Research ques+ons: • Which spa*otemporal quali*es of arm gestures convey emo0on? • How are these cues subjec*vely represented in percep*on? • Are different cues related to different emo*on dimensions?
Predic+ons: Major dimensions of emo0on expression can be perceived and influence different perceived spa0otemporal quali0es of arm gesture.
High arousal ac*ve, fast Pleasant gesture percep*on fluent, located high in space High potency forceful, expansive
What makes bodily expressions emo+onally salient?
Perceptual ra+ng of emo+onal expressions
Conclusions
Scale Verbal description RSB
Arousal Physical excitement in response to the situation 0.98
Potency Power/control over the situation 0.91
Valence Pleasantness of emotional experience 0.89
Intensity Impact of situation on person 0.97
Emo+on dimensions
References [1] Dael, N., Mor0llaro, M., & Scherer, K. R., (in press). Emo0on expression in body ac0on and posture. Emo0on.
[2] Pollick, F. E., Paterson, H. M., Bruderlin, A.,& Sanford, A. J. (2001). Perceiving affect from arm movement. Cogni0on, 82, B51-‐B61.
[3] Fontaine, J. R. J. Scherer, K. R., Roesch, E. B., & Ellsworth, P., C. (2007). The world of emo0ons is not two-‐dimensional. Psychological Science, 18, 1050-‐1057.
[4] Scherer, K. R. (2001). Appraisal considered as a process of mul0level sequen0al checking. In K. R. Scherer, A. Schorr and T. Johnstone (Eds.). Appraisal processes in emo0on: Theory, Methods, Research (pp. 92-‐120). New York and Oxford: Oxford University Press.
[5]Bänziger, T., Mor0llaro, M., & Scherer, K. R. (in press). Introducing the Geneva mul0modal expression corpus for experimental research on emo0on percep0on.Emo*on.
S*muli Geneva Mul0modal Emo0on Portrayals corpus (GEMEP)[5] 12 emo0ons x 10 professional actors (N = 120 video clips) Felt experience enactment of social interac+on scenarios Mul0modal expression
face: blurred voice: muted pseudo sentence body: co-‐speech gestures
Raters 43 undergraduate students (10 males)
Valence
Potency high/low
Posi0ve Nega0ve
Arou
sal
High
Elated joy Amusement
Pride
Hot anger Panic fear Despair
Low
Pleasure Relief Interest
Cold anger Anxiety Sadness
Scale Verbal descrip*on and scale labels RSB
Amount Amount of arm movement: ac0ve / passive 0.99
Speed Speed by which the arms are moving:slow / fast 0.99
Force Physical effort put in the arm movement: strong / weak 0.98
Size Spa0al extent or span of arm movement: compact / expansive 0.98
Fluency Smoothness of speed changes: abrupt / fluent 0.94
Height Ver0cal loca0on of the arms: low / high 0.96
Gesture dynamics
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
Nega0ve Posi0ve
Mean Speed
Valence
High Potency
Nega0ve Posi0ve Valence
Low Potency High Arousal
Low Arousal
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
Nega0ve Posi0ve
Mean Force
Valence
High Arousal
Nega0ve Posi0ve Valence
Low Arousal High Potency
Low Potency
• Both emo0on dimensions and gesture dynamics can be reliably perceived • Arousal and potency strongly affect all perceived spa0otemporal quali0es of gesture
• Arousal boosts the effects of potency on gesture percep0on
• Valence has a subtle effect gesture that is difficult to disentangle from the effects of arousal and potency
• Gesture is an integrated part of nonverbal emo*on communica*on, not only serving cogni*ve-‐linguis*c but also affec*ve-‐expressive func*ons
Body movement (gait, dance, posture, everyday ac0on) conveys reliable informa0on about the emo0onal state of the expressor [1], which can be recognized by others [2]. Limle is known on the func0on of gesture to express and perceive emo0on. Emo+on is encoded in movement categories but also in dynamic quali0es [1, 2]. Arousal is posi0vely related to movement speed, accelera0on and jerk, but no other indicators have been found for emo0on dimensions or components [2]. At least three func+onal dimensions are relevant for emo0on expression and experience: Arousal/ac0va0on, Valence/pleasantness, and Potency (power/control) [3, 4].
Design Emo0ons grouped according to level of arousal (A), valence (V), and potency (P) Effects of A, V, P on ra0ngs of • arm movement • emo0on dimensions ANOVA with A, V, P as within-‐subjects variables (2 levels for each scale)
Table 1. Interrater reliability of emo0on scales Table 2. Interrater reliability of movement scales
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
Arousal Valence Potency
Mean ra*n
g of m
atching dimen
sion
high low high low pos neg 0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
High potency Low potency
Mean He
ight
Valence
High Arousal
High potency Low potency
Valence
Low Arousal Nega0ve
Posi0ve
p< .001, η2 = .93
p< .001, η2 = .66
p< .001, η2 = .44