Month-To-Month Fluctuations and Seasonal Cycles
Olive Oil The accounts give eleven years with at least three monthly prices (table 5.1). The first series comes from 304 B.C. , for which the accounts preserve three prices for the first three months, and then, after a gap of five months, two consecutive prices in Metageitnion and Bouphonion. The early prices are the highest known from Delos, but after the gap the price has plummeted 41 percent, and it suffers a further staggering fall in the next month of 44 percent. The total decline for 304 B.C. amounts to
[17] I ignore many intricacies here. For an introduction to the construction of economic indices, see S. N. Afriat, The Price Index (Cambridge, 1977); Jacqueline Fourastié, Essai sur la mesure des quantités économiques (Paris, 1972); and U.S. Department of Labor, Bureau of Labor Statistics, BLS Handbook of Methods, vol. 2, The Consumer Price Index (Washington, D.C., 1984). See also Kent, 309 with n. 209; Roger Bagnall, Currency and Inflation in Fourth-Century Egypt (Atlanta, 1985), 1–8.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
about 67 percent. Of the remaining ten years, six offer a series of prices well distributed across the year. Five show a distinctive pattern in which prices undergo an adjustment in the late summer or early fall (250, 246, 224, 179, and 174 B.C. ). In 250 and 246 B.C. , prices rise; in 224 and 174 B.C. , they fall, 179 B.C. shows a mixture of the two patterns. Six years show another adjustment in the late winter or spring. In 250, 218, 200, and 179 B.C. , prices rise once in Galaxion or Thargelion and then return to former levels. In Artemision 231 and Panemos 174 (already early summer), prices fall for a month and then resume previous levels. Exceptionally, the prices of 218 B.C. remain elevated after a rise in Thargelion. Finally, three years have a slight but interesting drop in price toward the end of the year (250, 179, and probably 174). I shall postpone discussion for the present (see pp. 136–38 below). The year 200 B.C. is unexampled in its stability.[18]
The spring and fall price adjustments are probably best seen as reflections of the exigencies of ancient transportation and of the seasonal cycles of the olive. As we saw in chapter 3, the sailing season "par excellence is from 27 May to 14 September . . . and . . . the outside limits are 10 March to 10 November."[19] Except for occasional forays to and from nearby neighbors, enabled by a spell of good weather, the Delians could not count on imports outside the sailing season. Supplies sufficient to tide them over the winter must have been on the island by the end of the autumn at the latest.
The annual cycle of the olive, which has not changed since antiquity, is well known.[20] Collection of fruit fell into three periods: fallen olives were picked up in August, green fruit was taken in September, and the real har-
[18] I can offer no good account of this year. It stands out among all the series by its absolute stability of price. Perhaps a benefactor provided oil for the year at a fixed rate (although there is no mention of such benefactions anywhere in the documents); perhaps the hieropoioi experimented with contracting a year's supply from a single supplier at a fixed price, as they did for the sharpening of tools in 279 B.C. (IG XI 2.161A107–8); or perhaps prices were simply extraordinarily steady. The remark that under the arkhon of 200 B.C. "there was health and prosperity" (IG XI 2.128.2) means nothing, since the sentiment is commonplace (cf. e.g., IG XI 2.105.2, 108.2, 109.1, etc.).
[19] Casson, Ships, 270–72, quotations from 270–71. See chapter 3, pp. 54–55, above. This chapter is in part a modest response to Casson's remark that an "important subject that has never been treated is the extent of the economic dislocation that all port towns had to suffer because of the limited sailing season" (271 n.4).
[20] See A. S. Pease, RE 17 (1937), s.v. Oleum, 2454–74, and s.v. Ölbaum, 1998–2022. Amouretti, Pain, 177–95, catalogues the many uses of olives and their oil in great detail. The bibliography on ancient olioculture is large and growing; forrecent contributions, see Amouretti, Pain, passim; Sandy, Production and Use of Vegetable Oils in Ptolemaic Egypt, 72–82, with 82 n. 18; M. Helzer and D. Eitan, eds., Olive Oil in Antiquity (Haifa, 1987); H. Camps Fabrer, ed., L'Huile d'olive en Méditerranée (Aix-en-Provence, 1985); brief overview, Isager-Skydsgaard, 33–40. On the life cycle of the olive, see Raymond Lousset and Gérard Brousse, L'Olivier (Paris, 1978), 47–77.
