Cognitive Daily reports nearly every day on fascinating peer-reviewed developments in cognition from the most respected scientists in the field.
Have you ever seen Singin' in the Rain? One of the movie's most hilarious moments is when the beautiful silent movie star Lina Lamont is asked to start making "talking pictures." As soon as this gorgeous screen siren opens her mouth, the illusion of her beauty is shattered: her squeaky voice instantly transforms her from a glamorous leading lady into a cartoonish boor. Threatened with losing their box-office cash-cow, the studio chiefs frantically enroll her in voice and etiquette lessons, but nothing helps, and eventually they're forced to substitute the voice of the attractive and lovely-voiced Kathy Selden.
But what makes one voice attractive and another unattractive -- and how much effect does a voice have on our overall perception of attractiveness? Recent research has found that women with attractive faces also tend to have attractive voices (which may explain why Lamont's hideous voice in Singin' in the Rain is so surprising).
In general we perceive higher voices as more feminine. Faces with exaggerated feminine features are also perceived as more attractive. Can we say the same about feminine voices?
A group of researchers led by David Feinberg recorded the voices of 123 young women as they pronounced five vowel sounds: ah, ee, eh, oh, and oo. Then ten male volunteers rated each voice for attractiveness. Here are the results:
The graph compares the pitch of each voice with its attractiveness rating. There is a lot of variation in the ratings for a given voice pitch, and many low-pitched voices are rated as more attractive than higher-pitched voices. Yet the overall pattern is shown by the diagonal line, the linear model, which indicates that high-pitched voices are seen as significantly more attractive than low-pitched voices. The dashed line represents the alternative explanation for the data: the idea that average voices are more attractive. While statistical analysis shows this model is also significant, the linear model fits the data significantly better. Higher-pitched voices are more attractive.
So will simply raising the pitch of a female voice make it more attractive, or are there other factors involved? Maybe high-pitch is simply an artifact of some other vocal feature. In a second experiment, three groups of five voices were chosen from the 123 original voices: five low-pitched, five medium-pitched, and five high-pitched. Then a computer program was used to artificially raise and lower the pitch of each of these voices. Then hundreds of volunteers listened to high- and low-pitch versions of each voice and indicated which was more attractive. Here are the results:
No matter what the starting pitch was, significantly more male raters found the higher-pitched version of a voice more attractive. The same result was true for female raters, for all except the highest-pitched voices. While these results support the first experiment, they also suggest that for women at least, there may be some limit to how high a voice can go and still be considered attractive.
Overall, however, it seems that as long as a voice is within the normal range of pitches, the pattern is for higher-pitched women's voices to be perceived as more attractive. If there is a limit to how high a woman's voice can be and still seem attractive, my anecdotal research suggests that that limit is somewhere below Lamont's pitch in Singin' in the Rain.
Feinberg, D.R., DeBruine, L.M., Jones, B.C., Perrett, D.I. (2008). The role of femininity and averageness of voice pitch in aesthetic judgments of women's voices. Perception, 37(4), 615-623. DOI: 10.1068/p5514
Brain & Behavior
Comments
Nice pitch!
Posted by: Ian | July 3, 2008 12:54 PM
The scatterplot suggests a lot more going on..
Wouldn't pitch be indicative of physical attributes such as body size? Also, there is an interplay between fundamental frequencies and harmonics in the human voice that is responsible for its character. It would be interesting to find out how that interplay relates to physical attributes and overall health..
Maybe the resonating body cavities responsible for those additional frequencies need to be in a certain size range, or in a certain ratio, to sound 'well', and deformities to such spaces might cause them to sound 'wrong'.
Posted by: Jan-Maarten | July 3, 2008 1:23 PM
Anybody who's been to Japan has noticed that women who serve the public, as in department stores, cultivate very high-pitched voices. Perhaps culture has more to do with these data than the investigators think.
Posted by: JakeR | July 3, 2008 2:18 PM
I also think that cross-cultural studies are needed here. Personal experience tells me that physical attractiveness is affected by cultural perceptions -- but I haven't heard much about cross-cultural studies into this area. Similarly, the voice-pitch phenomenon might also have cultural elements involved.
Posted by: GAC | July 3, 2008 2:30 PM
As you mention, there is an upper-Lina limit to the pitch, before it starts sounding cartoonish. Maybe related to higher=more attractive is a higher=more feminine preference. Without being able to see the speaker, guys might just go with the more feminine sounding voice in case they think they are being tricked into choosing a male voice.
Posted by: Dan | July 3, 2008 4:24 PM
This is sort of an aside, but I remember a voice teacher of mine once saying that women today in general have lower pitched talking voices because of WWII, when women were asked to enter the factories to make up for all the men being shipped overseas.
