Digenes Akrites Criticism - Essay

John Mavrogordato (essay date 1956)

(Classical and Medieval Literature Criticism)

SOURCE: Mavrogordato, John. Introduction to Digenes Akrites, edited by John Mavrogordato, pp. xi-lxxxiv. London: Oxford University Press, 1956.

[In the following excerpt, Mavrogordato examines the discovery of Digenes Akrites, compares the poem's many versions, and surveys assorted critical analyses.]


In the middle of last century nothing was known of the Byzantine epic of Digenes Akrites; but the atmosphere had been prepared by the publication of several ballads of what is now called the Akritic Cycle (a name first used by Legrand in 1874). The discussion of these—(in particular a paper by Büdinger who had used the headline ‘A Greek Mediaeval Popular Epic’, although the Song in question, The Sons of Andronikos, was only seventy lines in length)—had opened the way for further revelations of an heroic age of mediaeval Greece.

(i) The discovery was made at Trebizond. Manuscripts could not be photographed at Trebizond, and after considerable correspondence the precious work was sent to Paris by post; and in 1875 Sathas and Legrand published Les Exploits de Digénis Akritas—épopée byzantine du dixième siècle … d'après le manuscrit unique de Trébizonde. The manuscript is said to be not earlier than the sixteenth century. There are several gaps in it. The poem is divided into ten books of which the first and the last are missing altogether; there remain 3,182 lines of it.

This manuscript was afterwards returned to the hands of Savvas Ioannides, the Trebizond schoolmaster, author of a statistical history of Trebizond, who in May 1868 had received it from a monk at the monastery of Soumelá. (For the present state of the monastery of Soumelá see D. Talbot Rice in Byzantion, v. 72 ff.) He published another edition of it at Constantinople in 1887,1 and then deposited it in the library of the Filologikos Syllogos. All the archives of this institution are said to have been removed to Angora by the Turkish Government and the fate of the Digenes manuscript is unknown; but in the opinion of Kyriakides (conveyed in a private letter 7 April 1936) it must, for the present at least, be regarded as lost.

(ii) Before the learned world had had time to digest this, several other manuscripts were discovered; beginning with the version found in Andros in 1878 and published by Mêliarakes in 1881, a manuscript of the sixteenth century practically complete in 4,778 lines, conveniently filling the gaps in the Trebizond manuscript which it closely follows. The Mêliarakes edition was reprinted in 1920. This Andros manuscript is now in the National Library at Athens, and is referred to by Kalonaros as the Athens version (which he prints first in his collection).

The Mêliarakes introduction refers to a Greek newspaper of 23 November 1880, reporting that Legrand had discovered in Constantinople, and proposed to publish, another version, the oldest known, written in iambics and containing the name of the poet. At that time (towards the end of 1880), Legrand must have known of the existence of the Grottaferrata and Oxford manuscripts besides the Andros manuscript, which was still in the press; he may also have heard about the description of other versions in an unpublished Athos manuscript of the monk Dapontes just then discovered (see below). Legrand, discussing his hopes of one or all of these in a conversation with reference to his own edition of the Trebizond manuscript, may have originated this newspaper paragraph of which nothing more seems to be known.

Another short version, in prose, said to have been discovered in Constantinople by Dr. Mordtmann, is mentioned in Legrand's preface to Grottaferrata (pp. xi, xxii—‘signalée par le Dr. Mordtmann à Sabbas Ioannidis’); and is also referred to hopefully by Polites (Peri tou Ethnikou Epous, p. 5); by Hesseling in his introduction to ESC (Laografia, iii. 551); and by Ioannides himself in the introduction to his Constantinople edition of 1887.

(iii) Next came a manuscript of the fourteenth century discovered at the Greek monastery of Grottaferrata near Frascati in 1879, which was published by Legrand in 1892 (reprinted 1902).

(iv) The incomplete ‘Madrid’ version of only 1,867 lines was discovered by Krumbacher in the Escorial library in 1904 and was published by Hesseling in 1912.

(v) A rhymed version was published by Lampros in 1880 from a manuscript at Lincoln College, Oxford, which had the advantage of being signed by the writer, a monk of Chios, by name Ignatios Petritzes, who finished his task on 25 November 1670.

(vi) A prose version, written by Meletios Vlastos of Chios in 1632, discovered by Mr. D. Paschales in Andros in 1898, was not published till 1928; it is now in the library of the department of Folk-lore at the University of Salonica.

(vii) Finally there is a very incomplete Russian version—fragments of a (linguistically) thirteenth-century version assembled from two different eighteenth-century manuscripts and from the quotations of the historian Karamzin (the third manuscript, from which he took his quotations, was burned in Moscow in 1812). This version was published by Speransky in 1922 and reproduced in a French translation by M. Pascal in 1935; and in a Greek translation in the edition of Kalonaros in 1941. A third and fuller eighteenth-century manuscript of this version was recently discovered in Russia and was published in 1953.

