9
AESTHETIC PREFERENCE AND PICTURE ASYMMETRIES Marie T. Bsnlcht, Wendy Hellets, and Jerre Levys (Dept. of Psychology, University oI lllinois ar Urlana;Qhampaign; 2Depr' of - Piychiatry, universily of chicag6: tDept. of Behavioral Sciences, University of Chicago) IhITRoDUcTToN Aesthetic preference appears to be one of the nany aspecs of cognitive and emotional preessing that is influenced by asymmetric organization of the brain. Levy (1976) reported that when a group of right-handers preferred one orienta- tionof a compiex picture over its mirror image, the center of interest wasjudged by another gtoup of right-hand€rs to be displaced to th9 right. In contrast, pictutes thai right-handers preferred equally well in both orientations were judged to have the center of interest displaced to the left. Suggesting that asym- meiry of hemispheric functioning plays a role in producing such effects, orien- tation preference of left-handers was unrelated to asymmetry of picture con- t€nt. It is important to consider, however, that other aspects of composition besides lateral location of content may influence the overall aesthetic impact of a picture. Additional structurat components of a picture are also likely to affect judgement of aesthetic value, and may conceivably be asymmetric (e.g. a road in a scene might meander off to the teft or meander off to the tighQ. In fact, Freimuth and Wapner (19?9) found that pictures with implied motion from left-to-right a1e preferred to those with implied motion from right-to-left. It seems possible, therefore, that brain organization might affect how such asymmetric form var- iables influence aesthetic judgement. If so, then overall aesthetic appeal of a picture would be influenced not only by perceived asymmetry of content but also Ly perceived asymmetry of other aspects of composition, and preferred _oriel- titibn may depJnd on the extent to which a subject's choice is influenced by the various aipe"ls of asymmetric information. The effect of these other aspecls might dilute or enhance the influence of asymmetry of content on a subject's orientation preference. A lack oi concordance between different aspects of asymmetry may explain some of the contradictory results reportd since Levy's original paper' Freimuth and Wapner (19?9), foi example, found no relation between the orientation preferences of right-handers and the lateral asymmetry of content in 8 paintings' inA in one of their two studies, Mclaughlin, Dean and Stanley (1983) found that picture asymmetry did not predict preference. In both cases, the researchers preselected slides solely on the basis of asyfiunetry of content and then investi- gated whether one orientation was preferred over its mirror image. Other aspects Cortex, (1989) ?5, 187-195

Aesthetic Preference and Picture Asymmetries

Embed Size (px)

Citation preview

AESTHETIC PREFERENCE AND PICTURE ASYMMETRIES

Marie T. Bsnlcht, Wendy Hellets, and Jerre Levys

(Dept. of Psychology, University oI lllinois ar Urlana;Qhampaign; 2Depr' of -Piychiatry, universily of chicag6: tDept. of Behavioral Sciences, University of

Chicago)

IhITRoDUcTToN

Aesthetic preference appears to be one of the nany aspecs of cognitive and

emotional preessing that is influenced by asymmetric organization of the brain.Levy (1976) reported that when a group of right-handers preferred one orienta-tionof a compiex picture over its mirror image, the center of interest wasjudgedby another gtoup of right-hand€rs to be displaced to th9 right. In contrast,pictutes thai right-handers preferred equally well in both orientations werejudged to have the center of interest displaced to the left. Suggesting that asym-

meiry of hemispheric functioning plays a role in producing such effects, orien-tation preference of left-handers was unrelated to asymmetry of picture con-t€nt.

It is important to consider, however, that other aspects of composition besides

lateral location of content may influence the overall aesthetic impact of a picture.Additional structurat components of a picture are also likely to affect judgement

of aesthetic value, and may conceivably be asymmetric (e.g. a road in a scene

might meander off to the teft or meander off to the tighQ. In fact, Freimuth andWapner (19?9) found that pictures with implied motion from left-to-right a1e

preferred to those with implied motion from right-to-left. It seems possible,

therefore, that brain organization might affect how such asymmetric form var-iables influence aesthetic judgement. If so, then overall aesthetic appeal of apicture would be influenced not only by perceived asymmetry of content but also

