The Repeated Name Penalty and the Overt Pronoun Penalty in Japanese

This research investigated the Repeated Name Penalty (RNP) and the Overt Pronoun Penalty (OPP) in Japanese. The RNP was first reported by Gordon, Grosz and Gilliom (1993), who observed that English sentences with repeated name subject anaphors were read slower than sentences with overt pronoun subjects when the antecedents were either the grammatical subject or the first-mentioned surface-initial noun phrase of the previous sentence. The OPP has been reported in studies of Spanish (Gelormini-Lezama & Almor, 2011) in which sentences with overt pronoun subject anaphors were read slower than sentences with null-pronoun subject anaphors for subject antecedents. A concern with the RNP and OPP is that, in most previously tested languages, the grammatical subject of a sentence is typically also its discourse topic. Thus, it remains unclear whether these effects are attributable to the anaphors’ subject-hood or to their topic-hood. To address this question, the present study examined the RNP and the OPP in Japanese, a language that distinctly utilizes the topic-subject and non-topic-subject. In addition, while the existing studies controlled saliences of antecedents by grammatical status (i.e., subject vs. object), the present study utilized the information-structurally different antecedents, namely topic vs. nontopic (both are grammatical subjects), and the non-linguistic aspects, empathy-locus vs. nonempathy-locus (both are grammatical subjects). The present study conducted six self-paced sentence-by-sentence reading experiments with native Japanese speakers. The results showed that when antecedents are subjects or objects, the RNP and the OPP were detected only when anaphors were the non-topic with ga, not with topic anaphors with wa. The outcomes indicate that the topic-hood of anaphors activated immediate anaphoric interpretation. In addition, when the antecedents were the topic, the processing of anaphors showed the marginal RNP, which is attributable to the topicalizing function of the topic-postposition wa. Finally, although the different empathy statuses of antecedents elicited neither the RNP nor OPP, a closer observation of the results finds the RNPlike trend where the antecedents were the empathy-locus. Overall, these results indicate that the topic-hood of anaphors should be considered in studies of the RNP and OPP and that the saliences of antecedents may be dependent on nongrammatical aspects such as discourse topic or notion of empathy in addition to grammatical status.