Prior Scholarship
Albert Lord began the description of thematic structure in 1938 with his study of narrative inconsistencies in Homer and South Slavic epic, and thirteen years later he focused on thematics as a compositional process in both poetries (1951a).[2] But while these articles proved influential in the spread of oral theory to other literatures, his major statement on the theme was to be, and remains, the fourth chapter of The Singer of Tales (1960). For Lord the most important characteristic of the theme is its multiformity : as a protean narrative pattern of consummate utility to the composing oral poet, the theme provides both an outline of related actions balanced from within and a kernel of verbal correspondence among its various instances. Apart from the external profile a scholar assembles to help keep track of elements associated with the pattern, the theme has no "norm" or archetype; it takes shape only in the variety of its instances, with all occurrences being equivalent (if differing) variants of the unit.[3]
The first of Lord's definitions in Singer (p. 68) is the most general: themes are "the groups of ideas regularly used in telling a tale in the formulaic style of traditional song." But although this unit of narrative exhibits a formula-like quality, it is most essentially a multiform and can take many (theoretically an unlimited number of) shapes. As he puts it (p. 69), "the theme, even though it be verbal, is not any fixed set of words, but a grouping of ideas." Lord describes and shows how variation can occur and how the degree of phraseological correspondence among instances may differ according to the given theme, singer, song, and circumstances of performance. We learn in addition that the pattern may be broken down into smaller narrative elements, a composite structure of motifs or minor themes that has attracted the attention of other scholars seeking to explain the structure of Serbo-Croatian epic.
As he did with the formula, Lord emphasizes (p. 86) the usefulness of the theme, noting that "it is always at hand when the singer needs it; it relieves his mind of much remembering, and leaves him free to think of the plan of the song itself or of the moment of the song in which he is involved." By the same token, certain multiforms are termed "ornamental," since they broaden or further articulate a scene or description without necessarily moving the action forward.[4] Whatever the case, however, there exists a discernible order in the theme, and the poet senses or follows that order even if he does so in an idiosyncratic manner, for within the general plan there is much room for individual variation:
In all these instances one sees also that the singer always has the end of the theme in his mind. He knows where he is going. As in the adding of one line to another, the singer can stop and fondly dwell upon any single item without losing a sense of the whole. The style allows comfortably for digression or for enrichment. (p. 92)
The order and balance so characteristic of the unit may proceed from any number of sources—a question-and-answer series, the logical sequence of events that are comprised in the dressing or arming of a hero or the caparisoning of a horse, and so on. The actual verbal means by which the singer travels these cognitive pathways varies from performance to performance, from singer to singer, and from one local tradition to another.
In their combination with other such units in a song, narrative themes present a double aspect. They have a certain associative attraction to other themes, and yet at the same time they maintain an integrity and completeness of their own. This quality of relative independence allows the kind of ubiquity that constitutes the lifeblood of an oral narrative tradition—without such a general application, any unit would prove of little utility in traditional composition; the same ubiquity, however, can also lead to narrative inconsistencies when two or more juxtaposed themes are not thoroughly adapted by the singer to their role in the given song he is composing.[5] As Lord remarks (p. 94), "the theme in oral poetry exists at one and the same time in and for itself and for the whole song."
Of Lord's studies since The Singer of Tales , the most important for the present purpose is "Perspectives on Recent Work on Oral Literature" (1975).[6] In the midst of a survey of scholarship on various traditions, he takes up in a precise manner the question of verbal correspondence among instances of a theme. His aim is to distinguish the more usual literary sense of the term, which implies a subject only, from the special sense in which he means it—as
the recognizable narrative unit of oral epic composition that assumes a similar verbal form in every occurrence. He thus argues (p. 20) that the theme as subject has no real bearing on oral literature research, "but if by theme one means a repeated narrative element together with its verbal expression, that portion of a poem, an aggregate of specific verses, that tells a certain repeated part of the narrative, then we find ourselves dealing with elements of truly oral traditional narrative style." Thus, although Lord does not demand a verbatim correlation among instances of a multiform (else the multiform would become a mere literary fossil and cease to be useful to the singer as he composes in the traditional way),[7] he does demand some concrete verbal dimension as a requirement of the unit.
David E. Bynum's complementary studies of the South Slavic epic theme began with his 1964 dissertation, in which he established a directory of recurrent multiforms based on seventy-five Wedding Songs drawn from the Milman Parry Collection and six published sources. In doing so he defined the theme (p. 39) as "a conglomeration of narrative matter in oral epic tradition which recurs in the tradition, and which is discrete because some of its occurrences have no consistent sequential relationship with other such units. Defined internally, it is a conglomeration of narrative the parts of which, if they are present, occur regularly together." It will be noted that Bynum includes no general requirement of verbal correspondence, that his concept of the theme is of a group of ideas that agglutinate and persist as a unit in oral tradition. A few years later, in a study devoted to themes of the young hero in Serbo-Croatian epic (1968), he would reiterate this concept in its essentials and point out that the order of elements or motifs within a multiform is irregular and further that associative clustering rather than sequence seemed to be the ruling force within the theme.
