 |
| Photo by Sara LeeAnn Banevedes |
I don't think Brian Alexander is a bad guy or a misogynist. He writes the Sexploration column for MSNBC, so sure, his job is all about selling sex stories to the public. He even wrote a book about American sexuality. But I don't personally think he has a burning hatred for women, or views them as objects placed on this Earth for the sexual satisfaction of men. However, I very easily could, given how he chose to report on a recent study published in Science about men's physiological responses to the chemicals present in women's tears.
The headline alone was enough to make me gag —
"Stop the waterworks, ladies. Crying chicks aren't sexy." The sarcastic bitch in me just couldn't help but think Why THANK YOU Brian! I've been going about this all wrong. When I want to get some from my honey, I focus all my thoughts on my dead dog or my great grandma and cry as hard as I can. No WONDER it isn't working!
I didn't even want to read the rest of the article.
But I did.
It doesn't get better.
Alexander's reporting of the actual science was quick and simplistic, and couched in sexist commentary (like how powerful women's tears are as manipulative devices). 
And to finish things off, he clearly states what he found to be the most important find of the study:
"Bottom line, ladies? If you're looking for arousal, don't turn on the waterworks."
It's no wonder that the general public sometimes questions whether science is important. If that was truly the aim of this paper, I'd be concerned, too!
Of course, Brian Alexander missed the point. This paper wasn't published as a part of a women's how-to guide for getting laid. Instead, the authors sought to determine if the chemicals present in human tears might serve as chemosignals like they do for other animals — and they got some pretty interesting results.
In many species, chemical signals run rampant. Scents, pheromones, and other chemical cues are deliberately and unconsciously given off to tell other individuals anything from "Back Off - MY Tree!" to "Hop on and ride me, baby!" But despite how common they are in the rest of the animal kingdom, the function of chemical signals in humans is hotly debated. Years of searching has yet to find human pheromones (no matter what those websites tell you), and while scent seems to play a role in communication in people, there is still relatively little knowledge as to what chemicals and why.
Given that tears are known to serve as sexual signals in mice, it isn't strange at all that Noam Sobel and his team chose to look at the physiological responses to tears. The Israeli team designed an impressive and unbiased set of experiments to determine if the tears produced by women when sad elicit physiological responses in men separate of the visual or auditory stimuli of a woman crying.
To find out if tears alone acted as chemosignals, the scientists collected tears from women watching tear-jerkers, and as a control, compared their effects to saline rolled down women's cheeks. Men sniffed the solutions without any knowledge as to what they were during a series of different experiments. In the first, men with a tear-soaked pad under their nose were asked to rate the sexual attractiveness and mood of female faces. While the smell of saline had no effect, men inhaling Eau de Tears consistently rated women's faces as less attractive, though this had no impact on whether they found the faces happy or sad.
For the second experiment, men sniffed tears before watching a sad movie. While doing so didn't affect their mood, the smell of tears did elicit a physiological response: men's faces became more conductive to electricity, which happens when we sweat and is indicative of a psychological reaction. Furthermore, the men self-reported less sexual arousal, which was reflected in their bodies as a 13% drop in saliva testosterone levels.
But to really get to the meat of it, the team threw their male test subjects into an fMRI machine and scanned their brains for activity while sniffing tears. Researchers saw much less activity in the hypothalamus and the fusiform gyrus, both of which are thought to be involved in sexual arousal. All three experiments lead to the same conclusion: the chemicals in women's emotional tears reduce male sex drive.
The real question, though, is why? Why do men's testosterone levels tank at the smell of a woman's tears? The overwhelming answer given by mainstream media (as Rheanna pointed out) is that tears just aren't sexy. When women cry, so the journalists say, it's a chemical signal that they don't want to have sex. Because evolution is all and only about sex... right?
Sorry to burst their bubble, but even when it comes to evolution, it's not all about sex. Selection also favors survival — because, you know, you can't have sex when you're dead*. Thus women's tears are not necessarily evolutionarily intended to turn guys off. For example, Ed Yong brings up the hypothesis that tears might be used to downplay aggression. Think about it: we cry when we're sad or physically in pain. In both cases, we're more vulnerable. Getting others, especially angry men, to be less aggressive towards us in that moment could certainly be a benefit to survival.
