Content area
Full text
Abstract
The Latin perfect is characterised by a variety of synchronically unpredictable verb stem formations. From an Indo-European point of view, these formations derive from distinct diachronic sources, and there is no agreement in the literature as to which factors determine their selection by individual verbs. In this paper, a new hypothesis is presented for perfect stem selection in Latin, focusing on a kernel of verbs which have so far proven most recalcitrant to systematic explanation. Specifically, this paper presents evidence that congruence between the stem vocalism of the present and perfect stems was a significant factor in determining whether an inherited reduplicating perfect was retained or replaced by a productive sigmatic perfect. Based on this hypothesis, the paper lays the basis for a chronologically layered (pre)history of the Latin perfect system.
Keywords
Latin - Italic - aorist - perfect
1 Perfect stem selection: a long-standing challenge in Latin historical linguistics
As any beginning student of the Latin language swiftly discovers, only a subset of Latin verbs allow their perfect stems to be straightforwardly predicted from their present counterparts. Thus, while 'weak' verbs use a regular perfect with a productive v-morpheme (such as amo ~ amāvī), many of the most common verbs in the language have elusive 'strong' forms (rego ~ rēxī do ~ dedi) which cannot be inferred and must simply be memorised (Gildersleeve and Lodge 1895, 93-94). Diachronically, this phenomenon has a straightforward explanation: the Latin perfect is derived from the syncretism of the Proto-Indo-European aorist and perfect (Rix 2002, 14; Zuin 2020, 23).1 As these two ancestral aspect stems fused into a single stem, they left a mosaic of different exponents, and Latin verbs do not universally agree as to whether they ought to be using perfect or aorist morphology.
In Classical Latin we can broadly identify four such 'strong' perfect formations, depending on how these stems are characterised relative to the present. Although the evolution of these formations is obscured by analogy and sound change, their relationship with Indo-European precursor categories can be summarised as follows:
- Reduplicated perfects such as cano, cecinī are derived straightforwardly from Proto-Indo-European perfects (Weiss 2020, 435), although the possibility of a few holdovers from the reduplicated aorist cannot be excluded (Meiser 2003, 89).
-...