This is the first volume of the long-awaited life’s work of Dominique Mulliez, who began studying the Delphic manumissions in 1980 and has published on them steadily ever since, in numerous articles that revised old readings, presented new inscriptions, adjusted dating sequences, and analyzed aspects of this centuries-long Delphic practice. Four volumes are planned: a second one, which publishes the remaining 619 inscriptions, appeared in the spring of 2023; a third, which promises to be extensive indices, complete with comments on its entries; and a fourth, an analytical synthesis, to appear under the auspices of the Bibliothèque des Écoles françaises d’Athènes et de Rome.
There is a certain feel of raw data to this volume, which has only a short Introduction (11–17) laying out the principles of chronological classification and epigraphical publication before plunging into the inscriptions themselves. The dating is done by the thirty-five Delphic priesthoods (priests served for life, 205 n.3), themselves divided into yearly archonships; these in turn were subdivided into half-years, the term of the bouleutes (first indicated in no. 162 [178/7 BC]; these officials did not become annual until the year 100 BC, 11 n.8), and finally organized by month. This scheme follows the dating principles set out by Georges Daux, whose results the author has been able to “nuance, make more precise, or correct” in some particulars (11). The major adjustment to Daux’s criteria for dating is Mulliez’s identification of the Delphic secretaryship as an annual office: Daux had thought it could be held for successive years (11–12). The structure chosen for the lemmata, not in itself innovative (as Mulliez says), is then laid out (12–16): description, bibliography, text (the Leiden system of transcription is followed, to which Mulliez adds four new sigla himself, two to indicate [vertical and horizontal] breaks of text over adjoining or damaged stones, one to indicate that an element of the formula has been skipped over [and therefore that the continuity of the text cannot be assured], and one to indicate a [parenthetical] insertion into the mostly standard formulae of the text). This is followed by an apparatus criticus, a translation (not in every case, given the formulaic quality of the texts; a helpful list of fully or partially translated entries is given on 655), and short commentaries on interesting or unusual aspects of the texts, where these exist. The apparatus criticus is especially meticulous at noting erasures, corrections, spelling variations of the same word in one inscription (especially vowels), how the cutters navigated around cracks in the blocks or gaps between them, and (of course) where previous scholars got readings right or wrong.
Each section of the book (= Delphic priesthood) begins with a discussion of the sequence of archons within that priesthood and a photograph of the first inscription in the series. The order of the priesthoods is not in doubt, and Mulliez flips only two archons in Daux’s sequence for priesthood I (39–40), while following (after discussion) Daux’s sequences for priesthoods II (49), III (139), V (333), VII (493), and VIII (513). The archon list for priesthood IV is adjusted more substantially (205–209), following a rationale based on external links, internal links, and spatial affinities; these arguments were previously published in 1998, and are here reprised with one correction and one modification. Daux’s sequence for priesthood VI is adjusted using many of the same criteria (371–376); Mulliez published his version (in which the dating of five archonships is certain and the other five are plausibly arranged) in 2006, with two minor adjustments to that argument here. Mulliez also notes that it was during this priesthood that inscribing expands from the great polygonal retaining wall (for the temple of Apollo) to include a supporting wall of unknown location (first is no. 502 [between 151/0 and 147/6]), and starts to fill in the section of the great polygonal wall behind the Stoa of the Athenians (375–376).[1] Finally, the sequence of priesthood IX was very uncertain under Daux (who felt he could do no more than simply order thirteen of the fifteen archons alphabetically, 543). Mulliez improves on this, assigning dates to five of the archonships, associating three more with the first half of the priesthood and the remaining eight (in alphabetical order) to the second part (547). All this matters because more precise dating allows attempts to connect the treatment of slaves and the behavior of masters (and the history of the sanctuary) to the wider historical narrative, and establishes a more firmly grounded chronology for the people who lived at Delphi or came from elsewhere. Moreover, other sorts of historical trends now become more visible: how inscribers used the walls and monuments; how the formulae of the inscribed documents change in small ways over time; how changes in the contents (such as the presence or absence of conditions placed on sales to the god) vary over time; and the number of years individuals in these inscriptions (now also linked to each other) were epigraphically active. These are not new questions, but the answers to them can now be more precise and more definitive.
