Tears as chemical signals – smell of female tears affects sexual behaviour of men

By Ed Yong | January 6, 2011 2:00 pm

In an Israeli laboratory, Shani Gelstein is harvesting a woman’s tears. The volunteer is watching the end of the boxing film The Champ. As she weeps, she holds a vial under her eyes to capture the fresh drops. This might seem ghoulish, but Gelstein has used the collected tears to understand why people cry during emotional times. She thinks they’re a chemical signal.

Gelstein used several different techniques to show that the smell of a woman’s emotional tears could reduce a man’s sexual arousal. The men never actually saw anyone cry, and they didn’t know about what they were smelling. Even so, their sniffs reduced their testosterone levels and they lowered the activity in parts of their brain involved in sexual desire.

“Smells are an effective non-verbal means of communication,” says Noam Sobel, who led the study. “It is intuitively obvious to us that smell signals can be used to attract (sexually and otherwise). It is just as helpful to have smells that will do the opposite.” Sobel and Gelstein don’t think that this effect is unique to the tears of women, but for reasons that will become obvious later, they only studied female tears.

Many other mammals use chemicals – pheromones – to influence each other’s behaviour, and Gelstein thinks that human tears contain just such a signals. After all, they are more than simple salty secretions – they contain many proteins, fats, salts, and even traces of drugs. For example, Kazushige Touhara at the University of Tokyo found the tear glands of male mice produce a protein called ESP1 that makes females more sexually receptive.

If similar signals exist in humans, they’re very subtle ones. After collecting the tears from several volunteers, Gelstein confirmed that they had no obvious smell. Men couldn’t tell the difference between them and drop of saline that had been trickled down the cheeks of the same women.

But the drops did provoke a reaction. Gelstein asked 24 men to sniff a jar containing either fresh tears or saline, and to wear a pad on their upper lip soaked in the same chemical. Each volunteer smelled tears on one day and saline on another. Neither they nor Gelstein knew which was which until all the results were in. With the smell of tears wafting into their nostrils, the men found pictures of female faces less sexually attractive, although no more or less sad. Saline didn’t affect them either way.

In a second experiment, Gelstein asked 50 men to sniff tears or saline before watching a sad film. In this explicitly sad context, the tears didn’t influence the volunteers’ mood any more than saline did. But when the men sniffed tears, their skin became better at conducting an electric current (a sign of sweat and psychological arousal) than after sniffing saline. As before, their sexual arousal dipped afterward, according to their answers on a questionnaire. Their saliva even backed up their claims, for it contained less testosterone.

As a final test, Gelstein scanned the volunteers’ brains while they took a whiff of tears. She specifically focused on parts of the brain that are involved in sexual arousal, such as the hypothalamus, which controls several basic bodily functions, and the fusiform gyrus, which helps us to recognise faces.  She found that these areas were less active when the men watched a sad film, if they had previously sniffed tears instead of saline.

Gelstein focused on emotional tears, because they contain different chemicals to those we shed to lubricate our eyes and remove irritating substances. These differences were discovered by William Frey II around 30 years ago. However, Ad Vingerhoets, who studies emotions at the University of Tilburg says, “I could not replicate that finding twice with much more sophisticated methods.”

Even if emotional tears are different to other types, Vingerhoets thinks that Gelstein should have compared sad tears to irritated ones, as well as to saline. “It would be intriguing also to harvest ‘positive’ tears, associated with feelings such as admiration or elevation,” he says. Touhara agrees that “some important controls are missing”; for a start, he wants to see what male tears would do.

So why didn’t Gelstein study male tears?  “In one word: feasibility,” she writes. “In the West, and perhaps even more so in Israel, crying is (currently) generally more acceptable for women than for men.” She wanted to work with fresh tears and she needed people who could cry on demand. To recruit them, she placed an ad asking for volunteers who could cry regularly and perhaps unsurprisingly, the responses came almost entirely from women. Male tears, however, are top of the list for future research.

