On the origin of 2nd person prefix #tV- in Trans-Himalayan languages

Two 2nd person verbal indexation forms are reconstructed back to Proto-Trans-Himalayan (PTH): a suffix #-n(a) (Bauman 1975; DeLancey 1989, 2014; H. Sun 1983, 1995; van Driem 1993; Watters 2002), and a prefix #tV- (Watters 2002; Jacques 2012; DeLancey 2011a, 2014). While #-n(a) is in paradigmatic distribution with other suffixal forms, the prefixal position of #tV-raises the question of its functional origin. DeLancey (2011a, 2014: 23) suggests hypothetically that #tV- finds its origin either in a non-finite nominalization or in an irrealis nominalization, “ideal for an impersonal 2nd person use: ‘One might [speak]’ rather than ‘You will [speak].’” Through the comparison of cognate tV- possessive and tV- nominalizer prefixes found in rGyalrongic, Kuki-Naga (Ao), and Sinitic, I propose that 2nd person #tV- can be traced back to a nominal possessive modifier such as one’s or someone’s: a man-type of R-impersonals (Malchukov & Siewierska 2011). Spreading to verbs, #tV- was used as a nominalizer indexing the notional S or A argument. With nouns and verbs alike, the referent of the possessor, whether 1st, 2nd, or 3rd person, was specific (Siewierska 2011: 62), accessible in discourse, anaphorically or deictically, regardless of the presence of a coreferential overt noun-phrase. The prefix #tV- started to be used in particular to address 2nd person indirectly, a development likely triggered by pragmatic motivations, i.e. politeness, before being reanalyzed as a 2nd person indexation marker.