Emotion and the Person in Nationalist Studies

Michael Roberts, reproducing an essay that was written in 1993 for a conference in Japan and one which has appeared in Japanese in The Shinso, Jan. 1993, special edition on Nationalism Today ed. by T. Aoki, pp. 127-50; but has not seen print in English. Note that the path towards this essay was prompted by the literary piece which I  drafted in 1991: “The agony and ecstasy of a pogrom: southern Lanka, July 1983.” This was reproduced in Exploring Confrontation, Reading: Harwood Academic Publications, 1994.

Victim of communal pogrom, Varanasi, 23 December 1992

Perhaps the most dominant strand in the analysis of ethnicity and nationalism within social sciences has been that which can be described as “utilitarian,” “transactionalist” or “instrumentalist.” The means:ends relationship construed in terms of economic (and thus political) advantage is the dominant principle in capitalist rationality, so this is not a matter for surprise. The instrumentalist perspective comes in several variants. In one version, favoured especially by those emerging from the social sciences in U.S.A., ethnic groups are viewed as interest groups (e.g. Paul Brass 1974 and 1991). Another version is found within the transactionalist emphasis favoured by several anthropologists who emerged from British universities, such as Fredrik Barth (1975), Freddie Bailey (1966 and 1969) and Abner Cohen (1969 and 1974). Emphasis by highlighting is my imposition today.

Burnt Buildings, 1983 Colombo

Let me use Cohen’s work to elaborate upon the theoretical slant associated with this approach. Cohen’s ideas were initially clarified through his ethnographic monograph (1969) on Hausa migrants in the Sabo quarter of Ibadan, Nigeria, a region where the Yoruba were the host majority. His subsequent essay on Two-Dimensional Man(1974) provides a considered and convenient summary of his position. In keeping with the dominant paradigm in British social anthropology Cohen adheres to a structural functionalist line:

… although ethnicity involves the extensive use of old customs and traditions, it is not itself the outcome of cultural conservatism or continuity. The continuities of customs are certainly there, but their functions have changed. Within the contemporary situation ethnicity is essentially a political phenomenon, as traditional customs are used only as idioms and as mechanisms for political alignments (1974: 3-4, emphasis added).

Thus, the Hausa in Sabo are presented as an “ethnic group [that] adjusts to the new realities by … developing new customs under traditional symbols, often using traditional norms and ideologies to enhance its distinctiveness within the dynamic contemporary situation.” In such a context men (sic) may ridicule the customs of the other tribal/ethnic groups, but they “do not fight over such differences alone. When men (sic) do … fight on ethnic lines it is nearly always the case that they fight over the distribution of power, whether economic, political or both, within the social system in which they take part” (1974: 93, 94). On the foundations of his case study Cohen then proceeds confidently towards an universalist proposition: “ethnicity in modern society … is the result of intensive struggle between groups over new strategic positions of power within the structure of [a] new state” (1974: 96, emphasis added).

What we have, then, is that universal figure, the Strategising Human. Such a line of emphasis could mesh neatly with the interest group theory favoured by the social sciences in North America during this period. This is revealed, for instance, in the work of Cynthia Enloe (1973) and Judith Nagata (1974); and in all those writings that regard the play of ethnic identity as “situational” — sometimes to the extent that it is treated as if it were a garment that could be discarded or draped on by actors as a matter of strategic choice (e.g. Okamura 1981; Vincent 1974).

The emphasis on situational fluidity is also in tune with more recent post-modern writings that stress “rupture” and “fragmentation”— though post-modernist readings of the world in anti-holistic ways would not be comfortable with the highly structured schemata favoured by Abner Cohen et al. However, what I would like to highlight here is the continuing force of a more venerable intellectual tradition, namely that of Marxist orthodoxy and its different but related form of instrumentalism. A classic case is Eric Hobsbawm. His writings on the subject (1972; 1984, and 1990: 54-57, 89, 117-19, 155-62) are permeated by a form of utilitarian functionalism as well as a fatal effort to reach sociological universals from empirical material assessed in positivist as well as a-sociological ways (e.g. 1990:5ff).

In describing the emergence of nationalism in Central and Eastern Europe from 1870 onwards, Hobsbawm lays particular emphasis on the role of those whom he calls, variously, “the lower middle strata” or “lesser middle strata” (1930:116ff). In this view, linguistic nationalism is a vehicle for their advancement. It is a means:end issue. Little weight is attached to the emotional attachment to a vernacular, a vernacular that was, more often than not, viewed by the literati of Western Europe as inferior to the languages of the so-called “historic nations” in Europe – an attitude of superiority that could only rouse the ire, and thus the nationalist sentiment, of those attached to their folk languages.

