Electrophysiological correlates of remembering emotional pictures
Research highlights
►Three ERP correlates of recognition (P2, FN400 and LPC) are enhanced by emotion. ►This effect is specific to items accompanied with a feeling of remembering. ►Emotions probably modulate multiple recognition subprocesses.
Introduction
Emotional memories have markedly different phenomenal characteristics when compared to neutral memories. For instance, emotional events tend to be remembered with a greater amount of perceptual and sensorial details, as well as an enhanced sense of vividness and confidence in the accuracy of memory contents (Schaefer and Philippot, 2005). This phenomenon, which will be referred to hereafter as the emotional enhancement of remembering (EER), is thought to play an important role in the etiology of several mental disorders (Lanius et al., 2003, Philippot et al., 2003), in eyewitness testimony (Brown et al., 2008) and in the formation of social representations (Luminet et al., 2004).
Despite being a common feature of everyday life, the cognitive processes underlying the EER phenomenon are still largely unknown. Two possible explanatory frameworks draw from a conceptual distinction between memory strength and recollection, two processes thought to be involved in retrieval memory. Item memory strength, which can be defined as the strength of activation of an item in memory, is thought to vary in a continuous fashion and might produce internal signals forming the basis of feelings of familiarity (Gonsalves et al., 2005). Recollection is defined as the ability to retrieve an item alongside the retrieval of information linked to the context in which it was encoded (Yonelinas et al., 2005, Rugg and Curran, 2007). A common assumption is that EER is at least in part driven by an emotional enhancement of recollection processes enabling an improved reinstatement of contextual and sensorial information present at the time of encoding (Ochsner, 2000, Schaefer and Philippot, 2005). Evidence in favor of this account is provided by studies showing that emotion enhances performance in source memory tasks, that is, tasks requiring the accurate retrieval of elements of the context in which the main to-be-remembered item was learned. For instance, one study has found (Doerksen and Shimamura, 2001) that the color of emotional words was better remembered than the color of neutral words. Other studies have since shown the effect of emotion on other variants of source memory tasks (D'Argembeau and Van der Linden, 2004, D'Argembeau and Van der Linden, 2005, Koenig and Mecklinger, 2008). An alternative account suggests that EER is driven by an emotional enhancement of item memory strength that can result in metacognitive feelings of vivid remembering and memory confidence (Phelps and Sharot, 2008). Evidence for this account is provided by well-known findings that emotion facilitates encoding of the central aspects of a scene to the detriment of peripheral contextual details (Christianson, 1992). Additional evidence in favor of this account shows that emotion can produce highly vivid, confident and apparently detailed, yet inaccurate memories (Talarico and Rubin, 2003), which suggests that the phenomenal richness of emotional memories might be the result of a reconstructive process decoupled from the actual retrieval of contextual details.
A possible model that reconciles these two apparently contradictory views is an account in which EER would emerge from the simultaneous emotional modulation of multiple retrieval processes including both recollection and memory strength. This account could explain how the EER is seemingly linked to recollection of contextual details while also occurring in the absence of accurate retrieval of contextual information. Further, this account would be compatible with the fact that emotion is known to modulate several distinct memory subprocesses (Philippot and Schaefer, 2001, Kensinger and 2008). Having multiple redundant causes would facilitate the occurrence of EER and could therefore maximize its adaptive gains. For instance, one of the putative adaptive advantages of the EER is to provide a sense of memory confidence that facilitates quick adaptive action when facing an object previously associated with threat (Phelps and Sharot, 2008). If recollection of actual contextual details of the previous encounter with the object fails, metacognitive feelings caused by enhanced item strength could still trigger a sense of vivid and confident (yet probably inaccurate) recollection of the previous encounter, which would in turn facilitate quick decisive action.
A useful approach to examine these different accounts of the EER phenomenon can be provided by the electrophysiological old–new effect, a well-established neural index of memory function relying on scalp event-related potentials (ERP). The old–new effect is observed in a classical recognition task in which participants have to discriminate between “old” items (i.e. seen in a previous list) and “new” items (unseen in a previous list). The key feature of the ERP old–new effect is that ERP waveforms triggered by correctly classified “old” items are more positive than waveforms triggered by correctly classified “new” items. This effect is widely thought to reflect retrieval of information in long-term memory (Wilding and Rugg, 1996, Rugg et al., 1998, Curran et al., 2006, MacKenzie and Donaldson, 2007, Rugg and Curran, 2007).
