CHAPTER SEVEN

A Perfect Storm

Rescue and Revisionist Numismatics

Over the past thirty years (1980–2010), the ongoing political and military crisis in Afghanistan has forced many scholars to explore the unfortunate but necessary methodologies of rescue numismatics. Hand in hand with the despoliation of archaeological remains during this period (see chap. 5, above), vast troves of numismatic evidence have been dispersed or destroyed by a perfect storm of poverty and lawlessness in league with supply and demand. Some sense of this problem can be gleaned by reviewing the offerings of Bactrian and Indo-Greek coins in sales catalogues over the past thirty years, where nearly ten thousand newly found specimens have been auctioned worldwide.¹ These coins represent the tip of a much larger tragedy, since many additional examples surely have left no detectable trace in the market literature. For scholars, this brings to mind the haunting vision of ancient Bactria's lifeblood being spilled before its history can be saved.

This crisis affects every category of numismatic evidence. Consider the special tetradrachms issued by Agathocles and Antimachus to honor their royal predecessors, discussed above in chapter 2. In 1984, a comprehensive study of the known examples of these commemorative coins produced a corpus of 37 tetradrachms discovered since 1843.² That number has since grown to over 120 coins in this latest flood tide of largely unprovenanced evidence, taking us from an average yield of fewer than three commemorative coins per decade to over thirty examples per decade. The problem is that not one of these new specimens has come from a controlled excavation, and few have ended up in museums. Of course, we are now facing the added worry that even those artifacts formerly considered safe in an Afghan museum may be in jeopardy. Of the fifty-seven recorded hoards from Bactria and India found between 1821 and 1979, only six were recovered under controlled circumstances and transported substantially intact to a museum.³ Of these, the first more or less complete Bactrian hoard to reach the National Museum in Kabul was the Qunduz treasure of 627 coins found at Khisht Tepe in 1946.⁴ To this were added two hoards excavated at Ai Khanoum in 1970 (containing 677 Indian karshapana coins and six Indo-Greek coins) and in 1973 (containing 63 Greek coins).⁵ These three hoards, half the total of intact deposits preserved from ancient Bactria and India, were not protected for long; in the turmoil since 1980, many of these coins have been pillaged from the National Museum, so that only those hoards housed outside Afghanistan have thus far escaped modern plunderers.⁶ The number of Bactrian and Indo-Greek coins (as opposed to Indian punch-marked coins) derived from secure contexts amounted, therefore, to fewer than a thousand specimens, of which many have now been scattered to the winds by way of theft and the antiquities market.⁷ Consequently, the past three decades have been especially damaging to scientific research on a scale that exceeds the worst days of the old Great Game in Central Asia.

Confronted by these grim realities, some experts must now devote at least a portion of their energies to rescue and recovery in order to make future research possible. Hoards in general provide essential guidance in the study of chronology, patterns of circulation, and political and military conditions—so much so, that considerable pains have been taken to record all such discoveries from throughout the homelands of ancient civilizations.⁸ This is true even though most of these hoards derive from non-archaeological contexts. In fact, only about 8 percent of all reported Hellenistic coin hoards from Greece to India have been recovered scientifically.⁹ To triage the other 92 percent, scholars must employ a makeshift methodology fraught with epistemological if not also personal dangers. The first hints of such a hoard usually stem from rumors in the coin trade about some spectacular find, triggered by any sudden influx of artifacts on the market. By this point, the discovery has probably been irrevocably compromised in any purely scientific sense and the scholar can only try to work backward to piece together the key data about provenance and composition: Where was the hoard discovered, and what was originally in it? The find spot cannot always be reliably ascertained; nor, in Afghanistan, can the actual finder (or finders) easily be interviewed. Everything depends on word-of-mouth reports, hampered by the suspicion that investigators may be deliberately misled by those who wish to protect a clandestine source of revenue. As for composition, hoards generally must be reconstructed based on the lamentable but necessary process of placing some trust in the timing of entry into the market. In other words, scholars must watch for the possible contents of a hoard as they show up, in dealers' shops or in auctions, following the rumors of a discovery. Here the problems multiply the more distant these potential contents appear in time and place from the original find. Hoards tend to travel quickly through bazaars and auction houses, where there is little reluctance to break them into lots for easier sale. The most collectible specimens may be culled for the connoisseurs, and extraneous merchandise may be pitched into the mix to satisfy everyone else. This may continue for months, or even years, and span several continents or even spill into cyberspace by way of electronic merchandising. Sometimes, special circumstances (such as a distinctive patina) may help in identifying the scattered coins of a given hoard, but normally the researcher must balance probabilities. Although this methodology is patently flawed and potentially very misleading, the alternative is to surrender the study of Bactria to only six intact, published hoards (three of them now pillaged).

