Neural and Cognitive Mechanisms Underlying Human Categorization

Categorization is one of the primary mechanisms underlying human perception and cognition, but how humans categorize complex multimodal sensory inputs from the real-world is still an open question. In this dissertation, I present three studies that examine how the human brain processes and represents categorical information of natural scenes, and how such categorical information shapes cognition, influencing what humans can learn from sensory inputs. Chapter 2 presents a study that shows scene information from either visual or auditory cues is represented in the prefrontal cortex independently of sensory modality, suggesting that conceptual aspects of scene category information are represented in this brain area. Chapter 3 describes a study that examined how scene representations in the prefrontal cortex are modulated by task context. This study demonstrated that scene category information is represented in the prefrontal cortex regardless of task context, unlike the global properties of scenes which were only represented in the prefrontal cortex when behaviorally relevant. These findings suggest that category information is automatically, without deliberate effort, processed and represented in the brain. In Chapter 4, we demonstrated that category information that observers extract from sensory inputs broadens the scope of learning, allowing humans to show the learning effect to novel stimuli based on category-level information. Taken together, these findings demonstrate that category information is automatically processed and represented across different areas in the brain, both sensory-specific and modality-independent brain regions, and interacts with such as learning and memory.Ph.D