1Time, memory, and the obliteration of memory have been central topics in recent scholarship on early imperial Roman culture. We have seen sophisticated treatments of changes in the conceptualization of time, the relationship between past and present, particularly with the advent of monarchy, and the ways in which modes of marking time and memory reflect shifting senses of community. However, these treatments remain predominantly metropolitan, or at least focused on «the Romans» as actors and agents, strongly interventionist in the cultures of empire. Thus, for Mary Beard, one of the functions of the «Roman ritual calendar» was «to define and delineate Roman power, Roman history, and Roman identity». For Denis Feeney, «The systems under examination include… above all, the experience of empire, by which the Romans took the temporal consciousness of a city-state and meshed it progressively with foreign time systems as their horizons expanded to embrace the entire Mediterranean, and beyond»1.

2The subject of memories of the past in the Roman imperial world offers an excellent opportunity to unpack such notions of «Roman power, Roman history, and Roman identity». Explicitly or implicitly, the relationship between Roman and local pasts has tended to be modeled on a combination of modern regimes in which pasts have been deliberately obliterated or overwritten, and, the exaggeratedly coherent and unifying products of modern European nation-states. Notorious examples include: the depiction of 1975 as Year Zero by the Khmer Rouge in Cambodia, itself a self-conscious echo of the French Year One of the First Republic; or the condemnation of India as inhabited by «peoples without history» before British rule and the collection and composition of «Indian history» as a recognizably modern western genre of writing; or the French educational system that Vietnamese children living under French rule in Saigon, to recite, «Nos ancêtres sont les Gaulois»2. It is thus especially interesting that, as Karl-Joachim Hölkeskamp has recently complained, the explicit discussion of the «cultural memory» of Rome, in all its theoretical complexity, has been largely restricted to the scholarship of continental Europe, and particularly to French and German work3.

3These have largely been the default models for the Roman empire, tending to lead to assumptions about the Roman potential and will to «erase» certain local histories, or alternatively about Roman domination as the sole raison d’être for historical consciousness. Indeed, models such as these are suggestive for the Roman imperial world up to a point. Roman imperial power could certainly be expressed as the assertion of control over time and history as well as geography. The destruction of Carthage and Corinth in the same year has rightly been seen as an assertion of influence on both local and global memories4. Occasional spectaculars are one important and effective feature of Roman imperial rule, one aspect of the empire’s «economy of force»5. The relationship between individual festival calendars and any «master-list» in the early imperial period remains tantalizing, with annual rituals of oath-taking and anniversary celebration one of many strategies to assert connection with the emperor and to create an imperial centre6.

4In many contexts, however, the division between «Roman» and «local» time, or «Roman» and «local» pasts is far from clear-cut. Notions of what we might think of as «Roman» time are often complicated precisely because they are individual local expressions, embedded in the local experience: we might think of the multiple different «provincial eras» commemorated by individual cities7. Not one but multiple different interpretations of «Roman time» are inscribed in local calendars. The most notorious example is that of the resolution of the assembly of the province of Asia in c. 9 BCE to begin the year on the birthday of Augustus, make adjustments to align the length of year to that of the Julian calendar, and rename the first month «Caesar». This most self-consciously «Roman» calendar has very little correspondence with the civil and religious calendar of the city of Rome, which begins on the kalends of January. «Synchronization» in the Roman imperial world was at best partial, probably very much so in lived experience8. As for «Roman history» as a literary exercise, it is worth reflecting on the fact that classical narratives of the «fall of the Roman Republic» and the advent of monarchy that are canonical for us are largely rooted in the Greek accounts of provincial writers, such as Plutarch, Appian and Dio9. These accounts raise interesting questions about who has a stake in Roman history, to whom Roman history «belongs», and what makes history «Roman».

5Thus, a unifying model of «Roman time» is frequently inappropriate. At the same time, influential studies of the invocation of the past in modern nation-states have also shifted notably in recent years, less comfortable than they once were with leaving unchallenged projections of unifying memories and histories. The multi-volume work directed by Pierre Nora, Lieux de mémoire, the first section of which was published in 1984, has been very influential for modern historians, and the tag of the title is something of a catchphrase for classical historians. The content and concerns of Nora’s final section, «Les France», published in 1992, with its jarringly plural title, has been the least influential for our field, perhaps because of Nora’s odd omission of the perspective of France’s colonial subjects here, despite upfront engagement with diversity of views within France proper10.