Ly perceived asymmetry of other aspects of composition, and preferred _oriel-titibn may depJnd on the extent to which a subject's choice is influenced by thevarious aipe"ls of asymmetric information. The effect of these other aspecls

might dilute or enhance the influence of asymmetry of content on a subject's

orientation preference.A lack oi concordance between different aspects of asymmetry may explain

some of the contradictory results reportd since Levy's original paper' Freimuthand Wapner (19?9), foi example, found no relation between the orientationpreferences of right-handers and the lateral asymmetry of content in 8 paintings'inA in one of their two studies, Mclaughlin, Dean and Stanley (1983) found thatpicture asymmetry did not predict preference. In both cases, the researchers

preselected slides solely on the basis of asyfiunetry of content and then investi-gated whether one orientation was preferred over its mirror image. Other aspects

Cortex, (1989) ?5, 187-195

188 Marie T. Banich Wendy Heller, and Jerre L*y

of asymmetry in the pictures were ignored, which could account for their weakeror null results. In contrast, Lely selected slides that right-handers significantlypreferred in one direction, a method which was likely to have isolated thosepictures in which various asymmetries, including content asymmetry, were rela-tively concordant in their influence on preference. Beaumont (1985), whoobtained results similar to Levy's, did not preselect his stimuli for agreement oforientation preference among right-handers, but his simple stimuli consisted onlyof one large and one small object. They varid, therefore, along only one asym-metric dimension, ccntent asymmetry. No other asymmetries were present thatcould have masked or overridden the effects of content asymmetry on prefer-ence-

In the present study we examined how preferencejudgements were influencedby two aspects of composition, asymmetric content and asymmetric motion(left-to-right, right-to-left)- Based on previous findings we predicted that thereshould be a strong relation between orientation preference and content asym-metry for pictures that right-handers significantly prefer in one orientation. Forthese pictures, we also predicted that different kinds of asymmetric informationwould be relatively concordant, thus eliminating or reducing conflicts betweenthe effects of the various asymmetries. How€ver, when slides are selected on thebasis of asymmetric content alone, the relationship between asymmetry andorientation preference should be weak, because different aspects of asymmetry ina slide may not be concordant. Furthermore, we predicted that the influence ofboth these asymmetric aspects of composition would be related to lateral brainorganization. In particular, we predicted that right-handers' preference for apicture in a particular orientation would be related to both perceived asymmetryof content and perceived direction of motion, but that left-handers' preferencewould not. To investigate these predictions we isolated (a) a subset of picturesthat right-handers significantly preferred in one orientation, and (b) a subset ofpictures selected solely on the basis of asymmetric content. In general, weexpected that different aspects of asymmetric composition would combine toinfluence the subject's overall aesthetic assessment, and that lateral brain organ-ization, as indexed by handedness, would influence the orientation preferencethat results.

MerEnlers er.lo Mrrgoo

Stimuli

The stimuli were 100 vacation slides of scenic vistas from around the world, onepresented in the original orientation and one presented in the mirror orientation. Eachslide had been previously rated for asymmetry of content [I) more interesting content onthe right, 2) more interesting content on the left, 3; symmetricl by 32 right-handed judges,and for asymmetry of motion [ ] right-to-left. ?) left-to-right, 3) symmetricl by 36 differentright-handed judges.

Using Levy's {1976) method. subjects were asked to make the content judgement bydetermining on which side the more interesting or important content was located. or bydetermining which side of the slide was "heavier". Subjects were also asked to decidewhether important objeca themselves, or the relation or arrangemenr of objects or con-stituents of objecs appeared to move right-to-left. left-to-right or neither Instructionswere as follows: "In this task we want you to determine whether the predominant direction

Aestlvtic pre!*ence ad pictute asymneties 189

of movem€.nt or implied move$€nt in each slide is.from.left-to-right or-from right-to-lelt.d*;"y 6JingJttis i* Uy tooking at the direction in which imporiant objects app.li t9 !?1noui"i1 nriottrEr way to-j"udt; rhfi ii Uy tooting at the relation' between6bjects or parrs ofoiliecfi;Aa;i;;mtii"sii tfiev sem t6 move ii a particular direction. For example, lines

ot"tt " *"t"t misht suc;est waies moving from left-to-right. Sometimes there may see-m to

be Darts of the Etiae diLt are moving from left-to-right and others that are moung lromrinf[t:ti:f"it.-lt v* Oi"t that thesitwo directions-of movement are equally dominantciftll r"ittrc", but if at all possible ry to decide between the two."

