Chapter 7—
Toward a Delian Economic History

Summarizing the economic relations between the temple of Apollo and the city of the Delians in the fifth and fourth centuries B.C. , Jacques Coupry wrote: "A great sanctuary tends only secondarily to be an economic power as such, whether nationally or internationally."^[1] This judgment applies with equal force to independent Delos. The Delian economy, as we have seen, depended largely on local suppliers to satisfy its demands, and Delos could best be characterized as a center of the local Kykladic economy, extending to Karystos on Euboia, and sometimes fed by Rhodos or its allies.

The distinction between temple and island is not always easy to draw, however. The temple, which provides the economic data I have used to illuminate an economy confined largely to Delos's immediate Kykladic neighborhood, had international connections too. It dealt with other religious centers elsewhere in the Greek world; received embassies from kings and cities; enjoyed the patronage of wealthy public and private donors, who adorned its treasuries with gifts and its temenos with buildings; displayed treaties and decrees from many cities; hosted meetings of the Island League and the bivouacs of armies. The international scope of Delian Apollo's religious and political reach may obscure the very limited range of his economic authority, but it is crucial to distinguish between the two to assess the character of the Delian economy and to trace its development across the years of independence.

The History of Delian Economy

Despite many gaps and uncertainties, three separate periods of Delian economic history can be identified. Different kinds of economic activity, levels

of price, and apparent strategies of investment characterize each. To some degree they correspond with the periods distinguished by Fritz Heichelheim, J. A. O. Larsen, and Michael Rostovtzeff,^[2] but their characters and the explanations for them are quite different; and, of course, I make no claim for the validity of this "economic history" for any of the Greek world outside of Delos and its immediate neighbors.

The three periods are (1) the early years of independence (314–290 B.C. ), characterized by high estate rents and declining oil prices; (2) a period of stability down to the late third century, when commodity prices were generally level, estate prices gradually declining, and house prices roughly steady (272–212 B.C. ); and (3) a period from the late third century to 167 B.C. , when firewood and pig prices rose to a new plateau, oil remained steady, estates except for Group I continued to fall, and house rents rose appreciably (207–167 B.C. ). This last period seems to have been one of readjustment in the economy during a time of growth and change; perhaps misleadingly, I call it the "new prosperity."

Period I (314–290 B.C. ):
Uncertainty and Adjustment

The removal of Delos from the Athenian orbit and the creation of the Island League seem to have spelled disruption for certain segments of the local economy, which needed years to recover and adjust. Athenian citizens had played an important role in the economy of the temple as renters of the sacred estates and houses, and perhaps (as I have suggested) as suppliers of the island's demand for olive oil. Once the Athenians were gone, locals competed vigorously for the estates. The competition to invest in landed property brought Apollo unexpected gains, but also unwelcome risks. Defaults climbed; renters' property did not always suffice to cover the losses. The situation must have provoked considerable debate among the Delian elites, who both profited from the business and, as temple administrators, felt responsible for protecting the god's patrimony. By 301 or 300 B.C. , a compromise had been worked out and embodied in the hiera syngraphe. Henceforth estate rentals would be tightly controlled, and defaulters treated severely; a series of defaults and confiscations of property showed how serious these regulations were. Because the new regulations effectively debarred potential renters who could not find guarantors or whose personal wealth could not weather a default or a few bad crops, only the highest levels of Delian society—and a handful of Kykladic islanders who had connections among the Delian landholding elites—had any hope of renting the estates. The immediate results were a drastic decline in rents in

290 B.C. and a tendency for estates to remain in the same hands or the same families for years afterward.

The response of estate rents to this change depended radically on the products the estates produced. Group I estates, which required the heaviest investment from the renter because of their lack of vines, suffered the severest reduction in rents. Those estates that had vines, and therefore demanded a more modest commitment of funds up front from renters, weathered the transition better; the estates of Group IIA, which may predominantly have been vineyards, underwent only a very moderate decline. Presumably the predictable demand for wine coupled with low initial investment continued to make these properties an attractive investment.

Olive oil also showed a continuous, steady decline in prices from the late fourth century down to ca. 270 B.C. I have suggested that a dependence in the fourth-century amphiktyonic period on Athens for oil may have given way after 314 to imports from Rhodos, and that the desire to have local sources encouraged plantings on Delos and neighboring islands, which resulted finally in substantial, reliable local supplies and a steady, low price. The facts fit this view well, especially in view of the correlation between the great siege of Rhodos (305–304 B.C. ) and extraordinarily high oil prices in 304 B.C. This interpretation also accounts in a satisfying way for the change in numbers of Rhodian amphorae recovered on Delos (always assuming that some of those amphorae contained oil). In contrast, the traditional view, which attributes high prices to the new demand of the Greek East for home products, looks much less appealing, especially because of its difficulty in explaining the low oil prices of the last months of 304 B.C.

Other sectors of the economy displayed little change. To the extent that pig and firewood prices can be traced, they do not seem to show levels much different from those common twenty or thirty years later. The exceptions, like the high pig prices of 302 B.C. , were linked to military operations in the islands. Citizens of Kykladic islands worked on Delos as contractors and laborers, imported goods, rented property, and borrowed money. Apollo lent funds (and rued it) to members of the Island League and to some few small neighbors just outside the Kykladic orbit,^[3] but Delian economic influence faded away at the boundaries of the Kyklades.

This reconstruction implies some interesting things about the Delian economy in these years. For one, rents for the agricultural estates display an ambiguous tie with local agricultural prices. We have seen that rents for Group I estates, and to a lesser extent Groups IIB and IIC, correlated well

in later periods with barley and olive oil prices. Although the data are too exiguous to permit analysis for this period, it is an apparently reasonable supposition that the high oil prices of this first period encouraged renters to offer the high rent bids to which (I have argued) social considerations impelled them too. Yet none of the estates had olives, and neither did renters plant them. The failure of renters to do so is unsurprising: olives need seven to twenty years to produce their first crop, whereas rental periods in 314–300 B.C. were never longer than five years. Moreover, the great collapse in estate rents in 290 B.C. had no connection with oil prices, which underwent no comparable radical, discontinuous change. This shows quite clearly the operation of extra-economic forces, or at least forces that did not affect oil prices. It can be argued that the instability of this sector of the economy resulted from a partial decoupling of estate rent levels from their normal indicators among agricultural commodities.