6While classical historians and archaeologists have engaged for some decades now in versions of postcolonial theory, recent studies of cultural interaction in modern empires might encourage us to move further beyond a tendency to think in terms of mutually exclusive categories of «Roman» and «native» and to interpret «native» cultural expressions as «resistance»11. Recent scholarship on the British Empire would go some way towards blurring the categories of «insiders» and «outsiders», and would emphasize the impulse shown by subjects both to «write the centre» and to appropriate and remodel cultures, symbols and ideologies of the imperial power to place themselves, to stake claims, with numerous «internal» or «local» dynamics rather than an exclusive focus on the imperial centre12. In terms of writing pasts, recent work on the Indian subcontinent would emphasize the drive towards points of connection between metropolitan and colonial histories, such as in explorations of liberty and liberalism in the Indian histories of nineteenth-century intellectuals such as Ram Mohan Roy, histories that invite parallels with contemporary British traditions13. Studies of this kind are interesting and suggestive for understanding the dynamics of the Roman empire, although I would want to insist ultimately on the particularity of the latter.

7Treatments of «local» pasts in the Roman world have regularly assumed a geographical division on very broad lines between «east» and «west», a division that has periodically been challenged but that still underlies much scholarship on the Roman empire. It is the west that is associated with «peoples without history», a tag that Eric Wolf had meant ironically in his Europe and the People without history. If for «the west», the assumption has been that Roman rule entailed the obliteration of local pasts (if indeed there were any to start off with), scholarship on the so-called Second Sophistic has traditionally suggested quite the opposite, the deliberate denial of the politically compromising last centuries of the Hellenistic or Republican age, with a strong preference for the remote, glorious, collective «Greek» past14. Scholars have tended to suggest that Roman time in the Greek East was restricted to the present tense, and some examples play this theme out well, including Guy Rogers’ important study of C. Vibius Salutaris’ «foundation» of Ephesus in 104 CE, re-enaction of the essential, local, and Greek roots of the city that could bolster up her sacred specificity for a present in which Roman institutions and a Roman centre were so clearly visible15.

8Greg Woolf, over the course of a number of different articles and his book Becoming Roman, has argued that we should think of patterns of commemorating and forgetting pre-Roman pasts that follow what he sees as a dichotomized «eastern» versus «western» geography of imperial dynamics. Thus, in «The uses of forgetfulness in Roman Gaul», he argues that the western provinces are to be characterized by «the absence of any independent memory of a past before their conquest by Rome», their intellectuals «effacing» their own identity as provincials. The choice between particular «independent» pre-Roman pasts and a «wholly», «common» Roman past is mutually exclusive. His metaphor of «forgetfulness» is in line with his more general view that western peoples, internalizing Roman notions of humanitas, more or less voluntarily elect to bring themselves up to Roman standards of civilization16.

9Woolf’s use of the terminology of memory and «forgetfulness» locates his discussion within modern studies of collective memory, but it also seems valuable to consider briefly the Latin and Greek terminology of remembering and forgetting that is quite frequently applied to the question of origins, and that appears in some interesting discussions of cultural continuity or disruption through contact with other peoples. The two aspects come together in Dionysius of Halicarnassus’ argument that the Romans «forgot» many, but not all, of the Greek customs owing to their ancestry through the relatively late influx of barbarian peoples, even if they were not totally barbarized (Ant. Rom., 1, 89, 3). This is rather remarkable in Dionysius’ view, and he cites a counter example of forgetfulness, Eleans who settled by the Black Sea and became the most savage of all barbarians, despite their original, impeccably Greek heritage17. We might be reminded that, as Harriet Flower has shown recently, in the case of the dead, forgetting and oblivion is the default position in Roman culture: preserving memory requires effort and positive action18. We can already begin to see some distance from Woolf’s use of «forgetfulness» to characterize western peoples, and from the dynamics traditionally posited in studies of modern empires and nations: for Dionysius, «forgetting» can be partial as well as total, while Augustan Rome is hardly a «colonized» people according to modern conventions. The question of whether or not a given people retains memoria of their origins and, if so, of what kind, is a regular, and obviously laden, question in Roman ethnography. This question seeks to find out where to place peoples according to the long-standing location of peoples in broad relationships to each other via descent from heroes, reflecting the relatively integrative mindset of societies in the ancient Mediterranean19.

