Truth in Tears: The Function of Crying

Can evolution and signaling theory explain why we cry?

When I was a kid, I had a beloved record called Free To Be You And Me. I listened to it probably a thousand times, much to the irritation of my (very understanding, it is true) parents. I memorized the lyrics to each and every saccharine song.

One of my favorites, It’s Alright to Cry, started this way:

It's alright to cry
Crying gets the sad out of you
It's alright to cry
It might make you feel better

Raindrops from your eyes
Washing all the mad out of you
Raindrops from your eyes
It's gonna make you feel better

I loved the song, in part, because it felt true. Crying did sort of make me feel better.

Now, of course, I’m interested in why crying makes us feel better. What is its function?¹

By “crying” I mean the raindrops from your eyes accompanying by the distinctive facial expression that occurs when one is experiencing strong emotion, usually negative, such as sadness, but occasionally positive, such as joy.

If emotions measure and motivate, why do some emotions lead to tears? It’s easy to see why fear—measuring danger—leads to fleeing. That makes sense. Running away does something useful when faced with threat: getting some distance between you and the threat. Tears don’t seem to have an obvious use.

Now, I’m holding aside uses such as keeping the eyes lubricated, washing away irritants, and, in the case of our crocodilian friends, getting rid of excess salt. These functions of tears-as-wiper fluid are common among many species, including humans. My interest is in emotional tears.

Swamped with emotion.

Crybaby

Babies cry.

Babies cry because they are small and helpless and need things from others but lack language to clarify what: food, sleep, wet wipe, etc. Crying is, notoriously, difficult to interpret, though now there are apps that purport to translate. Still, judging from all the times I’ve seen crying babies rewarded with food or a clean diaper, the tears seem to be effective. From this pattern—combined with the fact that crying doesn’t seem to do anything useful in and of itself—it’s reasonable to guess that crying, in babies at least, is a signal.

This is helpful because there is very good theory about signals. First and foremost, signaling theory² tells us that signals evolve because they confer benefits to the signaler. This very simple fact can be easy to lose sight of, in part because the advantage to the signaler is not always obvious. (For example, how does it help me to signal to others that I am in awe? Great question!) This makes sense for babies: the tears tell parents the baby is in need, and the baby is better off because it gets help.

This line of reasoning probably explains why crying diminishes as we age. Just as in the case of the laughter associated with tickling, being able to talk reduces the need for non-verbal communication. Once someone has language, tears become less important because the person can use their words to say, “I’m hungry,” “my elbow hurts,” or “I tremble with fear for the future of the planet.”

Still, adults do cry, if less often than babies.

Why?

The Crying Game

Let’s return to signaling. Signaling theory as developed in the field of animal behavior (and economics) posits that communication between organisms often involves signals that are costly to produce or are otherwise "unfakable," making them credible indicators of the signaler's intentions or quality. The expense or difficulty in producing the signal ensures that only those who can "afford" to send the signal do so, making the signal less prone to deception. For example, the extravagant plumage of a peacock is a costly signal that makes getting around considerably more of a pain but demonstrates genetic fitness to potential mates; the effort and resources required to maintain such a display mean it's an honest signal of his health and vigor because a lesser peacock could not produce the display and manage to survive. This theory helps to explain why many forms of communication in the animal kingdom are effective.

Signaling theory also suggests that nature of signals depends on the alignment of the interests of the signaler and the receiver.³ In the peacock case, interests are not perfectly aligned. The male would like to manipulate the female into mating with him, even if he isn’t of high quality. The female’s interests, of course, are served by choosing only high-quality males. This conflict of interest leads to the costly (can’t-be-faked) signal. This is a general pattern. When interests aren’t aligned, signalers often have to put on a whole song and dance or receivers will simply ignore them. This idea explains why so many signals are large, loud, and active.

However, when interests are aligned, the receiving organism doesn’t need as much convincing because there is, by definition, less incentive for the signaler to lie. The closer the alignment of interests, therefore, the smaller the signal. For example, fireflies (lightning bugs) face a problem: mating with members of their species—and not other bugs who happen to be flying around. They use distinctive patterns of flashes to signal their species to potential mates. Importantly, they use them only intermittently, keeping as dark as possible to reduce the chance of being seen by things that would like to eat them.

You light up my life…

A similar example is the somewhat counterintuitive dance of predator and prey. If a prey has seen a predator, then it is in the interest of the prey to signal, “Hey, I see you over there.” The reason is that if the predator believes it has been detected, the chance of a successful hunt goes down and the predator is better off trying to find another prey, one it can take by surprise. How can the prey signal it has seen the predator? One way is for the prey to direct its gaze at the predator. This is especially effective as the predator moves around. Related, you might have seen prey mirroring the movements of predators, moving left when the predator moves right, in a little dance. This subtle signal lets the predator know it is being watched and carefully attended to, since the prey could not otherwise mirror the movements.

