Welcome to the CoDeNeuro lab!
We study how foundational cognitive abilities interact with developmental experiences to give rise to uniquely human knowledge. Currently, we are using behavioral, neural, and computational approaches to study mechanisms underlying the acquisition and emergence of our ability to conceptualize numbers, among various other cognitive science topics. Our research has been supported by the National Science Foundation.
So, why do we study number concepts?
What do you know about natural numbers (1, 2, 3, …)? We understand that each number represents a particular quantity of a set composed of discrete elements. We understand that a number has a particular relationship with another number. We also understand that numbers never end. All these are quite straightforward, making the concept of number likely the simplest abstract concept. However, we argue that it is actually one of the most difficult concepts to learn. As a matter of fact, it takes nearly a decade of one’s early life time to fully acquire these number concepts. Furthermore, without proper cultural and linguistic resources as well as education, a child may never learn the basic number concepts that we take for granted. At the same time, even with extensive training, no other species on this planet is capable of acquiring this seemingly simplest idea. Thus, number represents a core aspect of uniquely human cognitive capacity, and it can tell us about how foundational cognitive abilities interact with developmental experiences to let us humans be us, providing a unique opportunity to gain scientific insights into humanity. This is why we study numerical cognition.
What are some examples of what we study?
Using both EEG and fMRI, we identified neural representations of numerosity in early visual cortex (Fornaciai & Park, 2018).
We developed a new task that evaluates children’s understanding of the base-10 syntactic structure of number words and demonstrated that children’s acquisition of numerical syntax is crucial for their generative number concepts (Guerrero et al., 2020).
We developed a computational model that explains how a simple feedforward neural network implementing divisive normalization naturally becomes sensitive to numerosity by removing the effects of size and spacing, the dimensions that are orthogonal to numerosity (Park & Huber, 2022).
Check out our Publications for more exciting research on numerical cognition, concept development, vision science, and more!