What, then, are the realities of rescue numismatics in a place like Afghanistan? A good example may be found in the so-called Ai Khanoum Hoard III, recorded and studied under less than ideal circumstances.¹⁰ In October 1975, curator Nancy Waggoner of the American Numismatic Society was shown by a New York dealer a large group of silver coins said to have been unearthed in Afghanistan. She quickly penned some notes about the lot, which at that time contained 135 tetradrachms and 6 drachms. In January 1976, Waggoner forwarded these notes to Claire-Yvonne Petitot-Biehler in Paris, who published a summary of them in her article about the 1973 hoard excavated at Ai Khanoum by Paul Bernard (AK Hoard II, found on 7 October).¹¹ According to Petitot-Biehler, Bernard had heard rumors of a clandestine hoard while in Afghanistan in 1974, which had been found near Ai Khanoum and then quickly spirited away. These were presumably the coins seen by Waggoner in New York. Thus, it appeared that some Afghan had dug up and sold a large cache of coins just a few months after the DAFA excavated a very similar hoard from an ancient house outside the walls of Ai Khanoum. (See chap. 9, below.) This led to immediate confusion. When the editors of Coin Hoards (vol. II, hoard 88) first tried to publish these two hoards in 1976, the finds were conflated into a single discovery reportedly made at Balkh! This error was caught and partly corrected in the next issue of Coin Hoards, for 1977 (vol. III, hoard 53).¹²

Meanwhile, the coins seen by Waggoner continued to travel the world, with stops in London and Switzerland. She examined them in New York again in 1976, noting that three specimens were now missing and that four others had been added, producing a slightly different record of 142 coins. The dealer agreed to supply Waggoner with black-and-white photographs of the coins, which she later consigned to this author while he was a seminar student at the ANS in 1980. (One example, a coin of Demetrius I, is shown here in fig. 18.) As it turned out, the coins in these photographs did not all tally with the two lists made by Waggoner, and one coin seems to be a forgery and another may be an intrusion, as noted in the final publication of the hoard.¹³ Furthermore, there are no known weights or die axes for the 139 coins in the photographic record, a serious deficiency remedied only in cases where a given specimen later appeared in an auction catalogue furnishing such data.

In addition to the somewhat protean nature of the hoard as it moved through the market, eventually to be sold piecemeal, there appear to have been portions separated from the main lot before it reached New York. Henri-Paul Francfort published two commemorative coins of Agathocles that passed through the Kabul bazaar in the summer of 1975, one of them a heretofore unknown issue honoring Pantaleon Soter.¹⁴ Then, in April 1976 during a visit to England, Parmeshwari Lal Gupta happened to notice three commemorative coins of Agathocles in an unidentified auction catalogue (which turned out to be Bank Leu 15), one of them also in honor of Pantaleon.¹⁵ This last coin, unfortunately broken during its travels, turned out to be the same one reported by Francfort before it was damaged; the other three coins were different. Although Gupta did not know that one specimen was common to both lots, he nevertheless suggested that all of them “might have originated from one and the same find.”¹⁶ To these four coins can be added several more, since one of the specimens spotted by Gupta in the Bank Leu auction had previously been in the company of seven other Bactrian coins.¹⁷ Although some of these tetradrachms had not yet even been cleaned for sale, they were all fine specimens that outshone the lot photographed in New York.

Figure 18. Tetradrachm of Demetrius I from Ai Khanoum Hoard III, seen in New York City in 1976.

The methodologies of rescue numismatics use this kind of circumstantial evidence to posit the existence of a large hoard found in Afghanistan, perhaps near Ai Khanoum, in the winter of 1973–74. Before its contents ever left the country, dealers in Kabul extracted the most prized specimens, including all the highly collectible commemorative coins of Agathocles. Some of these gleanings ended up in Switzerland while the bulk of the hoard passed from country to country, its contents subject to adulteration, until all the coins were dispersed over the course of several years. Fortunately, Nancy Waggoner at the ANS was able to examine the main lot on two separate occasions and to secure a partial photographic inventory, which led to the publication of Ai Khanoum Hoard III in 1981. This represented the first foray into the field of rescue numismatics for the period 1980–2010.