Ladies started to develop more masculine attributes on purpose to fit in with managers and eventually the returning soldiers. One such adjustment was voice pitch, and it's just sorta stuck since then. Her idea is that this is why pre-1940s starlets sound to modern ears like they've been sucking on helium.
There's no scientific basis for this as far as I know, but it always sounded to me like a fun theory to test!
Posted by: koffeekat99 | July 3, 2008 5:39 PM
Counter-example: Lauren Bacall.
Posted by: Ahcuah | July 3, 2008 5:54 PM
I have noticed that high pitched voices attract much less focused attention than lower pitched voices. In a circle of scientists discussing a recent lecture of publication, lower pitched voices will almost always dominate.
Posted by: pamela ronald | July 3, 2008 5:54 PM
More important than pitch is lack of the nasal quality that many women have in their voices, especially some female "commentators" in American TV. Like Nancy Grace. That kind of voice can drive you nuts. (Bad example perhaps, because the woman is nuts to begin with.)
Posted by: temizlik | July 3, 2008 7:20 PM
Whenever I spend time with a cat, my allergic reaction causes me to speak in a husky low pitch like the actress Brenda Vaccaro. Anybody old enough to remember her? Ex., Tampon commercials. Anyway, I feel less than attractive in that condition.
Posted by: --Lisa S. | July 3, 2008 11:24 PM
Rerun this test with Dutch participants.
Posted by: false_cause | July 4, 2008 2:27 AM
I love how they draw a curve through what is essentially a random scattering of data points on their chart...okay, I know there are linear regressions, multivariate analyses, and all that, but it doesn't look like a nice simple relationship to me. And, as JakeR points out, attractiveness can be and is very strongly led by culture. Although the rationale for high-pitched shop assistants in Japan is not necessarily to do with attraction it could just be that people want to spend quickly and get out of the shop...
Posted by: David Bradley | July 4, 2008 3:35 AM
The dashed line represents the alternative explanation for the data: the idea that average voices are more attractive. While statistical analysis shows this model is also significant, the linear model fits the data significantly better. Higher-pitched voices are more attractive.
To be consistent with the facial attractiveness studies (which show combining several faces together--resulting in an averaged/composite face--results in a more attractive face), it's likely that the nonlinear regression is closer to reality. Obtaining more data points above the 240HZ range might strengthen the nonlinear model, since it looks like removing the data above 240 HZ would result in a linear regression line with a steeper slope that fits the data from 180-240HZ better. The nonlinear plot would also be consistent with a more recent facial attractiveness study ( (Science gives beauty some of its mystery back) suggesting that if one continuously digitally abstracts out the facial characteristics identified as making a face beautiful, the face is judged to be more attractive up to a certain point, and then it is judged as significantly less attractive.
To address the issue of cultural biases, it would be interesting to conduct this study with infants. Facial attractiveness research with infants shows that beauty (at least in terms of the face) may be objective since infants around 6 months old show a preference for faces adults judge as attractive. So if repeating the current studies with infants yields the same result, it might suggest no cultural bias. This would especially be interesting in light of child-directed speech (formerly called motherese), whereby adults tend to raise their voice pitch when talking to children.
Posted by: Tony Jeremiah | July 4, 2008 12:24 PM
How do we know the men heard the same thing? Those recent articles on tone-deaf genetics like my family carries suggest to me that we aren't all hearing the same qualities in a sound.
Posted by: Mary | July 4, 2008 7:05 PM
There's the overall pitch of a voice, and there's the placement of vocal formants within the frequency spectrum.
Not the same thing, but there's a middling correlation between the two.
So maybe listeners are responding to the formant pattern, whose definite-but-not-strong correlation with pitch produces the scattering on the chart.
Posted by: Kapitano | July 5, 2008 7:04 AM
I'm sorry, but there is no way to justify drawing a line through that data. You could draw anything through that mess. How do they justify claiming any correlation?
Posted by: tonyl | July 5, 2008 12:59 PM
To those focusing on the first graph, remember that it is supported by the second experiment, which I think provides a much clearer and stronger bit of evidence supporting the contention that higher-pitched voices are more attractive.
I would agree, however, that extending the line all the way out to the 280 Hz frequency doesn't make much sense. If you look at the actual data points above about 240 Hz, you don't see much support that voices that high are more attractive than lower-pitched voices.
Posted by: Dave Munger | July 5, 2008 2:37 PM
The wide variance in this data suggests there may be different types of listeners -- those who fall for high pitch, and those who don't. The variance may narrow down significantly if we could figure out what high pitch lovers have in common.
My place of employment is about 80% women. Females ask other females for favors ("could you please file this") using their high voices. If a male makes the same request, using the same words but in a low male voice, they are perceived as rude.
