« Non vi è nulla nel passato che in certe circostanze non possa provocare passioni nel presente. »

Arnaldo Momigliano, Le Radici classiche della storiografia moderna.

I

1One of the persistent recent trends in South Asian historiography has been to oppose the ‘State’, with its murky and violent history and destructive logic, to another rather more wholesome social site, which is ostensibly characterised not merely by relative stability but by an absence of constitutive violence, namely the ‘Community’. Recent authors as diverse in other respects as Burton Stein and Partha Chatterjee, have thus proposed to their readers that the locus of not merely healthy opposition to the state, but of a sort of pre-modern innocence, might be (or might have been) the community. To Stein, the long historical trajectory of South Asia could thus be schematised as the progressive passage from a situation of “communities without states” (prior to 800 BC), to one of “communities as states” (to 300 AD), to one where communities and states coexisted (to about 1700 AD), and finally to a modern situation of “decorticated communities” after about 1700. It was this last situation, in Stein’s view, that permitted the emergence into full bloom of modern-day communalism, as “what had been historically vital and changing community formations [were reduced] to decorticated shells of ideology” (Stein 1998: 21-35, cit. on p. 21). Less pessimistically perhaps, Chatterjee (1998: 277-82) has proposed that communities still form the basis of a possible project for the “internal transformation” of society in South Asia, if only they can renegotiate their relations with the state. In his view, such small-scale community-formations represent, even today, “a domain of politics that is located neither within the constitutional limits of the state nor in the orderly transactions of bourgeois civil society” (Chatterjee 1999: 117).

2While it may still be possible to see the disembodied spirits of Henry Maine and Karl Marx behind such formulations (in particular that of Chatterjee), that continue to oppose stable communities to relatively ephemeral states, this is surely not the most constructive manner in which to approach this literature. The present essay attempts therefore to take a different tack, arguing that in order to do justice to these arguments, it is necessary to investigate the historical processes that underpin the formation of communities, rather than taking them to be natural or organic outgrowths of the ‘essence’ of society. Processes of community-formation shall be treated here, for the most part, as processes of collective identity-formation, the theme that forms the principal preoccupation of the present volume. The central focus of the exercise will be to examine the relationship between violence, and the memory of violence, and the constitution of communities and hence of identities. The present-day relevance of such an exercise should be a mystery to no one, since the harnessing of past wrongs and past violence to present ends is very much a staple of contemporary South Asian political discourse. In keeping with my own area of competence, I shall primarily focus on violence in the late pre-colonial centuries, thus pushing further along an analysis of sectarian violence and the state that I had published some years ago (Subrahmanyam 1996: 44-80).

3We may begin with a rather simple observation: the formation of collective identities in contemporary India seems to be based in a very large number of instances on a foundational narrative or a myth of origin dating back to the precolonial period, and that is extremely violent in character. Such is the case with the foundation myths of a large number of sectarian groups and urban centres dating from pre-colonial times, but it is also all too often the case with the socalled caste-purāṇas. Nevertheless, what remains to be examined is the extent to which such narratives and sites of collective memory are transformed into loci of exclusion, and thus simultaneous of identity and of alterity; it is equally of significance to examine the means by which such shifts take place.

4A very well-known and rather controversial hypothesis in this regard is that of the English historian C. A. Bayly in an essay entitled “The pre-history of ‘communalism’?”, first published in 1985 (Bayly 1985: 177-203). Concerned for the most part with Hindu-Muslim relations, the most recent version of this essay, that appears in a collection by the same author from 1998, mildly modifies the earlier version but retains much of its construction and logic, while responding parenthetically to critics.1 Bayly notes to begin with, the existence of a very wide (but questionable) consensus in the historiography, to the effect that “the quality and incidence of communal violence changed dramatically in the last third of the nineteenth century”, and suggests on the contrary that “fairly extensive documentary evidence [exists] of Hindu-Muslim riots in North-Indian cities during the first half of the nineteenth century”, which, he suggests, “point back in turn to a range of conflicts which had taken place in the mid-eighteenth century.” Bayly is anxious not to deny the existence of what he terms a “widespread Hindu-Muslim symbiosis” of the pre-colonial and early colonial period, but proposes that such a symbiosis did not preclude the possibility of violence along communal lines. The point for him then is to tread the relatively narrow space between two extreme interpretations, the first the rosy view of standard Indian nationalist historiography, that before British ‘Divide and Rule’ policies raised Muslim communalism as a counter to Indian nationalism, such phenomena did not exist; the second, the view attributed by him to “some colonial writers and […] Maratha and Muslim communalists of the present century”, to the effect that already in the eighteenth century, a number of regional powers had risen up against the Mughals motivated by “exclusive strategies of religious revivalism”. In place of these two views, Bayly proposes that by an accumulation of evidence of conflicts on sectarian lines already in the pre-colonial period, one can identify the existence of “preconditions in [South Asian] social structures for sustained communal violence”. However, he also emphasises that “preconditions are not causes”, even if “they provided pre-existing lines of social fracture”.

