Acquisition of object clitics in child Polish: Evidence for three developmental stages

• The study tests the theories of object clitics acquisition in child Polish.
• Both clitic production and clitic comprehension data are collected and compared.
• Results show early clitic omission and a strong age effect.
• Clitic production is found highly unlikely without evidence of clitic comprehension.

Cross linguistic variation in L1 clitic acquisition is limited and well-governed, and has been attributed to an underlying syntactic mechanism, such as the Unique Checking Constraint (UCC) in connection with clitic-past participle agreement (Wexler et al., 2003), or a pragmatic constraint, such as Failed Referentiality (Schaeffer, 2000). The present study seeks to validate the claims following from the above theories by looking at the clitic acquisition facts in child Polish in two experiments: clitic production and clitic comprehension. The paper argues that claims following from the two acquisition theories are not supported by Polish L1 data due to an initially high clitic omission rate (60%) and the evidence of early clitic comprehension which precedes clitic production. By comparing clitic production and clitic comprehension results, three developmental stages are identified. A maturational account is adopted attributing non-adult-like structures in child grammar to a discourse-linking mechanism (Borer and Rohrbacher, 2002).