Time's Arrow: Violence and Ethnohistorical Surrealism in the Lost Colony



Standard ethnohistorical accounts of colonial encounters downplay violence and genocide in the interest of holistic ethnographic accounts. In particular, “upstreaming” is a technique that attempts to sublimate violence into interpretations of ethnological understanding. An argument is made for “ethnohistorical surrealism” as a methodology for preserving the ruptures of colonial contact. The principle of incommensurability is an important driver for the transformation of contact zone exchange systems into structures of violence. Drawing on Bataille's reading of the potlatch, exchange systems based on unreturnable gifts produce abjection in the recipient, which results in cycles of violence and ultimately subjection.

“To Dream a Race”

There they go, to the day's work, with their heads stretched back. I was puzzled at first but now I know why they do it, why they stretch their throats like that. They are looking for the souls of their mothers and fathers, their women and their children, gathering in the heavens—awaiting human form, and union. . . . The sky above the Vistula is full of stars. I can see them now. They no longer hurt my eyes.

These familial unions and arranged marriages, known as selections on the ramp, were the regular high points of KZ routine. It is a commonplace to say that the triumph of Auschwitz was essentially organizational: we found the sacred fire that hides in the human heart—and built an autobahn there. But how to explain the divine synchronicities of the ramp? At the very moment the weak and young and old were brought from the Sprinkleroom to the railway station, as good as new, so their menfolk completed the appointed term of labor and ventured forth to claim them, on the ramp, a trifle disheveled to be sure, but strong and sleek from their regime of hard work and strict diet. [Amis 1992:123]

So the writer Martin Amis, in his novel Time's Arrow, imagines the last conscious moments of a Nazi doctor who had worked at Auschwitz. Having moved to the United States he is able to fashion a new life, and by the time of his death in the 1990s, he has had a long, successful career as a physician. The title of the novel is ironic, as, contrary to the inexorability of time's movement, the old man's life is imagined as a film played backward. Not only time but also the absolute meaning and ethical value of actions are reversed. Men and women are restored to health, “gifts” of gold for their teeth and jewelry are freely offered by the camp guards. In the end, the “preternatural purpose” of all this is, he says, “To dream a race. To make a people from the weather. From thunder and from lightning. With gas, with electricity, with shit, with fire” (Amis 1992:120).

To “dream a race” is in its way a project of much of ethnohistory and historical anthropology, especially that of indigenous peoples of the Americas. William Fenton, the distinguished Iroquoianist, averred that he “never intended that all ethnohistory should be written backwards,” but one could well argue that this is precisely what Fenton's key methodological concept, “upstreaming,” has led to (1962:12). Reversing time's arrow—in effect, imagining all those reunions among relatives and loved ones, living communities long since extinct—withdraws as well the material arrows, musket balls, and dagger points.

This erasure of the violence of the contact zone is not only ethically suspect but also elides the essential nature of the relationship between Europeans and natives that characterized it (see Ferguson and Whitehead 2000). Indeed, violence so saturates the encounter between Europeans and indigenous peoples that it can be said to be constitutive of it. The view of violence as somehow extraneous to the fundamental relationship, the notion that a purely “ethnological” level can be recovered, as Fenton proposed, has an air of futility like Lady Macbeth's hand washing.

Violence at Roanoke

In the historiography of the Lost Colony of Roanoke Island—the first British attempt at colonization on mainland North America—distinct phases of the treatment of violence exist. In the 19th century, the typical rhetorical maneuver was to examine the issue as a moral one located in the savage breast. Reflecting broader cultural constructions of savagery and nobility, this viewed the “problem” of violence as one of a lack of evolutionary advancement (see Berkhofer 1978; cf. Jahoda 1999). Thus, the presumed death of the colonists at Roanoke becomes an icon of martyrdom. In her famous and popular epic poem, The White Doe, Sallie Southall Cotten (1909) depicts the death of Virginia Dare—the first and likely only child born in the Colony—as an act of transubstantiation. Having been transformed by sorcery into a doe, she is pursued by hunters, intent on consuming her white flesh (implying both cannibalism and miscegenation). On being shot by an arrow, her heart blood fertilizes the very earth with the essence of civilization, bearing an immediate fruit of the scuppernong grape, and a more long-term one of the promise of civilization. Thus, her death, tragic although it is, will redeem those future inhabitants of a land now no longer savage. The Christological aspect of this (doe, rather than lamb, the redemptive quality of blood) reminds us that within the Western tradition violence connects more directly with religious and aesthetic concerns than even with politics.