We thus have now manuscripts of five metrical versions of the Digenes Epic:

Trebizond MS. sec. xvi, 3,182 lines, 10 books
Andros MS. sec. xvi, 4,778 lines, 10 books
Grottaferrata MS. sec. xiv, 3,749 lines, 8 books
Escorial MS. sec. xvi, 1,867 lines
Oxford MS. 1670, 3,094 lines, 8 books

One prose version:

Paschales MS. 1632, 10 books

And one Russian version:

Speransky MSS. sec. xviii
Kuzmina MS. 1761.

I usually cite these versions by the first three letters of these titles (TRE, AND, GRO, ESC, OXF, PAS, SPE), giving the number of the line and of the book where necessary for the metrical versions, and for the prose versions the number of the page of the volumes of Laografia and Byzantion in which the Greek prose version and the French translation of the Russian version were first published.

The Ballads of the Akritic Cycle are more difficult to deal with because no approximately complete collection has yet been published and they must be hunted out in various journals and anthologies. Polites once said that his own private collection contained more than 1,300 ballads; but after eliminating versions which differ in only a few words or only by combination or contamination with fragments of other versions, it might be permissible to guess that the number of ballads deserving consideration as ‘Akritic’ must be less than a hundred; of which not more than fifty have any great significance.

Each version includes some details or names which must have been added by the author or transcriber of the version in which they occur, and could not have formed part of the supposed archetype, the original Digeneid, which is generally believed to have been put together from a great variety of sources in the tenth century.

The Digenes Akrites is a romantic epic of between 3,000 and 4,000 lines, narrating and celebrating the parentage, education, exploits, and death of its eponymous hero—whose name implies the burden of the story, for it may be translated as Twyborn the Borderer. His father was a notable Arabian emir who in a raid over the Byzantine frontier carried off the daughter of a Greek general; after embracing Christianity he was permitted by the family to marry her; and settled down as a law-abiding subject of the ‘Roman’ Empire. The son of this union of two races and two religious soon showed in the chase signs of exceptional heroic ability—hunting and athletics having been in the Greek Empire the pride and privilege of a ruling class. He imitated his father by carrying off—but single-handed—the daughter of a Greek general. After his marriage he left the parental castle, and with his bride and a few personal attendants lived a nomadic life among the lonely places of the border; making it his special business to exterminate the bands of robbers and cattle-drivers who haunted it (all of whom seem to have been Greeks, and one of whom was a woman). He built himself a palace on the Euphrates (although the earlier books seem to imply an imperial frontier in Kappadokia, where his parents had their castle till they came, before their death, to live with him). There he soon fell ill and died surrounded by the honours of the whole empire in Asia.

The epic can hardly have been officially inspired for it seems to embody no definite propaganda. If the imperial authorities had wanted to promote a new ideal of peace they would have displayed it in a less indefinite and more popular form, and would have employed a more accomplished and a more metropolitan poet. The bare outline just given suggests an attempt to arrange a bunch of local adventure stories into the likeness of an epic embodying ideals of tolerance and peace. It is not a romance in spite of many borrowings from Hellenistic romances; and it is untouched by Western influences in spite of many reflections, through such Hellenistic writers, from the Mediterranean world which was later to inspire any number of Italian and French storytellers. J. B. Bury (Romances of Chivalry, pp. 18, 19) praises the epic comprehensiveness of Digenes, ‘which justifies us in naming it along with Homer and the Nibelungenlied’.

It is a heroic poem of provincial origin intended for private reading or for recitation not in the market-place but in banqueting hall or refectory.



The Grottaferrata version is probably the earliest we possess. The narrative is clear, simple, and concise. Irrelevancies are omitted and effective detail often added. It omits altogether the rather silly story about Ankylas. It begins with the Emir instead of with the once-upon-a-time King and Queen who had a beautiful daughter. It omits the earlier visit of Digenes to Philopappos. In the episode of the Emperor's visit it names the Emperor Basil (instead of Romanos as in TRE and AND) and it adds the detail that Digenes catches a wild horse and kills a lion by way of display for the Emperor's entertainment; it also names the Emperor Basil (instead of Romanos) in the passage in Book IV which refers to the banishment of Eirene's father, and in the same passage, by a sort of attraction or association set up by the name Basil, gives the Emperor himself the surname of Akrites, calling him ‘Basil the blessed, the great Borderer’ (cf. GRO iv. 55 with TRE 835 ff.). It omits the later references to honours conferred on Digenes by the Emperor Nikeforos which appear in the penultimate books of AND and TRE. From the same part of the narrative it also omits many redundant details of his wealth and daily life, and it omits the excessive lamentations on the death of his mother. It mentions that his final illness was the result of a chill after bathing; and it specifies that his tomb was built on a hill near Trôsis. It develops at greater length than the other versions the incidents of his courtship and marriage.

There is more Moslem colour in GRO than in other versions (e.g. a knowledge of Moslem miracles (GRO iii. 139) and of the Moslem shrine at Palermo (GRO i. 101)); but there is also a greater sincerity, one might almost say a greater savagery, of Christian morality.

