Expectable Phraseological Makeup

As noted above, we shall be concerned in this chapter with the theme as defined and illustrated by Arend, Lord, and Edwards, that is, as a self-contained

[9] In a footnote (4n4) Nagy (1979) adds: "Even from a descriptive point of view, I will consistently argue that Homeric epithets are indeed appropriate to the themes associated with the words that they describe."

[10] Cf. the observations of Hansen (1978, 18): "The forms of themes are usually not so fixed that the poet need only insert the appropriate names into the proper slots. It is frequently not possible to classify themes (or formulas) neatly, since the one gradate, into the other, and a given passage is often a blend of materials elsewhere appearing independently. Like the verbal formula, the theme is an analytic convenience."

[11] See further Edwards 1980a, 27; Edwards 1980b.

unit describing a single event. The more extensive patterns, from Fenik's more heterogeneous "typical scenes" of battle to Hansen's "narrative sequences" and on to the tale-type or "story-pattern," lie beyond the scope of this chapter.^[12] Although these other patterns are worthy subjects and, to the extent that they also illustrate the central role of multiformity in oral and oral-derived epic, are certainly relevant to the concerns of the present volume, we shall be concentrating on scenes that are readily comparable to the units found in Serbo-Croatian and Old English epic, and thus shall limit our analysis to the theme in the more restricted sense of the term.

Returning now for a moment to the findings of chapter 4 on Homeric phraseology, let us consider what we might be able to predict about the verbal dimension of these narrative structures. We established that the diction of the Odyssey , which developed in symbiosis with the hexameter, was a tradition-dependent diction that took its shape from the idiosyncratic prosody to which it was partner. Thus the phraseology was shown to be based on the inner metric of the hexameter, a complex poetic line with a complicated system of colonic shapes.

While the normative unit of meter turned out to be the colon, the minimal normative unit of phraseology proved to be the hemistich. Nonetheless, our analyses of elements of diction were unanimous in denying primacy to any one "length" for either "the formula" or "the formulaic system." From epea pteroenta through the twenty-one-line passage from Book 5, the evidence was the same: the diction simply could not be fairly reduced to a single type of phraseological pattern but had instead to be understood as a spectrum of phraseology. Hemistich, whole-line, and multi-line units were encountered; what is more, in many cases a given line seemed to be related to more than one of these levels in the diction, so that it was impossible to point to any single phrase as the primary unit, the core around which the others were formed.

Since (quite arbitrarily) denominating one level of structure as primary necessarily meant shortchanging the relative importance and the dynamics of other levels, I proposed a new solution to the riddle of the phraseology that avoided giving false prominence to any one proposed configuration and thus avoided reducing the plasticity and vitality of Homeric diction to a collection of mechanical parts. This solution involved the derivation of a number of traditional rules for the formation and maintenance of the phraseology from the principles of comparative prosody developed in chapter 3. Along with the Indo-European principle of right justification, we demonstrated the significance of word-type localization for the structure of hexameter phraseology. It was observed that these rules allowed us to explain the shape of formulaic

[12] See chapter 10 for a discussion of story-pattern.

and (on available evidence) non-formulaic lines from the Odyssey , and in so doing to reach well beyond conventional formulaic analysis.

Of course, Homeric phraseology, like other traditional phraseologies treated in this volume, was in a constant state of evolution, and we recognized this developmental fluidity by noting cases in which formulaic structure has superseded minimal word-type localization. In these cases, usage has created families of formulaic elements or patterns that in their group coherence deflect the percentage occurrence figures for certain words away from the average percentages for overall word-types. That is, the individual lexical elements that make up these formulas and systems, although admitted to the phraseology on the basis of traditional rules, now skew the average figures because of the utility of the larger phrases in which they are embedded. This phenomenon also contributes to the varied spectrum of traditional phraseology.

Thus, in turning to the three analyses of thematic structure to follow, we may conceive of the verbal data for those narrative units as a spectrum of different kinds of elements rationalized by traditional rules. These rules provide the matrix through which all phraseology must pass in order to enter the tradition; they will call for right justification, word-type localization, and observance of the other features of the hexameter's inner metric.^[13] Once having entered, these elements may evolve in virtually countless ways: they may form composite phrases, formulas, and systems; they may remain in their original form, at least for a time; they may disappear from the repertoire of a singer, local tradition, or the tradition at large; or—perhaps—they may come to be marshaled by thematic structure.^[14] Whatever the individual case, we may be sure that the general relationship between thematic structure and phraseology will be at least as complex and many-sided as the diction itself, and further that the diction used to express the traditional ideas of themes will be overseen by traditional rules.