Abstract
Scholars of Indian history, who have come to expect great things of Irfan Habib, will see their expectations largely realised in this magisterial volume. The work is impressive, not only for the depth and breadth of its scholarship, but also as a physical production, measuring 31 cm x 65 cm and weighing almost three kilograms. However, from a cartographic perspective, as will be shown, it is seriously flawed. The authors—including Irfan’s son, Faiz, who was chiefly responsible for the cartography—describe the atlas as follows:
Atlas of Ancient India presents the most detailed mapping yet attempted of India during ancient times, from the arrival of man to c. AD. 750. The text aims at providing substantiation through adequate referencing, of what has been depicted on the maps. It also clarifies many problems of archaeology and history necessary for interpreting the mapped information. The character of the information varies, to a considerable extent, according to the nature of the sources available. While in the first four maps the atlas delineates the areas of archaeological culture, in subsequent maps it becomes possible to show approximate boundaries of dynastic states. Economic information is furnished for all periods; and linguistic changes are also traced to the degree possible. Over 1075 sites and places are plotted on the maps, and precise coordinates are furnished in an index. There is a full bibliography, listing all sources used. The atlas is thus designed to serve as a work of reference for the researcher and as an aid to the general reader interested in the ever-changing human geography of our ancient land. (endnote, p. 135)
Except for the claim regarding ‘the general reader’, the foregoing description is essentially correct. This work is an eminently worthy successor to Habib’s similarly executed work, Atlas of the Mughal Empire, published in 1982.
The research embodied in the work, extending over several decades, is formidable. It draws upon original sources in Sanskrit, Prakrit, Pali, Avestan, Sogdian, ancient Greek, old Iranic, Arabic, Chinese, and other ancient and modern languages; and, on most maps, it indicates the pronunciation in accordance with a necessarily complex array of diacritical marks. On several maps, however, diacriticals are omitted (as they will also be in this review, even in otherwise faithful quotations).
The bibliography cites nearly 500 sources, overwhelmingly in English. More than a 100 frequently cited sources are indicated in the text by abbreviations (for example, EI for Epigraphia Indica). This saves considerably on space, but can constitute an impediment to the work of relatively neophyte scholars. This reviewer could not help but note that while the 1978 edition of his Historical Atlas of South Asia was singled out for notice, the 1992 reprint, containing extensive addenda relating to the prehistoric and ancient periods, was not mentioned at all (Given India’s excessively strict and counter-productive rules on what maps can be imported into the country, the 1992 work was presumably unavailable to the Habibs).
The Habib atlas provides twelve original maps, each extending over two adjacent pages. Ten of these maps relate not only to the whole of South Asia (that is, the contemporary SAARC region), but also an adjacent territorial fringe varying somewhat from map to map. This has led to the use of eight slightly differing map scales, ranging from 1:9.181 million to 1:8.126 million. One map, focusing on the Harappan period, covers the north-western areas only and is plotted at a much larger scale, 1:5.587 million; and the introductory map, relating to the evolution and diffusion of humankind, covers most of the ‘Old World’ at an equatorial scale of 1:30.5 million. Additionally, Oxford University Press (OUP) has provided a richly coloured physical map of contemporary South Asia and adjacent areas. This map, which abounds in toponymic neologisms (for example, Hazaribag Plateau, Rohilkhand Plain), also includes national boundaries, the boundaries of Indian states, and national and Indian state capitals. Strangely, however, the names of Indian states are omitted, as are the boundaries and names of the provinces of Pakistan.
The original maps are limited mainly to information derived from painstaking examination of primary sources: verifiably datable (largely through radio-carbon analysis) archaeological remains, inscriptions, rock edicts, coins, surviving texts, etc. The information portrayed includes settlements, regions, tribes, other peoples, dynastic names, rivers and other physical features, economic products, and numismatic evidence. The dates to which the finds relate, and, where relevant, the language of the source are meticulously noted. The authors’ standards for admissibility of evidence are high. Hence, many features whose provenance is questionable are intentionally omitted, as are many sites in areas with a high density of datable features for which the inclusion of further information would serve no scholarly purpose.
Supplementing each original map, but physically separated from it, is a text chapter containing detailed notes keyed to each of the places, physical features, economic products or other pieces of evidence. Chapters range from three to twenty-five pages in length. Although map and chapter titles are usually alike, several of the latter are more inclusive. They may be cited as follows: (1) Evolution and Diffusion of Mankind; (2) Neolithic India; (3) The Indus Civilization and Antecedent and Contemporary Cultures; (4) Archaeological Cultures, c. 1800–600 BC; (5) Historical Geography of India, 1800–600 BC; (6) India, 600–320 BC; (7) Mauryan India; (8) The Political Geography of India, 200 BC–AD 300, from Inscriptions; (9) Economic Geography, AD 1–300; (10) India, Political, AD 300–550—Based on Inscriptions; (11) Political Geography, c. AD 550–750; (12) Economic Map of India, AD 500–800.
