hindu nationalist movement, 361.

³⁶¹ Ibid., 425, 486–87, 531.

³⁶² For a letter written by the Arab nationalist Khalīl al-Sakākīnī (d. 1953), a Palestinian Christian, to his son, in which he complains that in Muslim eyes “as long as I am not a Moslem I am naught,” see Kedourie, “Religion and politics,” 339–40; for Sakākīnī’s career, see 318–19.

³⁶³ Jordens, Dayānanda Sarasvatī, 111. For some reason he thought the Jains came from the region of China.

³⁶⁴ For this, see Jordens, Dayānanda Sarasvatī, 12–13.

³⁶⁵ Savarkar, Hindutva, 81.

³⁶⁶ Ibid., 38.

³⁶⁷ Jordens, Dayānanda Sarasvatī, 229.

³⁶⁸ Eckert, Charisma of direct action, 220.

³⁶⁹ Golwalkar, We or our nationhood, 10, 66.

³⁷⁰ Mishra, RSS, 96. He also includes in this mission “Lord Mahavira,” the founder of Jainism.

³⁷¹ Malkani, RSS story, 205.

³⁷² Kakar, Colors of violence, 155–56 (the Jains and Sikhs are likewise represented). For a sample of her anti-Muslim venom, see 165.

³⁷³ Jaffrelot, Hindu nationalist movement, 347.

³⁷⁴ Seshadri, RSS, 137–38.

³⁷⁵ K. W. Jones, Arya Dharm, 205–7.

³⁷⁶ Ibid., 135–36.

³⁷⁷ Ibid., 136–39. The star role in this outcome was played by the zealous Paṇḍit Guru Datta (d. 1890).

³⁷⁸ Savarkar, Hindutva, 125; see also 126.

³⁷⁹ Golwalkar, We or our nationhood, 10, and see also 58, 66.

³⁸⁰ Jaffrelot, Hindu nationalist movement, 343, 345.

³⁸¹ Frykenberg, “Hindu fundamentalism,” 245.

³⁸² Seshadri, RSS, 105–7; Gold, “Organized Hinduisms,” 575 and 590n200.

³⁸³ Jaffrelot, Hindu nationalist movement, 198.

³⁸⁴ Ibid., 361.

³⁸⁵ Ibid., 315–16, 325. Cf. also 326 on ʿĀrif Beg.

³⁸⁶ Ibid., 325, 509.

³⁸⁷ Ibid., 423, 430.

³⁸⁸ Seshadri, RSS, 141–42.

³⁸⁹ Jaffrelot, Hindu nationalist movement, 303–4.

³⁹⁰ Ibid., 327.

³⁹¹ Ibid., 503.

³⁹² See 95–96.

³⁹³ Golwalkar, We or our nationhood, 45.

³⁹⁴ Savarkar, Hindutva, 91–92, 99, 100–101.

³⁹⁵ Ibid., 111, 113, and cf. 102. As Muḥammad Iqbāl (d. 1938) had written some years earlier, “In as much as the average man demands a material centre of nationality, the Muslim looks for it in the holy town of Mecca” (quoted in Shaikh, “Millat and mazhab,” 386).

³⁹⁶ Savarkar, Hindutva, 114–15.

³⁹⁷ Ibid., 139.

³⁹⁸ Jaffrelot, Hindu nationalist movement, 325. Incidentally, the idea that an Indian Muslim could be proud of the Hindu cultural heritage received a measure of endorsement from no less an authority than Jamāl al-Dīn al-Afghānī (d. 1897), see Keddie, Islamic response, 58–59; Keddie, Sayyid Jamāl ad-Dīn “al-Afghānī,” 159–60.

³⁹⁹ Jaffrelot, Hindu nationalist movement, 57.

⁴⁰⁰ McKean, “Bhārat Mātā,” 259–60.

⁴⁰¹ Subramanian Swamy, “How to wipe out Islamic terror.” This politician would like to see the imposition of a uniform civil code, the compulsory study of Sanskrit, the renaming of the country as “Hindustan,” and a law prohibiting conversion from Hinduism to any other religion. I owe my knowledge of his views to Bernard Haykel.

⁴⁰² Cf. the Shiv Sena website in 2000: “It is Shiv Sena’s belief that whatever may be our religion, whatever may be our form of worship, our culture is Hindu” (quoted in Eckert, Charisma of direct action, 93; and cf. Katzenstein et al., “Rebirth of Shiv Sena,” 378).

⁴⁰³ Eckert, Charisma of direct action, 185.

⁴⁰⁴ Ibid., 92; and see Katzenstein et al., “Rebirth of Shiv Sena,” 378.

