Time is Money? A Second Look at Production Quantification and Chronology in the Late Seleucid Period more

OLIVER D. HOOVER TIME IS MONEY? A SECOND LOOK AT PRODUCTION QUANTIFICATION AND CHRONOLOGY IN THE LATE SELEUCID PERIOD Introduction At the end of 2007, Historia published an article in which it was suggested that the die usage estimates for the late Seleucid kings at Antioch on the Orontes might be a helpful tool for determining the relative chronology of rulers for whom the textual sources offer poor evidence. 1 This suggestion was based on a tetradrachm die study begun alongside Arthur Houghton and which will hopefully see publication in association with Panagiotis Iossif, and the use of the statistical method for estimating dies presented by Warren Esty in 2006. 2 In this paper it will be argued that tetradrachm sequence used in 2007 for the two reigns of Antiochus X should probably be inverted. The reversal of the coins attributed to the first and second reigns of this king creates a problem because the estimated number of dies seems inappropriate for the periods of rule involved. A possible explanation for this phenomenon may be found in the large-scale recoinage that took place under Philip I. The Case for a Stable Rhythm of Production at Antioch Looking at the data and estimated numbers of dies for the dated tetradrachms 1 Hoover 2007. I am grateful to François de Callataÿ for the opportunity to present these concerns in Rome as well as to Panagiotis Iossif and Peter van Alfen, who read an early version of this paper, to Carmen Arnold-Biucchi, who championed it at the conference, and to Michel Amandry and Warren Esty, who passionately critiqued it. I also wish to thank Catharine Lorber for the long and occasionally heated discussions of late Seleucid coinage that ultimately led me to reappraise my earlier position on Antiochus X. Nevertheless, all conclusions are the sole responsibility of the author. 2 Esty 2006, 359-364. QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it 252 OLIVER D. HOOVER of Antioch under Demetrius I and Alexander I in Table 1 (figs. 1-2), 3 one gets the impression of fairly stable tetradrachm production from one year to the next except in times of major conflict in Syria. Fig. 1 Fig. 2 During the relatively peaceful years 155/4, 154/3 and 153/2 BC, Antioch used roughly 10-16 or maybe a few more dies per year to produce tetradrachms for Demetrius. However, the numbers suddenly double in 152/1, probably in response to the serious military threat posed by the rival king Alexander I in that year. In the summer of 153 BC Alexander had obtained a senatus consultum recognizing him as the legitimate Seleucid king and by 152 BC he was operating against Demetrius from Ptolemaïs-Ake with the support of Ptolemy VI and a coalition of foreign kings opposed to the Demetrian regime. 4 The severity of the danger and the need for funds to combat Alexander in 152/1 is illustrated by the sudden production of gold staters for Demetrius I in that year (fig. 3). 5 Fig. 3 Tetradrachm die usage appears to fall back to pre-152/1 levels in 151/0, which is a little unexpected as this was the year in which the struggle for Syria was most acute for Demetrius. However, this may be explained by the fact that Demetrius Houghton, Lorber & Hoover 2008, nos. 1641 (Demetrius I) and 1781.3-1784 (Alexander I). Diod. 31.32a; Jos. AJ 13.35-36; Just. 35.1.6-11; Polyb. 33.18.6-14; Eus. Chron. 1.255; Ehling 2008, 151-153. 5 Houghton, Lorber & Hoover 2008, no. 1627. 4 3 QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it TIME IS MONEY? 253 was killed in battle, probably in June of 150 BC. 6 As the Macedonian Seleucid calendar employed in Syria began the year in October/November, this means that for at least a quarter of 151/0, Antioch struck no coins in the name of Demetrius. Bearing this in mind, if we add a quarter to the estimates for 151/0 we arrive at an estimate of 16-21 dies. These numbers would be in keeping with the high production of the preceding year and are what we might expect to see considering that the depth of the crisis for Demetrius I in 151/0 was so great that Antioch struck a series of emergency gold coins in that year including octadrachm, tristater, 2 ½ stater, distater, and hemistater denominations (fig. 4). 