- Email: [email protected]
Equivalent perceptual asymmetries for free viewing of positive and negative emotional expressions in chimeric faces
002X 3932 93 ShOO+OO c 1993 Pergdmon Pres,Ltd
NOTE EQUIVALENT PERCEPTUAL ASYMMETRIES FOR FREE VIEWING OF POSITIVE AND NEGATIVE EMOTIONAL EXPRESSIONS IN CHIMERIC FACES STEPHEN
D. CHRISTMAN* and MICHELLE D. HACKWORTH
University
of Toledo,
Toledo,
(Receilaed 21 July 1992; accepted
OH 43606, U.S.A. 21 January 1993)
Abstract-Research employing chimeric stimuli (in which smiling and neutral half-faces are paired) has demonstrated greater influence of the left half-face in determining perceived intensity of expression. To date, no studies have examined how emotional expressions other than happiness are perceived in this format. Right-handed subjects viewed chimeric faces depicting both positive (happiness, pleasant surprise) and negative (sadness, anger) emotions in a free vision task. Results indicated a left half-face bias for all four emotions, supporting the hypothesis of a greater right hemisphere role in emotional perception. The lack of differences in strength of left half-face bias as a function of the specific emotion depicted suggests that results obtained with typical chimeric half-face paradigms can be generalized to emotions other than happiness.
INTRODUCTION RESEARCH IN THE AKEA of brain lateralization generally indicates a predominant right hemisphere role in the perception of emotion [7,8, 141. One of the more recent and innovative methods to address this issue involves the use of chimeric faces in a free-vision task. The most common type of chimeric face stimulus combines an emotive half-face with a neutral half-face to produce a composite which is viewed by subjects under free viewing conditions. When subjects are asked to compare the intensity of emotional expression for a pair of chimeric stimuli which are mirror-images of one another (thus differing only with respect to the side of the face in which the emotion is shown), the result is most often a left half-face bias, presumably reflecting a right hemisphere response [7]. To date, virtually all free-vision studies utilizing chimeric face stimuli have employed only positive, happy halffaces paired with neutral half-faces. Thus, the perception of negative emotions in the context of chimeric faces has been relatively unexplored. Those studies that have employed emotional expressions with negative valences have exclusively paired happy and sad half-faces together simultaneously [2, IO], thus precluding independent examination of the roles of positive vs negative valence. Evidence suggests that emotional valence may influence the degree or direction of lateralization of emotional processing, with research indicating that the left vs right hemispheres may be specialized for positive vs negative emotions, respectively [I, 3, 123. This raises the possibility that findings obtained with positive emotions may not generalize to negative ones; for example, if negative emotions are more strongly represented in the right hemisphere, then the strength of the left half-face bias may be stronger for negative than positive emotional expressions. The present study extends the free-vision chimeric method by employing both positive and negative emotional expressions to pursue possible hemispheric differences with regard to emotional valence and to assess the generalizability of the results obtained with chimeric face stimuli depicting positive emotions.
*Address for correspondence: OH 43606, U.S.A.
S. Christman,
Department
621
of Psychology,
University
of Toledo,
Toledo,
621
NOTI
METHOD
Four poscra. I*() malt and two female. were photographed, each dijplaqing a neutral cxprcsslon and four emotional expressions (happy, pleasant surprise, bad and angry). I-or each poacr. prints were made depicting each crpresion in tither a normal or mirror-revused orientation. F-allowing the procedure of HruIx and LI L’Y [4]. the photographs were cut down the midllne and recombined to make tuo diflicrcnt types of chimera\: ln one, the emotional expression \\as present on the left halfofthe poser’s face and the neutral exprcwon on the right half: in the other. thir was rc\cr\ed. For any given stimulus pair. the emoti\c half-face in both mcmhers of the pair had been produced by one side of the poser’s fact and the neutral esprcwon by the other. with each side of a poser’\ face occurring equally often a\ the left and right half-facea. This \~a?,done to control for poaihlo facial as>mmctries of the posers. A total of32 pairs wcrc constructed. )xlded hq tho faclorial comhinatiun offcour posers h> four emutlvns h! two vc‘rslons (half-face from poser‘s left 1s Iright) Each of the 32 chimcric pairs ucrc arranged \crtlcallq and ph~)tocopied to prclduce hooklcts L\ hich were administered individually to subjects. The order of ~qpe\ of emotion. poser idcntit,, and position ofchimeric stimulus with emotive left half-face (ahow or belw ) %a:,Irandomwed in the hooklct.