The Delian evidence also casts new doubt on long-standing claims about generally high prices in the Greek world in the late fourth century. The facts that prices of individual goods did not move together and that rents for estates rose at least partly independently of prices and of each other undermine any attempt to attribute these "high prices" uniformly to inflation brought on by the influx of eastern wealth or other such causes. Moreover, many of the "data" from other parts of the Greek world are equally suspect. The evidence from Athens, collected long ago by William Ferguson and Rostovtzeff, does not hold up under close examination. "Prices" quoted by Theophrastos (Char. 3.3, 23.5, 8) have no claim to represent market prices, and the enormous dowries attested by New Comedy have been deflated by M. I. Finley's careful study of the Athenian horoi.^[4] This whole subject requires reexamination.

These first few decades of Delian independence also saw the evolution of the Kyklades into an economic unit focused on Delos. While the archipelago certainly always formed something of a region, during the fourth century many of the islands enjoyed important relations with the outside world independent of their links to Delos. The western islands in particular cultivated relations with western neighbors, including the cities of Euboia and Athens. Especially before 350 B.C. , Athens played an important role, running the second Athenian sea league, to which most Kykladic states belonged, and controlling the temple of Delian Apollo. Sometimes allies

like the Andrians participated in this Athenian control, and individual citizens of other Kykladic states also sometimes benefited from it. The establishment of the Nesiotic League in 314 B.C. , however, entailed a marked reorientation of the Kyklades toward their center at Delos and away from the outside world. The political and military relations of the islands with their hegemones —first Antigonos Monophthalmos and Demetrios, then the Ptolemies—were mediated through the structures of the league and through the royal officials, such as the nesiarkhoi, the oikonomoi, and Philokles, king of the Sidonians, put in charge of it. The league knit the islands into a unit in a way that had not been seen before. Furthermore, the removal of Athens from the central Aegean seems to have opened economic opportunities for the islanders on Delos. The result was the creation of an economic region for which Delos became the natural exchange center through a combination of geographic centrality, religious magnetism, and housing the apparatus of the Nesiotic League.^[5]

Period II (CA. 290–230 B.C. ):
The Steady Economy

By the beginning of this period—poorly defined as it is—the Delian economy began to achieve a balance that would persist for about half a century. Commodity prices were quite stable over the long run. Estate rents fluctuated, but in fact all four groups recovered after the debacle of 290 B.C. , with rents in 269 B.C. higher than, and in 249 B.C. about the same as, those of 290 B.C. Group II estates, distinguished by their vineyards, benefited particularly from this recovery, a phenomenon I have attributed to the insularity of the Delian economy, which created a good market for local vintages. Except for the very unusual rental period of 247 B.C. , house rents remained roughly steady.

The economy continued to be largely a local matter. Renters were Delians or their Kykladic neighbors; imports (where attested) came from nearby sources. Some years the Delian government intervened in the purchase of grain during the winter, but nothing indicates that this was anything but an irregular practice. Visitors and passing merchants were mainly from nearby islands. The economy felt the impact of military operations in the neighborhood—for instance, the naval battle near Andros in 246 B.C. —but was insulated from more distant disruptions. There was some transit trade in grain (IG XI 4.1049), but the amount does not seem likely to have been great, given the evidence we have for the level of port activity in the second quarter of the third century (see below), and the

unidentified city importing grain, as I have argued in chapter 4, was almost certainly a close Kykladic neighbor.

The accounts for 279, 278, and 250 B.C. give a measure of the port activity on Delos during this period. In 279 a portion of the 2 percent harbor tax, which was regularly farmed out, was paid to Apollo to cover debts the city owed. The amount was 14,910 dr, of which 5 percent was the   or "sales tax" charged on the 2 percent tax; the net 2 percent tax collected therefore came to 14,164 dr 3 ob. The payments for the next year totaled 18,800 dr, or 17,860 net. In 250 B.C. a tax called the  , which Vial interprets as the portion of the 2 percent tax on goods coming into or going out of the city, brought in 5,250 dr, or 4,987 dr 3 ob net. Vial concludes from these figures that two-thirds of the tax was collected on goods in transit and that the whole value of the farmed-out tax is represented.^[6]

These inferences are far from certain. The payments of 279 and 278 were made to cover debts—probably loans—the city owed Apollo. Public loans were frequently secured on public income,^[7] but nothing assures that the total amount collected went toward the debt. The actual income might have been higher. Nor is the identity of the   secure. It may, as Vial thinks, represent the 2 percent tax on the portion of the goods that came into or out of Delos itself, as opposed to purely transit goods, but it might also be an alternative name for the general 2 percent tax. The accounts are not consistent in reporting taxes and their names. While a grain tax ( ) and a tax on rents ( ) are reported in both 279 and 250, the   appears only in 279. A tax called the  , which was perhaps a "fishing rights tax," was recorded in 279 and 278, but a 10 percent tax on fish ( ) in 250 (IG XI 2.161A26–27, 287A9–10). These two might be the same taxes under different names or different taxes imposed at different times. The law regulating the sale of wood products (ID 509 = SIG³ 975), which was passed probably in the 230s, makes no provision for separate recording of the pentekoste on wood sold in the city (11. 13–14:   [i.e.,  

 ). (Of course, the pentekostologoi may have known to register imports and transit goods separately.) The Delians reformed the administration of temple business several times during independence, and they may have done the same for city finances.^[8] Even if Vial is right, the transit:import ratio must have varied considerably from year to year (figures for 279 and 278 B.C. prove that), and although it would in the long run have fluctuated around some typical value (assuming no sustained growth or decline), there is no warrant that the ratio of two to one for goods in transit to imports that Vial derives was typical.^[9]

Moreover, if the figures for the   do represent payments on goods in transit, then each unit of goods would have paid the tax twice, and the actual value of goods in transit would have been twenty-five, not fifty, times the amount of tax collected on transit goods. If Vial has inferred correctly that the tax value of goods imported into the city—thus paying the   only once—amounted to one-third of the tax value of transit goods (paying the tax twice), then the total value of goods in transit would equal the tax paid divided by 0.06.^[10] On this calculation, Delos saw 236,075 dr (= 39 talents) of transit goods in 279 B.C. and 297,667 dr (= 50 talents) in 278 B.C. If Vial's results are rejected, and the full value of the tax is treated as deriving from transit trade, the figures are 354,112 dr (= 59 talents) and 446,500 dr (= 74 talents). Finally, if the full value of the tax is treated as deriving from imports, the results are 708,225 dr (= 118 talents) and 893,000 dr (= 149 talents). The same considerations for 250 B.C. yield figures (for the transit trade) of 247,875 dr (= 41 talents) or 123,937 dr (= 21 talents), or, taking the full figure as imported goods, 232,875 dr (= 39 talents). Since however some of the goods coming to Delos must have been consumed there, some portion of the total tax collected must have been on those imported goods. If we accept Vial's distinction between the simple   and the   as the best explanation of these terms, the total annual value of goods in transit on Delos would have remained fairly stable across the first half of the third century—roughly 39 talents in 279, 50 talents in 278, and 41 talents in