Since then, the exertions of Osmund Bopearachchi in particular have been fundamental to these endeavors. Born and educated in Sri Lanka, Bopearachchi completed his advanced studies at the Sorbonne working, in part, under the tutelage of Paul Bernard. Thereafter, he commenced publishing major public collections (Bibliothèque Nationale, Smithsonian Institution, American Numismatic Society) and private holdings (Aman ur Rahman) of Bactrian and Indo-Greek coins.¹⁸ At the same time, he began investigating the numerous hoards from Afghanistan and Pakistan that were spilling onto the coin market. Materials that might otherwise have disappeared were instead partly documented, as conditions allowed, in a growing list of his books and articles. Most notable among these hoards was the so-called second deposit from Mir Zakah, Afghanistan (MZ II).¹⁹ Stumbled upon by villagers in 1992, the contents of MZ II soon reached the Shinwari bazaar, at Peshawar, and thence the broader antiquities market. Throughout 1993, talk of this treasure confounded the numismatic community.²⁰ In March 1994, Bopearachchi personally examined a number of 110-pound (50-kg) sacks groaning with artifacts from MZ II during a visit to Pakistan; he found in them an unimaginable array of about 550,000 coins as well as jewelry, statuettes, vessels, and other artworks (fig. 19).²¹ This single hoard remains the largest deposit of gold, silver, and bronze coins ever recovered from the ancient world, and its ill-fated contents have glutted the coin trade in America, Europe, Asia, and the Middle East.²² Even so, tons of coins from MZ II are rumored still to be unsold, lying in locked storage containers somewhere in Switzerland. Eventually, these will increase the number of marketed Bactrian coins recently auctioned far beyond the ten thousand currently known from sales catalogues.

Figure 19. Part of Mir Zakah Hoard II, the largest cache of ancient coins ever found.

Although the integrity of this spectacular treasure has been irrevocably compromised, Bopearachchi managed to identify and publish some of the important specimens it once contained.²³ He did likewise for smaller hoards, ranging in size from a few coins to several thousands, coming from Afghanistan, Tadjikistan, Pakistan, and India.²⁴ One or more hoards containing a total exceeding fifteen hundred coins, about half of them examined by Bopearachchi, has been published as Ai Khanoum Hoard IV.²⁵ The composition of this hoard mirrors that of the other finds from Ai Khanoum. In 1996, a similar treasure of more than eight hundred coins came to light at Kuliab, site of the discovery of the Heliodotus inscription. (See chap. 6, above.) Learning of this find through a private collector, Bopearachchi managed to study about a fourth of the coins directly or through photographs.²⁶ (See plate 5.) Many of these specimens are small denominations (drachms and obols), yet the composition of the hoard corresponds to those found at Ai Khanoum (AK II–IV), which extend from the reigns of Alexander to Eucratides.²⁷ A similar hoard of Bactrian tetradrachms spanning the reigns of Diodotus I to Agathocles was discovered at Tokhmach Tepe, near Bukhara, in 1983 and published by archaeologist Edvard Rtveladze.²⁸

During regular investigative missions to Pakistan during the 1990s, Osmund Bopearachchi recognized an interesting phenomenon of rescue numismatics. Many Indo-Greek hoards were being reported from regions hosting large throngs of Afghan refugees. In the process of setting up camps and digging foundations, often on unrecognized archaeological sites, busy shovels were striking silver and gold.²⁹ Indeed, any increase in population or resource development tends to produce new coin finds through digging, dredging, or plowing. In 1992, a broken pot in the Swat River near Saidu-Sharif yielded about eight hundred coins, some of them unique. In October 1993, a hoard of the same size entered Western markets from Bajaur, where two similar deposits had been uncovered in 1942. Another eighty-three coins were unearthed in Mian Khan Sanghou (Dec. 1993), plus about 1,220 coins in the village of Wesa (Jan. 1994). At Sarai Saleh, near Haripur, a bulldozer leveling the ground for the tomb of a village leader struck a bronze vase in 1994. About two thousand Indo-Greek and Indo-Scythian coins tumbled out and were sold in the bazaars of Haripur, Islamabad, and Peshawar. Subsequent investigations revealed that a substantial ancient city once stood on the site, and archaeological excavations commenced there in 1995. Likewise, another set of finds in a farmer's fields at Siranwali produced not only a hoard of four hundred coins (mainly of Menander) but also fragments of Hellenistic pottery suggesting the presence of an ancient settlement.³⁰

Rumors spread in April 2001 about an unusual discovery of gold coins minted by the first independent kings of Bactria: Diodotus I, Diodotus II, and Euthydemus I.³¹ Villagers digging clay from the Ganges near Vaisali, deep in India, allegedly hauled up a clay pot containing the treasure. Some of the artifacts were precipitously melted down by local jewelers, but within a few months others went to European and American dealers; this trove is still being actively marketed. The estimated number of specimens actually found has been greatly reduced over time, from over a thousand to fewer than two hundred, but this discovery may yet help to rewrite the history of these kings.

Out of these many recent finds have emerged previously unknown rulers (such as Nashten and Heliodorus), unexpected coin types, monograms, or denominations (of Alexander, Eucratides, Menander, Amyntas, Artemidorus, etc.), and examples of new overstrikes of one coinage over another (such as Euthydemus on Diodotus, and Agathocleia on Menander). All this information is valuable, even if so much more might have been learned had these hoards been subject to careful scrutiny in a public collection. As it stands, many of the coins from these hoards will never be properly cleaned, weighed, photographed, or published. They are hostages in an ongoing war against error, a struggle for truth waged by scholars armed with the clumsy but essential weaponry of rescue numismatics.

