The list-composition effect in memory for emotional and neutral pictures: Differential contribution of ventral and dorsal attention networks to successful encoding

Abstract

The Emotional enhancement of memory (EEM) is observed in immediate free-recall memory tests when emotional and neutral stimuli are encoded and tested together ("mixed lists"), but surprisingly, not when they are encoded and tested separately ("pure lists"). Here our aim was to investigate whether the effect of list-composition (mixed versus pure lists) on the EEM is due to differential allocation of attention. We scanned participants with fMRI during encoding of semantically-related emotional (negative valence only) and neutral pictures. Analysis of memory performance data replicated previous work, demonstrating an interaction between list composition and emotional valence. In mixed lists, neural subsequent memory effects in the dorsal attention network were greater for neutral stimulus encoding, while neural subsequent memory effects for emotional stimuli were found in a region associated with the ventral attention network. These results imply that when life experiences include both emotional and neutral elements, memory for the latter is more highly correlated with neural activity representing goal-directed attention processing at encoding.

Local context influences memory for emotional stimuli but not electrophysiological markers of emotion‐dependent attention

“…For example, in our previous work, we showed that arousal had direct effects on early long‐term free recall tests of memory even when attentional effects were covaried out statistically (Pottage & Schaefer, ; Talmi & McGarry, ; Talmi, Ziegler et al, ) or through experimental manipulation using divided attention (Talmi & McGarry, ). In our fMRI study, which used an identical paradigm to the one reported here, we have also not observed any evidence for reduced attention to neutral stimuli in mixed lists (Barnacle et al, ). How should we understand the decoupling of attention and memory, which are typically closely linked?…”

“…The emotion modulation of the LPP and SW is typically considered to be a result of bottom‐up modulation of visual attention; but when participants intentionally encode neutral items for a subsequent test, they could well recruit visual attentional resources through top‐down means. Indeed, in our recent fMRI study, the successful encoding of neutral scenes in mixed lists relied on top‐down attentional resources to a greater extent compared to the successful encoding of neutral scenes in pure lists (Barnacle et al, ). It was therefore entirely possible that the emotional modulation of these ERPs would be attenuated in our task, where participants allocated their full attention to intentional encoding (Holmes, Mogg, de Pockert, Nielsen, & Bradley, ).…”

“…Behavioral work has established that emotional memory enhancement in free-recall tests of early long-term memory is readily observed when emotional and neutral stimuli are encoded and recalled together in "mixed" lists, but is weaker when these scenes are encoded separately, in "pure" lists. This effect can sometimes disappear completely when pure lists are additionally controlled for confounding factors such as differential organization and attention (Barnacle, Montaldi, Talmi, & Sommer, 2016;Hadley & MacKay, 2006;Sommer, Gläscher, Moritz, & B€ uchel, 2008;Talmi, Fuentemilla, Litvak, Duzel, & Dolan, 2012;Talmi, Luk, McGarry, & Moscovitch, 2007;Talmi & McGarry, 2012). The context dependence of the effect of emotion on memory is surprising because a pure list of emotional scenes is effectively an operationalization of a real-life event that consists of a number of emotional aspects, for example, witnessing a traffic accident where one might observe injured persons, damage to property, medical personnel, and so on.…”

A retrieved context model of the emotional modulation of memory

“…In that study, temporal and emotional context effects were confounded because the three item categories were studied in three separate lists, but emotional clustering appears also in free recall of mixed lists that contain both emotional and neutral items. For example, Barnacle et al (Barnacle, Montaldi, Talmi, & Sommer, 2016) presented participants with mixed lists of emotional and neutral pictures which were equally semantically related, and found clustering around the emotional/neutral category; the degree of clustering predicted memory performance. Long and colleagues (Long, Danoff, & Kahana, 2015) presented participants with mixed lists of words, some of which were negative and some positive, and found that participants tended to retrieve positive items after other positive items, and similarly for negative and neutral items.…”

Alpha Oscillations Prior to Encoding Preferentially Modulate Memory Consolidation during Wake Relative to Sleep

“…Emotional content influences attention during encoding (Barnacle, Montaldi, Talmi & Sommer, 2016). Regions of the brain activated by emotional content, such as the amygdala, promote hippocampal-dependent memory consolidation (LaBar & Cabeza, 2006;Murty, Ritchey, Adcock & LaBar, 2010;Smith, Henson, Dolan & Rugg, 2004).…”