Til "otent "ry**"try-rcoiJfot

eah slide was rhe number of sgb.lqts 1n9juf g1f lfeslide to have heavier con6nt on the right minus the number of subjects who.luclgecl lleiiia" i" tiave treauicr *rrt*t on tlre leit, divided by the total number of ju-dges (N=32)IA=(R-L) /321t-OI th€ l00slides,4gwerejudged tobeasymmetric_(q<.qt'two-tarled)'iA ad btt-biaseil and 2l as right-biased. We designated thcse 49 slides the Slrdes wrth

Asvmmetric Content.-Tlrii gioup of slides is similar to the set of stimuli used byi\d;;;hfi

"t ir"?i-gbsi

"oa-riiih"th and Wapner ( 1979). Each slide wry qlso assigned

;-;ffi;tti;*iiioni6.!, which was the numlber oi subjects whoju{ged the motion to

il;fiil[I miioi irre o"inuet"'t ojudged it to be left-t&right ovir the total number ofjudgEs [(R to L)-(L to R)/35].

APParatus

Two Kodak Ektagraphic IIIA carousel prd_etors, onelocated above the other, were

urd l" Aiptay tnJ iftOis. One projector w-as iesignated A, the other B.

Pruedure

One coov of a slide was placed in tray A and one copy in tray B' Slide placemenl

tr"Vi-*"r li"aotn "xc"pt

f6r the cotrsdaint that hatf ihe slides in each tray weretnln

s. In tesring whetlrcr there is a significanl prcferencc

for a slidc ii a iartic-glar oricira-tioo ovcr irs mirtor irnaga rhc null hypothcsis $t"tP th"i.qf 1y.P1_9-l::.S.T:T"r"f* rhc sti tc'in thc oricntatfun s'itb riStrt-birscd contenl is cquivdcnt to thc numbcr o-IJudgcs w-Do Preler ur€ susc

l" t|rcori*rs6o"wirh bft-bias€dconrcnr Thus,oneis rcsring: lL: (Nr- Nr)/Nr-+ L=O'whcr€ Nr rs trl€ DumDeroI

iJec" orcrcrrim rhc ricbt-bi;;;rdt"d*;d x"i" ttt" i'uo6"ioiluagEprefening the left-biased orientation'

###tffi,ffi ffi*mn*ql*rys*y,til##;F,ftfi'til's*$ffiffi;'ift t;rtdffi-r* iliiJt-Gi."a "qiqqtim

Tluq I 12 Nx-would have bccn scorpd 1s tigttt'biasef

#rffiW,*r'ffi 'yrfi f,ilfis'J:*a'ip'*ffi ffi ci

fii;;ii"-;r"ft'db*rrlftii.t-it*i,L-rffiia*r.lrirtreaumuiibtotscrvea rior'biasedchoice.s) + I /2

N' (hss rlc numbcr ';f:m;'Iffi;;'-1i,G' ii;-;;ii -!'vp"q*it -predicts thqt ltll..+

lrhiifr+u'.tsl&;#*ltl]l*"t1:#:til,*r#tffi 'ffiS#1ifiitt[,,,+'it

i"C*rqffi ffiffi#ffi#;;;fraue r-rr"-Fi*i. Onl *asiuigcd as rlght-biasod bfooe viarer. as lcft'biascd by 2 vi

;;;.-rb"-;ih*;*i"isfr; t ri$;; uy. rs 'i"*un .s 'igr-uig1%t115y*-"1.a *'v-qfl'9.qv.91i