Bactrian evidence comes at such a price that even the oldest of discoveries must from time to time be rerescued. At the British Museum, a special initiative was launched in 1993 to recover and catalogue the vast collections made by Charles Masson.³² This explorer's sixty thousand coins sped through the arteries of nineteenth-century British numismatics, many ending up in private hands and others at the East India Company, the British Museum, the India Office Museum, the Ashmolean Museum, the Fitzwilliam Museum, and so forth, until a portion actually made the return trip to India. The archival research necessary to retrieve this data is no mere parlor game or academic's pastime; the results matter a great deal to our understanding of the sites stripped by Masson, such as the city of Begram.³³ Similar work is also progressing on the early collections assembled and then dispersed by the French officers serving Ranjit Singh, such as General Court.³⁴

The surge of new evidence drawn from these many facets of rescue numismatics has pushed Bactrian studies far beyond the compass of narrative numismatics, even though a public that prefers its facts in story form still accords writers like Tarn and Narain a powerful hold over Bactrian history. Rescue numismatics has led to revisionist numismatics, an approach that more or less abandons the crafting of elaborate narratives for the specialist's emphasis on technical analysis. The aim is to revise earlier views by tackling specific problems in light of fresh evidence and emerging technologies, including computer-enhanced analyses and advanced laboratory methods (X-ray fluorescence spectrometry, scanning electron microscopy, atomic absorption spectrometry, etc.).³⁵ Electronic archives and databases are fast becoming the essential tools of the academic trade. Meantime, the rudimentary, now outdated catalogues of the nineteenth and early twentieth centuries have been superseded on the shelf by more scientific, comprehensive reference works on the model of Bopearachchi's Catalogue raisonné of the much-expanded collection in the Bibliothèque Nationale.³⁶ Centuries-old problems, such as sorting out the coinages of Diodotus I from those of Diodotus II, have now been tackled with some success.³⁷ Tarn's imaginative interpretation of the so-called pedigree coins has been proven false by closer analysis of these issues.³⁸ Epigraphic evidence has been tied to the coinages in order to rethink the chronologies of the Indo-Greeks.³⁹ Sequences of currency are being refined by the study of overstrikes.⁴⁰ All this progress results from a notable increase in the amount of specialized research being published on Bactrian numismatics.⁴¹

One welcome aspect of revisionist numismatics is its underlying emphasis upon proper methodology. In 1992, Osmund Bopearachchi announced a new approach to the study of Bactrian history, one that challenged the works of Tarn, Narain, and others.⁴² Addressing one of the oldest problems in Bactrian numismatics, Bopearachchi proposed a different methodology for the study of monograms. He noted that the total number of these markings had been vastly inflated by the gradual accumulation of errant readings. Cutting that figure by more than half, he then grouped the monograms into units with their hypothetical variants, so that the control marks and might be officinae within the mint represented by . Bopearachchi then tried to localize these units based on find spots.⁴³ In the cases of long-lived monogram groups, Bopearachchi then sought new data such as die links and overstrikes to posit a better chronology.⁴⁴ Drawing on a wide range of evidence, he argued that previous scholars had been completely mistaken in their reconstructions of the final period of Greek rule in India.⁴⁵

More recently, Peter Mittag has again addressed the methodological Kernproblem in Bactrian studies, namely the handling of material and written sources.⁴⁶ In order to unravel the basic questions still surrounding ancient Bactria (Where, in what order, and exactly when did its forty-five known sovereigns reign?), Mittag discusses a series of access points to each problem. For example, chronologies can be established by way of a few coins possibly bearing dates (Plato, e.g.), the series of commemorative issues (Erinnerungsmünzen) struck by Agathocles and Antimachus, nearly three dozen recorded overstrikes, kinship (as attested in texts or surmised from coins), die links, iconographic borrowing, hoards, stylistic differences, letter forms, monograms, and the legends on coins. Mittag rightly stresses that not all these methods are of equal value, with overstrikes clearly ranking as the most deterministic, since they establish a relative order of issues. Less reliable in his view are chronological conclusions drawn from the study of iconography, hoards, style, or letter forms. Geographic areas controlled by specific rulers might be reconstructed from find spots, the languages used on the coins, monograms, coin types, style, and letter forms; however, none of these is considered particularly reliable. Mittag's hierarchy of evidence serves as his basis for a schematic reconstruction of the Bactrian and Indo-Greek reigns.⁴⁷

Only one attempt has been made to address these matters in a comprehensive epistemological fashion. In 1987, the noted archaeologist and numismatist Olivier Guillaume embarked upon this ambitious mission in his book L'analyse de raisonnements en archéologie: Le cas de la numismatique greco-bactrienne et indo-grecque, a work so important that it was quickly reissued in an English edition.⁴⁸ Its author proclaimed:⁴⁹

The purpose of this study is to sum up the present situation of numismatics within the field of Graeco-Bactrian and Indo-Greek studies from a methodological point of view. We shall examine how numismatists describe the coins in their catalogues and how historians use the catalogues to establish their historical reconstructions.

