ИСТИНА

I discuss the derivation of impersonal sentences in Old Slavic Languages and defend the so called 'Autonomy Hypothesis', an early version of the non-configurationality hypothesis. Old Russian lacked the distinction of impersonal vs personal verbs in the lexicon and probably licensed the derivation of impersonals from verbs of all groups, including modals and verbs of perceptions. Standard Modern Russian bans or restricts the derivation of transitive impersonals with features [+ Experiential], [+ Animate], [+ Causative], while substandard and archaic varieties of Russian arguably license many impersonal derivative with [+ Causative] and [+ Animate] features, though impersonal derivates from true experiential verbs are not attested. Disclaimer. This presentation includes some polemics against the wide-spread view that grammar, especially the syntax of predicate-argument relations, directly represents the referential situation and the mental state of the speaker. I argue that subject-predicate analysis and the procedure for identifying ellipsis (coordinate deletion and other recoverable phenomena) vs impersonal sentences (non-recoverable elimination of the grammatical subject) must be carried out irrespective of the beliefs of the speakers that Odinn or Zeus were responsible for the rain, fire or flood. The idea that ancient Indo-Europeans could fairly well identify the animate subject behind the scenes in sentences like O.RUSS. МОЖЕТ ИДТИ ''<One> can go/travel' or РОЗДРЕТЬ МЯТЕЛЬ "<if somebody> tears the cloake apiece" is sound, but does not substantiate the claim that in such sentences one can recover the link to a deleted subject.