By the 20th century, more concern with English culpability in the violence of the Lost Colony was exhibited. A 1921 silent film carries the caption: “Thus are sown the seeds of unfriendliness between the Indians and the English from which the White men reap a rich harvest of famine and death” (Anonymous 1921). In reference to the harsh collective punishment ordered by Governor Ralph Lane for the theft of a single silver cup, this seems an accurate summary. Of course, locating the roots of violence in the European breast, although a distinct improvement on the discourse of savagery, has its limitations as well. Michael Oberg (1999) has argued that much depends on the sort of Englishman one had. Those who were more “cosmopolitan,” the educated classes coming mostly from London and the Thames estuary, were less likely to foment violence, whereas those whose careers had been forged on the “frontier,” especially in Ireland, were more likely to do so. This model has some explanatory value. Certainly those, like Ralph Lane, who had served Walter Raleigh in Ireland, were most often the instigators of violence in the Colony. However, the model also suffers from its own formality. It is precisely this sort of dualism that was applied to the Indians in the 19th century. Thus, Manteo and Wanchese were seen as diametrically opposed examples of Indianness, much as we would tend to view John White—the artist and governor of the colony—and Ralph Lane. Although one could well remember Marx's dictum that people make history, but not of their choosing, the issue here has as much to do with the limitations of formalism as those of methodological individualism. Clearly, the locus of violence must be located in the analytic space between individual and group, structure and event, and in the dialogic space between groups.

Arguing along these lines, Seth Mallios (2006) has recently documented discrete examples of violence at Roanoke, Jamestown, and the Jesuit outpost at Ajacan. The nub of Mallios's argument is that in these cases of contact with mid-Atlantic Algonquians, and by extension in all cases of intercultural “first contact,” relations between the two sides can be captured in their material trade, which tends to begin with gifts among elites, moves toward a broader commodity exchange, and ends with violence. Along the way, the system is nudged from one steady state to the next by “exchange violations,” such as failure to reciprocate or receive gifts, theft, violation of trading protocol, and so forth. In essence, there is a decline along the scale of reciprocity, as defined by Marshall Sahlins (1972), from generalized to balanced to negative—with, in some cases, an eventual collapse altogether. One conflict that arose in these three cases is that the Algonquians tended to value ceremonial gift exchange more than did either the English or Spanish, who were in fact eager to push the relationship in the direction of commodity exchange. This led to a series of conflicts with the local people, especially the werowances (hereditary chiefs), who were mostly interested in establishing alliances and acquiring the spiritual power of the English leaders through a series of gift exchanges. The colonists were hungry and in all three cases were part of an enterprise designed to turn a profit, in terms either of land and commodities, or of professions of faith and fealty among the natives.

Thus, violence is a transaction, which is similar in type, if not in value, to gifts, barter, and theft. One can move along the scale in either direction. (As several examples at Jamestown and Roanoke attest, it was in fact possible to repair damaged relations with appropriate gifts.) The economic nexus is surely important in understanding violence, if for no other reason than the ironic fact that much gift exchange consisted of weapons or potential weapons (e.g., iron implements). Moreover, it is worth recalling the underlying notion of Mauss's theory of gift economy: the establishment of a bounded moral community. Gift exchanges at Roanoke, Jamestown, and elsewhere were thus an exercise in social imagination. To locate a person or group within the bounds of reciprocity was to imagine them as equivalent to self.

Abjection, Incommensurability, and Ethnohistorical Surrealism

Although this perspective serves us well, it takes us only so far in understanding the phenomenon of violence arising from exchange regimes. The religious and aesthetic dimensions of violence—so important in both European and Algonquian culture—are left out of this sort of model. I am sympathetic to the call for a “poetics of violence” to recapture some of these additional meanings, which must have driven human behavior at least as much as economic concerns in these contact zone situations (Whitehead 2004). Much if not most violence at this early stage of English colonialism consisted of performative acts, employing a principle of synecdoche through the selection of a sacrificial victim. The notion of sacrifice reconnects with exchange, as the culmination of cycles of production and exchange (Mauss 1990:16). Indeed, I am in full agreement with Mallios (2006) that violence is rooted in exchange. By teasing out the implications of Mauss's theory, in particular the question of what happens when the gift cannot be returned, we can approach an understanding of contact zone violence.