Small groups ol’ubjccts (.\ =4 8) \bcrc arwnhlcd in a classroom. Each was plvcn a stimuli hoohlct containing ;I handcdncs~ questionnaire. 32 pages ofchmleric stmul~ pairs. and ;I mood scale. Suhyxth were asked to examine each pair of chlmeric stimuli and to dctcrmlne which face loohcd more emotive (c.g. “Which fact looks happier’! Anpricr’l Sadder’? More surprised’!“). Thq were allowed to give a “can’t decide” response ifthe) feh the IWO faces wcrc cwcll~ equal in intensity ofemotional expression. Left-biased (right-biased) responses uere those In uhich the chimera with the emotion to the wewcr’s left (right) was Judged by the \uhjccts ah being more emotive. blood ~a\ evaluated using :I procedure developed by I>,\\II) L2] in which suhjccts marked a pwtlon wrrcspondlng to thclr prexnt mood on ;I 10 m sale Inbclcd “mo\t deprcsscd au” at fhc left end (scored as /cro) and “mo\l happy c\er” at the extreme right (acored a 10). Intermedlatc poaltion\ wrc wored :,b the distance 1111 cm I from the left end of the talc. The midpoint \\a\ not Indicated on the wale.. Thih mood scale ~a\ uxd to ewminc an) powhle relationb betv,een >ublect\ mood and their rlpht or left blah 111the judgcmcnt of em~>tlonal intcnalt!
RESULTS The Iateral bus scow wcrc the number of chlmcrlc pa”‘\ for each cmotion m which ;I leftward hlas wab shown minus the number of pairs 1n which a rightward hias v,as bhwn. divided by the total number of pair\ for cxh emotion (,v= X). Posltlve \aIues indxatea left half-fact hiasand ncgativr \alucs indicate;, righf half-face hIa\. Thcsc w~rcs are presented in Table I. Preliminary analyses induted no maln cffccts or intcractlvns involving gender. 40 ~uhsequent analyses collapsed xro\s this variable A one-!\a!, wlthin buhyxth .4NOVA \\a~ used lo ewmlnc an> dlfferences in nature of the half-face bins hetuecn the four cmotions. No G_eniticant dllTercnccs uure found LF< I 1. f-test5conducted IO contirm the prescncc of a half-fact lxa (i.e. a non-xro scow) yielded signllicant left half-face buses for all four cmotlons: happy [I (571-3.17. P
Mean bias wore C‘orrelatlon with mood score
0.2 IO 0.30
0.145
0.154
0.096
0.039
0.145
-o.oxx
NWk
623
DISCUSSION Results indicated significant and equivalent left half-face biases for emotional expressions of both positive and negative valence. in the free viewing ofchimeric stimuli, indicating that emotional valence is not an important factor in free vision chimertc face tasks. The general nature of these results is consistent with previous studies showing a left bias/right hemisphere specialization for the perception of positive emotions 17.8, 131. More important. the current study extends these findings to emotions other than happiness. demonstrating that previous findings from studies employing chimeric stimuli composed of smiling and neutral half-faces can be generalized to other emotions of both positive and negative valence. The correlational data indicated a negative correlation between subjects’mood and the magnitude of the left bias judgcment of happy chimerlc faces. These results may be compared with the findings from DAVID’S study [2] reporting a positive correlation between subjects’ mood and their choice of facial affect; namely, the more positive the mood, the greater the tendency to label the ambiguous chimeric faces as “happy”. However, the magnitude of the perceptual bias was not affected by subjects’ mood. The present study, on the other hand, indicated that subjects reporting a more negative mood exhibited a stronger left bias