250 B.C. Like all of our calculations, these must be taken with a grain of salt.

We must appreciate what a piddling trade these figures represent. Kaunos and Stratonikeia brought Rhodos 120 talents each in tribute in the second century. We do not know how these payments were calculated, but on the reasonable assumption that they cannot have been more than about 10 percent of total economic activity, these two cities each enjoyed a "gross city product" 24–40 times greater than that of Delos. Delos practically vanishes compared to truly great economic centers. For Athens, Andokides reports an Athenian pentekoste of 33 talents in 402/1 B.C. and just over 36 talents the next year, representing a transit trade worth perhaps 550 and 600 talents.^[11] In 165/4 the famous embassy the Rhodians sent to complain about the creation of a free port at Delos claimed that before 167 B.C. their harbor dues used to fetch a million drachmas, representing 25 million dr worth of goods in transit alone (= 4,167 talents), or 83–106 times the Delian income.^[12]

The other taxes recorded for these years are low. The   brought in 530 dr 279, 543 dr in 278; the  , 1,850 dr in 250; the 10 percent tax on grain, 120 dr in 279, 410 dr 3 ob in 278, 120 in 250; the  , 2,056 dr 4 ob in 279; the  , 600 dr in 279, 1,690 dr in 250. Each tax includes a surtax of 5 percent (the  ) totaling 62 dr 3 ob in 279, 47 dr 4.5 ob in 278, and 445 dr in 250.^[13]

These taxes give a sense of the level of economic activity on Delos in the second quarter of the third century. If the hypotropion really was a fishing rights tax, analogous to the funds collected from leasing porphyry fields, the full value of fish caught must have exceeded 600 dr in 279–278 B.C. .; by how much it is impossible to know. The 10 percent tax on fish from 250 B.C. represents the sale of fish worth 17,575 dr. At the mean rent for Apollo's houses in 250 B.C. of 55 dr per annum, the tax on rents that year of 16,055 dr represents about 290 properties, which may have housed 20–80 percent of the population.^[14]

The first two periods of the Hellenistic Delian economy were characterized by three features: the adjustments consequent on the liberation of the island from the Athenians, the stability of prices for goods (like wood and pigs locally available) that had not been in Athenian hands, and a generally low level of economic activity. The population, although large by Kykladic standards, was nevertheless relatively small in the scale of Greek cities. Local needs could be satisfied largely from local sources. Aside from the situation with olive oil in the two decades after 314 B.C. , there was certainly very little regular importing from farther away then nearby neighbors. Occasional dearths may have forced the Delians to rely on long-distance imports, although it must be admitted that there is virtually no evidence for such activity, in striking contrast to Athens, Samos, and many other similarly well-documented cities. Transit trade remained at a very low level even in comparison to inland cities like Stratonikeia, and such transit trade as there was certainly was moving through Delos from one Kykladic island to another. Delos's role as a pan-Hellenic sanctuary, visited and patronized by kings and wealthy citizens from all over the greater Greek world, had no noticeable impact on its economic life.

Period III (CA. 230–167 B.C. ):
The "New Prosperity"

From the 240s and especially the 230s, the economic scene on Delos began to show signs of change. On Delos itself, new residential building started after midcentury and seems really to have blossomed in the last third of the century in new construction north of the temple. Excavations there are not complete, and the material far from fully published, but the results that have appeared so far point consistently toward new growth. The sanctuary also shows new building, especially after 200 B.C.^[15]

The crucial period for this change was the last third of the third century B.C. A frustrating gap in the data, a result of the loss of all but a few

fragments of the accounts between ID 290 of 246 B.C. and ID 351 of 220 B.C. , obscures the character of the changes evident from a comparison of the evidence from either side, but it is nevertheless apparent that the Delian economy underwent important adjustments in those years. Around 220 B.C. , mean firewood and pig prices rose by about 20 percent. The rises in firewood and pig prices may be connected, and the rises in firewood prices are surely related to important new regulations governing the sale of wood and wood products on Delos. Both phenomena suggest a rising demand for wood, which would tie in nicely with new building. The import regulations suggest that the Delian government was eager to guarantee itself a full share of the higher profits.

Another set of evidence points to increased local transit trade. I have already reviewed the evidence for the grain trade, which increased in the last quarter of the third century. The permanent sitonia fund created by 209 B.C. may have reflected increased pressure on local supplies. The Delians also established a new commission to oversee renovation of the docks, spending almost 10,000 dr on the project between 217 and 171 B.C. Rents rose for sacred houses, most of which were probably devoted to commercial ends, and many of which were located in the harbor district. The first evidence for foreign financiers like Eutykhos of Khios and Athenodoros of Rhodos appears in this period. After 200 B.C. , foreign bankers become more prominent.^[16] The second century also saw an influx of new metics. Prominent among these were Phoenicians. In 187–175 B.C. , "the warehousers and shippers in Laodikeia in Phoenicia" ( ) erected on Delos a statue of Heliodoros,   and minister of Seleukos IV (IG XI 4.1114.4–5 = Choix, 72; cf. 1112–13 = Choix, 71). This inscription offers evidence, far more dependable than the proxeny decrees, of real activity on Delos by merchants going back and forth between Delos and the Levant. Nevertheless, we must be cautious. The numbers involved are impossible to calculate; even if M.-F. Baslez were right that proxeny decrees awarding only enktesis represent metics who had come to live on Delos, the number preserved would reflect only a very modest growth. Yet this growth, a result of the new attractiveness of Delos as a local distribution center for the Kyklades, might account in part for some of the changes seen after 220 B.C. , like the rise in rents for the sacred houses.^[17]

Apollo's estates also show important adjustments at this time. The rents

for Group I estates, producers of grain and livestock, rise from 220 B.C. on. This corresponds so nicely with the simultaneous rise in pig prices that it is tempting to postulate an increased local demand for animal products. The vine-growing estates of Group II suffer a real decline in the same period, which has been attributed to changes in the demand for and availability of various vintages. The introduction of cheaper and more desirable Koan wines diminished the demand for local Delian products, which were not only less popular but rarer and therefore more expensive. Rhodian wines may have figured also in the equation, given the rise in numbers of amphorae after ca. 220 B.C. Local Kykladic vintages may also have benefited, if the evidence for some increased production of amphorae on Paros and some other islands has been interpreted correctly.