vest of ripe olives ran from October through December. In the modern Kyklades, the harvest typically lasts from late September through November.[21] The "market year" for olive oil thus lasted from autumn to autumn. Two additional factors affected the pricing of olives. Trees produce good yields only every other year; during off years, the crops tend to be only half of the previous year's yield. Moreover, orchards over vast regions tend to synchronize: biennial fluctuation in yield can be seen in harvest figures for the Peloponnesos, and even for the whole of Greece. Finally, although mature olives are generally reliable producers, regional shortages were not unknown, and once trees reach the age of about 200 years, as Theophrastos remarked (Hist. plant. 4.13.5), production tends to fall off markedly.[22]
Delos must have expected its first cargoes of the new sailing season each year in the period between March and May, which so frequently shows an adjustment in price level. Price rises may, therefore, reflect the final sales of depleted local stocks just before the arrival of fresh shipments; consumers would be willing to pay higher prices in the face of immediate local but temporary "shortages" and of uncertainty about prices of oil to come (250,
[21] Geoponika, ed. Beckh, 9.17.1; Pliny Nat. hist. 15.4; Theophr. Hist. pl. 1.11.4 and De caus. pl. 1.20.4. Amouretti, Pain, 73. For Greece, Theophrastos (De caus. pl. 6.19.3, 8.1, 8.5) indicates that olives ripen after the rising of Arcturus, which falls in September; cf. also Pliny Nat. hist. 15.3: "Augetur oleum ad Arcturi exortum a. d. XVI Kalendas Octobris: postea nuclei increscunt et caro." For the modern agricultural calendar in Messenia, see Aschenbrenner in MME, 51 (olive harvest November–January); for Melos, Malcolm Wagstaff and Siv Augustson in Island Polity, 121, fig. 10.7 (olives in November–January); for Amorgos, Christopher Connell, In the Bee-Loud Glade (Nafplion, 1980), 47.
[22] Aristotle knew the pattern (De plant. 1.7 [821b15–17]); his student and successor Theophrastos observed a three-year cycle at Olynthos (De caus. pl. 1.20.4). For modern data, see Aschenbrenner in MME, 53. On the shortage at IG II 903, see Philippe Gauthier, REG 95 (1982): 275–90. "The notorious uncertainty of the olive crop might account for fluctuations in price in antiquity as in modern times, when a good harvest may be followed by a complete and utter failure,—a few days of rainy weather at the critical time of blossom may bring disaster," says W. K. Pritchett (Hesperia 25 [1956]: 184 n. 34). On senescence, see Marie-Claire Amouretti in Agriculture in Ancient Greece, 80; Lousset and Brousse, L'Olivier, 63.
218 [but see further below], 200, and 179). If the previous year had been good, expectation of reduced yields might also contribute to upward pressure. Since olives were harvested and processed in the late fall and early winter, after the sailing season had closed, Delians probably lacked complete information about the size and quality of the harvest. Arrival of the first shipments calmed the market in those years when harvests were typical; in years with good harvests, or simply of higher preseason anxiety, the appearance of fresh oil may even have depressed the price temporarily (231, 174).
Of the two patterns, price declines are clearly more important. Except for 250 B.C. , the rises are of little importance (about 12 percent) and of brief duration; in 200 B.C. , the change affected only a single spot purchase in the month of Galaxion: another purchase the same month cost the same as in the preceding and following months. The declines of 231, 179, and 174 B.C. persisted through the summer. This difference suggests that years in which the price fell and stayed down were benefiting from a good harvest the previous autumn, which had only come onto the market and affected prices with the opening of the spring sailing season. We can guess that harvests in 232, 180, and 175 B.C. had been abundant. Conversely, poorer harvests preceded those springs in which the opening of the sailing season led to no permanent downward readjustment of prices (251, 219, and 201 B.C. ). This further implies that Kykladic olives followed a cycle of good-poor harvests in even-odd years (by the Julian calendar), with the occasional exception like 175 B.C.
In 218 B.C. , prices did not readjust after the rise in Thargelion. The persisting price of 112.5 is the highest summer price for any year except 304 B.C. (whose prices are exceptional: see below); it is matched only by prices from the late autumn or winter of 250, 179, and 169 B.C. This unique pattern probably represents a poor harvest year. The cycle derived above would make 219 B.C. a year of low yields anyway, but these high prices imply that it was poorer than usual.[23]
The adjustments of the late summer or early fall correspond to the olive harvest and the closing of the sailing season. Merchants typically returned to winter ports in September, although again some might continue to sail as late as November. The olive harvest was at its height from late September till December, with oil produced continuously. If harvesting began long enough before the sailing season ended, increasing supply might have
[23] The shortage of 218 B.C. was trivial, however, compared to the prices of 304 B.C. These prices, which form a crucial piece of evidence in the inflation-demand theory espoused by Friz Heichelheim and others, receive detailed discussion below.