7 Fig. 4                                                                                           Table 1. - Dated Tetradrachms of Demetrius I and Alexander I at Antioch. A roughly similar pattern occurs at Antioch under Alexander I. Die usage is fairly stable, if comparatively low in 150/49 and 148/7 BC, but more than doubles in 149/8 and 147/6 BC. The increase in 147/6 can be attributed to Alexander’s need to raise an army against Demetrius II who arrived in Syria with a force of Cretan mercenaries in that year and gained the support of Alexander’s former patron, Ptolemy VI. 8 However, the reason for high production in 149/8 is unclear. 6 Demetrius’ defeat and death is mentioned in a Babylonian astronomical diary dated to June/July 150 BC. See Van der Spek 1997-1998, 167-168. 7 Houghton, Lorber & Hoover 2008, nos. 1628-1632. 8 Diod. 32.9c; Jos. AJ 13.103-110; 1 Macc. 11.1-12; Just. 35.2.3; Ehling 2008, 159-164. QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it 254 OLIVER D. HOOVER                                                            Table 2. - Dated Tetradrachms of Demetrius II and Antiochus VI at Antioch. The evidence of the dated tetradrachms produced by Demetrius II during his first reign and those of his nemesis, the young Antiochus VI (figs. 5-6), also gives the appearance of stable annual production hovering at about 9-14 dies per year (see Table 2). 9 Fig. 5 Fig. 6 The case is similar for the second reign of Demetrius after his return from Parthian captivity. 10 Although the Antiochene tetradrachms are undated, we are on fairly solid ground in attributing all of his tetradrachms to the single year 129/8 BC. The dated bronzes of the pretender Alexander II Zabinas show that Demetrius held Antioch for less than a full year after his return. 11 Demetrius’ estimated 11-15 dies for his brief second reign at Antioch is generally in keeping with the estimates for the his earlier dated tetradrachms as well as those of Demetrius I, Alexander Balas, and Antiochus VI that we have already looked at. The general stability of annual die usage is also apparent at Damascus under the late Seleucid king Demetrius III (see Table 3 and fig. 7). 12 Estimated annual usage hovers around 2-7 dies. Damascus, of course, was a much smaller mint than Antioch, and therefore the die usage there is far smaller than at the Syrian capital. The astonishingly large estimates for 91, 90 and 88 BC are likely to be artifacts of the poor coverage of the sample and therefore should Houghton, Lorber & Hoover 2008, nos. 1906-1907 (Demetrius II) and 2000 (Antiochus VI) Ibid., nos. 2164-2166. 11 Ibid., no. 2229 is dated SE 184 (129/8 BC). 12 For the numismatic peculiarities and historical reconstruction of Demetrius III’s reign at Damascus, see Hoover, Houghton, and Veselý 2009, 306-316. 10 9 QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it TIME IS MONEY? 255 Dest 1.06 5.42 3.54 4.70 2.11 13.80 9.29 2.31 4.71 Low Conf. 0.7 4.3 2.2 3.2 1.6 7.5 10.8 1.4 4.9 High Conf. 1.6 6.8 5.7 7.0 2.8 26.8 50.9 3.9 24.2 Reign Demetrius III Dates 97/6 96/5 95/4 94/3 92/1 91/0 90/89 89/8 88/7 n 7 28 11 15 14 15 4 8 3 d 1 5 3 4 2 8 3 2 2 d1 0 1 0 1 0 4 2 0 1 Cest 1 0.964 1 0.933 1 0.733 0.5 1 0.667 Table 3. - Dated Tetradrachms of Demetrius III at Damascus. Fig. 7 not be taken as evidence of increased production. On the other hand, we might expect a higher estimate in 91 BC, after an entire year with no new coinage at Damascus. Likewise, 88 BC was the year that Demetrius probably marched north to briefly seize Antioch and challenge his brother Philip for control of Syria. 13 Unfortunately the samples are far too small for the dated issues of Demetrius’ Damascene successors Antiochus XII and Tigranes of Armenia to judge with any accuracy whether the mint continued this rate of die usage. 