Gelstein’s study adds a new possible role for tears to an already varied list. Frey suggested that people cry at emotional times to remove chemicals that build up during stress. This is why we feel better after a good sob (and Hippocrates has a similar idea). Robert Provine thinks that they are simply a visual sign of sadness. Oren Hasson thinks that tears are a deliberate handicap. By blurring a person’s vision, they provide an honest signal of submission or helplessness, triggering sympathy from friends or mercy from an enemy.

Vingerhoets has a similar idea. He thinks that tears allow babies to influence the behavior of nearby people, after they’ve been attracted by the sounds of crying. Parents would be prompted to care for the child; strangers would feel a stronger social bond; aggressors would be appeased.

In fact, Vingerhoets thinks that the smell of tears could also make men less aggressive, which would fit with their falling testosterone levels.  Their reduced sexual arousal could just be a side effect. Sobel also says, “I expect that the signal in tears will also lower aggression (as it lowered testosterone).  Lowering aggression in the person you are interacting with is an obvious interest.” The field is clearly open for debate.

An even more contentious question is whether humans have pheromones at all. In 1998, the answer seemed to be yes. Martha McClintock at the University of Chicago found that a woman’s sweat can lengthen or shorten the menstrual cycles of other women, depending on the time of the month when the sweat was collected. She billed it as “definitive evidence of human pheromones.” And Sobel previously found that male sweat contains chemicals that influence the level of hormones in women.

Both studies seem to show that we can secrete chemicals that alter each other’s behaviour  but they have drawn their fair share of criticism. Sobel thinks the debate is “more of a semantic argument than anything else” and he wants no part of it. He makes no claims about human pheromones in the paper, and the word is barely mentioned. “There is no clear agreement of what is or isn’t a pheromone,” he says. “I prefer to stick to the important information: that tears contain a chemical (or chemicals) that sends meaningful messages to [members of the same species].  Some will call this a pheromone, and some may not.”

Does Gelstein’s study change anything? Certainly, everyone I spoke to praised its design but Touhara is still sitting on the fence until the team actually finds the specific chemical in tears that affects male behaviour. The team have already acknowledged this challenge. It’s on their list, along with all the other controls that Vingerhoets and Touhara suggested.  They want to see if the tears of men and children have a similar effect, and how tears affect people of the same gender. They want to see if irritant tears can act as signals. And they want to understand if female tears say anything besides sexual disinterest.

Reference: Science http://dx.doi.org/10.1126/science.1198331 If the citation link isn’t working, read why here

More on emotions:

Twitter.jpg Facebook.jpg Feed.jpg Book.jpg


Comments (26)

  1. I shouldn’t say it, but. . . I’m sure John Boehner could help her out. (I know, I know, I’m sorry). :)

    (& quick explanation for that U.S.-centric comment: Boehner is our new Speaker of the House; he cries constantly).

  2. Absolutely fascinating and great write-up, Ed. Very clever design, too – good job Gelstein et. al.

    I find it interesting that tears reduced male sexual interest. Quite a few men report being turned on by crying. Now, obviously, this could just be a paraphilia standing orthogonal to these findings, but it’s worth thinking about nonetheless.

    And then there’s the Roman proverb that “when a woman weeps, she is setting traps with her tears.” It would be fascinating to test whether the tears themselves (not just the act of crying) can be manipulative.

  3. Very interesting!

    Two quick thoughts: first, I bet there will be a lot of population variation in this, depending on the cultural relevance and meaning of crying. Second, I wonder why it is that crying has to be a signal — the more important element of crying, to me, is how useful it is to the individual (that is, many people across many cultures feel better after crying, and my understanding is that this is because more hormones are cleared from the body through the act of crying — through tears and sweat brought on by that physiological process). I know they found this depression in testosterone with smelling the tears, I’m not saying that evidence isn’t there. I just wish there was more research on why crying is good for you!

    Also (okay, this is a third thought), I wonder why, even in cultures where crying babies are not harmful (that is, they aren’t signaling there are some tasty humans around to a nearby predator), we are so anti-crying in babies. It’s probably very useful, physiologically, for them to cry. Aletha Solter has a cool paper on this from the early 1990s.

  4. Ooh — instead of testosterone, wouldn’t it have been cool if they’d looked at oxytocin and vasopressin? Just thinking aloud. Done now.