Guillaume used the logicist analysis then in vogue among proponents of the so-called New Archaeology (see chaps. 5 and 8), inspired by theoreticians such as Jean-Claude Gardin, who had also worked at Ai Khanoum.⁵⁰ The approach is therefore described as a kind of calculus that first exposes how the numismatic data have been processed in a sample of six standard catalogues: Cunningham (1884), Gardner (1886), Smith (1906), Whitehead (1914), Lahiri (1965), and Mitchiner (1978). These seemingly straightforward compilations of hard evidence are shown, in fact, to be laden with all sorts of ambiguous language and reasonings. Thus, the development of Bactrian studies has depended not only on the expansion of numismatic evidence but also on how that material has been arranged and described: one cataloguer's Bactrian coin may be another's forgery or barbarian imitation; one expert's Apollo may be another's Artemis or Anaitis.

In the second half of his book, Guillaume examines how these catalogues have been used by historians to create their narrative accounts, specifically those of Tarn (1951) and Narain (1957).⁵¹ The goal is to unveil the sequence of propositions (P1, P2, P3…) that leads from the cataloguers' data to the historians' reconstructions, based on eleven variables such as metals, overstrikes, and epithets.⁵² For example, the use of nickel (or cupronickel) for some coinages is a fact stated in most of the catalogues, which then becomes another kind of fact in the works of Tarn and Narain, who draw different conclusions from it regarding trade and conquest, respectively.⁵³ The same is true of overstrikes, which Tarn and Narain take as proof both of conquest and of a desperate shortage of metals, depending upon the needs of their narratives.⁵⁴ In terms of epithets, each author explains a given title according to preconceived notions.⁵⁵ For Tarn, the epithet (“Just”) was taken by Heliocles in order to win favor among the Euthydemids, whereas Narain saw it as Heliocles' honorific for killing Plato, the murderer of their father, Eucratides. The same epithet, , on the coins of Menander has a possible Buddhist connection in the eyes of Narain, but not for Tarn.

In the end, Guillaume concludes that the foundational works of Tarn and Narain are both flawed by weak methodologies and ideological biases.⁵⁶ He suggests that “the literary form of the narrative is singularly inadequate” for Bactrian studies.⁵⁷ In fact, it is characterized as a dead end: “as long as historical reconstructions are presented—as they are—as narratives one cannot hope for any progress.”⁵⁸ Instead, Guillaume calls for closer attention to numismatic typologies and hypothesis testing as the best ways to move forward. To some extent, revisionist numismatics has answered that appeal for more scientific, problem-oriented investigations.⁵⁹

Yet some authors still cling to the worst enticements of narrative numismatics. As one example, in his recent books about Bactria, anthropologist Homayun Sidky offers what has been called a “critical narrative” to supersede the outdated, impressionistic classic by Tarn.⁶⁰ Curiously, Sidky takes a hostile tone toward professionals, specialists, and experts—terms he usually sets apart derisively in quotation marks; the methods and interests of these unnamed and allegedly “obtuse” specialists are deemed antithetical to Sidky's self-professed emphasis upon a “narrative approach” and descriptive imagery.⁶¹ Without conducting any original research of his own, Sidky shuffles the basic reconstructions of others, particularly Tarn and Narain, into a synthetic narrative that perpetuates the kinds of problems identified by Guillaume. Thus, for example, Sidky tells the story of Demetrius's fight against Eucratides:⁶²

And so, Demetrius, who as a young man alongside his father had faced Antiochus the Great, whose conquests south of the Hindu Kush considerably increased the size of the Greek kingdom, and whose military successes had earned him the epithet Aniketos, “the Invincible,” met his end on the field of battle. This was not at all a shameful death for a rash adventurer, with more ambition than skill, as Narain makes of Demetrius.