brought prices down. Conversely, if the harvest of the olives or the processing of the oil was delayed until after the merchants had put their ships up for the winter, tight supplies and the anticipation of no new imports till spring might have fueled a price rise. The data seem to reflect these patterns perfectly. For all years, the price adjustments are permanent, unlike those of the spring; this reflects the combined impact of the harvest and the end of sailing (and incidentally makes it very unlikely that these adjustments should be attributed to the adventitious arrival of a winter shipment made possible by temporarily favorable weather). Three years show rises in the fall (250, 246, and 179), which probably mean either that the sailing season was effectively over by the time the harvest was well under way, or that the harvest was poor (247 B.C. , for which we have only two prices, also follows this pattern; see table 5.2). In the two years showing declines (224 and 174), a good and early harvest probably overlapped with the sailing season enough to permit distribution of some new oil before winter set it.
There is one additional factor that must be taken into account, and that is local production. The conventional view is that Delos lacked olive trees. The inventories taken every ten years by the hieropoioi of the capital equipment of the estates owned by Apollo show no olives until after 237 B.C. , when new estates on Mykonos came into the god's possession. The early traveler J. Pitton de Tournefort, who visited Delos in the middle of the eighteenth century, remarks on its grain crop but says nothing about olives; they were, he observes, "peu" on Mykonos. No olives grow on Delos today.[24]
But these arguments are hardly conclusive. No inventories have been preserved for three estates, so that we do not know what was cultivated on them, and the Delian
[24] Philippe Bruneau and Philippe Fraisse, BCH 108 (1984): 721; Brunet, 147; Isager-Skydsgaard, 197; Alison Burford, Land and Labor in the Greek World (Baltimore, 1993), 110, following Kent, 288. For Kykladic oil production, see ID 366B18–23 (Mykonos); Galen De simp. med. temp. 11.872K; Herak. Lembos, ed. Dilts, frg. 41 (Peparethos); Athen. 67a (Samos). For some modern accounts, see Wagstaff and Augustson in Island Polity 111–13 (Melos); Paul Halstead and Glynis Jones, JHS 109 (1989): 51, and Connell, In the Bee-Loud Glade, 43–47 (both Amorgos). The rarity of olives on early modern Keos, often noted by travelers, may have been because of concentration on commercial production of acorns (velanidia); see J. Bennet and S. Voutsaki in Landscape Archaeology, 377, S. B. Sutton in ibid., 387, and T. M. Whitelaw in ibid., 447–49.
[25] There are no inventories for Akra Delou, Lykoneion, and Sosimakheia. On Phytalia, see chapter 6, pp. 207–8. For private farms, see P. Bruneau, BCH 112 (1988): 569–73; Michèle Brunet, BCH 114 (1990): 706, 113 (1989) 754–61,112 (1988): 787–91, and 111 (1987): 644–47; Brunet, passim; G. Reger, Phoenix 46 (1992): 322–41.
that the accounts preserve no mention of olives before 237 B.C. , for Apollo himself owned at least one olive tree. Under Apatourion 250 B.C. , the hieropoioi report an income of 3 dr 2 ob from two sales "of wood from the wild olive" (
Price drops in 224, 179, and 174 B.C. may thus reflect a good local harvest; the same decline in Poseidon 250 B.C. may also belong to this pattern, and olive yields in the Kyklades may have been synchronized to give good harvests in even-numbered years and poor in odd. Furthermore, the occasional association of declining winter wood prices with low early winter olive prices—250 and 174 B.C. ; in 179, wood prices remained stable (see table 5.3)—lends credence to the view that Delos did have some olives. Another argument, explored below, also implies that Delos ought to have had some olives.
The data speak quite eloquently and in considerable detail about the character of the annual market for oil. Prices correspond well with the behavior of olives (and their producers) in conjunction with the exigencies of supplying a small island far from self-sufficient in oil and subject to closure to merchants for a good part of the year. The patterns we have seen also justify some guesses about seasonal price changes for years with only limited data (table 5.2). In both 272 (or 271) and 247, prices reflect typical end-of-year changes. For 279, 269, and 265-255, prices look instead like those before and after a spring adjustment (assuming they are reported in chronological order). In this context, it might be interesting to estimate the total annual demand for oil on Delos. Much of this demand must have been met by imports, and an estimate will help to delineate the scale of trade in oil.
