The distinction between foregrounded and backgrounded clauses began with the observation that a narrative can be subdivided into clauses which narrate a sequence of events and those that do not, termed narrative and non-narrative clauses. Non-narrative clauses may present events that are out of sequence, such as flashbacks, or may not narrate events at all, such as descriptions or explanations. Since a narrative tends to be about a sequence of temporal events, it seemed natural that the narrative clauses would be the most salient, hence the term foreground. The non-narrative clauses were considered less salient, and hence background. However, to what extent does this distinction hold as we move away from the narrative genre?
Here is where the approaches of Longacre and Niccacci begin to diverge. To understand Niccacci’s approach, it is important to begin with the work of Harald Weinrich. Weinrich approached language from the perspective of text-linguistics and made a fundamental distinction between two registers of text which reflect the orientation of the author to the subject: Erzählen and Besprechen. The former is translated as narrative, and the latter is variously translated as comment, discussion, discourse, etc. I will use discourse here, since that is what Niccacci seems to prefer, but note that it differs from Longacre’s definition of discourse which is more similar to Weinrich’s “text”, ie the largest unit for analysis. Also, Weinrich is interested in texts so Besprechen doesn’t refer to actual spoken language, but rather to when an author makes use of more conversational language. Narrative is impersonal and tends to be related in the third person and past tense, while discourse is more intimate, bringing the author and reader into the situation by using first and second person along with present and future tense.
Within each of these groups, Weinrich identifies two further axes which motivate the choice of verbal form. Perspective is something like relative tense, and depends on whether the event is contemporary (called neutral or null degree), anterior, or posterior to the reference frame. For narrative the temporal reference is past tense, so the simple past is the null degree form, while in discourse the normal null degree form is the present. Lastly, Weinrich describes relief, which is the use of tense forms to distinguish foreground from background. The parade example is French where the passé simple is the narrative form appearing in foregrounded clauses while the imparfait appears in background clauses.
As far as I can tell, Weinrich only discusses the use of specific verbal forms for expressing relief within narrative. This is because narrative is the special case. As seen in the French example and in Biblical Hebrew, among others, it is not uncommon for languages to develop special forms for narrative. Schneider, through whom Weinrich’s ideas impacted the study of Biblical Hebrew, specifically states that in discourse foreground and background are not expressed by the use of verb tenses, but by other means:
Vordergrund und Hintergrund der Rede werden – anders als in Erzählung – nicht durch die Tempora – sondern durch andere Zeichen (Satzstellung, Partikeln, Hinweise auf die Sprechsituation) bezeichnet (Grammatik §188.8.131.52, 188).
Niccacci, however, extended the idea to non-narrative texts. Of course, since there are many more tense forms available in discourse, the system becomes much more complicated. For the present tense, the normal clause type is the simple nominal clause. This type is used for both foreground and background, which must be distinguished by other means.
In the past tense, the system is the same as in narrative, except that the initial verb is an (x)-qatal form, and then the following foregrounded clauses use wayyiqtol. Background is again expressed by x-qatal, non-verbal sentences, x-yiqtol, and wəqatal.
For the future, Niccacci distinguishes between indicative and volitive moods. A future indicative text begins with an x-yiqtol (Niccacci argues that all clause initial yiqtols are volitive) and the foreground verbs then switch to wəqatal. Background information is signified by the switch wəqatal > waw-x-yiqtol which is analogous to the shift wayyiqtol > waw-x-qatal in narrative. The future volitive begins with a volitive form (cohortative, jussive, imperative). Niccacci argues that the following foregrounded verbs then switch to wəyiqtol if the volitional mood is to be continued, but to wəqatal if the mood switches to indicative future, ie as a succession of events that will naturally follow.
Longacre has also extended the correlation of verbal forms with grounding beyond narrative, but with slightly different parameters. He has not followed the distinction of narrative and discourse, but instead suggests two basic parameters: contingent temporal succession and agent-orientation. For our purposes, the more important is contingent temporal succession, which is basically the existence of a chronological backbone to the text. Texts without such a backbone are organized logically or thematically. Thus a prophetic text is similar to a narrative, only with a future orientation. Instructional and procedural texts describe how something usually is or should be done, and also follow a sequence of steps. In all three, Longacre argues that wəqatal is the primary tense while x-yiqtol is used for secondary themes.
If we synthesize these two views, you will notice that there are three basic forms used for foreground. In narrative it is the wayyiqtol, while in future/modal contexts wəqatal and wəyiqtol are the foreground forms. The main secondary forms are x-qatal and x-yiqtol respectively. Note that the foregrounded forms are all clause-initial, while backgrounded forms are not. Again, I think that this reflects iconicity. In a sequential context we expect the events to be given in order of occurrence, and in NW Semitic it is the clause-initial position that is iconic for sequence. However, when the text does not have a sequential backbone, the distribution of the verbal forms no longer corresponds strongly to the distinction of foreground from background. Instead, as Schneider stated, other means are used.
As we move to poetry then, my working hypothesis is that the correlation of verbal forms to grounding will only be useful to the extent that the poem reflects contingent temporal sequence.