in the judgement of happy chimeric faces. This result is in contrast to results from subjects suffering from clinical depression. who exhibit a waker left bias [S]. This discrepancy may derive from basic differences in the neurological substrates ofclinical depression vs the overall less beverc negative moods exhiblted by the normal subjects in the current study. Our results could also possibly reflect a relation between negative mood and some form of greater right hemisphere activation, although its selective occurrence with only happy faces suggests either a more tonic response to the perception of positive emotion or the possibility of a Type I error. Overall, the lindings from this study suggest that the almost exclusive use of happy emotional expressions in chimeric free-vision tasks is not problematic. Emotional valence appears to not influence the perception ofchimeric faces. Thus, the happy chimeric faces used in most previous studies serve as valid indicators of lateralization. regardless of the limited range of emotional valence used. Indeed. there is evidence that at least part of the basis for left-sided bias in the perception ofchimeric face stimuli derives from factors unrelated to the processing of emotional or even facial information, as other studies have demonstrated the presence of left-sided biases in judgements of f&al identity and in spatial judgements of nonface chimeric stimuli [6.9]. The lack of effects of type of emotion in the current study aupport the notion that the left bias obtained with chimeric face stimuli is driven not by emotional processes per se, but rather by more basic perceptual factors associated with spatial processing.
A[.~,lr,\~/rd~rmrnrs -This research was supported in part by an Academic Challenge Grant to the Department of Psychology at the University of Toledo. Portions of this paper are based on a Senior Honor’s Thesis conducted by the second author. The authors gratefully acknowledge the assistance of Hongkeun Kim in the preparation of the stimulus booklet.
REFERENCES I AHFKN, G. L. and SCHWAKTZ, G. E. Differential lateralization for positive and negative emotion in the human brain: EEG spectral analysis. Neuropsycholoyia 23, 745-755, 1985. 2. DAVID, A. S. Perceptual asymmetry for happy-sad chimeric faces: Effects of mood. Neurops~~llologiu 27, 1289-1300, 1989. 3. DIMONL),S. J., FARRINCTON, L. and JOHYSON. P. Differing emotional response from right and left hemispheres. Nature 261, 69&692. 1976. 4. HULLER. W. and LEVY. J. Perception and expression of emotion in right-handers and left-handers. Npurop.s?“‘h”logi(( 19, 263 272. 198 1. 5. JAEGER. J.. BOROU, J. and PESELOW, E. Depressed patients have atypical hemispace responses in the perception of emotional chimeric faces. J. Abnorm. Ps~ckol. 96, 321 324. 19X7. 6. KOLH. B., MILNEK. B. and TAYLOK, L. Perception of faces by patients with localized cortical excisions. c‘cm. .I. Ps~~c~/Io/.37, X 18. 1983. 7. LEVY. J., HCLLEK, W., BANICH. M. T. and BUKTOX, L. A. Asymmetry of perception in free viewing of chimeric faces. Bruin C‘o
I?. RLL TI-K-LOKLNT. P. and DA\‘IIXOI*~.R. J. Differential contribution of the two cerehrxl hemisphcrrs to the perception of happy and sad faces. ,V~,u~[,p.slc,h~j/r,Hio19, 609 613. 198 I 13. SCHWAKTZ. M. and SMITH. M. L. Visual asymmetries with chimeric faces. hll’~~~(~p.\~.~~h~~l~~~/i~~ 18, 103 106. 19X0. 14. SLWKI, M. and MCKLTVFK. W. F. Diffeiential right hemispheric memory a&age oiemotional and nonemotional faces. Nru~ops~~holoclin 15, 757 76X. 1977.