The language is fairly correct literary Greek of, probably, the eleventh century, with a noticeably large number of words from the Septuagint. Such, however, is the unity of the Greek language, and so insufficient is our present knowledge, that, if allowance be made for ecclesiastical and literary influences (for familiarity with the Septuagint and with historical sources) on the one hand, and for the revisionary habits of later copyists on the other, it would be difficult, judging by language alone, to say much more than that the Grottaferrata version was written between the tenth and fourteenth centuries.


TRE and AND are a pair and follow the same story in all its details so that AND is useful for filling up the numerous gaps in TRE (especially the whole of the beginning down to the pursuing brothers' search for their sister's body in the Emir's camp, and the whole of the end from the speech of Digenes to Eudokia on his deathbed).

The language of TRE is also inclined to be literary but is distinctly less correct than that of GRO; the syntax is often in a state of dissolution so that the semi-classical manner seems less natural.


AND tells exactly the same story as TRE and has many identical passages. But it has been written up in a later and more romantic manner; many passages suggest that it was copied and rewritten as late as the fifteenth or sixteenth century. (The writing of the manuscript is said to be of the sixteenth.) The end of Book I from line 247 onwards actually breaks into rhyme and produces a lyrical peroration in roughly rhymed octosyllables. Although the extant parts of TRE are, as has been said, identical in incident with AND, it is permissible to suspect that the beginning of TRE, if we had it, might differ considerably; for the first book of AND (‘once upon a time a King and Queen’), while consistent with the romantic manner of the AND narrative and language, would have seemed incongruous in the distinctly more epic manner of TRE. A romantic monk of the fifteenth century seems to have been rewriting a more learned text which was almost that of TRE, and often took over its lines as they stood. (For the introduction of rhyme see Krumbacher, p. 700.)

Perhaps this romantic monk was Eustathios, who is said in the opening lines to be writing the story for a dear young friend called Manuel. It has been surprisingly believed that these lines do in fact give us the name of the original author of Digenes. A Greek scholar, A. Hatzês, has tried without much success to prove by linguistic arguments that the Digenes was written by Eustathios (or Eumathios) Makrembolites, author of the Byzantine prose romance Hysmine and Hysminias. (See Byz. Neugr. Jahrb. ix (1931), pp. 256 ff.; and for Grégoire's criticism and list of Hatzês's works, Byzantion, vi. 482.) It ought to have been impossible for anybody to suppose that the heading of AND implies that Eustathios was the original author. Petritzes, the writer of OXF, uses even more definite language—esyntaxa kai 'synthesa to, he says—yet we know that he was only the versifier. Eustathios may have been the redactor who added the rhymes and other romantic embroideries. As for the author of Hysmine and Hysminias, he lives in quite a different world of literary preciosity.


The ESC version is extremely incomplete, beginning at about the same point as TRE and omitting altogether many incidents and innumerable details, while many of the episodes and lines are in the wrong order. It is written in an extremely but not uniformly popular style, with a Cretan flavour, which combines with many repetitions and confusions to give it a striking if superficial resemblance to some of the Akritic ballads. It has become a commonplace, started by N. G. Polites, to say that the Ballads are more ‘poetical’ than the Epic, because they are full of magic and confusion. Consequently Kyriakides and others declare that the ‘breath of life’ blows through the muddle of the ESC because it resembles the Ballads. It has even been argued that, assuming the Epic to have been assembled from scattered Ballads or Lays, the fact that ESC most resembles the extant Ballads shows that it is nearer to the source and consequently earlier than all the other versions. A similar argument was at one time (October 1941) produced by Grégoire in support of the Russian version. This argument ignores the considerations that the qualities of an Epic are not the same as those of a Ballad; and that we cannot conclude that an epic text is closer in time to its component lays because it resembles what other lays or ballads have become after a thousand years of oral transmission. So much a priori. But Grégoire has now shown (Byzantion, v. 339) that the details of Arab raids borrowed by GRO from the chronicles of Genesios are not absent from ESC; and ESC even exceeds other versions in bookishness by borrowing the tomb of Digenes from Arrian's description of the tomb of Cyrus (Anabasis, vi. 29. 4-8) and the parrots in the garden from Achilles Tatius.

Actually the disorder of the ESC version, the innumerable omissions, the innumerable additions of irrelevant tags out of ballads, the repetitions of the same line, the duplications, the phonetic confusions (e.g. 719, amêras, ‘the emir’, for Omêros, Homer), the way the words continually overflow the metre, and the metre breaks off into half-lines or into the rhythms of a chanted speech—all these lead to the conviction that this was taken down, not before the end of the fifteenth century, the date of the manuscript, from the dictation of a wandering Cretan ballad-monger who was trying to recite from memory in the musical recitative which still survives in Crete a version which contained a few original details (e.g. that Digenes made a bridge over the Euphrates). (The same overflowing of the metre is noted by Kyriakides (p. 119) as a sign of dictation in the Ballad of Armoures.) Whenever his memory fails he repeats as a catchword one of the characteristic or operative lines of the episode he is trying to recall, or he improvises, or marks time with a tag or two out of his repertory; and his version certainly contains vestiges of a literary original. Anyone who doubts the possibility of memorizing even imperfectly the whole of the Digenes may be reminded that only a few years ago villagers could be found in Crete who professed to know by heart the whole of the Erotokritos which is nearly three times as long as this (9,956 lines).

