Physically emotional: the role of embodied emotions from encoding to memory

Theories of embodied cognition hold that the perception of an emotional stimulus can trigger a simulation of the correspondent state in the motor, somatosensory, and affective systems. Amongst other bodily reactions, it is thought that such embodied simulations are also reflected in facial expressions in accordance to the emotional connotation of the presented stimulus – a phenomenon also referred to as facial motor resonance. Chapter 1 reviews the theories of embodied cognition, in general, and facial motor resonance, in particular. The aim of the present thesis was to further define the function of embodied simulations, reconciling previous inconsistent results concerning the level at which embodied simulations affect the processing of emotional information, and to explore uncharted aspects of embodiment theories such as their role in memory for emotional information. In Chapter 2, I investigated the hypothesis that embodied simulations play a key role in processing only emotional information (happy and sad sentences and faces), which is low in emotional intensity or difficult to encode. This hypothesis was tested in a behavioral experiment involving a group of participants undergoing subcutaneous cosmetic injections of Botulinum Toxin-A (Botox) compared with a matched control group. The results confirmed the hypothesis: participants in the Botox group, but not those in the control group, rated emotional sentences and faces as less emotional after the Botox treatment. Furthermore, they were slower at identifying sad faces as sad after the treatment. The critical nuance of these findings was that only stimuli with moderate emotional intensity were affected. Upon considering the findings of Chapter 2, the question arose as to whether facial motor resonance, in addition to playing a role in the initial processing and recognition of emotional content, also determines its retrieval. This topic was investigated in the study reported in Chapter 3, in which eighty participants underwent a memory task for emotional and neutral words. The task consisted of an encoding and a retrieval phase. Facial muscles were blocked by a hardening facial mask in one of four conditions: during encoding, during retrieval, during both encoding and retrieval, or never (control). The results showed that memory for emotional words decreased significantly if embodiment was blocked at either point in time during the experiment (during encoding, during retrieval, or during both), in contrast to the control condition. These results suggest that facial motor resonance is involved in the encoding and retrieval of emotional words. In Chapter 4, this line of research was extended and applied to the processing of emotional content in a second language (L2). In a classical memory task involving an encoding and a retrieval phase, thirty-two Spanish/English late bilinguals were presented with emotional (happy and angry) and neutral words. Electromyographic (EMG) activity and skin conductance (SC) were recorded during the encoding phase. The results suggest that the emotionality of an L2 appears to be not only reduced as compared with a first language (L1), but also to be less embodied. This was suggested both by the absence of the Enhanced Emotional Memory (EEM) effect in L2 as well as by partially decreased and delayed EMG and SC activity in response to emotional words in L2 as compared with L1. If facial motor resonance is involved in the recollection of emotional information, what is its role in forgetting emotional information? This question was pursued in the study reported in Chapter 5, employing the directed forgetting paradigm (DF), which involves the presentation of a stimulus (e.g. a word), followed by a cue to “remember” (R-cue) or to “forget” (F-cue). Twenty-one participants were instructed to remember or to intentionally forget neutral, negative, and positive words. EMG from the zygomaticus and corrugator muscle was simultaneously recorded with event related potentials (ERPs). The behavioral results showed that both neutral and emotional words were forgotten at equal rates. However, the type of word and cue instruction interactively modulated facial motor resonance, as measured by EMG. Upon R-cues, the muscle activation patterns for both negative and positive word types were significantly enhanced, in contrast to the facial motor resonance evoked by F-cues. It was speculated that the increase in facial motor resonance reflects active rehearsal, whereas the decrease is associated to active suppression mechanisms. This assumption was supported by the ERP data, indicating that the successful forgetting of affective words required more active suppression, as was indexed by enhanced frontal positivities. In contrast, intentional encoding of emotional words followed by R-cues seemed to be facilitated by an enhanced P3 and late positive potential (LPP) components emerging from centro-parietal areas. These components have been hypothesized to reflect rehearsal and memory consolidation processes. Overall, the present results suggest that embodied simulations help with the processing of indefinite emotional information and assist with the formation of enduring representations of emotional stimuli. The implications of these findings for theories of embodied cognition, in general, and for emotion processing, in particular, are discussed in Chapter 6.