(As an aside, if you’re wondering why babies’ cries are so loud, given that this seems to be a case of alignment, my guess is the answer lies in parent-offspring conflict. Broadly, offspring want more investment than is optimal for parents—just ask any parent—so they fight about it. The loud signal is, in fact, a good indication that interests aren’t aligned.)

What Are You Crying About?

While people sometimes cry for joy, I’m going to focus on crying when you’re sad.

Deep sadness and grief typically occur after someone has borne a very big fitness loss or after seeing someone else who has borne such a loss. These losses are often, sadly, in the form of harm to (or death of) a romantic partner, family member, or close friend. Those are not the only sources of sadness, of course. People mourn the loss of jobs, status, or anything that they valued for whatever reason. Importantly, the loss does not have to be one’s own loss—we cry when friends suffer. And it doesn’t even have to be a real loss: people cry in movies, mourning the deaths of fictional characters.

Sadness seems to be triggered by real and fictional loss and harm. And, of course, we can feel sadness even when—sometimes, especially when—the entity in pain is not human. Many of us are, for whatever reason, reduced to tears when we see a dog in pain even more than when we see a person in pain.⁴

Having said that, this analysis is complicated by the fact that it’s not exactly clear why humans experience the debilitating emotion of grief. On this Substack, we have talked about how emotions guide adaptive behavior when you’re faced with a threat or opportunity. In the case of sadness and grief, generally the bad thing has already happened and there’s nothing to be done. So what’s the adaptive value of the emotion? Here are three possibilities, and the right answer might be some combination of all of them, or none of them.

First, consider the deep sadness caused by the death of a spouse, close friend, or family member. Such people often play important roles in our lives and we spend considerable time planning with them in mind. After the loss of such a person, all those plans need to be rethought, all the roles they played reassigned. Perhaps grieving shuts down the rest of life so that we can figure out new plans given the new situation we find ourselves in.⁵ Or maybe our entire model of the world—the scents we expect to smell, the sights we expect to see—was structured under the premise that this person exists, and now we have to painstakingly revise all those expectations, one by one.

A second possibility derives from the fact that humans are very good at simulating. So, we can imagine taking this course or that one. When we simulate, we have to have some way of evaluating our imagined results. Some have argued that sadness and grief are there so that if you imagine a situation in which a loved one dies, that feeling is evolution’s way of telling you to avoid that path at all costs. Grief is a deterrent for choices that lead to fitness-terrible outcomes. Then, when the bad thing really does happen, we actually feel that sadness.

A third possibility is that there is value in signaling to others that we have just borne a great cost. This possibility links to the case of babies, above. According to this view, we are embedded in social networks and it can be to our advantage to signal that we are in acute need as the result of having just suffered tremendous loss, as a means to gain comfort and aid.

This third possibility might help to explain tears. The thing about tears is that they don’t seem to be under our conscious control. While, I am told, skilled actors can cry when they need to, most of us cannot. Therefore, when we see tears in others, we can be reasonably sure that they are genuinely experiencing something very costly. If people could cry on demand, then the signal would be of no value. The signal can only remain useful as long as it is kept honest, in this case by the decoupling of the tears from conscious control.⁶ Blushing is also like this, an involuntary reaction that can be used by observers to draw inferences. When someone blushes, we don’t typically doubt that they are genuinely embarrassed.

According to this view, tears are—just as in the baby case—a literal cry for help. They are somewhat subtle as signals go—though sometimes they are accompanied by loud sounds, to be sure. From the fact that they can be relatively subtle, I judge that these tears are signals aimed at people whose interests are aligned: family, friends, partners, and allies. They are requests for succor, kept honest by the fact that the tears flow outside of our control.

Crying Out Loud

There is another important feature of crying. The diversity of stimuli that make people cry is tremendous. One person’s tears are another person’s joy. In Conan the Barbarian, the eponymous Conan was asked what is best in life. His answer (my emphasis): “To crush your enemies, see them driven before you, and to hear the lamentation of their women!”

Now, holding aside literal barbarians such as Conan, there are some stimuli that I would guess elicit tears more or less universally. Like many people, I visited Dachau when I was in Poland. I was able to stave off tears until I reached the room where thousands of shoes were piled up—a silent testimony to the different human beings who wore each pair, killed by the Nazi machine in the camp. And of course each of us experiences overwhelming sadness and grief when our loved ones are in pain or die.