In a number of cases in the accounts, the hieropoioi record oil bought for the Hieropoion (
[26] E.g., IG XI 2.154A14–15, 161A108; ID 316.79, 354.59. On the religious duties of the hieropoioi (quite limited), see Vial, 216–32. On the Hieropoion, see now Jacques Tréheux in Stemmata, 377–86, esp. 383–86.
| ||||||||||||||||||||||||||||||||||||||||||||||||||
consumption to the average per person for the population as a whole established in chapter 4, this comes to 1. 08 kh /mo/person, or 13 kh /yr.[27] This figure corresponds well with an independent estimate offered by Marie-Claire Amouretti, who, beginning from rather different suppositions, puts the annual consumption of a "free man, citizen, who frequents the gymnasion" at 55.5 liters, or 17.8 khoes (= 1.5 kh /mo). Adjusted to account for less consumption by women and children, Amouretti's figures yield an annual demand per person of about 15.6 khoes, or about 1.3 kh /mo. This figure is only 18 percent higher than my estimate, and, as Amouretti herself admits, her figure "must be taken as an unusual consumption."[28] At my lower estimate of 13 kh /person/yr, the estimated Delian population of 2,700–9,100 would have had a total annual demand of 2,925–9,860 metretai.
But this estimate is probably too high. Because the hieropoioi came from the highest social levels of Delos, their standard of living must have surpassed that of most of their fellow citizens.[29] Oil was expensive in antiquity, so the level of consumption was no doubt linked to social standing. Poorer citizens, without the resources, leisure, or responsibilities of their wealthier fellows, must have consumed appreciably less oil. We can figure consumption minima by reckoning from caloric intake. We saw in chapter 4 that
[27] See purchases for 279, 250, 231, and 218 B.C. , all years when only two hieropoioi served (Vial, 163–64, with discussion at 172–83). For the ratios, see chapter 4, pp. 85–86, above.
[28] Amouretti, Pain, 195, approved by D. J. Mattingly, J. of Roman Arch. 1 (1988): 159.
[29] On the social level of the hieropoioi, see Vial, 187–91, 253–61, 262–69.
a reasonable guess for the average daily caloric intake of the Delians is 2,460 cal/day/person. Thomas Gallant has estimated that roughly 5 percent of a person's needs were satisfied with oil, which implies a monthly per capita consumption of about 0.15 khous and a total Delian demand of 405–1,365 met /yr.[30] These results are not even one-sixth of those calculated on the basis of the purchases of the hieropoioi. As in the case of grain, for which ancient rations consistently outstrip modern estimates of caloric need, other factors are likely to intervene. Oil served other purposes than food; it was used for lighting, washing, perfumes, medicines, ritual, and in other contexts.[31] Demand estimates based purely on consumption as food will ignore these important uses. Moreover, Gallant's caloric estimate of roughly 5 percent seems low. Diets in which about 28 percent of daily caloric needs were satisfied with oil are attested for modern Greece; consumption at such levels in fact comes very close to 13 kh /person/yr. Other surveys yield, rather consistently, a consumption of about 12.6–12.9 percent of calories in oil.[32] Granted that the hieropoioi who bought oil for themselves—and Amouretti's "free man, citizen, who frequents the gymnasion"—consumed considerably more oil than the typical Delian, let us strike a compromise and reckon that, very roughly, minimum consumption was about 4.5 kh /yr (corresponding to 12.6 percent of calories in grain) plus half again as much for other needs, for a total of 6.75 kh /yr/person. Taking the consumption at the level of the hieropoioi (13 kh /yr/person) as a maximum, and assuming that only 10 percent of the population consumed at this level, gives a typical annual consumption of 7.4 kh /person, or, for Delos as a whole, roughly 1,665–5,611 met /yr.
Figures compiled by Lionel Casson for the carrying capacity of ancient merchant ships suggest typical cargoes of about 1,500 metretrai. A wreck of about 275 B.C. recently recovered off the southwestern coast of Turkey carried at least six hundred amphorae in two sizes, with mean capacities of 38.0 and 10.87 liters, for a total minimum cargo of about 580 metretrai. (These amphorae may not have been typical; a recent study of capacities found standard Rhodian containers to hold 24–26 liters.) By Casson's reckoning, this was a small ship; but annual Delian demand would have filled
[30] See chapter 4, pp. 85–86, above; Gallant, 72–73. On the caloric value of olive oil (8,073 cal/l, or about 25,188 cal/khous for a khous of 3.12 l), see Catherine F. Adams, Nutritive Value of American Foods in Common Units (Washington, D.C., 1975), 102; Foxhall-Forbes, 85, table 1.
[31] Amouretti, Pain, 181–95.