Really, the idea that tears are intentionally used as a turn off is a hard sell to an evolutionary biologist. What benefit do women get from not having sex when crying? Does it somehow make them have healthier or more babies? Not for any reason I can think of.
There is, instead, an even more intriguing explanation, one that makes a whole lot more sense. Many who wrote about this paper (including Brian Alexander) mentioned that tears are known to contain a variety of compounds, including prolactin, the hormone which is responsible for making a guy cool his jets after he gets off. But prolactin does much more than ensure a guy stops going at it — it's a hugely important hormone for nurturing behaviors. In fact, the connection between reduced testosterone and nurturing/bonding behaviors may be the real reason as to why men's testosterone levels dip upon sniffing tears.
Numerous studies have shown that parental and nurturing behaviors are mediated by prolactin while inhibited by testosterone. For example, research has shown that prolactin levels positively and testosterone levels negatively correlate with a father's impulse to respond to a baby's cry. Furthermore, men's prolactin levels spike and testosterone levels drop in the weeks before their partner gives birth.
It goes beyond babies, too. Decreased testosterone and increased prolactin are strongly implicated in establishing and maintaining relationships. Monogamous men have significantly lower testosterone levels and higher prolactin levels than their single brethren. Furthermore, studies have directly shown that artificially increasing testosterone in a double-blind, placebo-controlled setting makes men less generous to strangers and reduces a person's empathy for others.
Perhaps prolactin or other chemical signals in tears are directly targeting and activating the nurturing pathway in men's brains. Being taken care of or protected when in emotional or physical pain would definitely benefit an individual's survival. Personally, I would like to see this study of tears replicated to determine women's responses to the scent as well as men's reactions when using men's and children's tears, as well as looking at the levels of prolactin, oxytocin, and other well-established bonding and empathetic hormones. My bet is the response isn't limited to men, and isn't limited to emotional secretions from women.
While Brian Alexander and the rest of the sensationalists seem to suggest the signal is "I'm not in the mood," its likely that the message has nothing to do with having or not having sex. Women aren't saying "back off" — they're saying "help me."
Why do I care so much? It's not just that they got it wrong. It's that their interpretation of research isn't labeled as opinion. It's that the vast majority of people who have any interest in science news are going to read inaccurate (if not downright insulting) news articles and think studies like this one are either misogynistic or frivolous. It's that dumbasses like Brian Alexander undermine good science for the sake of attention grabbing headlines. And as a scientist and a writer, it's a double insult.
Gelstein, S., Yeshurun, Y., Rozenkrantz, L., Shushan, S., Frumin, I., Roth, Y., & Sobel, N. (2011). Human Tears Contain a Chemosignal Science DOI: 10.1126/science.1198331
* I can hear the comments now, you sickos, so let me clarify: you can't have baby-producing sex when you're dead.
Also, thanks to Kira Krend for the thoughtful and hilarious discussion on this topic!
Other References:
Haga S, Hattori T, Sato T, Sato K, Matsuda S, Kobayakawa R, Sakano H, Yoshihara Y, Kikusui T, & Touhara K (2010). The male mouse pheromone ESP1 enhances female sexual receptive behaviour through a specific vomeronasal receptor. Nature, 466 (7302), 118-22 PMID: 20596023
Fleming, A. (2002). Testosterone and Prolactin Are Associated with Emotional Responses to Infant Cries in New Fathers Hormones and Behavior, 42 (4), 399-413 DOI: 10.1006/hbeh.2002.1840
Storey AE, Walsh CJ, Quinton RL, & Wynne-Edwards KE (2000). Hormonal correlates of paternal responsiveness in new and expectant fathers. Evolution and human behavior : official journal of the Human Behavior and Evolution Society, 21 (2), 79-95 PMID: 10785345
Burnham, T. (2003). Men in committed, romantic relationships have lower testosterone Hormones and Behavior, 44 (2), 119-122 DOI: 10.1016/S0018-506X(03)00125-9
Zak, P., Kurzban, R., Ahmadi, S., Swerdloff, R., Park, J., Efremidze, L., Redwine, K., Morgan, K., & Matzner, W. (2009). Testosterone Administration Decreases Generosity in the Ultimatum Game PLoS ONE, 4 (12) DOI: 10.1371/journal.pone.0008330
HERMANS, E., PUTMAN, P., & VANHONK, J. (2006). Testosterone administration reduces empathetic behavior: A facial mimicry study Psychoneuroendocrinology, 31 (7), 859-866 DOI: 10.1016/j.psyneuen.2006.04.002
- Log in to post comments
Fantastic post, Christie!! One of the best analyses of this article that I've read.
