Research project Towards a rational theory of resource utilization in human visual working memory

The two most recent decades have seen enormous progress in our understanding of the limitations and structure of visual working memory. However, current theories still suffer from two fundamental shortcomings. First, they are all built on the assumption that memory capacity is fixed, which is contradicted by recent empirical findings showing that capacity is task-dependent. Second, current theories are purely descriptive: they are aimed at providing accurate quantitative descriptions of working memory limitations, but offer no deeper insight into the factors that shape these limitations or the question why capacity is limited in the first place.

This project addresses these shortcomings by using a rational-analysis approach to the study of capacity limitations. The key idea is that since storage of memory is costly to the brain, a rational memory system would adapt the amount of cognitive resources it spends on storage to the demands of the task, similar to how the human body adapts its heartrate and energy expenditure to the demands of a physical task.

Behavioral studies will be conducted to obtain better insight into the factors that affect memory capacity, such as task difficulty, reward, and the memory fidelity and duration required by the task. The empirical results will be used as a bases for the second part of the project, which aims at improving current mathematical models of visual working memory.

The results of this project are expected to produce novel insights into the flexibility of visual working memory and may lay the foundations for a general and rational theory of how it utilizes its resources.