Bataille's famous reading of Mauss is particularly apt here. Although balanced reciprocity represents a bounded moral community in a state of homeostasis, other forms of giving and taking are linked to ritual violence, which is their ultimate meaning. Thus, excessive giving and sacrifice, as in the potlatch, is for Bataille a means of creating abjection in the person of the recipient, who has no means to reciprocate (1985:123). It is clear that such abjection may be created not simply by an overwhelming quantity of goods, but by extraordinary quality as well; thus, in the potlatch, the famous “coppers” have enormous value, owing to their individual histories and to what Marx called “social relations of things” (i.e., commodity fetishism; see Marx 1976:165; Mauss 1990:24). These were objects of such great and individualized value that they could never be returned. That is, they were incommensurable with any possible return gift (see Harkin 1997:127–132).

This is directly relevant to the colonial context, as small items of European manufacture and material were the preferred gifts from the English to natives. These were in a concrete sense incommensurable with anything the native people could themselves offer, to such an extent that, when terms of trade were agreed on, they exacted a huge price in terms of native food and labor (i.e., the “beads for Manhattan” phenomenon). That is, they created a state of abjection among the natives, because of their inability to return the gift without paradoxically removing their ability to participate in future exchanges (i.e., by losing their land and/or ceasing to be a distinct social entity). Bataille's perceptive interpretation of this type of exchange is that it represents the creation of class distinction, and specifically the creation of a permanently abject “working” class (1985:125). That is, a quantitative or qualitative difference in gift exchange results in a qualitative difference between donor and recipient, where only a quantitative difference had previously existed. To translate this into the colonial context, we have the creation of a subaltern identity.

In a similar way, human sacrifice, as may have existed in the potlatch and certainly did in the contact zone, creates an irreparable breach between the roles of donor and recipient, or “magister” and “minister.”1 This, too, is accomplished through the medium of an incommensurable good, in this case a human life. The killing of certain individuals, by both sides, is thus part of a larger strategy to place the other side in a permanently inferior status position. Kidnappings and murder or the threat of murder (as in John Smith's case) were used strategically in early postcontact North America to attempt to assert domination over the other. Of course, in the long run it was only the English who successfully accomplished this.

This process is precisely congruent with the explicit goals of the Roanoke Colony. As the first civilian English colony in North America, it was above all a commercial enterprise. The combined efforts of the governor and artist John White and the priest and writer Thomas Hariot were trained on portraying the lands as fertile and the people as amenable to some form of benevolent servitude (Campbell 1999:55; Harkin 2008:125–126). At the same time, the stakes of this asymmetrical exchange—and, no doubt, of the painting, surveying, and cadastral mapping going on—were clear to the native people, who intentionally disrupted the ongoing relations through acts of theft, according to a cultural logic also present two centuries later at Kealakekua Bay (Sahlins 1981:42).

Two moments in the history of the Roanoke Colony illustrate the two poles of this relationship. In the engraving by Theodor de Bry of John White's watercolor, which de Bry titled A cheiff Ladye of Pomeiooc, we see a young girl holding an English-made doll and rattle, while her mother, the chief's wife, looks on forebodingly, a crooked finger subtly gesturing to the girl to be still. In the background, typically for de Bry, is an array of resources, in this case laden fishing boats and ducks or geese flying overhead. The girl cannot resist the attraction of the toys, which are clearly incommensurable gifts, linked to the entire productive landscape (and, we can see, to the future generations of the coastal Algonquians themselves). This eerily prescient picture sets up an absurd equivalency, which nevertheless is enacted. This is one example of what we might call “ethnohistorical surrealism”—a condition based on the proposed equivalency of incommensurable and wildly disproportionate values and actions. I am intentionally drawing on James Clifford's notion of “ethnographic surrealism,” characterized by “fragments, curious collections, unexpected juxtapositions” (1988:118). However, I differ with Clifford in my understanding of the relation of ethnography to surrealism. Although it is true that some intellectual figures trained by Marcel Mauss in the French ethnographic tradition (Bataille, Métraux) possess surrealist sensibilities, it seems to me that the main thrust of ethnography within all national traditions is to erase surrealism by reference to some model of holism (see Thornton 1988). Similarly, ethnohistory, especially the technique of “upstreaming,” attempts to repair the breach created by surrealism. What I am arguing here is that in certain circumstances, Clifford's notion of surrealism well characterizes the experience of historical moments from the native point of view.