It may feel like raw data, but what raw data it is! For those approaching this body of material for the first time, there is a multitude of interesting practices on offer, as well as many (implicit) human-interest stories behind these sales made to the god Apollo that entrust him also with the future freedom of the slave. Money changes hands (the exact place in the sanctuary is sometimes specified: nos. 1, 4, 16, 17, and 102). The sale can be compounded by a requirement that the slave continue to serve (perhaps serving but elsewhere, for which no. 108 is the sole example), or “stay and serve” (the more frequent phrasing, when this clause appears), or provide for, or perform funeral rites for, or crown funerary statues of, the former master, mistress, or even third parties (e.g. nos. 192, 193, 196, 202, 219, 255, 294, 316, 388, 492, 532, 617, 653, 671). When such service is complete, the slave is to be free to do what he or she wants, and to go wherever he or she wishes, although still (it would appear) the property of the god: does the god manumit, then, after a time? Or is the initial sale a “fictive” sale, and the slave freed immediately? What is the legal status of a slave in paramonē, the requirement of “staying”? These questions have provoked vigorous debate for more than a century. In this volume Mulliez holds his cards close, only intimating through translation (and the occasional comment, e.g. 131 on no. 131 or 502 on no. 572) that slaves (and explicitly those who are “of the god,” a phrase that appears a handful of times, nos. 1, 23, 28, 37 [hiera esto], 41, 44, 65, 143) during or after paramonē “sera propriété du dieu” and therefore suggesting that the sale is not “fictive.” Moreover, the sale to the god is to be guaranteed (if challenged) by seller(s) and their required guarantor(s), and punishments (sometimes a specified fine, but mostly left as “in accord with the law(s) of the city”) are exacted if they fail to do this; this reinforces the reality of the sale to the god and its terms. Exactly what a master holding an enslaved person in paramonē is allowed to do in the face of a slave’s non-compliance can be specified, but significant to this question of what the god’s “propriété” means is the requirement that the master cannot sell that (former) slave (nos. 202, 255, 357, 452 [where, uniquely, specific legal steps to stop this are laid out], 484, 489, 503, 555, 556, 567, 587, 606, 617, 638, 653, 658, 702; 673 [uniquely] specifies “no re-enslavement,” which is virtually the same thing, since former masters holding slaves in paramonē do not own them). If there is disagreement about the performance of the person in paramonē, the priests of Apollo (plus one other: nos. 1 and 4) or a commission of three (nos. 209, 246, 311, 388, 433, 496, 500, and possibly 563) can be empanelled to come to a decision. If a slave’s subsequent free status is challenged, anyone is “legally empowered” (kurios) to protect him or her, without fear of indictment or fine. In one case (no. 4), the seller swears a “lawful [nomimos] oath” not to harm two slaves, who swear a counter-oath to serve; this is the only time such oaths are mentioned in the entire corpus.
Behind these documents are human relationships. We can sense despair and exploitation: a mother was compelled to place her daughter into slavery (no. 32); slaves could be renamed (e.g., no. 355), sold multiple times (nos. 182, 184, 190, 211, 228, 414, 459, 462, 506, and 520 identify previous owners), or required to provide income through prostitution or some other means (nos. 205, 672). If a child is born to a woman in paramonē, the child will be free only when the woman is (nos. 259, 578, and 606), or a woman must provide two boys for the master before she leaves his service (no. 354, the first [160/59 BC] and only example in this volume, but this requirement will become more common later [308]), and in no. 372 if a woman wants to leave paramonē early, she must supply a girl “of the same age”). One inscription reveals that the slave had been either a war captive or kidnapped (no. 97: the only one in the corpus), while in two others the slave must repay the ransom her sellers paid to recover her (nos. 398, 722). In one case (no. 155, written by the master himself in koinē), a man requires that an enslaved woman accompany him to Macedonia before she is allowed to be free (why?). But one wonders whether kinder circumstances are not (also) sometimes possible, as when a slave is assigned to a third party “to be as a daughter to her” (no. 214; no. 397 is similar). Freed slaves can also inherit property from their masters (nos. 41, 532, 617, 671) or, in one case, be designated as epinomos (“légataire universelle,” no. 653). Some of the freed will sell their own slaves to Apollo later (nos. 110 and 214–386; 172 and 397; 214 and 436; 263 and 458 [and 522]; 337 and 532, 383 and 612), with their former masters as guarantors of the sale (nos. 242 and 501; in 103 and 202, and 263 and 500, the guarantor is the son of the former master); one master requires a freed slave to support that slave’s freed mother (no. 305). One also gets glimpses of the world of a great sanctuary when an overseer of the Pergamene king Attalus II’s property, travelling to Delphi some months after the king’s death, sells a “royal girl” (basilika paidiska) to Apollo (no. 8), and shortly thereafter a man from Pergamon, no doubt part of the same embassy, volunteers his services as a witness (no. 10). A member of another group, an embassy of Thessalian Amphiktyons, sells two men, allowing Mulliez (following Daux) to reconstruct the peopling of that embassy from their service as guarantors and witnesses (584–584, on no. 651).
