Bryn Mawr Classical Review 94.10.19


D. R. Macdonald, Christianizing Homer, the Odyssey, Plato, and The Acts of Andrew. New York: Oxford University Press, 1994. ISBN 0-19-508722-4.


Reviewed by Robert Lamberton, Washington University (rdlambert@artsci.wustl.edu).

The Christian appropriation of Greek polytheist culture in the second century of the common era was, on the whole, not a pretty sight. The principal players were ham-fisted, self-styled philosophers of the order of Justin Martyr and Tertullian, whose claims to teach philosophy amounted to little more than the eviction of the traditions of Greek philosophy from what they defined as the search for truth, and their replacement by a monotonous, scriptural rhetoric, professions of faith, and such inane and ultimately useless equations as "Christ is the logos". Few and far between are the Christian texts that bear witness to any depth of knowledge of polytheist texts, whether philosophical or literary. All in all, the century and a half between the day Paul quoted Aratus to the Areopagites and the time of the confrontations with Greek tradition of the scholarly Alexandrians Clement and Origen offer little to suggest that the nascent Church found time to read the classics or put on a veneer of culture. It appropriated what it could and trampled the rest -- the bulk of the demon-ridden culture of its paranoid vision -- into the mire. The texts of the period are grim and shrill, and even when we reach the richer cultural atmosphere of Alexandria and the higher intellectual standards of Clement and Origen, we search in vain for genuinely protreptic texts, seductive texts that attract rather than proselytize, invite imaginative and intellectual engagement, rather than belabor the all-too-familiar threats, warnings, and injunctions. There is little to give the lie to Lucian's description of his Christian contemporaries: benighted, gullible "poor bastards (kakodaimones) who've convinced themselves that they're going to live forever" (Peregrinus 13). They seem to have been the sworn enemies of any possible pleasure of the text. They generated what is surely the most unsympathetic, in-your-face literature in the Western tradition.

Dennis Ronald MacDonald has been working for some years on a text that goes far to counteract this picture. If the lost original of the apocryphal Acts of Andrew was anything like what he claims it was, and if it was in fact composed in late-second-century Alexandria, then we will simply have to acknowledge that a second-century Christian could and did produce a tale of wit, fantasy, and sophistication, weaving into it themes, motifs, and whole episodes from Homer and Plato and "transvaluing" them into a Christian romance, a deliberate and self-proclaiming fiction of a richly rewarding sort. In his new book, MacDonald presents his reasons for believing that the Acts of Andrew was such a text. I have serious doubts about a great deal of what he claims, but beyond the range of my scepticism enough remains in his arguments to make this an important book that anyone concerned with the literature of the high Empire should read.

Let us first be clear about what we are dealing with here. The New Testament apocrypha as a whole are a textual critic's nightmare, and the text known as the Acts of Andrew (the brother of Paul, an obscure figure in the canonical NT, but in the apocrypha designated apostle to Achaea) has not been seen intact since the ninth century. By that time, versions of it were in circulation in Coptic, Syriac, Armenian, and Latin, representing states of Andrew's story that predate the surviving Byzantine Greek versions. Especially important to all reconstructions of the original is the Latin epitome composed ca. 593 by St. Gregory of Tours (Miracula sancti Andreae). The task of collating all of this material was undertaken by Joseph Flamion early in this century (Les Actes apocryphes de l'apôtre André. Louvain, 1911), and two reconstructions of the "original" Acts of Andrew, presenting the relevant sources and providing translations, have appeared almost simultaneously in the past few years: MacDonald's own (The Acts of Andrew and the Acts of Andrew and Matthias in the City of the Cannibals. Atlanta, 1990), and that of Jean-Marc Prieur in the Corpus Christianorum, Series Apocryphorum (Acta Andreae. 2 vol., Turnhout, 1989). MacDonald spells out the relationship between the two reconstructions in the introduction to the new volume, and at somewhat greater length in that to his own edition. Aside from Prieur's more extensive apparatus and more complete descriptions of the source materials, they differ principally in their treatment of the story of Andrew's rescue of Matthias from the khora ton anthropophagon, an anonymous locus in the Greek versions, but in the Latin (and the Anglo-Saxon!) traditions of the tale designated as "Myrmidonia" or something of the sort (35). Gregory's sixth-century epitome included the story as its opening episode and called the home of the cannibals Myrmidonia (or Mermidona or any of five other variants). Flamion, followed by Prieur, considered The Acts of Andrew and Matthias in the City of the Cannibals to be distinct from the original Acts of Andrew, an accretion already welded onto the Acts in the Latin translation epitomized by Gregory. Prieur therefore omits the episode (with the result that the only recent edition of most of the Greek text, with translation, is to be found in MacDonald's edition of the Acts). Thus the reader of MacDonald's reconstruction will encounter a quite different, longer, more fabulous version of Andrew's adventures than the reader of Prieur's.