&-a1.sYifi 'Hff#ffiHffiffi f..#"rg.:ulr,,.r,#lt'J#,Tfi iHvicwcd thc sli& as alil,DMl-ii*""ariiileilf f f= -Ofj'f*5gr.conitstiOc. However,underihisscoringsysterl thcfirstslide.wtuch tne vast""ffiffibT;fi;fi""i-;-,ffi;E;;ortd b"*"sdd"d rnon asymrnei--ri6 than thesocond stidc. obviouslv'

iffi#'Jaftil.dfifl *r tfiy-rcni.rini rle pcrccprion of rhc sli<[: by.themajoritv of viewers.*-i;;l;;"riii"*".plr,ii,oidi;iltiail"ae:Aurro-viciliiasrighr-ftascd-.{ouizzasut-uiascd.undcra

scoring syscm rhEr ooly r"&iy*iitriii.lE"i#sinto aacounrlthe score-for this slide would be -'313Il}-?2\tS2lrphich is approilm,aic$y cqual ui rd scorc ttat *o,rtO be obtain€d for the firsr sli& in thc P.r?uiousexamotc- - -33310-21/31. fi;"riir; rt"{ r"ft*tid suU.locts; perception bccausc although rhis latter slide rras

il&?i;*)ffiil;i;iih;ffidry"f ii;"*iiiist-"*fics"*asyinmerryscorcasthefirstslidein theprevious

;:#*mfu ;mflffi r*x"g,r'Ufi f'T5:fi &m:;"'Wffi 3#"itr'iii*'igf;of - .3?5 t{ lg-22llrzf ioaioiils l*i',r* irai'*ii-i,;oF ro u" *oo

""ym."tric rf,an rhc stidc with I right-biascd'

i'rcrilui]"1!-"lf idiatG-iuo'gi"ttts 'wtrictr

would -have rcccivcd a score of - '032'

t90 Marie T. Banich, Wendy Heller, and Jere L*y

the original orientation ad half were in the nirror orientation. Psition of the prqiectors(A-top, Fbottom; A-bottom, Ftop) was counterbalanced across subjects. Subjects wereseatd in central areas of the room. Preliminary analyses comparing subjects seated to theleft of picture midline to subjects seated to the right renealcd no differences between thetwo groups, so they we,re subsquently combined.

Subjects, in groups of approximately 6 to 10, were shown each of the 100 pairs of slidesfor l0 seconds per pair. They were asked tojudge whether they preferred the top, bottom,or neither slide by marking the column labelled T, B, or N on their answer sheet.

Subjects

Thiily-nine right-handed {19 female, 20 male) and 19 lefi-handd (9 female l0 male)individuals served as subjects, none of whom made asymmetric motion or asymmetriccontent judg€ments. Handedness was determined by seU report and by the exlrcrimenter'sobservation of which hand the subject used for signing his/her name. All subjects wereuniversity studenis who were paid for their participation.

Elimination of Subjects with a Position Eias

Using a chi-square test with one degree of fredom, we examined whether each subjecthad a bias to choose either the top or botton slide, which would indicate that a subject'schoices were dependent on a slide's position rather than content. These analyses revealedthat 9 of the 3&right-handed subjects had a significant position bias (p<.05) (7 for thebo*om position,2 for the top position) as did 2 of the 19left-handed subjects (both for thebottom position)- The data from these ll subjects were therefore excluded from subse-quent analyses, reducing the right-handed group to 29 subj*ts (12 female, l? male) andthe left-handed group to l7 subjects (8 female, 9 male).

Selection of Slides for Which Right-handers Have a Signiticant Orientation Preference

Of the 100 slides, there were I I for which one orientation was significantly preferredover the other (p<.05). This set will be designated the Slides with a Preferred Orientalionto distinguish them frorn the set of 49 Slides with Asymmetric Content. The proportion ofslides for which one orientation was significantly preferred over the other, I I of 100, isquite similar to that reported by Levy (1976) in her origind study, in which she found 14 of97 slides to be significantly preferred in one orientation over the other. Of the I I Slideswith a Preferred Orientation, 3 were also members <rf the set of the 49 Slides withAsymmetric Content.