The economic changes Delos underwent in the last third of the third century transformed it from a consumer dependent on its neighbors to something of a local transshipment point, through which much of the local Kykladic traffic passed. This new role would have increased the availability of these other vintages, and probably also, although at first marginally, of Rhodian and other wines as well. If these imported wines were preferred to local vintages, then the rents for estates dependent on wine may have declined without any general fall in wine prices. Indeed, the great volatility of agricultural prices from year to year—well attested, too, for Koan and Knidian wines bought for the Posideia, whose prices could fluctuate by 100 percent—should have masked any gradual decline, only the availability of either cheaper or preferable wines on Delos can account for the fading appeal of Apollo's estates.

To what should the new prosperity be attributed? It seems clear to me that the economic changes I have catalogued are not consistent with a simple rise in local population, whether owing to natural local increase—unlikely on other grounds^[18] —or to an influx of metics,^[19] because population growth ought to prevent any agricultural decline. Higher population should fuel the demand for olive oil, wine, and grain, guaranteeing at least stability in prices, if not a rise. Yet oil prices remained low and fairly stable, and the demand for local wine evidently fell off. On the other hand, live-

stock prices clearly rose, as attested directly by pig prices and indirectly by the rent histories of Group I estates. Unlike wine, sheep, goats, and pigs vary little in desirability according to their origin, and although they were often shipped by sea in antiquity, transportation of livestock was certainly more cumbersome and more expensive than that of wine.^[20] This fact may explain the continuing appeal of Group I estates. But why should prices for livestock have risen when those of other agricultural products did not?

The answer may lie simply in increased local prosperity. The consumption in antiquity of domestic livestock, whether as food or for sacrifice, depended very much on social standing and wealth. If there was a real increase in local wealth on Delos after ca. 240–220 B.C. , the demand for pigs and other livestock may have grown without a concomitant growth in the demand for oil and local wine, of which the Delians already consumed enough. A growth in the local transit trade could also account for the rising rents of warehouses Apollo owned, and increased local wealth could have prompted new construction—even of residences, as families that used to live in the harbor district moved away,^[21] or as the increasingly wealthy demanded bigger, finer homes.^[22] Both Claude Vial and Robin Osborne have argued that the Delian elite became more accepting of wealth generated through trade and business as opposed to agriculture. I would place the most important changes in their attitudes exactly in the period of the "new prosperity."^[23]

Local political changes seem to have accompanied this economic transformation. Vial has carefully examined the political careers of Delians. Be-

fore 240 B.C. , the authors of public decrees—the men cited in the   clause—came from many different groups: they were renters of estates, holders of office, hieropoioi. The number of different men attested as authors was high. After 230 B.C. , the pattern was completely different. Very few individuals moved most of the decrees, and no renters or hieropoioi appear among them. The phenomenon culminates with the remarkable Telemnestos son of Aristeides, who completely dominated Delian politics after 200 B.C. , no doubt thanks in part to his ties with Rhodos (see below).^[24] The economic changes of the second half of the third century thus entailed political changes as well. The rising wealth the economic evidence implies seems to have concentrated political power too. Decline in the value of most estates (those of Group II) perhaps contributed to a loss of influence for the people who rented them, a group that—as we saw in chapter 6—tended to be rather closed after 290 B.C. The stronger local transit economy, moving goods in and around the Kyklades, bringing in traders from neighboring islands to buy or sell grain or slaves or wood, put money into the accounts of Delians who participated in the business. It may well have been these people who in turn began to flex their political muscle, replacing the more traditional landed, or agriculturally based, aristocracy in the halls of the boule or at the bema of the ekklesia. With wealth outstripping their rivals, they tended to monopolize policy decisions.

This analysis in turn suggests, as I have already argued, that the change in the Delian economy was owing to the growth of the Delian port into a local transshipment center for the Kyklades. The first decades of this growth, which did not alter the embeddedness of the Delian economy in the local Kykladic economy, fell during a period when the Kyklades were largely free of outside domination. The Ptolemies were gone, the Antigonids were preoccupied in Greece and in the islands ringing the Saronic Gulf, and the Rhodians were kept busy, first, dealing with the impact of the great earthquake, then with fighting Byzantion and the Kretans. Emblematic of both the absence of the Rhodians (whose active presence was always signaled by a campaign against pirates) and the prosperity of the islands is the rise in piracy, which I have already discussed in chapter 2.

There is, however, another element to the role of piracy in the Kyklades: piracy contributed to the economy. Pirates did not simply raid, kidnap, and flee; Demetrios of Pharos, whose purpose was indeed to withdraw wealth from the archipelago, provides a poor model for the full local impact of piracy. In the first place, pirates undoubtedly recruited heavily from the

inhabitants of the islands. In early modern times, under the Tourkokratia, the Kyklades were notorious as nests of pirates. While we are virtually devoid of evidence for the origins of pirates, a handful of literary sources provide some hints. Alkiphron's letter of Eukolymbos to his wife Glauke, although obviously a literary production, nevertheless suggests the appeal of piracy to a poor fisherman who could hope to get gold and nice clothes without having to murder or stain his hands with bloodshed (1.8). Asklepiades' Samian prostitutes took names from types of boats often used by pirates.^[25] The pirates who raided Syros and Siphnos in the early first century were probably based on the little island off Siphnos where they took refuge with their captives (IG XII 5.653.28–29). Probably around 154 B.C. , Kretan pirates attacked Siphnos: they kidnapped people in the polis center and raided the temples (Diodoros 31.45).^[26]

When these pirates wanted to dispose of their booty or ransom captives, they turned to local economic centers. In the 230s B.C. , pirates who had captured citizens of Theangela in Asia Minor had recourse to Delos as a slave market, as we know from a fragmentary inscription in honor of the Delian Semos who ransomed the Theangelian captives from the block (IG XI 4.1054; cf. SEG 3.666). This sale need not, of course, imply a great "slave market" on Delos in the late third century comparable to the business Strabo describes in the wake of late-second-century Kilikian piracy (14.5.2, C668–669); there was plenty of local demand for slaves.^[27] The Delians went so far as to honor Boukris son of Daitas of Naupaktos as a proxenos of the temple and the Delians for long-lasting good will toward them; he was the notorious pirate who raided Attike, captured "many citizens and others from the city," and disposed of them in Krete.^[28] The king of

Sparta, Nabis, who figures in official Roman and Akhaian propaganda as a pirate who worked with Kretans (cf. Polyb. 13.8.2; Livy 33.44.8; 34.32.18, 35.9, 36.3), was a Delian proxenos (IG XI 4.716 = Choix, 58). Pierre Brulé has suggested that he and his Kretan friends sold booty on Delos; this is nothing but a guess, but on the face of it not an unreasonable one.^[29] Piracy was part of the economic revival of the last third of the third century; emblematic of this upswing are the raids of the outsider pirates Demetrios in 219 and Dikaiarkhos in 205 or 204 B.C. , both of whom found the archipelago satisfying hunting grounds.

