On the Typology of the Null Subject Parameter: A Proposal on Turkish

Abstract

Since the introduction of Principles & Parameters Model (Chomsky, 1981), the null subject parameter (NSP) has drawn the attention of quite a number of generativists as one of the parameters which differentiates languages from each other. The NSP determines whether the subject position can be left empty or not. With respect to this, null-subject languages allow phonologically unrealised subjects in tensed clauses (e.g., Spanish, Italian, Portuguese, and Greek). Conversely, non-null subject languages do not allow null subjects (e.g., English, German, and French). In null-subject languages, the empty category was labelled as pro after Chomsky (1981) and the licensing and identification of pro has been extensively studied to figure out how null subjects are allowed. However, concerning the literature on the NSP, different typological accounts on how null subject and non-null subject languages realise their subjects have been given, which suggest that languages have further means to mark subjects either null or overt. In the light of these observations, the aim of the current study is to review these typological accounts to bring rationale to the discussions raised in the field. Drawing on the NSP typologies suggested by Tomioka (2003) and Roberts and Holmberg (2010), the study aims to discuss and suggest a proposal on the possible place of Turkish (a classical null subject language) in the NSP typology. By doing so, the study emphasizes the agreement-based and discourse-bound role of null vs. overt subject distribution in Turkish. © Peter Lang GmbH.