⁴⁰⁵ Jaffrelot, Hindu nationalist movement, 317, quoting a statement from a BJP manifesto of 1984; the phrase already appears in a Jana Sangh manifesto of 1951 (Graham, Hindu nationalism, 50).

⁴⁰⁶ Hansen, Saffron wave, 225. Compare the presence of “important Dharma Sutras from all the religions of India,” including Islam, in the Bhārat Mātā Mandir at Hardvār (McKean, “Bhārat Mātā,” 275).

⁴⁰⁷ Hansen, Saffron wave, 180.

⁴⁰⁸ Jaffrelot, Hindu nationalist movement, 40–45.

⁴⁰⁹ Ibid., 39.

⁴¹⁰ Embree, “Rashtriya Swayamsevak Sangh,” 629–30.

⁴¹¹ Jaffrelot, Hindu nationalist movement, 26, 51–54.

⁴¹² Ibid., 37. The music has changed since India became independent (Seshadri, RSS, 260–61).

⁴¹³ Jaffrelot, Hindu nationalist movement, 39.

⁴¹⁴ Ibid., 41n128 (quoting Goḷwalkar).

⁴¹⁵ Eckert, Charisma of direct action, 16, 211; Banerjee, Warriors in politics, 152.

⁴¹⁶ Banerjee, Warriors in politics, 119. “Śiv Senā” means “Army of Śiva.”

⁴¹⁷ Ibid., 1, 119, 152–53; cf. Graham, Hindu nationalism, 152; Jaffrelot, Hindu nationalist movement, 417n22.

⁴¹⁸ For a brief description of the VHP, see Jaffrelot, “Vishva Hindu Parishad.”

⁴¹⁹ See 297, 428.

⁴²⁰ The idea of such an assembly went back to the nineteenth century, making its first appearance with Dayānand’s teacher Virjānand (Jordens, Dayānanda Sarasvatī, 37, and cf. 155–56).

⁴²¹ Jaffrelot, Hindu nationalist movement, 198, 346, and cf. 373.

⁴²² Ibid., 201.

⁴²³ Devahuti, Harsha, 180–81.

⁴²⁴ Ibid., 181n1; Lamotte, History of Indian Buddhism, 60, 292.

⁴²⁵ Veer, “Hindu nationalism,” 666. A curious exception is the VHP’s dislike of the use of Hindu names by Christians (A. Sharma, Modern Hindu thought, 247n21); a Hindu nationalist should surely welcome this practice as a sign of assimilation of Hindu culture.

⁴²⁶ Wolpert, Tilak and Gokhale, 179, in a speech of 1906.

⁴²⁷ Jaffrelot, Hindu nationalism, 40–41; and see K. W. Jones, Arya Dharm, 285–86.

⁴²⁸ For brief surveys of Hindu nationalist activity in the diaspora, see Jaffrelot, Hindu nationalism, 361–64, and Therwath, “Far and wide.” Therwath suggests that a decision was taken in the 1990s to “emulate Muslim transnationalism” (417); if so, the emulation has not been very successful.

⁴²⁹ Thus one of the component processions of the Ekātmatā Yātrā set out from Nepal (Jaffrelot, Hindu nationalist movement, 360). See also Seshadri, RSS, 36–37; Gellner et al., Nationalism and ethnicity, 80, 166 and 183n8, 504–5.

⁴³⁰ For the Hindu nationalist conception of territory, see Jaffrelot, “From Indian territory,” 207–13.

⁴³¹ Literally “unbroken India” (akhaṇḍ Bhārat).

⁴³² The term “cultural nationalism” is applied to Hindu nationalism by Lal (“Economic impact,” 416); cf. also Keddie, “New Religious Politics,” 710.

⁴³³ Jaffrelot, Hindu nationalist movement, 49n166.

⁴³⁴ On the cult of Bhārat Mātā, see Kinsley, Hindu goddesses, 181–84; McKean, “Bhārat Mātā,” 250–65.

⁴³⁵ See, for example, Savarkar, Hindutva, 91–92; also Jaffrelot, Hindu nationalist movement, 26–27.

⁴³⁶ Savarkar, Hindutva, 3.

⁴³⁷ See 110.

⁴³⁸ Savarkar, Hindutva, 89, and see 4.

⁴³⁹ Golwalkar, We or our nationhood, 22.

⁴⁴⁰ Ibid., 40–41.

⁴⁴¹ Ibid., 24. See also Jaffrelot, Hindu nationalist movement, 54–55.

⁴⁴² G. K. Gokhale (d. 1915) in Parvate, Gopal Krishna Gokhale, 192, quoted in A. Sharma, Modern Hindu thought, 188.