This period of prosperity, which based on price rises must have begun between 250 and 230 B.C. , corresponds perfectly with the years of Kykladic independence. As we have seen, the great powers retreated from Delos after 245 B.C. , not to return until the Second Makedonian War. Of the many consequences the absence of overlords had, probably the most important was that wealth generated in the islands no longer migrated to royal coffers. Payment of tribute ended, or at least diminished (the Delians surely continued to award crowns from time to time to kings). Garrisons were withdrawn from some islands, although others still supported them. The wars of the second half of the century imposed far less frequently on the Kyklades, which probably contributed fewer soldiers, paid less money, and witnessed less fighting than they had under their former sovereigns.^[30] The result was a period of prosperity unparalleled earlier in the Hellenistic period.

The role of the Rhodians in this remains difficult to assess. Rhodian traders certainly plied the Kyklades in these years, although they remain frustratingly hard to identify.^[31] The striking rise in numbers of Rhodian amphorae on Delos after 220 B.C. or so attests explicitly to the import of

Rhodian products, whether or not they all were transported on Rhodian bottoms. Politically and militarily, however, the Rhodians maintained no permanent presence in the Kyklades until after 200 B.C. Their sweep against Demetrios of Pharos was just that, a sweep, mediated by larger political considerations. The First Kretan War, fought because of Kretan pirate raids against Rhodos and its immediate neighbors,^[32] probably played a more important role in turning Rhodian attention toward the central Aegean. Philip's aggression against Asia Minor, which threatened Rhodian possessions there, made them only too happy to join Attalos and the Romans against him. The Kyklades were their prize.

The Rhodians

In his discussion of the preliminaries of the great siege of Rhodos that earned Demetrios Poliorketes his ironic nickname, Diodoros of Sicily treats the close economic connection between the Rhodians and Ptolemaic Egypt; the Rhodians, Diodoros claims, provided the bottoms to export Egyptian grain and import other goods. The speaker of the Demosthenic speech Against Dionysiodoros imputes similar ties in the late fourth century when Kleomenes was in charge of Egypt.^[33] This tie with Egypt served as the foundation of the role Rhodos played in the economic life of the Hellenistic world.

But economic power (without being more specific about what that was) did not necessarily spell political influence, and neither can it be assumed that Rhodian political decisions were inevitably based on economic considerations. This point should be obvious, but modern commentators are fond of statements such as "It perhaps goes without saying that one mechanism of Rhodes' control [of the Kyklades in the second century] was its tremendous economic influence among the islands,"^[34] which in fact are very hard to justify. The role of the Rhodians in the Second Syrian War offers a very good example of the embarrassments that arise from equating political and economic interests. The Rhodians chose to fight against their long-standing Ptolemaic allies, whom they defeated ca. 258 in the battle of Ephesos. This defection from "their most important trading partner" has issued in theories of a Rhodian policy to avoid a "world-monarchy": Rhodian "policy makers were farsighted enough to realize that the predominance of

any one power, even Egypt, would almost certainly mean the end of the republic's autonomy and influence."^[35]

This sounds as if the Rhodians had anticipated the outcome of the war against Perseus by almost one hundred years. Moreover, the Rhodians had no trouble supporting the Ptolemies during their remarkably successful years from 287 to 267, when they liberated Athens, controlled the Kyklades, seized Samos and bases in Asia Minor, held Koile Syria, and fought the Seleukids in the First Syrian War. Surely, if the Ptolemies looked about to recreate a world empire in 260 B.C. , they had already started on that road by 279. Why did the Rhodians not act to stop them earlier? Or again, in the Third Syrian War, which started out looking like a sure defeat for the Seleukids, which would have delivered a huge land empire into Ptolemaic hands, the Rhodians stayed loyal. Given the assumption that economic considerations loomed large in Rhodian policy-making, choosing to oppose their premier trading partner would at any time have been exceedingly risky.

Once again, local political concerns are the better place to look for an explanation. Since 279, the Ptolemies had controlled numerous cities in Karia, Pamphylia, and elsewhere in littoral Asia Minor. The causes of the Second Syrian War remain very obscure, but it is clear at least that the revolt of Ptolemaios the Son resulted in the loss to the Ptolemies of many possessions in Asia Minor. When the war began, then, it looked, not as if the Ptolemies were about to recover Alexander's empire, but rather as if instability and unwelcome tyranny were spreading right in Rhodos's back yard, making the region ripe for exploitation. Antiokhos II had, of course, the same idea; by working together, he and the Rhodians could carve up former Ptolemaic possessions. This, it seems to me, provides a much better explanation of Rhodian opposition to the Ptolemies. It also, at least partly, decouples political and economic policy.^[36]

Our examination of Rhodian activities in the Kyklades reinforces this position. In the late fourth and early third centuries, the Rhodians stepped in to replace Athenian suppliers of olive oil to the Kyklades. Yet throughout the years of these activities, there is not the slightest indication of Rhodian attempts to obtain political jurisdiction over the archipelago. There is no sign of Rhodian political or military activity either in 311, when Polemaios revolted from Antigonos Monophthalmos, or in 308, when Ptole-

maios I took advantage of the confused situation, or in 307, when Demetrios came back through the islands, or in 288, when control of the archipelago was up in the air.

The same is true for the last two decades of the third century. Amphora finds on Delos, although not yet fully published, demonstrate a striking interest in Delos beginning after 220 B.C. From ca. 275 to 220 B.C. , an average of only 0.6 handles/yr has been found on Delos; for the years from ca. 220 to 175, the figure leaps to 9.9 handles/yr.^[37] Yet again, this economic interest, which may have been fueled largely by a demand for better-quality wine (see above), led to virtually no political involvement. Rhodian interventions against Demetrios and Dikaiarkhos (in the context of their First Kretan War) are their only attested activities in the Kyklades before the decisive events of the Second Makedonian War. In the one case, they were acting to expel an outsider whose presence had larger political and military implications; in the other, they were expanding military activity that had begun as a result of Kretan raids in the immediate territory of Rhodos, not in the Kyklades.^[38] These events may well have alerted the Rhodians that the central Aegean, a region close to their own shores, was essentially a political vacuum, absent any outside hegemon,^[39] and that the islands had become fairly prosperous since the last Rhodian incursion. But such a realization, if indeed it occurred, did not eventuate in action until the support of the Attalids and the Romans facilitated intervention in the archipelago after 200 B.C.

