Review of the Cambridge Economic History of the Greco-Roman World

From William A. Percy
Jump to: navigation, search

Review on The Cambridge Economic History of the Greco-Roman World

Walter Scheidel, Ian Morris, Richard Saller

Cambridge University Press, 2008, 958 pp., $204

When I turned randomly to inspect a chapter in the Cambridge Economic History of the Greco-Roman World, I was shocked to discover some egregious errors within the few pages in the chapter entitled ‘The Early Iron Age in the Western Mediterranean’ by Michael Dietler. First, “there may have been a Phoenician metal trading system in the area of Spain before the conquest of Tyre by the Assyrians in 572” (p.246), when, in fact, the Medes captured Nineveh in 610 BC and the Assyrians made no conquests thereafter. (Is he perhaps referring to the destruction of Tyre by the Babylonian king Nebuchadnezzar in 573 BC?) As shocking as that chronological gaffe is, we are informed, shortly afterwards, (p.250) that Marseilles, a central focus in that chapter, was “one of the last well-protected harbors between the Rhône and the Pyrenees,” when in fact it is well to the east of the Rhône, between it and the Alps, and became part of Provence, the earliest region of Transalpine Gaul to be dominated by Rome. Finally, the claim that Marseilles, with 15,000 to 20,000 inhabitants was “vastly larger than any other settlement in the western Mediterranean (colonial or indigenous) until the Roman period” seems exaggerated to me. Dietler never clarifies what he means by “the Roman period” (apparently he liked such imprecisions as "in the area of Spain" above) - i.e. when and where it actually begins. A few sentences later, he admits that Marseilles may have been “relatively small by the standards of Greek colonies in southern Italy or Etruscan cities” (p.250), but his generalization about Marseilles’ great size (no date given) remains problematic, and he never seems to consider other large settlements in the Western Mediterranean, such as Ostia, Naples, Palermo, and even New Carthage.

Such sloppy editing, whether by Richard Saller and Walter Scheidel, who are supposedly experts at least on demography, or by any other editors or proofreaders of such a press at one time so fastidious instills little confidence. Statistical estimates in the chapters by Scheidel entitled “Demography” (p. 38-86), and that by Saller “Household and Gender” (p.87-112 ) are way off. I argued in Pederasty and Pedagogy in Archaic Greece, (Illinois, 1996), that in of their innovations of the Greeks after 620 delayed marriage for their upper class males from around 20 to about 30, but Saller astonished me by upping the age to 33 for all eras without any proof. They repeat the numerous errors they they have made in these two chapters of their's. They misinterpret Latin epitaphs to misconstrue the ages at which ancient Romans first married.

Astonishingly, they claimed that Latin males married on average at nearly 30 to females at 19 based on their “proof” that husbands' fathers on average ceased to be the main commemorators of husbands when those sons of theirs turned 28 and of their daughters until they turned 19, in both cases to be replaced by spouses. Arnold Lelis, Beert Verstraete and I showed in The Age of Marriage in Ancient Rome (Edwin Mellen Press, 2003), and I elaborated again on my website ( after Scheidel tried to refute us with later additions and improvements that marriage was, not relevant to Roman dedicators because fathers, as long as they lived, always dedicated the tombstones of their sons, whether married or not. As long as they lived, fathers also commemorated their daughters, even if married, unless those daughters of theirs had living offspring at the time of their deaths. If the wife had died before she left a living child, the husband had to return her dowry to her father who did the commemoration. If the wife died with at least one living child, her husband kept the dowry and set up the epitaph. In Ancient Rome, unlike in modern America, marriage did not make spouses the significant other of each other. In Rome, males married , as always thought before Saller and Shaw began to later misinterpret epitaphs a few decades ago, at ages 18 or 19 and females at 14 or 15.

The scholarly consensus is that if Greek or Roman parents reached 30, it was likely in normal time (that is ones without die-offs or excessive overpopulation) that they would live, on average, into their late fifties, at least for those having wealth, as the actuarial tables of Ulpian, for example, indicate. No agreement whatever, however, exists about ages at first marriage for either Greeks or Romans. After reading our book, Saller grudgingly admitted that in the highest class, girls married at 14 or 15 to males at about 24. Traditionally, scholars, led by German authorities from the 18th to the early 20th century, placed the average ages of first marriage for males at 18 to 19 and for females at 14 or 15, though law allowed males to marry at 14 and females at 12. Their children would be on average 30 years younger than their fathers whose own fathers died on average at 58 (30+28=58). Poor girls left the alimenta at 14 only presumably for immediate marriage. With a pregnancy every two or two and a half years and half of infants dying, it took a bride on average till 19 and a husband till 24 to have a living child. Even in the 1950s, Keith Hopkins of Cambridge University estimated that Roman males married in their early 20's, brides from 14 to 15.