Also, your comments are working now. :)
Thank you!How refreshing to read a thoughtful,multi-faceted approach to this study,when every mainstream media outlet goes with the simplistic and insulting (to men and women)"OMG tearz killz bonerz!1!!"-narrative.
On top of everything else, so much science journalism is just plain corny, and acts like we have the attention span of goldfish (although real goldfish, of course, actually remember things from at least three months ago). They feel the need to keep rephrasing the basic idea over and over in the most inane language possible, as if reading science as science is just too daunting for us stupid Americans.
"Yes, scientists say it looks like your pet goldfish, far from being the aquarium airhead, might be able to reminisce about events as far back as last October."
It's worse than those action movies where the hero feels the need to follow every dispatching of a guard with something like "Well, it looks like he won't be catching the 7:30 Express." And this MSNBC article looks like one of the worst offenders on that.
Thank you for this post!
This is an amazingly written piece. I am a graduate student in biology and, as such, I'm the go-to-person in my group of friends when someone finds some interesting science article. I hate when people bring me something of this sort, with a catchy headline, decent science, but a stupid interpretation that most average chump believes because it came from a "scientist". It irritates me that this kind of thing gets published.
Alternatively, whatever causes the physiological changes in men isn't in tears for any particular reason (ie the null hypothesis, which is rarely considered for these studies)
Good post, though.
very nicely done. it's unfortunate (and dangerous) that mainstream media needs to make some titillating high-school hee-hee gossip approach to every damn story to attract readers. sigh.
I'm a non-scientist, and every time I see an article that has anything to do with gender, I automatically assume it's crap. And since that sort of thing is always popping up on mainstream media, I am weary of science articles in general, through the lens of non-science journalists.
Also, why are all these articles about science seeking to validate the current views of men and women? I wish it was "edgy" to try to DISPROVE them.
Great post, delightful amount of snark.
great article ! Could not agree more with what you wrote.
Wow, what a sexy photograph you've posted there .. oh, no wait, SHE'S CRYING!!!!!! Run away!!!
(Actually, that is a really sad photograph, well done to the photographer and model/sad person.)
This story is a classic. When I read the part of your post where you report the headline, I thought that maybe this was a case of the editors doing something an author would not do, but it does appear that the article as a whole takes all the cheap shots one might expect with a story like this.
I've not read the original paper yet, but I'm a little suspicious of the whole thing. Steroid hormone amounts operate in a person on the basis of a baseline that may vary a great deal from one person to another. I'm not sure how this works in calibrating as an interpersonal signal. Also, said hormone amounts have a lot of important informational qualities. How does a bodily fluid like tears happen to have an amount of what essentially becomes a pheromone in relation to that?
A test to see if this is real is to look for a mechanism of regulating hormone levels in tears, snowing that their presence and levels are not just incidental to the fact that steroids have the run of (most of) the body.
Anyway, lovely post. Brought a tear to my eyes.
Awesome analysis.
Next, they need to do a study on why tears have no effect on abusers. But maybe it just wears off after a while?
Excellent analysis of the paper, Christie.
I wonder, however, whether the pheromone actually has to hit the olfactory senses for the ardor to cool, so to speak. Surely there are plenty of other emotional cues, including visual and auditory, which might take effect even before the tear smell reaches you? Of course, the smell can reinforce the effect, but I wonder if the physiological response starts sooner. For example, what happens if men watch that sad movie without sniffing anything in advance? I haven't read the paper, so I don't know if they had that "control" in their experimental design - but doesn't watching a sad movie about complete strangers (i.e., without any immediate physical/chemical connection to the person crying) also suppress arousal? Or else men might be less averse to weepie chick-flicks, no?
One more thought - what about violence, which is also linked to testosterone? I'm not sure there is much evidence that crying helps women escape domestic abuse, for instance - or does it?
Ah... I see Sarz raised the same question as I was typing my response; why no effect on abusers?
My gut reaction to the original MSBC article was also ... what about rape!
And I was so disappointed in the scientists summation which I felt contributed to the misinterpretation by similarly seeming to exploit "sex" sensationally ... which in fact the research did not do.
