r IS THE ESCORIAL AKRITES A UNITARY POEM ? This paper covers ground that has been much trodden by others, and in attempting only the broadest sort of answer to the above question its aims are modest ('). But it will be evident, I hope, that there is room for the question to be asked, if only as a small contribution to an important debate (2). The figure of Digenes Akrites has undergone many changes through the centuries, but in the view of this writer, the most momentous transformation took place at the point at which five poems relating to the hero were gathered (perhaps in the twelfth century) into a manuscript which is the ancestor of the Escorial MS known as E. It was at this point, and not before, that Akrites acquired a biography. In stressing this version to the exclusion of the Grortaferrata version (G), despite the lateness of the MS, it will be evident that I am following the arguments of Professor Stylianos Alexiou that E, not G, is closer to the archetypal Akrites ; the most persuasive being those based on the presence in E, and E alone, of proper names from the Eastern borders, military terms and other lectiones difflciliores which are passed over or smoothed over in G (3). An aspect of Alexiou's (1) This paper is based on a lecture delivered to the Faculty of Medieval and Modern Languages at Oxford in October 1987 ; much of the material was presented piecemeal to the Byzantine Text Seminar at Binriingham University in 1987-88 ; I am very grateful to colleagues for their questions and comments. I owe a particular debt to those who took the trouble to read a draft of this article : Prof. S. Alexiou, Dr. R. M. Beaton, Dr. D. W. Holton, Prof. M. J. Jeffreys and S. MacAlister. Naturally I am responsible for any errors of fact or emphasis that remain. (2) This article takes as the basis of discussion the edition of S. Alexiou, BaoiAeioc Atyenfc Axpitnc (Athens, 1985), to the extent that it may almost be considered a review-article. My references to the huge secondary bibliography are sparing : I restrict myself largely to newer or supplementary material. On the fortunes of Akrites in later literature we now have a lucid treatment by G. Kechagioglou, "Tuxec, trig PuCavrivflg axpmxrig rcoir|ot|g orr| veoeAXnvixn. XoyoTEXvia: oraOuoi xcu XPnoeig. A7iOTiuiioei(;, 'EUnvixa, 37.1 (1987), 83-109. (3) S. Alexiou, 'Axpmxa (Herakleion, 1979). It should be noted, of course, that Alexiou's view has not found universal acceptance .- for a sophisticated alter- about the escorial AKRITES 185 treatment in his monograph Anpmxh which is particularly thought-provoking, and potentially revolutionary for the study of the subject, is his hint at the possibility that what we have in E may not be a single poem at all (*1). In 1985 Alexiou published his edition of E, the first which actually makes the work readable (s). The student of Byzantine vernacular poetry is tempted to breathe a sigh of relief and take the view of Jowett: "Don't dispute about texts. Buy a good text." (6). With great ingenuity and good sense Alexiou has largely restored the text metrically and explicated numerous difficulties. And yet the very readability of this new E, and the fact that the reader now has a starting-point for reflection, brings some new problems to the fore. My reservations concern, not so much the surface of the text - on this one might sometimes be more conservative - as its structure (7). It is perhaps the case that, in arguing for E's authenticity, Alexiou has gone native account developing Alexiou's findings in another direction see R. Beaton, The Medieval Greek Romance (Cambridge, forthcoming ; I am grateful to the author for showing me a typescript in advance of publication). In general, I would by no means to exclude the possibility that G contains authentic material from the archetype (see esp. N. Oikonomides, "L"epopee' de Digenis et la frontiere orientale de Byzance aux x* et xf siecles", Travaux et Memoires, 7 (1979). 335-397) — but here I draw solely on E for such evidence. (4) Alexiou, Axpitixa, p. 87 : "The text does not appear organically unitary, and perhaps we shall be able to show that it consists of more than two (Emir-Lied and Digenes-Roman) parts". See also his article, "'0 Aryevis|c, 'Axpixric. tou •EoxopidX", npaxttxa rfjc Axadnpiaq ABtjvwv, 58 (1983), 68-83, esp. p. 80. (5) The first published edition, by D. C. Hesseling ("Le Roman de Digenis Akritas d'apres le manuscrit de Madrid», Aaoypcupia, 3 (1911-12), 537-604) looked chaotic and deterred further study ; that of P. Kalonaros, BaoiAetoc Aiyevt)c 'Axphag, ra fyperpa xetpeva (Athens, 1941), made the text look uninviting by placing it among the more readable versions. (6) In G. Madan, Notebooks (ed. J. A. Gere and J. Sarrow, Oxford, 1980), p. 61. (7) On questions of the Alexiou text see the review by Beaton in Journal oj Hellenic Studies 106 (1986) 271-273 and the same author's article, "AxpiTqg xai ot xpmxoi: cpiAOAoyixA xcu exSonxd npoPXriuara", in H. Eindeneier (ed.), Neograeca Medil Aevi (Cologne. 