In two later works Bynum emphasized this associative dynamic and turned his model to acts of criticism. A 1970 article on thematic pattern illustrated how the poet and tradition transform a generic character by making him the agent or focal point of certain multiforms. Far the more ambitious project, The Dæmon in the Wood (1978) considers a worldwide mélange of oral fable from the point of view of motifs involving trees. Bynum argues persuasively for a single tradition of fable by treating diverse sources with a taxonomy founded on the sort of theme he had been describing in earlier studies. Themes are "the bounded kind of motival clusters," as opposed to an unbounded cluster not under the aegis of narrative pattern; the generic motifs involved in a theme arc "like the fibres of a timber or the atoms in a molecule, capable
of analytical separation from patterns but incapable of entering into the composition of traditional stories except as integrants of patterns, where they are held indissolubly in their orbits of association with other motifs by the powerful cohesive force of story-telling custom" (pp. 79-80).[8] This last point will prove an important one: as presaged in Lord's phrase "tension of essences" (1960, 97), themes—and the motival elements that comprise them—cohere not by the mechanistic imposition of a narrative latticework but rather through the mutual attraction of the involved units for one another, an attraction memorialized by thematic multiformity and ubiquity within the tradition. It is a cohesion that emerges over time, as the singer and tradition use and re-use a theme, group its parts, and reconstrue its whole; the cohesiveness that appears to us to be synchronic fixity is in reality the associative pull of diachrony.
David M. Gunn also contributed to the study of the theme in Serbo-Croatian epic by comparing selections from volume one of Serbo-Croatian Heroic Songs with passages from the Iliad and Odyssey . In the first of two articles (1970) he argues for rejection of the verbatim memorization theory for ancient Greek epic texts and for adoption of the oral-dictation hypothesis, adducing an inconsistency in a song by Sulejman Fortic[*] as an analog to Homeric compositional practice. His comparison depends on Fortic's[*] apparent attempt to ornament a theme and employs the concepts and definitions developed by Lord.[9] His second and more extensive study (1971) presents more detailed thematic morphologies for the two traditions and works toward a claim of unity of authorship in Homer based on the distinctiveness of narrative design in the Iliad and Odyssey :
We find idiosyncrasies in vocabulary, in formulaic expression, in preference for certain elements, and in the particular sequence of elements whether couched in formulaic language or not. Moreover, in the instances of a singer's theme we can discern the degree to which the theme in whole or in part is "fixed" in the singer's mind. (p. 3)[10]
Gunn's emphasis on individually tailored occurrences of a theme thus harmonizes with Lord's ideas on verbal multiformity (although with an unambiguous basic correspondence among instances in terms of repeated phrases and words)
and with Bynum's observation that the sequence of elements within the unit may permute and combine in various orders.
This collective stress on avenues of flexibility within limits was continued by Mary P. Coote in her thoroughgoing analysis of themes in the Parry Collection guslar Camil[*] Kulenovic's[*]The Captivity of Vrhovac Alaga (1980). Coote (p. 202) prescribes two aspects for the multiform: the paradigmatic theme , "a general narrative idea which the singer restates in performance using specific motifs"; and the compositional theme , "a piece of narrative that repeats a set of formulas." Although she finds the Kulenovic[*] song-text entirely thematic, not all of its multiforms arc compositional—that is, not all of them are appreciably formulaic with respect to other occurrences. The wide variation in density of verbal correspondence leads her to posit not a single type of narrative unit but rather a taxonomy of themes (p. 233): those called formulaic demonstrate at least a 75 percent density of formulaic phraseology and consistency in the order of elements or motifs; marked multiforms reveal neither significant phraseological stability from instance to instance nor a consistent sequence of motifs but are characterized by certain recognizable phrases or clusters of phrases; neutral themes express a paradigm without the features of the marked pattern; unusual multiforms also exhibit a common paradigm but include rare motifs; and rare themes are those used infrequently by a singer. Coote shows that Kulenovic[*] depends chiefly on compositional, or formulaic, themes for his song-making, and that compositional and marked multiforms together constitu e about 85 percent of the 3,388 lines in the Vrhovac Alaga .[11] In other words, the singer employs most often those themes that have a heavy or at least discernible formulaic content. In this respect Coote corroborates Lord's insistence on significant verbal correspondence as a necessary hallmark of the narrative theme.
In sum, research on the theme in Serbo-Croatian oral epic has produced several points of consensus. First, there is general agreement that a "formulaic" narrative unit, multiform and therefore suitable for many different contexts, underlies the guslar's compositional practice. Second, all investigators concur that the traditional quality of this unit—both its integrity as a complete pattern in itself and its associative links to other multiforms—make it a crucially important feature of the singer's craft. As for the actual structural makeup of the theme, however, we find some disagreement. Lord prescribes, most forcefully in his later writings, a necessary aspect of verbal correspondence; without this phraseological core, he stipulates, we have no theme. Bynum defines the multiform somewhat similarly but without the criterion of verbal stability, seeking in his most recent work to relate narrative themes across traditions to a worldwide pattern of fable. Gunn follows Lord's model rather
closely in his comparative analyses, but Coote enlarges the concept to include five kinds of themes, a taxonomy founded on differing formulaic content and on adherence to an order of subunits.
Clearly, then, there has been much evidence brought forward for the existence of a narrative multiform and enough unanimity of scholarly opinion registered to make possible more than a rough sketch of the unit and its traditional importance and behavior. Just as clearly, however, we can discover no consensus on its internal structure, specifically on the questions of its phraseological texture, the significance of the order of its constituent parts, and the exact relationship among various occurrences. In pursuing answers to these questions, it will be well to turn to the texts themselves as the primary witness on thematic structure. Let us begin where the Stolac guslar Ibro Basic[*] leaves off—at a narrative seam incised by his pause for rest before resumption of performance and story.