Skip to main content

Counting to One: A Step toward Understanding the Homeric hapax ezeugmena

The Homeric hapax ezeugmena (noun-epithet combinations that occur only once) are an important element in epic composition; they were gathered in a section called “Repertory II: Frequency Lists” in Iuncturae Homericae (Dee 2010) but left without further analysis. This paper will examine from several different perspectives this category, which, perhaps surprisingly, comprises almost two-thirds of the 6005 iuncturae in the epics, concentrating on certain types of expressions and assessing their significance for the “oral-formulaic theory,” the most extreme forms of which suggested that the entirety of the two poems was formulaic. That is a position few would endorse now, but the sheer mass of these hapaxes requires attention; the ca. 4000 two-unit phrases together with “entrained” prepositions, particles, and other little words, must account for more than 10,000 of the 200,000 words in the corpus—a quantity that strongly implies the possibility of “coinage” by the poet.

Because of the obvious impossibility of dealing in detail with each of ca. 4000 hapaxes in a fifteen-minute paper, the focus will be on a few specific avenues of approach. One is the “degree of formularity,” culling from the full set those that exhibit parallelisms or analogies with established formulas. The linkage may be metrical (occupying identical cola) or semantic (serving as near-synonyms or variants). At the opposite end of the spectrum are expressions that seem like improvisations or at least appear formed for a specific and unique context—and thus deserve consideration as coinages. If one of the words in the iunctura is a hapax legomenon (extensively studied in Kumpf 1984), a relatively rare occurrence within the four-thousand-member set, that fact might increase the probabilities; but most of the phrases in question are harder to evaluate. A third category that needs examination are those iuncturae that might be called “vanilla,” since they have a generic or “all-purpose” quality that might lead us to expect more than a single instance in more than 27,000 hexameters. Some of these are, or are part of, elements labelled “under-represented formulae” (Hainsworth 1968 and the Cambridge and Oxford commentaries). No broad generalizations can be drawn from such select subsets, but there will be an overview of the rôle of this category as a critical element in the Homeric narrator’s toolbox.

The paper will conclude with some observations based on the particular examples given and will describe a larger and more challenging project that would attempt to gather and organize, not phrases or formulae, but full syntactical units (sentences anchored by a main indicative verb), as a way of reaching a deeper understanding of the nature of Homeric narrative at the level that matters most, viz. the minimum core of the epic composer’s technique, the complete independent utterance. A tentative beginning appeared three decades ago (Visser 1987), but there is a rich vein waiting to be explored.

A handout will offer a summary of the paper’s principal data and a catalogue of the iuncturae that are discussed.