Kyriakides (op. cit., p. 75) compares three extracts—ESC 806-23, TRE 1207-37, GRO iv. 380-95—with a view to showing that ESC, allowing for a lacuna, follows the fullest tradition. He argues that there must be a lacuna because the mother of Digenes prays to the Virgin twice; and that therefore something to account for her second prayer—actually, as we see from TRE, the fact that Digenes couldn't eat his supper—must have been lost. But the whole passage shows that ESC doubles nearly everything; he repeats things because he is trying to remember what comes next; e.g. 831, Digenes plays his tampouri as well as his labouto; 918, 919, he calls out to Doukas to arouse him, but Doukas is aroused by hearing the gallop of the horse; there is another repetition in 836, 845, and another muddle in 844, again defended by Kyriakides in a long note (p. 79) on the kailyard principle that whatever is laïkôteron must be poiêtikôteron. It is true that this version has beauties and originalities of sound and surface that have been produced by oral transmission; but they are the qualities of a seventeenth-century ballad, not of an eleventh-century heroic poem.

The ESC manuscript which contains this version also contains a version of the Lybistros and Rodamne romance written in the same hand and disfigured to a smaller extent by faults of the same kind. The latest editor of Lybistros and Rodamne (Mme J. A. Lambert, Amsterdam, 1935) dismisses the hypothesis of dictation in the case of the Lybistros, in spite of the phonetic evidence, on grounds which seem insufficient.


There is no such problem when we come to the OXF version. We know that it was composed in 1670 by the monk Ignatios Petritzes of Chios, who puts into rhyme a version which substantially resembles that of AND; but as it is in eight books instead of ten he may have been working from one of the versions seen on Mount Athos by Dapontes (see below, p. xxi). In so doing he humanizes and to some extent rationalizes the story. His tale is well-proportioned and not savage or sanctimonious; it is set in a world which is more ‘civilized’, or at least less mediaeval. He brings in priests and bishops to celebrate weddings and funerals. The Arabs become Turks; the emir Haplorrabdes becomes Abdullah; and Petritzes is the only redactor who bothers to give Abdullah's wife and daughter Moslem names, Aissé and Fatouma—names easily found, for these were the favourite wife and youngest daughter of Mohammed. In the same episode he explicitly denies that Digenes, after rescuing the deserted bride, helped himself to his own reward, as the earlier versions allege, or did anything to be ashamed of. When he comes to Maximo the Amazon, whom he calls Maximilla, he remarks that a woman's place is in the home; and now he allows Digenes to yield to temptation, thus distinguishing Fatouma from Maximilla, who had, literally, asked for it. The name Maximilla may have been familiar to him as that of a prophetess who was head of the Montanist heresy at the end of the second century.

The fact that the earlier versions seem to see no difference between the treatment of the unwilling Fatouma and that of the willing Maximo, may be connected with the oriental seclusion of Evdokia (who was never seen by any of the servants; only a little boy waited on them at table) to show that the author of Digenes was exceedingly provincial in his outlook; there was nothing like this at Constantinople (or in Hysmine and Hysminias); but the author of Digenes was Asiatic; he seems never to have heard of Europe; and only mentions one place in Europe—the Moslem shrine of Palermo (except for AND 2419, sta merê Ahaias).

Ignatios Petritzes was not only six hundred years later; he was writing in Chios. The whole is written in good popular Greek of the seventeenth century without pretentiousness or affectation.2 We are surprised only by quotations from Aristophanes (Plutus), Theocritus, The Song of Songs, Bion, and Euripides (Hecuba). (OXF 237-42 = Aristoph. Plut. 3-7; OXF 1563-6 = Theoc. iii. 15, 16; OXF 1576-80 = Bion, 1. 7-20; OXF 1052 ff. = Song of Solomon vi. 8; OXF 1593-6 = Eur. Hec. 600, 601.) The chief interest of this version is in the personality of the author Ignatios Petritzes. (There are four other manuscripts copied by him in the library of the Greek patriarchate at Jerusalem. See GRO, introd. p. xii.) His humble lines of dedication at the end of his work, and his hopes that it may some day be printed, are modest and attractive.


D. Paschales who discovered the prose version in Andros in 1898 waited thirty years before he published the manuscript. When at last it appeared in 1928 it was a sad disappointment. It had been hoped that a prose version might throw some light on problems of date and origin; or at least that it might correspond to one of the versions which the eighteenth-century monk and polygrapher Caesar Dapontes (1714-84) describes as then existing in the library of the monastery of Xêropotamou on Mount Athos. (This reference was first discovered by M. Gedeon in the manuscript of a Byzantine chronicle in verse called Biblos Basileiôn. See Lampros, Romans grecs en vers, introd. p. xcix. For a notice of Dapontes see R. M. Dawkins, The Monks of Athos (1936).)

