Semantic knowledge helps listeners interpret degraded speech signals

Evidence

Our results show that processing natural speech comes with a timely integration of semantic information, which in turn reduces lexical competition. Listeners are then able to pre-select a displayed target based on its semantic coherence with the context, and this allows listeners to reduce the effort involved in lexical competition. When processing degraded speech the integration of semantic information is delayed, as is also lexical competition. This implies that semantic integration is not able to reduce lexical competition, which by itself is longer and occurs later. — Dijk, Pim van, Physiology, Psychoacoustics and…, 3953 (#complicates, via linked claim)

Categorization

• XREF: connects to: Signal degradation delays semantic integration and raises listening effort — Dijk, Pim van, Physiology, Psychoacoustics and…, 3951; Linguistic context lets listeners understand speech at lower acoustic intelligibility; Degraded speech delays semantic integration and blocks contextual relief of effort — Dijk, Pim van, Physiology, Psychoacoustics and…, 3953 (+8 more)

Broader Connections

• Linguistic context lets listeners understand speech at lower acoustic intelligibility #corroborates (0.95)
• Semantic-context benefits to word recognition vanish under signal degradation #contradicts (0.90)
• Degraded speech delays semantic integration and blocks contextual relief of effort #complicates (0.85) — Dijk, Pim van, Physiology, Psychoacoustics and…, 3953
• Priming from previously heard clear speech can boost intelligibility of degraded speech #expands (0.80)
• Degraded speech processing raises baseline cognitive effort independently of semantic context #complicates (0.80) — Dijk, Pim van, Physiology, Psychoacoustics and…, 3978
• Speech models built on optimal listening neglect constant signal degradation #complicates (0.75)
• Beyond audibility, spectral shape of loss drives which speech cues fail #qualifies (0.70)
• Speech processing deficits can impair understanding independent of hearing #qualifies (0.70)