NOTE EQUIVALENT PERCEPTUAL ASYMMETRIES FOR FREE VIEWING OF POSITIVE AND NEGATIVE EMOTIONAL EXPRESSIONS IN CHIMERIC FACES STEPHEN
D. CHRISTMAN* and MICHELLE D. HACKWORTH
University
of Toledo,
Toledo,
(Receilaed 21 July 1992; accepted
OH 43606, U.S.A. 21 January 1993)
Abstract-Research employing chimeric stimuli (in which smiling and neutral half-faces are paired) has demonstrated greater influence of the left half-face in determining perceived intensity of expression. To date, no studies have examined how emotional expressions other than happiness are perceived in this format. Right-handed subjects viewed chimeric faces depicting both positive (happiness, pleasant surprise) and negative (sadness, anger) emotions in a free vision task. Results indicated a left half-face bias for all four emotions, supporting the hypothesis of a greater right hemisphere role in emotional perception. The lack of differences in strength of left half-face bias as a function of the specific emotion depicted suggests that results obtained with typical chimeric half-face paradigms can be generalized to emotions other than happiness.
INTRODUCTION RESEARCH IN THE AKEA of brain lateralization generally indicates a predominant right hemisphere role in the perception of emotion [7,8, 141. One of the more recent and innovative methods to address this issue involves the use of chimeric faces in a free-vision task. The most common type of chimeric face stimulus combines an emotive half-face with a neutral half-face to produce a composite which is viewed by subjects under free viewing conditions. When subjects are asked to compare the intensity of emotional expression for a pair of chimeric stimuli which are mirror-images of one another (thus differing only with respect to the side of the face in which the emotion is shown), the result is most often a left half-face bias, presumably reflecting a right hemisphere response [7]. To date, virtually all free-vision studies utilizing chimeric face stimuli have employed only positive, happy halffaces paired with neutral half-faces. Thus, the perception of negative emotions in the context of chimeric faces has been relatively unexplored. Those studies that have employed emotional expressions with negative valences have exclusively paired happy and sad half-faces together simultaneously [2, IO], thus precluding independent examination of the roles of positive vs negative valence. Evidence suggests that emotional valence may influence the degree or direction of lateralization of emotional processing, with research indicating that the left vs right hemispheres may be specialized for positive vs negative emotions, respectively [I, 3, 123. This raises the possibility that findings obtained with positive emotions may not generalize to negative ones; for example, if negative emotions are more strongly represented in the right hemisphere, then the strength of the left half-face bias may be stronger for negative than positive emotional expressions. The present study extends the free-vision chimeric method by employing both positive and negative emotional expressions to pursue possible hemispheric differences with regard to emotional valence and to assess the generalizability of the results obtained with chimeric face stimuli depicting positive emotions.
*Address for correspondence: OH 43606, U.S.A.
S. Christman,
Department
621
of Psychology,
University
of Toledo,
Toledo,
621
NOTI
METHOD
Four poscra. I*() malt and two female. were photographed, each dijplaqing a neutral cxprcsslon and four emotional expressions (happy, pleasant surprise, bad and angry). I-or each poacr. prints were made depicting each crpresion in tither a normal or mirror-revused orientation. F-allowing the procedure of HruIx and LI L’Y [4]. the photographs were cut down the midllne and recombined to make tuo diflicrcnt types of chimera\: ln one, the emotional expression \\as present on the left halfofthe poser’s face and the neutral exprcwon on the right half: in the other. thir was rc\cr\ed. For any given stimulus pair. the emoti\c half-face in both mcmhers of the pair had been produced by one side of the poser’s fact and the neutral esprcwon by the other. with each side of a poser’\ face occurring equally often a\ the left and right half-facea. This \~a?,done to control for poaihlo facial as>mmctries of the posers. A total of32 pairs wcrc constructed. )xlded hq tho faclorial comhinatiun offcour posers h> four emutlvns h! two vc‘rslons (half-face from poser‘s left 1s Iright) Each of the 32 chimcric pairs ucrc arranged \crtlcallq and ph~)tocopied to prclduce hooklcts L\ hich were administered individually to subjects. The order of ~qpe\ of emotion. poser idcntit,, and position ofchimeric stimulus with emotive left half-face (ahow or belw ) %a:,Irandomwed in the hooklct.