14 All of this tends to support the recent differentiation between “peacetime” and “wartime” emissions, but the apparent stability of production from year to year might seem to tell against the controversial theory of the Seleucid financial administration consciously “topping up” the circulating medium. 15 In a strictly Seleucid context, it may even lend some degree of credibility to the dubious practice of adducing annual die usage by averaging the number of estimated dies for undated coins over the known years of a king’s reign. Hoover 2007, 293-295. For the 1-4 coins known per year of Antiochus XII’s reign, see Hoover, Houghton, and Veselý 2009, 335-336. For the 1-2 coins known per year of Tigranes’ reign at Damascus, see Nercessian 2006, 80-82. 15 Aperghis 2003, 237-246; Aperghis 2001, 90-95. For criticism of the theory of replacement coinage and “topping up” the circulating medium of the Seleucid Empire, see Le Rider and de Callataÿ 2006, 217-221. 14 13 QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it 256 Reign Antiochus VIII (1) Antiochus IX (1) Antiochus VIII (2) Antiochus IX (2) Antiochus VIII (3) Antiochus IX (3) Dates 121/0-114/3 113/12 113/12-111/10 110/09 109/8-c. 97 c. 97/6-96/5 n OLIVER D. HOOVER d 32 20 30 16 78 17 d1 8 2 10 4 27 4 Cest 0.968 0.988 0.882 0.941 0.915 0.945 Dest 37.2 21.3 39.7 19.1 100.0 20.1 247 168 85 68 319 73 Confidence Interval 34.7-39.9 19.7-23 32.2-48.9 15.9-22.9 91.5-109.3 16.9-23.9 Annual Average 4-5 20-23 11-16 16-23 7-9 8-12 Table 4. - Undated Tetradrachms of Antiochus VIII and Antiochus IX at Antioch (Periodized by Associated Dated Bronzes). When we reach the three alternating reigns of Antiochus VIII and Antiochus IX at Antioch (figs. 8-9), we get a similar impression of relative production stability from year to year. Fig. 8 Fig. 9 However, the margin for error is much greater since the tetradrachms of these kings were not dated but can only be periodized by the association of their control marks with those found on dated bronze coins of Antioch (figs. 10-11). 16 Fig. 10 Fig. 11 Recognizing this element of imprecision, annual die usage for each reign of Antiochus VIII generally falls into the neighborhood of 4-16 dies, while that of Antiochus IX is between 8 and 20. Based on this apparently stable rhythm of production at Antioch, it was argued that one could get a feel for what the expected annual average die usage should be under the late Seleucid kings who issued no dated coins at Antioch. If one has an 16 Houghton, Lorber & Hoover 2008, vol. 1, 484-485, 499-506 with nos. 2297-2316 (Antiochus VIII), and 521, 533-538 with nos. 2363-2374 (Antiochus IX); Hoover 2007, 284-287. QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it TIME IS MONEY? 257 idea of probable annual die usage it is then possible to take the numismatic data for kings whose regnal periods are derived from the corrupt chronologies preserved by Eusebius and St. Jerome, dubious statements by Josephus, and a few dated market weights, and cautiously estimate the probable number of years in a king’s reign. 17 Using this methodology of estimating annual die usage in combination with the textual and metrological evidence yielded a revised chronology for the late Seleucids at Antioch that extended the second reign of Antiochus X down to 90/89 BC, reduced the reign of Demetrius III to 88/7 BC, extended the reign of Philip I from 88/7 to c. 75 BC, and reduced the reign of Tigranes to c. 74/3-69/8 BC. The limited textual evidence ultimately supports all of these major modifications. The end date for Philip and the start date for Tigranes is guaranteed by a statement of Cicero regarding the recognized legitimacy of Antiochus XIII in Syria in c. 75 BC and Josephus’ remark that the Hasmonaean queen Salome Alexandra (76/5-67 BC), was suddenly alarmed at the arrival of Tigranes. 18 A Quantitative-Chronological Problem Nevertheless, doubts have now arisen about the propriety of associating an anomalous bronze series (fig. 12) marked with the monogram – a variant form used by Seleucus VI at Antioch – with Antiochus X’s tetradrachms marked with ee and placed first in the regnal sequence by Newell (fig. 13). 