As Narain makes of which Demetrius? Here lies the hopeless mélange: Narain never says that Demetrius I, son of Euthydemus, met a shameful death at the hands of Eucratides because of rashness, ambition, or anything else. Narain chides that Tarn would have it thus, but that Demetrius II actually fought Eucratides.⁶³ Nor, of course, does any source report that either Demetrius died on the battlefield. That bit of descriptive imagery is a narrative builder. In fact, Sidky admits that the battle he envisions is not mentioned anywhere but asserts that one is required nonetheless to complete his story of Eucratides' conquest of India.⁶⁴

Even in the pages of an academic numismatic journal published both in Greek and in English, the shadow of Tarn and narrative numismatics still lingers. Jens Jakobsson criticizes “the far from satisfying status of current scholarship” and advocates for a better methodology.⁶⁵ He worries that the recent trend of specialized research (i.e., revisionist numismatics) has once again driven a wedge between coin collectors or numismatists on the one hand and historians or philologists on the other.⁶⁶ One solution, he proposes, is to revisit “the somewhat abandoned field of relations between different rulers.”⁶⁷ We return, therefore, to speculations regarding possible marriage connections between various kings. For example, Jakobsson interprets the commemorative coins of Antimachus as a record of actual ancestry, whereas those of Agathocles were not. This leads to the supposition that⁶⁸

Antimachus I was a younger son of Euthydemos I, who had been married to a daughter of Diodotos I, as could well be imagined. Antimachus was then the most legitimate heir to Euthydemos II, whereas Agathokles and Pantaleon belonged to a sideline.

Similarly, Jakobsson argues that “Eukratides I had probably been married to a daughter of Apollodotos I” and that Heliocles the Just “was not necessarily a relative of Eukratides, but was perhaps married to Eukratides' sister or daughter Laodike, or if Laodike was Eukratides' queen, Heliokles was her brother or father.”⁶⁹ It is difficult to see how these guesses are indeed any “better rooted in facts” than the “fanciful speculations” of W. W. Tarn.⁷⁰

There have appeared three larger studies of Hellenistic Bactria and India since 2005, all keen to revive the study of the written evidence for whatever clues may have been missed. The premature death of Stanislaw Kalita in 2000 during the early stages of his promising academic career means that we have only an undeveloped version of his book Grecy w Baktrii i w Indiach, redacted by his university colleague Edward Dabrowa.⁷¹ It was Kalita's aim to examine anew the history of the Greeks in Bactria and India “through the prism of the narrative sources” in order to compare more closely the written and numismatic evidence.⁷² This work therefore deals only with those problems for which there is textual evidence, such as the chronology of Bactria's secession from the Seleucids, the war between Euthydemus I and Antiochus III, and the reign of Menander. In its unfinished state, this study glosses over the career of Eucratides.

On the other hand, a well-researched and richly documented dissertation by Omar Coloru, a student at the University of Pisa, has recently been published as Da Alessandro a Menandro: Il regno greco di Battriana.⁷³ Although written very much in the tradition of narrative numismatics, with emphases on chronology and dynastic relationships, the work is revisionist in its up-to-date incorporation of epigraphic, archaeological, and numismatic evidence.⁷⁴ Like Tarn, Coloru treats the reigns of Euthydemus and his son Demetrius as the golden age of Bactria—followed by civil war and the disastrous usurpation of Eucratides. The wars of the latter king are reconstructed in great detail (pp. 209–29) along the lines suggested by Justin's epitome of Trogus, with special attention to affairs in the West involving the Seleucids, Mithridates, and Timarchus. Although the local religions of Bactria and India made some inroads into the Greek-ruled state, Coloru considers the area as very much a satellite of the Seleucid empire.

The work of François Widemann (director of research, CNRS) offers a sweeping treatment based on technical analyses of the coinages coupled with new archaeological and epigraphic evidence and a rehabilitative faith in the surviving scraps of literary sources.⁷⁵ The author accepts the impracticality of creating an “histoire événementielle” and addresses the many “hypothèses invérifiables” of Tarn.⁷⁶ The work, however, is (by the author's own admission) uneven, speculative, and essentially narrative in its approach. Widemann tells the story of Bactria idiosyncratically, his own interests reflected in a chronicle that is at some times meticulous and at others superficial. For example, he takes no notice of recent research published on the Diodotids; yet the short reigns of the so-called Nickel Kings (Euthydemus II, Pantaleon, Agathocles) receive twice as much attention as Diodotus I, Diodotus II, Euthydemus I, and Demetrius I combined. This is largely due to the author's fascination with the cupronickel coinages of Bactria.⁷⁷ Widemann contends (following Tarn) that Bactria suffered a perpetual “penury of precious metals” in spite of the fact that it minted such huge amounts—and huge denominations—of gold and especially silver.⁷⁸ The narrative sometimes assumes a Tarnlike precision in its chronology, as in the calculation of kings' ages and reigns, and in its reconstruction of family relationships.⁷⁹ Truly a Tarn for the twenty-first century, Widemann concocts from thin air a remarkable institutional structure for Bactria. In order to bridge the gaps between various dynasties, he imagines a powerful “conseil de Bactres” while admitting that there is no evidence for it (“mais aucune source ne renseigne sur ce point”).⁸⁰ He posits a formal system of epithets (“épiclèses”) peculiar to Bactria, which allegedly were conferred by this Bactrian council as legal indicators of rank among co-ruling kings. To make this work, Widemann must arbitrarily switch the epithets normally assigned to Diodotus I and II, and suppose that Eucratides I did not adopt the requisite epithet Theos because the council refused to grant it to him.⁸¹ There are even identifiable factions of Greeks, Indians, and Sakas that allegedly animate the politics of Taxila, and like Tarn's Euthydemids, Widemann's King Maues champions Alexander's alleged dream of a brotherhood of all mankind.⁸²