But then you came along with the BEST SMACK DOWN EVER of a piece of poorly researched journalism!
Your wit is priceless and I am still laughing at all of it ... while even more pleased by the thoughtful, thorough exploration of the research and it's implications.
This is a beautiful, witty article and I simply cannot stop smiling when I think of it.
Congratulations ... awesome!
I don't think the evidence suggests there's no effect on abusers. Implicit in Christie's explanation is the fact that people who aren't certain kinds of broken respond to signals of vulnerability and distress from people* with compassion and restraint. If abusers respond to those signals with feelings of power and gratification, then it's perfectly consistent, whatever else you'd say about it.
*specifically, from people they recognize as "people", as opposed to "that [minority]" or "that weak kid in homeroom" or "that slut who knows she wants it" etc.
Good points, Azkyroth, about the potential for different responses from different kinds of people. Which only strengthens my suspicion that there is a lot more to these kinds of responses than simple hormonal responses to pheromones in tears! If you can get opposite kinds of responses to the same stimulus, then it can't just be that one stimulus alone doing it now, can it?
BTW ... delighted by the * footnote! (priceless)
Christie, thank you for writing a blog that is intellectually responsible. Can't say the same about Alexander's.
Was surprised to see the NY Times (In Womenâs Tears, a Chemical That Says, âNot Tonight, Dearâ) get it wrong too ... stating: 'They had assumed chemical signals from tears would trigger sadness or empathy in others. But initial experiments found that sniffing womenâs tears did not affect menâs mood or empathy, but âhad a pronounced influence on sexual arousal, a surprise,â Dr. Sobel said.' when no test for empathy had been preformed, nor did they note that no chemical test for sadness was explored. (Nor had a chemical test for sadness been explored)
It's an annoying title as is its concluding summation: 'Dr. Sobel said whatever the evolutionary origin of emotional tears, he was not contending that his findings had modern romantic applications. âItâs not that I would recommend to any woman to cry to send a message,â he said. âIt would be much better to just say no.â'
Serendipitously, we were discussing this exact subject over on reddit a few days before this research came out. There's ample anecdotal evidence that the NYT article (and the associated research) isn't broadly accurate, or at least that there's a lot of room for exceptions.
Given that, as you note, solid evidence of human pheromones has so far proved elusive, finding them first in tears seems quite unlikely. I'm going to wait until this result is replicated a few times before assuming it's valid. And even if it is, the particular interpretation made in the linked article, ignoring for the moment its sexist assumptions, is only one of those possible.
Unfortunately, our media has become so tabloidized that the way this was reported (omigod, sex!) isn't surprising in the least. And the attachment to gender stereotypes isn't either.
Thank you so much. This is the first article that I've read about this study that made sense. I enjoyed it MUCH MORE than any of the sensationalized bull shit I've read before it.
Thank you!
"My gut reaction to the original MSBC article was also ... what about rape!"
Not tonight, dear.
Sorry.
The paper is obviously trying to create buzz by being offensive. If you can get people riled up, you'll get column inches.
For me, I would say a bigger effect is more like "if she's crying, maybe she's not interested in 'hide the sausage'". 'course some people may consider it an opportunity to be nice and cop off on a woman who is thankful for comfort, though whether this is right or wrong is hard to say (since the woman would be accepting of the arrangement).
And, obviously, those who want a power trip will want tears because it demonstrates that they have broken the other person.
PS anything on men crying? Or is it only women crying changes the hormones?
Awesome write-up -- best one I've read of this reseaerch.
I did read that indeed, they are planning future research to explore whether (as they anticipate) the same effect is seen with male tears and female sniffers.
Lucid and clear analysis. I read about the study but didn't know about the prolactin end of it. Thanks! I ignored all the idiot takes on it, but you're right that it is insulting on a number of fronts.
Thank you for this, it is always nice to see bad science journalism taken down a peg.
Like the fact you used the research to turn the interpretation of this finding from the male chauvinistic 'ME' sexual gratification consequences mindset to the reasons and causes based on hormonal behavioral facts. Male testosterone isn't just about sex but high levels induce aggression and violence. Secreting something to lower it does help in raising survival levels.
Here is what I have been posting in comments for several science blogs on how it's been inherently understood the effects of tears per violence, especial sexual violence.