Typically, Hobsbawm seizes upon Kumari Jayawardena’s work to extend this thesis to the emergence of Sinhalese nationalism in the mid twentieth century and the ensuing Sinhala:Tamil conflict (1990: 158; also see p.vii). Jayawardena had in fact anticipated Hobsbawm’s thesis in arguing that in Sri Lanka “language was mainly connected with certain class interests of the petty bourgeoisie” (1984: 164ff). In this view the adoption of a Sinhala-oriented language policy was a means by which the Sinhalese petty bourgeoisie could secure “the prestigious posts held by the English-speaking elite.”

Unlike Hobsbawm, however, Jayawardena is aware of the emotions attached to the language issue,¹ but only touches upon it (1984: 165). In other words, her awareness remains subordinated within a rigid Marxist schema and its reflectionist view of ideology. Thus, it is asserted that neither the big bourgeoisie nor the working class and peasantry had a deep interest in the Sinhala Only programme (1984: 164-65).

The latter part of this argument misrepresents the complexities of the situation, and fails to grasp the ramifying implications of the language issue — implications which reached into the pores of working class subordination just as much as they stimulated the anger of idealists drawn from all classes and/or status groups. This is not the occasion to develop my counter argument (see Wriggins 1961 and 1960: 215-70 & 326-69; Kearney 1967; Farmer 1965; Gunatilleke 1962; Roberts 1978 and 1979: 39ff, 59-64 & 70-83; and Roberts et al 1989: chap 1 for elements supporting this position). What interests me here is Jayawardena’s instrumentalist and materialist emphasis. Implicitly, the reasoning is that where there is no economic advantage in a language programme – “advantage” that is defined by the analyst with limited evidential support from the thinking of the protagonists — there can be no genuine commitment to a cause. In this vulgar Marxist line of reasoning,² I should stress, Jayawardena has not been alone among the Asian scholars grappling with what is called “communalism.”³ The writings of Bipan Chandra and Ashgar Ali Engineer, for instance, are framed on the same lines (e.g. 1979; 1984 and 1987).

The instrumentalist interpretation of ethnic conflict is also widespread in popular ‘explanations’ of such events in the twentieth century. It is a venerable part of Sri Lankan historiography. Ever since G.C. Mendis, from within a non-Marxist perspective, attributed the “communal conflict” to competition for jobs and political clout among “the middle class” (1943), a long line of scholars have emphasised the significance of economic competition in generating Sinhala:Tamil conflict — without, however, necessarily seeing it as the primary factor and without necessarily seeing it as an issue confined to “the middle class” or the “petty bourgeoisie” (e.g., Wriggins 1961: 314; Kearney 1967 and 1973: chap 5; Tambiah 1955; Fernando 1970; Wilson 1979; and Roberts 1979: 70-79. Also cf. K.M. de Silva 1981: 496-503, 510-24).

What we have here, then, is a loaves and fishes argument. Such an evaluation is often based on solid evidence, just as the popular wisdom on this point is based on experiential knowledge. Likewise, such instrumental objectives can be discerned within the incidents of ethnic violence in Sri Lanka known as “riots.” The best known of these are the Sinhalese attacks on the Mohammedan Moors in 1915 and the attacks on Tamils in 1958, 1977 and 1983. In each of these “riots” one has to allow for regional variation in the pattern of attacks. In all these “riots,” there was significant populist involvement, though state agents are widely believed to have participated in the violent activities of 1983.

Among the populist actions in all these “riots” was the practice of looting of immoveable property belonging to the victimised community. In Ratnapura District in 1915 looting was so predominant that a British administrator described it as “the loot” (Thaine 1916).⁴ Even in 1983 looting was a facet of the assailant activity,⁵ and investigations in Colombo indicated that on occasions policemen participated in looting as well.⁶

No scholar has reduced such “riots” to the looting objective. Quite rightly so. Many scholars, however, have emphasised the economic goals and the instrumental objective of Sinhalese seeking to weaken their competitors in trade and other business. Thus, with reference to the “riots” of July 1983, Newton Gunasinghe underlines the “organised attempts on the part of certain chauvinistic elements, both within and outside the regime, . . . [to break] the economic base of Tamil entrepreneurs,” a programme which resulted “not only in attacks against small Tamil shop-keepers but also in attacks against major industrial establishments.” Gunasinghe then draws “a co-relation” between the open economic policy” initiated by the U.N.P. regime in 1977 and “the frequent occurrence of rioting” from that date (1984: 198).⁷

Again, in analysing the “riots” of 1958 James Manor has called attention to the economic jealousy directed against Tamils for their relative success in securing government jobs, and the interests of Sinhala traders in looting the premises of their Tamil business rivals, as being among the factors contributing to the attacks (1989: 292-93). Likewise, with reference to the anti-Moor riots” of 1915, a number of scholars have pinpointed the economic advantages to Sinhala traders in burning the goods and premises of Moor traders; and the generalised hostility to the Moor traders and moneylenders for what many Sinhalese deemed to be “rapacious” business activity (Jayasekera 1970: 100-11, 271ff, 281ff & 321; Kannangara 1984: 132, 140; Jayawardena 1970: 228-29; Scott 1989).