Small groups ol’ubjccts (.\ =4 8) \bcrc arwnhlcd in a classroom. Each was plvcn a stimuli hoohlct containing ;I handcdncs~ questionnaire. 32 pages ofchmleric stmul~ pairs. and ;I mood scale. Suhyxth were asked to examine each pair of chlmeric stimuli and to dctcrmlne which face loohcd more emotive (c.g. “Which fact looks happier’! Anpricr’l Sadder’? More surprised’!“). Thq were allowed to give a “can’t decide” response ifthe) feh the IWO faces wcrc cwcll~ equal in intensity ofemotional expression. Left-biased (right-biased) responses uere those In uhich the chimera with the emotion to the wewcr’s left (right) was Judged by the \uhjccts ah being more emotive. blood ~a\ evaluated using :I procedure developed by I>,\\II) L2] in which suhjccts marked a pwtlon wrrcspondlng to thclr prexnt mood on ;I 10 m sale Inbclcd “mo\t deprcsscd au” at fhc left end (scored as /cro) and “mo\l happy c\er” at the extreme right (acored a 10). Intermedlatc poaltion\ wrc wored :,b the distance 1111 cm I from the left end of the talc. The midpoint \\a\ not Indicated on the wale.. Thih mood scale ~a\ uxd to ewminc an) powhle relationb betv,een >ublect\ mood and their rlpht or left blah 111the judgcmcnt of em~>tlonal intcnalt!
RESULTS The Iateral bus scow wcrc the number of chlmcrlc pa”‘\ for each cmotion m which ;I leftward hlas wab shown minus the number of pairs 1n which a rightward hias v,as bhwn. divided by the total number of pair\ for cxh emotion (,v= X). Posltlve \aIues indxatea left half-fact hiasand ncgativr \alucs indicate;, righf half-face hIa\. Thcsc w~rcs are presented in Table I. Preliminary analyses induted no maln cffccts or intcractlvns involving gender. 40 ~uhsequent analyses collapsed xro\s this variable A one-!\a!, wlthin buhyxth .4NOVA \\a~ used lo ewmlnc an> dlfferences in nature of the half-face bins hetuecn the four cmotions. No G_eniticant dllTercnccs uure found LF< I 1. f-test5conducted IO contirm the prescncc of a half-fact lxa (i.e. a non-xro scow) yielded signllicant left half-face buses for all four cmotlons: happy [I (571-3.17. P
Mean bias wore C‘orrelatlon with mood score
0.2 IO 0.30
0.145
0.154
0.096
0.039
0.145
-o.oxx
NWk
623
DISCUSSION Results indicated significant and equivalent left half-face biases for emotional expressions of both positive and negative valence. in the free viewing ofchimeric stimuli, indicating that emotional valence is not an important factor in free vision chimertc face tasks. The general nature of these results is consistent with previous studies showing a left bias/right hemisphere specialization for the perception of positive emotions 17.8, 131. More important. the current study extends these findings to emotions other than happiness. demonstrating that previous findings from studies employing chimeric stimuli composed of smiling and neutral half-faces can be generalized to other emotions of both positive and negative valence. The correlational data indicated a negative correlation between subjects’mood and the magnitude of the left bias judgcment of happy chimerlc faces. These results may be compared with the findings from DAVID’S study [2] reporting a positive correlation between subjects’ mood and their choice of facial affect; namely, the more positive the mood, the greater the tendency to label the ambiguous chimeric faces as “happy”. However, the magnitude of the perceptual bias was not affected by subjects’ mood. The present study, on the other hand, indicated that subjects reporting a more negative mood exhibited a stronger left bias