Whichever reconstruction one takes to better represent the original Acts of Andrew, it remains the case that Andrew's travels received more fabulous, novelistic, and esthetically attractive treatment than those of the other apostles of the apocrypha. It is nevertheless quite important for MacDonald's central argument that the cannibal episode be integral to the original work, and that their land bear the allusive name Myrmidonia as a deliberate hint to the reader, revealing the author's "hypertextual intentions". This is the episode that provides the Iliadic background for the author's appropriation and "transvaluation" of the Odyssey, and that single toponym (absent from all Greek testimony) is perhaps the most persuasive bit of evidence in MacDonald's large arsenal to support the proposition that this Christian text presented itself to its original, Christian audience as an explicit and deliberate reworking and "Christianizing" of Homeric material. Here, Andrew the Christian anti-Odysseus is summoned to the land of the savage, cannibalistic, (and, of course, polytheist) Myrmidons -- remember the Myrmidons as blood-dripping wolves (Il. 16.156-63) and Achilles' observation that he'd like personally to devour Hector (Il. 22.346-47) -- to rescue Matthias, whose mission there has left him imprisoned and being fattened for roasting. Circe-like, the Myrmidons give a drug to their captives, with the result that these unfortunates accept hay for food, facilitating the fattening process and marking their transformation, at least for culinary purposes, into beasts. Odysseus-like, Andrew is put ashore asleep -- not by Phaeacians but by angels, commanded by the ship's captain, who is Jesus in disguise. Andrew liberates Matthias and the other inmates, encounters an old man who offers the butcher/executioners his young son and daughter in order to save his own aging skin -- his similarity to Agamemnon is clearer to MacDonald than to me -- saves the children, is arrested and dragged (Hector-like) by a rope around his neck through town. Jesus comes to him in prison and patches him up, allowing him to flood out the ant-men by means of a water-spewing statue (dubiously compared to the Scamander of the theomachia), aided by the Archangel Michael in a fiery cloud, preventing their escape. Andrew sends the cowardly old man, along with the fourteen butcher/executioners, off into "the abyss" (though only provisionally), then revives and converts all the victims of the flood, and sets out to return (if, indeed, this is a nostos of a sort, as MacDonald claims) to Achaea. He will eventually be crucified in Patras (a city selected, we are told, in the absence of supporting evidence for Andrew or any other Christians there until a much later date, for its proximity to Ithaca).

I hope that my rather flippant paraphrase of MacDonald (who in turn has a tendency to proceed in much the same manner with the texts in front of him) will not appear unfair. It is intended, in any case, to emphasize both the strengths and the weaknesses of MacDonald's presentation of his arguments and his claims of "hypertextuality", each of which must stand or fall on the merits of its details. Some of the episodes in this fabulous tale do indeed sound a lot like echoes of Odyssey episodes, and they are enriched, for a reader capable of making the connections, by the reminiscence of those Homeric episodes. At the same time, a large number of the prototypes MacDonald finds lurking behind figures in the Acts of Andrew seem to me suggestive analogies at best, and quite a few of them have nothing at all to do with Homer. As has often been noted, Homer's Circe has a lot in common with the witch in the deep, dark forest encountered by such hikers as Hansel and Gretel -- and there's no doubt what all those little gingerbread boys and girls are for -- but at the same time, Homer's Circe is emphatically not a cannibal, any more than his Agamemnon is a child-sacrificer.

