Retrieval orientation for memories encoded in emotional contexts: An ERP study
Introduction
To reactivate the engram (i.e., records of previous experiences stored in the brain), the retrieval of past episodes requires ecphory—defined as the interaction between retrieval cues and memory traces (Tulving, 1983, Tulving, 1985). Retrieval success depends on how a retrieval cue is processed prior to and during the cue’s interaction with targeted memory traces. The processing of a retrieval cue can be varied in a pre-retrieval operation, referred to as retrieval orientation, that optimizes the compatibility between the cue and memory trace (Rugg & Wilding, 2000). Evidence for retrieval orientation has been furnished in studies that have manipulated various types of mnemonic variables, reflecting the flexibility of retrieval cue processing. However, few studies have examined the modulation of emotion—which has been demonstrated to considerably affect memory (for a review see Hamann, 2001)—on the processing of retrieval cues. Therefore, to fill this research gap, the present study examined how the emotionality of the encoding context affects subsequent retrieval orientation. Specifically, event-related potentials (ERPs) were analyzed in recognition memory tests, where ERPs recorded during the test phase for items that were encoded in emotionally valenced versus neutral contexts were compared.
The neural correlates of retrieval orientation can be revealed by comparing brain activities elicited by the same retrieval cues used to probe different types of memories (Rugg & Wilding, 2000). In a recognition test, corrected rejected new items are considered to serve as optimal retrieval cues for examining the retrieval-orientation effect; these items are optimal because the influence of the successful retrieval of different types of memory is minimal. Previous studies adopting this approach have reported that correct rejection (CR) trials were associated with different ERP patterns when new items were presented with studied items that were of different input modalities (Hornberger et al., 2004, Hornberger et al., 2006, Hornberger et al., 2006), material types (Robb and Rugg, 2002, Woodruff et al., 2006), and encoding tasks (Rugg et al., 2000, Johnson et al., 1997). The differential processing of retrieval cues has been reported to be associated with the enhancement of retrieval accuracy (e.g., Bridger and Mecklinger, 2012, Roberts et al., 2014) and to contribute to the strategic recollection of essential information (e.g., Herron & Rugg, 2003; Dzulkifli and Wilding, 2005).
Although retrieval orientation is useful in the study of memory retrieval, it remains unclear whether and how the processing of retrieval cues varies depending on the emotionality of memory. Emotional memories are more durable, vivid, detailed, and accurate than neutral memories are (Kensinger & Schacter, 2008). Many studies have examined encoding and consolidation processes that differentiate emotional memories from neutral memories (see Dolcos et al., 2017 for a review). However, relatively few studies have focused on the retrieval of emotional memory (Kensinger & Ford, 2020). An important concern in studying emotional memory retrieval is how emotion is manipulated. If emotion is manipulated by employing neutral and emotional materials as items to be remembered, the test items presented in subsequent memory tests will also comprise emotional values; this makes it difficult to discern the effect of emotion on memory encoding from that on retrieval (Maratos and Rugg, 2001, Smith et al., 2004). This problem can be prevented by presenting neutral items in an emotional or neutral encoding context during the study phase and without emotional contexts during the test phase (e.g., Erk et al., 2003, Erk et al., 2005, Maratos et al., 2001, Smith et al., 2004). Previous ERP studies adopting this approach have reported that memories for neutral items encoded in negative contexts were accompanied by greater recollection than items encoded in neutral contexts were (Maratos and Rugg, 2001, Smith et al., 2004, Jaeger et al., 2009 Jaeger and Rugg, 2012; Bowen & Kensinger, 2017). Functional MRI findings have also indicated the recapitulation of sensory and emotional processing during the recollection of memories encoded in emotional contexts (Fenker et al., 2005, Bowen and Kensinger, 2017). A recent ERP study further indicated that the effect of the valence of the encoding context was exerted during the very early stage of the old/new effects of ERPs, suggesting that recapitulation can occur in the early stage of the retrieval process (Bowen, Fields, & Kensinger, 2019). Although these studies have examined the recollection and recapitulation of emotional memories, they did not directly investigate retrieval orientation. Specifically, they could not do so because their findings could not differentiate between pre-retrieval cue processing and retrieval success.