19 Fig. 12 Fig. 13 Regular Antiochene mint practice at the end of the second and early first centuries BC was to duplicate controls on both silver and bronze emissions. 20 If the } bronzes are not paired with the tetradrachms, as seems likely in retrospect, this would then indicate that no royal bronze coinage was struck at Antioch in tandem Hoover 2007, 282-284.and 300-301 with Table 8. Cic. Ver. 4.61; Jos. AJ 13.419; Hoover 2007, 296-297. 19 Houghton, Lorber & Hoover 2008, nos. 2427 (bronze) and 2434 (tetradrachm); Hoover 2007, 295; Newell 1917-1918, 112. 20 It is this very tendency that allows us to reconstruct and date the reigns of Antiochus VIII and Antiochus IX, as noted above. 18 17 QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it 258 OLIVER D. HOOVER with the ‡ tetradrachms, yet bronzes are known sharing the controls of Antiochus X’s tetradrachm series (figs. 14-15). 21 Fig. 14 Fig. 15 Fig. 16 Since Antioch began to produce a dated civic bronze coinage beginning in 92/1 BC (fig. 16), the most reasonable conclusion is that the © series represents the first reign coinage of Antiochus X while the ‡ tetradrachms were struck during his second reign. 22 No royal bronzes were associated with the ‡ tetradrachms because they were not necessary. The Antiochene civic bronzes simply took the place of the earlier royal bronze coins struck at the mint of Antioch. It is tempting to speculate on a possible increased civic involvement in the production of the ‡ tetradrachms as the monogram on these coins includes elements of the city ethnic. Reign Antiochus X Antiochus XI Antiochus X Demetrius III Dates 94/3 c. 93 93/2-89/8 88/7 n 35 11 132 9 d 8 2 30 1 d1 0 0 7 1 Cest 1 1 0.947 0.889 Dest 8 2 35.4 1.7 Confidence Interval 6.4-10 1.4-2.9 31.2-40.2 1.1-2.5 Annual Average 6-10 2-3 6-8 1-3 Table 5. - Undated Tetradrachms of Antiochus X, Antiochus XI, and Demetrius III at Antioch per Hoover 2007. Reign Antiochus X Antiochus XI Antiochus X Demetrius III Dates 94/3 c. 93 93/2-89/8 88/7 n 132 11 35 9 d 30 2 8 1 d1 7 0 0 1 Cest 0.947 1 1 0.889 Dest 35.4 2 8 1.7 Confidence Interval 31.2-40.2 1.4-2.9 6.4-10 1.1-2.5 Annual Average 31-40 2-3 1-2 1-3 Table 6. - Undated Tetradrachms of Antiochus X, Antiochus XI, and Demetrius III at Antioch Revised . If the © series does indeed represent the first reign coinage of Antiochus X, it becomes difficult to understand how an estimated 31-40 dies could have been used in the single year 94/3 BC, even bearing in mind the incomplete coverage of the sample for this series. The difficulty is compounded when we consider that for part 21 22 Houghton, Lorber & Hoover 2008, nos. 2432-2433. For this civic bronze series, see Butcher 2004, 307-312. QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it TIME IS MONEY? 259 of this year Antiochus X was driven out of Antioch by his cousin Antiochus XI. The opposite problem is posed by the average use of 1-2 dies per year during the second reign of Antiochus X. In short, there now seem to be too many dies for the first reign and too few for the second reign to fit the rhythm of production identified for the preceding reigns of the later Seleucid period. A Numismatic Explanation? There are really only two ways to rationalize the revised die estimates for the reigns of Antiochus X given in Table 6. Either the intervening reign of Antiochus XI took place much later (c. 89 BC?), leaving only a brief period for the second reign of Antiochus X, or the die sample for Antiochus X is somehow skewed, resulting in a disproportionately low estimate for the second reign. The first possibility seems unlikely on the basis of Eusebius, who reports that only Philip I, the brother of Antiochus XI, and Antiochus X remained to contest the throne by the end of 94/3 BC, 24 and the existence of royal bronze coinage struck in tandem with the silver issues of Antiochus XI. 25 These royal bronzes do not make sense if they are dated much later than 93 BC, since the civic bronzes of Antioch begin in 92/1 BC and it has already been shown that these coincide with the second reign ‡ tetradrachms of Antiochus X. Since the chronographic tradition and the numismatic evidence appear to preclude drastically modifying the lengths of Antiochus X’s two reigns, there is little choice but to explore the possibility that the die sample has been skewed. A likely occasion for any compromise of the die sample is the massive tetradrachm recoinage undertaken by Philip I at Antioch immediately on the heels of Antiochus X’s second reign and an ephemeral reign of Philip’s brother, Demetrius III, probably in c. 88/7 BC. 26 After years of ruinous war, Philip needed to refill the royal coffers in order to fund the restoration of the greatly diminished Seleucid state and to prepare for the day when he would be faced with another rival for the throne. His younger brother Antiochus XII was a potential threat at Damascus until 84/3 as shown by a new tetradrachm struck at that city, although there is no evidence that he had any ambition to take northern Syria from Philip. 27 This is the sequence that was ultimately preferred in Houghton, Lorber & Hoover 2008. Eus. Chron. (Karst ed.), 207. 25 Houghton, Lorber & Hoover 2008, nos. 2442-2443. 26 For the date, see Hoover 2007, 294-295. 27 Houghton, Lorber & Hoover, no. 2472A (unknown before 2008) is dated SE 230 (83/2 BC). For discussion, see Hoover, Houghton & Veselý 2009, 316. Antiochus XII is entirely ignored by the chronographic and historiographic traditions informed by Antiochene sources. 24 23 QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it 260 OLIVER D. HOOVER Likewise, Cleopatra Selene and Antiochus XIII, the respective wife and son of Antiochus X, still lurked somewhere in Cilicia or Coele Syria, waiting for an opportunity to claim the Seleucid throne. 28 Philip appears to have melted down the Antiochene tetradrachms of his predecessors that were struck to a reduced Attic standard of c. 16.00g and reissued them with his own types and a lighter weight of c.15.65g. By so doing, Philip’s treasury profited by about half an obol on the new light system every time an older tetradrachm was recoined. 29 The recoinage may have been undertaken with some sense of urgency and an eye to economy is suggested by the apparent overstriking of earlier issues and the use of at least one recut die. One obverse die known from two coin specimens seems to have the royal portrait recut on a non-Seleucid die (figs. 17-18). 30 Fig. 17 Fig. 18 To date, the identity of the original type in this die has eluded identification, but it may perhaps be the turreted head of Tyche like that found on various Syrian civic tetradrachms of the first century BC (fig. 19). 31 Three projections similar to the turrets seem to be present above the portrait of Philip I. Fig. 19 28 Jos. AJ 13.420; App. Syr. 70; Houghton, Lorber & Hoover 2008, vol. 1, 615-616 with nos. 24842486; Hoover 2007, 297; Ehling 2008, 241-242; Bellinger 1949, 82. 29 Houghton, Lorber & Hoover 2008, vol. 1, 596 30 Both coins are illustrated here to show that the previous type was in the die and not an undertype on a restruck host coin. 31 For a review of the production, metrology, and circulation the Syrian civic tetradrachms of Seleucia in Pieria, Laodicea by the Sea, and Aradus, all of which featured a turreted bust of Tyche on the obverse, see de Callataÿ 2002. QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it TIME IS MONEY? 261 The use of various flan forms (broad and thin or thick and dumpy) indicates that Philip reused earlier coins as flans rather than melting them down to create new flans of a more uniform style. 