Ethnohistorical surrealism and the incommensurability principle are evident again in the second incident. In 1585, a party under the command of Richard Grenville explored the coastline of Pamlico Sound, to the south of Roanoke. While exploring Secotan country—which White and Hariot considered the exemplar of a civilizable colonial society—a silver cup was stolen. In retaliation, on July 16, the English burned down the village of Aquascococke. As the journal of the Tyger states the matter: “One of our boates with the Admirall was sent to Aquascococke to demaund a siluer cup which one of the Sauages had stolen from vs, and not receiuing it according to his promise, we burnt and spoyled their corne, and Towne, all the people being fledde” (Quinn 1991:191).

The precise nature of the cup is not known, but to Algonquian eyes it would have appeared similar to the silver chalice used by priests, obviously of great inherent value, and assimilable to Algonquian notions of sacred power inhering in material objects: mantoac (cognate of the northern Algonquian term manitou). Regardless, the material itself and its manufacture were like nothing they possessed. In this sense, the theft of the object was appropriate, because its incommensurability made reciprocal exchange impossible. At the same time, from the English perspective, the destruction of the village and its fields was a token of its true value, via Bataille's“principle of loss,” and prophetic of a colonial future (1985:118).2 Such objects functioned in part as demonstration pieces of the technological superiority of British society. Like the better-known demonstrations of weapons and firearms, these objects would have caused at least as much anxiety, especially given their connection to British magicoreligious practice.

Cultural Orders of Violence

Indeed, although violence is undeniably rooted in material relations, it is useful to consider it as a mode of semiosis and,3 in particular, as an utterance with pragmatic and metapragmatic dimensions.4 As Whitehead (2004:69) argues, violence may be considered part of a structure of symbolic reproduction, which ultimately relies on the ritual construction of outsiders for the purposes of identity. In other words, one needs a dialectically constituted category of otherness to produce notions of self. This may take a range of forms, including reciprocal exchange and alliance, but the heart of the matter is violence and its transformations (Bloch 1992:44). Thus, for coastal Algonquians, the English represented a powerful source of mantoac, or sacred, other-than-human power, particularly with respect to weapons and metal implements (which possibly could be refashioned into weapons). In addition, disease, which may be understood as microbial, but which tribal peoples have almost universally considered a type of sorcery, presented a real and material power to destroy. Finally, according to its practitioners, Christian ritual possessed the power to conquer death itself, similar to the power of Algonquian sorcerers. (These forms would be joined in later contact situations by biomedicine—see Harkin 1997.) The common theme among these valued forms of English mantoac was clearly the ability to deal death to others and to avoid it oneself.

For the Algonquians, violence and warfare were inherently sacred activities, and were accompanied by elaborate rituals (Quinn 1991:424–425). These rituals were both preparatory and celebratory. Within the indigenous political system, warfare was a common means of establishing relations of incorporation and vassalage, much as in feudal Europe. Thus, specific identities were always in a state of play, subject to the result of diplomatic and martial gambits. With the appearance of the English (and, to a lesser degree, the Spanish at Ajacan), the local tribes viewed them as both an opportunity and a danger. Attitudes among local chiefs varied, from the alliance offered by Manteo, to the opposition of Wingina, who nonetheless sought to obtain the benefit of English ritual and technology (Quinn 1991:377). Acts of violence against unarmed colonists, such as the unfortunate George Howe who was gathering crabs, must be viewed as acts of sacrifice, which had the multiple function of appeasing the god of chaos, Kiwasa, acquiring the sacred power of the English, and creating abjection among the colonizers with an eye to incorporating them, as was likely the case with the Roanoke colonists (Hariot 1972; Quinn 1974:432–481).

For the English, the meanings of violence were somewhat different. The product of centuries spent attempting to subdue the folk of the Celtic hinterlands; the English warrior class developed an ideology that foreshadowed later theories of cultural evolution, which justified the use of violence to suppress the “barbarous” elements in these societies, in places such as Cornwall, Scotland, and particularly Ireland. The historian Nicholas Canny has argued that a settled dichotomy between the kerne—violent mischief makers—and the chorle—dull-witted peasants well suited for drudgery—shaped English practice in both Ireland and North America (see Canny 1973, 1976:161–163). Violence was thus performative and demonstrative, and designed to defeat or to destroy the small elite in subject societies, making available their human resources, as well as opening up territory to colonization. This praxis was rooted in the Irish experience, but was refined and subjected to an increasing degree of metapragmatic self-awareness, to the point where the use of violence became a highly conscious utterance in the dialogue with indigenous peoples.

