Evidence acquisition time and result: the case of the Kildin Saami inferential Perfect

Ural-Altaic Studies. 2026. No. 2(61). P. 21–54.

Kosheleva M.

This paper is devoted to the Kildin Saami inferential evidential Perfect. I show that the inferential evidential Perfect requires the re sult of the inferred situation to exist at the evidence acquisition time (EAT) or at the utterance time (UT). The type of result required to exist depends on the verb’s actional interpretation: if the inferred situation is telic, it is the lexically determined result that is required to exist; if the inferred situation is atelic, it is some pragmatic result. The Kildin Saami data challenge the traditional view that lexical aspect cannot influence the behaviour of an evidential marker [Tatevosov 2019]. In line with [Fleck 2007], [Smirnova 2013], and [Lee 2013; 2023], the Kildin Saami data speak in favour of distinguishing an evi dential-specific temporal point — the evidence acquisition time. I argue that the Kildin Saami inferential evidential Perfect cannot be analysed in the style of [Arregui et al. 2017] as involving resultative aspect and conclude that the EAT should be incorporated into its semantics. I also show that the inferential evidential Perfect treats the evidence acquisition time and the inference time (IT) differently: if the result exists at the inference time (and not at the EAT or UT), the inferential evidential Perfect is infelicitous. Thus, the Kildin Saami data highlight the need to distinguish between EAT and IT when dealing with evidentials. I argue that for the category of evidentiality the EAT should be added to the neo-Reichenbachian system of temporal points [Klein 1994], though not all evidentials should necessarily use it.

Keywords: perfect, перфект, эвиденциальность, actionality, акциональность, кильдинский саамский язык, inferential evidentiality, Kildin Saami, evidence acquisition time, время получения свидетельств