While there is variation in what elicits tears, there is similarity in the signal. That is, while people might cry after seeing very different things, the crying itself is consistent.

This fact raises an interesting possibility. Consider that the social world is inherently ambiguous. Things like beliefs and values are invisible, hiding inside others’ heads. Nonetheless, it can often be helpful to signal to others that you share similar beliefs and values. Many groups, indeed, cohere on this basis. As always, just saying so won’t always do the job: after all, people can lie. Talk, unlike a costly signal, is cheap.  

Therefore, crying can be used to signal your values to others by demonstrating what you believe to be a great harm. To the extent that people might not know where your loyalties lie, crying allows one to signal loyalty to a group, cause, or idea. You say you’re a dog-lover but you watched Old Yeller without a tear, so I’m skeptical. On the other hand, we learn about the depths of the character Vivian Ward, played by Julia Roberts in Pretty Woman, when she cries at the opera. Over the course of our lives, we come to value people, ideas, and tribes. These values echo in our tears.

Let’s take two examples.

Two people crying.

In the picture on the left, a woman is describing the unbearable: the moment that she found out that her grandmother had just been killed. In the picture on the right is someone describing an experience at a bar where someone used a word⁷ that the person took amiss.

Both pictures include some of the features that frequently accompany crying. These include, of course, tears. Other features include swollen eyelids, flushing, a furrowed brow, and a downcast gaze. Some of these features are simply consequences of crying—reddening of the eyes is due to increased blood flow to the eyes and the salt in tears. Others, such as the downcast gaze, can be understood as a signal of deference.⁸

In any case, these two examples illustrate that very different stimuli can lead to very similar signals.

Love for relatives is universal. Some values are acquired as a result of the specific cultural contexts in which we find ourselves and it can be useful to signal to others what one takes to be harmful. These signals serve the dual function of soliciting aid as well as honestly broadcasting commitments to what we value.

The similarity of the expression in response to very different contexts supports the view that crying serves a broad signaling function, indicating the signalers relationship to the person, event, idea, group, and so forth.

CONCLUSION

A problem that people face is credibly signaling what they care about. It is one thing to say, “I have suffered a loss, and I am need of aid” or “I am committed to the idea of this country with all my heart.” Language makes these sorts of things easy—and cheap—to say.

It is another thing to shed tears in moments of tragic loss or when bearing witness to the defacing of the symbol of a nation.

Tears show your true, honest feelings.  

CODA

I asked someone to edit a prior version of this paper and they drew my attention to a paper under review by Sznycer, Gračanin, & Lieberman, Emotional tears: What they are and how they work. There is considerable overlap in the ideas in this post and that paper. I was not aware of paper when I drafted this. Readers should see that manuscript for a more thorough account. Their model has elements mine does not. To take one important example, they argue: “Tearing may function as an implicit plea for receivers to minimize the costs imposed on the tearer by nature, by third-parties, or by the receivers themselves.” I had not thought about this aspect of tears—as a plea for forbearance—which seems plausible to me. (I suggested something like this for the opposite of tears, laughter that accompanies ticking.) Sznycer et al. also suggest, as I do here, that “tearing may exhort receivers to infer and register which things the tearer values, positively or negatively.” Let the record show they thought of it first. And said it better.

REFERENCES

Dawkins, R., & Krebs, J. R. (1978). Animal signals: information or manipulation? In Dawkins, R., & Krebs, J. R. (Eds). Animal signals: information or manipulation? London: Blackwell.

Tooby J., Cosmides L. (2008). The evolutionary psychology of the emotions and their relationship to internal regulatory variables, in Handbook of Emotions eds. Lewis M., Haviland-Jones J. M., Barret L. F. (New York, NY: Guilford), pp. 114–137

1. Spoiler alert: this post does not, in fact, answer why crying makes us feel better. It’s about the evolved function of crying, and how it made our fitness better. Tomayto tomahto.

2. It’s old at this point, but Krebs & Dawkins, 1978, is just a great introduction to these ideas. A true classic. Also old but also good is Marc Hauser’s book on animal communication.

3. Again, this argument originated with Krebs & Dawkins, 1978.

4.  post on empathy for dogs versus people is in the works. Watch this space.

5. See Tooby & Cosmides, 2008, for a broad view of the adaptive role of emotions.

6. This raises the question of why there wasn’t evolution that favored tear control. I find this question vexing and I offer no answer. I agree it’s an issue.

7. The word was “lady.”

8. See the coda, below. The downcast gaze, signaling deference, seems very consistent with the idea that crying is plea.