Abstract:This study examined how the morphological typology of second language (L2) learners' first language (L1) affected neural processing of L2 morphosyntactic knowledge. We used functional near infrared spectroscopy (fNIRS) to examine adult L2 learners' processing of English Subject-Verb number agreement (e.g., duck swims, ducks swim) during a picture-sentence matching task. Two English learner groups with contrasting L1s, Spanish (with rich inflectional morphology, n = 16) and Mandarin (with a lexical morphology, n = 18), were compared to native English speakers (n = 19). Both L2-learner groups demonstrated comparable accuracy on the picture-sentence matching task. However, neural results revealed L1 influence on L2 morphosyntactic processing. Mandarin-speaking English L2 learners showed greater neural activity in the left middle temporal gyrus (L-MTG) for singular (e.g., the duck swims) versus plural sentences (e.g., the ducks swim). Mandarin relies on semantic, rather than inflectional, information to infer number and L-MTG is involved in lexical-semantic processing, suggesting L1 influence on L2 inflectional processing. Spanish-speaking English L2 learners showed greater neural activity in areas including the right MTG and prefrontal cortex for the plural versus singular sentences whereas native English speakers showed greater activity for singular versus plural sentences. The plural form is morphologically marked in Spanish and greater neural activation for the plural rather than singular form suggests L1 influence. Importantly, cross-linguistic influences were only observed at the neural level, revealing that different neural activation patterns underpin similar behavioral results. Both L2-learner groups showed different patterns of neural activation corresponding to the specific linguistic features of their L1, indicating that L2 processing is affected by L1 characteristics in linguistically principled ways. This study advances our understanding of how morphosyntactically-distinct languages are organized and processed in adult L2 learners.

Acquisition of Inflectional Morphology in Artificial Neural Networks With Prior Knowledge

Linguistically inspired morphological inflection with a sequence to sequence model

A Comprehensive Comparison of Neural Networks as Cognitive Models of Inflection

Transfer in a Connectionist Model of the Acquisition of Morphology

Unsupervised Inflection Generation Using Neural Language Modeling

What Artificial Neural Networks Can Tell Us About Human Language Acquisition

Morphological Inflection with Phonological Features

Pretraining with Artificial Language: Studying Transferable Knowledge in Language Models

Modeling Target-Side Morphology in Neural Machine Translation: A Comparison of Strategies

Cross-linguistic influences of L1 on L2 morphosyntactic processing: An fNIRS study

Neural Dynamics of Processing Inflectional Morphology: An fMRI Study on Korean Inflected Verbs

Category Clustering and Morphological Learning

Testing learning hypotheses using neural networks by manipulating learning data

Analyzing Linguistic Knowledge in Sequential Model of Sentence.

Are Human Learners Capable of Learning Arbitrary Language Structures

Neural Transductive Learning and Beyond: Morphological Generation in the Minimal-Resource Setting

The Grammar-Learning Trajectories of Neural Language Models

An Investigation of Noise in Morphological Inflection

What does the Knowledge Neuron Thesis Have to do with Knowledge?

Neural Basis of the Implicit Learning of Complex Artificial Grammar with Nonadjacent Dependencies

Linguistic Knowledge Can Enhance Encoder-Decoder Models (If You Let It)