Scoring Procedures

For each pair of slides the subject was assigned two scores, one for content asymmetryand one for motion asymmetry. For both aspects of asynmetry, scores for a slide reflecteda) whether the subject preferril tbe right-biised version, the leit-biased version, or neither,and b) the degree of asymmetry in that slide. Thus, for scoring of both content and motionasymmetry, a subject's choice of preferred slide for each pair was recoded so that when asubject prefened the right-biased version, the score for that slide was positive, and when asubject preferred the left-biased version, the score for thac slide was negative. If the subjectindicated no preference for one orierrtation over the other, a score of zero was assigned. Fordegree of content asymmetry, the value assigned to a slide was the content asymmetryscore for that slide as previously determined by the 32 judges [(R-L)/321. Likewise, fordegree of motion asymmetry, the value assigned to a slide was the motion asymm€try scorefor that slide previously determined by the set of 36 otherjudges [(R to L)-(L to R)/36].Therefore, the sign of this value (positive, negative) was determined by the version of theslide (right-biased, left-biased) that the subject preferred, and the value reflected thedegree of asymmetry in the slide as previously determined by a separate set of judges.

Aesthaie prefererce and piuwe asynmaries t9l

If our hypothesis is corret,slides with a Preferred Orientation sho-uld be significantly

riehr-biased for conten;;;;'iibi;; dlistlniOitection of mgti'n' However, for Slides

;f;t' /i;ffi;;;e;;il;;h;;i;;ft-ii uttt" o' "o

relationship betwen asvmmetrv of

content-and direction of motion-

Stides with a Preferred Orientation

,,:f !:K#,fl l?,J,iilJi;fi ul#-E!3.'i',#]-T;:tr'3,[iffi ',ru"Xt"riirffi,;h;;;; ;ffi;;;;;i;;rdir'g'..' iieieffea ii, its miror lmage by a sigrrificant

;;j"hi; ;l tight-handefr, p*t"ii*A stdei are judqd to have,the important content

afu.";d to rfie righl-F;; l'"1;:[;;?"d-*bjit ,'ptJo"nce on the whole appears to.be

iiiiffi,: l.e6;a":i;;;; il;i"ileol (s"" riule r). rhys, left-handers do not judge

slides that right-handil;di#;,ty |igferi" o"J"ti6"t"tion as having content that is

either right'or lef t'biasedAs anorher *"y o?E *ioing.lhe rglation betrvgen preference -and asymmegV for

rert-r,l'aor, *" iJ"raiJ'nore fiia* th+ ieft-handers.'sigrificantlv oreferred in one

orienrationover*r"orrr]. i;;A#ilp;Ti*ii-6;eiittteritn"t"oniv3 of the 100 slides

for which rhey showd "

**irlii,to-tiJntution ;;l;;;, the proporti"l. thll ol1Y.1lld;;;;r, bt;i;&;-. Tili, i"eg;;thut lcft-handers' oreference with respect to asvmmetry rs

random. Furthernore' ,#fr;; *;;;i: 'b57jfor fuiitt s slides did n6t significantlv differ

from zero.

;h,'#Jsil1#l3l;}**x*ffi ffi till;llLl#::1,"f"1?ff tllffi 1il"ff j:sn*:reiults confirm those of Levy (1976).

Judgements of Motion. ]rye predict$ rhat for stides with a preferred orientation, the

f, fi'""f s"'*#txl*:,*tlxl*l*nx"*l'mx''r"'hi6!:lfis$:lllliie oieferted orieniation, siides were judged to have a--significant asvmmetry ol motron

i.#';'fiiT*i;i;'rTlr.itii.i.i+:-s,:;:.00t il;;6EJo *nti"'t, scores. of reft-

h*i"*-aia *r ribtifii*'tfi-dUfii'fion tetir lt=1.81, i.f.=16, n.s-, two-tailed) (see

TABLE I

Average pretoence tor ight vs.

MeanStandard deviarion

Ie[t asvmtnetric contmrt, l2 I *r.r .062

.l3ll7

.036

.082l7

NAverage prelerence lor right+oJeft vs.

