Grouping by Proximity and Grouping by Similarity in Audition

By using visual dot lattices as a tool, Kubovy and van den Berg (2008) studied two classic Gestalt grouping principles --- grouping by proximity and grouping by similarity --- in vision and found surprising additive effects between the two principles.

This dissertation is aimed at building an auditory analogy to explore the effects of these two grouping principles in audition. We used auditory necklaces --- ambiguous auditory patterns that we have been developing (Yu & Kubovy, submitted) --- as a tool and studied grouping by temporal proximity, grouping by loudness similarity, and grouping by pitch similarity in audition. In Experiment 1, we examined the effect of grouping by temporal proximity alone. In Experiment 2, we examined the separate and conjoint effects of grouping by temporal proximity and grouping by loudness similarity. In Experiment 3, we examined the separate and conjoint effects of grouping by temporal proximity and grouping by pitch similarity.

The results showed that as individual grouping principles, grouping by proximity and grouping by similarity perform as lawfully in audition as they do in vision. We can predict the probability that a note is perceived as the starting point of a circular auditory pattern by using the strength of grouping by temporal proximity, grouping by loudness similarity or grouping by pitch similarity. When grouping by temporal proximity and grouping by loudness similarity were conjointly applied to the same stimulus, their effects were additive, as was found in vision. However, when grouping by temporal proximity and grouping by pitch similarity were conjointly applied to the same stimulus, people only relied on pitch similarity for grouping auditory necklaces and ignored temporal proximity.