32 Presumably the types of the original coins were filed down to reduce the weight to the new lower standard. Such processing would account for the lower weight standard, as well as the absence of any identifiable undertypes or edge alteration on Philip’s coins. If Philip’s recoinage was indeed capable of skewing die samples of earlier rulers to the point of affecting the mathematical estimation of the total number of original dies, we might expect the second reign issues of Antiochus X to be the most severely compromised. These coins would have entered circulation or the royal treasury most recently and would have been most easily recalled and recoined. The very small samples for the coins of Antiochus XI and especially Demetrius III (figs. 20-21) may also have been heavily skewed by the recoinage, although the textual evidence suggests very brief reigns for these kings at Antioch to begin with. 33 Fig. 20 Fig. 21 The effect of Philip’s recoinage is noticeable in the hoard record. Up until the death of Antiochus IX, Syrian hoards regularly included a variety of Attic-weight emissions struck by foreign cities and kings as well as issues of several different Seleucid monarchs. 34 However, this pattern grinds to a halt in the hoards buried during the reign of Philip I. With the exception of the Kamishliyye 1980 hoard, which was probably closed very shortly after Philip began his reign in Antioch, and Akkar 1951, which may also be early notwithstanding an intrusive Parthian tetradrachm of Phraates IV, 35 all known Syrian hoards with Philip I components lack the coins of earlier rulers. One might argue that earlier coins were largely Houghton, Lorber & Hoover 2008, vol. 1, 596. Hoover 2007, 289 and 294-295. 34 This is confirmed by a review of the Syrian hoards listed in Houghton & Lorber 2002, vol. 2, Appendix 3 and Houghton, Lorber & Hoover 2008, vol. 2, Appendix 3. 35 For this coin as a probable intrusion, see Houghton, Lorber & Hoover 2008, vol. 2, 101 contra Seyrig 1973, 107. 33 32 QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it 262 Hoard Kamishliyye 1980 (CH VIII, 515) Akkar 1951 (IGCH 1583) Commerce 2007 (SC 2) OLIVER D. HOOVER Contents Antiochus XI, Demetrius III, Philip I Seleucus VI (2), Antiochus X (9), Philip I (4), Philip (126), Antiochus XIII (1) Table 7. - Mixed Syrian Philip I Hoards. Hoard Aleppo 1940 (IGCH 1575) Syria 1980 (SC 2) Commerce 1991 (SC 2) Commerce 2003 (SC 2) Seleucia in Pieria 1932-1939 (IGCH 1573) Aleppo environs 1950 (IGCH 1578) Syria 1991 (SC 2) Bâb environs 1951 (CH IX, 589) Contents 20 lifetime 53+ lifetime 16 lifetime 10 lifetime 4 AE cores (1 lifetime, 3 Roman imitations) 23 (1 lifetime, 22 Roman imitations) 5 (1 lifetime, 4 Roman imitations) Philip I (26 lifetime, 50 Roman imitations) Table 8. - Syrian Philip I Hoards. driven out by the new light standard and Gresham’s Law, but then we should expect to find Syrian hoards of earlier Antiochene issues closing with the coins of Antiochus X, Antiochus XI, or Demetrius III. Such hoards are entirely unreported in the literature, probably because Philip’s recoinage was so thorough that only a handful of Antiochus XI, Antiochus X second reign, and Demetrius III tetradrachms escaped recoining within the borders of Philip’s kingdom. If Philip’s recoinage and lower weight standard had the effect of erasing the preceding issues of Antiochus XI, Antiochus X, and Demetrius III from circulation in Syria, this would help to explain the proclivity of the Roman governors of Syria for imitating his types for the provincial silver beginning with Aulus Gabinius (5755 BC) (fig. 22). 36 Fig. 22 A political explanation for the latter development is sometimes sought, with the claim made that following Pompey’s removal of Antiochus XIII in 64 BC, Philip was seen by the Romans as the last legitimate Seleucid king. 37 Because of this supposed special status, his types were therefore the most appropriate for the 36 For the imitative types, see Houghton, Lorber & Hoover 2008, nos. 2487-2491; McAlee 2007, nos. 1-28; Burnett, Amandry & Ripollès 1998, nos. 4124-4149. 37 Burnett, Amandry & Ripollès 1998, 606. QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it TIME IS MONEY? 263 Roman governors to imitate. While this is a possible motive for the imitation of Philip’s coinage, it is far more likely that it seemed appropriate to copy because there was so much of it around and other possible late Seleucid models had been destroyed by the recoinage. 