MacDonald's book is largely made up of claims similar to those evoked above, and as such defies adequate discussion in the scope of a review article. He makes no secret of the fact that he anticipates that many of his "similarities ... may seem strained", but insists that their occurrence "in the same order as in the epics" (39, cf. 307) constitutes evidence of the appropriation of specific texts by the author of the Acts of Andrew. In fact, he comes dangerously close to claiming that a large enough accumulation of marginally plausible "similarities" constitutes a demonstration of his thesis, which of course it cannot, and the arguments based on sequence are generally open to alternative formulations and interpretations. The verbal similarities lined up in his parallel passages are rarely compelling and his claims regarding proper nouns as hypertextual indicators generally fall flat. One might argue, of course, and with justification, that ancient "etymological" extraction of meaning has nothing to do with historical linguistics, but MacDonald's claim (for example) that "the name Varianus derives from the Latin participle varians and means 'changeable' or 'unstable'", and so might be translated "Mr. Fickle", "indeed an apt name for Zeus" (136) is a perverse distortion of the force of the adjectival form of the common name Varus -- an equivalent hermeneutic ploy might be the claim that every Germanic proper name ending in "-son" or equivalent endowed its bearer with solar attributes.

The figures in the Acts of Andrew also have an alarming habit of switching their Homeric identities in ways few readers, I think, could possibly follow. Thus Andrew, back in Patras, after curing the wife of the proconsul Aegeates (= Poseidon, Odysseus' old nemesis, since Aegae is associated with the cult of Poseidon), then converts her while Aegeates is out of town, so that she becomes an anti-Penelope, living out the consequences of her chastity to their (second-century Christian) limit and refusing to have sex even with her returning husband (now himself Odysseus). He then takes outrageous and unpopular revenge on the successful suitor (Andrew) and crucifies him on a cross stuck in the ground by the sea (= both Odysseus' mast and the oar he must "stick" in the ground far from the sea in Tiresias' prophecy, here filtered through the sort of salvational allegorizing the neoplatonist Porphyry, about a century after the date here proposed for the Acts of Andrew, practiced when he made that prophecy into the final cause of the action of the Odyssey).

It is only in his Conclusion that MacDonald finally presents arguments that attempt to prove, rather than simply illustrate, his hypothesis (302-316). His claim is that by the criteria of "density and order", "explanatory value", "accessibility", "analogy", and "motivation", the Acts of Andrew is best understood as a hypertextual transvaluation of the Odyssey, with a generous dash of Plato and other polytheist texts mixed in. The first issue turns, as I've said above, on the specifics of each "similarity", and there are just too many loose ends and dubious equations. Order proves nothing if the parts do not hold up. Nevertheless, MacDonald proves alarmingly (and sympathetically) willing to concede a great deal here (307): "Even if one were to disqualify three-fourths of the evidence argued for here, the remaining quarter would demonstrate more investment in transforming Greek mythology than one can find in all previous Christian narrative combined." Now, when we are talking about "mythology" -- about stories, story-patterns, motifs, situations -- MacDonald is much more persuasive. Simply to establish what is claimed in this sentence would be a great deal, and it could be accomplished quite efficiently. Frankly, the gymnastics required to argue MacDonald's further claim of "literary dependence" -- that a Christian author with certain books of the Odyssey in front of him proceeded episode by episode to rewrite them as tales of the apostle of a benevolent God victorious over the demons of the polytheists -- look in retrospect like overkill. If he feared that three-quarters of his specific "similarities" might be suspect to a sceptical reader, then the better plan would have been to scrap them and build a stronger and more persuasive case on what was left.

