Multisensory perception as an associative learning process.

Area of Science:

• Cognitive Psychology
• Neuroscience
• Philosophy of Mind

Background:

• Multimodal perception, where senses combine, is often assumed to involve automatic feature binding.
• This automatic binding is thought to link sensory inputs, like visual and auditory stimuli, into a unified experience.
• Previous research posits a default mechanism for integrating cross-modal features.

Purpose of the Study:

• To challenge the assumption of automatic feature binding in multimodal perception.
• To propose associative learning, specifically unitization, as the mechanism underlying multimodal feature integration.
• To explore the philosophical implications of this shift from binding to learning in perception.

Main Methods:

• The study presents new evidence questioning the automatic feature binding hypothesis.
• It introduces the concept of "unitization" as a learned process for creating multimodal perceptual units.
• The argument is supported by theoretical reasoning and implications for philosophical debates.

Main Results:

• Evidence suggests that multimodal feature integration is not an automatic process.
• Multimodal perceptions, such as a cymbal's jolt and clang, are proposed to be learned perceptual units.
• Unitization provides an alternative explanation for unified multimodal experiences.

Conclusions:

• The automatic feature binding mechanism for multimodal perception is contested.
• Associative learning and unitization offer a compelling alternative for explaining how sensory features are integrated.
• Understanding perception as a learned process has significant implications for philosophical questions about experience and consciousness.