That such instrumental goals directed the activity of some Sinhala activists who participated in the attacks on the Moors in 1915 is undeniable (Roberts 1994: 189-201). But the relevance of this argument should not be built up in ways that obscure or underplay the emotive force of procession disputes over the making of music in front of mosques. These disputes date from the 1890s. In an era marked by a religious struggle for hearts and minds, the several confrontations on this issue created intense resentment among some Sinhalese Buddhists. They considered the demands of the Moors, and the British policy on this point, to be an injustice (ayuttak) which interfered with their traditional practices of sabda pîjâ or “noise worship” (Roberts 1990 and 1994: chap.7).

The different perspectives which induce George Scott and Kumari Jayawardena, the one transactionalist (Scott 1989) and the other Marxist (Jayawardena 1970: 223-24, 228-29 and 1984: 122-24 & 133), to underestimate the force of the religious issue and the intense concern it caused⁸ in the years 1899-1915 is a pointer to the shortcomings of the instrumentalist lines of reasoning. The need to dwell on these shortcomings is rendered urgent by the complacency with which influential scholars take the primacy of utilitarian motivations as givens: “There is . . . a broad consensus amongst scholars that communal competition is grounded  in competition of material interests,” affirm Hamza Alavi and John Harriss in their edited work on South Asia (1988: 171) — surely a classic instance of self-referential confirmation supplied by one’s very own network of intellectual affines.

This essay seeks to displace rather than replace the instrumentalist line of causal reasoning. It is an exercise in displacement. It is an invitation to readers to step outside the means:ends rationality which permeates our world today and to also attend to the force of emotion during moments of ethnic tension.

In itself, attention to affectivity and emotion in the world of ethnic interaction is not novel. A.L. Epstein, among others, has developed this emphasis (1978: chaps 1 & 3; also see Nash 1989: 5ff, 33-39, 84-85) in what is in many ways a debate with his friend, Abner Cohen, an instrumentalist and materialist par excellence. Again, Kapferer has explicitly addressed this issue in his comparative study of Sinhalese and Australian nationalism (1988 and 1989).

Kapferer’s narrative mode, however, is highly cerebral. It is cast in the clinical mould of semiotic structuralism and Western philosophy, albeit modified by phenomenological emphases. In consequence, the intersubjective experiential world of actors, of assailants and their victims, appears in a relatively subdued manner in his book.⁹ Thus, in publishing an essay recently on “The Agony and the Ecstasy of a Pogrom: Southern Lanka, July 1983,” I presented it as an alternative mode that nevertheless complements the approaches of a Kapferer or a Richard Handler – as well as, say, the more empiricist perspectives embodied in a range of analyses of the 1915 pogrom (e.g. Roberts 1994: chap.8; Kannangara 1984; Rogers 1987: 188-94). “Agony” illuminates the range of passions that inspire “riots.” It is a move in the direction of understanding zealotry, while yet rendering vivid the agential complicity and the horrors perpetrated by chauvinists and all assailants. It is part of an ongoing project that is seeking to understand the social conditions and social relations which can so transform an individual as to overwhelm his/her body with emotions of rage or whatever — so that the emotion has manifest, felt, bio-physiological components which unify mind and body in powerful ways. I am thinking, here, of anger that inspires rapid-fire vituperation, of grief which induces total collapse, of fear which turns one’s legs to jelly, of rage which pushes one into acts of amok¹⁰ and of a Muslim in England whose body was overtaken by cold tremors when he heard what Salman Rushdie had said in hisSatanic Verses.¹¹

The latter, that Muslim, I shall call the “True Believer.” There were other such Muslim true believers, though — and this is important — not all Muslims in England were so transformed. The point here is that the activity of assailants during the riots in Sri Lanka was promoted by atrocity stories which circulated during the course of the riots, and perhaps even preceded the catalytic moment. The significance of such rumours has been brought out in both my previous essay on the subject (Roberts 1981) and Kannangara’s important article (1984).  Rumours are also known to have had devastating effects during several communal riots in India – for instance, before and during the anti-Sikh pogrom of November 1984 (Srinivasan 1990: 316; also see Spencer 1992: 267-71). My interest in zealotry, therefore, encompasses the conditions that made (make) some people, both men and women, receptive to the atrocity stories (mostly fabricated) which circulated (circulate) at the height of tension. My interest is in the openness to such stories, how a person becomes “wounded by hate” (a phrase borrowed from Pete Seegar) and how these emotions empower the retailing of such stories from true believer to receptive others. And thence leads, en masse, to arson, destruction, assault and massacre by the males.

