Bryn Mawr Classical Review

Bryn Mawr Classical Review 2003.01.14

Gabriella Carbone, Il centone De alea. Introduzione, testo, traduzione, note critiche, commento e Appendice. Studi Latini 44.   Napoli:  Loffredo Editore, 2002.  Pp. 156.  ISBN 88-8096-885-8.  EUR 15.00.  

Reviewed by Thomas Opsomer, Fund for Scientific Research - Flanders (F.W.O. -Vlaanderen) at K.U.Leuven, Belgium (
Word count: 2088 words

This book is concerned with the Vergilian cento de alea (Anthologia Latina 8 in Riese's edition1). Its core consists of an edition of this poem, followed by an indication of the original Vergilian verses and a few notes of textual criticism, a translation and a commentary. In addition, it contains a relatively extensive introduction concerning the cento in general and an equally substantial chapter on ancient dice games, as well as brief discussions of various topics such as the poem's manuscript tradition, the centonist's technical abilities or lack thereof, the structure of the poem, and its images and style. A bibliographical list rounds off the book.

The book's greatest virtue is that it tackles a very difficult poem that has received little attention until now (but then again perhaps deservedly so -- see below), which is an act of bravery in itself. It may well stir up the discussion on de alea, and in any case it cannot be overlooked in future studies of the poem. Its biggest flaw, on the other hand, must be its longwindedness -- perhaps a surprising statement on a relatively short book that seems to be of such a functional nature. But in my opinion, the booklet contains little more that is both relevant and new than could have been presented in a more concisely -- and more carefully -- written fifteen-page article.

I shall first discuss the book's individual parts in some detail (I), which will enable me to answer the fundamental question, whether or not the poem makes more sense in the interpretation proposed by Carbone [C.] (II). After that, I will evaluate the qualities of this booklet as a whole (III).


The introduction on the cento in general (pp. 7-34) covers all the major issues and on the whole presents matters correctly, but it contains little that has not been said before.2 C. takes her time to define the nature of the cento, going in a roundabout way from stating that the cento is a technique rather than a genre (pp. 9, 10) to calling it a genre anyway (p. 26), in a discussion which is fairly theoretical throughout and erratically structured.3 Perhaps the most interesting pages are pp. 20-25 on the 'falso d'autore', where C. argues that in a way the Vergilian texts of origin can also be considered centonic.

The two pages on the poem's manuscript tradition (pp. 35-36) are not particularly up-to-date: although C. quotes M. Spallone's article on the subject,4 she follows Riese's praefatio in stating that the Codex Salmasianus (A) was written in the seventh or eighth century and that it has not been corrected by other hands than those of the scribe himself (A1) and De Saumaise (a), whereas Spallone dated it to the end of the eighth or the beginning of the ninth century and recognized at least five -- probably six or more -- different hands, stating that De Saumaise's corrective interventions are much overestimated (pp. 36-49, esp. p. 44 for the date; pp. 51-57 for the corrective hands).

The text (pp. 37-40) closely follows Riese's edition, except for a few corrections proposed by Schenkl, Salanitro, and Palla;5 only those by Salanitro and Palla, being subsequent to Riese's edition, are marked by italics.

The pages that list the Vergilian verses of origin and the notes of textual criticism (pp. 40-64) also contain next to nothing that is new, except for a one-sentence description of each original verses's meaning or context. Textual criticism is limited to corrections proposed by others. The list of Vergilian verses contains only a few additions to Schenkl's6: two minor remarks on vv. 59 and 69, the correction by Palla on v. 62, and one remark which I consider to be of interest on a larger Vergilian context being imported at v. 100. C. fails, however, to mend some rather more significant gaps in Schenkl's list.7

I did not find C.'s defense of the cento's author (pp. 65-71) entirely convincing. She finds that the technical imperfections of de alea are all permissible liberties. Indeed one must realize that the standards set by Ausonius's Cento nuptialis are unmatched by any other cento, so we do have to make some allowances. However, those who have compared the centones from the Codex Salmasianus from a technical point of view have found de alea to be technically inferior within this group, and their arguments are not refuted by C. The centonist is generally condemned for his lack of technical abilities because of the large proportion of incomplete lines or lines composed from more than two source-verses, the many alterations to the Vergilian verses, the repetition of the same hemistichs and the lack of clarity of the resulting text.8 As the last aspect seems to be what tips the balance (C. does not consider verses composed from frustula to be flawed, for instance, if their meaning is not problematic, p. 68), one's judgement of the poem's technical merits may depend on the extent to which one finds the text intelligible after all (see below). But if that is the case, the sense of this chapter, which evaluates the technical aspects of the poem 'prima ancora di tentarne l'interpretazione' (p. 65) becomes quite unclear.