1987). pp. 75-84. It may be thaton a very few occasions Alexiou has wrongly deleted from his text: line 763, fie ra xaAhfra rov, may be worth keeping in order to stress that the hero goes out just on foot against the beasts j compare the phrase in "Armoures" 96a. ne&c. pi ra yovaria. (Perhaps E 763 originally read ne(bc pi to xaXlxiv. cf. E 1323). 186 d. ricks about the escorial AKRITES 187 rather further than he ought in special pleading for its unity. (Although remarks made since tend to modify this.) (8). Authenticity and unity are not the same thing, and in arguing that E is not just the disjecta membra of G or a precursor of G we must be careful to avoid the elision to the claim that E, though lacunose, is essentially a whole. It is indeed my view that, now that we can see E the more clearly, it is the clearer that it is not a poem at all but a collection — and that the attempt to bind the poems together is perfunctory and superficial. If I am right, the consequences will stretch far indeed — as much forward, to the after-life of the work, as backward, to its genesis. Locally, then — to the constituent parts of E — Alexiou has done an inestimable service: he has enabled a fairer judgement of their poetic quality, and no scholar will be able to ignore the work from now on or dismiss its virtues as the product of chance (9). The text has been largely liberated from an inadequate scribe and over-cautious editors without the imposition of a spurious uniformity: thanks to Alexiou our knowledge of medieval Greek continues to expand (I0). But the view that E, with its glaring incompatibilities of plot and ethos, is a single work implies a readership whose taste for quite sophisticated verse is accompanied by a blindness to contradictions which can only (8) On the question of structure Alexiou's remarks are rather brief. Only once is the question of pre-existing material raised (p. pP' n. 91) and the arguments on pp. xp'-Xy' are not compelling .- that each part of E begins with a name and a gnome and ends with a significant outcome does not entail any organic relationship between the parts. Alexiou in fact concedes that especially the first three sections "have a certain self-sufficiency and could ... be read or recited singly"; and by way of conclusion (p. pX0') concedes with understatement that E is "not entirely unitary" — leaving the door open for the point of view argued here. Since completing this article I have been able to note a development in Alexiou's views in the direction I follow here : in his valuable article on his editorial policy (Ticc Tnv ex5oor| tou AxpiTq xai too Apuoupn". MavraroipSpoc;, 25-26 (1987), 57-62) he now admits the possibility of more than one poet. (9) Like J. Mavrogoiujato (ed.), Digenes Akrites (Oxford. 1956), p. xix : the reaction against nationalist scholarship was needed, but one is entitled to be suspicious about the view that poetic beauties come by chance. (10) Alexiou, pp. oC-ttP'. There are occasions when a little more consistency in the text may be required, or the lack of it observed, e.g. 859 and 865 : 8ev ri&bpeu;, opp&na fiov, rd <; rwv 6v fiov xi eyib ij&vpw, opfi&na fiov, rd 0aae neaelv 6 npwrog /lovo/iaxog. (39) I. Karagianne, '0 Aiyevr\g 'Axpixr\g rov Eaxopiak, avufioAri axr\ fiekexn rov xeifievov (Ioannina, 1976), pp. 225-263. (40) See the list, section (ii) b, above ; for the original lay as a possible reflection of social tensions between akritai and landowners see G. L. Huxley, "Antecedents and Context of Digenes Akrites", Greek Roman and Byzantine Studies, 15 (1974), 317-338. (41) The beginning is hopelessly muddled and unmetrical, so much so that it may just have been cobbled together by an earlier scribe faced with an akephalic MS of section A' and made worse still by the semi-literate scribe of E. (The presence of the dragon could be deduced from the next bit, but the earlier part of the story was lost attempt at weaving a number of exploits — perhaps the bulk of the exploits known to traditional lays — round an Akrites who is modern enough to love his girl: the girl forms the narrative backbone as she provokes the successive attacks of ogre, lion, apelatai and Maxi-mou (42). The narration is in the first person, suggesting origins in a distinct genre ; but the beginning has been lost (43). Nonetheless, we may speculate with some confidence, that the beginning in E represents an attempt to cobble together a beginning for the story which was already lost by the time of G, and that the original story, in harmony with the fifth poem, relates the events that occurred on the young couple's journey home from the abduction. In other words, this fourth poem is rather an alternative version of the third (and this does something to explain why the third poem, the work of a sophisticated poet seeking emancipation as much from earlier acritic poetry as from the current vogue for Troy, excludes the apelatai altogether) (44). With the first person narration comes a quite different ethos from that of the third poem: after having his way with the defeated Maximou, Akrites reports back to his wife with a no doubt proverbial punch-line (1596-1598); husband and wife enjoy the joke together (45). We are some way from the pledges of troth in the third poem. 5. The fifth part of what has been arranged as a sort of biography of Akrites in a way which I cannot believe to have taken in any audience but which has gripped scholarly opinions, concerns the retirement and death of Akrites. Alexiou treats the two as separate, but for ever, and the idea of an unmotivated excursion of the hero and his loved one looks desperate ; nor is G any the wiser.) (42) E 1106, 1122, 1149, 1397. (43) Trapp's view, endorsed by Beaton. "Axpirrig", p. 79, that this is in imitation of the Odyssey holds for G, no doubt, but not for E i we do not have to suppose a classical origin for an embedded first-person narrative of this sort. As Kcr remarks (Epic and Romance, pp. 110-111) "In English poetry there are instances of stories dramatically introduced long before the pilgrimage to Canterbury. In Beowulf 'there are various episodes where a story is introduced by one of the persons engaged." (44) Except for one hint - teasing and appeasing the audience — when the hero prays to St Theodore rdv fieyav dnekarnv (891). We could not have a clearer indication, incidentally, that the notion of Akrites' being the implacable foe of the apelatai is utterly mistaken : he is a figure which is poacher turned gamekeeper. (45) With Alexiou's note, but not his text, 1 include 1599 in the final third-person statement. 198 d. ricks about the escorial akrites 199 I doubt if this can be correct; although the corresponding sections even of G are somewhat perfunctory, the ekphrasis of the castle there has some point, while in the E version it stands alone (and disproportionately brief); furthermore, the mention of the hero's tomb cannot but be placed here in order to lead on to the hero's death C6). (Here a rare and cautious recourse to the evidence of modern Greek folk song may confirm this view (47). The Pontic versions of the death of Akritas have the hero building his castle and garden ; the birds in the garden warn him of impending death ; and then Charos comes to take him (48). This song, with Charos apparently an innovation, is to be seen, it is to be stressed, not as a prototype of the written poem, but as a reflection, however abbreviated and distorted, of a manuscript tradition ("').) No poem of the Akrites collection is so easy to analyse into the parts of which it has been made ; but for this very reason our (46) It takes up a good deal of space (1160-1177), and any weakness of positioning may be the fault of scribes in what is a particularly confused section. (47) In this I mean, not the use of existing songs as supplying evidence for the materials out of which the written poems were long ago produced (a tempting line of enquiry followed by Gregoire, "Notes on the Byzantine Epic. The Greek Folk-Songs and their importance for the classification of the Russian versions and of the Greek manuscripts" in Autour de I epopee byzantine, London, 1975 ; see the remarks of Beaton, "Digenes Akrites and Modern Greek Folk Song : a Reassessment", Byzantion, 51 (1981), 22-43) but the use of existing songs as supplying evidence about lost MSS. The study of Greek folk song has been hampered by the traditional search for "pure" oral material, while it is evident that a good many songs derive from MSS or printed books. (On the general