In pursuing this interest my first step has been to jettison the label “riots” (Roberts 1988 and Roberts 1994: 184-85). This word has been part of the British legal lexicon for several centuries and historians have slavishly followed this terminology. “Riots” draws no qualitative distinction between a sustained assault on a segment of the population and a minor affray for an hour or so involving a few score people. The term “riots” must be replaced with that of “pogrom” — in the pre-Auschwitz sense of the term as it was widely used in Europe. In that usage, as a dictionary would tell us, “pogrom” refers to the organised massacre, often state sanctioned and/or inspired, of Jews (see K.M. de Silva 1988: 67). But the Oxford Dictionary was “compiled effectively from the 1880s to the 1920s” (O’Connor 1989: 53). And it does not tell us that the English word “pogrom” derives from an old Russian word that simply meant “destruction” and that, over time, this meaning was lost and “the word came specifically to denote the destruction of Jewish life and property” (Kochan 1957: 12).

My usage does not depend solely on considerations of etymological origin. Rather I am situating myself self-consciously in the 1990s, with all its retrospective advantages, when I propose a metaphoric extension of the label “pogrom” to describe widespread and/or intense attacks on a subordinate segment of the population by a dominant section¹²(with a further distinction being drawn between “pogroms” on the one hand and “genocidal programmes” of the sort perpetrated by the Nazis and the Khmer Rouge on the other). This extension is deployed as an act of consciousness raising; and endorses the use of this term by other writers on the subject (e.g. Piyadasa 1984; and Newton Gunasinghe and Shelton Kodikara in newspaper articles in Sri Lanka). The extension applies beyond Sri Lanka. Several events described as “race riots” by the historians of U.S.A., where Whites assaulted Blacks in a sustained manner (e.g. at Atlanta, 22 September 1906: see Crowe 1969), should, in this view, be re-categorised as pogroms. So, too, is the massacre of Sikhs at Delhi in November 1984 a pogrom (see PUDR 1984; Srinivasan 1990; Das 1994; Van Dyke 1996). It is likely that many such massacres, in the Balkans, in Turkey (of Armenians, Greeks), and elsewhere could also be so redefined.

The second step is to qualify the spirit of positivistic empiricism that governs the work of so many historians who investigate such events. Hobsbawm has recently asserted that a Fenian or an Orangeman could not produce “a serious study of Irish history” (1990: 13). I read this as a clarion call for dispassionate analysis, for distancing acts. Such distancing is also effected by forms of bourgeois respectability and/or intellectual rationality that lead both journalists and scholars to describe the assailants in pogroms as “bloodthirsty hoodlums,” “thugs,” “paranoids,” “rabble,” “savages” (e.g. Vittachi 1958: 67, 69: Engineer 1984 and 1987).

This is the language of separation by stereotyping, a displacement and a casting of responsibility upon those beyond the margins of normality. This method of separation, or displacement, is supported by the general tendency among so many people, whether victims, witnesses or commentators, to describe the crowds of assailants as “the mob.”

It is my contention that such a separation is misleading. Under specific conditions, killing is a potential within a normal human being, a pogrom is a potential within the normality of social relationships, a capacity within the human, “politics and peace by other means.” Thus, to side with Jonathan Spencer, some episodes of ethnic violence can be “one moment in a longer political process.” And some facets of killing and destruction, “far from being an irruption of unreasoning pre-social passions [may] display … a logic that is both moral and collective” (1992: 266, 271).

To grasp this, to address the specific conditions that draw out these potentialities and render it a practice, an analyst should attend to those occasions (if any) when he/she could have assaulted, and thus come close to killing, some other; or when he/she vituperatively wished the death of A.N. Other – and even fantasised the wish. Let me note here that, having indulged in such fantasies on the odd occasion, that has been my starting point.

This starting point may also assist one to qualify the belief, so pervasive among the intelligentisia, that the assailants in a pogrom are in a state of frenzy. In fact the range of emotions expressed by assailants in pogroms would seem to be quite varied. There are assailants who move about in a cool, calculated style, doing their job or work (see Roberts 1994: 317-19). There are moments of ecstatic, ebullient action. There are moments of rage, frenzy.¹³ Indeed the evidence from lynchings and race riots in southern U.S.A. suggest that some individuals move through a range of passions from rage to euphoria as the killing caps off the hunt (see Ginzburg 1961 and Crowe 1969).