I am not very familiar with ancient dice and board games, as C. clearly is, so I do not think I am in a position to assess the quality of the pages that explain the ancient terminology concerned (pp. 73-104). However, they do not seem all that relevant to the study of de alea. Apart from words such as numeros, nomina, ossa, casus, and loca, which are self-explanatory, the only technical terms in the cento are ima and collis, the meaning of which has to be inferred from another poem of the Latin Anthology.9 Therefore it escapes me why a treatment of about a hundred Greek and Latin terms (including the names of all kinds of throws, most of which are of unsure meaning), which do not even occur in the poem -- one must not forget that it is a cento and is thus restricted to Vergilian vocabulary -- is relevant to the topic. These pages may have some merit in themselves, but as a source on ancient board games they are not easily accessible because of the lack of indexes in the book.

How C. can infer with such certainty from the poem's limited evidence that the game described in de alea is the particular one called τάβλη, is equally unclear to me.10 In fact, the strongest indication that the poem is about dice games at all remains its title. In any case, this rather elaborate chapter seems to contribute very little to our understanding of de alea .

If one accepts C.'s interpretation of the poem (see below), the structure proposed on pp. 105-109 makes sense. Likewise, the translation (pp. 110-112) can generally be called reasonable if one accepts some less delicate hermeneutics to make the text interpretable (for instance: levium spectacula rerum (v. 10) is translated as 'gli schemi degli instabili dadi', ergo ubi delapsae (v. 17) as 'poi quando sono caduti i dadi', consilium quaerens (v. 18) as 'cercando di sapere il risultato' and ad vocem celeres (v. 40) as 'solleciti al richiamo (del gioco)'). Some indulgence is indeed called for: one should realize that the text makes no sense whatsoever when translated literally, so some bending and stretching is probably required. However, in some instances this is taken rather far (for instance, fremituque secundo (v. 61) is translated as 'con un tiro favorevole') and at times I cannot figure out how C. arrives at the translation given (for instance 'una facile fine si dà ai veri deboli' for veris facilis datur exitus umbris (v. 47) or 'attraverso le insidie del nemico' for per avia (v. 93)).11

The commentary (pp. 112-140) is especially long-winded and contains little more than paraphrases of the translation, repetitions of what has already been stated earlier in the book, and all too imaginative digressions. The smaller the branch one is standing on, the closer one should keep to the tree, so taking the interpretation to metaphysical dimensions seems particularly hazardous.12 C.'s thoughts on the ethical contrast in the behaviour of the two brothers -- we are already assuming then that the poem is about two brothers -- are so far from what can actually be found in the text that their relevance seems quite questionable (pp. 138-140). Concrete facts of the kind one is used to finding in philological commentaries, such as the observation that 'solidum' can mean 'money' in later Latin (p. 122), are extremely few and far between.

In the last chapter (pp. 140-149), the context of the Vergilian verses of origin is used as a starting point for a judgement of the poem's images and style. The fact that the centonist draws largely on the same stories in the Aeneid is valued positively by C. as a choice for 'coerenza d'immagine', but this might just as well be held against the centonist, as a token of a limited memory for source-verses.13

The bibliography on pp. 151-154 is eclectic and seems to focus on Italian publications.14


de alea is generally considered a very obscure text.15 After reading C.'s book, it has not become that much clearer to me. Her translation still makes little sense in places (e.g. vv. 47, 82, 95-96) even though it already stretches the Latin text to some extent. She certainly makes a story out of the poem, but she interprets it with such enthusiasm that she seems unaware of the precariousness of the whole enterprise, and an inclination toward Hineininterpretierung casts a shadow over many pages. Though C. makes her interpretation sound obvious, the case is not very cogently argued: her judgements often seem intuitive and her reasoning rather associative.16 In sum, her interpretation is supported by so little in the text of the poem that acceptance of it requires a leap of faith.

However, it would perhaps be unfair to condemn her interpretation for its lack of hard evidence in the absence of a better one. I think it reasonable to give some credit to an interpretation that can make sense out of a poem as hard to understand as de alea, and to consider it valuable until a better one can be found.17 An attempt to clarify the poem using information from other sources about dice games seems to be sound as such and in very broad terms, C. must be on the right track.