Relevant to the current study, the old–new effect can be separated in subtypes corresponding to distinct retrieval subprocesses, including recollection and memory strength (Finnigan et al., 2002, Rugg and Curran, 2007). There is a wide consensus that recollection is tapped by an old–new effect conforming to a late positive complex (LPC) with parietal maxima occurring in a 500–700 or 500–800 window (Wilding and Rugg, 1996, Rugg et al., 1998, Rugg and Curran, 2007). Several studies have shown that the LPC varies specifically as a function of factors linked to recollection, such as performance in source memory tasks (see Rugg and Curran, 2007 for a review). Item memory strength is thought to be indexed by two earlier old–new effects. First, memory strength is believed to be indexed by an early midfrontal negativity which tends to be larger for new items compared to old items. This effect is often called the “FN400” and is measured mainly between 300 and 500 ms post-stimulus onset (Finnigan et al., 2002, Gonsalves et al., 2005, Curran and Hancock, 2007). Second, memory strength is thought to also be linked to an even earlier (~ 100–300 ms) old–new effect observed in frontal sites and often taking the shape of a P2 component (Tsivilis et al., 2001, Curran and Dien, 2003, Gonsalves et al., 2005). There is a debate on whether the FN400 and P2 old–new effects index generic item strength leading to familiarity or whether they reflect modality-specific memory strength. For instance, it has been suggested that the FN400 is specifically sensitive to conceptual/semantic information (Voss and Paller, 2009b, Voss et al., 2010) and that the P2 old–new effect reflects perceptual priming (Curran and Dien, 2003).
We therefore sought to investigate whether these subtypes of the ERP old–new effect could be differentially sensitive to the EER phenomenon. For that purpose, we used a paradigm that has often been used to operationalize the EER phenomenon. Specifically, the task used in the current study was an old–new recognition test involving emotional and neutral pictures followed by a “Remember–Know” (R–K) task. The R–K task follows a classical old/new recognition task in which participants are asked to judge the quality of their memory for every item that they recognize. A “Remember” (R) judgment is described as a vivid recollection of information pertaining to the context in which the item was learned, and a “Know” (K) judgment is described as a feeling of familiarity triggered by the item, but with no recall of associated contextual information (Henson et al., 1999). Although the R–K task has originally been designed to dissociate recollection from familiarity, it has often been argued that it cannot reliably differentiate between these two processes (Dunn, 2004, Wais et al., 2008, Mickes et al., 2009). However, it is widely accepted that the R/K task is a useful technique to measure qualitative differences between retrieved contents, and more specifically, to assess the degree of phenomenal richness of memories (Talarico and Rubin, 2003, Sharot et al., 2004, Bayley et al., 2005). It is therefore important to note that the current study has used the R/K task primarily as a measure of phenomenal quality of memories, and also because it is a paradigm widely accepted as a measure of the EER phenomenon.
Consistent with the EER phenomenon, a common finding with the R–K task is that the frequency of R judgments is increased when test items have an emotional content, whereas K items tend to remain largely unaffected by emotion (Dewhurst and Parry, 2000, Ochsner, 2000, Kensinger and Corkin, 2003). In line with these findings, we first predicted that the EER phenomenon should be reflected by an emotional modulation of the ERP old–new effect only for R items. Further, we hypothesized that if the EER relies on distinct retrieval subprocesses, then this pattern of results should be observed for distinct subtypes of the old–new effect that can be temporally and topographically separated. In line with prior literature (e.g. Phelps and Sharot, 2008), we focused specifically on old–new effects thought to reflect recollection (LPC) and memory strength (FN400 and the P2 old–new effect).
In addition, we also tested whether the effects of emotion on the old–new effect are different according to the levels of emotional arousal (or intensity). This analysis was necessary to control for a possible confound between response type and arousal, characterized by the fact that R items tend to be more arousing than K items and any difference between R and K waveforms might potentially reflect a difference in arousal rather than retrieval processes. Therefore, ERP effects needed to be tested separately by distinct levels of arousal to control for this possibility. In addition, a previous study (Schaefer et al., 2009) has found that in certain cases emotional intensity has a nonlinear effect on ERP correlates of retrieval. Specifically, that study found that the old–new effect was enhanced by moderate levels of emotional arousal, whereas low (non-emotional) and very high levels of arousal were associated to a smaller old–new effect. The current study tested whether this effect can be observed for different qualities of memory retrieval using the R–K task.