issue see M. Alexiou, "Folklore : an Obituary?", Byzantine and Modern Greek Studies, 9 (1984-85), 1-28 ; for a familiar example see G. Morgan, "Cretan Poetry, Sources and Inspiration", Kpnnxd xpovixd, 14 (1960), 405-415. It is worth noting that in this respect much work on folk song represents a regression from the work of Fauriel, who was more open-minded on the issue : note his inclusion, with justification, of a passage from Erotokritos (Greek edition, An/tonxd Tpayovdia Tfjg ovyxpovov 'Ehkddog, 1956, pp. 264-265). The Cypriot song of Armoures is sufficiently close in one of its versions ('EMtjvixd dtj/ionxd Tpayovdia, pp. 46-51) to allow the inference that songs can quite faithfully preserve the contents of a MS, at least so far as concerns its broad lines (this does not, of course, include the preservation of proper names). (48) See G. Saunier, "Le Combat avec Charos dans les chansons populaires grecques. Formes originelles et formes derivees", 'EAAnvixd, 25 (1972), 119-152, 335-370. (49) Charos as an innovation : G. Saunier, Adikia: le mal et I'injustice dans les chansons populaires grecques (Paris, 1979). conclusions about the version originally included in the collection must be tentative (50). The strongly Christian element may have entered with the very idea of writing a poem about the hero's death, and the part relegated by Alexiou to an appendix is perhaps no more inauthentic than much of the rest (51). But it is clear enough that, whenever they came together, we are dealing with three layers. First we have the archetypal Akrites in retirement telling of his deeds and declaring to his men that there will never be another Akrites ("). (This version is preserved in a Euboean folk song collected before the (50) There are common features with the Spaneas ; see G. Danezis, Spaneas: Vorlage, Quellen, Versionen (Munich, 1987), p. 134 ; plus e.g. Spaneas 516-517 (ed. Zoras, Bvfavnvrj noiqan): 'O xoa/wg ev npoawpivbc, f/ftepeg vnayaivovv, 6 nkovxog, to Xoydpiov wg dvefiog diapaivei. Although it contains traditional material it appears to be a set piece on vanitas vanitatum. (If the Akrites vogue really was closely tied to the figure of Manuel, we might hazard that this poem would not have been written before 1180.) (51) E 1794-1867 and Alexiou, p. X'. Metrical variations, however, seem to occur in more or less any Christianizing passage in E, and just appear to be part of this idiom (e.g. 1753). And the expatiation in a Christian vein is not necessarily alien to heroic poetry (Ker, Epic and Romance, p. 50) and, moreover, shows close affinities with the Alexander Romance, as G. Veloudis has shown, Der Neugrie-chische Alexander (Munich, 1968), pp. 265-269 : Alexander says farewell to his warriors and his wife, and his wife dies shortly after. There are verbal correspondences between e.g. the 1521 prose version (V. L. Konstanttnopoulos, ed., Ps.-Kallisthenes: Zwei mittelgriechische Prosa-Forschungen des Alexanderromans, vol. 2, Konigstein, 1983) e.g. 198, 18-20 .- Kai vd jjfevpnc on rijv oi/fiepov ijfiipav r\ aydnn flag, onov ei'xa^iev 61 diio p,ag, x^pKsTat • eytb vnaya'tvio eig rdv "Abr\ xai eoha dq>t']vw fie rov &eov. & r\yam\iievn dydnr\. Compare E 1772, 1774-1775 : xaXi] wg eldeg an' dpxvg, five efyaftev rov nbQov ... Erjnepov xuptCbfieOa xai dnepxofiai eig t6v xbaftov rdv fiavpov, oxoTeivbrarov, xai ndym xaxw eig "A6nv. We cannot assume that what we have here in E is allusion rather than simply a traditional language ; but the death of Alexander may well be the inspiration of this part of E with its ecumenism quite alien to the archetypal Akrites e.g. 1790-91. (52) E 1709-1764. Much of this section is of course extraneous to the original lay, in which the point will not have been Christian faith but a defiant piece of heroic boasting in the vein of the Emir's words at 500-512. Krumbacher (quoted in Alexiou, p. pX') was surely right to see this as a traditional element. 