Scholia Reviews ns 8 (1999) 5.

D.P.M. Weerakkody, Taprobane: Ancient Sri Lanka as known to the Greeks and Romans. Indicopleustoi: Archaeologies of the Indian Ocean.. Turnhout: Brepols, 1997. Pp. xxii + 287, incl. 3 black-and- white plates. ISBN 2-503-50552-X. BEF2,500.00.

Grant Parker
Department of Classics, Princeton University

'Taprobane . . . was for a long time thought to be another world. The age and achievements of Alexander the Great proved clearly that it was an island. Onesicritus, commander of the fleet, wrote that bigger and more warlike elephants are produced there than in India.'[[1]]

Islands are good for story-telling, especially when it comes to stories about marvellous creatures and other natural phenomena -- as the elder Pliny well knew, and as Emilio Gabba showed in an important article nearly two decades ago.[[2]] To judge from this attractively-produced new volume by D.P.M. Weerakkody, Taprobane, which we may safely equate with the modern nation-state of Sri Lanka, was no exception. The author has 'attempted to bring together the references to Taprobane in Greek and Latin texts for the purpose of examining their value as sources for the study of ancient Sri Lanka' (p. vii). In the opinion of this reviewer (who, it must be said, comes to the book with a background in Classics rather than South Asian Studies), that effort is largely successful.

The Greek and Latin texts themselves, beginning with two references to Onesicritus, are given at the end of the book (pp. 197-222), with translations following (pp. 223-249). These texts fall easily into two main sections, the first giving geographical data (notably Strabo, Pomponius Mela, the elder Pliny, Ptolemy, Solinus, Martianus Capella), and the second detailing its flora and fauna (including Aelian, Dionysius the Periegete, and Cosmas Indicopleustes of flat-earth fame).[[3]] Only a rudimentary apparatus criticus is given, which is a little unfortunate in view of the textual problems plaguing Pliny's text, for one. (This means in practice that one still has to turn to the critical editions in order to quote any of the texts involved.) On the credit side, though, it is admirable that the treatise 'On the Life of the Brahmans' is given in Greek and in both of the Latin versions to survive. In general among the texts cited, greater effort might have been taken to indicate which edition is being used. The bulk of the book is devoted to a discussion of those texts -- as if, in linguistic jargon, the literary record may be taken as 'marked' and the broader historical reconstruction 'unmarked'. The strength of these chapters is their constant reference to South Asian archaeology and history, which is made possible by the author's command of sources in Tamil and Sanskrit.

The kinds of source-material in which Sri Lanka occurs are varied, and these are mostly treated in seventeen separate chapters, ranging in length from 5 (chapter 7, 'Pliny's Influence on Later Latin Notices of Sri Lanka', pp. 79-84) to 19 pages (chapter 15, 'Some Reflections on Roman Coins from Sri Lanka, pp. 151-70). The technical problems involved in the archaeological evidence are enormous, perhaps nowhere more so with the hoards of Roman coins found. In Greek and Latin verse the pickings are lean (chapter 11, 'Sri Lanka in Greek and Latin Verse', pp. 113- 17), the most substantial of these being the (probably Hadrianic) Greek hexameters of Dionysius Periegetes and the late Latin of Avienus and Priscian. The natural historical interest in Sri Lanka as the edges of the earth is seen especially in the elder Pliny (pp. 65-77) and in Aelian (pp. 105- 112). Given the amount of ground covered here, it has clearly been necessary to avoid detailed discussion of specific problems, but within this trade-off Weerakkody has been admirably succinct, and has done a good job of using footnotes to refer to debates and controversies.