5Of particular interest for our purposes is a part of the concluding section of the essay, in which Bayly sets out to present a review of the methodological terrain that he has covered. Here, he writes:

The notions of ‘identity’ or ‘consciousness’, so widespread in the literature now, whether applied to peasants or ethnic groups, seem most dubious on methodological and philosophical grounds, quite apart from the virtual impossibility of proving empirically that such entities ever existed. In some ways, the Annales term ‘mentalite’ seems much more acceptable, implying as it does a more variable, ambiguous and fragmented form of consciousness and one that is partly contingent on social and economic circumstances rather than constructive of them.2

6We shall return on another occasion to these comments, which are rather problematic in their own way, notably in their excessively sanguine view of the term ‘mentalité’, which is rather brusquely preferred to two other rather innocuous terms.3 Since the first publication of Bayly’s essay, the debate has continued, as we see from a recent critique by Asiya Siddiqi of a work by Lakshmi Subramanian on eighteenth-century Surat (which while it does not refer explicitly to Bayly’s position, is certainly aware of it).4 L. Subramanian describes in this work the emergence of what she terms an “Anglo-Bania Order” in Surat, that helped consolidate the Castle Revolution of 1759, and the English Company’s seizure of power in that declining port-city. Thereafter, she sets out to define the challenges that emerged to this ‘order’ in the last years of the eighteenth century, notably through the riots of 1788 and 1795, of which the latter is described in some detail. We are thus not in a pre-colonial political order but in the early colonial one, and the violence that is described, while relatively small-scale by twentieth-century standards, nevertheless endured several days in August 1795. The incidents followed on the seizure of a certain Khwaja Muhammad Riza Khan Bangali and his companion, on charges of attempting to break and enter into the house of a wealthy banker (sāhukār), Adit Ram Bhat. Since Khwaja Muhammad was the muezzin (ḵẖātib) of the Fath Masjid, the accusations against him were not deemed credible in many quarters. A crowd assembled to protest this, under circumstances that are not altogether clear, and eventually on the morning of August 6,1795, some of the main areas of the city where the Banias resided were attacked, and sacked, with some violence both to persons and property. Eventually, the forces of the English Company sprang into action against the rioters, and subdued and dispersed them. These events are presented by Subramanian as precisely those of a communal disturbance wherein a Muslim crowd rose up under provocation, and attacked the Hindu mercantile community, in order not only to express its resentment against very recent events but those of the past decades, in which the Anglo-Bania Order had been put into place to their detriment.

7In a brief critique of this work, Asiya Siddiqi has pointed to several problems at levels both of method and argument. This view of incidents in Surat, it is noted, is almost entirely constructed from the English Company record, and the work of Lakshmi Subramanian reproduces verbatim, and without a great deal of critical reflection, the terms and even the very phrases of the colonial archives. To Siddiqi, this explains why a communal construction is being placed on an incident, which seems to her far more easily explicable in class terms, as reflecting the conflict between poor weavers (some, but not all, of whom were Muslims), and rich merchants allied to the Company state (Siddiqi [review of L. Subramanian] 1999: 501-03). In contrast, to Subramanian, the class character of the rioting crowd is not at all evident, for where Siddiqi sees disenfranchised weavers, she perceives a “Muslim crowd”, drawn relatively indifferently from a Muslim community in Surat that laid great emphasis on “solidarity and cohesiveness”, through “daily congregations at mosques binding the entire community, the rich and poor alike”, and by other social manifestations like community dinners. Thus, despite admitting that “the elements that made up the city’s Muslim population were equally diverse” (as the Hindus), Subramanian is confident that the mobilisations of August 5-6, 1795, united the Muslims of the port-town, precisely because the bases for a single Muslim community already existed prior to the mobilisation, on account of such institutions as the “established tradition of annual pilgrimages from Surat to Mecca”, and in the more recent past, “the decline of Muslim political authority and patronage [which] served merely to reinforce the sense of religious solidarity” (L. Subramanian 1996: 211).

8This debate echoes, in rather curious ways, another recent exchange, this one around a well-known essay by Sheldon Pollock, seeking to demonstrate how the epic, the Rāmāyaṇa, came to be deployed as a political force in medieval India in a context where Muslims had seized power in many regions.5 Pollock had argued that the ‘mythème’ of the epic was “first deployed as a central organizing trope in the political imagination of India” from the twelfth century, and that once deployed, “the Rāmāyaṇa has served for thousand years as a code in which proto-communalist relations could be activated and theocratic legitimation could be rendered”. On the face of it, this argument differs considerably from that presented in the case of Surat, notably because Pollock refuses to engage in an analysis of the historical processes of community-formation, preferring to speak of the ‘imaginary’ in rather more mytho-literary terms. Yet, the most recent and most powerful of his recent critics, B. D. Chattopadhyaya argues that the brunt of the exercise is to demonstrate “how contemporary communal consciousness can be traced back directly to the early medieval phase of Indian history and that a construction of the history of communal consciousness does not have to make any reference to its colonial phase” (Chattopadhyaya 1998: 98-115, cit. on p. 115). To this, Chattopadhyaya adds a number of other telling critiques, both on matters of detail and larger substance, that cast serious doubt on the soundness of Pollock’s formulation as a whole. Yet, if one accepts Chattopadhyaya’s own position, the question posed by Bayly of a ‘pre-history’ of ‘communal consciousness’ (though, as noted above, Bayly is against the use of the term ‘consciousness’ in principle), before its ‘colonial phase’ remains unanswered. The debate thus straddles the divide without attempting a resolution, even as the question of the relative place of class and community in organizing violence in late eighteenth-century Surat opposes preconceptions rather than procedures. For Chattopadhyaya, the real issue is thus the “tension between non-monarchical and monarchical systems of governance”, and the emergence of state societies in the Gupta and post-Gupta periods. According to him, then, “it was in the logic of historical change in India […] that the discourse of monarchy was to hinge largely on the Rāmāyaṇa”.