[32] Reckoned from Gallant, 65, table 4.1. For the 28 percent figure, see Leland G. Allbaugh, Crete: A Case Study of an Underdeveloped Area (Princeton, 1953), 100–111, 131; criticized at Gallant, 64, but see Lousset and Brousse, L'Olivier, 22, giving an annual per capita consumption in Greece of 18.9 kg, the highest they cite.
only 3–13 such vessels, and at Casson's typical cargo capacity, as few as 2–4 shiploads could have satisfied the annual Delian demand.[33]
It is also possible roughly to calculate the number of trees that would have been required to meet Delian demand. At a biennial average of 4 kilograms of oil per tree, 16,848 to 56,777 trees would have sufficed. Planted at 40 trees per hectare, they would have covered only 421–1420 ha, a maximum of about 0.6 percent of the total area of the Kyklades; even at a much less dense planting of 6 trees/ha, which would have left almost all that land free for cereals, they would have occupied no more than 9,463 ha, or 4 percent of the islands' surface. No doubt Kykladic olives were clustered in orchards much more rarely than their cousins in North Africa (source of the planting densities) and were far more integrated into a small-scale, diversified economy.[34]
As we have already seen of cereals in chapter 4, total Delian demand for oil was a trivial portion of the local Kykladic demand, to say nothing of that of the Aegean as a whole. A handful of shipments sufficed to cover Delos's total annual need. This situation helps to explain the general stability of price across any one year. Once local oil retailers had sufficient stock to cover a year's demand, wholesale prices on Delos would drop. Merchants with oil for sale would divert elsewhere, perhaps until close to the end of the sailing season, when local dealers would want to replenish stocks before winter set in (and take advantage of the harvest if possible). Any fluctuations in the wider market for oil would bypass Delos, where prices would tend to be set across the whole year by the prices that had obtained when the original shipments arrived.
Firewood Month-to-month price changes for firewood appear in table 5.3. Two patterns are detectable. Many years show a shift, usually a decline, within the first few months of the year. Prices in 250, 231, 224, and 174 B.C. fell in Hieron, Galaxion, Artemision, or, in 174, after Thargelion. Two years (200, 179) have rises early in the year. Another adjustment occurred in the late summer or fall. In 250, 246, 231, 224, 200, and
[33] Cemal Pulak and Rhys F. Townsend, AJA 91 (1987): 31–49; Carolyn G. Koehler and Malcolm B. Wallace, AJA 91 (1987): 49–54; recent study: M. B. Wallace in Recherches sur les amphores grecques (Paris, 1986), 87–94, at 93; Foxhall-Forbes, 84; Casson, Ships, 184.
[34] Amouretti, Pain, 204; D. J. Mattingly, J. of Roman Arch. 1 (1988): 160. Ibid., 45. Amouretti in Agriculture in Ancient Greece, 79–80; Forbes in ibid., 93–98. Chapter 4, p. 107, above. At the planting ratios cited at Sallares, 475 n. 41 (present-day planting at 190 trees/ha and a figure of 100–130/ha for dry farming area with minimum annual rainfall of 500–650 mm, after Loussert and Brousse, L'Olivier, 10, 178), the surface covered would have been even smaller.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
174, prices rose between Metageitnion and Posideon, with rises most common in Bouphonion-Aresion. In only one case, 272 or 271 B.C. , did prices fall, by 7.7 percent from Aresion to Posideon. But 250 and 174 B.C. also showed declines in Aresion or Posideon after the more typical rise in the fall. Prices tended to stabilize in the late spring and summer, although subject to occasional monthly fluctuations (268, 250, 246, 231, 224, 218, 200, 179, and 174).
These patterns are amenable to explanations based on the seasonal demand for firewood, the exigencies of the sailing season, and the "harvesting" patterns for scrub. The demand for wood was certainly higher in the winter than in the summer; the nodes of change fell in the spring and the fall. Spring declines, which appear in four out of six cases, clearly fit the supposition of a reduced demand with the onset of warmer weather. The opening of the sailing season would have added to the downward pressure on prices, since now supplies from elsewhere could be brought in to replenish stocks depleted over the winter. The pattern appears nicely in 250, 231, 224, and 174 B.C.[35]
The summer pattern of lower prices subject to fluctuations responds to a reduced demand met by occasional imports; the transient price changes that occur across months, as in Panemos of 250 or Hekatombaion of 231 B.C. , probably reflect temporary changes in supply. The fact that these fluctuations are transient and do not follow a clear pattern lends further support to this view. In the fall, demand changed again. The approach of winter and the impending closure of the sailing season would have led to increased stockpiling, hence increased demand; six years reflect this change with increased prices at exactly this time.