The Rhodian domination of 199–167 B.C. had real political consequences. The Rhodians exacted alliances, reactivated the Island League, compelled the islanders to serve in a Nesiotic fleet, stationed troops on Tenos, encouraged or pressured islanders to adopt Rhodian-style constitutions, entangled the islanders in the wars they fought, and probably supported local figures like Telemnestos son of Aristeides on Delos, who has been implicated in a pro-Rhodian policy during the 190s and 180s.^[40] But it would be a mistake to impute a clear economic aspect to this policy. Am-

phora handles tell part of the tale: 92 handles dated ca. 220–210 have been found, or an average deposition rate of 9.2/yr; for ca. 210–175 the figure is 10.0/yr, a change of no significance.^[41] Despite the wealth of epigraphical evidence from Rhodos, only fifteen Kykladic islanders are attested there, and several of them date to after the end of Rhodian control of the islands.^[42] There was neither a "federal" coinage nor any attempt to impose Rhodian money on the islands. Unlike in the Peraia or at Kaunos and Stratonikeia, the Rhodians did not demand tribute or other payments. The development of the Delian and Kykladic economies antedated the Rhodian presence, and the forces that drove them continued to operate into the second century. The real break in Delian development fell not under the Rhodians—although there can be no doubt that the numbers of outsiders grew in the second century—but after 167 B.C. , when Delos was made a free port to punish Rhodos, and especially after 146 B.C. , when the destruction of Korinthos brought Delos's trade to levels it had not before experienced.^[43]

Politics and the Economy

This review of the interests of the Rhodians in the Kyklades leads directly to consideration of the larger connections between politics and economics on Delos, in the Hellenistic Kyklades, and in the ancient Greek world generally. Certainly no single picture characterizes the relation. Consider the Ptolemies. On the one hand, some of their activities must have had an economic impact. The kings and their families and officials dedicated phialai, crowns, and other precious objects to the god; the offerings of the Ptolemies and their wives and relatives amounted to perhaps 1,000 dr in gold, or about two talents of silver.^[44] Festivals they founded (several Ptolemaieia and a Theuergesia) used dedicated capital to make loans whose interest covered the cost of annual dedication of phialai or other gifts. The Ptolemaic official Philokles and the nesiarkhos Hermias created foundations with capital of 6,000 and 3,300 dr. Bakkhon the nesiarkhos and Patroklos the admiral of the Khremonidean War donated precious objects. In

this they stood in a long tradition that included the Antigonid admiral Polykleitos, Androkles the Amathousian king, and the Spartan Lysandros.^[45]

On the other hand, there is virtually no evidence that the Ptolemies attempted to integrate the archipelago into their economic circuit. Trade between the islands and Egypt seems to have been limited to small shipments of high-value goods, like Rheneian and Kythnian cheeses, Kimolian fuller's earth, or a few amphorae of Parian wine. The Ptolemies did not try to bring the islands into their monetary orbit, as they did with Thrake and Koile Syria, and neither did they build any important structures on Delos.^[46]

Indeed, the strongest indicators of some kind of economic impact by the Ptolemies are negative. The passage of armies frequently stressed local systems of supply and disrupted prices and rents, although this was always transitory. More important were demands for tribute (however paid), which probably absorbed an appreciable, although unquantifiable, proportion of local wealth. Recent study of the territory of the Keian city of Koresia, renamed Arsinoë during the Khremonidean War, which was port to a Ptolemaic fleet, has documented the start of the polis's decline in exactly this period; in contrast to Thera, where the imposition of a garrison had positive effects, Koresia apparently suffered. We must be careful in our analysis. The evidence uncovered by work in Koresia is not well enough dated to prove that the beginning of Koresia's decline did not correspond with the withdrawal of the garrison rather than its introduction; but it is nevertheless certainly true that the akropoleis of the city show no new building in the third century, in strong contrast to Thera, with its extensive remains of the Ptolemaic fortress.^[47]

These effects, or rather their absence, can be attributed to the nature of Ptolemaic interest in the Kyklades. As I have repeatedly argued, the Kyklades provided the Ptolemies with the necessary stepping-stones for project-

ing their power into Greece. Their intrinsic appeal was low. "Investment," in the sense of reconstructing the local economy or putting lots of money into building projects, made little sense. The Nesiotic League and Philokles bore the administrative responsibilities, freeing the court in Alexandria to devote its attention to its genuinely important possessions.

These two examples, the Ptolemies and the Rhodians, nicely illustrate the complexities that arise in the study of the interconnections between political activity and economic effects. Strict political hegemony involving taxation, garrisoning, recruitment of soldiers, and military operations can subsist perfectly comfortably with a commodious disinterest in economic matters. That the Ptolemies, who, despite some recent revisionism on the issue, did care deeply about the economic operation of their own country, evinced no concern whatsoever about the same issues in the Kyklades, shows perfectly the disjunctions between politics and economy with which Hellenistic rulers could happily live. The Rhodians illustrate the same disjunctions. They cared profoundly about the condition of their possessions on the mainland of Asia Minor, but despite being involved far more intimately and immediately in the Kyklades than the Ptolemies ever were, they too did virtually nothing to affect the Kykladic economy: they simply exploited the system already in place.

I have stressed the principle of discontinuity between politics and economics, but in the case of the Kyklades it must be admitted that strictly local considerations surely went a long way to encourage the disjunction. The islands formed an economic unit, largely self-sufficient in foodstuffs (especially after ca. 270, when local olive oil production seems to have risen enough to satisfy local demand), building materials (except for wood for construction), and other basic needs. The system yielded a surplus sufficient to support a large "parasitic" population on Delos, which paid for its needs by providing religious services (which in turn brought in external wealth in the form of dedications) and by acting as a convenient central point of exchange, a role that expanded once the islands were permitted after 245 to retain the wealth outside hegemones had formerly appropriated. But the islands must have been operating near the bounds of their carrying capacity. They supported large populations themselves, on land already thoroughly exploited, as recent surveys have made more and more apparent. Squeezing more wealth out of them would have been a taxing undertaking, hardly worth the effort. The impetus for greater exploitation simply was not there. This situation conspired with the interests of the outsiders—and here I am thinking especially of the Ptolemies—to mitigate against any intimate involvement with the local economic system.