In addition to the text, the authors provide three alphabetical indices, wherein virtually all places mentioned in the atlas are noted: (1) Places; (2) Tribes, Territories [and] Kingdoms; and (3) Ancient River Names. Precise latitude and longitude information is provided for each entry, along with the numbers of the maps on which each name appears. Cross-references are numerous.
For some maps, the text notes are grouped wholly according to sources consulted. The section heads for Chapter 6, for example, are: Early Buddhist Literature, Achaemenid Inscriptions, Herodotus, Arrian, Nearchus, and Archaeologic and Numismatic Evidence. For others the organisation is primarily by subject. Thus, in Chapter 9, after a brief discussion of sources, we find the following headings: Minerals and Pearls, Animals and Animal Products, Forest Produce [and] Crops, Craft Products, Technology, Coinage, Inland Routes and Sea Routes. Each of these sections is further subdivided; under Craft Products, for example, we find sub-sections on Cottons, Woollen Goods, Rouletted Ware, Glass Bead Manufacturing and Gold Work.
Notes relating to specific sites, regions, peoples or physical features vary from as little as a single line of text, to several densely worded, tightly reasoned paragraphs. Throughout the work the author’s sources are fully and scrupulously cited. I here provide a tiny sample from the plethora of perspicacious analyses that characterises the text:
The uppermost limit in time of the Rigveda is more or less fixed in at 2000 BC. Since the horse-drawn chariot, constantly referred to in it, is not archaeologically attested anywhere in Eurasia, including its most likely area of origin, the Russian and Kazakhstan steppes, before 2000 BC…Given the remarkably close linguistic affinities between the Rigveda and the Avestas, even if all the claims of Glottochronology [sic] to precision are ignored, the earliest composition of the Rigveda could hardly have been 500 years older than the Gathas. One would, therefore, be justified in putting the ceiling of 1500 BC for the Rigveda. (p. 20) Marching toward Bactria, Alexander crossed ‘the Indian Caucasus’, a name suggestive of an original Iranian form, Hindu-kakush. Is, then, the ‘Hindu-kush’, despite its popular and probably false etymology (Hndu+kush, Hindu-killer) so ancient a name for the famous mountain range? (p. 35) ‘Bhojas’ [a people], R.E. [rock edict] XIII. Literary texts suggest location in Vidarbha (Barua, Inscriptions of Asoka, II, pp. 261–65). The Hathigumpha inscription of Kharavela pairs Rathikas and Bhojakas, and thus suggests that they were in proximity to each other, and within the range of possible military operations from Kalinga. The position of the town of Bhojakata also needs to be considered in any discussion of Asoka’s Bhojas. (p. 43)
The atlas text does not shy from thoughtful (though not necessarily correct) criticism of previous work, including, in several places, that of this reviewer. Consider, for example, the following:
A notable feature of the Schwartzberg Historical Atlas of South Asia in respect to the coastline depicted on maps of ancient India, pp. 12–32, is that coastlines run impossibly further inland along the mouths of the Indus and of the Ganges–Brahmaputra than would seem to be justified by the net silt deposition of these rivers in modern times. That the Rann of Kachchh was part of the sea until the 12th century (see map on p. 32) seems hardly plausible. (p. xvi, fn. 6)
To indicate in a single review all the historiographic issues on which the Habib atlas sheds important light would be an impossible task. Let me here mention just a few: When did modern hominids (homo sapiens) first arrive in India? When did agriculture in India begin? Was the Indus civilisation the result of cultural infusions from the northwest or was it essentially an indigenous development? Where in South Asia did iron-working technology originate and how was it diffused? How does the controversial chronology of the Kushana era relate to that of the Saka era and others of ancient India? To what extent can we accept Kautiiya’s Arthasastra as indicative of the actual political and economic organisation of the Mauryan Empire?
Additionally, wherever political boundaries (or, more appropriately, fuzzy frontiers) are shown, how adequate is the epigraphic and other evidence for their position and to what degree and for how long did the writ of the dynasty in question actually apply. In this regard there are many significant differences between the Habib and Schwartzberg atlases, not to mention interpretations in other published works. The Habibs generally adopt a more cautious approach than other authors and elect not to show boundaries/frontiers at all in many cases where the evidence is dubious, even for states as important as, say, the Vakatakas, Pallavas and Calukyas. The Historical Atlas of South Asia, on the other hand, provides what might be described as educated approximations (with reasons for the same in the text) of the maximum extent of at least a dozen significant polities that the Habibs refer to only in textual notes. Readers would do well to weigh the two works against one another and form their own opinions.
