Part of my work for the WIVU last year included determining the mainline in the text for each of the chapters. That meant determining what clauses moved the story/discussion/dialogue further and what clauses were merely supporting and/or interruptive material. To make it a slightly less subjective activity, some of the ways to determine what fit into the mainline included: a repetition of verbal forms, repetition of person, gender, and/or number, logical connections (e.g. conjunctions and prepositional forms), and general connections between all of these elements.
If you have experience in creating clausal outlines of texts and/or trying to determine the mainline in texts, you will know that clausal outlines for narrative texts are easier than those for poetical texts. In narrative texts, there is a fairly obvious main verbal form (wayyiqtol), and clauses not containing this form are generally not included in the mainline. This makes clausal outlines for narrative texts fairly objective.
Poetic texts are much different. Many clauses do not have verbal forms and determining where clauses break is problematic. Moreover, there are often sudden shifts in verbal forms and in person, gender, and number. Furthermore, many ideas and concepts are repeated, making it difficult to see what might be the main ideas and what would be subordinate. On account of all these things, determining mainlines in poetry is often quite difficult. One might even choose, as many do, not to use the same syntactical concepts for narrative as for poetry – in this situation, the discourse should be seen in terms of parallelism and not in terms of a mainline.
After trying to sort out the mainline in the text of Ezekiel and coming across the apparent impossibility of doing that in some chapters, I came up with the hypothesis that, in some texts, there was more than one independent mainline. And those mainlines weave through each other at times. This allows one to take the syntax in the text seriously (and how different verbal forms and different participants interact), while also taking seriously the repetitions and parallels in the text. At the moment this hypothesis seems the best means I know of for providing a systematic and objective means of reading the discourse in the text.
Hopefully I’ll be posting some examples of how this works soon.