Sunday, September 27, 2015


Adam J. Goldwyn, Dimitra Kokkini, John Tzetzes, Allegories of the Iliad. Dumbarton Oaks medieval library, 37. Cambridge, MA; London: Harvard University Press, 2015. Pp. xxiv, 577. ISBN 9780674967854. $29.95.

Reviewed by Donald J. Mastronarde, University of California, Berkeley (

Version at BMCR home site

This new volume from the Dumbarton Oaks Medieval Library brings (on facing pages) text and English translation of a long and curious work (not represented yet in TLG) of the 12th-century polymath John Tzetzes. The editors/translators provide a short general introduction, and at the end minimal notes on the text, 35 pages of very short notes on the translation (the majority of which are cross-references to lines of the Iliad), a short bibliography, and an index of proper names.

Allegories of the Iliad, a poem in the 15-syllable πολιτικὸς στίχος (over 6600 of them!), is an odd mishmash that bears the marks of its checkered history. It is partly brief plot summary (ὑπόθεσις), partly extended paraphrase of the epic in simpler Greek, partly allegorical explanation of almost all divine and supernatural elements, and partly self- advertisement of the author. Tzetzes began the work as a commission for the empress Eirene (Bertha von Sulzbach, from Bavaria), wife of Manual I Komnenos, probably in the 1140s, but the elaborate Prolegomena (over 1200 lines) contain appeals for guidance from the dedicatee as to what exactly she wants the work to be. Apparently, she never gave a clear answer and eventually lost interest in the work (or in the author). The separate books in which Tzetzes summarizes and explains the books of the Iliad vary widely in length and content. After the long Prolegomena and a treatment of Book 1 in 375 lines, his Books 2-15 are much less ambitious (averaging only 149 lines each). Then at Book 161 it is revealed that the task has been taken up afresh for a new patron, Konstantinos Kotertzes, who was apparently willing to pay for more extensive and more learned material (the treatments of Books 16-24 average 330 lines each).

For classicists who know of Tzetzes from occasional forays into his scholia on Aristophanes or Lycophron (or the cranky notes he made in a Thucydides manuscript),2 or from following up references to his massive Chiliades, it is interesting to see the range of his styles in this work, from an overwrought ceremonial dedication at the beginning of the Prolegomena or sometimes convoluted allegories involving etymology, physics, and meteorology to rather smooth and simple paraphrase of the events of the plot. Hints of Tzetzes' distinctive personality emerge here and there, as in his strong identification with Palamedes, including alleged physical resemblance (Proleg. 724ff.), his use of astrological interpretations to explain the interventions of Homeric gods (e.g., 7.114ff.), defending the length of his treatment (18.645ff.), or his admiration of Homer as font of delight for simple folk and of knowledge for the wise (e.g. 20.33-47). In general, Goldwyn and Kokkini have done well as translators in what must at times have been a tedious task. Readers whose Greek is not up to handling Tzetzes on their own and readers without Greek will be able to get a good sense of Tzetzes' interests and methods, although the limits on annotation in this series mean that the reader receives less help than might be needed to recognize Tzetzes's uses of etymological punning in his allegories or his allusions to other texts.

I will offer some examples of occasional failures in the translation at the end, but first it is necessary to ask just what text this volume is translating. Coming to this work with no previous knowledge of it, but with an interest in Tzetzes and the possible influence of his notes and commentaries in the last period of Byzantine scholarship, I wondered about the textual history of this work. How many manuscripts exist? Who produced them, at what dates? What other evidence may there be for knowledge of this work? The translators are singularly uninterested in these questions. In their Note on the Text (p. 517) they report: "We have used the Greek edition prepared and published by Jean François Boissonade in 1851. It is very sound for a text edited some 150 years ago. Boissonade's edition was based on three manuscripts: Parisinus graecus 2707, dated June 12, 1300/1301 (A); Parisinus graecus 2705, fourteenth century (B); and Parisinus graecus 2644, fourteenth century (C)." The bibliography (p. 559) lists no other edition.

