Bryn Mawr Classical Review

John Pendlebury (1904-1941), archaeologist and hero in the Boy's Own Paper mold, was right when he warned of the passion that maps can generate: "Never if you can help it display a map before a Greek," he advises on the opening page of his Travelling Hints, explaining: "I did so once and the man leapt at me with a short howl, yelled for his friends who promptly appeared from nowhere, and with them tore the map from my hand. Within a minute they had settled down to a five days' quarrel over the places mentioned on it, a business no less keen because one half could not read Latin characters and the other half could not read at all. Within five minutes the map was unrecognizable."¹

To be sure, the present set of maps is far from meriting such uncouth treatment, even if it ultimately fails to fulfill the initial high expectations raised by the volume's handsome appearance. It is, after all, a substantial component of a major encyclopedia project (Der Neue Pauly), in outsize format, its maps lavishly presented in an extensive color palette, and its three co-editors all established figures in the fields of ancient history and cartography. In chronological scope the atlas aims to match DNP, with a span from the third millennium BC in Mesopotamia to the fall of Constantinople in AD 1453. The core comprises 127 pages of maps in color; these are grouped broadly speaking by successive time periods, except for a cluster at the start relating to worldview and exploration (pp. 2-9, the worldviews re-created very traditionally, and all oriented North according to modern convention). Typically, each map page is paired with a facing page of text and bibliography relating to the area or theme mapped; space is also found on these text pages to accommodate 44 other maps of varying size, all of them uncolored. Frontmatter is brief: preface (one page); table of contents (4 pp.); breakdown of the maps by area, culture and period (3 pp.); guidance on use of the atlas, and its transcription of alphabets (2 pp.); abbreviations (5 pp.). Endmatter comprises the continuation of bibliography and tables too extensive to fit within the core of the atlas (19 pp.); a map-by-map listing of the contributor(s) responsible for each, with (where appropriate) details of the previously published map adapted or reproduced here; finally, a gazetteer (33 pp.).

There is much to admire and praise. The mere provision of an atlas is a pathbreaking initiative in the evolution of the Pauly encyclopedia over the past century and more; none accompanies the behemoth Pauly, nor its handy successor Der Kleine Pauly. The amount of effort expended by Wittke and Olshausen (by far the principal contributors) and fifteen fellow scholars is manifestly prodigious. At the same time, thanks no doubt to equal dedication by Szydlak, the style of presentation for the maps (including their elevation tints) is impressively consistent, a boon to users. In fact this presentation matches that of the Tübinger Atlas des Vorderen Orients (= TAVO, 1969-1993), in which Wittke, Olshausen and Szydlak were all involved. The map "Das Palmyrenische Sonderreich (250-272 n. Chr.)" (221) well displays this cartography's strengths. Central placement of Palmyra within the frame highlights the city's strategic location in the zone where the Roman and Persian empires meet. Colors readily distinguish the areas claimed by each empire, with hachuring overlaid to demonstrate the remarkable extent of Palmyra's conquests. Route linework in further colors traces Odenath's offensives against Ctesiphon, Zenobia's forays into Asia Minor and Egypt, and Aurelian's retaliatory attack on Palmyra. Nowhere is the map cluttered; the scale adopted is appropriate to the quantity and range of names chosen for marking. Symbols and lettering fonts, too, are clear and suitably sized. The accompanying text summarizes first the nature and development of the Palmyrene state, and then its rapid rise and fall in the third century. In the paragraph summarizing the sources a further sentence would have been desirable, to draw attention to the fragment of Peter the Patrician which preserves the story that Odenath had sought to side with the Persians, only to be rejected by Sapor (frag. 10 = FHG IV, p. 187). Overall, given the unfamiliarity of Palmyra to most students of the third century, the map and text offer informative assistance.