"To me this implies an innate meme humans know and understand when you see/read writing in violent and horror stores where the male sexual abusers scream at their female victims to shut up and stop crying or try to keep from seeing their tears. It not only forces a sympathetic guilt but chemically shuts down male sick sexual arousal from violence. But women shouldn't have to start crying to stop male sexual violence."
Thank you for this. Thank you, thank you, thank you. Really. Thank you.
...And to the commenters here, who restore something of my faith that intelligent conversation is possible somewhere in the world....
"Male testosterone isn't just about sex but high levels induce aggression and violence."
Women have testosterone too.
Women tend to go for mental violence rather than physical. Society has "hen pecked" and such men are considered wimps. Society has "abused" wives and such women are considered tragic. This slant makes the mental game a much safer place to play for women.
I've come to this conclusion:
There are fewer differences between men and women than the men think and there are fewer differences between women and men than women hope.
Humans. Who'd have them?
"Personally, I would like to see this study of tears replicated to determine women's responses to the scent as well as men's reactions when using men's and children's tears, as well as looking at the levels of prolactin, oxytocin, and other well-established bonding and empathetic hormones. My bet is the response isn't limited to men, and isn't limited to emotional secretions from women."
I agree, and perhaps baby tears would have an even more dramatic effect on both sexes. Infants obviously have fewer ways to communicate, so crying is even more important for them than for those who can speak. Also, babies are more vulnerable than adults, so reducing aggression in mom and dad would be even more important. And the baby being the greedy guy that it is, would want to monopolize it's parents time, so reducing the frequency that dad pursues mom for sex would be good too.
Does it have to be mainly chemical in nature? How about the empathy of the human watching the tears? Or is empathy only enforced from outside? (and I thought I was a pessimist.
PS People pursue other people for sex.
Heck, not even humans pursuing humans. Every species boffs critters in heat from another vaguely similar species. "Girass: he's making one" as demotivational.com says.
Awesome discussion & writing, I had no idea there was so much science (and hormones) involved behind the response to tears. It is a bit chauvinistic to assume "Crying isn't sexy in chicks", what a title. I did find it funny in some ways,..like how part of the expt involved getting guys to sniff tears & saline while a watching a sad movie.
I can't remember the study :( but the authors of it looking at tears themselves were promoting how it's good to cry when you're upset. It's a release of a multitude of chemicals/hormones. In the test they had people cry when upset and when an onion was held up to their eyes. The same hormone cocktail did not appear in the onion induced samples.
I wonder if animals can sense it too? (Sorry to be far fetched..my dog used to be extra sweet and friendly if i was crying).
Excellent analysis. I just went over this paper for journal club, so I had a few additional thoughts.
I have a couple of issues with their testosterone analysis, and I'd love your take on it.
1) They didn't actually do a statistical comparison of tears vs. saline when it came to testosterone. Instead, they independently conducted tests for testosterone reduction in the tear-sniffing and saline-sniffing conditions, and found a significant reduction in the tears condition but not in the saline condition. But that does't really tell you whether tears have any effect beyond saline! It could be that, say, watching sad movies and smelling salt produces a marginally significantly reduction in testosterone, and the effect just happened to reach significance in the tears trials but not in the saline trials. (They made a similar error, I think, in analyzing self-rated sexual arousal.)
2) There's one huge outlier in the saline condition (figure 3H). It's not that obvious because they changed the scale of the graph axes between the tears condition and the saline condition, but see that guy up top? His after-movie testosterone level was about five standard deviations above the mean--50% higher than anyone else's in the sample! I don't know what's going on with him, but if you remove him from the analysis, it looks to me like you'd get a significant testosterone drop in the saline condition, just as in the tears condition. The authors should have at least acknowledged the guy.
Neither of these problems means that their conclusion is wrong, of course, but I'm definitely more skeptical of it than I was when I first heard about the research.
Yep. And that applies to their fMRI results too. The fusiform gyrus (left in particular) does light up in men seeing sexy images, but it also lights up when they see threatening images--angry faces, photographs of human/animal attacks. (See for instance Schienle, Schafer et al, 2005, "Gender differences in the processing of disgust- and fear- inducing pictures.") If tears generally suppress that activity, it's totally consistent with their being a signal for "help me" and/or "don't hurt me."
