Emotional Valence Precedes Semantic Maturation of Words: A Longitudinal Computational Study of Early Verbal Emotional Anchoring

We present a longitudinal computational study on the connection between emotional and amodal word representations from developmental perspective. In this study, children's adult were generated using latent semantic analysis (LSA) vector space model Word Maturity methodology. Some used to set mapping function with neural network ratings 9-year-old children. The was trained validated in child space. Then, resulting tested an data set. Samples of 1210 5315 words spaces, respectively. Results suggested that valence can be predicted even at stage, accurate propagation found representations. way, different propagative processes observed These findings highlight potential mechanism for early verbal anchoring. Moreover, multiple linear regression mixed-effect models revealed moderation effects performance model. First, maturation subsequent definition promoted propagation. Second, interaction effect age acquisition abstractness explain performance. theoretical methodological implications are discussed.