The pity is that the sound cartographic principles reflected by this map have too often been overriden elswhere. Surprising though it seems (and unexplained), only one map (8-9, "Fernerkundung in der antiken Welt") and the two endpapers (showing physical landscape) occupy a double spread. Otherwise all the maps are limited to a single page at most, and frequently less than that in practice because many Key boxes claim considerable space. Such a frame of no more than approximately 7.75 x 12 ins (20 x 30 cm) must have proven a source of repeated frustration. It is nowhere near adequate for presenting Alexander's campaigns satisfactorily, for example, or Rome's three wars against Mithridates VI, as the maps on 113 and 159 strive to do. Both in fact derive from TAVO sheets B V 1 (1985) and B V 6 (1981) respectively, where the main maps measure 12 x 26 ins (30 x 66 cm) and 17.25 x 22 ins (44 x 56 cm), with insets in addition for local detail. In the atlas, the olive-colored lines on the Alexander map (for various generals' forays) serve to confuse users more than inform them, while the congestion within Asia Minor on the Mithridates map renders the region meaningful only to the most conscientious of users. In other instances, too, the maker of a map has chosen to proceed oblivious to the inadequacy of the scale determined for it, overcrowding certain areas regardless rather than exercising restraint or adding an inset. As a result, users will be baffled, not enlightened, by Sicily on 93 (Karte B) and again on 138, for example; Cyprus on 115; central Greece on 119 and again on 135; the area south of Lake Tiberias on 123; and that of modern Tunisia on 147. The inadequate scale for showing the sources of marble (84) leaves the unfortunate impression that the Gebel Fatireh (= Mons Claudianus) quarries were on the Nile or very close, when in fact they are situated more than seventy punishing miles distant from the river.

The use of standard physical bases for maps is a prudent and practical expedient (also followed by TAVO), but flexibility in its application is sometimes lacking. On "Roms Kriege im Osten I (214-129 v. Chr.)" (151), the mismatch between the density of coverage eastwards to the Hellespont and the map's emptiness thereafter (for the full length of the Black Sea) calls for the frame and scale to be rethought. Similar rethinking is in order for the second of the maps on these wars (153) so as to enlarge the presentation of western Asia Minor there. By contrast, a full page for the Iberian peninsula on "Roms Kriege im Westen (206-101 v. Chr.)" (149) leaves that map too empty, because (as the accompanying text explains) the surviving sources lack sufficient detail for us to follow the course of these campaigns closely.

Several instances where the scale chosen could have permitted a more informative presentation are to be found in the maps of Roman provinces (either single ones, or an adjacent group). These maps are intended to illustrate shifts in the provinces' extent and boundaries from their annexation as early as the third century B.C. through the fifth century A.D. Altogether an astonishing amount of space (both map and text) is devoted to this theme, and the results hardly justify it. The case of Sicily (143) is especially striking. Nowhere else in the atlas is the island presented anywhere near as large, and yet here the map's sole concerns are changes in the boundary of the province (actually just a single, early change in over seven centuries, when Hiero II's kingdom was annexed) and the legal status of the leading cities under Roman rule; the use of color for this map is redundant therefore. So much more data could have been added, and to do that would have been easy because, ironically, the map is in fact a stripped down version of an uncolored forerunner "Die römische Provinz Sicilia (ca. 241 v. Chr. -- 535 n. Chr.)" which illustrates the DNP entry 'Sicilia'. There, rivers are named, villas marked, and roads and aqueducts traced. The villas include Piazza Armerina, which is nowhere marked in the atlas even though DNP includes an entry for it, with a plan too. Meantime a column and a half on Sicily's text page in the atlas (142) is left an unused blank. The bibliography there, moreover, is a miscellany that seems to be lifted uncritically from DNP 'Sicilia'; too many specialist items are retained, including an unpublished dissertation by Goldsberry (1973, not 1982), and a 1955 article about the Panormus-Agrigentum road even though the map omits roads. For the latter in any case it would surely be more appropriate to cite the monograph by G. Uggeri, La Viabilità della Sicilia in Età Romana (Rome, 2004). By the same token, further invaluable citations missing from the atlas bibliography (none of them included in the DNP entry either) are E. Manni, Geografia fisica e politica della Sicilia antica (Rome, 1981); G. Nenci and G. Vallet (eds.), Bibliografia topografica della colonizzazione greca in Italia e nelle isole tirreniche (Pisa and Rome, 1977-ongoing); and the periodic syntheses in Archaeological Reports (the latest by F. De Angelis, 2007).

