Dennis D. Hughes, Human Sacrifice in Ancient Greece. London: Routledge, 1991. Pp.xiv, 301. $42.50. ISBN 0-415-03483-3.
Reviewed by Richard Hamilton, Bryn Mawr College.
This revised dissertation (Ohio State 1986) achieves its goal "to collect, organize, and present ... and then to evaluate the evidence for the ritual killing of human beings in ancient Greece." The collecting is voluminous: about twenty-five archaeological sites are discussed in detail and over a hundred literary testimonia, this alone justifying the price of the book. The evaluation is less systematic: the basic tendency is to discount the possibility of ritual killing, with the usual, mostly reasonable array of arguments: insufficient evidence, unreliable (poetic, late, tendentious) testimony, incomplete archaeological report. What we lack is any consideration of the counter-argument, which might begin with the Carthaginian material, which shows what a combination of archaeological and literary evidence can look like in an undisputed case.1
The archaeological evidence is judged doubtful in virtually every case. (1) At the Minoan "sacrifice" at Anemospilia, featured in National Geographic in 1981, the building may not be a shrine while the blade, which is actually a spearhead, may have fallen on the boy from above; his leg may be accidentally curled not bound since the hands are not bound; and the difference in bone color is not due to blood loss. (2) The scraped bones at Knossos, especially in the Room of the Children's Bones, may signal cannibalism but may also show preparation for burial or reburial. (3) Late Bronze Age Burials in the Argolid: the Dendra tholos tomb is possibly "suttee" but maybe "simultaneous death of a man and wife from disease, accident, or violence"; the tholos tomb at Kazarma is not published fully enough to tell; the claim of human sacrifice for three of the four Argolid chamber tombs "is open to serious question" while the fourth, an apparently simultaneous burial of six persons "above the door of Tomb 15 in Mycenae ... may represent some sort of funerary ritual killing." (4) Seven of the nine alleged slave burials on Cyprus only show that "occasionally the citizens of Geometric Lapithos buried their dead in the doorways of tombs" but in two cases (Salamis Tomb 2, Lapithos Tomb 422) the figures buried in the dromos were bound and so qualify as "ritual killing." (5) Three Middle Helladic "dual burials" of a male and female do not look like "suttee" when set against the full array of burials at Lerna (228 graves, 16 with more than one burial, eight or nine being "dual" including the dual burial of two males). (6) The spectacular burial at Lefkandi, however, does, at least as far as one can tell from the preliminary publication.
The literary record is subjected to the same sceptical inquiry. First funerary killings: many of these are simple cases of revenge killing; some seem to be literary invention (Evadne, Polyxena, Lucian de luctu 14), and none is "a killing performed in a particular situation or on a particular occasion..in a prescribed, stereotyped manner, with a communicative function of some kind." [But if revenge killing is stereotyped, can it be excluded? Conversely, would war killing not fit the definition? H's use of "sacrificial" vocabulary as an important criterion is undercut by his own discussion of Herodotus' rather flexible language.] The Cypriot ritual killings (above #4) were probably revenge killings, as we seem to find at Istria where "three Archaic tumuli built over central pyres, contained peripheral burials of humans and horses." H then considers the testimonia to "mythical" and historical human sacrifice, concluding that the mythical ones can be dismissed as unhistorical and the historical ones as pseudo-historical. The sacrifice to save the city fits too regular a pattern to be historical; youths given to monsters (e.g. the Minotaur) are not technically sacrifices since they are exposed not killed and are better termed folktales than myths; initiation rituals of Artemis, which are repeatedly attested, bloody but do not kill. (Herodotus' confusing description of the Athamantids at Alos is said to describe a ritual drama of initiation, depite the author's earlier, correct warning against haphazardly labeling anything an initiation.) Military sacrifices either were not carried out (Agesilaus at Aulis, Pelopidas at Leuctra) or are from a fanciful source (Phainias on Themistocles, following Henrichs). The cult of Zeus Lykaios, where alone human sacrifice is repeatedly attested, is dismissed because excavation has revealed no human remains at the site. [Could they not have been intentionally removed?] Thus, although the list is long it can be drastically shortened. Moreover, "for more than half the historical human sacrifices we are indebted to two writers...of late antiquity" whose reliability H effectively undermines by showing what might result "if some writer of late antiquity had only Pausanias as his source." H concludes that tragedy was instrumental in promoting the idea of human sacrifice: "no human sacrifices in the Homeric poems...a few then in other early epic; in the fifth century, a sudden burgeoning of human sacrifices on the tragic stage; then, beginning in the fourth, the creation of new human sacrifices by historical writers ... and finally, human sacrifice as a convenient narrative device in the novel... It is little wonder that the belief that the 'ancients' practised human sacrifice is now firmly entrenched." [One might object that we have hardly any literature other than Homer before the fifth century, and Homeric epic may not have room for human sacrifice.]
The book ends with chapters on the pharmakoi and the Locrian maidens, where the evidence for actual killing is almost non-existent, and with appendices on Pylos tablet Tn 316 (hardly evidence for human sacrifice) and on other cut marks and mass burials (mainly in wells). The mostly bibliographic notes run to fifty-five pages, the bibliography to twenty. [I missed only Hastings' still useful Encylopaedia of Religion and Ethics (1908-26) Eliade's much newer Encyclopedia of Religion (1987), A. Henrichs on wineless libations HSCP 87 (1983) and Early Greek Cult Practice, ed. R. Hägg, Acta Atheniensis 1988.] There is a subject index and an index locorum.
The general conclusion is that "at present very few connections can be made between the written and the archaeological evidence.... In the past, the written evidence ... seems to have had an influence on archaeologists, who quite naturally expected confirmation to be forthcoming from the soil. But the curious and varied collection of texts studied here ... is more a testament to the capacity and breadth of the imagination of the Greeks than a documentary record of their practices." A depressing conclusion for those of us who believe that a future for Classics lies in bringing archaeology and literature closer together.
 See now Shelby Brown, Late Carthaginian Child Sacrifice (Worcester: JSOT Press, 1991).