9Thus, to sum up briefly, none of the four positions appears satisfactory for our purposes. One proposes that community identities exist in a latent form prior to the moment when they manifest themselves through events, because of the existence of institutions and loci of contact that appear neither necessary nor sufficient for the formation of such communities. A second reduces the issue of violence and community-formation to a simple problem of class-conflict. A third displaces the problem into the sphere of tropology, reviving in the process a curious form of structuralist argument (thus: the binary nature of the epic fits the binary nature of representation that the political imagination requires); while the fourth brings us full circle, to locate everything of importance where community-identity is concerned in the colonial period, while falling back on functionalist reasoning to explain all that is ideological in pre-colonial India.

II

10In view of the somewhat unsatisfactory state of play in the secondary literature then, it may be useful to turn to a series of primary materials directly concerned with issues of violence and collective identity-formation, to attempt to draw out from them a series of concepts with a wider application. The identities that will be focused upon shall, in the greater part of the cases, be rather more limited and restrictive than the very large identity-categories of ‘Hindu’, ‘Muslim’ and ‘Sikh’ treated for the most part by Bayly, as well as by the others whose work we have surveyed briefly above. Still these micro-narratives are not without their interest in my view, and can perhaps shed light on larger questions than their modest dimensions might suggest. Further, the identities that will be examined will not only be religious ones, but others, local, regional, even ‘ethnic’, in nature. The first narrative before me concerns a foundation-myth, that of the fortress-town of Chandragiri in Southern Andhra, located very close to the prestigious temple-town of Tirupati. The version I use comes from the celebrated Mackenzie Collection, put together in the early years of the nineteenth century, and is termed “Historical Account of Chandragiri [Chandragerry]”.6 It follows two other documents, one in the genre of a kālajñāna (pseudo-prediction), the other a statistical list of temples and water-works in the town and its environs.7 The text, which I shall closely paraphrase, runs as follows:8

After the long rule of the Cholas in the region, lasting some eleven hundred years, there followed the reign of Sriranga Yadava Rayulu, who founded the line of the Yadava Rayas. One of his descendants was Immadi Narasimha Raya, who was succeeded in turn by Surya Yadava Raya, who ruled for twelve years. This Raja once came to the area around Chandragiri to examine the state of the country there, and decided to build a fort on a hill that was to the East of Chandragiri, at two hours’ distance from it. He consulted his ministers who approved the plan, and so he began to build the fort. At this time, the hill and valley of Chandragiri were still overgrown with jungles.
It so happened that one day, two washermen brothers who worked for the Raja, had spread his fine clothes out to dry, and they were snatched up by the divine bird Garudalvar, and dropped on the branches of a tree on a hill where Chandragiri is now located. The older of the two brothers, called Chandra, ran after the bird, climbed up the hill, and located the clothes. While on top of the hill, he found two pools of water as well as a very fine tank, that was splendid enough to be frequented by the celestial nymphs. He then rushed back, prostrated himself before the Raja, and told him that the Western hill was far better than the site where he was constructing at present. It had more water (an important consideration in the area), and was considerably more spacious.
Surya Yadava Raya decided to inspect the other hill with his ministers. They then consulted and decided that the Western hill was indeed far better for their purposes. The Raja thereupon sent his workmen, who began a fort on the top of the hill, and another below it. However, each time they laid the foundation during the day, the part they had begun fell down during the night. The Raja was greatly alarmed at this, and began to worship Venkatesvara, the god at Tirupati. Three days later, the god appeared to him in a dream and told him that his project was indeed a good one. However, on the hill there were many bhutas, as well as inferior devatas, who had their place of residence. Two men had to be sacrificed to propitiate them, and only then would they permit the building to proceed.
The Raja awoke the next morning and recounted his dream to the ministers. He was filled with sorrow, asking himself whom he would sacrifice to placate the bhutas. At this, the two washermen brothers, Chandra and Giriya, appeared before him, and hearing the account of the dream, prostrated themselves at his feet and said: “O Lord! Do not trouble your mind so, over the need to sacrifice men for the success of your building. We cheerfully agree to it, on condition that the place be called after our names, so that our reputation shall forever flourish in the world.” The Raja was naturally overjoyed when he heard this offer. He gave great wealth to their families, and sacrificed them so that the building could proceed. Since the place was to bear the name of the two brothers, Chandra and Giriya, it came to be known as Chandragiri.

11At one level, this account, dated June 1802, holds no great surprises and resembles a great many other such texts collected in the region. Let us begin by noting the major common features in the body of texts such as this from the broad area, usually termed kaifiyats or sthala mahātmyams. The bulk of them are interested in tracing the origins of a place, and its subsequent history, and many of them seek out a royal founder, in the majority of cases one of the prestigious Vijayanagara rulers or their representatives (thus, a large number use Krishnadeva Raya as a foundational figure). In this respect, this text belongs to a minority group (albeit a statistically significant one) that lays stress on the pre-Vijayanagara period by claiming an enhanced importance for the Yadavas, or Idaiyars. It would appear that the eighteenth century witnessed the rise in importance of this group in Northern Tamilnadu, and Southern Andhra, and they sought to infiltrate themselves into the founding stories, or early histories, of a number of important places, typically taking up the intermediary position in terms of chronology between the Cholas and Vijayanagara. In this sense, this story is about the identity claims of a community in a process of social ascension.9 But going beyond this general view, it is of course striking that this is a narrative about a form of social violence that is regarded as perfectly justified. In this respect too, the story is not unique, for a very similar account appears in other texts, such as the socalled Velugoṭivāri vaṁśāvaḷi, where the name of the Recherla Gotra of Velamas is explained in very similar terms (an Untouchable servant called Recha or Echa, who offers himself for sacrifice, so that the lineage can accede to a treasure protected by a (vetāla) demon) (see Shulman & Subrahmanyam 1990: 225-48). As such, it presents us a rudimentary ‘folk’ etymology for the name Chandragiri, that derives then from the (probably) imagined (but still socially significant) sacrificial violence inflicted on a low-status group (the washermen), by the higher-status Yadavas, with the aim of placating popular deities and malignant powers. What is worth stressing, at the risk of labouring the obvious, is that we are dealing with the representation of a category of ‘incorporative’, rather than ‘exclusionary’, violence. At the same time, however, Mackenzie’s assistants also collected the rather more standard myth of origins of the place, which lays the stress elsewhere (oioc, Mackenzie Collection, General, vol. 25: 121-22, “Historical Memoir”). In this version:

The founder of the town and fort was Immadi Narasimha Yadava Rayalu, who while on his way to Tirupati, decided to settle at a place called Addakonda. Once more the washermen who are drying out his clothes (here, in particular, his red turban-cloth) find them carried off by a kite, which drops it by the pool of Agastya, on the top of a hill. The king begins to build there, but lacks a name for the place. At this stage, Brahmins from a nearby agrahāra inform him that his town is located at the very spot where Chandra (the Moon-God) had performed a penance on being cursed by Vishnu. He hence calls the place Chandragiri (Moon-Mountain). No sacrifice is called for, or mentioned.

12In such texts as these, where popular (or, at times, erudite) memory was called upon by later circumstance to describe the constitution of a community (here the community of a place), it was banal enough to find the use of violent motifs, which in some sense are regarded as sealing the ‘social contract’ of the place in question. This should take us to the first of our general observations, namely that the memory of social violence was not necessarily divisive, so long as it was not treated in the mode of a ‘grievance’. An even more striking case than this, one that I have dealt with elsewhere (and which will be the subject of a forthcoming monograph by Shahid Amin), concerns the site of Bahraich in Eastern Uttar Pradesh, where the foundation-story of the shrine of Salar Mas^cud Ghazi has been an important place of pilgrimage since the fourteenth century (Subrahmanyam 1996: 67-71; also see Mahmood 1989: 24-43). Yet, the community of pilgrims and residents (who are of very mixed religious affiliation) are not particularly perturbed by the fact that the standard tale of Salar Mas^cud portrays him as a nephew of the eleventh-century conqueror Mahmud of Ghazna, who was killed on the spot in disputes over disputes with local Rajputs concerning a Sun-Temple that he wished to destroy. This is once more a case of incorporative violence, for while the elaborate cycles of stories in circulation concerning Ghazi Miyan, from Bengal and Nepal, to the Punjab, never deny the core of this story, they choose to emphasise other, incorporative, pan-sectarian, aspects. As has been noted by recent historians like Shahid Amin, this has been the bane of a number of late nineteenth-century and early twentieth-century neo-Hindu propagandists (such as those of Arya Samaj), who have repeatedly attempted at the annual ^curs festival to convince Hindu devotees that they have no business being there, since Salar Mas^cud’s violence is directed at them (that is, at their ‘historical community’).10 We thus have an attempt to engineer a shift from the incorporative to the exclusionary, based on precisely the same core narrative of a violent past. In order to do so, it is proposed that the holding of grievance (and indeed, grievance-collection, an activity to which we shall turn below), is the only correct attitude towards past violence, whether real, imagined, or in some indeterminate epistemological space.

13This brings us to a second narrative, also from the same broad region, and the same broad-time frame as the story of Chandragiri, but with a very different character and trajectory. The Persian chronicle of Burhan Khan Handi, entitled Tūzak-i-Wālājāhi, was written in the 1780s to provide an authoritative version of the history of the family of Anwar-ud-Din Khan and his son Muhammad ^cAli Khan, semi-autonomous rulers (at times designated Nawwabs) of the region of Arcot, somewhat to the south of Chandragiri. Burhan for his text drew on an earlier chronicle in verse, but, as he informs us, he also incorporated “new facts which were discovered later and ascertained to be true”.11 In this text, the dominant organizing principle is genealogical, for the locus of memory is principally the family. The mnemonics of place, evident in texts of the kaifiyat genre, slide into relative insignificance here. Nevertheless, the term community (qaum, plural aqwām) is very much in evidence throughout the text, to speak not only of the Europeans, who are carefully divided up into several communities (English, Dutch, French, Portuguese and even Danish), but also of the other predecessors, rivals, and allies of Anwar-ud-Din Khan and his descendants. As for the members of the Walajah family itself, they are identified through their high lineage (with origins in the Quraish of Mecca), and their subsequent passage through Central Asia (in particular, Bukhara) before entering Hindustan. As such, Burhan Khan avoids giving them a clear ‘ethnic’ identity, but he is unsparing when it comes to others. Thus, referring to one of the early rulers of Arcot in the eighteenth century, Sa^cadatullah Khan, he notes that he was “from the community of the Nawayat (az qaum-i nawāyat ast)”, and adds:

The word Nawayat is the plural form of the singular nāit, a community of Arabs. There are different views about their origin. According to the view of the Tārīkh-i-Tabarī, they are of Bani Quraish. The author of the Tārīkh-i-Yamanī says that they come from the community (qaum) of sailors (mallāhīn). The writer of the Jāmi^c-ul-Lubāb says that they are nobles (shurafā’) of Kufa. [Tūzak, trans, by Nainar, pp. 64-65; Persian text, pp. 72-73]