However, the worst error in analysis by Saller and Shaw, more recently unequivocally endorsed by Scheidel, is having diligently and assiduously collected hundreds of epitaphs that showed that until 28, fathers predominated as dedicators of their sons and until 19 of their daughters, they concluded, or rather misconcluded, that upon marriage, spouses became the significant others of each other, and therefore erroneously claimed that first marriages occurred at those ages. (A modern and vague concept, not applying to the very formal tradition of Latin funeral dedications). Because no major journal would agree to print my thesis as a whole, Lelis, Verstraete and I resurrected the previous consensus about early marriages of both males and females, roughly quadrupling the previous precise citations of females at first marriage from 9 to about 40 and of males from 40 to about 160, whether we derived them from manuscripts or tombstones that were precise about ages at first marriage. On my website (, I since have presented much more material to disprove Scheidel's three attempts to dismiss out book (two on the Princeton\Stanford Working Papers in Classics website, Version 1.0 (Feb. 2007) followed shortly thereafter by Version 6.0, and "Roman Funerary Commemorations and the Age at First Marriage," Classical Philology vol 102 (2007): 389-402.).

Errors of facts and analysis also occurred in Sitta von Reden's "Classical Greek Consumption" pp. 385-408, which is both equivocal and vacuous. She asserts, "Painted pottery replaced silver and gold containers, and its mass production in Athens in the late sixth and fifth centuries, reflected the opening of the symposium and gymnasium to poorer people who emulated the former elite symbolic behavior." Poor Sitta couldn't even decide between erroneous theories, the first being outrageously incorrect, exactly the opposite of what actually happened, and the second the opposite of the first, also being incorrect because the painted pottery made thereafter was not designed for symposia, but for mainly for family dining.

On pp. 393-394, von Redden scribbles:

Despite restrictions put on individuals, per capita consumption of wine was probably significantly higher than, for example, in mediaeval times. To judge from the numbers of drinking cups and transport amphoras found in excavations, the consumption and trade of wine reached an unprecedented scale in the late archaic and early classical period.39 Most wine was marketed and consumed locally, but for connoisseurs it was shipped over long distances (see above, Chapter 13).40 Two explanations have been proposed for the increase in consumption. The first is that increasing democratization changed the symposium from an exclusive aristocratic gathering to a form of hospitality practiced by a wider group of citizens. Painted pottery replaced silver and gold containers, and its mass production in Athens in the late sixth and fifth centuries reflected the opening of the symposium and gymnasium to poorer people who emulated the former elite's symbolic behavior.4' The second is that the symposium remained largely aristocratic, while ordinary people frequented public taverns (kapeleia) that seem to have been abundant in cities and villages. Whatever theory is more valid, by the fifth century there was an extensive drinking culture supplied by shops, local markets, and foreign trade. It is worth noting that ordinary wine was called after the measure in which it was sold, so it was largely regarded as a commodity rather than a subsistence food.

As John K. Davies proves in Chapter 12, "Classical Greek Production," pp. 333-361, silver became plentiful after 480. In the days of Pericles, Nicias and Alcibiades, the Athenians actually controlled or at least sought to control not only the mines at Laurium in Attica, but also tried to control those near Amphipolis and also in the hinterland of Apollonia, the former colony of Cocyra, all of which brought about "a step change in the quantity of accessible bullion," followed by " the Athenian attempt in the 420s to impose the use of Athenian coins, weights, and measures throughout the [Delian] empire."

The large and heavy undecorated amphora exported full of wine and oil did indeed often return empty to Attica, acting as a ballast, but the decorated pots, in contrast, never returned and were clearly too light and not ever filled, in any case, to act as ballast, even on outward bound voyages. But this sale of painted pots was less significant economically than the export of silver bullion, so that cities without a source of their own silver could by importing bars strike their very own coins, which they did more for political reasons than for profits from minting. On the other hand, controlling silver mines, the Athenians profited significantly from minting their own drachmas.

