The fields of oral-poetic theory and ethnopoetics approach challenges involving the meter and arrangement of lines, but from different perspectives (Hymes 1994, 330). Organization of lines, according to Hymes, is one of the likely definitions of poetry. There may also be organization within each line (Hymes 1994, 331). Hymes further expects to identify some sort of intonation which would show a pattern, rather than a more straightforward, step-y-step, progression. Of course, he observes, written traditions seldom indicate intonation (Hymes 1994, 332). Hymes looks for patterns of connectinon, which he describes as a "series of arcs" especially notable in poetry which is classified in groups of three uits (Hymes 1994, 332). Hymes illustrates the phenomenon, drawing on the analytical work of Nessa Wolfsom (Hymes 1994, 333 ff). Breaking the narrative into small sense units shows clear parallelism and arrangement. This is very like Chinook narratives transcribed by Sapir in 1909 (Hymes 1994, 335-337).
Hymes finds these same linguistic patterns in play in Alaskan societies and those of Native Americans in New Mexico (Hymes 1994, 338). The sense units not only group together, but they follow patterns in the number of units presented in different stanzas. Hymes finds a clear organizational pattern (Hymes 1994, 340). This suggests to him that it may be a natural means of mental organization.
Based on this concept, Hymes suggests that manuscript studies need to be re-evaluated. There may be elements of a narrative which have been ignored because scholars who did not recognize the structural elements considered some statements irrelevant (Hymes 1994, 342). Breaking of a regular pattern may also be evidence of a transcriptional error (Hymes 1994, 343). Hymes illustrates in his own work how a failure to recognize a small structural element can result in a misinterpretation of the importance of various statements.
Hymes sees a furtehr interpretive issue revealed in his process. There may be works with lines or statements missing, and which can be understood better by the ability to fill in gaps once a pattern has been identified (Hymes 1994, 350).
The questions arises, then, whether the editor of oral collections should place a higher value on the specific order and words of the narrator or possibly on the intention, once a pattern or structure can be identified (Hymes 1994, 352). The process of editorial review becomes even more important when we attempt to "repatriate" a text to its appropriate historic context (Hymes 1994, 353). Hymes finds a tension between the Native American value placed on the current spoken word and the value placed on bearing the message of ancestors. This becomes a matter of even greater difficulty when dealing with works from extinct languages (Hymes 1994, 354). There may be elements of word play and cultural allusion which could easily go undetected. Hymes illustrates the difficulties at length. He also notes (Hymes 1994, 358) that collections of oral material are rarely presented in any sort of chronological manner. Different versions of the same material may be collected at different times with no apparent rhyme or reason.
Hymes urges throught his article a careful approach to oral materials, inclduing those which could be classed as poetry. The skills of a philologicst may well be of more value than those of a collector or a collator.