Dapontes gives in twenty lines a summary of the story which he unfortunately breaks off at the point where the hero's parents have been introduced; up to this point, that is to say up to about the end of the first book, he appears to follow closely the story as given in AND—except that the father of Digenes is said to have been an ‘emir of Egypt’ or ‘Sultan of Misir’ (Cairo): this difference is probably without significance: the Emir is actually called Soultanos in AND 307—and a Syrias may have been confused with Misiriou. Then interrupting his summary he adds:

‘The story is very long but interesting and sweet as sugar. It is a book of eight or ten quires and contains all his exploits. I have seen it in two forms with illustrations and without pictures. It is divided into eight books and it is very rare and difficult to find. At the beginning of each book it has five lines of verse containing the argument of each book: and it is always in manuscript. I have not seen a printed copy and it seems never to have been printed. They have printed Erotokritos, Sôsanna, Erofile and others: what a pity they have never printed Basil. If it is given me to live I mean to put this story into verse and send it straight to Venice. Happy the printer who prints it for it will bring him both profit and honour.’

Two points should be noticed in this quotation. First of all his intention of putting it into verse before having it printed, which implies that both the manuscripts he saw were prose versions. (It cannot be assumed that he saw two different versions. He may have seen two copies of the same version, one of them illustrated.) It has been suggested, quite untenably I believe, that he only meant to put it into rhyme, to add rhyme to blank verses. But his mention of the five lines of verse (and if they were five lines they were almost certainly unrhymed lines) prefixed to each book implies that the rest of the work was in prose. It is curious, however, that the library should have had two copies in prose and none apparently in verse. If Dapontes had succeeded in producing his own versified edition, we may be sure it would have resembled closely the Oxford version written a hundred years earlier by that other monk from Chios, Ignatios Petritzes, whose character, gentle, literary, and unheroic, must have been very much like that of Dapontes.

It was the curious habit of those who enjoyed and transmitted to us the works of middle Greek literature to rewrite, apparently each to his own taste, any work which they thought worthy of preservation. The Cretan play Erôfile even after it was printed is said to have been rewritten for the second edition by a patriotic Cretan. Of the Cypriot Chronicle of Mahairas the two sixteenth-century manuscripts (Venice and Oxford) differ so much, and yet are so much alike, that Professor Dawkins felt himself obliged to conclude—rather unsatisfactorily—that they were written by two independent authors working from identical materials.

This rewriting habit must have begun early. The text of Chaireas and Calliroe, a romance of the second-century novelist Chariton of Aphrodisias, depended on a single Florentine manuscript until the recent recovery from Egypt of three small fragments. Of these fragments two papyri of the second or third century generally confirm the Florentine text. But a seventh-century parchment palimpsest discovered near Thebes in 1898 differs so widely that in the words of the latest editor (W. E. Blake, Oxford, 1938) ‘rationem inter Thebanum et Florentinum haud aliter definire se posse crediderit vir doctus quam si duas memorias omnino inter se diversas poneret, quarum utraque suo modo ex ampliore exemplari contracta esset’. Before the discovery of these fragments the romance was usually attributed to the fourth century.3

Be that as it may, when D. Paschales at last published the prose version associated with his name, it turned out to be another seventeenth-century version which had little interest either as a prose romance of Digenes or as a specimen of the seventeenth-century language, because it was obvious that the writer had taken the trouble to turn into flat and literary prose a version very closely resembling that of AND, leaving embedded in his periods many undigested fragments of the original verses.4 The editor claims (Laografia, ix. 312) that it represents an independent tradition, and it is true that in numerous unimportant details he seems to follow TRE rather than AND and in a few details to have been following a version which differed from both. (See, for example, Laografia, ix. 350, compared with TRE 1128, 1133, lines which are different in AND; p. 358, where the name of the Saracen Soudales is omitted but details are inserted which are not in AND or TRE (lacuna), or GRO (where the Saracen is omitted altogether), but which are in ESC 930, 931; p. 359, where the detail of the girl's father and her two brothers pursuing does not agree with TRE 1267 or AND 2075 or OXF 1957 (where the three pursuers are her three brothers), or with ESC 969 (five brothers) but with GRO iv. 610, iv. 657; PAS p. 361, the twelve eunuchs, who are to be found in GRO iv. 925 but in no other version; p. 366, a detail which is in no other version about the confusion in the house of Haplorrabdes when his wife fell ill; p. 406, the doctors feel his pulse; and a few other details and misunderstandings.)

