Electronic Thesis and Dissertation Repository


Doctor of Philosophy




Köhler, Stefan


We live in a rich visual world, surrounded by many different kinds of objects. While we may not often reflect on it, our ability to recognize what an object is, detect whether an object is familiar or novel, and bring to mind our general knowledge about an object, are all essential components of adaptive behavior. In this dissertation, I investigate the neural basis of object representations, focusing on medial temporal lobe (MTL) structures, namely, perirhinal cortex, parahippocampal cortex, and hippocampus. I use what type of thing an object is, or more specifically, the broader category (e.g., “face” or “house”) or domain (e.g., “animate or “inanimate”) to which an object belongs to probe MTL structures. In the Chapter 2, I used fMRI to explore whether object representations in MTL structures were organized by animacy, and/or real-world size. I found domain-level organization in all three MTL structures, with a distinct pattern of domain organization in each structure. In Chapter 3, I examined whether recognition-memory signals for objects were organized by category and domain in the same MTL structures. I found no evidence of category or domain specificity in recognition memory-signals, but did reveal a distinction between novel and familiar object representations across all categories. Finally, in Chapter 4, I used a neuropsychological approach to discover a unique contribution of the hippocampus to object concepts. I found that an individual with developmental amnesia had normal intrinsic feature knowledge, but less extrinsic, or associative feature knowledge of concepts This decreased extrinsic feature knowledge led to abnormalities specific to non-living object concepts. These results show that the hippocampus may play an important role in the development of object concepts, potentially through the same relational binding mechanism that links objects and context in episodic memory. Taken together, these findings suggest that using object category or domain to probe the function of MTL structures is a useful approach for gaining a deeper understanding of the similarities and differences between MTL structures, and how they contribute more broadly to our perception and memory of the world.