10In contrast to Woolf’s model of «western» forgetfulness, it is only in rare cases that Roman authors assert that a people have no memoria of their origins, such as the Elder Cato’s criticism of the Ligurians, on the grounds of illiteracy and mendacity20. Tacitus in the Agricola does not use the language of forgetting, but leaves open the question of whether the Britons’ ancestors were indigenous or immigrant, a fact that cannot be pinned down sufficiently «as is the way amongst barbarians». As one cannot get far by asking them, Tacitus suggests that we can make deductions from the various physical types, German, Spanish and Gallic likenesses. Only in the case of the Britons most like the Gauls are we encouraged to note likenesses in customs as well as in physique, perhaps because the Gauls are reckoned to have something more closely approaching «culture» than the other peoples cited. Tacitus’ northern barbarians are variegated21.

11Although his Britons, that most northerly island people, are off the map in terms of ancestors, his Germans in the Germania are attributed not just an elaborate variety of myths of origins, but «ancient songs», a primitive historical form that Tacitus with perhaps ironic analogy describes as their only form of memoria and annales. These myths include the earth born god Tuisto (or whichever textual variation we accept), and encounters with Hercules and Ulysses, to whom allegedly an altar was dedicated, together with his father Laertes. These international heroes are, significantly, imagined to have been mere visitors of the Germans rather than their progenitors. The «earth-born god Tuisto», father of Mannus, himself the father of three sons for whom three German tribes are named, is neither an «international» hero, nor a straightforward aetiological construction working back from the name Germani. There are interesting implications both for Roman conceptual geography and for the possibility that we are looking at something more than a simple Roman invention of a German progenitor22.

12Enigmatic questions about whether there might be German self-accounts of their origins somewhere in the background of Tacitus’ versions of descent myths invite us to consider broader issues of local histories and their relationship with what is sometimes called «great history». These modern questions start with Felix Jacoby’s Atthis, and his exploration of the interrelationship between local histories of Attica and a more universal, or Greek «history» written by Hellanikos, Herodotus or Thucydides23. Recent work by Simon Price and Katherine Clarke has taken issues of interplay between «local» and «great» history down to the Roman period. There are great advantages in taking a long view of this phenomenon, not least that it is helpful to think beyond narrower characterizations that become too focused on the dynamics of Roman rule. While Price is concerned with the «Greek East», and Clarke bases her study on the Greek accounts of Jacoby’s fragmentary local histories, there are important implications that might extend beyond these particular geographical or linguistic contexts. Price would make much more subtle the distinction between centre and periphery and between panhellenic and local mythologies. «Manipulation» of mythology in creating and asserting accounts of origins is not by any means peculiar to the periphery; myth is a flexible means of conceptualizing the past that can accommodate the changing political focuses of Mediterranean world beyond the polis over long periods of time, whether we are thinking of inter-polis relationships, of Hellenistic kings or of Roman emperors24. Continuity of these processes into the Roman imperial period is more easily demonstrable in the Greek East, but not out of the question for certain western peoples at least. It is useful to extend more generally this subtlety of distinction between center and periphery, local and «great» to Roman imperial context, lest we become too preoccupied with the centralizing scope of some grand universal narratives of world history or Roman Republican history.

