Implicit acquisition of grammars with crossed and nested non-adjacent dependencies: investigating the push-down stack...

Area of Science:

• Neuroscience
• Cognitive Science
• Linguistics

Background:

• Distinguishing human language from animal communication often involves grammar complexity (finite-state vs. phrase-structure).
• An alternative view suggests non-adjacent vs. adjacent dependencies are more relevant for neurobiological language studies.
• Processing non-adjacent dependencies necessitates online memory resources, potentially involving a push-down stack.

Purpose of the Study:

• To investigate the neurobiological mechanisms underlying the acquisition of non-adjacent dependencies in language.
• To test the hypothesis that a limited push-down stack is essential for processing non-adjacent dependencies.
• To compare the learning of non-adjacent and adjacent dependencies in an artificial grammar learning paradigm.

Main Methods:

• Subjects learned artificial grammars with non-adjacent dependencies implicitly.
• The study compared the acquisition of non-adjacent dependencies with adjacent dependencies.
• Processing predictions of the push-down stack model were tested using nested and crossed dependencies.

Main Results:

• No qualitative differences were observed between the acquisition of non-adjacent and adjacent dependencies.
• Acquiring non-adjacent dependencies required more exposure but utilized similar mechanisms as adjacent dependencies.
• The push-down stack model was partially supported, indicating stack-like properties are among other mechanisms for online memory in language processing.

Conclusions:

• The acquisition of non-adjacent dependencies relies on the same fundamental mechanisms as adjacent dependencies.
• While a push-down stack model offers partial explanation, language processing involves a broader set of neurophysiological mechanisms.
• Online memory resources for structured sequence processing, including language, exhibit stack-like properties within a larger neural framework.