This unsatisfying presentation of Sicily in the atlas raises the larger issues of the broad goals envisaged for the volume and of its intended relationship with DNP. A claim to stand alone is made, but then immediately qualified (XIII): "Der Atlas ist als eigenständiges Werk konzipiert, kann aber auch gut in Verbindung mit den alphabetischen Bänden genutzt werden." The Vorwort (V), especially in its section "Voraussetzungen", affirms close linkage with DNP's scope, approaches and materials; indeed it is stated that about sixty per cent of the maps derive from the encyclopedia (which is printed without color throughout). In these entirely justifiable circumstances, therefore, the puzzle arises that a closer meshing was not arranged. Instead, time and again -- as in the case of Sicily -- we find bibliography appended to the alphabetic entries needlessly repeated in the atlas, not to mention occasional lists of rulers (216, 218, 264) which Supplemente 1 (2004) already offers in full. Equally, the excessive space found for tables of text references documenting shifts in the extent and boundaries of Roman provinces -- including the full nine pages 265-73 -- could have been put to far better use in an atlas with the provision of more maps. For example, the course of Xenophon's Anabasis is an ideal cartographic theme, and it is duly mapped in the DNP entry 'Xenophon', but the atlas declines the opportunity beyond incorporating an easily overlooked thumbnail sketch within the map "Fernerkundung in der antiken Welt" (8). Another natural choice is a map to illustrate the democratic process in classical Athens; again the atlas offers none, even though the map "Attische Phylen (nach 508/7 v. Chr.)" could be lifted from DNP 2 (1997) 237-38 and greatly enhanced by the addition of color. The inclusion of such a map could also only help to redress the imbalance that the atlas as a whole may be perceived to reflect in favor of Roman civilization against Greek. Altogether the number of city-plans offered by the atlas is disappointingly small (see list, XII), and curiously so when many more could just have been lifted from the relevant alphabetic entries and Nachträge, among them Delos, Hattusa, Olympia, Pergamum, Pompeii, Portus (Ostia), Priene, Syrakusai, Veii; even provision of a full list of such plans in DNP would be a most welcome enhancement of the atlas. Again, reproduction of DNP's upper plan of Troia (12/1 [2002] 859-60) with color added could help immensely to distinguish periods VI, VIIa and VIIb there.

Incorporation of maps already in DNP aside, the atlas presented greater opportunities for extending DNP's range of maps than were taken. True, the map "Die drei grossen Missionsreisen des Paulus" (228) is a useful such extension, though it is a puzzle that the final journey to Rome should be entirely omitted (despite its mention in the accompanying text), and a pity that the map could not be in color, allowing more distinctive differentiation of the three journeys. A clear instance of an opportunity for extension of DNP not taken at all is the lack of a city-plan to supplement the short entry for Thamugadi. Equally, Hadrian's travels remain without a map, even though DNP 'Hadrianus' includes a section "Die Sorge um das Reich -- Die Reisen." On 149, as mentioned above, the scale for illustrating Rome's second-century campaigns in the Iberian peninsula could readily have been reduced, and space found for an inset of Numantia ringed by Scipio's nine siege camps, which DNP's entry for the site specifically mentions but does not map. More generally, the atlas might have opted to overcome DNP's unexplained reluctance to admit battle-plans, but it silently preferred not to, and that will be further cause for disappointment to many (note that the Rezeption volume 15/2 includes the entry 'Schlachtorte'). Ironically, DNP 'Kalkriese' includes a half-page plan which would look even more appealing in color. The atlas, however, despite its German origin not only omits this plan, but also never even marks either Kalkriese or Saltus Teutoburgensis.