At first blush Strabo might seem the most promising topographical source. Yet, in reality Strabo is a classic case of the failure to capitalise on newly- available information -- a failure which was well diagnosed by Dihle in a landmark article some 35 years ago.[[4]] Indeed, in the end 'Strabo's all too brief notices of Taprobane are a sad reflection of what could have been achieved, had their author thought better of his contemporary informants' (p. 50). Strabo himself, with remarkable candour, explains that the low social cachet of merchants reduced the supposed value of such topographical information as they might bring:

'As for the merchants who now sail from Egypt by the Nile and the Arabian Gulf as far as India, only a small number have sailed as far as the Ganges; and even these are merely private citizens and of no use as regards the history of the places they have seen.' (15.1.4 C686, tr. H.L. Jones)
Given the stigma attached to mercantile activity in the Graeco-Roman world we must regard the Periplus Maris Erythraei as a rare and valuable document, not only of trade activity but also of the kinds of geographical information circulated. Most will now accept a dating that places it between AD 40 and 70. It is remarkable that a text that otherwise shows accurate information about the south Indian coast should say that the island extends on an east-west axis, and that it stretches nearly up to part of Africa ('Azania', PME 61). This is where you can get pearls, transparent gems, cotton garments and tortoise shell, says its anonymous author in a spirit of practical helpfulness that conceals his lack of information about it (ibid.). Sadly, the single important manuscript is corrupt at this point; it does seem to claim that the island was currently known as Palaisimundu (its Tamil name), formerly as Taprobane, in which regard it directly contradicts the more thorough-going Ptolemy (7.4.1) about six to eight decades later.[[5]] It is striking that the PME so overestimates the size of Sri Lanka, but then this was a mistake shared by Greek and Roman writers on the subject, and for that matter by South Asian and Arab writers too. Even Ptolemy was guilty of this, just as he was guilty of imagining an India which practically ignored the existence of its enormous subcontinent. Given this persistent tendency of Greeks and Romans to overestimate the size of Sri Lanka, it would have been good to read more discussion from Weerakkody on that phenomenon in itself, perhaps in the chapter on Ptolemy. Pliny looms large in the work, perhaps more so than two chapters on him would suggest. His influence would have been even more intensely felt had Weerakkody taken the story further into the western Middle Ages.[[6]] Pliny himself made extensive use of Hellenistic sources such as Onesicritus, Megasthenes and Eratosthenes, and is today important for attempted reconstructions of now lost works.

If one is to analyse Graeco-Roman ideas about Sri Lanka, India cannot be left out of the picture too long, if one looks at the sources cited and discussed. This fact may sit awkwardly with contemporary international relations and their attendant academic exchanges. Some Greek geographers even thought Sri Lanka was part of India, according to Mela (p. 96). As regards India, the work of Dihle (to cite another scholar whose research has produced remarkable results from the breadth of evidence used) has pointed to a substantial discrepancy between archaological finds and the literary record. We know that Buddhism originated with the life of Siddharta Gautama in the late 6th century or perhaps slightly later; and we know that there was substantial trade contact between South Asia and Egypt (and ultimately the Roman world) in the time of Strabo and Pliny. Yet it is not until Clement of Alexandria (Stromateis 1.305) in the late second or early third century that Buddhism is mentioned in a Graeco- Roman text. The lack of Graeco-Roman references to Buddhism is for Dihle a sure sign of the imperviousness of literary sources to new information. It is pleasing to see that Dihle's innovative work on India, which goes back to the 1960s but has only recently attracted attention amid a resurgence of interest in foreign peoples, is here given its due measure of credit.

If we are to put Weerakkody's labours in perspective, we might compare two other collections. J. André and J. Filliozat twelve years ago published a substantial collection of Roman sources on India.[[7]] True to form for the Budé series in which it appeared, this contained facing translations (into French), and 26 dense pages of notes at the end. That collection was intended as a sequel to the same editors' Budé of Pliny[[8]] -- a fact which perhaps accounts for the decision to focus on Latin sources to the exclusion of Greek. The strength of this team was that it brought to together the skills of a classical philologist with those of an Indologist. Given the substantial overlap between Latin and Greek sources on India, as shown by the notes there, this division seems ultimately somewhat artificial. While the editors' decision is certainly understandable in practical terms, it does mean, ironically, that modern scholars are much better served for Latin sources than Greek, even though the Greek sources (not least the Alexander historians) go much further back in time. In the end this is a doxographical problem: the Greek sources are generally much older than the contexts in which we know them (i.e. in both Greek and Latin texts). Thus, for example, our knowledge of the substance of Ctesias' Indika is known to us through Aelian's Historia Animalium and Stephanus of Byzantium, and is obviously subject to their interests in the processes of selection.[[9]] A major requirement facing any scholar of this material, then, is to take due account of the context of their often fragmentary sources. Weerakkody does a good job of this in the 17 thematic chapters which constitute the bulk of the book.

