When come and go go necessive

Deictic motion verbs as passive auxiliaries: the case of italian andare ‘go’ (and venire ‘come’).

“…As argued by Bourdin (: 128), however, both the abilitative and the deontic interpretation presuppose the existence of a modally neutral passive construction, which is indeed attested in Apabhraṃśa (cf. ex.…”

“…Moreover, such ambiguous cases are too sporadic in the corpus to be reasonably assumed to constitute the onset stage of the grammaticalization process at issue. Finally, as Bourdin (: 129–130) also notes, Giacalone Ramat's () explanation is too ‘local’ (i.e. too closely tied to the specific diachrony of Italian) and does not take into account that there are other passives having motion verbs (including ‘go’) as auxiliaries, as discussed in Section 1.…”

“…In Italian, indeed, the modal (deontic) meaning emerges through conventionalization of pragmatic implicature from the habitual/generic overtones that characterize the earliest occurrences of the passive with andare ; as such, it can be considered as a development that takes place after the passive construction has emerged out of a semi‐copular construction. Moreover, this implicature ( if something is generally done in a certain fashion, then it must be done that way ) is at work in a number of voice constructions across languages, and, as Bourdin (: 128) correctly observes, is not tied to any particular marking of passive voice (e.g. the auxiliary): in some Indo‐Aryan languages, for instance, deontic modality ‘can be supported by the synthetic passive as well as by the periphrastic passive involving the itive auxiliary’ (Bourdin : 128), and, as widely documented in the typological study by Narrog (: 110), ‘there is a diachronic path from voice constructions, specifically passive and spontaneous constructions, to event‐oriented [i.e.…”

“…Moreover, this implicature ( if something is generally done in a certain fashion, then it must be done that way ) is at work in a number of voice constructions across languages, and, as Bourdin (: 128) correctly observes, is not tied to any particular marking of passive voice (e.g. the auxiliary): in some Indo‐Aryan languages, for instance, deontic modality ‘can be supported by the synthetic passive as well as by the periphrastic passive involving the itive auxiliary’ (Bourdin : 128), and, as widely documented in the typological study by Narrog (: 110), ‘there is a diachronic path from voice constructions, specifically passive and spontaneous constructions, to event‐oriented [i.e. potential and necessive/deontic, AS‐AGR ] modal constructions’, in which irrealis (e.g.…”

“…generic or negated) contexts may have played the role of ‘bridging contexts’ (Narrog : 112) . Voice constructions, and in particular passive constructions, lend themselves particularly well to a modal (and specifically deontic/necessive) reinterpretation, since ‘one of the situations where backgrounding or suppression of the Agent makes most sense is when its referent is diffuse and this is certainly a common feature of statements divorced from the speaker's hic and nunc such as general prescriptions as well as descriptions of permanent properties or “dispositions”’ (Bourdin : 128). The border between generic/habitual and modal (deontic) interpretations of passive constructions is a vague one, and the fact that there are languages, such as Hindi, in which the deontic reading of the passive construction is denied by some authors (e.g.…”