200 d. ricks about the escorial akrites 201 written versions were discovered, and therefore to be considered immune from the attentions of nationalist men of letters (53).) Secondly we have the romantic Akrites as husband addressing his wife. (This has been preserved in some other folk songs (54).) Finally we have an overlay of Christian reflection on the vanity of things. The thing may have been done with scissors and paste, but it still presents sufficient incompatibilities and traditional elements that we may rule out its having been composed ab initio for the collection (55). (As to the question of authorship, only speculation is possible. With the second poem it is probably out of place to speak of an author at all; the emulation by the third poem of the first suggests, perhaps, a different author; so too does the possibility that the third poem is an alternative telling of the fourth ; and the language and priorities of the fifth poem clearly stand apart.) To summarize the view of E that I have put forward. It is true that the very idea of collecting existing heroic poetry either about Akrites or relatable to Akrites and laying it end to end had in it the seeds of a biography of Akrites. Indeed, a biographical tendency has crept into the interstices between poems by the time of G (e.g. on the death of the hero's parents) (56). But E is not the biography of Akrites. Reluctance to admit this has its origins in the initial (and persistent) characterization of the Akrites versions as they were discovered as epics ("). The question is: if E had been discovered before T, G and (53) N. G. Politis famously claimed 1350 songs of the Akritic genre, and was criticized for this by Psichari, "A propos de Digenis Akritas", Quelques travaux de linguistique, dephilologie et de litterature helleniques (Paris, 1930), pp. 41-45. There are plenty of songs into which we may suppose that the name of Digenes has been foisted (see e.g. the song collected by the notorious Lelekos in N. G. Politis, "'Axpvnxa Sauccra. 'O O&varog tou Avyevfj", Aaoypa, "The ignorant, God-bothering, greedy monk had never even heard the name of Digenes Akrites and wasn't even interested to learn it!" (75). But Akrites' lack of prominence in the folk tradition did not of course entail that the new literary Akrites, however much he revealed of his own time and the Great Idea, was destined to be a failure ; and yet, as Kechagioglou acknowledges, our conclusions here must be a little disappointing (76). Akrites was still in the province of the philologist when Palamas inaugurated the modern literary vogue with his famous lines from "Iafifioi xal dvaitataroi. Greece certainly needed some sort of resurrection after 1897, hence Palamas' concluding lines, ott] Cwri Zavavpaivonai xal Xaovq dvaataivw. Palamas appears to be making the same pun here as a folk etymology of "Digenes" ("). And yet the hero had no invigorating effect on Greek poetry : in retrospect the presence of Akrites in the poetry of the modern Greek state seems to flicker as briefly as the flame of irredentist sentiment which produced it: although the hero is spoken of as the "unfailing spring" of poetry, this is far from being the case (78). In the absence of a unitary and standard Akrites the possibili- (75) N. A. Bees, "Mopamxa rpayooSia tou dxpiTjxou xoxaou", Ilavadrj-vaia, 11 (1905), 33-37. (76) On the distinction between the old and the new Akrites see Beaton, Folk Poetry of Modern Greece (Cambridge, 1980), pp. 78-82. (77) N. G. Politis, Tlapaboaeic(Athens, 1904), vol. 1, p. 69 : yiaxi etyffe 8vb yeveatc,. Compare Sikelianos, 'O 6a.va.Toc tov Aiyevtj (Athens, 1948), p. II : x' eoi nob naxnaec rr\v an/inv dppwotia, yeia Lov — xapa Eov, Atyevf/! ...Xpifftbc dveorng (78) G. Zoras. "'O Arv6vn.c; iv rfj vEoeAXnvixfi woifjoei", 'EUnvixt] Amuovpyia 6 (1950 - hence the Cold War emphasis) 839-843. On the original connection with Greek irredentism see A. Bryer, "Han Turali rides again", Byzantine and Modern Greek Studies 11 (1987) 193-206 ; for a poem which puns on the name Digenes in a debunking of nationalism see N. Calas, "Aryevfj", 'Oboe Nixr/rcc Pavrov (Athens, 1977), p. 92. about the escorial akrites 207 ties for lyric presented for the modern Greek poets were but limited. The contrast with modern Greek poetry's exploitation of the Homeric inheritance is stark (79). And yet, even if the modern Akritic myth is a dated and largely failed one, that does not deny us the opportunity - indeed it may encourage us - to look again at the poems of the E collection. For in it we may see various poets at work with varying but far from contemptible results. Birmingham. David Ricks. (79) The connection of Akrites with Homer is favoured but highly inapposite, especially if we take the view of E which I have outlined j for an example see G. Papacharalampous, "Akritic and Homeric Poetry", Kvnpiaxai Enovbai, 26 (1963), 25-65. For a contrast with modern Greek poetry's drawing on Homer see D. B. Ricks. "Homer and Greek Poetry 1888-1940", (Ph.D. London. 1986) and "A Greek Poet's Tribute to Keats". Keats-Shelley Journal, 37 (1988), 35-42.