Another recent book to merit comparison is Stanley Burstein's collection, which examines the ancient African civilizations of Kush and Axum, using (mostly but not exclusively Greek and Roman) literary and epigraphic sources, ranging chronologically from Strabo to an 8th-century Arabic document.[[10]] Agatharchides of Cnidos' On the Erythraean Sea, which dates to the late 1st century BC, features large here. Arguably, Burstein does a better job of placing the texts themselves at centre-stage, but then there are fewer relevant primary texts, much fewer than those discussed by André and Filliozat.

If we were to go further back, we would have to take account of a number of texts edited by McCrindle towards the end of the last century.[[11]] It appears McCrindle never did apply his considerable energies to Taprobane specifically, hence the novelty of Weerakkody's effort. All of these were restricted to English translation of the Greek and Latin originals, with some notes. This was put together before Jacoby's pioneering work in the Fragmente der griechischen Historiker did much to define modern notions of the texts and authors involved. McCrindle's translations have been reprinted with merciless regularity, sometimes with a new foreword, but they cannot now be regarded as reliable. For example, his volume on Megasthenes (and Arrian) relies heavily on the now much outdated collection by Schwanbeck.[[12]] All this underlines the magnitude of Weerakkody's achievement in covering what is admittedly a smaller body of material. To conclude, a few comments of a technical nature. The number of typographical errors is higher than one might expect in a book of this nature.[[13]] In terms of style, there are many points where copy-editing would have produced a somewhat tauter prose style. The bibliography especially could have done with more careful editing. Physically this is a handsome book in its large format and glossy paper. It is a pity that no more than three plates are provided at the end, especially since a fair amount of space is wasted between chapters. It is also misleading to describe the map given as Plate 1 in the appendix as 'Ptolemy's Map of Taprobane'. Given the uncertainty over whether Ptolemy himself produced maps or merely provided readers with the information needed for them to produce their own, it would be more accurate to cite the early modern edition from which this illustration is taken.

It is one thing to cavil in this way about small points of realisation of this project; it is another to praise the industry and breadth of vision that brought the book into fruition. The very breadth of material covered, which makes the book hard to review, is its strength. The types of evidence adduced here -- literary, historical, archaeological -- are notoriously hard to compare, and there can be no doubt, on balance, that Weerakkody has done a fine job. Whatever smaller points of criticism may be levelled, it is indeed a great achievement to have taken such a wealth of disparate sources into account. Anyone interested in either literary representations of an eastern land, or in the historical realities of contact will find that the book repays close study.


[[1]] Pliny HN 6.81 = Onesicritus, FGH 134 F13, cf. Weerakkody p. 223.

[[2]] Emilio Gabba, 'True history and false history in classical antiquity,' JRS 71 (1981) 50-62. The question of marvel-mongering here raised has been taken further in the bipartite article by Guido Schepens and Kris Delcroix, 'Ancient paradoxography: origins, evolution, production and consumption,' in La letteratura di consumo nel mondo greco- latino (Cassino 1996) 343-460. See also William Hansen's two interesting new editions: Phlegon of Tralles' Book of Marvels (Exeter 1996) and Anthology of Ancient Greek popular literature (Bloomington 1998) esp. 249-71; and also the new anthology by Joseph Nigg, The book of fabulous beasts: a treasury from ancient times to the present (Oxford 1999), which begins with a generous helping of Homer, Ctesias and others.

