Emotional Arousal and Valence Effects on Memory

In recent years, the influence of emotion on memory has been studied extensively. Data seems to indicate that stimuli which evoke emotion are remembered better than neutral stimuli. However, emotion does not only enhance memory, it may also distort memory performance by inducing a "recognition bias" (Windmann and Kutas, 2001). The amygdala, a subcortical structure, has been linked to emotional arousal. Some studies have indicated that amygdala activity may modulate hippocampal and medial temporal lobe activity, causing memory effects (Dolcos and Cabeza, 2002). Other experiments suggest ventromedial/medial prefrontal region involvement in memory and learning (Windmann and Kutas, 2001). Memory effects are studied in conjunction with EEG recording and ERP analysis in an attempt to localize the region related to emotional memory and quantify its effects.

In recent years, the influence of emotion on memory has been studied extensively. Data seems to indicate that stimuli which evoke emotion are remembered better than neutral stimuli. However, emotion does not only enhance memory, it may also distort memory performance by inducing a "recognition bias" (Windmann and Kutas, 2001). The amygdala, a subcortical structure, has been linked to emotional arousal. Some studies have indicated that amygdala activity may modulate hippocampal and medial temporal lobe activity, causing memory effects (Dolcos and Cabeza, 2002). Other experiments suggest ventromedial/medial prefrontal region involvement in memory and learning (Windmann and Kutas, 2001). Memory effects are studied in conjunction with EEG recording and ERP analysis in an attempt to localize the region related to emotional memory and quantify its effects.

In 2002, Dolcos and Cabeza completed the first study of the "subsequent memory effect" for emotional pictorial stimuli. The subsequent memory effect is an ERP effect in which remembered stimuli cause more positive ERP's than non-remembered. In this experiment, a pool of 60 pleasant, 60 unpleasant and 60 neutral pictures are used. Subjects are asked to view the pictures and rate them according to their initial reactions. These ratings are ranked on a five point scale along the valence continuum from pleasant to unpleasant with three being neutral. The subjects are also instructed to memorize the image for a later memory test. After the study portion, the subjects are asked to write a descriptive sentence about each image. An EEG is recorded throughout for later analysis.

After an analysis of memory performance, the data shows a significant difference between neutral and non- neutral stimuli, but not between pleasant and unpleasant. Overall, ERP's for emotional pictures are more positive than for neutral pictures. This indicates that arousal is the key factor in emotional memory. During the early epoch (500-800 msec), however, the emotion effect is mostly sensitive to valence at the frontocentral electrodes. While in parietal electrodes the effect seems to be arousal related, the frontocentral electrodes record the emotion effect earlier for pleasant pictures than unpleasant ones. The memory data shows that, for both emotional and neutral stimuli, in frontocentral and centroparietal regions remembered images elicit more positive ERP's than non- remembered ones. However, during the early epoch, the subsequent memory effect is larger, centrally, for arousing than neutral images. This suggests that emotional pictures are encoded faster than non-emotional images.

As opposed to studying recall of pictorial stimuli, Windmann and Kutas' 2001 Electrophysiological Correlates of Emotion- Induced Recognition Bias experiment studies word recognition memory. According to the authors, emotional stimuli can cause a "recognition bias." This means that subjects become more likely to respond "old" to a negative stimulus than a positive or neutral stimulus. Whether the word is experimentally "old" or "new" has no effect on the subject response. The experiment studies the effect of the word's emotional valence on "response bias" and memory.

Windmann and Kutas' subjects are shown 70 neutral (~90%)/positive (~10%) verbs and 70 verbs with a negative connotation. After the initial study portion, a 30 minute distraction period and then a test period follows. During the recognition portion, 88 "new" words, both neutral and negative, are added to the original set. For the purposes of this experiment, there is no significant difference between neutral and positive word responses, so they are grouped together. The subject is then shown a random combination of "old" and "new" words and asked to identify whether or not he has seen each word before. An EEG and EOG are recorded during these tasks.

Results show that, compared to neutral words, emotionally negative words are more quickly and frequently classified as "old." They are not, however, recognized more accurately than neutral words. ERP's for correctly identified words do not show any effects of emotional valence until very late; while valence affected "old" response ERP's much sooner. The theory that intentional retrieval begins in the prefrontal cortex is supported by the fact that correct "old" recognition shows more positive ERP response in the frontal area. This early valence effect seems, then, to support the theory that the bias occurs at an unconscious, automatic memory level.

There seem to be a number of shortcomings in Windmann and Kutas' (2001) study, which Dolcos and Cabeza (2002) do take into account when designing their experiment. Most importantly, the recognition study does not differentiate between arousal and valence. The authors state, "we have referred to emotional valence rather than to emotional arousal throughout this report because we are interested primarily in emotion- related information processing patterns, not in process associated with emotional patterns... All in all, we believe that our stimuli probably did not induce any significant physiological arousal in our subjects" (Windmann and Kutas, 2001, pg. 587). They do recognize, however, that the negative words differ from neutral ones in arousal value because of their fight or flight association.

Also, according to Dolcos and Cabeza (2002), word stimuli generate a smaller overall effect than picture stimuli. As such, it is more difficult to find a clear difference between valence and arousal. Lastly, Dolcos and Cabeza (2002) theorize that previous experiments have not accurately studied valence because they only compare neutral and negative stimuli. This is evident in Windmann and Kutas' (2001) word categories; neutral combined with a few positive words, as compared to negative words.

Both the picture recall and the word recognition studies utilized ERP data to localize regions of emotional memory activity. Although there are many methodological and experimental differences between the two, both indicate prefrontal cortex in the pathway. The recall experiment is a little more specific, indicating frontocentral and centroparietal regions. This possibly results from a more focused and detailed experimental technique. In any case, these studies indicate that emotion plays a large role in memory processing and situational assessment. As more studies on memory and emotion are conducted, the pathway and cause of such effects will become more evident.

References
Windmann, S. & Kutas, M. (2001). Electrophysiological Correlates of Emotion- Induced Recognition Bias. Journal of Cognitive Neuroscience, 13:5, 577- 592.

Dolcos, F. & Cabeza, R. (2002). Event- Related potentials of emotional memory Encoding pleasant, unpleasant and neutral pictures. Cognitive, Affective & Behavioral Neuroscience, 2:3, 252- 263.

MG