Macrobius' Saturnalia is cast in the form of a learned dialogue between more than a dozen aristocratic pagans over the eve and three days of the December festival in c. a.d. 382. In this ‘colloquium liberale’ (Saturnalia 1.1.1) the interlocutors discuss a large number of subjects in rhetoric, philosophy and religion, and pay particular attention to the works of Virgil as both poet par excellence and the source of much other learning. Some decades ago, Alan Cameron convincingly re-dated Macrobius' composition of this work to the 430s, a half-century or so after its dramatic late-fourth-century setting (A. Cameron, ‘The date and identity of Macrobius’, JRS 56 [1966], 25–38); more recently, he has re-visited and adjusted some details of these arguments (A. Cameron, The Last Pagans of Rome [2011], pp. 231–72). Cameron's re-dating, adopted by most scholars including K., distances the author from his characters and makes his portrait of elite antiquarian culture all the more poignant; Macrobius was a learned, probably Christian writer, imagining pagans of an earlier, more fraught period of religious relations relating to, discussing and deploying still earlier classical erudition. The speakers in the Saturnalia thus represent what K. has called in the introduction to his recent Loeb translation a ‘highly idealized and highly stylized republic of learning’ (R. Kaster, Macrobius Saturnalia [2011], vol. 1. p. xviii). Important though this text is for understanding the state, style and ambitions of aristocratic learning in the fifth century, and its refraction of classical and pagan pasts, it has not been well served by all of its editors. Here, K.'s Studies on the Text of Macrobius' Saturnalia (2010) provides a comprehensive account of the history of modern editions of the text, and should be consulted alongside his excellent new edition here reviewed.
L. von Jan's 1852 edition was the first to make use of a large number of pre-humanist manuscripts but suffered from an extremely inaccurate critical apparatus. F. Eyssenhardt's edition for Teubner (1868, 2nd edn. 1893) was fatally flawed by its use of a remarkably limited number of manuscripts: two for the first three books, and only one for the remainder. In the mid-twentieth century, A. La Penna published an important article which both surveyed and added to Jan's list of manuscripts (A. La Penna, ‘Studi sulla tradizione dei Saturnali di Macrobio’, ASNP 22 [1953], 225–52); he was also the first to identify the three families of the medieval manuscript tradition, although the relations between these families have since been refined. J. Willis produced a new edition of the text for Teubner (published in 1963 and revised in 1970 and 1994) which has been the standard edition ever since. However, although it offered an improved text, it was criticised at the time for its patchy and inaccurate apparatus. K.'s landmark new OCT edition is a definitive corrective to this history; it is also in some senses an unexpected gift. He initially undertook to produce a new translation of the text for the Loeb series with the intention of using and improving on Willis' edition on the same lines as Marinone had done for his edition and Italian translation of 1967 (2nd edn. 1977). However, he discovered that Willis' apparatus was consistently and seriously inadequate in its failure to report errors in most of the manuscripts consulted; a sobering sample of a number of passages in which Willis misrepresented the text's documentary basis is provided in K.'s Studies, pp. 103–11. This dismaying discovery led K. to return to the manuscripts to produce this new critical edition which also provides the text (with a necessarily more minimal apparatus) for his 2011 Loeb translation.
The preface to K.'s edition contains an abbreviated summary of his findings on the medieval manuscript tradition of the Saturnalia, and a usefully detailed account of the manuscripts on which his text and apparatus have been built. There are only fleeting traces of the Saturnalia in the 500 years after its composition; the direct tradition first appears in the early ninth century on the continent and seems to have been restricted to France for at least the next century. Most of the extant manuscripts date from the ninth to the twelfth centuries. K.'s task was, then, to check and correct Willis' reports of thirteen manuscripts, to use collations of two manuscripts made by M.J. Carton after Willis' edition, to make a census of other pre-thirteenth-century witnesses, and to collate a further 21 pre-humanist manuscripts. K. then dispenses with two manuscripts (T and Z) used by Willis and makes use of a number of others ignored or not known by Willis. He confirms that there are three families of manuscripts descending from a single (probably late-eighth or early-ninth-century minuscule) ancestor, an identification first made by La Penna and refined by Willis who showed that two of the families, β1 and β2, were closely related to each other through a common ancestor, β, that stood between them and the archetype. After meticulously examining each family in turn, and reporting a selection of singular uncorrected and uncorrectable errors in each manuscript, K. constructs a comprehensive stemma (p. xxvi) which both adds to and refines that of Willis (J. Willis, ‘De codicibus aliquot manuscriptis Macrobii Saturnalia continentibus’, RM 100 [1957], 152–64 at 156).
K.'s preface contains a catalogue of errors in the citations of Greek works which he judges to be explicable by some kind of simple mechanical confusion on the part of copyists. Scribes from the Carolingian age onwards were often challenged by Greek, and thus when Greek citations were not omitted altogether, they were often copied with errors which must be put down to the copyists' failure to understand the text. K.'s decision not to clutter the apparatus of his text with the record of these kinds of errors but to preserve them separately is sensible, since it will surely, as he suggests, prove useful in the future to scholars interested in scribal capacities and behaviours. The reader interested in the complex business of dealing with Macrobius' citations of and allusions to a range of Greek and Latin texts, must consult K.'s Studies, pp. 65–84, for an account of the editor's guiding principles. Macrobius himself, as K. warns in his Studies, p. 65, must be regarded as something of a copyist. His editor must thus regularly reflect on whether he is correcting the author rather than his copyists, and attempt to distinguish between the kinds of error Macrobius was likely to have made, and those his later copyists were likely to have made. K. has constructed and consistently applied criteria for ‘constructing a plausible Macrobius’ (Studies, p. 67) which will be useful for scholars of this text and of others which offer analogous problems. Furthermore, K.'s painstaking identification and reportage of testimonia in the apparatus of this edition vastly improves on Willis' incomplete and inconsistent apparatus. Taken together, K.'s edition, translation and study of this text are a monumental achievement; his research has finally produced a text which inspires confidence. This edition will transform how scholars study Macrobius' command and use of classical learning, as well as allowing them to engage with important issues of transmission such as the levels of Greek of medieval scribes.