The present study investigated whether different cognitive operations were engaged in the processing of retrieval cues when the encoding contexts of emotionally neutral retrieval targets had different valences. We generated emotional memories by presenting neutral object pictures in encoding contexts (specifically, scenes) that were neutral, positive, or negative in different blocks. The Positive Affect Negative Affect Scale (PANAS, Watson, Clark, & Tellegen, 1988) was administered after each of the three encoding sessions to confirm that the valenced pictures triggered the target emotions during encoding. In the subsequent recognition tasks, only pictures of neutral objects were presented. Typically, the modulation of the encoding context valence on retrieval-orientation is examined by comparing ERPs associated with CR trials in three test phases. However, the ERPs associated with CR trials could also comprise brain activities related to the retrieval mode, a tonic state maintained during retrieval to ensure the stimuli are processed as retrieval cues (Rugg and Wilding, 2000, Tulving, 1983). Moreover, viewing a series of valenced stimuli has been demonstrated to affect subsequent cognitive operations (e.g., Bradley et al., 1996, Smith et al., 2005, Sutton et al., 1997). The ERP differences between CR trials in three test phases could also be due to brain activities associated with processes irrelevant to memory retrieval orientation, such as retrieval mode or emotional states after viewing a series of positive or negative pictures. To exclude this potential confounding factor, we presented blank trials (see Methods section for details) during the recognition test to record brain activities associated with processes irrelevant to retrieval orientation. ERPs associated with the blank trials were subtracted from ERPs recorded in the CR trials of the same test phase to examine the retrieval orientation effect.
An advantage of our design was the ability to differentiate retrieval-orientation effects specific to valence from those related to arousal. A study argued that emotional experience can be evaluated using the quality of subjective feelings (valence) and the intensity of physical arousal (Russell, 1980). Previous studies have shown that emotional valence and arousal processes can be dissociated at the neural level (Gerdes, 2011, Lewis et al., 2006) and may contribute differently to memory performance (Kensinger & Corkin, 2004). However, these studies have focused on the encoding rather than the retrieval stage, and no study has examined the valence and arousal effect on retrieval orientation. Thus, in our study, we matched the arousal levels of positive and negative context pictures (both have a higher arousal level than the neutral context pictures) to make them differ only in terms of the valence. Consequently, any retrieval-orientation effect exhibiting a difference between positive and negative encoding contexts can be considered an indicator of the influence of valence. However, any effect indicating differences between neutral and emotional contexts but not between positive and negative encoding contexts can be considered to indicate the effect of arousal.
Several studies have suggested the possible modulation of emotion on retrieval orientation with ERPs. Maratos & Rugg (2000) reported that the ERP old/new effects were smaller for negative words than for neutral words mainly because more positive-going waveforms were associated with the CRs to new negative items than to new neutral items. Bowen et al. (2019) similarly reported that the valence of the encoding context exerted an effect on the very early part of the ERP old/new effects over the parietal scalp region. In a study examining the effect of emotion on strategic retrieval, Herron (2017) reported that emotional context exerted an effect on an ERP positivity over the right frontal scalp region. Another ERP experiment that used emotionally neutral, pleasant, and unpleasant images (Lavoie & O’Conner, 2013) also demonstrated that in later time windows (1000–1400 ms) unpleasant new images elicited more positive-going waveforms relative to old images. Based on these findings, we speculated that the effect of an emotional encoding context manifests over the parietal and right frontal regions in the early and later parts of the recording epoch, respectively. In addition, memories for emotional materials have been suggested to be modulated by trait anxiety (e.g., Marzi et al., 2014, Nugent and Mineka, 1994). The Trait Anxiety Inventory (STAI-T, Spielberger, Sydeman, Owen, & Marsh, 1999) was therefore administered to investigate whether the modulation of emotion on retrieval orientation is affected by trait anxiety.
