Mind and Brain Lectures

ERA Chair Inaugural Lecture

Cognitive Neuropsychology: A Window into the Functioning of the Mind/Brain

Speaker: Era Chair Holder Alfonso Caramazza | Date: 21.04.2023

A major challenge in understanding the functioning of the human mind/brain, aside from its immense complexity, is the relative inaccessibility of the representations that are computed in the course of a cognitive operation, such as recognizing the objects in an event or scene, understanding a sentence, reaching to grasp an object, or deciding which route to take for a walk. Even something as “simple” as reading a word poses major challenges. What are the intermediate representations between the visual shapes of the letters and the recognition of the word and its meaning? We don’t have direct access to those representations and must infer their properties through indirect means such as, for example, measuring how quickly we recognize individual letters versus scrambled letters versus pseudowords versus words. The problem, simply put, is that since we don’t have direct access to intermediate levels of representation all we can measure and analyze is an input-output relationship: for example, the relationship between a visual stimulus and a response such as reading a word or naming a picture. But Nature occasionally creates opportunities for observing more directly the computational role of distinct neural components within a cognitive system through their selective damage in clinical situations. In such cases, we can observe the computational consequences of damage to a specific component of the overall architecture of a cognitive system and infer the nature of intermediate representations within that system. Here I discuss the value and limitations of this methodology – Cognitive Neuropsychology – through several illustrative cases of the insights we have gained about otherwise inaccessible cognitive operations and representations. I will make the case for its indispensability in the armamentarium of cognitive science and neuroscience.

How Children Learn

Speaker: Elizabeth S. Spelke, Harvard University | Date: 19.04.2024

Children may be the most prodigious learners on earth: With little to no instruction, they master the commonsense concepts and skills that their culture requires, and then go on, in school, to master highly demanding symbolic skills and systems of knowledge beyond both intuition and perception. How do they do this? Research on human infants, children, adults, and non-human animals, using diverse methods from the cognitive, brain, and computational sciences, provides evidence for seven early emerging, domain-specific cognitive systems: six systems of core knowledge that are shared with other animals and serve to represent places, objects, animate beings, social beings, number and geometry, and a seventh system that likely is unique to humans and underlies early learning of one or more natural languages. These automatic, unconscious, and fixed systems provide inputs to the malleable systems underlying our conscious perceptions and thoughts. All seven systems are functional at birth, and throughout life, they support the growth of our endlessly inventive human minds.

UNDERSTANDING THE NEURAL CODE FOR CONSCIOUS SYMBOLIC THOUGHT: A CHALLENGE FOR HUMAN COGNITIVE NEUROSCIENCE

Speaker: Stanislas Dehaene, Collége de France | Date: 20.06.2025

According to the global neural workspace hypothesis, the mechanisms of conscious access are similar in human and non-human species. Wherein, then, lies the singularity of the human brain? In this talk, I will propose that the contents of consciousness became markedly richer in humans as our brains acquired a capacity for compositional thought using discrete symbols. Recent comparative data from my lab show that humans possess unique abilities for symbolic learning and a mathematical “language of thought”. Even the mere perception of a square or a zig-zag involves a short mental program that captures the observed data in an internal language of geometry. Behavioral and brain-imaging experiments indicate that the perception of geometric shapes is poorly captured by current convolutional neural network models of the ventral visual pathway, but involves a symbolic geometrical description within the dorsal parieto-prefrontal network. I will argue that existing connectionist models do not suffice to account for even elementary human perceptual data, and that neural codes for symbols and syntax remain to be discovered.

References:

Sablé-Meyer, M., Ellis, K., Tenenbaum, J., & Dehaene, S. (2022). A language of thought for the mental representation of geometric shapes. Cognitive Psychology, 139, 101527. https://doi.org/10.1016/j.cogpsych.2022.101527

Sablé-Meyer, M., Fagot, J., Caparos, S., Kerkoerle, T. van, Amalric, M., & Dehaene, S. (2021). Sensitivity to geometric shape regularity in humans and baboons : A putative signature of human singularity. Proceedings of the National Academy of Sciences, 118(16). https://doi.org/10.1073/pnas.2023123118

Van Kerkoerle, T., Pape, L., Ekramnia, M., Feng, X., Tasserie, J., Dupont, M., Li, X., Jarraya, B., Vanduffel, W., Dehaene, S., & Dehaene-Lambertz, G. (2024). Brain areas for reversible symbolic reference, a potential singularity of the human brain. eLife, 12. https://doi.org/10.7554/eLife.87380.2