14Of particular interest for our purposes is Burhan’s portrayal of the Afghans, who are given some importance in his history. Thus we have an account of Da’ud Khan Panni, the predecessor of Sa^cadatullah Khan mentioned above, who is described as a member of the qaum-i afghān. Burhan then adds that “the genealogy of the particular group (farīq) to which he belonged” was a matter of dispute amongst historians, but goes on to note that Muhammad Qasim Firishta, in his Nauras Nāma, had stated that they were “Copts (banī qubtī), survivors of the followers of the Pharaoh (az qaum-i fara^cūn), who on account of their disbelief in the Prophet Moses, were drowned by a miracle in the river Nile. Instead of remaining in their native homes, as became true believers, they left them and came to Hindustan” (Tūzak, trans. by Nainar, pp. 62-63; Persian text, p. 71). A later section, dealing with the “expulsion of that community” (ikhrāj-i ān qaum) from the Karnatak Payanghat region provides us with a further narrative that is relevant to our purpose. Here, Burhan begins by reminding his readers that the Afghans, characterised by him as the “wicked people of the day”, had settled in the town of Arcot during the rule of Da’ud Khan, and further that they were given to all kinds of mischief (sar-kashī). They were thus in the habit of harassing people in town, in terms both of their honour (ābrū), and their wealth and money (māl-o-zar). The shopkeepers of the town were greatly troubled by them, and they could not care less about the officials to the point of attacking young women on the streets. Thus:

If they saw young women on the road, it became Baghdad for them, and they imagined themselves to be the Caliphs. They molested them, and sent them back to their husbands, and many women swallowed poison out of shame, while their husbands died of sorrow.

15This state of intoxication (mastī) is said to have continued into the times of Anwar-ud-Din Khan (that is, the 1740s), when a certain Afghan by name Yadul Khan, in the midst of a wedding-ceremony, was instigated to stab Sa^cadatullah Khan, the ten-year-old son of the deceased Nawayat ruler, Safdar ^cAli Khan. When this matter was brought to the attention of Anwar-ud-Din Khan, writes Burhan, he decided that the Afghans had at last gone too far in their “foul deeds and great mischief (af^cāl-i namīma wa fasādāt-i ^cazīma)”. So, the account continues:

As a retribution for this great crime, as a precaution against such wrong-doers, and as a lesson to all the murderers, [he] punished him [Yadul Khan] with the maximum penalty and destroyed him. The other Afghans who were the inhabitants of the city of Arcot and other places in Payanghat were expelled with great disgrace, unspeakable molestation, and troubles. On account of such severe punishment, and of their fear of imprisonment, that mischievous community (qaum-i sharīr) changed their manners and dress (tabdīl-i auzā^c wa libās karda) and styled themselves Shaikhs. But, from their mode of speech, they were recognised and became like thieves in the hands of superintendents. To be brief, in a short period, there was not even the name of Afghan in Payanghat. [Tūzak, trans. by Nainar, 1934, I, pp. 110-11; Persian text, pp. 116-18]

16Several aspects appear to me of abiding interest in this version of things. First of all, the group that is being described is not one that has a distinct religious identity, but rather an ‘ethnic’ one, which if anything might be considered even harder to seize. The stereotyping is sharp, and draws not merely on current behaviour, but on origins. Thus, the identification with the Copts, who would have done better to stay on the banks of the Nile, is not a matter of mere chance in terms of what follows. Further, in this view of matters, the Afghans are identified throughout as a community, with certain collective forms of behaviour, as well as of dress, and indeed a distinctive manner of speaking (tarz-i-gūyā ’ī). Even when they want to hide these characteristics, and attempt to change their identity through a transformation of their external appearance, they cannot, and are found out soon enough.

17Violence too plays a clear role in this account, both the violence of the Afghans and that done to them. This violence begins early on in the text, in the description of Da’ud Khan Panni himself and his vicious twin dogs Khizr Khan and Bassu; it is stated that “when criminals were brought before him, he set these dogs on them”. Yet, it is not as if any specific acts of violence are used to define the collective identity of the Afghans, since what is in question is their ‘innate’ aptitude for violence. Thus, the only act of violence that is described in any detail is one that is used, paradoxically, not to form the community, but to dissolve it. Here, it would seem that the logic is one where the violence of the community begets the violence of the state, resulting in a form of pogrom, accompanied by “great disgrace, unspeakable molestation, and troubles”. It is in similar terms that other groups are periodically treated in the writings of the period from Arcot, for instance the Bundela Rajputs, who are seen as inherently seditious, even when far away from their homeland, and relocated in the northern Tamil country (Subrahmanyam 1999: 69-113). In their context, we have the harsh judgement of an early eighteenth-century chronicler, Jaswant Rai Munshi:

And since it is futile to expect good fruit from a bad tree, or try to cultivate in a barren field, the grandson [Tej Singh] too trod the path of his forbear [Bir Singh Deo], as the [Qur’anic] saying has it: “Everything ultimately returns to its source.”

18The Bundelas too, like the Afghans, represent one of those small groups that had accompanied the Mughals into Southern India, but eventually fell foul of state authority. As with the Afghans, they were also made to disperse with some violence, having been identified and stereotyped as a collectivity with some undesirable characteristics.12 Unlike the Afghans of the region, whose dispersal seems to have been relatively effective, the Bundelas have however managed to keep alive a collective memory of this moment of dispersion, largely on account of a powerful oral-epic, which is paradoxically not transmitted for the most part by themselves. Thus, even if a collective identity was not forged by the passage into violence (but preceded it, in a crucial way), one can see the relationship between violence and identity as mutually supportive in nature, rather than constitutive (in either the exclusionary or the incoporative senses delineated above). In an analysis of the early history of the Safavid political order in fifteenth and sixteenth-century Iran, Kathryn Babayan has pointed recently to the complex relationship there between the transmission of memory and the formation of a community. In the case of Iran (and Khorasan), the existence of a medieval epic, the Abū Muslimnāma, precedes the formation of the community, which arises in a crucial sense around the performance of the epic (which is also one centering around sectarian rebellion, repression and a notable violence).13 Such cases can arguably be found in South Asia too, but the situations we are considering here seem to proceed from a reversal of this logic; in contrast, one could argue for the constitution of a community around an epic in cases such as the Palnāṭī vīrula kathā (Roghair 1982).