The brief discussion of "explanatory value" (308) does not get us far beyond the assertion that a number of decided oddities about the Acts of Andrew can be resolved "by reading the text against Euripides, Plato, and, above all, Homer" (308). In general terms, this seems to be the case, though I find it far easier to believe that the stories, the myths we know best in the literary form given them by Homer and Euripides, are here subject to appropriation rather than any specific literary version of those stories. The case of Plato is rather different and there is little doubt that a certain number of motifs do enter the latter portion of the Acts of Andrew from the dialogues -- most obviously, the metaphor of the philosopher (= apostle) as midwife, from the Theaetetus, and the figure of the imprisoned Socrates, lecturing to the last, from the Phaedo, clearly the prototype for the portrayal of the imprisoned Andrew (midwife: 218-22, 234-36; Andrew as Socrates: 252 ff.).

This brings us to the matter of "accessibility", third in MacDonald's list of criteria of "literary dependence" above. MacDonald's touchstone here is Clement, quite credibly the best-read Christian of his time, and the assertion that he knew numerous dialogues of Plato as well as Homer and Euripides needs no defense. He cites them, criticizes them, appropriates what he can and condemns much of what is left. But the author of the Acts of Andrew would not have needed to know a single one of those texts firsthand to have known enough about them to generate this novelistic biography. He was capable of reproducing the rhetoric of Christian Platonism in the speeches he wrote for the imprisoned Andrew, but there is little danger here of confusing mimetic skill and true knowledge. Furthermore, the figure of Socrates that we get in Plato's Phaedo was probably a topos by the time Plato wrote the dialogue (as Socrates' Apologia was not a single composition but a genre in the 390s). The evidence from the period of the Acts of Andrew, in any case, is abundant, both among polytheists (Lucian Peregrinus 37) and, as MacDonald shows, in Justin Martyr and Tertullian (249-51). In the rich resonances of the style of the Second Sophistic, there could hardly be a representation of an imprisoned philosopher who was not in some sense Socrates -- so much so that the penetration of the topos into Christian texts is hardly a surprise.

Under the rubric of "analogy" (310-14), MacDonald points to comparanda that increase the plausibility of his hypothesis. He correctly observes that the closest analogy to his text is to be found in the polytheist romances, permeated in turn by the influence of the Odyssey and of Euripides. Still, the romances we have do not appropriate Homeric or Euripidean material in the way MacDonald claims the author of the Acts of Andrew has done. Their appropriation, first of all, is purely esthetic and literary, and more important, it can rarely (if ever) be reduced to parallel passages and narrative imitation. Furthermore, we must ask again if even this sort of adaptation and appropriation, coupled with a "transvaluation" of the original, can be conceived as a Christian project of the second century. The only Christian appropriation of Homer MacDonald can point to at such an early date is the centones, and one would like a more convincing one. MacDonald promises (313) "a description of the role the Odyssey played in the Book of Tobit, the Gospel of Mark, and the Acts of Luke" and such a display would indeed lend plausibility to his claims about the dependence of the Acts of Andrew on Homer, but I remain sceptical that any such dependence can be shown.

Are any generalizations possible from this rich accumulation of detail? Prieur, in the section of his Introduction to the Acts dealing with "sources and literary influences" dispensed with "la tradition et la littérature antiques" in less than 150 words (403-4). MacDonald effectively expands his treatment of the same subject to 350 pages. Both cannot be correct and neither is wholly so, but we can say with certainty that MacDonald has sensitized us to an aspect of this wonderful text -- perhaps its most important aspect -- to which Prieur remains almost entirely insensitive.

Under his final rubric of "motivation", MacDonald reasserts his repeated hypothesis that we have here an author whose "thorough rewriting of classical Greek literature" was intended to urge "the reader to compare the worst of traditional pagan religion with the best of Christianity and to choose the latter over the former" (316). That was certainly a goal of second-century Christian intellectuals, from the heavy-handed Justin Martyr to the learned and subtle Clement. It would be surprising if the author of the Acts of Andrew did not have something of the sort in mind. What remains in doubt is whether he went about his task with his eye so firmly fixed on prominent polytheist texts as MacDonald would have us believe -- but what we can be more certain of, thanks to MacDonald's work, is that in what John Dillon (Middle Platonism [London, 1977], 379) called the "sea of bores" of the late second century, there existed alongside the sparsely distributed polytheist "islands of wit" at least one modest Christian archipelago.