I do not mean to say, however, that political actions did not frequently

have unintended economic effects. The creation in the late fourth century of the Nesiotic League by Antigonos Monophthalmos, its central role among his son Demetrios's possessions after the battle of Ipsos, and the decision of the Ptolemies, once they had seized control of the archipelago, to retain it as their instrument of administration of the islands, played a determining role in defining the character of the regional Kykladic economy in the first half of the Hellenistic period. I do not however believe that either the Antigonids or the Ptolemies intended this result. For the hegemones of the Kyklades, the league simply provided a convenient mechanism for administering many small poleis and for disseminating the rulers' propaganda. Likewise, the powerful impact of military activity in the islands, whether through transient troop movements and fighting or through the establishment of permanent or semi-permanent garrisons, was a direct result of political decisions taken by the hegemones or their allies or enemies. But again, the economic effects of this activity on the Kyklades played no role in the decisions of whether to fight a war, or, on a smaller scale, whether to send troops through Delos or Tenos.

As a coda, it should be said that after 167 B.C. the rules of the game changed dramatically. When the Romans declared Delos a free port, they set in motion a process that ultimately decoupled Delos from the rest of the Kyklades.^[48] The destruction in 146 B.C. of Korinthos, which had been an important mustering point for traders moving east or west, accelerated the process. Foreign traders had begun to appear on Delos in the first third of the second century, attracted by the new local prosperity, but their numbers swelled in the last fifty years of the century. The great expansion of city and port belong to this period—and well into the first century, until Mithridates' invasion and the pirate attacks of the midcentury drove away almost everybody. The wealth of this period came from well outside of the Kyklades, which themselves seem to have suffered a serious decline.^[49] The flood of Italians, Phoenicians, and other traders devoted to moving goods west to Italy and uninterested in the local economic scene undoubtedly swamped the local traders who had traditionally moved goods around the archipelago. The loss of their central local entrepôt cannot explain the de-

cline of the islands attested from ca. 200 B.C. , since the abandonment of rural settlement and the shrinkage of population centers seems to have been widespread in Greece in the last two centuries B.C. But the decoupling of Delos from its traditional Kykladic base may help to explain the failure of the sacred island to recover from the depredations of Mithridates and the pirates of the mid first century. If the neighboring islands had undergone an economic decline in the intervening century, they might no longer have been able to support a local trade at Delos; when the long-distance merchants fled, there was no one in the archipelago to take their place.

Further Directions

Where do we go from here? In many ways, this study is only the first step on a long road to a more nuanced understanding of the character of the Hellenistic economy. I would like to end with a few of the lessons it has offered and some suggestions for future investigation.

One serious problem has remained largely unaddressed in the preceding chapters, although I have often alluded to it. The economic data we have for Delos come almost exclusively from the records of its great pan-Hellenic sanctuary. I have generally assumed that these data may be extrapolated to the city as a whole; in the case of commodity prices, I have tried to show that such prices must have been fair market prices (chapter 1). But we have also seen that the temple was constrained by some considerations that might not have affected secular consumers. It had to buy sacrificial animals, no matter what the price; if some scholars' suggestion is right, it may have sometimes refused to rent sacred houses that failed to fetch satisfactory rents (chapters 5 and 6). Beyond this, however, we know virtually nothing specific about either the economic relations between the sanctuary and the city or the impact of the sanctuary on the local economic scene.^[50] There can be no doubt that the dedications of kings and their ministers and the stopovers of visitors great and humble brought in wealth (chapter 3); but how and to what degree did that wealth impinge on the local economy? Did any real local benefit flow from the phialai and crowns and thousands of other objects stored in the buildings of the sanctuary? Our understanding of these issues would be greatly helped by comparative study of other sanctuary sites. Delos was not the only small town in the Hellenistic world that had a great sanctuary nearby; besides the obvious case of Delphi, we may think of Samothrake, Tenos, and Labraunda near Mylasa. Elucidation of the economic impact of these sanctuaries on their cities would help us evaluate the situation on Delos.

Similar institutions at different places do not always produce the same results. The economic impact of garrisons illustrates this. On Thera the Ptolemaic garrison fostered prosperity, despite occasional conflict between the community and the soldiers; at Keos the imposition of a garrison either had no visible effect or may have correlated with the beginning of a period of decline. The difference may be explicable as the consequence of the permanence of the garrison at Thera compared to the transitoriness of the Keian one; the situation at Samos may have been comparable to that on Thera. But the point remains that it is dangerous to assume that every city reacted similarly to similar institutions.

It is a serious mistake to underestimate the carrying capacity of the land in antiquity. Typically, especially in the minor poleis that formed the vast majority of Greek states, the khora certainly did satisfy most local needs, even in apparently unproductive environments like the Kyklades. Regular consumption of barley, which requires much less rainfall than wheat, made an important contribution to the ability of the Kyklades to feed themselves. Intensive exploitation of the land by terracing, rotating crops, manuring fields, and constructing waterworks raised productivity enough to permit limited exports, which went to support the nonagricultural population of Delos. Regular long-distance importation of foodstuffs was not an important part of the economic scene, except during occasional periods of famine. Recent attempts to argue that trade in fact played a large role in creating income for many cities seem to me mistaken, although it may well be that the Kyklades were more prosperous in the fifth century than at any other time.^[51] Self-sufficiency was not merely a slogan but a reality for many Greek cities, notwithstanding a certain level of necessary imports.

The links between politics and economics remain very obscure. Despite sources like Pseudo-Aristotle's Oikonomikos, we know very little about the economic policies of most Greek cities, especially as they relate to purely internal matters. The economic imposts of outside powers present similar obscurities. It is clear, however, that the connections that the modern world sees as a matter of course between the two may not have been so palpable in the Hellenistic world. Hegemones seem to have had a very limited sense of what they wanted to extract out of the Kyklades: taxes first and foremost; perhaps troops when necessary or possible. They appointed officials to oversee the collection of these monies and settled disputes within cities that might impede the flow of tribute, but that was about it. The gifts of precious phialai or other dedications to the sanctuary of Apollo on Delos enriched the temple and the city of the Delians, but this effect

was purely coincidental. The motive behind the dedications was religious, not economic.