Finally, Mulliez makes pertinent observations about the acts of inscribing and archiving. Republishing the Delphic manumissions is not simply a question of re-transcribing inscriptions on walls: Mulliez has also sleuthed in the storerooms to match unattributed fragments to existing inscriptions, as well as to reconstruct the (inscribed) stone mini-pillars of the “fence” that cordoned off the eastern terrace of the Athenian treasury. He also publishes here 26 manumissions that are either new or to which new pieces have been added: nos. 154, 194, 238, 261, 350, 381, 514 (in part), 528 (in part), 540 (in part), 558, 565, 588, 590, 595, 611, 613, 619 (in part), 621 (in part), 625 (in part), 638 (in part), 687 (in part), 693, 700 (in part), 703, 707 (in part), 719 (in part), and 721. Although Mulliez does sometimes adduce the physical relationships of sales close in time inscribed close to one another as a contributory argument in his discussions of the sequence of archonships, he also notes when (not infrequently) inscribers have not followed an expected chronological order in their placement of the inscriptions. The act of inscribing may itself have cost six drachmas, an amount Mulliez deduces from no. 479, which was the (meager) total price paid in this sale. There are intriguing explanations for later inscribing of these sales (no. 701: one month; no. 705: seven to twelve months; no. 602: two-and-a-half years), and even much later inscribing (no. 557; 568, on no. 632, on the wall of the analemma [retaining wall] of the theater), and the rare erasure of paramonē-clauses is also noted (e.g. no. 16). Archiving is a more complicated issue. When the practice is referred to, the ōnā (in these phrases meaning “sale-document,” not just “sale”) or a copy of it is mostly lodged with a person, although once (no. 596) specifically “with the god in the naos” (temple). If the ōnā was in the hieron, the sanctuary (e.g., nos. 223, 244, 245, 251, 274, 285, 414, 698, and 719), did that refer to these inscriptions, or was there a double-archiving system (256–257)? This issue is one of several (e.g., the relationship of the original document and the inscribed version [592], and some prosopographical arguments) to be examined in either the analytical volume (vol. 4 of this project) or the extensive index, with commentary on names (vol. 3).
There are thus innumerable riches in this first volume of the Delphic manumissions. Those who have worked with this material before are of course well aware of what this dataset offers, but may find the volume initially difficult to work with (or at least to correlate with previous work or work in progress) because we must wait for volume three to see not only the indices but also the concordances to previous publications. Other than that annoyance there is much to admire and very little to complain about in this book. I have found only 31 trivial (e.g., punctuation) typographical errors (and only one in Greek: 635.1) in 657 pages, which must count as some kind of miracle, given the density of the text. I do note in addition a misplaced heading (p. 64: the archon should be Diodorus [L4], not Emmenidas [L2]), and 33 other errors that mostly appear in misquotations or mis-citations of English and German works, such as the fact that the author listed as “Wayne, Tucker C.” is actually named “Wayne C. Tucker” and so belongs under “T.” Such trivial errors in a big, highly detailed book that must have been highly complicated to typeset are (thus) to be noted but, in the end, not important. What is important is that these inscriptions are now properly published, that the remaining volumes will appear, and that scholars can dive into wonderful material whose texts are secure, whose placement in the sanctuary is established, and whose dates are now clear or very plausible, all with the wise guidance of Dominique Mulliez, who is to be congratulated on his magnificent achievement. I look forward to burrowing into volume 2—now out—when I next have three free months, and apologize to the author for the tardiness of this review.
Notes
[1] The dating of no. 380, inscribed on an unknown treasury, is tentatively listed as 158/7, but the restoration of this fragmentary inscription and its dating formula is extremely uncertain, and therefore one cannot “draw any conclusions about the chronology of places of inscribing” from it (327). No. 398 (156/5) is “for a long time” an isolated inscription on the polygonal wall that protects, to the north, the terrace where the Athenian treasury is located. No. 432 (154/3), on the north wall of the Athenian treasury, is unfinished; no. 440 (153/2) is a later reinscription.