38 In short, the Roman decision to imitate Philip’s types in Syria probably hinged more on economic, rather than political factors, much as had the earlier decision to continue the types and standard of the royal cistophoric coinage after the Attalid kingdom was bequeathed to Rome in 133 BC. Randomness in Recoinage The idea that a major recoinage could skew a sample in the manner we have proposed above and adversely affect the estimated total number of dies is likely to draw criticism on the grounds that the Esty formulae used to make the estimations take into account random loss. 39 While this is certainly a just criticism, it is based on the assumption that the Seleucid state lacked the mechanisms that could effect a successful general recall of a coinage and therefore die evidence destroyed by recoining as well as the coins that escaped recall can only be considered random. If the die sample derived from the earlier coins returned to the mint for recoining with Philip’s types can only be random, any missing dies should be accounted for by the formulae. However, the assumption of randomness may be doubted in light of the surviving tetradrachm coinages of the Seleucid usurpers, Molon (222-220 BC), Achaeus (220-214 BC) and Timarchus (163-160 BC) (figs. 23-25). Fig. 23 Fig. 24 Fig. 25 For the same basic conclusion, see also Baldus 1987, 127; Newell 1919, 80-84. My thanks to Michel Amandry and Warren Esty for raising this critical point when this paper was first presented in Rome. 39 38 QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it 264 OLIVER D. HOOVER The sample for Molon consists of a single specimen, while those of Achaeus and Timarchus involve four and three coins, respectively. 40 There can be little doubt that these coins exist in such small numbers because they were aggressively recalled and recoined by the legitimate kings, Antiochus III and Demetrius I. This is certainly the case for Timarchus, as his original types are sometimes still visible beneath the new types of Demetrius I struck over them. 41 These examples show that the Seleucid state was very capable of recalling undesirable money with an almost total success rate, thereby rendering the samples for Molon, Achaeus, and Timarchus completely useless for estimating their original die usage. This gives the impression of a systematic and remarkably thorough recall of these coins from circulation rather than of a haphazard and random removal of the offending issues. It is not very difficult to imagine how such a systematic recall could have taken place. If it had been announced that the coins of these usurpers would cease to be accepted (or suffer a reduction in value) in payments to the state after a certain date, but that holders of the coins could redeem them for acceptable currency, few individuals are likely to have held onto their Achaeus or Timarchus tetradrachms instead of returning them to the mint for new Antiochus III or Demetrius I coins. The only coins likely to have escaped a general recoinage of this type are those that may have been hoarded prior to the announcement of their demonetization (or value reduction) or that may have traveled beyond the borders of the Seleucid kingdom. It is difficult to know whether such escapees represent a random sample since they may have come from batches shipped outside of the kingdom as state-to-state payments or may represent fresh coins hoarded before entry into circulation. If the recall of older coinage and its recoining by Philip I was as systematic as it seems to have been under Antiochus III and Demetrius I, and if the small surviving samples for Antiochus XI, Antiochus X (second reign), and Demetrius III are not random, then one might expect to obtain artificially low die estimates from the Esty formula. 42 Conclusion The critical question of randomness in the samples of earlier coins surviving Philip’s recoinage must remain unanswerable since the coins come from disparate sources and lack their original hoard provenances. Nevertheless, there seems to be sufficient evidence to support the possibility that the samples might not be com40 Houghton and Lorber 2002, vol. 1, 343 and no. 950 (Molon) and 348 with no. 359 (Achaeus); Houghton, Lorber & Hoover 2008, vol. 2, 142 and nos. 1588-1589 (Timarchus); Fischer 1988, 15, no. 4; Houghton, 214. 