A number of studies have investigated the effects of emotion on retrieval activity using the event-related potentials (ERP) method. Many of these studies have found that emotional contents tend to produce an overall enhancement in ERP correlates of retrieval memory (Johansson et al., 2004, Inaba et al., 2005, Inaba et al., 2007, Langeslag and Van Strien, 2008, Schaefer et al., 2009, Van Strien et al., 2009). However, no study has to our knowledge investigated the effects of emotion on the ERP correlates of the R–K task, which could contribute to a better understanding of the effects of emotion on different subtypes of memory retrieval and the EER phenomenon in particular.
Section snippets
Participants
Thirty-nine healthy right-handed participants from the University of Leeds and surrounding community without history of neurological or psychiatric disorders (24 females, mean age of 21.9, SD = 5.8) completed two experimental sessions with a one-week interval. Twelve participants were excluded (five participants had a hit rate at chance level and seven participants did not have enough artifact-free trials in at least one relevant condition, either because of excessive EEG artifact or because of a
Behavioral data
As expected, emotional pictures were rated during encoding as being more negative than neutral pictures [SAM valence ratings for negative pictures: M = 4.0, S.E. = 0.31; and neutral pictures: M = 2.7, S.E. = 0.29, t(26) = 17.6, p < .001]. Emotional pictures were also rated as more arousing than neutral pictures [SAM valence ratings for negative pictures: M = 2.7, S.E. = 0.81; and neutral pictures: M = 4.1, S.E. = 0.68, t(26) = 14.6, p < .001].
In order to obtain a memory discrimination score, a Pr score [P(hits) −
Discussion
The main finding of the current study is that negative emotion enhanced three distinct subtypes of the electrophysiological old–new effect (P2, FN400 and LPC) specifically for old items associated with a “Remember” judgment. The finding of an emotional enhancement of the old–new effect only for R items is consistent with our behavioral results showing that emotion enhances recognition only when it is accompanied by a feeling of “remembering”. More importantly, the observation that this effect
Acknowledgments
A. Schaefer is supported by the UK Biotechnology and Biological Sciences Research Council (BBSRC). C. Pottage was supported by an undergraduate research bursary from the Nuffield Foundation at the time of data collection. The authors thank Dariush Saeedi, Beth Collins and Karrie Alexander for their help with data collection.
References (86)
- et al.
The neuroanatomy of remote memory
Neuron
(2005) - et al.
A multiple source approach to the correction of eye artifacts
Electroencephalogr. Clin. Neurophysiol.
(1994) - et al.
Measuring emotion: the self-assessment manikin and the semantic differential
J. Behav. Ther. Exp. Psychiatry
(1994) Effects of attention and confidence on the hypothesized ERP correlates of recollection and familiarity
Neuropsychologia
(2004)- et al.
ERP old/new effects at different retention intervals in recency discrimination tasks
Brain Res. Cogn. Brain Res.
(2004) - et al.
The FN400 indexes familiarity-based recognition of faces
Neuroimage
(2007) - et al.
ERP ‘old/new’ effects: memory strength and decisional factor(s)
Neuropsychologia
(2002) - et al.
Memory strength and repetition suppression: multimodal imaging of medial temporal cortical contributions to recognition
Neuron
(2005) - et al.
Neural evidence of effects of emotional valence on word recognition
Int. J. Psychophysiol.
(2005) - et al.
An electrophysiological comparison of recollection for emotional words using an exclusion recognition paradigm
Brain Res.
(2007)
ERP correlates of the incidental retrieval of emotional information: effects of study–test delay
Brain Res.
Corticosteroid effects in the brain: U-shape it
Trends Pharmacol. Sci.
Content dependence of the electrophysiological correlates of recollection
Neuroimage
Retrieving accurate and distorted memories: neuroimaging evidence for effects of emotion
Neuroimage
Age differences in the emotional modulation of ERP old/new effects
Int. J. Psychophysiol.