The writer gives his name as Meletios Vlastos and the date 1632. The editor on insufficient linguistic grounds says that he was a native of Chios. A Cretan monk of this name is said to have been one of the teachers of Cyril Loukares. Why did Meletios want to turn the story of Digenes, of which he had in his hands a perfectly good redaction in verse, into indifferent prose? His motive was not the same as that of the gentlemen who translate the New Testament into ‘modern English’. Verse is the natural speech of the peasant culture (cf. the story of Mrs. Flecker's cook in J.H.S. liii (1933), 1); verse aids the memory in recitation, and is no longer necessary when recitation gives place to reading.5 There are many French prose romances (e.g. Balin and Balan) made from earlier verse romances; and various prose versions of the Roman de Troie of Benoît de Ste-Maure from the twelfth century onwards. To Meletios there was something vulgar and uneducated about a story in verse; he wanted to have a good story in a form fit for a gentleman to read.


The Russian version (SPE) is composed of three fragments of a (linguistically) thirteenth-century prose romance, two from manuscripts of the eighteenth century, and a third quoted by the historian Karamzin from a manuscript probably of the thirteenth century which was burned in Moscow in 1812. These fragments were assembled and published by Speransky in 1922, and were edited in a French translation by Pascal in 1935 (in Byzantion), and in a Greek translation by Kalonaros in 1941.

Here, combined with many folk-tale elements—a book of fate, a magic horse, a spring of water with a light burning in it,6 and other fairytale wonders which suggest oral transmission—we can recognize the chief incidents of the Greek story.

Maximo the Amazon becomes Maximiana the daughter of Philipap, and Devgeny after pole-jumping over the river7 easily defeats them both and sends them home to his parents. He refuses to marry Maximiana because the Dream Book says that if he marries her he will live sixteen years but if he marries Strategovna (the daughter of the General) he will live thirty-six years. So he carried off and, after several bachelor parties, married the General's daughter, and settled down to a life of fame and hunting, until he was attacked by a certain Caesar called Basil; and then he jumps over the river again, defeats him, enters into the town (not named), and ascends the throne.

Grégoire accepts this rebellion as the first nucleus of the poem, and argues that we must see in it a revolutionary manifesto issued by the Paulician heretics of the Armenian border and their Arab allies against their arch enemy the Emperor Basil I the Macedonian (867-86)—from whom the hero was given his first name—and that all the Greek versions are descended from a loyalist revision of this seditious original officially prepared and circulated at the beginning of the tenth century. Grégoire's arguments, too numerous to be examined in detail, would be convincing only in an unreal society. (He has not noticed by the way that Philipap's army—SPE 318—are said to be ‘brave as Macedonians’.)

These arguments were strongly contested by Wartenberg (in Byzantion, xi (1936), pp. 320 ff.); it must be added, however, that they seem to have been largely accepted with other Gregorian hypotheses by Professor Arnold Toynbee (A Study of History, v. 252 ff.). The evidence leads us rather to believe that the Russian is descended by oral transmission from a sophisticated Greek version already combining the later Euphrates frontier—quite inconsistent with the reign of Basil I—with the earlier Kappadokian frontier; the Russian includes even the eulogy of the month of May which seems to derive from the eulogy of the Rose in the second book of Achilles Tatius. It is reasonable to suppose that the Hero of the Borders becoming Emperor in the City is a fairy-tale ending tacked on in the course of transmission—a common form of contamination in popular literature of any period. This conclusion seems to have been finally confirmed by the discovery of a third Russian manuscript dated 1761. It closely follows the earlier of Speransky's two eighteenth-century fragments, but is much fuller and preserves traces of an earlier text and vocabulary.8 It begins with the Emir and completes the story of the abduction of the daughter of Strategos, but omits altogether the final episode of the overthrow of the Emperor Basil. If, as seems probable from the superior text and archaic vocabulary, this new version in fact preserves the oldest Russian tradition, it follows that the Emperor's defeat must be the addition of a later copyist, and does not preserve any early Greek original put about by rebellious Paulicians. It was already impossible to understand, on Grégoire's hypothesis, why no version, Russian or Greek, shows any trace of feeling either for or against the Paulicians.


Before leaving this review of the various versions of the Digeneid and passing on to consider the story they present, something must be said about the Ballads, the Akritic Cycle of Folk-songs or tragoudia.

These ballads have been found to some extent in all parts of the Greek world (as have also ‘Castles’ or ‘Tombs of Digenes’), and it was clear that they belonged to the Byzantine period and to Asia Minor, on the fringes of which, and especially in Pontus and in Cyprus, the best of them have been collected. Their nature is well indicated in the well-known scholion, on a passage in Philostratus' Life of Apollonius of Tyana, written by Arethas (850-932), Bishop of Cappadocian Caesarea,9 which speaks of ‘wandering beggars, like the cursed Paphlagonians who now make up songs about the adventures of famous men and sing them for pennies from door to door’ (see Kougéas, in Laografia, iv. 236). This brings us at once to the place and also to the time of the Akritic ballads: the period which, as Grégoire has shown, is peculiarly the Heroic Age of Mediaeval Greece: the ninth and tenth centuries which produced the adventurers of the Amorian and Macedonian dynasties, the Andrónikoi and Constantines of the Doukas family, the Nikefóroi of the Fokâs family, and many others whom he has convincingly identified in the fragments of the existing ballads.