The Vorwort avows a commendable commitment to developing non-traditional cartographic themes, but again some striking opportunities are missed. For example, DNP 'Bodenschätze' offers a map "Mineralische Rohstoffe in der Ägäisregion (ca. 4000 -- nach 1100 v. Chr.)," derived from TAVO A II 2 (1990) which stretches on eastwards as far as Afghanistan and the Indus valley. The basis for extending the limited coverage of the DNP map in the atlas, and in color, was to hand therefore, but no such initiative was taken.² No more was the lengthy "Chronologische Tabelle antiker Wracks (2 Jt. v. Chr. -- 7 Jh. n. Chr.)" in DNP 12/2 (2002) 579-90 made the opportunity for a map -- an enhancement potentially all the more valuable because no map accompanies 'Unterwasserarchäologie' (DNP 15/3 [2003] 922-29) either, even though this is a new field to which the Vorwort (1 [1996] V) specifically commits DNP.

It is cause for regret that the makers of the atlas chose not to elaborate upon how they mean to serve their intended audience. According to their Vorwort, they envisage this as spanning an ambitious range, across the entire spectrum from classical scholars to students and non-professional enthusiasts. Everywhere along it, however, there is liable to be disappointment felt for a variety of reasons. Scholars consulting a work of reference reasonably hope to be informed about the nature and value of comparable existing tools, and about how the present one orients itself in relation to them. The Vorwort does draw attention to the Barrington Atlas of the Greek and Roman World (though not its Map-by-Map Directory) as an essential resource for topography, but nothing at all is said here of the Tabula Imperii Romani, for example, or Tabula Imperii Byzantini, let alone of The Helsinki Atlas of the Near East in the Neo-Assyrian Period (2001).³ Even TAVO is taken granted, although it is in fact unfamiliar to most scholars outside Germany; a paragraph or two about its nature and scope would be fully justified. More generally, no reference is made to the DNP entries 'Kartographie' or 'Kartographie (Rezeption)', still less to my analysis "Mapping the classical world: major atlases and map series 1872-1990," JRA 5 (1992) 5-38. Scholars may be irked when the texts in the atlas are no more than loosely related to the maps they are placed to illuminate, as in the case of the text for "Städtegründungen und Bildungsstätten (4. bis 2. Jh. v. Chr.)" (118-19), for instance. It may not please scholars either that the sections headed "Die Quellen" seem unduly prone to give primacy to texts, while considering other materials and sources of knowledge as secondary. Scholars may query the unexplained decision to maintain the coverage of the atlas so far beyond the end of antiquity to the fall of Constantinople. Granted, DNP itself does extend as far, but only for Rezeption. Thus there are entries 'Karolingische Renaissance', for example, and 'Paläologische Renaissance', but none for Karl der Grosse or individual Palaiologoi, or for the Crusades. In practice the coverage of the millennium from AD 500 that the atlas offers remains very limited -- with a map devoted to the first three Crusades (249), for example, but none of Charlemagne's empire. It might have been more useful, and still sufficiently in line with DNP's scope, to settle for a cut-off date well before 1453, and to expand the coverage for earlier periods.

One divergence from TAVO's practice that all users are likely to regret is the (again silent) decision never to state any map's scale in figures (1:2,000,000 aut sim.) as well as by means of a scalebar. It goes almost without saying that such a statement of scale is useful not only when examining any map but also for comparing it to related ones.