⁴⁴³ The unsuccessful Telengana movement of 1969–1971, which expressed the resentment of the Telugu-speaking districts of the former state of Hyderabad at having been merged into Andhra Pradesh in 1956 and thereby exposed to Āndhra domination, faced an analogous problem on a smaller scale: its mostly Hindu leaders could hardly claim a separate identity on the basis of having previously been ruled and influenced by Muslims (Bernstorff, “Region and nation,” 139–43).

⁴⁴⁴ D. E. Smith, India, 88.

⁴⁴⁵ Golwalkar, We or our nationhood, 59–60; and see 63 on “the serai theory.”

⁴⁴⁶ Gould, Hindu nationalism, 58–63, 71–73, 76–84; also the summary at 265–74.

⁴⁴⁷ Golwalkar, We or our nationhood, 44. “Hindusthān” is a Sanskritization of the Persian form “Hindustān.”

⁴⁴⁸ The abandonment of the axiom of linguistic nationalism that Hindus should speak Hindi, just as Germans speak German, also reflects the continuing demand for English in the educational marketplace, even among Hindi speakers; thus brochures about a Hindu nationalist boarding school in Madhya Pradesh emphasized the teaching of English as a selling point (Jaffrelot, “Hindu nationalism,” 206–7; and see Kumar, “Hindu revivalism,” 550, on English as a medium of instruction in the schools of the Ārya Samāj). The RSS also sponsors the teaching of spoken Sanskrit (Seshadri, RSS, 125, 234–36) but without any apparent ambition to see it displace the vernacular languages.

⁴⁴⁹ As noted by Varshney (Ethnic conflict, 71).

⁴⁵⁰ Jaffrelot, History of Pakistan, 13–16.

⁴⁵¹ Talbot, Pakistan, 78, based on the detailed account in A. K. Jones, “Muslim politics,” 184–200.

⁴⁵² For attacks on Hindus after independence, see Jaffrelot, History of Pakistan, 37–38, 56; Jaffrelot, Hindu nationalist movement, 96.

⁴⁵³ W. C. Smith, Modern Islām in India, 25.

⁴⁵⁴ Jinnah, Speeches and writings, 2: 404, drawn to my attention by Qasim Zaman.

⁴⁵⁵ So Jinnah in a speech of 1944 (ibid., 2: 24).

⁴⁵⁶ The “Hindu state” (Hindu Rāṣṭra) of the RSS (Jaffrelot, Hindu nationalist movement, 44, 50, 114, 115) is a state that exists for Hindus, not an embodiment of the political values enshrined in the Hindu heritage.

⁴⁵⁷ Compare the politics of the Hindu Tamils of Sri Lanka: the Tamil Tigers never had to face the danger of being outflanked by a religious rival. In Palestinian terms there was no Tamil Ḥamās to challenge the secular nationalists.

⁴⁵⁸ Jaffrelot, Hindu nationalist movement, 551–52.

⁴⁵⁹ Ibid., 552. A telling detail is reported from an election in Karnataka in 1991: the BJP activists were hoping for fair weather on election day, and it was their competitors who wanted rain to discourage less determined voters (Manor, “Southern discomfort,” 181).

⁴⁶⁰ Jaffrelot, History of Pakistan, 230, 242–43, 245, and cf. 37, 231, 241.

⁴⁶¹ Ibid., 231–34, 259–60.

⁴⁶² To go back no further than the nineteenth century, see the vivid account of a Śaiva persecution of the Vaiṣṇava sects in one of the native states of Rajasthan in the 1860s given in Clémentin-Ojha, Le trident sur le palais, 95–107 (353 in the English summary). Dayānand and others saw sectarian division as a major aspect of Hindu degeneration (Jordens, Dayānanda Sarasvatī, 37–38, 51, 69, 129).

⁴⁶³ Ritambarā in the preamble to her speech in Hyderabad has a string of invocations related to Viṣnu (Sītā, Hanumān, Rām, Kṛṣṇa, Vālmīki) but only one related to Śiva, namely Lord Viśvanāth of Kāśī (Kakar, Colors of violence, 155–56). Yet this may reflect no more than the fact that the current agitation was about the Rām cult in Ayodhyā (Kṛṣṇa, the only invocation related to Viṣṇu that does not tie into the Rām cult, is doubtless there for Mathurā, site of another mosque slated for destruction, and the reference to Kāśī has a similar thrust, see Jaffrelot, Hindu nationalist movement, 547–48).

⁴⁶⁴ See Pew Global Attitudes Project, Muslims in Europe, 3, 11, 27. The wording used is “in India fully 90% of the public self-identifies as Indian rather than Hindu” (11), leaving a mere 10 percent identifying primarily as Hindu. I take it that “the public” in this context refers only to Hindus, but this is not quite clear. See also Kull, Feeling betrayed, 34.