What's the gist? The influence of schemas on the neural correlates underlying true and false memories

Abstract

The current study used a novel scene paradigm to investigate the role of encoding schemas on memory. Specifically, the study examined the influence of a strong encoding schema on retrieval of both schematic and non-schematic information, as well as false memories for information associated with the schema. Additionally, the separate roles of recollection and familiarity in both veridical and false memory retrieval were examined. The study identified several novel results. First, while many common neural regions mediated both schematic and non-schematic retrieval success, schematic recollection exhibited greater activation in visual cortex and hippocampus, regions commonly shown to mediate detailed retrieval. More effortful cognitive control regions in the prefrontal and parietal cortices, on the other hand, supported non-schematic recollection, while lateral temporal cortices supported familiarity-based retrieval of non-schematic items. Second, both true and false recollection, as well as familiarity, were mediated by activity in left middle temporal gyrus, a region associated with semantic processing and retrieval of schematic gist. Moreover, activity in this region was greater for both false recollection and false familiarity, suggesting a greater reliance on lateral temporal cortices for retrieval of illusory memories, irrespective of memory strength. Consistent with previous false memory studies, visual cortex showed increased activity for true compared to false recollection, suggesting that visual cortices are critical for distinguishing between previously viewed targets and related lures at retrieval. Additionally, the absence of common visual activity between true and false retrieval suggests that, unlike previous studies utilizing visual stimuli, when false memories are predicated on schematic gist and not perceptual overlap, there is little reliance on visual processes during false memory retrieval. Finally, the medial temporal lobe exhibited an interesting dissociation, showing greater activity for true compared to false recollection, as well as for false compared to true familiarity. These results provided an indication as to how different types of items are retrieved when studied within a highly schematic context. Results both replicate and extend previous true and false memory findings, supporting the Fuzzy Trace Theory.

Interplay of the long axis of the hippocampus and ventromedial prefrontal cortex in schema‐related memory retrieval

“…The aHPC has been suggested to be more involved in global or gist‐like processing (e.g., Poppenk et al, ; Robin & Moscovitch, ), and our results provided supports for the role of the aHPC in schema‐related memory retrieval. Consistent with our results, Webb et al () investigated subsequent memory effects for objects that were encoded either in congruent or incongruent scenes and found increased aHPC activation for schema‐related recollection. Human fMRI studies have also suggested that activation in the aHPC is associated with constructing mental representations of scenes (Dalton, Zeidman, McCormick, & Maguire, ; Zeidman & Maguire, ).…”

“…The aHPC has been suggested to be more involved in global or gist-like processing (e.g., Poppenk et al, 2013;Robin & Moscovitch, 2017), and our results provided supports for the role of the aHPC in schema-related memory retrieval. Consistent with our results, Webb et al (2016) investigated subsequent memory effects for objects that were encoded either in congruent or incongruent scenes and found increased aHPC activation for schema-related recollection.…”

“…For example, when new streets were entered during navigation of a simulational city, the aHPC was more associated with a global transition structure across the street network and the pHPC was more associated with local path options (Javadi et al, 2017). Therefore, based on the functional distinction between the aHPC and pHPC along with previous findings (e.g., Bonasia et al, 2018;Webb et al, 2016), it is reasonable to hypothesize that the aHPC and pHPC play different roles in retrieving schema-related memory, with the anterior part activated more for schema-related and the posterior part more for schema-unrelated retrieval.…”

“…Although rodent studies have suggested a critical role of the vmPFC in schema-related memory (Tse et al, 2011;Wang et al, 2012), in human studies, its activation increased due to schema in some studies (Brod et al, 2015;Sommer, 2016;van Kesteren, Rijpkema, et al, 2010;Wagner et al, 2015), but did not change in other studies (e.g., Bonasia et al, 2018;van Buuren et al, 2014;Webb et al, 2016). The vmPFC is involved in evaluating and monitoring memory (Gilboa & Moscovitch, 2017), guiding memory retrieval by using relevant contexts to resolve conflicting information (Preston & Eichenbaum, 2013), and replacing the role of the hippocampus as memory ages (Nieuwenhuis & Takashima, 2011).…”

“…For the hippocampus, only a few human studies found its activation due to schema during retrieval but in opposite directions (e.g., increased in Webb, Turney, & Dennis, ; decreased in Bonasia et al, ). It is noteworthy that the increased hippocampus activation was located in the anterior portion (Talaraich space: y = −13) for schematic versus nonschematic scene recollection trials (Webb et al, ), and the decreased hippocampus activation was located in the posterior portion (Talaraich space: y = −40) for recalling congruent versus incongruent events (Bonasia et al, ; however, see Sommer, ). This pattern raises the possibility that the anterior and posterior hippocampus are differentially involved in schema‐related memory retrieval.…”