Variation in Gizey third-person subjects and the emergence of antilogophoricity

Abstract This paper is concerned with how grammatical functions emerge in natural languages. Specifically, it provides insights into the emergence of antilogophoricity, viz., the special coding of participants other than those whose speech or thoughts are being reported. The study is based on Gizey (Masa < Chadic). In Gizey, nonzero third-person subjects can be expressed with the Set A pronouns nàm SG.M , nà ʔ SG.F and sí PL , or the Set B pronouns mì SG.M , tì SG.F and àsí PL . In reported speech, the Set A pronouns are used logophorically. They mark coreference with the entity whose speech or thoughts are reported. The Set B pronouns, for their part, track non-perspectival participants. Outside reported speech, the Set A and Set B pronouns can be used interchangeably, but the Set B pronouns occur less frequently than their Set A counterparts, and they are disallowed in the topic and focus positions. In this paper, I surmise that Set A and Set B pronouns are the remnants of an erstwhile functionally determined variation, not originally related to logophoricity or antilogophoricity. I argue that the Set B pronouns show signs of specialization as antilogophors. This paper presents the first documented observation of pronouns potentially evolving into antilogophoric pronouns.