Re-reconsidering the role of temporal order in spoken word recognition

Area of Science:

• Psycholinguistics
• Cognitive Science
• Speech Processing

Background:

• Spoken word recognition models traditionally assume linear processing of word sounds.
• Recent research questions the strict linearity constraint in lexical access.
• The visual world paradigm has been used to study online spoken word recognition.

Purpose of the Study:

• To replicate the anadrome effect in spoken word recognition.
• To test the coarse grain similarity account against sequential match models.
• To investigate the role of spectral similarity in mapping spoken input to lexical representations.

Main Methods:

• Utilized the visual world paradigm to track listener eye movements.
• Compared fixation patterns for target words with anadrome competitors and phonologically unrelated distractors.
• Introduced a novel condition with competitors lacking vowel position overlap.

Main Results:

• Replicated the anadrome effect: listeners fixated anadrome competitors (e.g., "bus" for "sub") more than distractors.
• Fixation curves for anadrome competitors diverged significantly around 275 ms post-target onset.
• Competitors without vowel overlap did not elicit increased fixations compared to distractors.

Conclusions:

• Findings support the anadrome effect, challenging strict sequential processing models.
• Results suggest coarse grain spectral similarity plays a crucial role in spoken word recognition.
• The study highlights the importance of stimulus design in visual world paradigm research.