Recall of emotional states in posttraumatic stress disorder: an fMRI investigation
Biol. Psychiatry
The acute effects of corticosteroids on cognition: integration of animal and human model studies
Brain Res. Brain Res. Rev.
The modulatory effects of corticosteroids on cognition: studies in young human populations
Psychoneuroendocrinology
Dissociating recollection from familiarity: electrophysiological evidence that familiarity for faces is associated with a posterior old/new effect
Neuroimage
Scalp distributions of event-related potentials: an ambiguity associated with analysis of variance models
Electroencephalogr. Clin. Neurophysiol.
Role of adrenal stress hormones in forming lasting memories in the brain
Curr. Opin. Neurobiol.
Affective picture processing: an integrative review of ERP findings
Biol. Psychol.
Event-related potentials and recognition memory
Trends Cogn. Sci.
Differential time-dependent effects of emotion on recollective experience and memory for contextual information
Cognition
Context effects on the neural correlates of recognition memory: an electrophysiological study
Neuron
Noradrenaline mediates amygdala activation in men and women during encoding of emotional material
Neuroimage
Valence interacts with the early ERP old/new effect and arousal with the sustained ERP old/new effect for affective pictures
Brain Res.
Remembering and knowing: electrophysiological distinctions at encoding but not retrieval
Neuroimage
Finding meaning in novel geometric shapes influences electrophysiological correlates of repetition and dissociates perceptual and conceptual priming
Neuroimage
In what way does the parietal ERP old/new effect index recollection?
Int. J. Psychophysiol.
Event-related potentials and the recognition memory exclusion task
Neuropsychologia
Indexing strategic retrieval of colour information with event-related potentials
Brain Res. Cogn. Brain Res.
ERP correlates of item recognition memory: effects of age and performance
Brain Res.
Electrophysiological dissociation of the neural correlates of recollection and familiarity
Brain Res.
The nature of recollection and familiarity: a review of 30 years of research
J. Mem. Lang.
The relation between remembering and knowing as bases for recognition: effects of size congruency
J. Mem. Lang.
A facilitative role for corticosterone in the acquisition of a spatial task under moderate stress
Learn. Mem.
Glucocorticoid release and memory consolidation in men and women
Psychol. Sci.
Pharmacological manipulation of human working memory
Psychopharmacol. (Berl).
Eliciting person descriptions from eyewitnesses: a survey of police perceptions of eyewitness performance and reported use of interview techniques
Eur. J. Cogn. Psychol.
Emotional stress and eyewitness memory: a critical review
Psychol. Bull.
Follow-up of a cross-national comparison on flashbulb and event memory for the September 11th attacks
Memory
Differentiating amodal familiarity from modality-specific memory processes: an ERP study
Psychophysiology
Cited by (78)
Effect of the congruity of emotional contexts at encoding on source memory: Evidence from ERPs
2022, International Journal of PsychophysiologyCitation Excerpt :Clearly, as previously studies noted, emotional arousal stimuli can attract more attentional resources compare to neutral stimuli and gain more elaboration in encoding, and then exhibit an emotional memory advantage (Hamann, 2001; Weymar et al., 2009). Numerous studies have focused on the facilitation effect of emotion on item memory, which refers to the memory for the central item of an event (Christianson et al., 1991; Slotnick et al., 2003), suggesting that item memory is enhanced for emotional items compare to neutral items (Buchanan and Adolphs, 2002; Christianson et al., 1991; Dolcos et al., 2005; Otani et al., 2007; Schaefer et al., 2011). However, results from existing findings regarding emotionality's effect on source memory are mixed, source memory refers to the memory for details (e.g., temporal order, spatial location, color, and so forth) that accompany the central component of an event (Chiu et al., 2013; Johnson, 1993; Minor and Herzmann, 2019).
The influence of negative emotional intensity on dual-processing recognition
2021, Biological PsychologyCitation Excerpt :In addition, the frontal old/new effect was elicited by remember responses for neutral and mildly negative stimuli. The results indicated familiarity was affected by detail extraction, which is an important indicator of memory intensity (Schaefer, Pottage, & Rickart, 2011; Vilberg, Moosavi, & Rugg, 2006; Woodruff, Hayama, & Rugg, 2006). The frontal old/new effect between remember and new responses was continuously enhanced with the decrease of negative emotional intensity.