13Moving beyond the calibration of origins, we can think more broadly about the presence of absence of historical narratives before the Roman conquest for northern and western peoples. Once again, there is something of a sliding scale: recognizable historical narratives cluster around the Gauls in a way that they do not cluster around the Germans. This does not make the Germans exceptional, however. There are a number of examples of peoples who do not attract a chronological narrative, rather than a primarily ethnographic treatment within Greek literary traditions. The Etruscans are a prime example, perhaps being conceptualized in some way as timeless and thus without a past to be measured and chronicled, despite the attribution to them of saecula, a means of conceptualizing periods of time. According to ancient literary traditions, the Etruscans were as much an «eastern» people as they were a «western» people, and they do not really fit into the category of the pre-civilized in Greek and Roman thought: if anything, they are hyper-civilized, so on two counts they do not fit Woolf’s model for the selective «effacement» of barbarian pasts25.

14The relationship between Greek and Latin literary attribution of origins and memories and origins and memories actually recounted by local peoples raises tantalizing, and sometimes ultimately unanswerable, questions. At the very least, the subtlety of patterns attributed challenges more blanket modern characterizations of colonializing sentiments, or indeed of «Roman views» that can be correlated with imperial practices. But «lieux de mémoire» are of course regularly nonliterary, whether they are monuments («places» in a concrete sense) or rituals, ceremonies or cultural practices. It has been conventional to treat these phenomena in the northern or «western» provinces of the Roman empire quite differently from the Greek east, to seek sites of «Celtic» or «Druidic» cultural resistance, in part perhaps a legacy of the complex nationalisms of northern Europe, or else, as Woolf has done, following John Drinkwater and others, to posit «forgetfulness» of local distinctiveness. Thus, Woolf makes the point that we do not see the enactment of local pasts in the contexts that are familiar in the Greek East, «coinage, festivals and monuments, local histories or vernacular literatures»26.

15However, there are some clear indications of «memory», of a pre-Roman past, even if they do not necessarily conform to this select list of «memory sites». These include local terminology, such as the office of vergobret, attested in Caesar’s Gallic War, and in local coinage, Latin epigraphy and a Latin graffito, and increasing evidence for diversity in name-choices, even amongst Roman citizens, including more or less Latinized versions of pre-Roman names, gentilicial as well as cognomina27. In each case, past and present, local and Roman, are interwoven, as they are in the enigmatic example of the Coligny calendar, a juxtaposition of hoary and perhaps incomprehensible Gallic terminology and Roman form, long considered to be either a case of «Druidic» resistance, or as too Roman to count as a useful example of authentic Gallic mentalities28.

16Built monuments of the pre-Roman era also remained, and in a few cases seem to have been retained actively, such as the cult of Dea Bibracte and the maintenance of Bibracte itself as a site of religious activity «remembered» in the new city of Augustodunum29. There is also the tantalizing case of the monumental triple-arch at Reims, the so-called Porta Martis. There are heavy caveats: the decorative sculpture of the arch is unfortunately very damaged and survives in a very partial state, encouraging reliance on 17^th century engravings and drawings for its reconstruction. Fundamental questions about dating and agency are very difficult to answer. Some of the figures that have been identified with more confidence include a sculptural relief of the she-wolf and twins on the ceiling of the eastern arcade, and statues of Venus and Aeneas, Rhea Silvia and Mars in the niches of the arch. At a very obvious level, this seems a thoroughly «Roman» arch, but Ton Derks has argued that there is a local specificity to this engagement with the foundation story of Rome. The demonstrable link is the identification of Mars with the local god Camulus, hyphenated as Mars Camulus in the manner that is so familiar in the western provinces. The more conjectural link is the hypothesis that the Remi at some point early in the Roman imperial period identified Remus as their eponymous ancestor, thus literally twinning themselves with Rome. Unfortunately, this attractive conjecture must be left hanging, because the descent from Remus is attested only from the 10^th century ecclesiastical history of Flodoard: we simply cannot know whether or not Flodoard based this story on earlier traditions30.