Many of the early investigators thought that the poem of Digenes had been made by a ‘rhapsode’ who stitched together short lays or ballads of this sort which celebrated the glories of individual heroes, a method of composition once supposed to have produced the Iliad and the Odyssey, and the French Roland (as Lönnrott in fact produced the Finnish Kalevala in modern times (1835-49) from traditional lays collected by Topelius in 1822).

Here, then, we say as we approach the Greek songs of the Akritic Cycle, here is a wonderful chance to study in their raw state the materials of an epic poem. In these ballads we might expect to find, distorted perhaps by oral tradition but still easily recognizable, many of the episodes of Digenes. We are surprised to discover an entirely different world: a world of supernatural feats, magic weapons, and talking birds; in which Digenes is only one of a number of heroes we have never heard of before—Andronikos, Porfyrios, Armouropoulos, Konstantas, Theophylaktos, Xantinos—and is not by any means the most popular. Digenes indeed hardly appears at all except in the series of ballads describing his death,10 foretold by talking birds, his iron death-bed, and his wrestling with Death who comes to fetch him. This connected series and one or two which seem to refer to his carrying off, in entirely different circumstances, of Evdokia, are the only ones that can be fitted into the story of our Digenes at all; and a great number are commonly called ‘Akritic’ only because of their obvious antiquity. Some of the other heroes have been plausibly identified by Grégoire as historical figures, emperors or generals or pretenders; or as the eponymous heroes of imperial themes or regiments. But he has had less success in rationalizing the incidents.

A son of Andronikos born in captivity, his mother having been carried off before his birth, escapes from the Saracens and goes in search of his father and brothers. (In this episode it is worth noting that the hero is digenes, twy-born, in another sense, as being the son of a Christian father Andronikos as well as the putative son of the Emir at whose court he is born and brought up (see Passow 482, Kyriakides, Dig. Akr., pp. 35 ff., and Legrand, Chansons pop. grecques (1874), no. 87, esp. lines 7, 8).) A gargantuan Porfyres falls in love with the king's daughter and no chains are strong enough to hold him. The equally monstrous Xantinos liberates his son who has been yoked to the plough with a buffalo. Most of the songs which are not in the Pontic or Cypriot dialect have almost lost their narrative character and have been so contaminated with later ballads that they are only recognized as Akritic by the names of the heroes and by their obvious antiquity. It is sometimes suggested that Digenes was only one among a number of ballad-heroes and that the author of the epic took up the Digenes cycle and left the other heroes to the wandering ballad singers. But why should adventures which were good enough to attract the epic-maker in search of material have completely disappeared from ballad circulation? It must be understood that the relation between ballads and epic is not one between successive stages of composition, or between different treatments of the same materials, adapted in one form for street singing and in another for ceremonial recitation or private reading. It is rather a relation between different levels of interest in the same community. The study of one does not necessarily throw any light on the other; although a knowledge of both is necessary for an understanding of the society in which both were produced and developed.

What nearly all editors call the poetical vigour of the ballads as distinguished from the epic is really a radical difference of theme and treatment. A fourth-rate ballad, especially if it is recorded in the surprising dialect of Pontus, may be superficially more attractive than a second-rate epic. The songs as we have them today, after 900 years of oral transmission, cannot be regarded as the sources of the epic, and cannot be used as standards by which to judge the relative ages of the various versions or the various episodes of the epic. We do not know what these ballads were like when they were first sung. If they were recognizably the same as they are now, then they obviously have nothing to do with our Digenes; even the death-bed series are disqualified for comparison by their association with miraculous incident; and if they were entirely different—still less can we be allowed to draw any critical information from their present derivatives.

Only one of the ballads, the Son of Armoures (first published by Destouny in Russia in 1877 with a facsimile and reprinted by Kyriakides, p. 119), exists or used to exist in a manuscript said to have been of the fifteenth or sixteenth century. All others have been collected in modern times, although the sources are not always known. (See, for example, the obscure history of the Sons of Andronikos ballad, first published by Zampelios (1859) who says the manuscript was given him by Brunet de Presle,11 who had it from the unpublished part of the collection of Fauriel, who had copied it from a manuscript in the Bibliothèque Nationale; this is flatly denied by Legrand who says that Brunet de Presle when a student took it down from the dictation of Professor Hase (Zampelios, Pothen hê Koinê Lexis Tragoudô? (1859), p. 37; Legrand, Chansons pop. grecques, p. 183). Büdinger, who reprinted the text given by Zampelios, refers to his vain efforts to trace the supposed manuscript in Paris (Ein mittelgriechisches Volksepos (Leipzig, 1866), p. 3); see also Wagner who again reprinted it (Medieval Greek Texts (1870), pp. x, xiii n. 34, xxii). Legrand implies that his own text was taken from the same manuscript of Brunet de Presle as that which Zampelios had copied and ‘improved’.)