19The flexibility of this relationship, between a violent act, the constitution of a textual corpus (whether in the form of epic or history), and the process of ethnogenesis has attracted the attention of a number of researchers dealing with South Asia in recent times. In keeping with the constructivist mood of recent years, however, the emphasis has too often been on the constitution of identity by the passage into textuality. A characteristic example of this is Michael Katten’s recent claim that the Bobbili Kathā, a corpus of sung epics relating to the defeat and destruction of the warriors of the Andhra coastal town of Bobbili by the Raja of Vizianagaram and the French Company in 1757, more or less created the caste category of Velama in the region (Katten 1999: 15-47). Thus, he writes of “the Velama creation of an epistemology of self through the telling of the Bobbili Katha”, suggesting that the repeated narration over generations of the violence that was visited on the warriors of the clans assembled at Bobbili in 1757 eventually produced an identity. This requires Katten however to go out on several untenable methodological limbs, notably by excluding other earlier texts that may be seen as crucially constitutive of Velama identity (and that precede the events at Bobbili by several centuries), and also by denying contemporary evidence from the 1750s that suggests a fairly distinct sense of difference between the Velamas of Bobbili and the Kshatriya Raja of Vizianagaram (cf. Narayana Rao & Subrahmanyam 1999: 15-47). We thus revisit the earlier debates surveyed in the first section of this essay, for one finds once more an anxiety on the part of constructivist researchers to attribute all change, all significant processes of identity formation, to the colonial period, rather than a proposal to examine the nature of what existed before, and what was transformed over the nineteenth and twentieth centuries.

20This brings us, in fine, to the third of the concrete cases to be examined here, after those of Chandragiri and the Afghans of Arcot. As distinct from the materials discussed above, we are concerned in this case with a Northern Indian situation, albeit one from the same broad chronological span, namely the 1740s. In this period, the celebrated courtier and litterateur, Anand Ram Muḵẖlis, himself a member of the Khatri community (a scribal and merchant caste) resident in Shahjahanabad-Delhi, found himself accompanying the Mughal emperor Muhammad Shah (r. 1718-49), on a journey-cum-military expedition in the Gangetic valley.14 Mukhlis was a reluctant traveller, not only because he was rather overweight and found travel fatiguing, but because he broadly knew the terrain over which he and his party were travelling, on their way to the petty fort of Bangadh. Yet, as he made his way across this familiar landscape, various things struck his fancy, evoking a memory here, or provoking the composition of a verse there. As a member of the Mughal elite, Mukhlis was conversant with both Persian (his preferred language of expression), and Hindi, and also composed in rekhta, a mixed poetic language of the period incorporating elements of the two. (As a native of Lahore, it is also certain that he knew Punjabi, though he rarely if ever used it as a literary language.)

21Eventually, some weeks into the voyage, the party arrived at the town of Sambhal, a centre of secondary importance in the epoch, and at some distance from the main highway. Ever eager to collect popular lore, Anand Ram noted that the town’s main gate was very high, with a peculiar feature, namely that it had a huge iron ring, with a circular mill-stone fixed in it. The story had it that a strong acrobat (bāzīgar) of the past had in one leap fixed the ring in the door-frame, and with the second, the stone in the ring. The challenge was thus open to all comers to try and equal this feat from the past, or to bring the stone down. Only those who could meet the challenge could pass through the door, and if not they should pass elsewhere. So, no other bāzīgar had since tried to pass through the door.

22A large part of the town, formerly considered to be one of the more established ones of the region, was in ruins by this period. The people of the town were as cultured as those of Delhi, in Anand Ram’s view, and he also states that Amin-ud-Daula, a well-known notable (of7,000 mansab rank) from the early eighteenth century had originally been from there. He had his own havelī in town, made up of several buildings, besides having constructed a number of gardens and marketplaces; in fact, even his tomb was there. Amin-ud-Daula’s mansion had a hammām and an āyina khāna (that is, a sort of glass house) in it.

23This brings Anand Ram in turn to the town’s major feature, for on the other side from Amin-ud-Daula’s residence was a high dome, that was at one time a temple called Har Mandal. Of this temple, it used to be said (in the vernacular dictum reported by Anand Ram):

Bhāg baḍe sambhal ke
ke Harjī har maṇḍal āvenge

Great is the fortune of Sambhal
where Harji (Shiva) will come to the Harmandal.15

24This reflected a sort of millenarian legend that one day, Lord Shiva would return in the flesh to Sambhal, which was thus privileged by this association. When Babur took Hindustan, writes Anand Ram, he had given Sambhal as a jāgīr to his son Humayun Mirza, and he in turn had converted this ancient building into a mosque, so that it was now the jāma^c masjid of the town. Anand Ram now goes on to cite the chronogram16 inscribed on one of its arches, noting that it was constructed on Babur’s orders by a certain Hindu Beg. The inscription ran:

He who gathers in himself all buildings of virtue and excellence,
Who raises high the flags of kingdoms and communities,
Who spreads the carpets of peace and safety,
Who builds the edifices of knowledge and good acts,
That king of high stature, Muhammad Babur,
May God keep him safe, and raise him high,
When the candle of his power was lit in Hind,
Sambhal was illumined by its rays.
In order to build this mosque,
So that it would be protected from harm,
He issued an order to this humble slave,
Who was one of the important supports of the state,
When the wise and understanding Mir Hindu Beg,
Who was exemplary in his moral conduct,
in keeping with the farmān of the Emperor of the age,
Completed the work with God’s grace,
The year, month and day were such:
Yakum az shahr-i Rabi-ul-awwal.17

25Yet, in the face of this transformation, what is the attitude of the ‘Hindu’ courtier, Anand Ram? “Earlier too it was a place of worship (^cibādat kadah), and even now it is a place of worship”, concludes Anand Ram calmly, as if this were the most naturel of things. Further, he notes the existence of a tank, now in poor condition, but still thought to be holy, where people came and bathed. Brahmans and flower-sellers too still came there in numbers to sell flowers and recite shlokas. Though the tank was almost dry, people bathed in it even then, using the sludge from it. This leads Anand Ram to compose a rather ambivalent verse in Persian that States:

After this, in place of tears,
the heart cornes out,
When the pond’s water dries,
earth cornes out.

26The regret that seems to inhere in this verse is a gentle enough comment on Anand Ram’s part, in keeping with the ironical vision that this eighteenth-century courtier had of his own life and times, and that of the Mughal court. Anand Ram, as has been pointed out at greater length elsewhere, was one of those Khatris of the period, who participated in a multi-faceted literary and cultural world, but also was ironical in respect of his own community, a group whose identity was torn between commerce and politics. On one occasion this had prompted him to make the following aphoristic statement: “Commerce (tijārat) is many times better than statecraft (imārat). Statecraft makes one subordinate while in trade, one leads the life of a ruler. The wealth accumulated by a noble is a misfortune, whereas the money earned in trade is lawfully enjoyable (wajh-i halāl)’’18 On the other hand, on another occasion (this one on the same voyage that took him to Sambhal and beyond), he reflected on the ineptitude of the Khatris faced with a battle situation, despite their pretensions to straddle the worlds of warrior and merchant. Noting the departure of several of his Khatri friends for Delhi, once preparations for a battle with Rohila forces were underway in Bangadh, he wrote sarcastically:

O friends, now that matters have come to arrows and swords (shamsher-o-tīr), why should we stay here, for we are not soldiers? We are Multawi Mal and Pakodi Das [Postponement Mal and Cutlet Das]; why should we then not leave for the city to do business there?19

27Rather like another community that has been better studied, namely the Kayasthas (a scribal and merchant caste), the collective identity of the Khatris was thus formed in the course of a complex, centuries-long negotiation, with a series of political powers in first Northern, and then peninsular, India.20 This insertion in Indo-Persian courtly culture may of course provide a simple explanation for some of Anand Ram’s response to what he saw at Sambhal, but to my mind the real reasons lie deeper. It is certainly true, on the one hand, that Khatri identity did not emerge by a straightforward process of forcing ‘alterity’ on some other group (and thus by a play of mirrors), but rather by a complex process of incorporative synthesis. This fact is somewhat unfortunately neglected in that part of the literature which assumes that the creation of ‘identity’ is primarily a negative project, one of defining that which one is not, and of depositing those attributes of which one wishes to divest oneself on the ‘Other’. However, historically, in South Asia as elsewhere, processes of identity-formation for communities have very often hinged crucially on the accumulation of (real or imagined) attributes rather than their transference. This allows us to understand why the physical evidence of past violence, which Anand Ram was confronted with, did not produce in him a sense of grievance. The choices before him were not, on the one hand, memory and grievance coupled together, and on the other, amnesia and hence forgiveness-by-default. It was still possible for him to remember and confront an unfortunate past with irony and equanimity, just as he could reflect with those very same qualities, on his own identity and that of his (explicitly identified) community.

III

28Bearing in mind the examples cited above, one may turn to the concluding set of questions that this essay seeks to address, namely the relationship between past violence and what one may call the creation of a veritable ‘grievance industry’ with respect to South Asia. Of course, one must hasten to add that South Asia is by no means unique in this respect; a recent work by an American historian who sought to study the figure of Ali Pasha of Ioannina, the late eighteenth- and early nineteenth-century Ottoman-appointed governor of Northern Greece, points to how relevant this question is for other parts of the world too. As the historian in question, Katherine Fleming, writes: “If I was not defending Ali, then surely I was on the Greek side. Otherwise, I was clearly a Turkish apologist. If Ali was not to be made into a saint, then he should be demonized instead” (Fleming 1999: 4). Thus, it was sufficient to present a scholarly communication on the economy of the Aegean islands in the late eighteenth century, suggesting that it witnessed some improvements at the time of Ali Pasha, to elicit the commentary: “My, my, you love him a lot! What do you want, to canonize him?” Scholars of the Mughal empire have surely been accosted with similar comments in recent times in South Asia (as well as the United States), if they suggested that anything happened in that period besides Muslim misdeeds and mayhem.21 In part, this is a reaction to an earlier official historiography (referred to by Bayly as well), which through much of the first three decades after Independence presented a saccharine view of South Asian society, and communal relations therein, in the centuries before colonial rule. But there is surely more to the strident contentions of recent times, than a mere reaction to another set of explictly ideological and reductionist histories. One recent suggestion is that History itself, as a discipline, carries the blame for such an exacerbation of matters; the claim, made most notably by writers such as Ashis Nandy, is that there are other forms of memory (say, in the rather crude terms of Nandy, ‘epic’ or ‘myth’), which are more organic, more sound, and hence far less apt to generate conflict than is history, allegedly a purely Western imposition on India (Nandy 1995: 44-66).

