CCN Colloquium: "Picture That! The Mismeasure of Mental Images"

Mental imagery - the ability to visualize in the absence of external stimulation - plays important roles in our lives: from spatial navigation and mental rotation to reliving past events and simulating future scenarios. The subjective nature of mental imagery, however, poses serious challenges for scientists looking to understand and precisely measure this inner phenomenon. In this talk, I will share recent work from our laboratory showing different ways in which we can improve the measurement of imagery vividness, for example, by relying on physiological measures such as pupillometry. By leveraging behavioral tools from vision science, we can also better characterize visual properties of mental images, such as sharpness or saturation, and show how these relate to vividness ratings. Finally, large language models (LLMs) and their ability to complete classic mental imagery tasks despite relying on non-visual processing offer an opportunity to reflect on the difficulty of creating purely visual mental imagery tasks. Together, these studies help us develop a new, more robust scientific approach to studying our mind's eye.