MeanStandard deviationN

.t2229

Ieft+o-right usymmetry of motion$- .054..*.083

29

*' StqnificanttY diffsent from zcro, p<'002. tw+tailed'

"t'-!fiiiic."itv-orierenr from zero. p<'001, one tailcd'

i{oositive vatucs indicatc prcference for tbe orienution with right-biascd cont€lli and ncgativc values indicate

;ff;;; for thc oricntadon wirh kft-bias€d content)'

Vmsitivevalucsindiearcprclcrctcefororcorienradorrwithright.tolcfimotionaodncaativcvaluesindicatc;iU;cnce for thc oricntation rvith hfi+o'riglrt croti'on)'

Table I) indicaring that they did not sigruficantly- prefer the orientation that had been

xdn*l*i"t*::*-*'r':l$;,ff L1""*m:f;**',";1"1*1':l'ff lifl **aesthetic Preference.

t92 Morie T. Banich Wendy Heller, and Jerre l*ry

TABLE II

Sli&s with Asymmetic Conrent

To investigate the relation between asymmetry of content and asymmetry of motion,we examined scores for preference of asymmetry 6f content and asynnmetry o? motion forthe I I slides with a preferred orientation. Of these I I slides, 7 wire iudied to be riehr-biased and ro have righr-to-left motion, I was considered right-biasdt uXrh left-to-ri'shrmotion, I was considered left-biased with right-to-left moiion and I was consideiedleft-biased with left-to-right motion (the remaining slide had no asymmerry of csntent andleft-to-right mgtion). Thus. the majority of slides Gth a preferred orientation were likely tobe judged both right-biased in cohtent and right-biase,a in morion when viewed in deirpreferred orientation. Furthermore, 8 of the l0 slides were concordant for asymmetry ofcontent and asymmetry of motion, which just barely misses significance usihg a tesi ofbinomial probability (p =.055).

When the individual subjects' scores for preference of asymmetry of content andasymm€try of motion were correlated, the results were highly sigirificant (r:.486,d.f .=/J,.p<.01).Thus, thedegree to which a subject prefers the version with righr-biasedcontent !s a vqry good_predictor of the degree to which he preferred the version-judged tohave right-to-left motion, indicating that slides selecred for a preferred orientaiiori-werehrghly concordant in these two aspects of asymmetric informladon.

Slides with Asymmetry Content

. Judgements of Content. For the 49 Slides with Asymmetric Content, the mean score ofrrgh!-handeA sdbjects wassignificantly positive(r= 2.01,d.t.=28, p<.05, one-tailed)(SeeTab-le Il). However, a paired t-test indicated that for right-handers, rhe magnitude ofpreference for slides with right-biased content is considerably larger for the Slides with aPreferred Orientation than for Slides with Asymmetric Conient (t=4.77, d.f.=28,p<.001, one-tailed)., For left-handers, the mean score did not significantly differ from zero (t=-.561,

d.f. = 16. n.s.,two-tailed) (See Table lI). This indicates thai rightward bias in a slide doesnot predict preference for left-handed subjects. The randomness of left-handers'prefer'ence in relation to contenl asymmetry appean to be due to the great diversity of resionses,as evidenced by the fact rhat berween-subject variance for prefdrence scores

-was more than

eight times greater for left- than righr-handers (F (16,28)=8.23, p<.0001).

Judgements ol Motion. Although analyses revealed that slides selected for asymmetriccontent were viewed by right-handers as having a consistent direction of motion(t=2.08,d.f.=28, p<.05, two-tailed) (see Table II), tf,is effect was relatively weak and, again,

Right-banders Left-handers

Average prelerence for right vs. lelt asymmetric contenttMean .025r'Standard deviation .067N29Average preference lor riglu-ro left vs. lelt-ro-right asymmerry of motion$Mean -

.017.Standard deviarionN

*.026.l9l

17

.003

.05r17

.0t0429

' Significanlly differenr frorn zero. p<.05. r*o-raild." Significantly differenr from zcro, p<.05. one-railed.

f(positive v-alues indicatc preference for the orierrtation wirh righr-biascd conrent and ncgative vatr.res indicatepreference for the orieltation with lefi-biased content)

$(positive v-alues indicate prcference lor the orientation rrritb right-to-ldt rnotion and negative values indicateprefcrence for the orientarion with left-ro-righr motion)

