Read more...In this paper I took the semasiological approach departing from the form and not from a particular meaning. Drawing on Thomas (1957) and Seržant (2016) I claim that the PP construction codes the meaning of perfect, very frequently the resultative perfect in Tocharian. The reference time of the PP construction is the same as the moment of speech and not prior to it. The latter is coded by the preterit (and imperfect).
This perfect of Tocharian is not a young category. This conclusion follows from a number of aspects pertaining to both meaning and morphosyntax. First, the resultative meaning is by far not the only one albeit – most probably the original one. In many instances, the PP construction denotes meanings typical of a perfect such as the explanatory meaning, rarely the experiential meaning (§5.2). It thus shows progress on the developmental cline in (1) above. Moreover, the PP often functions as an anteriority converb which is equally unlikely if the PP would have had only the resultative meaning elsewhere because anteriority presupposes that the dynamic meaning component, namely, the preceding-action entailment must have become more foregrounded. Secondly, the combinability of the PP construction with cross-cutting categories, for example, by means of the co-occurrence with the past-tense or subjunctive auxiliary is an indication for advanced grammaticalization.29 Thirdly, there are no selectional input restrictions onto which lexical verbs may occur in the PP construction: all Vendler classes are found in this construction – a situation that may not be found with early resultatives.
At the same time, the perfect meaning of the PP construction cannot be too old as semantic extensions typically found with perfects in other languages are not found in Tocharian. For example, the evidential meaning is not found in Tocharian while the experiential meaning may be suggested only for some rare examples. I have not come across the universal-perfect meaning. Moreover, I have argued above that the imperfect auxiliary does not yield a distinct category (pluperfect) that would not be analysable in terms of the composition of the resultative/perfect meaning with the meaning of the auxiliary.