17It is unclear to me that these modes of evoking memories are qualitatively different substantively from modes in the Greek-speaking communities of the east. At the very least, I would urge greater subtlety in deciding what counts as «history» and what does not. A recent evaluation of historical writing in south India has offered a new perspective on the old question of whether or not the peoples of the Indian subcontinent had a concept of history before British rule, reminding us not to be so rigid and narrow in our definition of what constitutes historical consciousness31. This insistence on plurality and subtlety is salutary for the study of cultures of the Hellenistic world as well as of the Roman empire. We might think in terms of subtly different modes of evoking and monumentalizing the past, as well as subtly different modes of characterizing pasts of other peoples. In the case of written self-accounts, subtle differences in the authority as well as in the texture of historical content might explain why certain «local» histories never seem to make their way into the centre, perhaps because the content of their histories does not quite meld with the stuff of «great» histories.

18I have begun to suggest that there are cases in which a clearcut distinction between «local» and «Roman» pasts is hard to maintain, and in general it is perhaps more helpful to think in terms of a sliding scale rather than of an antithesis. Some eastern contexts illustrate very well this difficulty of drawing a distinction, as well as insisting on a long historical perspective in which Rome is palpably a comparative newcomer. Connections made via mythological progenitors continue in the Roman world to project relationships that honor both parties simultaneously, important in negotiating the power relations of the present day32. In the famous example of the eleven cities of Asia competing in 26 CE for the honor of a temple to the imperial cult, Tacitus gives the longest and strongest arguments of the two shortlisted cities, Sardis and Smyrna, both of which cite origins and distinctive historical relations with Rome. Smyrna ultimately «wins» on grounds of services rendered to Rome at difficult historical times, but Sardis cites kinship via the brothers Tyrrhenus and Lydus, and thus a role in the making of Italy, a pre-Roman Italy indeed, and a role in the colonization of the Peloponnese, a role in founding «Greece» as Tacitus puts it. This is a fascinating transformation of the «same» myth from Herodotus’ version, where Lydian colonization of «Tyrrhenia» is embedded in Lydian claims to have invented games, presumably a means of forging an essential connection between two supposedly luxurious cultures. The version given by Tacitus reflects and projects a very different world: the Sardians make claims simultaneously on Italy before the Romans, engaging with and contributing to increasingly complex and enmeshed traditions of a Roman-Italian past, and Greece before the Greeks, asserting their prime role in the prehistory of both political and cultural centres33.

19Moving from «mythological» to «historical» memory, at the material level, Christopher Jones has explored the intriguing phenomenon of monuments of imperial date to the Roman Republican past in individual communities of the Greek East. This phenomenon is of especially interest because it goes some way to close the gap between more remote past of origin myths and the somewhat presentist accommodation of monarchy, engagement with an imagined Roman centre within, for example, local structures of the imperial cult. More broadly, they raise important questions about the construction, ownership and memory of the Roman Republican past, a past that has more often been imagined in the geographically narrower terms of a metropolitan, Roman concern with a pre-monarchical past34. The monuments to the Roman Republican past in cities of the Greek East attest and assert the specificity of local relationships with Roman individuals, such as Lucullus, Sulla, Pompey, Caesar and Antony, especially through reciprocal relationships of services and honours, always particular and at some points markedly at variance with metropolitan patterning of the past. At the very least, this phenomenon frequently suggests the dynamic engagement of the present, imperial age with these local Roman Republican histories, links to be reasserted in seeking or maintaining privileges granted by Rome, such as the right of asylum. Simultaneously, it suggests the dynamics of more local competitions, combining the local and the Roman focus as patently as does the so-called «archive wall» of Aphrodisias35.

20As one might expect, there is less evidence for this phenomenon in the northern or western provinces, given the generally later Roman engagement with these areas. It is conventional to emphasize the more obvious strategies to «buy into» the Roman centre, such as claims of Trojan descent for the Aedui, perhaps beginning with the earliest diplomatic relationship with Rome in the 2^nd century BCE36. In the general context of the interplay between larger and smaller histories in the Greek, Hellenistic and Roman worlds, I begin to wonder how far we should see such strategies as somehow particularly characteristic of northern and western peoples, a sign of a peculiar degree of marginalization. The kinds of narratives and memorials that we have traditionally recognized and labeled historical are generated at charged interfaces between societies. The «Gauls» enter conversations of this kind comparatively late, but nevertheless their presence of course predates second-century diplomatic relations with Rome. Parallels drawn between the Gallic assault on Delphi in 279 BCE and that on Rome in 390 BCE encouraged a considerable amount of dialogue between authors of a number of provenances in the Hellenistic world, and historical narrative connecting these two events. One example is the argument recorded by Strabo between Posidonius and Timagenes about whether or not the gold that Q. Servilius Caepio took from the sacred lake of Toulouse in 106 BCE came from Delphi37. This case nicely complicates questions of whose history is whose, and contextualizes a Roman history of direct engagement with Gaul within the longer, complicated dynamics of a world in which Rome had only recently become the major player.

