Talker-specificity and token-specificity in recognition memory

Area of Science:

• Cognitive Psychology
• Speech Perception
• Auditory Neuroscience

Background:

• Recognition memory experiments often use identical speech tokens, creating a confound between talker and stimulus identity.
• Talker-specificity effects, where listeners recognize words better from a familiar talker, are foundational in speech perception research.
• The extent to which listeners encode high-level acoustic characteristics of a talker's voice remains an open question.

Purpose of the Study:

• To investigate the distinct roles of token repetition and talker repetition in continuous recognition memory.
• To determine if talker-specificity effects persist with novel, non-identical speech tokens.
• To explore how variations in talker characteristics influence episodic encoding and memory.

Main Methods:

• Two continuous recognition memory experiments were conducted.
• Experiment 1 used a single talker with either identical or novel repeated tokens.
• Experiment 2 involved two demographically matched talkers, presenting same-voice repetitions as either identical or novel tokens.

Main Results:

• Classic talker-specificity effects were observed for both identical and novel tokens.
• Recognition memory for identical tokens was sometimes stronger than for novel tokens.
• Recognition memory varied across talkers, indicating differential episodic encoding.

Conclusions:

• Novel speech tokens should be the default in recognition memory studies to avoid confounds.
• Listener memory is sensitive to talker-specific acoustic information beyond exact stimulus repetition.
• Accounting for talker variation is crucial for advancing the understanding of speech encoding and memory processes.