Section snippets
Participants
Thirty-five undergraduate and postgraduate students (aged between 18 and 24 years) from National Central University, Taiwan, participated in this experiment. All participants were right-handed native speakers of Mandarin Chinese with normal or corrected-to-normal vision. They were paid NT$500 for their participation. The data of 11 participants were excluded from analysis because they contributed insufficient (<16) valid ERP trials in one of the critical experimental conditions. Of the
Affective rating scores
For the PANAS administered after the study phases, the mean positive affect (PA) scores (with SD) were 21.7 (7.04), 22.6 (7.82), and 29 (7.73) in the cycles of negative, neutral, and positive encoding contexts, respectively. A one-way repeated-measures ANOVA on PA scores indicated the significance of the main effect of encoding context (F2,46 = 10.73, p < 0.001 = 0.32). A post-hoc analysis with Bonferroni correction indicated that the PA score following the positive encoding context was
Discussion
Our study aimed to examine how retrieval orientation can be modulated by the emotionality of encoding context. By pairing neutral items with emotional and neutral backgrounds, we manipulated the encoding context to have a positive, negative, or neutral valence. During the test phase, only neutral items were presented for recognition. We hypothesized that if retrieval orientation can be adjusted according to the emotionality of the encoding context, then the ERPs elicited by CR items in the test
Limitation and conclusion
In the present study, we employed blank trials to prevent any mixing of the retrieval orientation and retrieval mode with the emotion carry-over effect. However, a better connection between the retrieval mode and blank trial is required to justify the representativeness, which can be achieved by adding another session without a retrieval task (e.g., Herron and Wilding, 2004, Xia and Evans, 2020). Another concern is the use of the subtraction method wherein we subtracted ERPs related to blank
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This research was supported by grants from Ministry of Science and Technology, Taiwan to SKC (MoST 109-2410-H-008-023-MY2, 106-2420-H-008-003-MY3) and TLL (MoST 107-2628-H-008 -002 -MY3). We thank Dr. Tzu-Yu Hsu from Taipei Medical University for her comments on the manuscript.
References (64)
- et al.
Discrimination and response bias in memory: Effects of depression severity and psychomotor retardation
Psychiatry Research
(1997) - et al.
EEGLAB: An open source toolbox for analysis of single-trial EEG dynamics including independent component analysis
Journal of Neuroscience Methods
(2004) - et al.
Electrophysiological indices of strategic episodic retrieval processing
Neuropsychologia
(2005) - et al.
Emotional context modulates subsequent memory effect
NeuroImage
(2003) - et al.
Emotional context during encoding of neutral items modulates brain activation not only during encoding but also during recognition
NeuroImage
(2005) - et al.
Is mood congruency an effect of genuine memory or response bias?
Journal of Experimental Social Psychology
(2001) Cognitive and neural mechanisms of emotional memory
Trends in Cognitive Sciences
(2001)- et al.
The relationship between the right frontal old/new ERP effect and post-retrieval monitoring: Specific or non-specific?
Neuropsychologia
(2008) - et al.
Right dorsolateral prefrontal cortex is engaged during post-retrieval processing of both episodic and semantic information
Neuropsychologia
(2009) - et al.
An electrophysiological dissociation of retrieval mode and retrieval orientation
Neuroimage
(2004)
fMRI correlates of retrieval orientation
Neuropsychologia
ERP correlates of the incidental retrieval of emotional information: Effects of study-test delay
Brain Research
The late posterior negativity in ERP studies of episodic memory: Action monitoring and retrieval of attribute conjunctions
Biological Psychology
Variations in retrieval monitoring during action memory judgments: Evidence from event-related potentials (ERPs)
International Journal of Psychophysiology
Recognition memory for emotionally negative and neutral words: An ERP study
Neuropsychologia
Reconstructing the past: The late posterior negativity (LPN) in episodic memory studies
Neuroscience and Biobehavioral Reviews
Strategic retrieval processing and the impact of knowing when a memory was first created
Brain and Cognition
Strategic retrieval in a reality monitoring task
Neuropsychologia
Retrieval processing and episodic memory
Trends in Cognitive Sciences
State anxiety and affective physiology: Effects of sustained exposure to affective pictures
Biological Psychology
fMRI correlates of the episodic retrieval of emotional contexts
NeuroImage
The late posterior negativity in episodic memory: A correlate of stimulus retrieval?
Biological Psychology
Electrophysiological dissociation of the neural correlates of recollection and familiarity
Brain Research
Prior emotional context modulates early event-related potentials to neutral retrieval cues
Journal of Cognitive Neuroscience
Memory-related functional connectivity in visual processing regions varies by prior emotional context
NeuroReport
Picture media and emotion: Effects of a sustained affective context
Psychophysiology
Electrophysiologically dissociating episodic preretrieval processing
Journal of Cognitive Neuroscience
The bank of standardized stimuli (BOSS), a new set of 480 normative photos of objects to be used as visual stimuli in cognitive research
PLoS ONE
False recognition of emotional word lists in aging and alzheimer disease
Cognitive and Behavioral Neurology
Recapitulating emotional context: Activity of amygdala, hippocampus and fusiform cortex during recollection and familiarity
The European Journal of Neuroscience
Brain activations to emotional pictures are differentially associated with valence and arousal ratings
Frontiers in Human Neuroscience
Cited by (1)
The power of negative and positive episodic memories
2022, Cognitive, Affective and Behavioral Neuroscience