21Some accounts go further, are more sympathetic and more intertwined. Pompeius Trogus, as represented in Justin’s epitome of the Historiae Philippicae, and Livy, reappraise Gallic migration over the Alps into Italy to a considerable degree. They engage with «international» narratives of migration, colonization and urbanization as well as with the notion of Hercules as a paradigm for crossing the Alps, and make a real attempt to historicize circumstances within Transalpine homelands38. More strictly local, rather than «great», Gallic Roman Republican histories revolve notably around the memory of Caesar. Notoriously, Julius Sabinus, the Treviran co-conspirator with Iulius Civilis, claims to be descended from Caesar himself via an affair with his great grandmother, a neat historicization of the integrating mythological motif of marriage between a wandering hero and the daughter of a local king. As it happens, Diodorus explains the origins of the Gauls in precisely these terms: the original «Galates» was the offspring of Hercules, founder of Alesia, and the daughter of the king of Celtica39.

22Finally, I will consider briefly the ways in which we might complicate a more conventional notion of «Roman history», the kinds of «great» narrative on which we have tended to base our modern accounts of Roman Republican history. What is particularly interesting is the degree to which self-consciously provincial authors perceived themselves to have a stake in «Roman history», especially in a Republican past before the kind of provincial integration that is reflected in the careers and stances of such authors. Is Roman history «our» history, and if so, in what sense? The tendency to make Rome a focal point of an increasingly centralized world is apparent as early as the 2^nd century BCE, but more striking in early imperial historiography, where this centralization is capped by the advent of hereditary monarchy at Rome. However, the relationship between the local and Rome in such accounts remains extremely interesting than ever. We might think of Katherine Clarke’s important work on the gaze of Strabo, on the tensions of perspective that are revealed when one focuses on the significance of phrases meaning «in our time». This personalized phrase has spatial implications in that it refers to the «intellectual life of the Greek East» in which Strabo places himself, but it is avoided when the political history of Asia Minor is under discussion, and particularly the context of the Mithridatic Wars, in which his own family members were implicated «on the wrong side». He seems studiedly to maintain some distance by avoiding personalizing time and space in these contexts40. Studies of the «ethnographic» gaze of late Republican and early imperial authors are also very revealing, particularly when that gaze is directed towards Rome. It is salutary to note that much of our more empirical detail of «constitutional» aspects of Roman provincial government comes from authors such as Polybius, Strabo and Dio41. However closely entwined the experience of the present and the perception of pasts may become, the distancing entailed in passages of this kind draws attention to the complexity of the question of whose history this is.

23The subject of Roman imperial expansion invites a broad range of responses from provincial writers of Roman history on the question of how far this is in some sense «our history». Polybius and I Maccabees vividly illustrate the manner in which Roman expansion can be narrated in very different ways, from different perspectives, as the latest, but greatest episode according to a canon of Greek history writing, and of pressing universal interest, or as embedded within a more particular local history42. Somewhere between the two is Velleius Paterculus’ simultaneously global, Roman and Italian history, with its exceptionally conciliatory «digression» that happily juxtaposes Roman colonization with extensions of the citizenship and leaves out Roman imperialism altogether43. Appian’s Roman History seems at first sight to be very different, precisely in its stark insistence on the «Roman-ness» of this history. This oddness has not been particularly noticed, perhaps because it looks so close in texture and interests to one ideal of traditional, modern empiricism, and, also close to one of the dominant modern ways of modeling the relationship between «Roman» and «local» pasts. As is well known, Appian divides his account by theatre of war or, occasionally, by protagonist, and he frequently signals the awkwardness of these divisions44.