The time has certainly come to stop seeing the "Hellenistic economy" as unitary. Differences were great from region to region, even from polis to polis. Within the Kyklades, variation in climate, resources, topography, and geography combined with differing social and political traditions to create substantial differences within a restricted and superficially uniform region. The divergent trajectories of the four poleis of Keos, an island covering only 103 km² , elucidate the point nicely. But the examples can easily be multiplied: Andros, with its relatively abundant water and strategic location; Tenos, with its remarkably flat (hence easily cultivated) landscape and its sanctuary, second in importance only to Delian Apollo's in the islands; Naxos, with its high mountains, extensive woods, and tradition of local political dominance since the Archaic period; Thera, with its commodious harbor; Amorgos, split virtually into separate islands by its tortuous landscape, straddling the sea routes to Krete—in each case, we are presented with a unique combination of circumstances that helped determine separate economic fates for each island, however much the archipelago as a whole may look like a unit. In the same sense, individual poleis were not unitary. Delos certainly offers the best example, tied as it was to the entire Greek world through its sanctuary, while its economy was virtually entirely parochial. But the other islands present their own contradictions. Through its garrison, Thera maintained connections with distant parts of the Greek world; it might even serve as a retirement home for a citizen of distant Perge. Parian marble had an assured market outside the Kyklades, although in other ways the island retained its local ties (including selling stone to nearby Delian Apollo). The strategic situation of Andros assured constant attentions from foreign powers wanting to secure the entrance to the Saronic Gulf. Full elucidation of these differences is crucial for a nuanced understanding of the economic life of the Hellenistic period.

The research and writing of this book brought home to me with some force a list of desiderata for future study of the ancient economy. Difficult as they may be, there is a great need for more quantification studies. Many places offer no data, of course, and the desire to squeeze as much information as possible out of the sources that do exist certainly can lead to over-interpretation, distortion, and generalizations based on inadequate evidence. The envy historians of ancient Greek economies feel for historians both of ancient Mesopotamia and of early modern Europe will never go away. But there are some bodies of evidence. The most abundant, and still very inadequately studied, comes from Ptolemaic Egypt. A recent collection of data for the Roman period has been treated virtually without statis-

tical analysis, despite tables of data that run in some cases literally for dozens of pages. I do not mean to trivialize the formidable difficulties the Egyptian material presents; I was lucky on Delos to have had data from a very limited time frame and all from one place, so that issues of comparability of prices did not arise. Nevertheless, the Egyptian data would reward close work.^[52] Likewise, the Greek world presents some of the kinds of data that Richard Duncan-Jones exploited for his quantitative studies of the Roman imperial economy. Despite the limitations of his approach and the justifiable reservations one may have about it, similar work would be worth doing for the Hellenistic period.

Amphora data need to be published thoroughly and consistently. It is especially regrettable that we still lack complete data for great sites like Delos. The French amphora project, which aims to computerize all amphorae known from all sites, will be a great help when it finally becomes available. In the meantime, traditional publication of all finds must go on.^[53]

The archaeological surveys of Melos and Keos cited so frequently in this book have been an invaluable aid to my understanding of the economic history of the Kyklades. There can no longer be any doubt that surveys have proven their value.^[54] The Kyklades offer an especially attractive arena for further survey work. The islands are small enough to be covered in a relatively brief time, with relatively small teams, at relatively low cost. The publication of two previous surveys offers plenty of comparative evidence, and the results of these surveys suggest that the history of the islands was unexpectedly rich, varied, and interesting. I would hope, however, that future survey teams will be characterized by greater cooperation from the beginning between archaeologists and historians. For obvious reasons, the designers and directors of surveys tend to be archaeologists (and increasingly anthropologists and ethnoarchaeologists). Historians could learn a great deal by participating in surveys from the initial stages; in turn, their participation could help ensure that surveys are designed in part to explore specific historical questions, especially questions that might seem at first glance to be unamenable to archaeological investigation.^[55]

One question that more surveys might help to answer comes immedi-

ately to mind. The results of surveys conducted not only in the Kyklades but in many other parts of Greece point to a decline in the rural population in the later Hellenistic period.^[56] There are two related problems here. One is the interpretation of the phenomenon. A decrease in rural habitation sites—which is really the evidence on which the proposition rests—is amenable to a number of explanations. Perhaps the rural population simply retreated to local polis centers, giving a nucleated settlement pattern. There need then be no actual decrease in the numbers of people. Recent work at Koresia on Keos, however, suggests a decline in this urban center at about the same time as the rural decline started.^[57] This would appear to mean that the rural population did not retreat to urban centers; but in the absence of studies of the other poleis of Keos—especially loulis and Karthaia, which absorbed Koresia and Poiessa sometime in the Hellenistic period (cf. Strabo 10.5.6 [C486])—it is impossible to be sure whether Koresia's contraction was localized or a local example of a general phenomenon. Further survey work, especially surveys that cover the whole of an island, would help to settle this matter.

The other problem is the exact date of the population decline, which currently cannot be dated more closely than to the third to first centuries B.C. Now the late third century, as we have seen, is exactly the period when Delos and its Kykladic neighbors show a rising local prosperity. I have attributed this prosperity to the absence of outside control, and as far as it goes, the explanation is reasonable. If, however, it could be shown that a real fall in Kykladic population had begun in the third century, it might be possible to attribute Kykladic prosperity in part also to expanding carrying capacity of the land as the absolute numbers of mouths to be fed declined. As the process continued, however, loss of labor would have reduced production even as economic activity was expanding. This might help account for the appeal of Delos to the Rhodians and even to the decoupling of Delos from the local Kykladic economy after the mid second century: a depopulated Kyklades no longer offered the opportunities for enough local activity to support the Delian population; the solution was sought in long-distance trade. This in turn would account for the failure of Delos to recover as a local entrepôt after 69 B.C. The low population and low productivity of the islands could no longer support even the functions Delos had carried out in the third and early second centuries. But all this will remain speculation unless survey work can provide evidence to support or refute it.

Finally, we need more regional studies. Only by piling up specific detail about limited regions can we hope to create a fully articulated, fully nuanced picture of the history of economic life and economic change in the Hellenistic world. The surveys of Keos and Melos, Roland Etienne's study of Tenos, Michèle Brunet's work on rural farmsteads on Delos and Thasos, Lila Marangou's excavations at Minoa on Amorgos, to mention only a few: these quite specific regional or polis studies are vital. In this regard, I think M. I. Finley was mistaken to condemn regional studies as uninteresting or largely antiquarian: they are, rather, the lifeblood of the study of the ancient economy.^[58]