41 Houghton, Lorber & Hoover 2008, vol. 2, 209 and 216-220 with nos. 3-4. 42 Esty 2006, 362. QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it TIME IS MONEY? 265 pletely random and therefore yield inaccurately low die estimates. If this is indeed the case, and if the monogram sequence for Antiochus X is correctly understood as © followed by , then it is no longer possible to argue for die estimates as a potential chronological tool, at least in the context of late Seleucid coinage. The case of Antiochus X tends to suggest that there is no way to establish annual usage with relative security since a major recoinage may have had a severe impact on the samples available for earlier monarchs. This conclusion is troubling regardless of how it is received. If it is accepted as correct, or at least plausible, then it must raise questions about the basic validity of the quantitative method for developing models for understanding ancient coin production and monetized economies. On the other hand, if it is erroneous, all reasonable avenues for explaining the very low estimated die usage for the second reign of Antiochus X appear to be exhausted. Bibliography G. G. Aperghis 2001, Population-Production-Taxation-Coinage: A Model for the Seleukid Economy in Z. Archibald et al. (eds.), Hellenistic Economies, London-New York, 69102. G. G. Aperghis 2003, The Seleukid Royal Economy, Cambridge. H. R. Baldus 1987, Syria, in A. M. Burnett and M. H. Crawford (eds.), The Coinage of the Roman World in the Late Republic, Oxford, 121-151. A. R. Bellinger 1949, The End of the Seleucids, Transactions of the Connecticut Academy of Arts and Sciences, 38, 51-102. A. Burnett, M. Amandry, and P. P. Ripollès, Roman Provincial Coinage Volume 1: From the Death of Caesar to the Death of Vitellius (44 BC-AD 69), London/Paris. K. Butcher 2004, Coinage in Roman Syria, 64 BC-AD 235, London. Fr. de Callataÿ 2002, La production des tétradrachmes civiques de la Cilicie jusqu’à la Palestine à la fin du IIe siècle et dans la premiere motie de Ier siècle av. J.-C., in C. Augé and F. Duyrat (eds.), Les monnayages syriens. Quel apport pour l’histoire du Proche-Orient hellénistique et romain? (Bibliothèque archéologique et historique 162), Beirut, 71-91. K. Ehling 2008, Untersuchungen zur Geschichte der späten Seleukiden (164-63 v. Chr.) (Historia Einzelschriften 196), Stuttgart. W. W. Esty 2006, How to Estimate the Original Number of Dies and the Coverage of a Sample, NC, 166, 359-364. Th. Fischer 1988, Molon und seine Münze (222-220 v. Chr.)(Kleine Hefte der Münzsammlung der Ruhr-Universitäts Bochum 9), Bochum. A. Houghton 1979, Timarchus as King in Babylonia, RN, 21, 213-217. A. Houghton and C. Lorber 2002, Seleucid Coins, Part I: Seleucus I through Antiochus III, 2 vols., New York. A. Houghton, C. Lorber, and O. Hoover 2008, Seleucid Coins, Part II: Seleucus IV through Antiochus XII, 2 vols., New York. O. Hoover 2007, A Revised Chronology for the Late Seleucids at Antioch (121/0-64 BC), Historia, 56 (3), 280-301. QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it 266 OLIVER D. HOOVER O. Hoover, A. Houghton, and P. Veselý 2008, The Silver Mint of Damascus under Demetrius III and Antiochus XII (97/6 -83/2 BC), AJN, 20, 305-336. G. Le Rider and Fr. de Callataÿ 2006, Les Séleucides et les Ptolémées. L’héritage monétaire et financier d’Alexandre le Grand, Paris. R. McAlee 2007, The Coins of Roman Antioch, Lancaster/London. Y. T. Nercessian 2006, Silver Coinage of the Artaxiad Dynasty of Armenia, Los Angeles. E. T. Newell 1917-1918, The Seleucid Mint of Antioch, AJN, 51, 1-151. E. T. Newell 1919, The Pre-Imperial Coinage of Roman Antioch, NC, 79, 69-113. H. Seyrig 1973, Trésors du Levant anciens et nouveaux, Paris. R. J. van der Spek 1997-1998, New Evidence from the Babylonian Astronomical Diaries concerning Seleucid and Arsacid History, Archiv für Orientforschung, 44-45, 167-175. QUANTIFYING MONETARY SUPPLIES IN GRECO-ROMAN TIMES - © 2011 · Edipuglia s.r.l. - www.edipuglia.it