If C.'s interpretation is not entirely devoid of merit, it is still regrettable that it came to fill a 150-page booklet. This has led to the inclusion of a great deal of information which is not quite relevant or not so carefully elaborated as it could have been and to a tiresome succession of repetitions, burying those points that are of interest under a tedious mass. The same topics keep coming back in the various chapters: notes of textual criticism are recapitulated in the pages on board games (e.g. pp. 51 and 96), many ideas that occur in this chapter reappear in the commentary (e.g. pp. 91 and 116-117); the context of the source-verses indicated on pp. 40-64 also resurfaces there (e.g. pp. 115, 120, 127) and is discussed at length in the stylistic analysis, which, like the commentary, reiterates thoughts on the cento in general (e.g. pp. 117, 141-142). The repetitions which occur throughout the book are often literal or semi-literal.18

These repetitions make the absence of indexes (an index of technical terms and an index locorum would have been particularly useful) all the more regretable. The book is further marred by a deplorable number of misprints,19 ranging from innocent blemishes that can be attributed to the press (e.g. odd word-divisions such as sorti-sque or althou-gh) through incorrect use of italics or underlining when these do have a special meaning (indicating the exact words borrowed from Vergil, for instance) to wrong numbers in references and misquotations which can be seriously misleading.

The booklet thus puts the reader's patience to a serious test, which only those of us who really need to know everything there is to know about de alea are likely to pass. Still, if you were to consult but one book in order to get to know as much as possible about the poem, it cannot be denied that this book must be Carbone's, in spite of its many imperfections. Therefore, I think Italian-speaking students will most likely be the principal readership.