How can the declines at the end or very beginning of some years (272 or 271, 250, 179 [Lenaion], and 174) be accounted for? I see three possibilities, which should be supposed to have acted in conjunction and in different combinations in different years. Some years must have had milder winters than others, just as the winter of 169-168 B.C. was exceptionally severe (Livy 44.20). In such years, Delians may have overstockpiled firewood and then sought to unload the excess in midwinter: hence occasional price declines in Aresion or Posideon. In other years, transient declines may have resulted from unexpected shipments from neighboring islands during a
[35] This view is not affected by the disagreement about the use for which wood was bought for the Hieropoion between Philippe Bruneau, BCH 105 (1981): 94 (wood for heating and cooking), and Jacques Tréheux in Stemmata, 385 (ritual only). The accounts cannot settle the matter, but the hieropoioi did not always buy wood for themselves each month, which may mean that their needs varied; some personal use may therefore be supposed.
spell of good weather. Perhaps 250 B.C. offers the best candidate in the strikingly anomalous price of 100 in Galaxion.
The third possibility has a structural aspect. As noted above, coppices could be harvested at any time of year. Olive prunings, which were also an important source of fuel, became available only during and immediately after the harvest. Olive presscake—the residue left after the fruit has been pressed for oil—was another important energy source also available in quantity only immediately after the harvest.[36] In years of heavy pruning and/or good harvest, these materials would begin to reach the market at the end of the Delian year, in Apatourion, Aresion, or Posideon. In sufficient quantity, they may have acted to counter the normal winter pressure keeping prices high. In conjunction with a mild winter, the sudden availability of olive by-products might account for the 12–25 percent price decline seen in these three years. Moreover, our investigation of oil prices suggested that the olive cycle in the Kyklades brought good harvests in even years, poor harvests in odd. In modern practice, pruning is heavier in years with good harvests. All four years showing wood price declines at the end of the year should also have produced good harvests (counting the low price of Lenaion 179 as evidence for 180 B.C. ).[37]
The level of olive production on Delos must have been low, at least compared to the total demand for oil. It is possible that the reduction in price of firewood at the end of many years reflected the availability to the temple of prunings from its own trees; this would account for the absence of other instances of sale of cuttings.[38] In any case, year-end declines in both oil and firewood prices are rare enough that local production can have covered only a fraction of the demand, and it is probably appropriate to imagine that, in the years where we see such reductions, harvests were especially good and other factors—like a mild winter—also helped.
[36] Aschenbrenner in MME, 54, on pruning in a contemporary Messenian community: "Some of the prunings are burned green soon after pruning, but most remain to dry and are used through spring, summer, and the next fall." For advice on presscake I am indebted to Lin Foxhall (per. comm.); cf. also Frederick R. Matson, Advancement of Science 23 (1966): 152.
[37] Aschenbrenner in MME, 54: "If the [olive] tree is loaded [with fruit], the pruning is very severe and several large limbs . . . are cut off, as well as many smaller ones." Pruning is necessary because olives bear fruit only on branches two years old; see also Sandy, Production and Use of Vegetable Oils in Ptolemaic Egypt, 73. Although IG XI 2.219 + 220 might belong to either 272 or 271 B.C. , J. Tréheux considers the early date more likely: REG 99 (1986): 301.
[38] At the prevailing prices for Bouphonion and Aresion 250 B.C. , Apollo had sold off between 4.3 and 5 talents of wood (IG XI 2.287A73, 80). No purchase of firewood for the temple is recorded in Apatourion itself. Could this mean that Apollo was able to supply all his own needs that month from prunings?
The great exceptions to this model for the annual price history of firewood are 200 and 179 B.C. Both years showed prices rises at the beginning of the year, high but stable prices across the spring and summer, and no further rise in the fall. Below I explore the possibility that, for 200 B.C. , prices may have been affected by the constant military activity in the central Aegean. For 179 B.C. , however, military operations cannot be invoked, since the Delians celebrated the year as one of peace (IG XI 2.130.1). Another possibility, also treated below, relates to the structural change in mean annual firewood prices that occurred around 218 B.C. ; it is possible that these structural changes in long-term price trends overwhelmed the annual pattern typical for the first eight decades of the third century.
In general, however, annual firewood prices conform with very satisfying consistency to a price model based on a typical annual demand fluctuation, the exigencies of the sailing season and of summer trade, and the occasional infusion of additional supplies at the end of the year after the olive harvest.
Pigs Monthly pig prices (table 5.4) show less patterned behavior than either of our other two goods. Fluctuations are sometimes extreme, sometimes mild; prices may rise or fall in two or three consecutive months, or they may zigzag; occasional stability across two or three months may give way to wild fluctuations.