  5. I was relieved to read that it reduced male sexual interest. I’m not sure that anyone is against babies crying in theory, it’s that when babies cry it is instinctive to soothe them. I think that’s the point really. A baby crying and crying with no let-up is an indication that something is wrong that isn’t being addressed. I remember as a new parent, the helpless feeling when I couldn’t figure it out, and the relief when I finally did.

  6. Lilian Nattel wrote: “A baby crying and crying with no let-up is an indication that something is wrong that isn’t being addressed. I remember as a new parent, the helpless feeling when I couldn’t figure it out, and the relief when I finally did.”

    I have a toddler myself, so I am also very familiar with that helpless feeling and the relief when they stop! :)

    First, unfortunately I don’t think it’s just an instinct to soothe — when my daughter was an infant and she would cry my family would get visibly upset. I really think a lot of people are bothered, angered sometimes even (like in a public place), by infant crying. But that’s a whole other topic I suppose!

    What I was thinking about is what happens when no matter what you do — nurse, soothe, rock, hug, whatever — the baby keeps crying. This is at least as common, sometimes more common, than need-based crying (for food/closeness/etc). My daughter was perfectly normal and healthy, but once a day for the first few months she had a screamfest that I couldn’t really impact — this is nearly universal among other parents I know. My sense is that crying is a release that gives them a space to express themselves, and it’s good for them to do so if all other needs have been met and they are supported by a loved one (not left to cry it out). Again, there is some evidence for this in the Solter article, but very very little work’s been done so I’m not basing this as much in the evidence as I’d like.

    But then, I study physiology not behavior so my mind ALWAYS goes to mechanistic level explanations. :)

  7. @KHBC – Hooray! I was hoping someone would mention oxytocin. Here’s a quote from Vingerhoets that I had to cut out of the piece:

    “I suppose that, if anything, tears will stimulate oxytocin in observers (or: those who are exposed to the “smell”) (actually: this is well known in breast feeding mothers). Oxytocin, more or less, may be considered as the antipode of testosterone in its behavioral effects: it reduces aggression and stimulates social bonding. In my opinion these are the two key functions of crying/ producing emotional tears. I would like to replicate the study with the focus more on aggression, social bonding (and oxytocin).”

    Also I mentioned the idea that crying clears hormones from the body. 6th paragraph from the end.

  8. Man, whenever I hear about this type of study, I always assume it’s going to be vastly over-interpreted by someone. It’s sounds like the authors in this case are not doing so and are, in fact, being reasonably cautious with interpretations – and “Yay!” to that.

    And double-yay for them stomping on the “pheromones” thing.

    Also, I think you accidentally omitted the third picture.

    I do think assuming that this effect is direct – and therefore, somehow, the evolutionary purpose of tears, as opposed to being a side-effect or downstream effect (as Vingerhoets mentioned) – is what I imagine the media will do to this story. That’s doubly a shame when the authors appear to be taking pains to avoid such over-interpretation.

  9. What third picture?

    And yeah, I admired them for the cautious interpretation. I think they made it clear that they’ve found something interesting but can only speculate as to *why* it happens. We have the proximate cause and not the ultimate one.

  10. Actually I meant the fourth picture (if you count the header of the post). It was a joke that, but one that maybe only I get. Apologies there.

  11. Hi Ed! Sorry, yes, I did see that you wrote that — I was commenting more on the study’s main aims. Sorry for not being clearer!

    And what they want to do next sounds really cool.

  12. Katharine

    Which chemical does this, and can we make a synthetic aerosol version?

    (Not only is this intrinsically interesting, I am seriously pissed off after nearly being assaulted at the metro station. What I wouldn’t give to have something to spray to lower the testosterone of an aggressor to either make them go away or put them off their guard.)

  13. @Katharine – I think the chemical you’re looking for is known as “tear gas”. Betcha it would lower testosterone levels…

  14. Katharine

    Actually, Ed, I think I may have found a possible answer: human tears, period, contain prolactin, which counteracts the effect of dopamine, which functions in male and female sexual arousal.

    If they tested how male tears affect female sexual arousal, they might have an answer.

  15. TGAP Dad

    I’m reminded of an episode from the 1980’s series Tales From the Darkside: The Tear Collector. I even remember the titular character holding a vial to a woman’s cheek for the collection process. Once again, life imitates art.