1.   F. Buecheler - A. Riese (Edd.), Anthologia latina sive poesis latinae supplementum. Pars prior: carmina in codicibus scripta, recensuit A. Riese. Fasciculus I: libri Salmasiani aliorumque carmina. Editio altera denuo recognita, Bibliotheca scriptorum Graecorum et Romanorum Teubneriana (Amsterdam, Hakkert, 1973 [= Lipsiae, Teubner, 1894]). The centones were not included in D.R. Shackleton Bailey's 1982 Teubner edition of part I,1 of the Latin Anthology (a decision rightly deplored in the reviews by Lebek [CR 35 (1985), 293-296 (pp. 293-294)] and Reeve [Phoenix 39 (1985), 174-180 (p. 175)]).
2.   The ground was already covered by introductions such as those by Giovanni Salanitro, Osidio Geta: Medea. Introduzione, testo critico, traduzione e indici. Con un profilo della poesia centonaria greco-latina, Bibliotheca Athena 24 (Roma, Ateneo, 1981), pp. 11-18 and 33-60; Rosa Lamacchia, 'centoni (centones)', in Umberto Cozzoli e.a. (Edd.), Enciclopedia Vergiliana, Vol I: A-DA (Roma, Istituto della Enciclopedia Vergiliana, 1984), pp. 733-737; Giovanni Polara, 'I centoni', in Guglielmo Cavallo - Paolo Fedeli - Andrea Giardina (Edd.), Lo spazio letterario di Roma antica. 3: La ricezione del testo (Roma, Salerno, 1998), pp. 245-275.
3.   On p. 141, the circle is closed, C. hesitating again to call the cento a genre. The pages on 'le regole del gioco' (pp. 30-34) are rather abstract and do not mention the truly technical aspects of the centonic process (for instance the size of the pieces borrowed, imported enjambements, vox communis, defects in metrical suture etc.); these are briefly mentioned on p. 14 under another heading.
4.   Maddalena Spallone, 'Il Par. Lat. 10318 (Salmasiano): dal manoscritto alto-medievale ad una raccolta enciclopedica tardo-antica', in IMU 25 (1982), 1-71.
5.   See 'Probae cento. Recensuit et commentario critico instruxit Carolus Schenkl. Accedunt tres centones a poetis christianis compositi', in Poetae christiani minores. Pars I, Corpus Scriptorum Ecclesiasticorum Latinorum 16 (New York, Johnson Reprint Corp., 1972 [=Vindobonae, Tempsky, 1888]), pp. 511-639 (p. 533); Roberto Palla, 'Risvolti di tecnica centonaria', in CCC 6 (1983), 279-297 (pp. 294-295); Giovanni Salanitro, 'Tunc nel codice Salmasiano', in Sileno 16 (1990), 313-315 (that the Codex Salmasianus often reads tunc instead of tum was already remarked by Schenkl p. 532 n.1.).
6.   Schenkl 1888 (see previous note), pp. 532-533. Incidentally, C. does not make it clear that Schenkl already published a complete list of source verses.
7.   At v. 14, she states just like Schenkl that omnibus extemplo is an adaptation of Aen. VIII, 178 omnibus in templis, whereas a Vergilian verse starting with omnibus extemplo does exist: Aen. VII, 276. At v. 94, she follows Schenkl in citing Aen. VII, 591 as a source, but omits just as he did Aen. III, 670 (AL 8, 94 is, in fact, a contamination of the two, differing from each by a single word taken from the other). (On p. 66 also, the Vergilian origin of both extemplo and dextra is ignored.) At v. 66, neither Schenkl nor C. mentions that Aen. IV, 285 and Aen. VIII, 20 are identical.
8.   See David F. Bright, 'Theory and Practice in the Vergilian Cento', in ICS 9 (1984), 79-90 (p. 87); Maddalena Valloza, 'Rilievi di tecnica compositiva nei centoni tramandati con la Medea dal codice Salmasiano', in Studi in onore di Adelmo Barigazzi. Vol I (Roma, Ateneo, 1986) [= Sileno 10 (1984)], pp. 335-341 (p. 336).
9.   AL 194R2. C. is unsure about the meaning of collis and ima (she proposes two hypotheses on pp. 95-96). Alfred Johannes Baumgartner, Untersuchungen zur Anthologie des Codex Salmasianus [Diss. Doct. Zürich] (Baden, Köpfli, 1981) -- whose 43-page account (pp. 95-141) of board games in the Latin Anthology is not mentioned anywhere by C. -- already noted the uncertain meaning of these two terms in his commentary on AL 194 (p. 124). Imum may be a technical term in v. 47 (ima) whereas in vv. 7 and 55 (imo) it simply means 'bottom'.
10.   For instance, one of the steps of her reasoning is her interpretation of v. 10, per varios casus: "È evidentemente riferito alle cadute dei dadi, alle loro combinazioni, al loro ordine e non al loro punteggio complessivo. Da questa indicazione comprendiamo che il centonario sta descrivendo un gioco in cui il movimento delle pedine è connesso all'ordine di caduta dei dadi e ai singoli punteggi, e non al totale raggiunto dalla loro somma" (p. 93). This reading eludes my interpretative skills.
11.   In some places, another translation seems preferable to me: e.g. nati melioribus annis (v. 37) means 'born in better times' rather than 'nel fiore dell'età', and at vv. 43-44 (nec mora missus adest fati sortisque futurae. / scinditur incertum studia in contraria vulgus.), I would punctuate differently (after adest) and connect fati sortisque futurae to incertum instead of missus: "and without delay, the throw is at hand. Unsure about fate and future, the people are torn into contrary sympathies".
