Post by Shireen Parimoo
What's the science?
Memories guide our eye movements and how we explore visual information. For example, people look at altered regions of familiar scenes more than they look at these same regions in familiar but unaltered scenes – a phenomenon known as the manipulation effect. This effect is observed only when people can also identify the change, suggesting that the eye-movement behavior might reflect awareness of the change, rather than an automatic and unconscious response. Regions of the medial temporal lobe (MTL) – particularly the hippocampus – play a key role in creating and retrieving recent declarative memories, which are memories about facts and events that we can consciously recollect. However, the role of the MTL in the manipulation effect as well as the relationship between the MTL and knowledge about the manipulation is not understood. This week in PNAS, Drs. Smith and Squire investigated the role of the MTL and declarative memory in the manipulation effect by comparing healthy participants and patients with MTL lesions.
How did they do it?
Participants included four patients with hippocampal lesions, one patient with broader MTL damage, and six healthy adults who served as controls. They viewed a series of scene images across three blocks while their eye movements were recorded. The same scenes were presented in the first two blocks and participants simply viewed the scenes. In the third block, half of the scenes were altered, and half were unaltered (repeated). After viewing each scene, participants were asked if the scenes were the same as before or if they had changed. They were then shown the altered scenes and asked to describe how those scenes had changed and where the change had occurred (i.e. the critical region). The authors divided the participants into three groups based on their knowledge of the altered scenes: those who correctly answered all three questions (whether the scene was altered and what and where the changes were) had robust knowledge (awareness) of the change, those who answered some of the questions correctly had partial knowledge about the change, and those who didn’t answer any question correctly were unaware. Eye movement data were analyzed to examine how often participants looked at the critical region of the altered scenes in the third block and how much time they spent viewing this region.
What did they find?
Compared to control participants, patients were unable to discriminate between the repeated and the altered scenes. Moreover, control participants had robust awareness of the changes in 60% of the scenes and were unaware of changes in 18% of the scenes. The reverse pattern was observed in patients, who showed awareness for only 11% of the scenes and were unaware of changes in 56% of the scenes. Thus, patients with MTL lesions had poorer declarative memory than the healthy controls.
Overall, the manipulation effect was observed only in control participants, who viewed the critical region of the scenes more when it was altered than unaltered. However, when participants had robust awareness of the changes in altered scenes, both patients and controls showed the manipulation effect. In other words, when they could identify that a scene had changed, what the change was, and where the change occurred, both controls and patients viewed the critical region more in the altered scenes than in the unaltered scenes. When participants had only partial awareness or were unaware of the changes, they spent a similar amount of time viewing the altered and repeated scenes, indicating that the manipulation effect is linked to conscious awareness for what has been learned.
What's the impact?
This study found that viewing behavior is related to the conscious recollection of memories, and that the manipulation effect could be driven by regions in the MTL, such as the hippocampus. Although previous studies have also linked viewing behavior to different memory-related eye movements, these findings help us better understand the role of awareness in declarative memory retrieval. Further research is needed to determine the conditions under which declarative or non-declarative memory supports different types of viewing behavior.
Smith & Squire. Awareness of what is learned as a characteristic of hippocampus-dependent memory. Proceedings of the National Academy of Sciences of the United States of America (2018). Access the original scientific publication here.