Davies adds (p. 356):

However, we can also identify ways in which the classical period saw increases in money supply. One was via plunder from warfare out of region, the most substantial being the gains made at the expense of the Persian empire between 480 and 450. A second source was mercenary service out of region. This was an old custom on a small scale, but grew in importance from the late fifth century when both the Persian empire and its adversaries resorted to hiring Greek soldiers. A third comprised spasmodic, politically motivated consignments from non-Greek rulers, such as the payments made by Persia to one side or another for a century from the 420s till Alexander's conquest. Yet the impact of these three sources on the money supply was minor compared to new bullion from the silver mines. The main sources exploited in the classical period after the flooding of the workings on Siphnos were Laurion in south-east Attica, the Pangaion range by Amphipolis in the north Aegean coast, Thasos, and the hinterland of Apollonia. Unfortunately, the rate of bullion flow into the Greek economy cannot be reliably quantified, and certainly experienced high annual variation, while the routes by which it entered circulation changed during the period under review.

Von Redden's first hypothesis, derived from Vickers' extreme theories, is the more absurd. "In fact, "the Greeks" in general, and the Athenians in particular, after 480 grew much richer, especially in silver, than ever before. The Athenian elite, first went over to the use of silver at symposia after that date. Before then, only tyrants had used it, and not even very many of them could afford it before about 580, roughly the midpoint of the Greek Renaissance. References to the use of gold in Homer and from finds from before 1200 are very rare indeed, rarer still from the Greek Dark Ages (1200-800) and still very rare before 630. The best pottery was made for symposia between 630 and 480. Afterward, the elite supped and dined from silver. By that time, however, many of the not so rich, being richer than before, began to afford painted pots for the first time. For this large class, painted pots were produced in much greater quantities than ever before, but henceforth, to please a bourgeois dining en famille with wives and daughters. Only the elite, and only they, thereafter bought silverware, which I believe was often erotic. Her second theory is probably partially true, but hardly explains the huge amount of painted pots produced after 480. Some were doubtless for taverns, but most must have been for middle class homes which could thereafter have first afforded such luxuries on a large scale.

In chapter 10 "Archaic Greece," pp. 277-301, Rita Oswald dismisses the quarrel between Vickers and Boardman about the price of Greek ceramics with "fine pottery is unlikely to have formed of itself a complete cargo of any vessel, even those will admit who have championed a high value for it" (p. 285). The export a silver sympoticware after 470 could bring in much more money and, though occupying even less space than pottery, could and I think did seriously affect the balance of trade and payments -- to the great advantage of Athens. Unlike painted pots, silver vases mattered enormously but not as enormously as the export of bullion for coinage elsewhere or the drachmas struck at Athens itself, for the total value of exports.

In Chapter 13, "Classical Greece:Distribution," pp 362-384, Astrid Möller dodges the antitheses of Boardman and Victers, trying some sort of weak avoidance, alleging that some of the amphoras were reused in Egypt "for bringing water into the desert -- certainly not the only occasion for reuse." However, he does refute Johannes Hasebroke's "picture of the Greek traitor" as poor, foreign, and illiterate, and also "that at Athens, metics and foreigners exercised trade while Athenians financed it." He gets it right that "Under Pericles, state revenues were no less than 1000 talents a year (roughly 600 from the allies and 400 from Athenians sources.) at its height, (probably in the 430's B.C.) Athens had a reserve of 9700 talents of coined silver deposited on the Acropolis."

In short, many of the articles in this tome are full of confusing charts and graphs, signifying nothing. Such incredible repeated errors of chronology, geography and demography are shameful for both distinguished editors and for a press that in the not some remote past was exemplary for the accuracy of its publications. They represent a larger dangerous trend in classical studies of anachronistically simplifying and generalizing historical facts in order to portray cultures in a manner compatible with preconceived notions of how thing are now, and this must always have been, done by overly specialized scholars who fail to understand the overall context of Greco-Roman society, discount the work of their eminent predecessors,those who sought Altertumwissenschift, the total understanding of whole classical culture, and seek novelty for its own sake. Don't waste your money on this expensive volume of inaccurate analysis.

And finally, people, including Scheidel, have argued that my figures on Roman marriage ages are statistically insignificant, but a distinguished statistician and University of Massachusetts Professor of Mathematics Emeritus with a Ph.D. from Princeton and two textbooks about statistics to his credit, says that Scheidel's conclusions are statistically impossible. (Introduction to Probability with Statistical Applications by Giza Schay, Birkhäuser Boston, 2007, pp. 237-238)

Personal tools