The Developmental Trajectory of Nonadjacent Dependency Learning

Abstract

We investigated the developmental trajectory of nonadjacent dependency learning in an artificial language. Infants were exposed to 1 of 2 artificial languages with utterances of the form [aXc or bXd] (Grammar 1) or [aXd or bXc] (Grammar 2). In both languages, the grammaticality of an utterance depended on the relation between the 1 st and 3rd elements, whereas the intervening element varied freely. High variability of the middle element is known to contribute to perception of nonadjacent dependencies (Góomez, 2002), but the developmental trajectory of such learning is unknown. Experiment 1 replicated the study of Gómez with a younger age group and a more subtle variability manipulation. Twelve‐month‐olds failed to track nonadjacent dependencies under conditions tested here (Experiments 2a and 2b), but by 15 months, infants are beginning to track this structure (Experiment 3). Such learning has implications for understanding how infants might begin to acquire similar structure in natural language.

The Role of Prior Experience in Language Acquisition

“…Based on these findings, Santelmann and Jusczyk (1998) suggested that increases in infants' working memory capacity support an enhanced ability to detect nonadjacent structure. Gomez & Maye (2005) found further evidence of developmental differences in the ability to learn nonadjacent relationships. They found that 15-and 17-month-old infants were able to track nonadjacent relationships in an artificial language, while 12-montholds were unable to do so.…”

“…the sentence "The boy is running" is grammatical, but "The boy can running" is not) if they are separated by no more than 3 morphemes (Santelmann & Jusczyk, 1998). Gomez and Maye (2005) found that 15-month-olds, but not 12-month-olds, were able to learn similar nonadjacent relationships when exposed to simpler language materials.…”

“…Twelve-month-old infants can learn co-occurrence relationships between word categories in an artificial language (Gomez & Lakusta, 2004), while infants this age seem unable to track nonadjacent structure (Gomez & Maye, 2005). The current experiment tested whether 12-month-olds given prior experience with adjacent dependencies can recognize those dependencies when they are nonadjacent.…”

“…This experience could enhance their sensitivity to nonadjacent relationships between such dependent elements, perhaps by attuning them to the presence of predictive relationships. Because Gomez and Maye (2005) found that the ability to track nonadjacent relationships does not emerge until 15 months of age, even under the most conducive circumstances, we tested whether 12-month-old infants could learn them if given relevant prior experience.…”

Do as I say, not as I do: A lexical distributional account of English locative verb class acquisition

“…Other verb bias alternations such as the noun phrase/sentence complement alternation (e.g., the man wrote [that] the story [was interesting]; Garnsey, Pearlmutter, Myers & Lotocky, 1997) also involve variability in the post-verbal position (e.g., optional that), suggesting that such variation may play a role in verb class acquisition more generally. Humans also find it difficult to learn non-adjacent dependencies in artificial grammar learning tasks and variation can modulate this ability (Gómez, 2002;Gómez & Maye, 2005;Newport & Aslin, 2004). This is particularly relevant given that SRN models provide some of the best accounts of learning in these tasks (Chang, Janciauskas & Fitz, 2012).…”