"Long before it's in the papers"
June 04, 2013


Science gives beauty some of its mystery back—for now

Dec. 22, 2007
World Science staff

What is phys­i­cal beau­ty, in a per­son?

In re­cent years, much sci­en­tif­ic opin­ion has gath­ered in sup­port of a some­what dreary an­swer: beau­ty is just the av­er­age. 

There’s something to that, stud­ies show. Vol­un­teers gen­er­ally rate “av­er­age” faces, crea­ted by com­pu­ters as composites of many others, as bet­ter-look­ing than un­u­sual faces. Blend­ing even just a few faces—even un­at­trac­tive ones—tends to pro­duce sur­pris­ing im­prove­ments. (Try it your­self here).

(a), a com­pos­ite of 60 faces; (b), of the 15 "most attractive." (Cour­te­sy L. De­Bru­ine et al.)

But if it’s dis­ap­point­ing to think that our fan­tasies cen­ter on a qua­li­ty so, well, av­er­age—take heart. A new study may have re­stored a tou­ch of the old mys­tery that beau­ty once had. 

Psy­chol­o­gists have found what would seem to be a slight but def­i­nite dif­fer­ence be­tween av­er­age faces and the most love­ly, as rat­ed by vol­un­teers.

“There are spe­cif­ic non-av­er­age char­ac­ter­is­tics that are par­tic­u­larly at­trac­tive,” wrote the re­search­ers, de­tail­ing their find­ings the De­cem­ber is­sue of the Jour­nal of Ex­pe­ri­men­tal Psy­chol­o­gy: Hu­man Per­cep­tion and Per­for­mance. “Our re­sults sug­gest that while av­er­age­ness is one com­po­nent of at­trac­tiveness, there is at least one oth­er... not ex­plain­a­ble by av­er­age­ness.”

The study did­n’t delve in­to deeper ques­tions beau­ty rai­ses: for in­s­tance, just what that non-av­er­age com­po­nent could be; and what beau­ty’s ev­o­lu­tion­ary func­tion might be (many sci­en­tists agree it probably helps sig­nal good health, though ex­actly how is un­cer­tain.) 

But the new find­ings did high­light the pos­si­bil­ity that even the lim­it­ed sci­en­tif­ic agreement sur­round­ing such is­sues may be prem­a­ture. A pre­vi­ous stu­dy, for in­stance, sug­gested beau­ty is that which the brain finds easy to pro­cess. But that con­ten­tion de­pended largely on the as­sump­tion that beau­ty is an av­er­age; the brain pre­sumably finds that easy to grasp, be­cause av­er­age is typ­i­cal.

In chal­leng­ing the beau­ty-as-av­er­age­ness hy­poth­e­sis, the new study was­n’t break­ing en­tirely new ground. At least one past study has al­so chal­lenged it. But crit­ics had ques­tioned the con­clu­sions of that work. The new study claimed to shore up some of the weak­nesses that spurred those doubts.

In that 1994 stu­dy, re­search­ers from the Un­ivers­ity of St. An­drews, U.K., av­er­aged 60 im­ages of faces with a com­put­er. They then sep­a­rately av­er­aged the 15 faces that vol­un­teers had judged as best-look­ing of the bunch. It turned out peo­ple rat­ed this smaller av­er­age as hand­som­er than the full-group av­er­age, rais­ing doubts about beau­ty as a sim­ple av­er­age.

But oth­er sci­en­tists raised ques­tions. Could­n’t it be—they asked—that the top 15 faces were in­deed so av­er­age, that their com­bina­t­ion re­flected the av­er­age of the hu­man race even bet­ter than the whole 60? The new­est study at­tempted to set­tle the ques­tion by re­peat­ing the ex­pe­ri­ment, but adding sev­er­al oth­ers to serve as strict tests of the in­i­tial re­sult.