One ought to be immediately suspicious of an old edition that relies only on some manuscripts in the editor's home town. The site Pinakes lists about 40 manuscripts for this work; but in some (or most?) manuscripts the work is split up into sections and incomplete. The Paris B manuscript cited by Goldwyn and Kokkini is accessible online at the Gallica site with a very thorough description (dated 2012). This description refers to the edition of the Allegories in P. Matranga's Anecdota Graeca e manuscriptorum bibliothecis Vaticana, Angelica, Barberiniana, Vallicelliana, Medicea, Vindobonensi deprompta (Rome 1850), which also contains Tzetzes' scholia on his own work (pp. 599-618, 749) and some variant readings from six Vatican manuscripts (pp. 709-748). Matranga is more forthright than Boissonade in indicating that his edition is based on only some of the many available manuscripts: (p. 9) Quum nulla fere sit mss. bibliotheca, in qua Tzetzae allegoriarum partes non asserventur, Vaticanae quamplurimi codices, ut integrum opus evadat, commodam mihi praestitere utilitatem. Although Boissonade's version is frequently correct where Matranga's presents a corrupt or modified text, the latter's edition has the correct text in almost all the places where Goldwyn and Kokkini report they have corrected errors (mostly typographic) in Boissonade (listed pp. 519-521). More important, Matranga's version contains more lines in some books than Boissonade's version (for instance, 60 additional lines in Book 11, and 19 in Book 12).3 Many of these plus-verses paraphrase Homeric similes. Clearly, the meaning of this discrepancy cannot be firmly established until someone does the hard work of surveying a good number of manuscripts.4 But I see nothing in the style or content of the lines that is inconsistent with Tzetzes as author, so it is possible that the author himself tinkered with his work in the course of his long career and that we are dealing with two editions; or the shorter version may be the result of abridgement. While the translators cannot have been expected to explore what is apparently a complex textual tradition, it is a major shortcoming not to have compared Matranga's version and informed readers of the uncertain extent of the original Tzetzean text.

Boissonade's Greek text is elegantly reprinted with great accuracy, but I noticed the following errata or mistaken "corrections":

Proleg. 8: read τῷ for ᾧ at beginning of line; the type in Boissonade's edition was damaged here, but was obviously originally τῷ.
10.77: κατεκλίνθησαν should not have been changed by the editors to κατεκλίνησαν, since aorist ἐκλίνθην is found in Homer and other poets as well as in Byzantine authors.
16.223, 17.107: subscript missing in μάχῃ.
18.148: ἔζευξαι (phonetic error αι/ε) should have been corrected to ἔζευξε (read in Matranga's edition); the translation is correctly third person.
18.304: it is probably incorrect to change neuter παμμέγιστον to masculine παμμέγιστος, since the Homeric scholia treat πέλωρ αἴητον as a neuter phrase in apposition.
18.545: read ὂν for ὃν; again damaged type in Boissonade; the translation rightly gives "being."
18.640: Boissonade's ill-advised emendation ταὐτὴν should not have been adopted and translated ("the same"); demonstrative ταύτην gives the correct sense here.
18.746: there was no need to change εἵατο (not a form of ἐάω as the editors think) to ἥατο.
20.337: Boissonade's συνήκει is printed at the end of the verse, but the syntax requires emendation to the infinitive συνήκειν, which is indeed what the translation assumes.
20.246: the editors delete οὐ from the sentence, but that makes it unmetrical (14 syllables); the solution is probably that τίς is to be changed to indefinite τὶς, "one is not able to discern" (scribes' accentuation of τίς/τις is often unreliable).
21.55: read Δίες (as Matranga prints it), not δῖες; Tzetzes (like Eustathius, and earlier Chrysippus/Plutarch) uses this nominative plural of Zεύς.

Problems in the translation are scarce, but there is a cluster in the allegorical explanations in the later books. Some errors result from mistaking idioms common in Byzantine paraphrasing and etymologizing.