[[3]] It is not clear why a third section is required for 'miscellaneous notices of Taprobane' (pp. 218-22 and 246-49): here we find extracts from Plutarch's weird essay 'On the Face of the Moon', the Historia Augusta, Ampelius, Philostorgios, Orosius, Jordanes, Julius Honorius and the shadowy geographer 'Aethicus'. A miscellany indeed, but it seems that all the material there cited could have been grouped under the foregoing geographical or natural historical sections. Any look at Aelian, Strabo or Pliny shows how varied and overlapping were the ancient categories of topography and (even more so) natural history.

[[4]] Albrecht Dihle, 'The Conception of India in Hellenistic and Roman Literature,' PCPhS ns 10 (1964) 15-23. This was reprinted in his Antike und Orient: Gesammelte Aufsaetze (Heidelberg 1986), which is not mentioned in the bibliography.

[[5]] Lionel Casson (ed.), The Periplus Maris Erythraei (Princeton 1989) 231; cf. Weerakkody, 17-25, esp. 20.

[[6]] Arno Borst, Das Buch der Naturgeschichte: Plinius und seine Leser im Zeitlalter des Pergaments (Heidelberg 1994).

[[7]] Jacques André and Jean Filliozat, L'Inde vue de Rome (Paris 1986).

[[8]] J. André & J. Filliozat, Pline l'ancien: Histoire naturelle. Livre VI 2e partie (Paris 1980).

[[9]] P. A. Brunt, 'On Historical Fragments and Epitomes,' CQ 30 (1980) 477-494. See now the thought-provoking collection, Glenn W. Most (ed.), Fragmente sammeln: Collecting Fragments (Goettingen 1996). A notable feature of several of the essays (esp. G. Schepens and G. Bowersock) is the criticism levelled at Felix Jacoby's Fragmente der griechischen Historiker.

[[10]] S. Burstein, Ancient African civilizations: Kush and Axum (Princeton 1998).

[[11]] John Watson McCrindle (1825-1913) produced the following collections of translated testimonia, with introduction and commentary, most of them published in Calcutta, where he taught at one time: Ancient India as described by Ktesias the Knidian (Calcutta 1882); Ancient India as described by Ptolemy (Calcutta 1884); Ancient India as described by Megasthenes and Arrian (Calcutta 1887); The Commerce and Navigation of the Erythraean Sea (Calcutta 1879); The Invasion of India by Alexander the Great (1896, perhaps earlier); The Christian Topography of Cosmas, an Egyptian Monk (London 1897). Finally he published a volume of texts that had fallen between the cracks: Ancient India as described in Classical Literature (London 1901). Reprints are legion, from India, the UK and the US, and reflect varying degrees of editorial addition. Many of these collections first saw light in The Indian Antiquary: a Journal of Oriental Research, beginning in the 1870s.

[[12]] E.A. Schwanbeck, Megasthenis Indica (Bonn 1846).

[[13]] The following list, which is not exhaustive, is given here in the hope that it might prove useful in the event of a revised reprint. Where Greek names begin with a vowel the breathing is consistently misplaced, approximating more closely the preceding word than the initial vowel. 'McCrindle' is alphabetically misordered, and 'MacDowell' should be written as one word not two (pp. 153, 260 etc.). The bibliographic entry under 'Swell' should fall away (cf. 'Sewell', which is given at the correct place). At p. 159 l.6 no comma is needed after '1995'. An umlaut is omitted from 'Beitraege' (p. 253) and 'Delbrueck' (p. 256). There is a problem of some kind with the following: 'Liebeschuetz' (p. vii), 'Begley' (p. 7, twice), 'Fraser' (pp. 37 n.48 and 257), 'Mattingly' (pp. 166, 260, etc.), 'Dictionary' (p. 188 n.24), 'scholia' (as plural, p. 211), 'cosmography' (p. 248, twice), 'Dihle' and 'Dudley' (p. 256).