The re­search­ers, Li­sa M. De­Bru­ine at the Un­ivers­ity of Ab­er­deen, Scot­land and col­leagues, again made an av­er­age of 60 faces—all white fe­males—and a sep­a­rate av­er­age of the 15 “pret­ti­est.” View­ers again rat­ed this as more al­lur­ing than the full-group com­pos­ite.

Fur­ther tests, the re­search­ers said, so­li­di­fied the con­clu­sion that av­er­age and gor­geous weren’t quite the same. For ex­am­ple, view­ers them­selves in­de­pend­ent­ly rec­og­nized the larg­er av­er­age as be­ing “more av­er­age” than that of the smaller group, ap­par­ently con­tra­dict­ing the idea that the smaller com­pos­ite might have been the tru­er ul­ti­mate av­er­age.

In anoth­er test, De­Bru­ine and col­leagues sub­jected their hap­less vol­un­teers to a bar­rage of at­trac­tive-face im­ages. Pre­vi­ous stud­ies had found that look­ing at many im­ages of one type, tem­po­rarily skews what view­ers con­sid­ers “av­er­age” to­ward that type. Thus, the re­search­ers rea­soned, if the beau­ty-is-just-av­er­age hy­poth­e­sis is cor­rect, then the vol­ley of win­some im­ages should have per­suaded vol­un­teers to see ad­di­tion­al beau­ti­ful faces as even more at­trac­tive—be­cause they look more av­er­age. 

In­stead, they wrote, the op­po­site hap­pened: the view­ers saw new im­ages of si­m­i­lar faces as slightly uglier than be­fore.

Yet anoth­er test served as some­thing of a probe of just how far from av­er­age “i­deal beau­ty” might be, to view­ers. The ap­par­ent ans­wer: even furth­er than the ear­lier ex­per­i­ment sug­gested.

The sci­en­tists used a com­put­er to iden­ti­fy the dif­fer­ences be­tween the or­di­nary av­er­age and the “at­trac­tive” av­er­age, then ex­ag­ger­ate those dif­fer­ences. That is, the ma­chi­ne took the “at­trac­tive” com­po­site, and dis­tort­ed it by am­pli­fy­ing what it had cal­cu­lat­ed as the beau­ti­ful char­ac­ter­is­tics. By de­grees, the chi­n got smaller; the nose nar­rower and more button-like. Event­ual­ly, the face started look­ing just bi­zarre. 

But be­fore that, some­thing in­ter­est­ing hap­pened. The study volunteers kept lik­ing the im­ages more and more un­til they were dis­tort­ed by some­where be­tween one-and-a-half times, and twice, the ini­tial “av­er­age-beau­ti­ful” diff­er­ence. Only after that did peo­ple start to call the pic­tures uglier in­stead of pret­ti­er. 

In oth­er words, the re­search­ers wrote, “at some point, car­i­ca­tur­ing an at­trac­tive shape will re­sult in a face that is so ab­nor­mal that con­cur­rent pref­er­ences for av­er­age­ness will out­weigh pref­er­ences for the at­trac­tive shape di­men­sion.” To put it more simp­ly, Plain Jane is not with­out her charm.

* * *

Send us a comment on this story, or send it to a friend

Sign up for

On Home Page         


  • Meet­ing on­line may lead to hap­pier mar­riages

  • Pov­erty re­duction, environ­mental safe­guards go hand in hand: UN re­port


  • Was black­mail essen­tial for marr­iage to evolve?

  • Plu­to has even cold­er “twin” of sim­ilar size, studies find

  • Could simple an­ger have taught people to coop­erate?