Pretty much all of Gelstein's results are consistent with the "attachment theory of crying" advanced (as Ed Yong mentions) by Randolph Cornelius, AJJM Vingerhoets, Michelle Hendriks and others: namely, that weeping functions to gain assistance and defuse aggression from others.
Oren Hasson has a theoretical evo-psych paper ("Emotional Tears as Biological Signals," Evolutionary Psychology, 2009) which advances pretty much the same theory--but using more of the vocabulary of animal behavior--and nicely reviews the literature on the topic while he's at it.
Linda, once when I was really upset and crying, a strange cat came up to me to cuddle. Once I calmed down, I took a good look at the cat-- I had been blurry-eyed from the tears before, and saw it had an infect injury around one of its eyes. I pulled myself together to see if I could get it to the vet, but as soon and my emotional focus changed, it turned skittish and went into the underbrush.
So I don't know if it's smell or what, but that cat was definitely giving kindness to a stranger.
I wonder if the effect is less pronouced when the cryer isn't a stranger. My husband would wait until I was more or less calmed down, but the comforting hug tended to morph into a grope, much to my annoyance. Maybe he's just an outlier, but otoh, maybe that's why tears don't always calm anger: a more familar tint to the smell diminishes its effect?
There is a problem with this whole premise. The questions shouldn't be what reproductive benefit this gives to women, but what reproductive benefit it gives to men. It is men that the tears are having an effect on, and doing something to those men which would seemingly reduce the reproductive success of those men (making them less apt to want to have sex with a crying woman).
I suspect that it relates to the reasons why women cried during evolutionary time. If she was crying because her mate/owner beat the crap out of her (probably something extremely common), then a man not her mate/owner getting it on with her might be dangerous (to the man) because her mate/owner would likely kill him (and her too).
This is essentially what happens in âhonorâ killings, a woman does something that one of her male relatives doesn't like, so he kills her and any non-related males she happens to have done it with.
I suspect that this may be part of why there is such reluctance to try and intervene in domestic violence against women. The men who perpetrate such violence are apt to take it out on anyone who would intervene. In the absence of a state police authority, crossing the perpetrator would be dangerous. Over evolutionary time, mechanisms might evolve to avoid such situations. Having male libido reduced by female tears might be one such mechanism.
daedalus2u @35
I disagree pretty strongly. Pheromones are not always useful to the recipients (nor are, say, the chemicals delivered in seminal fluid); I'm only really familiar (vaguely) with insects but hijacking the other player's endocrine system is not an uncommon ploy. Crying, however, makes no sense. It wastes resources and blurs your vision in situations where you'd think visual acuity might be useful; it is a frankly dangerous and confusing response, so it needs more explanation.
But what really throws me is that the researchers apparently started with a sex division. Even if you assume females cry more often, why not assume that they cry in front of other women as well? Intrasexual competition and aggression do happen. I'm at home, so I can't access the paper, so is there any explanation from the authors as to why they went that way?
Does there HAVE to be a benefit? All that there needs to be is not so great a disadvantage as to reduce the opportunities for breeding.
Empathy helps because it allows planning for hunting or group cohesion of great variety and, for a poorly armed weak ape like hominids, the group action is VERY successful.
Empathy however allows anthropomorphising and that makes us less likely to act correctly in the face of danger like when finding lion cubs.
Aaaaw, cute!
Is not a useful reaction when mum's about and cheesed off.
As to the lowered sex drive, is that not possibly a reaction to empathy: we have sex because WE LIKE IT. If someone is crying, they're probably not going to like it and our empathy kicks in: it's better sex when two people (or more!) are enjoying it.
And association will be made with the visual look, the sound (cat mewls sound a lot like baby cries intentionally) and the smell.
Yes, emotions will be driven by chemicals, but not in a linear causal relationship. Memory is chemical in nature. And that can cause emotion.
What an amazing article! Thank you, thank you, thank you! And the comments have been great as well.
I disagree pretty strongly. Pheromones are not always useful to the recipients (nor are, say, the chemicals delivered in seminal fluid); I'm only really familiar (vaguely) with insects but hijacking the other player's endocrine system is not an uncommon ploy. Crying, however, makes no sense. It wastes resources and blurs your vision in situations where you'd think visual acuity might be useful; it is a frankly dangerous and confusing response, so it needs more explanation.