Smashing verb learning through parental sound symbolic input in preterm and full-term children

Abstract

ABSTRACTSound symbolism, the iconic link between speech sounds and meanings, helps children’s verb learning. In sound symbolically rich languages such as Turkish, hearing sound symbolic words might facilitate early verb learning and later language-specific expressions of motion events, by providing an easier way to map verbs onto events. These links could be much stronger for children who had difficulties in word-referent mapping (e.g., preterm children). The current study examined the association of Turkish-speaking parents’ sound symbolic input at 20 months with children’s concurrent verb knowledge and later motion event descriptions at 48 months. Ninety-one parent-infant dyads (Mage = 19.92 months, SD = 1.38; 41 preterms) were observed at 20 months, where we measured sound symbolic input and verb knowledge. Sixty-four of these dyads participated in a motion event description task at 48 months (25 preterms, Mage = 48.79 months, SD = 1.68), where we coded how children use path (the trajectory of motion) and manner (how an action is performed) of a motion. Results from the robust regression models suggested that for preterm children, parents’ sound symbolic input used in adverb forms was positively and concurrently related to children’s verb knowledge. Only for full-term children, sound symbolic input (used as adverbs) predicted full-term children’s use of both path and manner information in their descriptions at 48 months. Lack of replication of these findings with outlier-removed analyses could only suggest a trend toward the differential contributions of sound symbolic input in different child populations and the importance of sentential cues in sound symbolic input in verb learning. AcknowledgmentsWe thank everyone in the Language and Cognition Lab and Language and Communication Development Lab at Koç University for their valuable feedback for this research. We appreciate the efforts of Işıl Doğan, Aslı Aktan-Erciyes, and Salih Özdemir for this project. We especially thank Işıl Doğan, Nurgül Arslan, Nurdem Okur, and Süeda Vardar for data collection, Cansu Kılıç for data coding, and the families who participated in our study. Special thanks to Metin Sabancı Healthcare Center for Family Counseling for the institutional support, Prof. İpek Akman for her continuous help in the project, Dr. Nihan Hande Akçakaya, Banu Bingöl, and Rahime Gökboğa from Metin Sabancı Healthcare Center to their support for the project, El Bebek Gül Bebek Foundation for Premature Birth and Gymboree Classes for helping us reach out the families.Disclosure statementThe authors report there are no competing interests to declare.Data availability statementThe data used in this study can be accessed from https://osf.io/74ftz/?view_only=2a2bf3c235b04c1da5dace95e1b3d616Notes1 The sound symbolic input data of 34 FT children were also used in Kızıldere et al. (Citation2022).2 Parental education had three levels: not holding a college degree, holding a college degree or equivalent, and holding a graduate degree.3 The performance of children who participated in the study via computer did not differ from the performance of children who did not participate in the study via computer in the motion event description task, Path Any: t(62) = -1.426, p = .159, Manner Any: t(62) = 1.570, p = .122, Path + Manner: t(62) = .854, p = .397.4 Fifty-eight children were with their parents while six children were alone during the procedure alone. Participating alone or with a parent was not associated with children’s performance in the motion event description task: Path Any: t(62) = .088, p = .930, Manner Any: t(62) = .034, p = .973, Path + Manner: t(62) = -.049, p = .961.5 To indicate the specific importance of adverbial use of sound symbolism, we also tested the model with sound symbolic input used in verb-related roles (Verb, Auxiliary Verb, Adverb). Results indicated neither a significant main effect (B = 8.50, SE = 5.48, p = .125), nor an interaction with neonatal condition (B = -17.72, SE = 11.62, p = .131).6 The interaction between sound symbolic input used in verb-related roles (Verb, Auxiliary Verb, Adverb) and neonatal condition, was also a significant predictor of Path + Manner (B = .128, SE = .047, p = .009).7 Since maternal education differed across PT and FT children, we also ran all models reported in this paper including maternal education as a control variable. The results from the models with maternal education was not substantially different than the results of the reported models..Additional informationFundingThis work was supported by the James S. McDonnell Foundation Scholar Award (Grant no: https://doi.org/10.37717/220020510) to Tilbe Göksun.