New Study: If a Dude Sounds Strong, He Probably Is

By Joseph Calamia | June 16, 2010 5:47 pm

armIt’s pretty clear that–in a fight–Darth Vader would crush Jar Jar Binks, Optimus Prime would beat Starscream, and Batman could pummel the Joker. Though some of these fictional characters don’t even look like humans, when it comes to strength, their voices give it all away. New research seems to confirm this: humans, like other animals, can accurately predict physical strength from voice alone.

In a study appearing today in The Proceedings of the Royal Society, researchers asked subjects to evaluate the upper-body strength of speakers from four distinct populations and language groups just by listening to their voices. Even when unfamiliar with a speaker’s language, listeners could tell which men might be good in a fight. The men they judged as sounding brawny were in fact physically stronger as measured by tests of hand grip, chest strength, shoulder strength, and bicep circumference.

As lead author Aaron Sell told Discovery News:

“Information about male formidability would have been important for both sexes over evolutionary time,” said Sell. “Both men and women would have benefitted from knowing who would likely win fights in order to make prudential alliances and for other reasons. Men would need this information to regulate their own fighting behavior. Women would also need this information in order to make effective mate choices.”

They study failed to make a similar link between women’s voices and strength. The study’s authors speculate that this is because early men were more likely to spar. The researchers also couldn’t determine what it was about certain male voices that made them sound strong–it wasn’t just a deep timbre–and say listeners may respond to a complex mix of cues.

For men, the finding proves especially interesting given the non-menacing statement researchers asked English speakers to say: “When the sunlight strikes raindrops in the air, they act like a prism and form a rainbow.” Apparently this sentence is from a passage that contains almost all the sounds of the English language, but those certainly aren’t fighting words.

Related content:
Discoblog: Speaking French? Your Computer Can Tell
Discoblog: Penn State’s Football Stadium: Now 50% Louder!
Discoblog: NCBI ROFL: Did Gollum have schizophrenia or multiple personality disorder?

Image: flickr / ~ggvic~

  • Carl

    While the result itself is interesting, the purported explanation is complete nonsense: the usual kind of post hoc “just so” story of evolutionary psychology. Imagine for a moment that the experiment had turned out to show no correlation. The 3rd paragraph would then have gone like this:

    “Disguising male formidability would have been important for both sexes over evolutionary time,” said Sell. “Men would have benefitted from disguising who would likely win fights in order to deter aggression and dominance by stronger males and for other reasons. Women would also want this information in order to make effective mate choices, so weaker men would have greater reproductive success if they could eliminate signals that give away weakness. It’s another example of adaptive mimicry so often found in nature.”

    Good science starts with a hypothesis, a prediction, and then a test. When your hypothesis — “evolutionary success shapes complex behaviors” — allows you to predict completely contradictory outcomes, then cherry pick the one that matches the outcome, you really don’t have much of a hypothesis.

    A more likely explanation? Probably there is a correlation between hormonal levels and strength, aggressiveness, and vocal characteristics.

  • Chris

    @Carl: From the sound of it, their ideas as to the cause were purely speculative, in a “future work” section, or similar. Your hypothesis for the cause is also good. It just means that further studies are needed if we want to actually test what the cause might be.

  • Joseph Calamia

    Carl and Chris,

    Thanks for the comments. You both make a good point: the researchers don’t seem to know why certain voices correspond to strength measurements. The test can’t tell them why; it can only show that they do correspond, statistically.

    To me, what is also interesting is that the researchers don’t seem to specifically know what it is about the voices that differs; they aren’t necessarily deeper. As Chris says, these all seem like areas for further research.

    Carl, your thought about what the authors would have said if the study showed no correlation is also interesting. This paper mentions that previous studies tried to link voice to height and weight and showed no significant link. The study also showed no link between women’s voices and their strength.

    For more, the whole study is up here:

    Thanks again for commenting.

  • Aaron Sell

    Evolutionary Psychology is frequently accused of making “just-so” stories when we document new findings. This is a curious criticism because it seems to apply less to EP findings than any other area of social science. For example, the debate on self-esteem and aggression fits neatly into any social psychology framework; e.g. if those with high self-esteem are aggressive it is because they feel entitled, if low SE is linked to aggression it’s because they’re lashing out at others to compensate.

    What makes EP different is that our hypotheses must pass a strict logical test: the features we posit must have plausibly led to increased reproductive success in our evolutionary past. It is the same test that theories in evolutionary biology must pass.

    When people are unfamiliar with evolutionary biology, animal behavior and models of natural selection it often looks like those who use the theories are just making them up. This is not the case. The literature on animal assessment and aggression show that animals, prior to aggression, exaggerate or enhance cues of fighting ability, but they do not hide them. This was noted by Darwin in his Expressions book and has been replicated, documented and predicted in dozens of species since then.

    Models of animal aggression show that mutual assessment is selected for. Indeed long “escalating conflicts of assessment” take place between animals in which relatively safe mutual assessment can occur.

    If I had studied any species other than humans and documented this ability no one would think to call it a just-so story. It is a clear prediction from a model of evolutionary selection that has already predicted and found this sort of ability in other species.

  • sunny

    That sentence does not contain all English sounds; for example, there is no “ing”, and nothing else that sounds like it.

  • Sheila Best

    The 70s party theme seems to go on forever.
    There’s music, films and fashion to remind us.
    The site here is full of all outfits worn by
    people of the time.


  • Eliza Strickland


    You’re right, I inadvertently inserted that error while editing this post. The sentence the men read is actually part of a longer passage that’s commonly used in speech tests, called “the rainbow passage.” The passage was designed to contain *almost* all the sounds in the English language — apparently it’s got everything except the palatal fricative and the glottal stop. I’ll fix the error in the post.

    — Eliza, DISCOVER online news editor


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!


Quirky, funny, and surprising science news from the edge of the known universe.

See More

Collapse bottom bar