in the judgement of happy chimeric faces. This result is in contrast to results from subjects suffering from clinical depression. who exhibit a waker left bias [S]. This discrepancy may derive from basic differences in the neurological substrates ofclinical depression vs the overall less beverc negative moods exhiblted by the normal subjects in the current study. Our results could also possibly reflect a relation between negative mood and some form of greater right hemisphere activation, although its selective occurrence with only happy faces suggests either a more tonic response to the perception of positive emotion or the possibility of a Type I error. Overall, the lindings from this study suggest that the almost exclusive use of happy emotional expressions in chimeric free-vision tasks is not problematic. Emotional valence appears to not influence the perception ofchimeric faces. Thus, the happy chimeric faces used in most previous studies serve as valid indicators of lateralization. regardless of the limited range of emotional valence used. Indeed. there is evidence that at least part of the basis for left-sided bias in the perception ofchimeric face stimuli derives from factors unrelated to the processing of emotional or even facial information, as other studies have demonstrated the presence of left-sided biases in judgements of f&al identity and in spatial judgements of nonface chimeric stimuli [6.9]. The lack of effects of type of emotion in the current study aupport the notion that the left bias obtained with chimeric face stimuli is driven not by emotional processes per se, but rather by more basic perceptual factors associated with spatial processing.
A[.~,lr,\~/rd~rmrnrs -This research was supported in part by an Academic Challenge Grant to the Department of Psychology at the University of Toledo. Portions of this paper are based on a Senior Honor’s Thesis conducted by the second author. The authors gratefully acknowledge the assistance of Hongkeun Kim in the preparation of the stimulus booklet.
REFERENCES I AHFKN, G. L. and SCHWAKTZ, G. E. Differential lateralization for positive and negative emotion in the human brain: EEG spectral analysis. Neuropsycholoyia 23, 745-755, 1985. 2. DAVID, A. S. Perceptual asymmetry for happy-sad chimeric faces: Effects of mood. Neurops~~llologiu 27, 1289-1300, 1989. 3. DIMONL),S. J., FARRINCTON, L. and JOHYSON. P. Differing emotional response from right and left hemispheres. Nature 261, 69&692. 1976. 4. HULLER. W. and LEVY. J. Perception and expression of emotion in right-handers and left-handers. Npurop.s?“‘h”logi(( 19, 263 272. 198 1. 5. JAEGER. J.. BOROU, J. and PESELOW, E. Depressed patients have atypical hemispace responses in the perception of emotional chimeric faces. J. Abnorm. Ps~ckol. 96, 321 324. 19X7. 6. KOLH. B., MILNEK. B. and TAYLOK, L. Perception of faces by patients with localized cortical excisions. c‘cm. .I. Ps~~c~/Io/.37, X 18. 1983. 7. LEVY. J., HCLLEK, W., BANICH. M. T. and BUKTOX, L. A. Asymmetry of perception in free viewing of chimeric faces. Bruin C‘o
I?. RLL TI-K-LOKLNT. P. and DA\‘IIXOI*~.R. J. Differential contribution of the two cerehrxl hemisphcrrs to the perception of happy and sad faces. ,V~,u~[,p.slc,h~j/r,Hio19, 609 613. 198 I 13. SCHWAKTZ. M. and SMITH. M. L. Visual asymmetries with chimeric faces. hll’~~~(~p.\~.~~h~~l~~~/i~~ 18, 103 106. 19X0. 14. SLWKI, M. and MCKLTVFK. W. F. Diffeiential right hemispheric memory a&age oiemotional and nonemotional faces. Nru~ops~~holoclin 15, 757 76X. 1977.