Aesthetic pref*ence ond picture asymmeties 193

sienificantly less than for slides selected for a preferred orientation (t=2.21, d.f -=29,o?.OZS, one-railed). Thus, for slides selccted for asymmetric content, right-handery thgwLnlv a very weak pieference for slides with right-to-left motion, which contrasts with thettiiirt" sis.iificanr'ric,ht-to-left motion in slidEs selected for orientation preference. Left-frafiOirs;-preferencei*ere random with respect to asytnmetry of motion (t= '24, d.f. = 16,

n.s., two.taild) (See Table II).---- More impo|intly, there was no correlstion between a subject's.preference for rhe

orienraiion <if a sUOiludged to have right-biased content and theorientation ju{ged to -

[a"" *[tr1-t*teft *otiottj r - - .257, n.i.1. Thus, for-slides selected for asymmetric con-

tent, aierage content asymmetry was unrelated to th€ av€rage mo.tion.asymmetry i.v,e,lthoush. for-rieht-handeis as a gioup, there was a weak preference{or right-biased slldes

ft fffrliA;i,iitr rigt i-to-t"ft-motib'n. This conlraslr yith the significani concordance of;th;-i;Gitmoiion ind right-biased content in the I I slides selecied for having a preferred

orientation-

DtscussloN

Our data show that asynrmetries in picture composition influence the aes-

thetic judgements of rigbi-handers, and that such asyrnmetries are not just

limited to isymmetries oicontent. In particular, our data indicate that slides withright-biased content and right-to-lefi motion are favored by right-hande1s. fncontrast, left-handers'prefeiences appear to be uninlluenced by asymmetries inpicture composition, indicating that brain organization affrcts judgements ofiesthetic valle. Furthermore, the results of this study strongly suPport the pre-

diction that when aspects of asymmetric information in a slide are concordant,subjects prefer that ilide in a particular orientation. When a slide is selected so

ttrai rigtri-tranders prefer it in one orientation over its miror image, asymmetry ofmotioi and asymmetry of content are concordant. In contrast, for slides selected

for content asymmetry only, content asymmetry and motion asymmetry were

unrelate.d, and preferinces of right-handers for right-biased content and forright-to-left motion were weak.-

These results are consistent with the findings of previous studies. When

content is the only dimension that varies asymmetrically (e,g. Beaumont, 1985) orwhen the selection process isolates pictures in which content asymmetry and

motion asynmetry are likely to be concordant (e.g. by selecting only those slides

preferred io r pattict tar oriintation), right-handers show a strong prefeten€e forright-biased siides. However, when complex pictures are selected solely for

"Stn*etry of content, content asynmetry and motion€symmetry may be con-

.ordut t, discordant, or unrelated. Under these conditions, preference may be

only weakly related to content asymmetfy, since direction-of motion may oppose

or attenuaie the influence of content asynrm€try on preference.

The preference for slides with apparent motion frorn right-to-left deserves

co**eoi. First, it appears that left-to-right scanning habits induced by reading

do not induce prefeiince for asymmetry of motion in our slides, beause right-handers preferred slides with right-to-left motion and because the orientationpreference of left-handers was unrelated to asymmetry 9! motion- Second, our

iindittgt differ from those of Freimuth and Wapler {l-979)' who found-thatpreferied slides werejudged to have relative motion from left-to-right. The slides

194 Marie T. Eanich lVendy He&er, and Jerre Lcry

in Freimuth and Wapner's study, however, did not have a significant asymmetryof content. It may be that right-to-left motion is only preferred when there is asignificant asymmetry of content, as in our study.

The preference for slides with right-to-left rnotion may occur because subjectsprefer compositions in which attention is drawn to the center of the picture.Given that right-handers prefer slides with the more interesting content on theright and that such slides induce a rightward gaze deviation (as demonstrated byB€aumont, 1985), motion from right-toJeft might serve to draw a subject'sattention back toward the center, and hence to the rest of the picture. In contrast,when content is right-biased, motion from left-to-right would only serve to drawattention out of the picture frame. In fact, it has been reported (Beaumont,unpublished data) that subjets preferred pictures of horses when they werejumping towards the certer of the picture. If the horse was placed to the right ofc€nter it was preferred when jumping to the left; conversely, if the horse wasplaced to the left of center, it was preferred when jumping to the right.