Not one of the Akritic ballads or fragments can be regarded as belonging to the category of Folk-chronicles or Historical Ballads, a class represented by the Cretan historical ballads of the eighteenth century like the well-known Daskalogiannes; and it is from the historical rather than from the romantic ballad that we should expect the maker of the Digenes epic to have drawn some of his materials.12 The strange absence from the extant Akritic ballads of any recognizable incident of the epic is paralleled to some extent in English ballad literature. The Gest of Robyn Hode is a small epic of 1,824 lines divided into eight ‘fyttes’, first published in the middle of the sixteenth century, and said to have been composed ‘by a poet of a thoroughly congenial spirit’ from ballads which had begun to circulate about a hundred years earlier. Not one of the ballads from which it was made up is extant in a separate shape, and ‘some portions of the story may have been of the compiler's own invention’ (Child's Popular Ballads, ed. Sargent and Kittredge, p. 225). …



An attempt must be made to summarize the interpretations and estimates of the Poem of Digenes put forward by a succession of scholars from the first editors Sathas and Legrand down to M. Henri Grégoire, who, in a series of articles beginning in 1930 (assisted on questions of the relations of Byzantium and the Arabs by the researches of E. Honigmann, A. A. Vasiliev, and Marius...

(The entire section is 18259 words.)

Hugh F. Graham (essay date 1974)

(Classical and Medieval Literature Criticism)

SOURCE: Graham, Hugh F. “Digenis Akritas and the Devgenievo Dejanie—A Reappraisal.” Studies in Medieval Culture 4, no. 3 (autumn 1974): 483-95.

[In the following essay, Graham analyzes the Russian version of Digenes Akritas, the Devgenievo Dejanie.]

The fascinating comparative study of the manuscripts of Digenis Akritas and the Devgenievo Dejanie has continued to attract international scholarly attention, although it suffered a great loss with the recent death of M. Henri Grégoire, who numbered it prominently among his scholarly interests and who developed many provocative hypotheses concerning it. At present our Soviet colleagues are...

(The entire section is 7810 words.)

Roderick Beaton (essay date 1981)

(Classical and Medieval Literature Criticism)

SOURCE: Beaton, Roderick. “Was Digenis Akritas an Oral Poem?” Byzantine and Modern Greek Studies 7 (1981): 16-27.

[In the following excerpt, Beaton analyzes the influence of the oral tradition on Digenes Akrites.]

The first question to be asked, as we come to consider the composition of Digenes, is: Which version is the most authentic? The opinion is now generally held that G represents the oldest version of the poem as well as the oldest text, although the claims of Kyriakidis and Grégoire for the superior authenticity of the E version have recently been restated by Stylianos Alexiou. Alexiou has brought forward new evidence that elements in...

(The entire section is 4714 words.)

Andrew R. Dyck (essay date 1983)

(Classical and Medieval Literature Criticism)

SOURCE: Dyck, Andrew R. “On Digenis Akritas Grottaferrata Version Book 5.” Greek, Roman and Byzantine Studies 24, no. 2 (summer 1983): 185-92.

[In the following essay, Dyck analyzes the bride-theft motif in Digenes Akrites.]

Only two versions of the epic Digenis Akritas narrate the hero's encounter with the daughter of the Syrian emir Haplorrhabdes: Book 5 of the Grottaferrata version (G) and Book 6 of the version Z which the latest editor, E. Trapp, reconstructs from the Trebizond manuscript and the Andros manuscript now in Athens.1 For this material, however, as Trapp notes,2 Z is likely to represent a contamination of its...

(The entire section is 2941 words.)

Michael J. Jeffreys (essay date 1983)

(Classical and Medieval Literature Criticism)

SOURCE: Jeffreys, Michael J. “Digenis Akritas and Kommagene.” In Popular Literature in Late Byzantium, edited by E. M. and M. J. Jeffreys, pp. 5-28. London: Variorum Reprints, 1983.

[In the following essay, Jeffreys discusses the origins of Digenes Akritas and examines some of the variations between its various existing manuscripts.]

Few byzantine texts offer more interest to a wider variety of scholars than Digenis Akritas. It provides material for literary historians at different levels of Greek, for students of traditional literature, for specialists in Arabic, Armenian and Russian studies, for codicologists and art-historians as well as for the...

(The entire section is 7721 words.)

Angeliki E. Laiou (essay date 1993)

(Classical and Medieval Literature Criticism)

SOURCE: Laiou, Angeliki E. “Sex, Consent, and Coercion in Byzantium.” In Consent and Coercion to Sex and Marriage in Ancient and Medieval Societies, edited by Angeliki E. Laiou, pp. 198-221. Washington D.C.: Dumbarton Oaks Research Library and Collection, 1993.

[In the following excerpt, Laiou examines Digenes Akrites for what it reveals about twelfth century laws regarding abduction and social attitudes concerning sexual consent.]

What can literature contribute to an inquiry such as ours? If handled carefully, it can instruct us as to contemporary concerns and underlying attitudes. It can also be vastly misleading, for literature is not necessarily a...

(The entire section is 10157 words.)