The Bilingual Language Interaction Network for Comprehension of Speech

Area of Science:

• Cognitive Science
• Computational Linguistics
• Psycholinguistics

Background:

• Bilinguals co-activate both languages during speech comprehension, leading to cross-linguistic interactions.
• These interactions impact language system structure and spoken language processing mechanisms.

Purpose of the Study:

• To examine cross-linguistic interaction effects in bilingual speech comprehension using computational modeling.
• To present the Bilingual Language Interaction Network for Comprehension of Speech (BLINCS) model.

Main Methods:

• Developed a connectionist model (BLINCS) using dynamic, self-organizing maps for interconnected processing levels.
• Simulated bilingual language processing in a controlled computational environment.

Main Results:

• BLINCS accounts for cross-linguistic interaction across multiple processing levels.
• The model explains cognate facilitation and audio-visual integration in speech comprehension.
• BLINCS demonstrates language separation without a global language-identification system.

Conclusions:

• BLINCS offers a novel computational approach to understanding bilingual speech comprehension.
• The model provides insights into the mechanisms of cross-linguistic interaction in bilinguals.