Acquisition of relative clauses in Turkish

Abstract

This study analyzes the acquisition of relative clauses in Turkish. Turkish is a language with a dual system of relative clause formation. It differentiates between subject (SR) and nonsubject (NSR) relatives. SR makes use of the nominalizing particle -(y)An and NSR the particle –DIĞ which are selected on the basis of the nature of the relationship of the head noun to the relative clause in Turkish. The choice of correct morphology cannot be an easy task for Turkish-speaking children as they would be challenged by the dual system of relative clause formation in the language. In the present study, experimental data is used. Using a picture-cued elicitation technique, 48 monolingual Turkish-speaking children in three developmental groups (age range: 3;3-8;2) are tested. The experiment is designed to elicit relative clauses in a relativization site of various syntactic positions: subject; direct/indirect/oblique objects. The semantic (reversibility) and syntactic (transitivity) properties of sentences with RCs as well as the position of the head noun in the main sentence are also diversified to see the role of these variables in children’s production of relative clauses. The findings have revealed a big asymmetry in children’s performance of SRCs and NSRCs. Children, regardless of their age groups, performed better on SRCs than NSRCs. The interesting finding of the present study, however, has been that children used subject relativization strategy to avoid nonsubject relativization and this use constituted half of all the nonsubject relative constructions used by each child, regardless of age. A further pattern that emerged has to do with the massive use of resumptives (RP) by children, the use of which is ungrammatical in adult Turkish. The diverse results of the experimental data found in this study have been explained in an account that considers –(y)An as an unmarked relative clause participle in Turkish-speaking children’s early grammar. –(y)An strategy is claimed to be less costly since it requires the least computation both morphologically (no subject-verb agreement) and syntactically (no A-movement). I further claimed that the resumptive pronouns and NPs encountered in the child data can be considered as a device that Turkish speaking children resort to so as to disambiguate nonsubject relative clauses from subject relative clauses. In sum, it has been shown that resumptive use is triggered by children’s use of –(y)An participle as an “All-purpose Relativizer” in Turkish.