Windows on ancient claims

Paul Cartledge on a work to inspire students of classical cartography.

"Biography/ Is different from Geography./ Geography is about maps,/ While Biography deals with chaps". So ran the original and perhaps less memorable version of Edmund Clerihew Bentley's well-known clerihew. His comparativist approach is to be applauded. But is geography only, or even essentially, about maps? We live, after all in an age of cognitive geography and hodological space, as Peregrine Horden and Nicholas Purcell notified insular classicists and medievalists last year in their stimulating The Corrupting Sea: A Study of Mediterranean History . In any case, what is a map? If we are to believe the website of David Woodward, co-editor with Brian Harley of the multi-volume The History of Cartography , he and his collaborators have developed "the broadest and most inclusive definition of the 'map' ever adopted in the history of cartography". Such grand claims would seem to go with the territory. The ambitions of cartographers, and it appears historians of cartography, can be not only literally global but also stratospheric, not to say immodest. What really is new?

Technology, especially satellite and digital, for one thing. This clearly is part of the explanation for the appearance (in more senses than one) of the Barrington Atlas , and geeks of all stripes will find plenty to satisfy their cravings in the "MapQuest production data" section printed in the main atlas. To cater for the rather more technologically challenged, the atlas also has its own dedicated website ( www.unc.edu/depts/clu/latlas </a> </a> ), including a frequently asked questions list. From this and from a number of other sources, we learn that the Barrington Atlas was begun in 1988 on the initiative of the American Philological Association.

It has involved a formidably large and international team of academic contributors over the years. Besides the editor, there were ten "vicars", no doubt reverend gentlemen all, but in this case regional coordinators. More than 70 "compilers" from most of the recognised continents of the globe worked for them. There were also some 90 "reviewers", who allied their combined skills and expertise with those of a more compact but doubtless equally formidable team of hot-wired specialist cartographers working for what is now called MapQuest.com, Inc. of Lancaster, PA. Some others who were none of the above are also thanked for their assistance. Whatever else may be said about this atlas, its achievement as a collaborative project is cause for satisfaction and congratulation.

Apart from its principal individual and institutional sponsors, Talbert singles out for special commemorative commendation Carl Müller (1813-1894) and Barbara Bartz Petchenik. Both duly appear in Jeremy Black's very informative Maps and History: Constructing Images of the Past (1997). But there is rather more to the subtitle of that book than may at first meet the untutored eye. A standard alternative term for the making of maps would be "compiling". "Construction", in current historiographical parlance, can often be code for more or less subjective, even arbitrary invention or fabrication, making up rather than just making. So it is here. The very title of another of Black's recent monographs, Maps and Politics (1998), almost makes the hint explicit by itself. There is a politics as well as an aesthetics and a semiotics of map making, and map using.

When in about 500BC the Greek dictator of Miletus brought along to Sparta perhaps the first recorded ancient Greek map, inscribed on a sheet of bronze, he meant it to be a persuasive not a descriptive tool. It, in fact, failed in its intended purpose, but that is beside the point. Likewise, Claudius Ptolemy's universe, as Liba Taub graphically reminded us in 1993, had natural-philosophical and ethical, not purely geographical and "scientific" foundations. Or as W. H. Auden so sagely put it in his "New Year Letter of 1940", "maps & languages & names/ Have meaning & their proper claims".

So what are the meaning and claims, proper or not, of the present atlas, produced in an avowedly multiculturalist new world to record an ancient and Eurocentric one? Talbert prefaces the main volume with an introduction that deals with the project's origins, objectives, bases, scales, scope, production and compilation. Among the five objectives, two key ones are the comprehensiveness and the comprehensibility of the finished product. Or rather products. For besides the 102 maps followed by a gazetteer in the main volume, there is also a map-by-map directory either in CD-Rom format or in hard copy (two volumes, 1,500 pages).

