Abstracts

Abstract:  It is well known that over a dozen areas in the early layers of the human visual cortex exhibit spatial organization that mirrors the layout of the retina. It is much less clear what this means at the level of representation. Though such regions look like literal pictures in the brain, philosophers and neuroscientists alike have often been skeptical of such appearances. In this talk I will argue that that, in fact, retinotopic areas not only produce iconic representations, but representations that function very much like pictures. Drawing on parallels between ventral stream processing and neural network models of computer vision, I will suggest that early visual computations exploit metric correspondences between functional space in the brain and space in the world, in a manner characteristic of iconicity.

Abstract: This paper develops a theory of iconic representation. We begin by examining prior proposals that appeal to the parts principle, according to which iconic representations are those where parts of the vehicle represent parts of the content. Then we develop our positive theory, according to which iconic representations are “locatively structured” collections of analog representations. We also examine how the resulting theory identifies the relationship between the iconic and the analog, precisifies the notions of a functional space and functional part, and generates an interesting taxonomy of representational systems.

Abstract: How are concepts mentally represented?  I will approach an answer via a case study of three different ways animals represent the abstract relations same and different. I begin by sketching evidence that human infants, and a wide range of animals, including insects, have representations of these relations.  I then sketch evidence for evolutionary and ontogenetic discontinuities in mental representations with these contents.  Three different formats of representation of same and different can make sense of these data: (1) Implicit representations (in which the content is carried by computations); (2) Iconic representations; (3) Discursive representations.  I present evidence that animals and babies have only the first two.  The capacity for symbolic concepts for these relations may depend, both in evolution in ontogenesis, on the human capacity for natural language.

Abstract: We often look to externalized language as a source of evidence, or at least hypotheses, for the structure of symbolic representations in the mind, such as predicate-argument structure, the interaction of logical operators, quantification, modal operators, etc. I'll focus in this talk on the way that symbolic and iconic representations interact, and critically also the ways they fail to interact, in externalized language, drawing evidence from both signed and spoken languages. Based on these observations I'll argue for a critical information structural role that discrete symbolic representations provide that iconic analog representations cannot, with an eye toward the question of what this means for such representations outside the context of information exchange/externalized language.

Abstract: In ‘subitizing’ tasks, infants accurately discriminate small collections, up to a set-size of ~3, after which performance falls to chance. It remains unclear, however, why performance consistently falls to chance under these conditions given that infants possess an equally well-attested capacity to approximately enumerate larger collections. I call this The Biggie Smalls Problem. The present paper clarifies The Problem, notes that it is exacerbated by influential ways of thinking about infant numerical cognition and argues that existing ‘solutions’ to The Problem are unsatisfactory. It then develops an improved solution, which turns on independently motivated claims about the format of the representations involved and the signature limits of infant working memory. Beyond generating testable predictions, this improved solution has ramifications for the architecture of numerical cognition, the structure of perceptual representations, and the ways in which perceptual states refer.