To use one’s own experience to understand, albeit in part measure, the embodied emotions of zealots is not to be subsumed totally in their cause (as Hobsbawm seems to understand this course). Even should this come to pass, Frantz Fanon’s Wretched of the Earth (1963) and Black Skin, White Masks (1968) reveal the insights that can emerge from such involvement. One does not, however, have to go as far as Fanon. A modus vivendi is feasible, as Renato Rosaldo has sensitively revealed in his Culture and Truth (1989).¹⁴ In writing “Agony” I set out on such a journey in a highly engaged manner — a journey which I commenced in Virginia in the autumn of 1991 well before I read Rosaldo, a journey into which I was driven by an urge within me, rather like the “thirst” which inspired Basil Fernando to pen his indictment of the 1983 events in blank verse (see Wijesinha 1988: 21-24 or Roberts 1994: 320-22; and letter from Fernando to Roberts, 19 Jan. 1991).

Two moves have been outlined thus far in this attempt to comprehend the inspirations that motivate and condition the violent activities of nationalist chauvinists and zealots – namely, their receptivity to atrocity story and their subsequent perpetration of atrocity. There is yet another step, less essential perhaps but useful. We require a narrative mode that goes beyond the typical empiricist commentary that is widespread today.

The latter, speaking broadly, is of two types, types that sometimes overlap. There is, on the one hand, the annalistic reportage of “what happened” by journalists and other historians, reports which usually adhere to a chronological framework (e.g., Sinha Ratnatunga 1988: 26-69; Piyadasa 1984: 78-98; Hyndham 1984; PUDR 1984). There is, on the other hand, the analytical coverage provided, with varying measures of insight, by social scientists, for instance Blackton 1970; Jayawardena 1970; Jayasekera 1970; Ameer Ali 1981; Kannangara 1984; Rogers 1987: chap.5; Tambiah 1996: chap. 3; and Roberts 1981 & 1994 for the 1915 pogrom; and, for aspects of the 1983 pogrom, Obeyesekere 1984; Tambiah 1986 and 1996: chap. 4; Jeganathan 1997 and Kapferer 1988. That such narratives can be contribution to knowledge is attested by the degree to which I have been aided in my understanding of racial violence in southern U.S.A. by essays written in this empiricist style by Crowe (1969), Prather (1984) and Williamson (1984); and by the insights on communal violence in India provided in the articles by Srinivasan (1990), Veena Das (1994) and those within the covers of the book edited by Paul Brass (1996).

But having granted this, my project here is directed towards urging the social science community to make itself open to other modes of representation — as ways of telling that complement empiricist tales, semiotic decodings and other abstract approaches by literally laying bare, revealing, the agential engagements in ethnic violence. One obvious path is that of the poem (e.g. Krysl 1997, Fernando in Wijesinha 1988). Another would be that of collages made up of visual scenes, eyewitness tales and commentary.15 But there also are narrative forms of the sort essayed by Taussig (1987), Trawick (1990 and 1997), Sarvan (1998), Daniel (1997) and myself (1994: chap. 13).

It is the latter essay, “Agony,” drafted initially in 1991, which has inspired me to make this plea. It is not that I self-consciously set out to work up a different style of prose narrative that was a counterpoint to the conventional academic analyses of riots and pogroms. I was not moving in a reasoned way towards a vivid literary evocation. It is only in retrospect, in part prompted by the responses from a range of scholars (mostly non Sri Lankan: see fn. 16), that I contend that “Agony” represents one of the alternative modes of telling, something that is in the line of old-fashioned storytelling, able thereby to recapture the immediacy of the moment of pogrom. “Agony” depicts the searing currents that run through a pogrom by presenting vignettes, chilling vignettes, from that of July 1983. These cameos, hopefully, should sensitise readers to the force of emotional affect — thereby foregrounding the limits of instrumental rationality in comprehending such affects.

These are immodest claims. The issues surrounding ethnic violence are too critical for modesty to submerge the paths to better understanding. What needs emphasis, here, is that I did not set out in deliberate fashion to carve out a telling literary representation. “Agony’s” effects are accidental. Its moral, nevertheless, lies in the medium. The narrative style is itself an argument. When I penned those lines, I felt urgent.16 I was a driven person. If it has succeeded in moving its readers, if it has conveyed some sense of the immediacy and horrors of that week in July 1983, it is because there was fervour in my storytelling.