visual cortex
Abstract
Emotionally salient stimuli such as naturalistic scenes, facial expressions and emotional body language elicit increased neuronal activation in associative sensory cortex, as well affective cortical and sub-cortical networks. It has been suggested, that the affective quality of these stimuli is largely a result of evolutionary predisposition and that their selective perceptual encoding may thus be biologically prepared. The present study tested whether comparable neuronal processing patterns are also elicited by expressive hand gestures, bearing distinct emotional yet highly symbolic meaning.
During functional scanning, 33 subjects passively viewed blocked picture presentations of pleasant, neutral and unpleasant hand gestures. Consistent with a priori assumptions, emotionally meaningful gestures were associated with increased BOLD activations in extrastriate visual cortex, as well as in parietal areas. These results suggest that emotional significance detected by the brain is not exclusively founded in the genetic heritage of the human species but may also get attached to highly symbolic stimuli by social learning.
Introduction
An increasing number of studies confirm that the brain systematically discriminates between emotionally salient and neutral environmental stimuli. From an evolutionary point of view, this is highly profitable for organisms because as a result optimally adaptive behavior enhances the organism’s survivability and reproductive success (Cacioppo et al., 1999; Lang et al., 1997; Öhman et al., 2000a). Evidence in support of such selective emotion processing is provided by brain imaging studies showing that emotional as compared to neutral stimuli are associated with increased neuronal processing in various brain structures. In particular, enhanced brain activity is consistently observed in several areas implicated in visual information processing, including the occipital and fusiform gyri, as well as the vicinity of the superior temporal sulcus when contrasting emotional and neutral picture contents (e.g., Bradley et al., 2003; Junghöfer et al., 2005; Lang et al., 1998b; Sabatinelli et al., 2004;
Sabatinelli et al., 2005). Furthermore, recent studies observed that the signal changes in the visual cortex varied strongly with self reports of emotional arousal. Specifically, pleasant and unpleasant pictures high in arousal were associated with strong and widely distributed visual cortical activations compared to low arousing emotional or neutral materials (Bradley et al., 2003; Junghöfer et al., 2005). Moreover, affectively enhanced activation of associative sensory cortex has not only been observed in the visual, but also in the auditory modality (Grandjean et al., 2005). Finally, emotional modulation of neuronal activity is also reported for several structures of affective cortical and sub-cortical networks, including the amygdala (e.g. Hadjikhani & de Gelder, 2003; Hamann et al., 2004; Junghöfer et al., 2005; Norris et al., 2004;
Sabatinelli et al., 2005; Vuilleumier et al., 2001; Whalen et al., 1998), the insula (e.g., Junghöfer et al., 2005; Simmons et al., 2004), as well as orbito-frontal cortex (e.g., Karama et al., 2002; Murphy et al., 2003; Nitschke et al., 2004; Vuilleumier et al., 2001).
A great number of these studies used picture materials depicting contents of high evolutionary relevance, such as mutilated bodies, threatening animals and erotic scenes. However, being a species relying heavily on complex social interactions humans not only need to identify and react to environmental stimuli signaling immediate physical danger or reproductive opportunity. Instead, it is also of eminent importance to perceive socially relevant signals, in effect allowing for optimized
behavior in a complex social environment. Accordingly, verbal and non-verbal social cues provide information about the motivational and emotional states of interaction partners, as well as the surrounding environment (fearful facial expressions may e.g.
indicate immanent danger). Consequently, they can be used to predict intentions, and likely future interactions (see Keltner & Kring, 1998). Because of this, it has been stressed that social and emotional processes are strongly interdependent and that socially relevant stimuli may thus be characterized by inherent emotional significance (Keltner & Kring, 1998).
In the context of emotion research, special importance has been attributed to communicative aspects of the face and the eyes as this class of non-verbal behaviors is regarded to have a major impact on the expression of emotion and on the regulation of inter-personal interactions (see Ekman & Oster, 1979; Öhman, 1986). Like naturalistic pictures, emotional facial expressions are processed preferentially as compared to neutral ones. In behavioral experiments, schematic threatening faces are recognized faster than friendly ones within an array of neutral distractors (Öhman et al., 2001b).
On the neuronal level, studies using EEG-measures found distinct electro-cortical indices of selective emotion-processing to be sensitive to the emotionality of faces (Schupp et al., 2004c). Also, a great number of studies have shown that emotional faces automatically attract attention (see e.g. Pessoa et al., 2002b; Vuilleumier, 2002).
Corresponding to results obtained using naturalistic scenes, emotional faces elicit heightened neuronal processing already in visual sensory areas, as well (e.g., Breiter et al., 1996; Pessoa et al., 2002a; Vuilleumier et al., 2001). Interestingly, this increased neuronal sensitivity to emotional facial expressions is not unique to the human species.
Face-sensitive neurons in the temporal visual cortex of macaque monkeys also respond with an increased firing-rate when viewing emotionally aroused as compared to neutral faces (Sugase et al., 1999).