It seems, therefore, that Bactrian studies remain partly mated to the old methodologies of narrative numismatics even while rescue and revisionist numismatics struggle to chart a new path for the twenty-first century. The numerous finds of fresh archaeological, epigraphic, and numismatic evidence may someday make clear the subtleties of Bactrian history, but for now a fundamental impasse remains. Bactrian studies began in the eighteenth century with compilations of king lists and other ostensible facts derived from classical literary sources, as was the norm for most aspects of ancient history. Old texts were deemed authoritative and scarcely needed validation; at most, scholars hoped to illustrate the content of those sources by means of ancillary monuments, art, and coins recovered from the earth. Gradually, the material record became important in its own right and sometimes served as a counterweight to the writings of poets, geographers, and historians. Whether the subject was Egypt, Greece, or Rome, the past revealed itself through an increasingly sophisticated system reliant upon what we might call the generation, verification, and integration (GVI) of data. Wherever source materials have been plentiful and diverse, the GVI process has worked quite well, but we have seen that Bactria is not yet such a place. Our ability to generate, verify, and integrate purported facts is severely hampered and seems incapable so far of sustaining the kinds of histories the public craves.

The literary texts, for example, claim that King Eucratides ruled a thousand cities, but neither numismatics nor archaeology can verify this. We know from Strabo that one of those cites was named for Eucratides himself, but we cannot confirm the belief of many that the archaeological site of Ai Khanoum is that place. It is no easy task to verify and integrate the results drawn from one kind of source with those from another. In the main, we have generated a large number of isolated literary, epigraphic, numismatic, and archaeological facts, but these seldom lend themselves to verification or integration on any scientific basis. Thus, the evidence about Bactria remains compartmentalized and little stronger than its unconnected pieces. As a result, almost any potential fact generated by even a respected source such as Polybius can be doubted if there is no corroborating testimony.⁸³

Take the unusually multifaceted case of Demetrius I. Historians can generate seven or more facts about this man based on ancient literary sources (L1–7), two facts based on epigraphy (E1 and E2), eleven facts derived from numismatics (N1–11), and a pair of archaeological facts (A1 and A2):

 

L1:	Demetrius was the son of Euthydemus I (Polybius).
L2:	As a young man (neaniskos) in about 206 B.C.E., Demetrius confirmed the treaty negotiated by Teleas between Antiochus III and Euthydemus I (Polybius).
L3:	Demetrius was deemed by Antiochus III worthy of kingship based upon his appearance, dignified bearing, and conversation, and Euthydemus was allowed to call himself king (Polybius).
L4:	Demetrius was promised a daughter of Antiochus III as a bride (Polybius).
L5:	Demetrius the son of the Bactrian king Euthydemus conquered part of India (Strabo).
L6:	A Demetrius (I or II?) king of India (“rex Indorum”) fought against Eucratides (Justin).84
L7:	A city named Demetrias existed in Arachosia (Isidore of Charax).85
E1:	Demetrius was the son of Euthydemus I (Kuliab Inscription).
E2:	Demetrius was hailed as Kallinikos, “Glorious in Victory” (Kuliab Inscription).
N1:	Demetrius was given the epithet Anikētos on the commemorative coins of Agathocles.
N2:	Demetrius issued Attic-standard silver portrait coins showing the draped king with a diadem and wearing an elephant-scalp headdress.
N3:	These coins of Demetrius used the reverse coin type of Hercules crowning himself while holding a club and lion pelt.
N4:	Demetrius issued Attic-standard bronze coins of three varieties: (obverse) Hercules,(reverse) standing Artemis; (obverse) elephant head, (reverse) caduceus; (obverse) shield, (reverse) trident
N5:	Demetrius's coins all bear the legend .
N6:	Silver coins with this same inscription but different portrait and coin type (standing Athena) may belong to a later king.
N7:	Demetrius did not strike gold or cupronickel coinage.
N8:	Demetrius's silver coins are relatively few in hoards as compared with his father's: Ai Khanoum Hoard II (Euthydemus 27, Demetrius 3), Ai Khanoum Hoard III (Euthydemus 81, Demetrius 8), Bukhara Hoard (Euthydemus 51, Demetrius 0), Qunduz Hoard (Euthydemus 12, Demetrius 8), although the partly reported Kuliab Hoard may be an exception (Euthydemus 29, Demetrius 50)
N9:	Demetrius struck his coinage on a fixed 12:00 axis.
N10:	Demetrius used over a dozen monograms, but mostly and .
N11:	A reverse coin die of Demetrius may derive from the area of Ai Khanoum.86
A1:	Demetrius's bronze coins are relatively few as compared with his father's in the Ai Khanoum excavations (Euthydemus 49, Demetrius 5).
A2:	A bronze statuette of Hercules crowning himself, excavated at Ai Khanoum, resembles Demetrius's coin type.

