On the Origin of the Epic Preterit

Abstract The use of past tense in narrative discourse at first glance seems to imply that the narrated events are lying in the past as compared to the act of narration. However, this intuitive notion was doubted by several scholars in the mid 20th century, among them Käte Hamburger, Harald Weinrich, Émile Benveniste, and Ann Banfield. This article investigates the temporal constitution of literary narratives from a perspective of the biological evolution of Human cognition. My analysis begins with Hamburger’s most disputed claim that in epic fiction the past tense »loses its grammatical function of designating what is past« and proceeds by testing a derived hypothesis against a cross-cultural sample of (mostly oral) folklore. Hamburger denied any temporal relation between the speaker and that which he speaks of, assuming instead a fictitious neverland in which the narrated events are situated and to which the preterit refers. If Hamburger’s model is correct, so the derived hypothesis goes, then the use of past tense in fictional narratives is merely a cultural convention, and different narrative traditions should expose different conventions. Indeed it can be shown that in other cultures stories are told in the present tense, in infinitive verb forms, or in forms indicating abstractness or remoteness. It can be followed that Hamburger was right at least in presuming that reference to the past is not a necessary constituent of verbal storytelling. Actually, instead of referring to the past, the epic preterit rather seems to indicate a change in the modality of speaking, thus adhering to the category of grammatical mood rather than tense. In some languages of oral cultures, however, this presumed mood shows up instead as a way of indicating the source of information. This kind of source information – fully grammaticalized in a quarter of the world’s languages – is called ›evidentiality‹ by linguists. In a phylogenetic perspective on the evolution of cognition, source information only becomes necessary with extended inferential and communicative capabilities and may thus have emerged as a cognitive tool in early humans when entering the ›cognitive niche‹. Evidentiality markers in language may thus be the linguistic reflex of a very ancient cognitive scope category in the innate architecture of the human mind, one which served to separate first-hand experience from reported knowledge. In oral storytelling, evidentiality is marked not only by specific verb forms but also by specific formulas (›they say‹/›it is said‹), intonations, or rhetorical devices. From this perspective, the phenomenon observed by Hamburger and others can be said to originate in the beginning of Tradition – that is, of verbal transmission of cultural knowledge. My hypothesis is that literary narratives in literate cultures still use this ancient cognitive scope operator of ›tradition‹ when employing the epic preterit. Admittedly, in literate cultures it often suffices to put »A novel by« on the title page in order to signal the categorical otherness of narrative fiction. Yet still, authors employ additional means to evoke the atmosphere of a ›murmuring conjuring‹ – as Thomas Mann once called it – that creates the impression of an objective world of tradition behind the individual story told. I point toward examples in literary first-person narratives, because homodiegetic narration – in contrast to Hamburger’s classical case of heterodiegetic narration – shows a continuous spatio-temporal relation between speaker and that which he speaks of and thus requires additional means or efforts to signal a break between the ordinary world of first-hand experience and the world of the literary. Since Hamburger once treated the epic preterit as a signal of fictionality, I briefly discuss the notion of fiction in the last paragraphs of my paper. I consider ›fictionality‹ to be a late cultural concept in literate societies that is not identical with the cognitive category of ›tradition‹ but is ultimately made possible by the existence of the latter.