As with our other two goods, at least some pigs are likely to have been imported to Delos,[39] but evidence for any impact of the opening and closing of the sailing season is elusive. Some years in which prices fall begin with low prices, others with high. Three years show high end-of-year prices compared with prices earlier in the same year (250, 174, and 171), which might be attributed to the closure of the sailing season, except that eight years show high prices in the summer (269, 246, 224, 218, 200, 179, 174, and 169), and in four of those years, prices declined in the fall (246, 224, 200, and 174). There is therefore no persuasive evidence for price changes in response to the close of the sailing season.
The opening of the season may have left rather more, although still tenuous, traces. Six years registered a price decline across either Galaxion-Artemision or Artemision-Thargelion, and no year recorded a price rise
across both pairs of months (250, 231, 224, 200, 179, and 174). This pattern looks promising as evidence for a price decline at the opening of the season except for one major problem: in only two (218 and 179), or possibly three (including 174), years did prices stay down. In the others they rose in the very next month, usually to levels at least as high as if not higher than they had occupied before. While the opening of the sailing season may have had some effect every year—and certainly seems visible in 218 and 179, when prices fell and stayed down—other effects that kept prices up seem generally to have swamped it.
The impression of atypical price behavior at the usual nodes for change is reinforced by the level of prices at other times of the year. High summer and low winter prices are particularly unamenable to explanation by the ebb and flow of the sailing season. Of the several factors that may have intervened to overwhelm the pattern seen with our other commodities, two are difficult to evaluate, although they are bound to have had their impact. Swine are especially prone to disease, a fact of which Aristotle was fully aware. Swine diseases spread rapidly and may wipe out most of the animals in an infected herd. On small islands like Delos and Rheneia, however, the population may not always have been large enough to sustain a disease, whether epidemic or endemic. Thus while there may have been occasional
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
epizootics that wiped out swine on Delos or its neighbors and so raised prices, in general the fluctuations we see are probably not to be attributed to this cause.[40]
The second matter has to do with requirements for sacrifice. Sacrificial animals had to be "perfect."[41] If different administrators at different times and under different circumstances imposed different standards, widely varying numbers of animals might qualify; very picky hieropoioi might feel justified in paying substantially higher prices for animals they judged to be better suited to Apollo. Particular circumstances might also have an effect. Early in the third century, a strange portent occurred "when grape cuttings appeared in the spring" (
Two other possible factors seem more important to me, because they would have been structural and cyclical: the impact of the life cycle of the pig, and the unusual demand structure for pigs as opposed to oil or firewood.
Pigs were reckoned in antiquity to be able to produce two litters a year, with a gestation period of about four months and two months for weaning.[44] A
[40] Hist. anim. 8.21. Sallares, 221–93.
[41] See n. 13 above. On the Delian boonai, appointed to buy oxen for sacrifice, see ID 399A7, 14, 17, 18, and Vial, 243–44. Although explicitly attested only in this one year, these boards are likely to have functioned throughout independence, since the sale of hides and payments to priests—two of the duties of the boonai —are attested frequently: see IG XI 2.274.24, 287A24, A113–14.
[43] Prices for older pigs also show violent month-to-month fluctuations. Two animals both bought in the same month in 246 B.C. cost 14 and 8 dr; in 247 B.C. , 18 and 16.5 dr; and 20 dr in 224 B.C. (ID 290.88, 291b23–24, 338Aa59).
[44] Arist., Hist. anim. 8.6; cf. Orth, RE, n.s., II.2 (1923), s.v. "Schwein," 801–15. This corresponds very well with modern experience: see W. J. Carmichael and John B. Rice, "Variations in Farrow, with Special Reference to the Birth Weight of Pigs," University of Illinois Agricultural Experiment Station Bulletin no. 226 (Urbana, 1920), 68–71; Dennis L. Meadows, Dynamics of Commodity ProductionCycles (Cambridge, Mass., 1970), 54; J. L. Krider, J. H. Conrad, and W. E. Carroll, Swine Production (New York, 1982), 172 (111–17 days, with average of 114).
and younger than the
Fish aside, pork was probably the commonest meat consumed in the ancient diet,[46] and young pigs were the least expensive pork. They were also a favored offering, especially for poorer people.[47] Demand for pigs,
[47] Plato Rep. 378a; cf. also Herod. 2.48, Xenoph. Anab. 7.85; [Demos.] 54.39, Hippon. fr. 40 (Bergk), Heniokh. fr. 2 (Kostel-Austin) = Athen. 9.396D. For scenes of household sacrifice on Delos, see W. Déonna, La Vie privée des Déliens (Paris, 1948), 103, citing M. Bulard, La Religion domestique dans la colonie italienne de Délos d'après les peintures murales et les autels historiés (Paris, 1926), 18, 57–58, 86 (which was unavailable to me). See also A. Plassart, BCH 40 (1916): 176, fig. 10;177, fig. 11; and esp. 213, fig. 29; and, most recently, BCH 112 (1988): 765–66, with fig. 36, a private votive plaque dating to before 69 B.C.