Regrettably, the cartography in the atlas is not reader-friendly. It reflects numerous poor design decisions, chief of which were the failure to make good use of available space and the inept use of colour. Given India’s shape, somewhat over 40 per cent of the earth’s surface portrayed on most of the maps lies in the Arabian Sea, the Bay of Bengal and areas north of the Himalayas for which the atlas provides little or no useful information. Yet, on many maps the land areas of India are frequently so cluttered with information that it becomes virtually impossible to see the forest for the trees. To fit in all the information deemed to be necessary, the authors decided, in literally hundreds (if not thousands) of cases, to put that information in legend boxes connected to the place to which it related by lead arrows, several of which had a map length in excess of 300 kilometres. The information burden borne by the map could have been substantially lessened by including inset maps for selected types of information in the relatively empty spaces noted earlier, together with additional empty space in the excessively wide margins at the top and bottom of each map.
Additionally, the legibility of several maps was greatly compromised because of decisions to combine incompatible types of information on a single image. Map 5, for example, shows ancient cultural zones and ancient tribal and territorial names, relating to a period from 1800 to 600 BC, along with contemporary boundaries of linguistic families and modern isohyets (lines of equal annual rainfall). Had the authors decided to include maps of the unquestionably useful linguistic and rainfall data as introductory reference maps, complementing the aforementioned physical map, the reader would have been much better served. Apart from this, it would have helped to divide the referenced period (some twelve centuries) into two, and to increase the number of maps accordingly. The decision to use only twelve original maps for the broad time-span and breadth of detail covered by the atlas strikes this reviewer as excessively parsimonious.
Apart from increasing the number of maps and, in some cases, decreasing the time-span to which a given map relates, the problem of clutter could have been mitigated by narrowing the range of data (for example, physical features, or economic particulars, or ethnic groups) for each map. Maps could thereby have been rendered clearly at significantly smaller scales than in the present work. Greater use of insets (as was done, in fact, by the authors on Map 8) to show information at an appropriate scale in areas with an especially high density of data would also have been in order.
Another way to make good use of under-utilised space would have been to complement the rather lifeless maps with related graphic media: photos or drawings of such artefacts as stone tools or pottery for the prehistoric periods, plans of archaeological remains of places like Harappa or Taxila, chronograms (relating, say, to climate change), time bars showing dynastic successions, and so forth.
Design choices in respect to colour were particularly uninspired. About the only use of colour extending over a broad area was to show the sea in solid blue. On land, the use of colour is confined mainly to typography, say to differentiate sites of inscriptions or archaeological finds according to their language or cultural provenance. The result is a weak image, from which the detection of meaningful patterns is all but impossible. Using solid colour (in pastel hues) to indicate cultural or political regions (wherever data are adequate) would have been helpful.
Exceedingly small type size is also a concern, an inevitable consequence of the general problem of clutter. On most maps I would estimate that a substantial proportion of the map text was rendered in difficult-to-read 4 to 7 point type. When names are rendered in black the problem is somewhat mitigated for readers with good eyesight; but the frequent use of small-size type in blue, pink, green, or tan can present a serious problem. A partial solution would have been to make better use of symbols and colour. For example, one could consistently indicate crafts by squares, crops and other vegetation by triangles pointing upward, animals and animal products by triangles pointing downward, numismatic evidence by ovals, and so forth, colouring each symbol according to the relevant time period, language or culture. Symbols might then be supplemented by abbreviations to provide greater specificity. To take the case of animals, C might stand for cattle, Cm for camels, E for elephants, G for goats, etc. Making each symbol at least several square millimeters in size to allow the colour to be read would have made discernment of geographic patterns much simpler.
Shifting much of the indigestible map text to the chapters keyed to each map would also be a boon to the reader. Doing so might have been facilitated by resort to an alphanumeric grid system (a common atlas convention) wherein each letter might indicate a four-degree longitudinal band and each number a four-degree latitudinal band. Let us suppose that the reference for Nagarjunikonda were to be E6 (on all maps on which that site appears). ‘E6’would be added to Index 1, supplementing the more precise indication there of latitude and longitude. Atlas text references to Nagarjunikonda would be followed by the notation ‘[E6]’. The additional space requirement would be negligible. For readers uncertain about the location of Nagarjunikonda, this expedient would preclude the necessity of going from the text to Index 1 and then from Index 1 to the map(s) under consideration. While in the present work it is usually relatively easy for the reader to go from map to text, the reverse process is not simple.
To conclude, the Habibs merit much praise for their truly stupendous research accomplishment. It is, however, a pity that the utility of their work is seriously diminished by the inadequacy of its graphic portrayal.