  16. Juicyheart

    I wonder if this is why men don’t cry as much? Especially if the tears have the same effect regardless of which gender cries them.

  17. RSBR

    cheap and natural contraceptive.. haha

  18. Juicyheart

    Could the tears be operating as a self-chill-out mechanism ? Rather than a communicative function? How close do the tears have to be, to be affective? Could tears be away of calming ourselves down, that can be literally wiped away?

  19. Juicyheart

     I guess I’m seeing the drop in testosterone, as the main finding here  The lack of sexual interest would be a result of that, but wouldn’t there also be a drop in aggression and perhaps flight/fight response?  Maybe testosterone is only one of many flight/fight hormone levels affected by tears.  And if the signal doesn’t travel far, the self might be the target. If crying babies are any clue, this would be more of a nudge, than an on/off switch (what was the percentage drop of testosterone levels?). But some times a nudge is all that’s needed to get things going in the right direction. 

  20. Great post! One thing I’ve always wondered about these kinds of studies, though, is how relevant they are to real-world situations. Having a tear-soaked tab taped under your nostrils/ sniffing a vial is one thing, but how often do we get that close to someone’s tears (or armpits, for that matter, as in the sweat studies that have been done)? Surely the concentration of these molecules at normal social interpersonal distances must be pretty low? And how volatile are they anyway? Did the authors talk about that anywhere?

    But given that they contain nearly 500 different proteins, (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1779605/?tool=pubmed) there is clearly more to tears than meets the eye.

    I’ll get me coat….

  21. I love my commenters. I always know that if I cut something out of a piece, someone will ask about it.

    @Claire – yes they covered that! I wrote a paragraph about it but had to cut for length/flow. They say that in the real world, crying often leads to physical contact. You’d try to console people who are crying, by hugging etc. That would bring your nose into close contact with far greater amounts of tear chemicals than the amounts they used in the experiment. Hence, real-world applicability.

    I’m not entirely sure how convinced I am with that, if only because I assume there must be a lot of cultural diversity in terms of how people react to crying. Do all cultures react with sympathy and close contact?

  22. @Ed — Editors are just as bad. If you cut something from your draft for length/ flow, they will ask you to put it in. Then they will go and take it out again at a later stage. For length/ flow. Then subs come along and complain the missing info is needed to improve the flow of the story. Back in it goes, until the piece is laid out on page and is found to be too long, so out it comes again. Then once it’s published, in come the readers’ letters…

    Thanks for answering my question! I’d be interested to know the answer to the cultural diversity question too. I also wonder whether pets, whose noses are more sensitive than ours, might also pick up on these chemical signals. IME dogs get very agitated in the presence of a crying human. Given the long history of association between our two species, it might not be so surprising if they do.

  23. KN

    Good place to get male tears: India!

  24. Kat

    HHHMMMMM — this study is a good start for a military chemical ordinance.

    Toss in a gas grenade loaded with the chemical that causes men to lose testosterone levels / interest and you have a more pacified enemy!!

    Talk about cheap weaponry ! No more guns & ammo — just gas grenades. The Military should start to back this type of study.

    Our people wearing gas masks throwing a handfull of grenades toward the enemy — wait 10 min. War OVER!!

    Or even better – make the platoons entirely out of females and they will not be affected by the chemical. They can be the Military’s new Shock and Awe Troops!

    who hopes people realize the above is written as SATIRE!!!

  25. I wonder if the presence of these chemicals in tears may be a vestigial evidence of a previous defense mechanism against sexual assault or abuse.

  26. Avinash Limaye

    I had Thyroide problem during my child hood school days, 5th to 8th – Age 8 to 12 years, Whenever any teacher point me or asks any question, wen i know the answer , Tears used to roll over, as if I dont know the answer and afraid of teacher or learning ? What is this phenomena, can any medic answer ? Now its ok after I passed that age !


Discover's Newsletter

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

Not Exactly Rocket Science

Dive into the awe-inspiring, beautiful and quirky world of science news with award-winning writer Ed Yong. No previous experience required.

See More

Collapse bottom bar