Confirming others, we found that slides prefered by left-handers, unlikeriglrt-handers, had no asymmetry of content. In addition, we found that slidespreferred by left-handers did not have a consistent asymmetry of motion. Thelack of a significant preference by left-handers for asymmetric content is con-sistent with Levy's findings, but the direction of this non-significant preference,which was for left-biased slides (see Table II), is consistent with the findings ofMclaughlin et al. (1983). As we demonstrated, left-handers are highly hetero-geneous in their preferences, and it may be that between-subject variability is toolarge to obtain a consistent effect, especially when relatively small sample sizesare used.

Previous explanations linking brain organization to aesthetic preference haveconcentrated on a$ymmetry of content. Lely (1975) proposed that the task ofjudging the aesthetic value of a picture differentially activates the right hemi-sphere, which produces an attentional bias towards the left. In consequence,right-asymmetric pictures are preferred, since they cornpensate for the attention-al imbalance and provide for more perceptually balanced pictures. Beaumont(1985) offered the alternative explanation that prefereace for right-biased pic-tures occurs because they induce a riglrtward gaze deviation, allowing more of thepicture material to fall within the left visual field and be projected to the righthemisphere. Our data indicate, however, that right-biased asymmetric contentdoes not ensure preference by right-handers, because slides with asymmetriccontent wcre not strongly preferred in the right-biased direction. Rather, pre-ference for a slide in a particular orientation was assmiatcd not only withright-biased content, but asymmetry of motion from right-to-left as well. Forleft-handers, preference for a slide was associated with neither a particularorientation nor a particular direction of motion.

Although brain organization clearly affected preference for one compositionover another, the effect was not due to the influence of one asyfitmetric factoroperating in isolation. Thus, although the exact mechanism by which brainorganization influences aesthetic judgement remains unknown, an adequateexplanation must take into account the interaction betwe€il the neuropsycholo-

Aesthetic prelerence ad picure asymmetries

gy oftion.

the viewer and at lcast two asymmetric components of picture composi-

ABSTRACT

Previous rcsearch has shown that when righr-handed subjects prefer oneorientadon-of

ffi :ilfi :;"ffi #ei'f *'::r#:"xn'##1ffiil':3;fill:;,iJ"ffi *lil;;;ffifi;ii.* i" a"iiiioi"r Jhether the relationship between aesthetic preferenqe

;ffi ;r;#t ;'nr*"t'y 6 iiraiea bv tl* interaction of dfi f f erent f PTtt- Pf:Y.91:,"information in a su$];ffi;;;;'stuJv-"iumi""a both asy.mmetry of content and

#ffin#rf ffi :';**i'ril*l$**tidrx*:[T.'1"r.*:1ffi :,*t'11two asymn."t*c components Jilici"r" compositi5n interact with brain organization to

infl uerice aesthetic j udgements.

Acknowledgmenrs. We would like to acknowledg-e the supprt of the Soencer Founda'

tion in the form "r "

gt*iiolh; rhia;;r#,;E1rri a*tiitan"e of Lisa strauss in da.ta

collection and analvsi?.Til;; td;"k K"d ir1t Nancy Wagner, two reviewers and the

Aii;; fr; ilkilt fielpful comments <rn earlier drafts of this paper-

RsreRBNcEs

BErUuoNt, J.G. Lateral orgrnization and qp$9rryp^1cferencc: The irrportancc of pcriphcral visuat

r*mryU,*frg-g*9ffi "**til

iiilit" organization in the evaruation or paintings'

,;:l:y,wr"ffi**T"$3.1:'3l:td#.New.qsychotogia,14:43t-tA5,r9?q.trii:"r;"*,-, j.a;ffiil E':;;-$;,ifr;, ii eotr,ai"'piercrefte ia dortrals and sinistrals.

Newapsychotogia, 2 l'- 147'153' 1983'

MarieT.Bsnict\DepsftilcnrofkJlchobs|'UaivasiryofltlinoisatUrbana-ChrrnpaiEr'603F:stDsniclStraer'Champaigr, tL 61820' U.S.A'