12.   The commentary abounds with phrases like "Non accetta [sc. la Fortuna] alcun atteggiamento irreverente, non sopporta l'orgoglio, e diventa una giusta dispensatrice di beni, quando qualcuno si crede esperto del gioco e fa diventare merito personale il favore che il destino gli ha concesso" (p. 122); "il rimprovero di una lunga riflessione, che muove ben oltre il piano del gioco per allargarsi a considerazioni esistenziali" (p. 131); "Entrambi, in quest'attimo di silenzio, sentono tutta la gravità della situazione, tutta la colpa del loro comportamento precedente" (p. 139).
13.   C.'s is not a novel approach: see José-Luis Vidal, 'La technique de composition du Centon virgilien Versus ad gratiam Domini sive Tityrus (Anth. Lat. 719a Riese)', in REAug 29 (1983), 233-256 (pp. 237-256), who uses charts of the proportional origin of the borrowed verses as the starting point of his analysis.
14.   Bright 1984 (see note 8) may not be indispensable, but the omission of Baumgartner 1981 (see note 9) is quite disconcerting. Anne Friedrich's book, Das Symposium der XII sapientes. Kommentar und Verfasserfrage, Texte und Kommentare 22 (Berlin - New York, De Gruyter, 2002) appeared to late to be taken into account by C. (the observations on pp. 81-99 on the monosticha de ratione tabulae (AL 495-506R2) may be of interest for de alea as well). Argonautica is the title of Françoise Desbordes' book, not a journal of which issue 19 would contain her article ['Le corps étranger. Notes sur le centon en général et la Médée d'Hosidius Géta en particulier', in Françoise Desbordes, Argonautica. Trois études sur l'imitation dans la littérature antique, Collection Latomus 159 (Bruxelles, Latomus, 1979), pp. 83-108].
15.   The fact that C. thinks the poem deals with a dice game, while Ermini considered it to be about a bloody combat, is a tell-tale sign [Filippo Ermini, Il centone di Proba e la poesia centonaria latina (Roma, Loescher, 1909), p. 42]. Most recently, Scott McGill called it a caligo densa ['Tragic Vergil: Rewriting Vergil as a tragedy in the cento 'Medea'', in CW 95.2 (2002), 143-161 (p. 144)]. C. herself admits as much in places: "il passagio è abbastanza difficile ed oscuro" (p. 137); she talks of a 'senso di smarrimento' in the reader, who feels the need for more clarity, a clarity which is hindered by the Vergilian origin of the verses (p. 148). But she also states repeatedly that for the ancient reader, the text must have been as clear as crystal (pp. 116, 147).
16.   For instance, on p. 119 C. writes that rapere should be understood as if spolia victi were the object of the verb. She supports this with a note referring to AL 194R, where spolia victi does indeed occur, but rapere does not. On p. 126, she writes that "dal bussolotto continuano a presentarsi punteggi-fantasma, se fosse vero che anche nel termine umbra sia da vedere una parola del linguaggio ludico". The logic behind these statements does not seem to be compelling.
17.   C.'s ambitions are formulated in a similarly modest way on p. 109.
18.   The most prodigious example is perhaps note 135 on p. 94, a three-line footnote which also occurs literally on p. 121 as note 157 (except for the altering of 'del resto' in 'tra l'altro'), in both cases supporting the statement that v. 17 is to be explained as ergo ubi tesserae delapsae sunt.
19.   On p. 17 line 29: read "imitation" instead of "imitaion"; p. 18 l. 34: read "voleur" instead of "vouleur"; p. 19 l. 37: read "connaissait" instead of "conaissait"; p. 36 l. 22: read "Schedae Divionenses" instead of "Schaedae Divionenses"; p. 39 l. 2: "risuque" should not be italic; p. 42 l. 11: read "II, 125" instead of "VII, 125"; p. 42 l. 18: underline "atque"; p. 44 l. 20: read "VII, 446" instead of "IV, 477"; p. 59 l. 8: underline "socios"; p. 62 l. 13: read "volte" instead of "vole"; p. 65 l. 16: add "57"; p. 92 l. 28: read "volvuntur" instead of "volvulntur"; p. 93 l. 3: read "7" instead of "6"; p. 94 l. 17: read "4" instead of "3" and "vario" instead of "varo"; l. 22: read "43" instead of "42"; p. 95 l. 19: read "5-6" instead of "3-4"; l. 25: read "47" instead of "46"; p. 98 l. 19: read "54" instead of "53"; p. 99 l. 11: read "ac pugna" instead of "et pigna"; l. 33: read "cauculus" instead of "calculus"; p. 107: l. 37: read "84-87" instead of "85-87"; p. 123 l. 14: read "contulimusque manus" instead of "contulimus manibus"; p. 125 l. 31: read "facilem" instead of "facile"; p. 128 l. 13: read "56-60" instead of "57-60"; p. 132 l. 4: read "72" instead of "71"; p. 133 l. 21: read "murmura" instead of "murmure"; p. 135 l. 6: read "tum" instead of "tunc" (see note 5); p. 143 l. 29: read "III, 56" instead of "VII, 56"; last line: read "78-79" instead of "78" and "107-108" instead of "107"; p. 144 l. 1: add "II, 61" between "II, 316" and "I, 218" and read "316-317" instead of "317"; p. 145 l. 8: read "v. 57" instead of "1" (I presume); last 3 lines: vv. 6, 71, and 104 have a source verse referring to Aeneas or Turnus, but not in "Aen." XI or XII; v. 90 has no source verse referring to Aeneas or Turnus; p. 147 l. 11: read "55" instead of "54" and "44" instead of "55". This list is not meant to be exhaustive.

Read Latest
Index for 2003
Change Greek Display
Books Available for Review

HTML generated at 13:28:22, Friday, 03 April 2009