Additional evidence for the preferential processing of affect-laden social cues is provided by studies examining ‘emotional body language’ (EBL, de Gelder et al., 2004; Hadjikhani & de Gelder, 2003; Meeren et al., 2005; Stekelenburg & de Gelder, 2004; for an overview see de Gelder, 2006). EBL involves hand and arm movements, as well as body posture, in combination providing revealing insights into somebody else’s emotional condition. Corresponding to the processing of emotional faces, the perceptual analysis of emotional body language as compared to neutral
postures is associated with increased neuronal activation of higher-order visual cortical areas (de Gelder et al., 2004; Hadjikhani & de Gelder, 2003). This was also confirmed by a recent study by Grosbras and colleagues (2005), reporting increased BOLD activity in temporo-occipital cortical regions when subjects viewed hand movements carried out in an angry as opposed to an emotionally neutral fashion. Even though no direct electro-cortical differentiation between emotional and neutral body postures has been reported to date, ERP-measures are still sensitive to the emotionality of body language, as demonstrated by an emotional congruency effect of faces and body postures apparent at the P1-component, starting at around 100 ms and observed over occipital leads (Meeren et al., 2005).
Altogether, emotional facial expressions, body language and naturalistic scenes share an emotional significance that may inherently be rooted in the phylogenetic heritage of our species. This idea was formulated early on by Darwin (1872) who proposed that emotions would not be private mental episodes but would instead have adaptive behavioral value, promoting favorable actions under given situational circumstances. In correspondence with this assumption, the evolutionary significance of such environmental cues promotes selective perceptual processing, attention capture and rapid aversive learning (see Öhman & Mineka, 2001). As Öhman and Mineka (2001) suggested, such stimuli might be considered to be ‘biologically prepared’. Specifically, whereas their inherent emotional significance may not be entirely innate, it may rather determine a general readiness to respond to fear-relevant stimuli such as angry faces or snakes and to quickly learn the association with according emotional states in aversive contexts. The observation that phobias most prevalently occur in dependency of evolutionarily significant fear cues also points in this direction (e.g., Marks, 1969). Furthermore, these stimuli appear to be greatly characterized by cultural universality. More specific, pictures of mutilated bodies or snakes may evoke aversive emotional reactions in people all around the world, regardless of their respective cultural background. Also, emotional facial expressions, at least involving those considered by Ekman to represent basic emotions, have been reported to be unambiguously identifiable for individuals of widely differing cultural background, including people of European, African and Asian origin (Ekman & Oster, 1979). Therefore, certain environmental cues as well as species-specific expressive emotional non-verbal behavior having its roots in adaptive actions (see also Schmidt &
Cohn, 2001) might be considered to be ’biologically prepared’ (see Bradley et al.,
2001; Lang et al., 2000; Öhman, 1986; Öhman & Mineka, 2001) and may thus be an integral part of each species’ evolutionary heritage.
Based on these considerations, it remains an open question however to what degree the signature of neuronal emotion processing may not only reflect the evolutionary preparedness of the eliciting stimuli, but may also rather be a function of culture-specific social learning. Specifically, to what extent are evolutionarily insignificant stimuli whose emotional meaning is culturally mediated also subject to preferential processing in the brain? Moreover, does such selective encoding resemble typically observed processing patterns, i.e. regarding involved brain areas and time course of the underlying processes? Accordingly, we suggest that another class of non-verbal communicative signals may serve as an adequate means to investigate these questions. More specific, simple static hand gestures are widely used across all cultures all over the world (for a comprehensive overview about worldwide use and meaning of numerous expressive hand gestures see Morris, 1994). Similar to the aforementioned social cues they are also carriers of socially important and emotionally relevant information. Moreover, meaningful hand gestures have been shown to be processed semantically (Wu & Coulson, 2005) and are associated with distinct neuronal activation patterns (Gallagher & Frith, 2004; Nakamura et al., 2004).
However, as opposed to the emotional non-verbal signals investigated in prior research, the meaning attached to hand gestures is highly symbolic and conventionalized, being strongly dependent on a shared cultural context between the sender and the perceiver of the non-verbal message. Therefore, whereas a particular hand gesture may be associated with a strong emotional meaning in one cultural communication context, it may not be associated with any semantic meaning at all in another culture.
Thus, in the present study we investigated the neuronal correlates of the processing of expressive hand gestures by means of functional magnetic resonance imaging. We assumed a priori that hand gestures posses culturally mediated emotional significance and that this would be reflected both at the behavioral, as well as the neuronal level. Behaviorally, we expected a differentiation between the pre-selected gesture categories on self-report scales judging the emotionality of the used pictures.
In terms of hemodynamic responses elicited by the different hand gestures, we expected an increase in neuronal activation while viewing emotional as compared to
neutral gesture categories. Specifically, since various classes of emotional stimuli have been consistently reported to be associated with stronger activation in associative sensory areas, we formulated our strongest anatomical hypothesis for the extrastriate visual cortex. Additionally, increased activation during the viewing of emotional gestures was also expected in affective cortical and sub-cortical networks frequently reported in previous studies, including the amygdala and the insula, as well as orbito-frontal cortex.
Method
Participants
33 subjects participated in the study (16 male). All of them received monetary compensation or course credit and a copy of their structural T1-MR-image for their participation. Participants were between the age of 19 and 33 years (M = 23.1) and were naïve with respect to the study’s aims and hypotheses, i.e., had no prior experience with the used picture stimuli. All of them were familiar with the meaning of the middle finger jerk and thumb up gesture. However, three subjects (all female) reported exclusive primary meanings of the finger points related to threat (pointing towards a companion) or punishment and were therefore excluded from further analyses. Subjects were also selected with respect to the specific requirements set by the MR-safety procedures, i.e. no non-removable magnetic objects of any kind within their body and no history of epilepsy and claustrophobia. Also, in order to fit the stimulation goggles over the subjects’ eyes within the head coil, only subjects were selected with average or small head size. Before the experiment, subjects read and signed forms of consent containing detailed information regarding MR-safety-procedures and possible risks linked with magnetic resonance imaging examinations.