a1 Corpus Christi College, Oxford
In Latin Hexameter Verse, his 1903 manual for composers of Latin hexameters which is still useful as a guide to Vergil's metrical and prosodic practices, S. E. Winbolt states that a hexameter ‘must not end with an adjective preceded by a noun with a similar short ending, e.g.…flumina nota’ unless the adjective is emphatic, ‘i.e. strongly distinctive, predicative or antithetical’. Whether or not his distinction between emphatic and non-emphatic adjectives in this position is wholly workable (predicative adjectives are clearly distinguishable, but it is not clear that the other types are), Winbolt here rightly detects a strong tendency in Vergil and other Latin poets towards avoiding endings of this general kind, which we can conveniently call the ‘Discordia taetra’ type after one of its earliest and best-known instances in the Annales of Ennius (225–6 Skutsch ‘postquam Discordia taetra/Belli ferratos postes portasque refregit’). The rarity of this type of line-ending is clear in Vergil; there are only 16 examples, regardless of whether the adjective is emphatic or not, in the 9890 lines of the Aeneid. Such a select and easily-defined phenomenon might prove a yardstick of some interest in the history of the Latin hexameter, for it seems to raise a number of questions to which the answers would be significant and useful. Is this type of ending avoided equally by all poets? Is there an increasing tendency to avoid it as time goes on? Is it associated with any particular genres of hexameter poetry? Do poets tend to use in it the same words or phrases as their predecessors? To discover the answers, this article will look at the ‘Discordia taetra’ phenomenon in Latin hexameter poetry, defining it as the instance where a noun ending in a short vowel (in practice, in ‘-a’) is immediately succeeded by an adjective of similar ending and in agreement at the end of the hexameter, and where such a noun is not a substantivised adjective and such an adjective is neither predicative nor a participle.
* My thanks to Professor R. G. M. Nisbet and to the editors of CQ for helpful discussion and criticism.