Usage-based theories of language acquisition emphasise the social nature of the phenomena of language. Children learn to speak listening to their caregivers. Therefore, statistical learning is one of the core abilities for language acquisition. Children are sensitive to distributional characteristics of the language they hear. Frequent syntactic frames are recurring constructions present in mothers’ speech. Stoll et al. (2009) analysed child directed speech (CDS) of English, German and Russian mothers and found that frequent syntactic frames were present in 70 up to 86% of mothers’ utterances. In the current study, their presence in Polish CDS is studied. Polish is a language with relatively free word-order in comparison to the languages previously analysed. This difference can influence the repetitiveness present in mothers’ speech. What is more, previous studies that searched for frequent frames in mothers’ speech, were based on manual analysis of relatively sparse language corpora (inter-recordings periods up to three weeks). The present study analysed a dense child directed speech corpus using an automatic procedure. Two Polish-speaking mothers were recorded while taking care of their daughters (1:08,25 – 3:0,21 years old), each mother during two 3 – 5.5 week periods. Mothers recorded themselves, in their homes for about an hour almost every day (inter-recording periods were not longer than couple of days). Data collection resulted in about 60 recordings per mother, at average 381 – 434 analysable utterances per recording. The results of the frequent frames presence analysis show that the observed overall percentage of the utterances including frequent frames in Polish is lower than that in the languages studied before. In Polish CDS between 50 and 60% of mothers’ utterances included constructions recognized as frequent syntactic frames. The reasons for that may stem from differences in the grammars of Polish and languages studied before, as well as in the methodology of searching for the frequent frames (for example the density of the corpora analysed).
Stoll, S., Abbot-Smith, K., & Lieven, E. (2009). Lexically restricted utterances in Russian, German, and English child-directed speech. Cognitive Science, 33(1), 75–103.