Is this the ideal atlas of the Graeco-Roman world that we have all been seeking? Well, yes, and no. Funding was admittedly very generous (some £3 million), but not limitless. So Talbert and co, and his backers, decided not to aim at producing a straight update of, say, William Smith's 1874 Atlas of Ancient Geography (on which Müller worked) or of the more recent single-volume historical atlases such as the Grosser Historischer Weltatlas (1953 and subsequent editions) or Pierre Vidal-Naquet's Atlas Historique (1987). Instead, within a single volume they aimed to offer a fully comprehensive series of the very best, that is most accurate and user friendly in purely geographical terms, maps of the regions for which Greek and Roman penetration can be documented, with only some attempt in some cases at building in political and other contingently temporal features.

The maps' coverage extends from the Atlantic to the eastern limits of Alexander the Great's Indian expeditions, and from Nubia to Rome's furthest northwards expansion in Scotland (sorry, Caledonia; Latin is the preferred language of the atlas). Of the devices available to cartographers for emphasising their chosen issue or goal - scale and contouring, perspective, colour and shading, etc - the first two, largeness of scale and clarity of relief, were therefore considered important above all others. The page size is large, 33.2 by 46.2cms or 13 by 18.25ins, and there are key maps printed on the endpapers. The series begins with six overview maps, and continues with 90 regional maps at two standard scales of 1:500,000 and 1:1,000,000, and three detailed city maps at 1:150,000, showing Athens, Rome and Constantinople and their respective environs. As a sort of hors-série bonus, there are three maps of the Roman empire to end with: provinces of the empire at the death of Trajan; dioceses and provinces of the empire according to the early-4th century Verona List; and dioceses and provinces of the empire according to Hierocles (late 5th/early 6th century).

The three last are "historical" maps. "Ideally", however, every map should have been, since as the editor himself concedes, the extensive physical and cultural changes occurring within the chronological span of the atlas "should be conveyed, say, by a succession of maps for the same area, each devoted to a limited period only". Experto credite : he is himself the editor of the still very serviceable, if monochrome Atlas of Classical History (1985) that presumably put him in (the firing) line for editing the Barrington Atlas in the first place. The same point was made at much greater length and with no less force a century and a half ago by William Hughes in his Illuminated Atlas of Scripture Geography (the relevant passage is quoted in Black's Maps and History ). But as Gibbon once said, where error is irretrievable, repentance is useless, and anyhow, this was not so much an error as a calculated decision to go for a different solution to a different set of problems.

Does the atlas succeed then on its own terms? So far as the scholarship that underpins the ultimate finished products is concerned, there can be no question but that it is utterly sound. The directory in particular makes that quite clear, if indeed the names of the various levels and types of contributors had not already provided sufficient reassurance. I doubt that it will ever be superseded as such. We are unlikely to find so many new data as to require an entire rewrite. Conversely, there is nothing quite like it for bringing home to us our invincible ignorance. Suppose, for the sake of example, one wanted to locate Salmous in Carmania, mentioned by Diodorus of Sicily as the site of a theatre where Alexander the Great staged dramatic competitions. The Barrington Atlas soberly includes it among the depressingly long list of "unlocated toponyms" for Map 3.

What of the competition? Staying with Alexander, the key figure bridging East and West and the classical and post-classical Greek worlds, we might for instance compare and contrast the Barrington version of the geographic scene of Alexander's empire with the dedicated map, The Empire of Alexander the Great , compiled by Richard Stoneman and Richard Wallace for the Classical Association in 1991. The latter is admittedly a wall map not a table map, though it is not all that suitable for classroom use. Rather astonishingly, too, it omits lines of latitude and longitude. On the other hand, it does plot a representation of Alexander's route of conquest from 334 to 323BC, and the routes of Craterus in 325 and Nearchus in 325/4BC, using different colours. Against that, however, its scale of 1:4,000,000 compares very unfavourably with the more-than-a-dozen relevant Barrington maps, the smallest of which is at 1:1,000,000.

Besides, the clarity and sheer beauty of the maps in the Barrington Atlas , for which Princeton University Press and the printers in Palladio's Vicenza deserve the highest credit and praise, make the main volume a joy to handle. The fold-out of the entire ancient Mediterranean world, Map 1 "Mare Internum", is to die for. At all events, whatever else this remarkable atlas achieves, it has made a major contribution to re-establishing cartography as one of the basic sub-disciplines within classical studies. It has also incidentally put Great Barrington, Massachusetts, firmly on the map.

Paul Cartledge is professor of Greek history, University of Cambridge.