15th Speech in Noise Workshop, 11-12 January 2024, Potsdam, Germany 15th Speech in Noise Workshop, 11-12 January 2024, Potsdam, Germany

P12Session 2 (Friday 12 January 2024, 09:00-11:30)
Dividing attention bilingually: The benefits and costs of spatial separation between talkers in two different known languages

Emily Rice, Sarah Knight, Angela de Bruin, Sven Mattys
University of York, UK

Energetic masking (EM) refers to spectrotemporal overlap between a to-be-attended target and a competing masker, disrupting the target at the auditory periphery. Spatial separation between target and masker reduces EM, increasing target intelligibility. This benefit is known as spatial release from EM (SREM). When listeners attend to both sounds rather than one, spatial separation reduces EM, but increases cognitive load, as listeners must shift attention between two spatial locations. Nevertheless, spatial separation still appears to improve performance, demonstrating that SREM’s benefits outweigh cognitive costs. However, existing research has focused on monolingual participants presented with one language. It is currently unclear how these effects operate when two different, but known, languages are spoken simultaneously, which is a regular occurrence for bilinguals. Furthermore, it remains an open question how the costs and benefits of spatial separation differentially affect the first language (L1) and second language (L2).

In the current studies, unbalanced Spanish-English bilinguals completed a selective listening (SL, N=98) and/or a divided listening (DL, N=80) task. Results are reported here for the 66 participants who completed both tasks. On each trial, participants heard one Spanish (L1) and one English (L2) sentence simultaneously, either collocated (diotic) or dichotic (one sentence in each ear). In the SL task, participants were told in advance which talker to transcribe. In the DL task, they were not told which talker to transcribe until after stimulus presentation, so had to listen to both to perform successfully. Working memory was measured using a visual Letter Number Sequencing task.

In both tasks, performance was better for the L1 than the L2 sentences and in the dichotic than the collocated condition. There was no interaction between language and spatial separation. Thus, SREM was beneficial for both languages, and neither language benefited from SREM significantly more than the other.

However, in the L2 condition, the SREM benefit was smaller in the DL than the SL task, potentially because the increased cognitive demand associated with attending simultaneously to two spatially separated talkers had a greater impact on transcription of the less proficient language. Furthermore, working memory scores were positively correlated with listening accuracy in the DL, but not SL task, supporting an increased need for cognitive control when processing two languages simultaneously.

Last modified 2023-11-29 15:13:43