24This arrangement is indeed very different from other «Roman» world geographies and histories, notably from Strabo’s world. One might have imagined that Appian’s arrangement would encourage emphasis on local cultures and histories, rather in the tradition of Herodotus’ account of the Persian wars. However, Appian most surprisingly seems to turn away from local particularity, and the fact that he signals that he is doing this implies that it is quite self-conscious. He very rarely engages in any ethnographical observation, and generally refuses explicitly to engage in the «prehistory» of lands and peoples before the arrival of Rome45. Thus, in an early chapter of the Iberike, Appian agrees to touch briefly on the identity of the early inhabitants of Spain, but distances himself as being concerned with ta Rhomaion, Roman history. Similarly, he passes over the identity of the temple of Hercules at the straits of Gibraltar, saying that he will leave such matters to the palaiologoi, the antiquarians46. In the preface, he even talks about when «the end» comes to each ethnos at a point subsequent to first Roman involvement47.

25The only real exception to this is his account of the Seleucid dynasty at the end of his Syriake, with its highly coloured story of the future Antiochus I, lovesick for his stepmother Stratonike48. This exception is perhaps conditioned by his exceptional interest in Rome’s imperial predecessors, something that one might have expected from his programmatic statement at the end of the Preface of his own authority and stake in Alexandria and Egypt, along with the stake and involvement in Roman rule that is implied by his mention of his honorary procuratorship49. In terms of a broader sense of ownership of this «Roman history» of Appian, I am interested in the intriguing question of whether there is relationship between the author’s clearly heightened interest in Hellenistic kingdoms, and the context within which surviving papyrus fragments of his work have been found, two small 3^rd century pieces apparently both from the Iberike, found in the originally Seleucid and always borderline settlement of Dura-Europus on the Euphrates50.

26In contrast with Appian’s staged lack of interest in prehistory before the Romans, the kinds of themes that do seem to excite him include taxation, for this author a very historical subject that reveals much about the relationship between her subjects and Rome. Most notably, Appian’s Sulla declares that the Ephesians’ payment of five years’ tax upfront, in recompense for their support for Mithridates, will preserve the Greek genos, their onoma and their doxe, their race, their name and their reputation: to preserve their identity, they must literally subscribe to the Roman empire51. Taxation systems are markers of peculiarly Roman ways of demarcating and considering territories and peoples, a historical reconfiguration of current identities. Appian’s interest in tax might be thought of at least in part as aetiological: current local situations are a lasting «memory» of the individual narratives of Roman conquest and subjection with which he is concerned in the «ethnic» books. If Appian in some ways comes close to some modern notions of history beginning with imperial colonization or conquest, or traditional modern national histories, his Roman histories defy any sense of a unitary past: they remain localized, stranded, in a very particular way.

27Earlier in this paper, I suggested that the works of Ram Mohan Roy are comparable to some degree with the ways in which I would characterize provincial investment in both «Roman» and «local» history. An intellectual employed by the East India Company, Ram Mohan was actively involved in the project of religious conversion in India through his role in co-writing a fake «ancient religious book» that would demonstrate that the god of Christianity had been known in India for thousands of years, concealed as Brahma. Ram Mohan Roy was not a radical, but at a time when others hoped for direct local political participation, or even for self-government or separation, his more modest hopes, for Indians to serve as jurors in the courts, for example, mesh with an intellectual project to assert the rationalism of a recognizable Indian «constitution» based around the panchayat, the local judicial body, his insistence on India (significantly not «Hindustan») as a national entity, and his stand against sati, incompatible with ancient Indian liberal traditions. In a number of ways, the case of Ram Mohan Roy is rather suggestive. But while his reflections on the past play beautifully around interfaces with European and specifically British political, religious and intellectual traditions, he writes the past of «India» rather than «Britain»52. I can think of no really close parallels to the degree of implication in the Roman empire that is implied by provincial subjects’ engagement in commemorating and writing Roman histories in which they might simultaneously be viewers and participants.