29Yet, more serious work than that of Nandy on the harnessing of collective memory towards contemporary political (typically, nationalist) projects, does not confirm the view that where history is unhealthy, other ‘traditional’ forms of memory act as an antidote.22 Still, the prejudice against history remains, even if in the case of much ethnic conflict, ethnography and anthropology cannot quite be seen as innocent disciplines either. In this context, the recent reflections of Valentine Daniel in relation to what he terms an “anthropography of violence” in late twentieth-century Sri Lanka are both problematic and instructive. In his extensive introduction to a brilliant, and in many respects chilling, work, Daniel argues that an useful contrast is to be made between ‘history’ and ‘heritage’, and contends that these are “two distinctive dispositions towards the past, each identified with one of the two ethnic groups in Sri Lanka, the Sinhalas and the Tamils” (Daniel 1996: 14). On to this already differentiated (or potentially differentiated) landscape is projected a new actor, namely ‘modem history’, which unlike ‘history’ as practised and understood earlier by Sinhalas, is “in its origins and development a largely European practice” (ibid.: 46). It is this ‘modem history’, for Daniel, despite his highly nuanced understanding of the society he examines, which is in many respects the villain of the piece, when taken together with its necessary companion, the modern nation-state. In this view, the equilibrium that had long existed between Tamil notions of ‘heritage’ and Sinhala notions of ‘history’ is disturbed by the irruption of modern history in the nineteenth century, and ‘heritage’ gradually enters into retreat, being by its very nature a non-event-centred view of coming to terms with the past. The key then is the fixation on the event, which for Daniel, is the object that “is held in common […] by all of European history since Herodotus and Thucydides”; and it is the tyranny of the event that leads in turn to “the erosion of heritage’s ontic moorings in its quest after an epistemic history”. Significantly, he argues that this process “is best illustrated by what has come to be known as the Ayodhya-Babri Masjid crisis, in which a section of India’s Hindu right destroyed a sixteenth-century mosque in trying to establish it was the birthplace of Lord Rama” (ibid.: 47).

30Without wishing to transform the discussion into a battle over disciplinary turf, this seems to me a rather questionable view of how history has functioned in twentieth-century India, and what its rôle in fact was in the events of December 1992. The Sangh Parivar-controlled group that tore down the Babri Masjid was certainly not doing so because it was “trying to establish it was the birthplace of Lord Rama”; the feeble archaeological evidence ‘produced’ (in every sense of the term) after the act could, at a stretch, have shown the existence of a medieval temple on the spot, not the fact that Rama was born there. The issue was in any case not an epistemological one, despite the wrong-headed attempts by some historians, leftist activists and journalists to turn the argument around the existence or non-existence of a medieval Rama temple on the spot. Rather the question was centrally one of whether such events (and I use the term with full cognisance of its implications) as the tearing down of a temple to build a mosque in the sixteenth century can at all be considered to be significant ‘grievances’ in the late twentieth century.

31Now, if this were a matter to be discussed in a court of law, one might conceive of the invocation of the epistemological spectre of history, since history and legal procedure have for long centuries shared affinities in matters of proof (and not only in the Islamic and Christian contexts). Yet, once more, this was not the issue in debate. Rather, the question was one of how to create a community of grievance, which would be done by mobilising groups (often highly privileged groups in today’s society) into a collectivity through a procedure of grievance-collection. In view of the rather heterogeneous community that was in the process of emerging, and gaining an identity (however ephemeral), all sorts of mechanisms were set in train for this, some based on an explicit mimicking of history, yet others on other manners of mobilising the past (including routes that worked through notions of what Daniel might term ‘heritage’). It is certainly true that it is the ‘historical’ (or ‘archaeological’) arguments that have most engaged the attention of professional historians, alarmed at the rise of a grievance-community of national proportions, one that was becoming a major force in politics besides. These would include such exercises as listing the ‘Hindu temples’ destroyed by the ‘Muslims’ (thus, the Hindus are naturally the aggrieved party, the Muslims those who must pay), or even more absurd still, the grotesque claims (reproduced in a section of the French press as well) that “between the years 1000 and 1525 alone, 80 million Hindus were killed directly or indirectly [...] by the Muslims”.23 Yet, these works, by writers such as Sita Ram Goel and Arun Shourie, can scarcely be thought to have more than a very limited potential for building the identity of an imagined community.24 These activities and procedures must be seen then as allied to a number of other, far more significant, modes of mobilisation, through rituals, processions, mass media, as well as the powerful neo-militarist organisations whose spread has been documented in a number of valuable studies (many of which are too well-known to require extensive citation here).25

* *
*

32To conclude, it would seem that neither history, nor indeed any other mode of memory, by itself can be seen as a driving element in the construction of communities, and group identities. The past, and the existence and recognition of violence therein, can serve various purposes, ranging from the incorporative, to the exclusionary, passing through a number of intermediary positions. It has been the contention of this essay, that in general, the perpetration of violence by one group on another need not lead to the automatic constitution of the idea of grievance, and from there of retribution. Nor does one need to pose amnesia as the sole meaningful alternative response. The historical passage, whether partial or complete, from other forms of the transmission of memory, to history, does not appear to me necessarily to have the effect of being the muscular midwife to the notion of systematic grievance either. It is certainly true that hubris-ridden historians (and, even more so, philosophers) of an earlier génération talked far too easily of history as catharsis, and the lack of historical knowledge of the past as condemning groups to repeat earlier errors (in the phrase turned cliché of the American philosopher George Santayana). This is a position that few will defend with confidence today. On the other hand, it is surely high time that we ceased blaming history for what is wrong with our relationship with the past.*