Current Research
Our research focuses broadly on the organization of human memory. To approach the general question of how memory is implemented in the brain, our research attempts to delineate the component processes of memory and how different brain regions subserve these component processes. This approach uses cognitive neuroscience techniques to associate activity in particular brain regions with particular cognitive processes involved in the performance of various memory tasks. Our research emphasizes a multimodal neuroimaging approach, using functional MRI (fMRI) to obtain information about anatomical localization and electroencephalography (EEG) to obtain information about the relative timing of memory processes. We are also interested in attempts to integrate spatial and temporal information by integrating across neuroimaging modalities, such as through the use of anatomically-constrained EEG source analysis.
Current Research Projects
- Neural Basis of Recognition Memory
Recognition memory is the process of discriminating between stimuli that are novel and those that have been encountered previously. Item recognition is thought to underlie the “feeling of familiarity” that we get when re-encountering such items, such as when we see a familiar face. Single-cell recording in rats and monkeys has indicated that neurons in cortical areas adjacent to the hippocampus signal that an item is familiar via a reduction in firing rate for familiar relative to novel items. In one experiment (Gonsalves, Kahn et al., 2005), fMRI and MEG signals from MTLC were correlated with parametric levels of perceived familiarity in a face recognition test. Results showed decreases in neural responses as a function of increasing familiarity, consistent the idea that these brain regions signal the familiarity of an item through a reduction in activity. Furthermore, MEG results showed that these MTLC familiarity signals occurred early, as soon as 200 ms after the presentation of the faces, consistent with theoretical assertions that familiarity is assessed relatively rapidly. Ongoing studies in the lab are attempting to replicate and further characterize this familiarity-related response suppression, including how signals from this region interact with other medial temporal lobe regions, especially the hippocampus, how this familiarity signal may be “read out” into other cortical regions, such as frontal lobe regions mediating control processes important for the assessment of familiarity, and whether the assessment of item familiarity reflected in these signals reflects a relatively automatic process or one that is affected by task demands.
- False Memory
Memory does not constitute a literal reproduction of past events. Rather, memory for the past is a constructive process, one that can be prone to systematic errors. We are interested in studying the brain mechanisms for various sorts of systematic memory errors, with the goal of better understanding how memory works in general by studying how it goes wrong.- Reality Monitoring errors
One mechanism for false memories is a confusion between things that we actually perceived and things that we only imagined, often called errors in "reality monitoring". We are studying the cognitive and brain mechanisms of such errors using ERPs and fMRI by inducing people to confuse memories for objects that they imagined for memories of objects that they actually saw. - Misinformation paradigm
Another kind of false memory can occur when people are given misleading information about a previously experienced event, and unintentionally incorporate this misleading post-event information into their memories. We are using fMRI to study the brain mechanisms whereby such misleading information gets incorporated into the original event memory.
- Reality Monitoring errors
- Memory for Political Candidates'
Issue Positions
An interesting phenomenon in the political science literature is that people with a large amount of political knowledge, so-called "political sophisticates", make systematic memory errors when remembering the issue positions of fictitious political candidates to which they have been exposed in an experimental setting. In particular, these people will tend to remember issue positions that are consistent with a candidate's ideology, but that the candidate never explicitly stated. We are using ERPs to gain a better understanding of how these errors arise.
- Interactions of Emotion and Memory