Posted by: Shay Jones | July 7, 2008 6:46 AM
I have some friends who I see weekly. One of them is female. Recently, a younger, more attractive male joined the group, and a friend in the next room pointed out that the female member's voice was noticeably higher-pitched that day. I didn't notice it myself, but find it very amusing.
Posted by: Dread Polack | July 7, 2008 4:29 PM
I think the study is way too simplistic. With a topic such as language and sound perception, you can't really gain a whole lot of understanding through a study like this, and the data shows it. There is a slight relationship, but the variability of the data suggests there is way more going on. Some of the comments allude to some possible factors. I'd say it's very similar to how we perceive the sounds of instruments: many play the same pitches, but have completely different timbres. Certain timbres have inherent attractiveness. The perceived attractiveness of timbre alone is topic that pulls in a lot of possible factors.
Posted by: Eric Weber | July 7, 2008 10:53 PM
Assuming attractiveness arises from perceptions of health and therefore reproductive potential it could be that outside of cultural preferences an attractive voice is one that comes from a healthy body. This could help explain the apparent link between pretty voices and pretty bodies.
If I'm unwell, cold, overtired or smoking too much my voice pitch is lower and more gravelly. Perhaps a set cues about overall heath are communicated this way and it is their absence, rather than the absolute pitch, which cause these results.
Posted by: strix | July 8, 2008 1:27 PM
As others have noted, the cultural influence cannot be ignored. Cross-cultural studies may either corroborate their conclusion or prove there wasn't much of a relationship at all. However, I would note that many screen sirens were/are known for their husky voices, and some men find that sexy. I saw something recently that indicated that testosterone levels during and after sex can make women's voices drop in pitch; thus, lower voices can remind men of sex. I also read in a recent interview with Scarlett Johansson that casting directors did not want to cast her as a child because of her low voice.
Something else to think about, as temizlik points out: nasal quality. Pitch isn't everything, and if the voices the men listened to spoke more than just vowel sounds, characteristics of an accent might also influence the attractiveness rating. The cultural data provided by movies indicate that men apparently find Eastern European women's accents very sexy.
Jan-Maarten: That's a fascinating idea.
Posted by: Elle Rayne | July 9, 2008 6:50 PM
Children exposed to electronic gadgets, high TV exposure, with very little communication communication at home are actually training themselves - without their knowledge- to a smaller bandwidth. In the next 10 years children may not be able to recognize low pitched sounds of animals and even human speech.
Posted by: Benedict | July 12, 2008 8:52 AM
In addition to the pitch of voice, scientists should also have explored another physical quality; timbre. I mean a sound of a particular frequency in the audible range may sound different with different waveforms. 100 Hz sound of sawtooth, sinewave or square wave are essentially different. Another thing that loudness (decibel) or a combination of loudness with pitch should also have been taken into account as high pitched voice spoken louder may not sound as sweet than when spoken low.
Regarding the 'vowels', I guess any woman would speak just as sweet!
Posted by: Amiya Sarkar | July 14, 2008 11:02 AM
Wow - very interesting!
It seems to me that higher-pitched voices are more indicative of younger women. And as we're genetically predisposed to "spread our seed" far and wide, a child-rearing aged female would be more appealing to men, especially young men.
It was my understanding that preferences for light colored hair are driven by this same predisposition.
Posted by: Drew | July 27, 2008 1:31 AM
Most of the men I know report the opposite -- lower, huskier voices are more attractive on women. My ex-husband used to prefer my post-cold voice to my "normal" one ... said it was far more exciting. Additionally, I am only 5'1" tall, and in my experience as a short woman, I have found that I seem to command more respect (professionally and in social situations) when my pitch is lowered.
Posted by: P. LaRocque | August 5, 2008 1:47 PM
It should be noted this is a generic study, and doesn't tease out the differences between pleasant vs. sexy vs. respected. E.g., feminine/pleasant is not the same as feminine/sexy. One should consider that deep voices may not only indicate testosterone levels or health, but also aggression. Growl at your dog (or cat or pet bear)and you will see what I mean. A deep voice can be threatening, a high one is less so, and therefore more pleasant. Just try asking for something with a lowered voice pitch compared to a raised voice pitch.
Personally, I find a deep voice in a man very sexy, and a deep, mellifluous voice will cover a variety of sins. A man with a high pitched voice, no matter how handsome, smart or successful, is not on my 'to do' list.
I also notice that my high voice is not heard well by men, especially older men (my father complains!). And I've noticed, like other women here, that having a lower voice gets you more respect in a group of arguing men.
Posted by: Leslie | August 11, 2008 5:21 PM
Very interesting! I think there was similar studies that related to men's voice, specfically that physically attractive men were found to have more attractive voices.
Perhaps voice, like body odor, can be a good indicator of mate quality in that healthier or higher value mates will produce more attractive voices.
Posted by: Nol | November 27, 2008 7:04 PM