Feature parsing: feature cue mapping in spoken word recognition

Area of Science:

• Psycholinguistics
• Auditory Perception
• Phonetics

Background:

• Word recognition requires mapping phonetic cues to phonological representations.
• Speech assimilation alters phonetic features, posing challenges for listeners.
• Understanding how listeners process assimilated speech is crucial for auditory perception models.

Purpose of the Study:

• To investigate the information listeners extract from assimilated speech segments.
• To determine how listeners interpret these altered phonetic cues.
• To explore the role of auditory grouping mechanisms in processing assimilated speech.

Main Methods:

• Three experiments were conducted using form priming and phoneme monitoring tasks.
• Experiment 1 used form priming to assess activation of underlying word forms.
• Experiments 2 and 3 examined the influence of assimilated segments on postassimilation context perception.

Main Results:

• Listeners activate the underlying form of the assimilated word (e.g., CONE), not its altered neighbor (e.g., COMB).
• Assimilated segments facilitate the perception of subsequent speech sounds.
• Context effects are independent of familiarity with specific assimilation processes.

Conclusions:

• Listeners recover both the underlying form of a modified word and information about the following context from assimilated segments.
• These findings support the role of general auditory grouping mechanisms in speech perception.
• The study provides insights into the dynamic processing of phonetic information during word recognition.