  • Diff­erent cul­tures’ mu­sic matches their spe­ech styles, study finds


  • F­rog said to de­scribe its home through song

  • Even r­ats will lend a help­ing paw: study

  • D­rug may undo aging-assoc­iated brain changes in ani­mals

What is physical beauty, in a person? In recent years, much scientific opinion has gathered in support of a somewhat dreary answer: beauty is just the average. There is evidence to back that up. In studies, volunteers usually rate “average”-looking faces as more appealing than atypical ones. Computer-aided blends or “morphs” of several faces—even unattractive ones—tend to look surprisingly good. (Try it yourself here). But if it’s disheartening to ponder the idea that our cherished fantasies revolve around something so, well, average—take heart. A new study may have restored a touch of the old mystery that beauty once had. Psychologists have found what would seem to be a slight but definite difference between average faces and the most lovely, as rated by volunteers. “There are specific non-average characteristics that are particularly attractive,” wrote the researchers, detailing their findings the December issue of the research journal Journal of experimental psychology: Human perception and performance. “Our results suggest that while averageness is one component of attractiveness, there is at least one other component of attractiveness that is not explainable by averageness.” The study didn’t delve into deeper questions beauty raises: for instance, just what that non-average component could be; and what beauty’s evolutionary function might be (many scientists agree it probably helps signal good health, though exactly how is uncertain.) But the new findings did highlight the possibility that even the limited scientific agreement surrounding such issues may be premature. A previous study, for instance, suggested that beauty is that which the mind’s eye finds easy to understand. But that contention depended largely on the assumption that beauty is an average; the brain presumably finds that easy to grasp, because average is typical. In challenging the beauty-as-averageness hypothesis, the new study wasn’t breaking entirely new ground: at least one past study has also done so. But critics had challenged some of its conclusions. The new study claimed to shore up some of the weaknesses in that previous work. In that 1994 study, researchers from the University of St. Andrews, U.K., averaged 60 images of faces with a computer. They then separately averaged the 15 faces that volunteers had judged as best-looking of the bunch. It turned out people rated this smaller average as handsomer than the full-group average, raising doubts about beauty as a simple average. But other scientists raised questions. Couldn’t it be—they asked—that the top 15 faces were indeed so average, that their combination reflected the average of the human race even better than the whole 60? The newest study attempted to settle the question by repeating the experiment, but adding several others to serve as strict tests of the initial result. The researchers, Lisa M. DeBruine at the University of Aberdeen, Scotland and colleagues,again made an average of 60 faces—all white females—and a separate average of the 15 “prettiest.” Viewers again rated this as more alluring than the full-group composite. Further tests, the researchers said, shored up the conclusion that average and gorgeous weren’t quite the same. For example, viewers themselves, unaided, recognized the larger average as being “more average” than that of the smaller group, apparently contradicting the idea that the smaller composite might have been the truer ultimate average. In another test, DeBruine and colleagues subjected their hapless volunteers to a barrage of attractive-face images. Previous studies had found that looking at many images of one type, temporarily skews what someone considers “average” toward that type. Thus, the researchers reasoned, if the beauty-is-just-average hypothesis is correct, then the volley of winsome images should have persuaded volunteers to see additional beautiful faces as even more attractive—because they look more average. Instead, they wrote, the opposite happened: the viewers saw new images of similar faces as slightly uglier than before. Yet another test served as something of a probe of just how far from average “ideal beauty” might be, to viewers. The scientists used a computer to identify the differences between the ordinary average and the “attractive” average, then exaggerate those differences. That is, the machine took the “attractive” average image, and distorted it by amplifying what it had calculated as the beautiful characteristics. By degrees, the chin got smaller, the nose narrower and more button-like—till eventually, the face started looking just bizarre. Finally, the scientists put the resulting series of pictures before volunteers. What they found: viewers kept on liking the images more and more until they were distorted by somewhere between one-and-a-half times, and twice, the original difference between average and “beautiful.” Past that, people began to call the pictures uglier instead of prettier. In other words, the researchers wrote, “at some point, caricaturing an attractive shape will result in a face that is so abnormal that concurrent preferences for averageness will outweigh preferences for the attractive shape dimension.” To put it more simply, Plain Jane is not without her charm.