however, may have been very sensitive to small price changes; that is to say, the demand for pigs may have been relatively price-elastic. If so, demand would have responded quickly to small price variations, leading to a quick price rise. High prices will naturally attract sellers, and may force the price back down if the supply exceeds the demand. The year 250 B.C. offers an excellent example of such a pattern.
The year begins in Lenaion with a price of 91.7. Suppose poorer buyers abstain, and by Hieron the price has fallen by 18.2 percent to 75. The low price attracts buyers, and demand pushes the price up again to 100. A high price encourages producers—this early in the year we should think of Delians, not importers—to bring more pigs to market, so that by Artemision the price has fallen to 91.7, but rises again within a week to 112.5 (for a pig bought for the Artemisia). Demand slackens again in the face of a high price, and by Thargelion pigs fetch only 100. This would be a perfect example of a price curve for a commodity whose demand curve is highly price-elastic.
Five other years show extended periods of similar response (224 for nine months; 200 for seven; 246, 231, and 174 for five); other years might also display the pattern if the data were complete (218, 179, and possibly 247). It seems evident that pig demand was extremely sensitive to price, and that this sensitivity overwhelmed the impact of both the natural cycle of the animals—hardly a serious problem, since producers could easily afford to withhold animals for a month or two if prices were low—and of the opening and closing of the sailing season, which seems to have exhibited no more than a modest and occasional effect in the spring and hardly any at all in the fall.
The failure of pig prices to show much response to the sailing season has important implications. It seems very likely that much of the local demand was satisfied, not from imports, but out of local production. Some pigs were clearly raised on Delos. Although the inventories of temple-owned estates, which are rich in words for agricultural buildings, never mention pigsties, a chance reference in another context proves Apollo's responsibility for at least one. This building may have belonged to one of the estates for which we lack inventories. Livestock, including pigs, were surely raised on private estates on Delos; Michèle Brunet has recently suggested that the peninsula of Patinioti at the northeastern tip of the island was
entirely devoted to ranching.[48] We also happen to know the names of a number of persons who sold pigs to the temple. Two, Timesidemos and Aristodikos, who sold a pig each in Bouphonion and Apatourion 269 B.C. , may well be identical with the brothers who shared the lease of Khareteia in 269–260 B.C.[49] It is tempting to suppose that they raised the pigs there. In any case, the
At the same time, total demand on Delos may well have been relatively low. The temple is a poor guide here. Its demand was largely inelastic, for the hieropoioi had to buy and sacrifice a pig on the first of every month. It is exactly where some slight "elasticity" in temple demand emerges, however, that some very striking corroborating evidence arises for the sensitivity of price to demand. The Delian Thesmophoria was held every Metageitnion, and this ceremony required the hieropoioi to buy four pigs of different condition, one of which was a
[49] IG XI 2.203A52, 53–54; 203A20. Cf. Kent, 335, no. 223, 323, no. 44; Vial, 218–19. As Khareteia had vines (see chapter 6, p. 194, below), I am reminded of Varro's advice (Rust. 2.4.6) on raising porci sacres for sacrifice: "Si fundus ministrat, dari solent vinacea ac scopi ex uvis."
[51] IG XI 2.287A68–69. Full citations at Bruneau, 287; see 285–90 on the Delian Thesmophoria.
[52] This rise is so great that I suspect some other factor, such as especially picky hieropoioi, must also have been at work.
to an increase in temple demand from one to five pigs apparently delivered a substantial boost to the price.
This price behavior also reinforces the notion that the total monthly demand for pigs on Delos could not have been very large. Except for the rarified stratum who could afford meat on a regular basis, individual Delians probably bought pigs only very rarely: for household sacrifice, for public ceremonies in which they participated,[53] and for feasts. Total demand cannot have approached that for firewood or olive oil, both staples in every Delian household.
In many respects the behavior of pig prices, tied partly of course to the natural history of the animals, has changed remarkably little in the past 2,300 years. This constancy suggests a possible explanation for high summer pig prices. Modern pig farmers, who also get two litters a year from their animals, prefer to avoid farrowing in cold weather. They therefore arrange for sows to farrow in May and September. Piglets born in May will not be weaned until July or August. If ancient farmers also favored late spring farrowings,