In view of the unavoidably high price of the atlas, it may seem optimistic of its makers to hope that it will see much use by students and enthusiasts. Should these learners gain the chance, however, they are likely to have mixed reactions. Maps showing all or part of the Italian peninsula they should find instructive. These often incorporate the physical landscape informatively and tend not to be overloaded with cultural data: the map "Sprachen im alten Italien vor der Ausbreitung des Lateins" (67) is an attractive example. Less useful by contrast are maps where the complete lack of elevation omits a vital element in learners' understanding, as in the pair "Die hellenistischen Königreiche Indo-Baktriens im 2. und 1. Jh. v. Chr." (133), which stretch from the Indus delta high up across the Hindukush and then north down again to the Aral Sea. Learners will certainly appreciate the four maps which chart the growth of the Kingdom of Pergamum from 241 to c. 185 B.C. (125). All but the most methodical, however, will be deterred by maps which seek to distill an indigestible succession of developments. As many as six Syrian wars (275/74-168 B.C.) form the subject of a single map (123, derived from TAVO B V 16.1), for example. "Die territoriale Entwicklung des Imperium Romanum in republikanischer Zeit" (141) distinguishes nineteen stages by date, and "Die Entwicklung der römischen Provinzen in Kleinasien (2. Jh. v. Chr. bis 5. Jh. n. Chr.)" (183) tops that with twenty-one. Today, such sequences are obvious themes for presentation electronically with overlays that can each be added to a map or removed at will. Almost all students are sufficiently conversant with the use of computers to be aware of this capacity, but they will search this atlas in vain for even a passing reference to digital mapping, let alone for even the briefest discussion by its makers of how they see the new technology influencing and reinforcing their aims. In the era of Powerpoint, GIS and Google Earth, such complete detachment seems unfortunate to say the least. When Rome's Marble Plan is cited (174), no reference is made to the indispensable Stanford Digital Forma Urbis Romae Project. Similarly, in connection with the impressive overviews of roads across the Roman empire (195-99, mostly from DNP 'Strassen' and 'Viae Publicae'), attention is not drawn to the various versions of the Peutinger Map accessible on the web. Much less is there recognition of the interactive maps now available that enable students to trace the course of campaigns and battles in particular (see notably The Mapping History Project).

Ever since 1579, when Abraham Ortelius created a supplement ('Parergon') of classical and biblical maps and attached it to his innovatory Theatrum Orbis Terrarum, the makers of historical atlases have struggled to achieve a satisfying balance of timeframe and themes, map and text. Their efforts are a dominant theme of Walter Goffart's detailed study Historical Atlases: the First Three Hundred Years, 1570-1870 (Chicago, 2003), which confirms -- predictably enough -- that no ideal solution is attainable. Every atlas, like every map, is perforce a set of compromises, the product of multiple choices. In their ambitious endeavor, Wittke, Olshausen and Szydlak make many sound choices and offer much of value. But the sense lingers of opportunities missed, potential not fulfilled, mismatch of material and its intended audience. More effective integration with DNP could have been developed by the historical atlas intended to supplement it; text and bibliography could have been subordinated more to maps. Last but far from least, especially in an atlas which proclaims its commitment to new ways of approaching antiquity, all engagement with information technology and digital mapping should not have been evaded. For a masterly appreciation of how the computer revolution only complicates the bewildering array of choices that the creative mapmaker already has to navigate, one could hardly do better than re-read Denis Cosgrove's (1948-2008) wide-ranging Introduction to his edited volume Mappings (London, 1999). Even so, as Cosgrove himself recognized, the new challenge is one to be welcomed and exploited; a twenty-first century atlas which sets it aside falls short.

1.   From the volume John Pendlebury in Crete, p. 1, printed by Cambridge University Press in 1948 for private circulation. For his exploits, see now Imogen Grundon, The Rash Adventurer: a Life of John Pendlebury (London, 2007).
2.   Note further in this connection A. Orejas (ed.), Atlas historique des zones minières d'Europe (2 vols., Luxembourg, 2001, 2003).
3.   Note the review of the Atlas by S. Richardson, Journal of Near Eastern Studies 65.2 (2006): 125-27.