The English conception of violence is of course no more practical than that of the Algonquians. Both notions were founded on highly symbolic ideologies of self and other, which were performed largely through acts of violence. Moreover, both cultures placed violence at the heart of their religious beliefs. Renaissance Christianity, in both its English and Spanish forms, emphasized the cult of sacrifice. Whether that referred to the scapegoating of “barbarous” elements within indigenous society or the death of Europeans seen as martyrdoms (the Jesuits at Ajacan, Virginia Dare at Roanoke), blood sacrifice played a central role in the colonization of Virginia.

Sacrifice and Sentimental Identification

As many analysts of sacrifice have noted, a strong paradoxical identification between sacrificer and victim is often produced (Bloch 1992:36; Richardson 1994:83). This leads to the perverse effect of displaced feelings of empathy with the victim. Although this tended to be less the case in historical practice (“examples” were usually made with little empathy), it does appear quite soon after the fact. National literatures of settler colonial societies are often founded on the displacement of existing indigenous groups with imagined ones, who play the role of fictive ancestor to a pseudoindigenized settler society (Berkhofer 1978:86–96; Goldie 1989; Harkin 2007).

In the broader context of English conquest of Virginia, and its accompanying genocide, this exists in the history writ large, for that evolutionary ideology on which English colonization rested permitted no survival of native society. The purported disappearance of native tribes (despite claims by numerous groups in tidewater Virginia and the Carolinas) transforms them into the sacrificial victims par excellence. Pocahontas, much beloved by contemporary Virginians, is well suited for this role (and plays a direct counterpoint to her predecessor Virginia Dare). Ironically, at a fair marking the tercentenary of Jamestown, the existence of contemporary indigenous communities was overshadowed in the public eye by the historical associations of the Powhatans and especially Pocahontas (Gleach 2003). Pocahontas remains an oft-claimed ancestor of contemporary Virginians, in part because of the “Pocahontas exception,” which removed the taint of miscegenation from otherwise white descendants of the Indian princess. Such a pedigree represents an important status marker in modern Virginia society, especially during the 20th century (Ellinghaus 2006; Harkin 2007). As Amis's doctor states the matter: “My position on the Jews has always been without ambiguity. I like them. I am, I would say, one of nature's Philo-Semites” (Amis 1992:152).


  • 1

    The notion of sacrifice in the context of exchange has been a vexed issue within anthropology. Mauss sees sacrifice as establishing a reciprocity with the gods, a point that Godelier vehemently refutes (Godelier 1999:194; Mauss 1990:24). According to Godelier, sacrifice “is not a business deal,” and the gods remain fickle regardless of what is done to appease them. My thinking is closer to that of Bataille, who stresses the interpersonal dimensions of sacrifice. Bataille's reading suggests that it be explained by the “principle of loss” (Bataille 1985:118). That is, irredeemable loss is created by certain activities, such as sacrifice, which are paradoxically parasitical on productive activity and which at the same time result in heightened production and exchange.

  • 2

    It is telling that the terrorism of destroying native villages persisted for over three hundred years in British North America—through the 1880s in British Columbia (see Gough 1984).

  • 3

    Here, I argue against Foucault's position that the semioticization of violence is a form of sublimination: “a way of avoiding its violent, bloody, and lethal character by reducing it to the calm Platonic form of language and dialogue” (Foucault 1980:115; see Goldie 1989:85). On the contrary, I wish to recapture the specificity of violence in the historical record, as against upstreaming discourses.

  • 4

    Silverstein's (1976) notion of metapragmatic discourse is quite relevant to understanding these sorts of events. That is, linguistic metapragmatics represents an awareness of the rules of pragmatics, and a self-awareness of the effect of one's utterance; one can see how that sort of awareness is necessarily a component of such actions. Indeed, the British colonial experience from Ireland onward consisted largely of refining the pragmatics and metapragmatics of violence. One could probably construct a definite set of rules governing the conditions under which certain actions (kidnapping the son of a chief, firing on a village with cannons) were to be taken.