Proleg. 35-36: infinitive ἐξυλακτεῖν goes with οἰστρούμενοι, not with ἀναχαιτιζέσθωσαν.
Proleg. 644: κτείναντα τῇ πατρίδι (modifying τὸν Τεῦθιν) is translated as "was executed in his homeland"; "died in his homeland" (θανόντα) is the expected sense, for Tzetzes must be referring to the story in Pausanias 8.28.4-5 (and compare line 646), and we seem to have one of those places where antonyms have been confused.
8.175: ἐξ ὧνπερ ἀπαρύεται ἥλιος, γῆς ἐκτρέχων is not "due to which the sun sets, having traversed the earth" but " from which the sun draws off (moisture) as it runs out of the earth
12.26: ἐκτύπει is "resounded" (κανάχιζε in the original), not "splintered."
12.106: καὶ πάντα τὰ κριώματα μοχλεύοντες ἐρρίπτουν, not "and charged forward with all the battering rams, using them to break through" but "levering up all the projecting support-beams (of the wall), they threw them down" (cf. Il. 12.259 στήλας τε προβλῆτας ἐμόχλεον); Tzetzes apparently uses κριώματα in the sense of ἰκριώματα, not as a synonym of κριός.
13.78: ἀνθ' οὗ τοῦτον συνέλαβε πρὸς θάλασσαν ἡ μήτηρ, not "by whom his mother conceived him at sea" but "because his mother conceived him by the sea."
18.55: ἐν ταῖς παρύγραις μάχαις, not "in nautical battles" but "in battles by the water" (Tzetzes probably knows πάρυγρος from astrological texts; see TLG).
18.81: (στοιχεῖα) ἄποια, not "inert" but "without distinct qualities."
18.162-163: not "then he told his mother about the water divination etc." but "in this passage the poet means by 'mother of this man' the water divination etc."
18.190: μὴ μέντοι τοῦ ἀέρος γε has to be "not, however, of the air."
18.709-710: the translators introduce incorrect parentheses and misconstrue the nominatives in 710: the sense is "as the astrologers say, and also now/here the divine old man [Homer] (says), who except for Orpheus preceded all the astrologers."
20.85: Διῒ φωνὴν τὸν ἦχον λέγων, not "addressing Zeus in his own voice" but "(Homer) speaking of the echoing noise (of the sea) as if it were (Poseidon's) voice (addressed) to Zeus."
21.233: ἀπάτη δὲ τῶν νέων not "a new kind of deceit" but "a source of deception of the young" (that is, of those who are not guided to an allegorical reading by a learned teacher).
23.56: Δήλιος ὢν καὶ φανερός, καὶ ἥλιος ἐκθλίψει, not "being Delian and bright, and the sun <because it> will oppress" but "[Apollo] being Delian and clear to see [etymological play on δῆλος], and being (called) Helios by extrusion [of the letter delta from Δήλιος]."
23.78-79: not "Understand, just as Tzetzes, that the wind blowing from the west, the best of winds, is stormy, not an adverse wind" but "Do not understand the word δυσαῆ to mean 'blowing unfavorably,' but, as Tzetzes does, understand it as the wind blowing out of the west [etymological play of δυσ- and δυσμῶν], the fairest of winds."


1.   The seven lines that acknowledge the new patron are printed and counted as the first lines of Book 16 in Boissonade's version, but in Matranga's edition they are printed as the final lines of Book 15.
2.   For the comments on Thucydides, see the fascinating detective work done by Maria Jagoda Luzzatto, Tzetzes lettore di Tucidide. Note autografe sul Codice Heidelberg Palatino Greco 252, Bari 1999 (Paradosis, 1).
3.   Here are the line numbers for verses in Matranga's edition that are absent in Boissonade's: Proleg. 534-535, 642, 720 [and Matranga lacks line 120 of Boissonade]; 9.103-104; 11.60, 67-68, 77-81, 97-118, 133-135, 158-159, 164-166, 174-175, 183-190, 224-234 [one missing]; 12. 28, 34, 80-84, 152, 162-172; 13.64-66, 76-79; 15.213-215; 16.190, 198-199; 18.447; 20.172-173, 335-336; 21.91, 270-271; 24.11. (Both editions may be read at or
4.   A good dissertation topic for a Europe-based graduate student with strong training in palaeography and textual tradition.

No comments:

Post a Comment