Eye fixations during encoding of familiar and unfamiliar language

This study examines gaze patterns of monolinguals and bilinguals encoding speech in familiar and unfamiliar languages. In condition 1 English monolinguals viewed videos in familiar and unfamiliar languages (English and Spanish or Icelandic). They performed a task to ensure encoding: on each trial, two videos of short sentences were presented, followed by an audio-only recording of one of those sentences. Participants choose whether the audio-clip matched the first or second video. Participants gazed significantly longer at speaker's mouths when viewing unfamiliar languages. In condition 2 Spanish-English bilingual's viewed English and Spanish, no difference was found between the languages. In condition 3 the task was removed, English monolinguals viewed 20 English and 20 Icelandic videos, no difference in the gaze patterns was found, suggesting this phenomenon relies on encoding. Results indicate people encoding unfamiliar speech attend to the mouth presumably to extract more accurate audiovisually invariant and highly salient speech information.