 

This list provides us with the best-attested fact in all Bactrian studies: Demetrius I was the son of Euthydemus I, as verified independently by two literary sources and an inscription (L1, L5, and E1) and not contradicted by any known numismatic or archaeological evidence.⁸⁷ No other relationship, not even that of Diodotus I and II, can attain this level of certainty, much less the dozens of reconstructed family connections and invented persons commonly found in modern histories. This does not mean that Heliocles and Laodice were not the parents of Eucratides, or that Pantaleon and Agathocles were not brothers; but these are conjectures at best, with numerous scholars vigorously disagreeing about them. The sundry rulers identified as additional sons of Euthydemus (Euthydemus II, Antimachus, Agathocles, Pantaleon) are not corroborated anywhere; all written sources mention only young Demetrius. One source gives his epithet as “Glorious in Victory” (E2); another says “Invincible” (N1), but his own issues offer neither (N5). Dare we link the military epithets to Demetrius's alleged conquests in India (L5, L6) and Arachosia (L7)? The elephants on his coins have been logically associated with these wars (N2, N4), although not always with due consideration for the chronological crux of when these conquests took place. The coins indicate that any such campaigns were completed before Demetrius began to strike coins, since the elephant scalp is featured from the very outset. We might assume, therefore, that Demetrius either struck his father's coinage in the beginning of his reign (N8, A1) or that he achieved his military successes before his accession.⁸⁸ The latter view is supported by Polybius's remarks that Euthydemus I (but not his son, though worthy) was granted the royal title (L3), that Demetrius was Kallinikos when not yet king (E2), and that his father alone was identified as the ruler of Bactria at the time when Demetrius invaded India (L5). Yet there are no Indo-Greek coins attributed to Demetrius I, and the jury remains out on whether he could be the same “rex Indorum” who fought Eucratides (L6).⁸⁹

Most of these twenty-two facts, therefore, do not verify each other, nor are they readily integrated into a reliable narrative of Demetrius's career. From Polybius to Chaucer, we may say that no Bactrian ruler enjoyed a better reputation than young Demetrius, but we cannot with any certainty explain what he did to earn that acclaim. There is no mention of him in the actual fighting against Antiochus the Great, and he was not yet twenty when his diplomatic skills impressed the Seleucid overlord. He was pledged a royal wife, but no one knows if he got one. He was thought kingly in manner, but apparently did not share that title with his father, Euthydemus. What sort of military glory did he have in the eyes of the man who commissioned the Kuliab Inscription or designed the commemorative coins? Must we seek that glory hidden somewhere in the king's elephant headdress? Did Demetrius restore Bactrian honor by invading India during his father's reign, in spite of the treaty between the Indians and the Seleucids, in order to grab elephants, which were the only symbolic spoils of war taken from Euthydemus by Antiochus? Or did Demetrius merely accompany Antiochus to India at the head of the Seleucid's elephant corps? In the end, we see that generating facts can be relatively easy, but not verifying them independently or integrating them into a sound narrative.

Even a case seemingly custom-made for GVI fails us in Bactria when On the Circumnavigation of the Red Sea, a literary source, appears to generate a blatantly numismatic fact: “Even now in Barygaza [modern Broach, Bharuch], old drachms stamped with inscriptions in Greek lettering come to hand, the coins of Apollodotus and Menander, who were kings after Alexander.”⁹⁰ This one should be simple; but alas, there are no significant finds of these coins in the region of modern Bharuch that would substantiate the testimony of the anonymous author of this sailing manual. Only a handful of Menander coins have ever been recovered from this area, and most of the Apollodotus (II) examples derive from a single hoard dispersed from Gogha.⁹¹ Even these specimens were unknown to the historian Tarn, who nevertheless sided with the texts and integrated this fact into his account as a reliable guide to the extent of Apollodotus's realm: “Consequently Apollodotus' rule in Barygaza cannot be in doubt.”⁹² The numismatist Narain, on the other hand, noted the paltry coin data and integrated this contrary fact into his narrative as evidence that Tarn's Apollodotus I never reigned there and probably never even existed at all.⁹³ The tremendous efforts of rescue and revisionist numismatics have yet to resolve these problems.⁹⁴ One possible solution may be sought in another approach, one that uses coins to answer a different set of questions.