Tears have two jobs, and for most of the day the eye is kept wet by a steady basic supply, and grit or onion fumes trigger a flush of reflex tears. Emotional tears use the same glands with no irritation at all.

That is the part people find confusing. Nothing has touched your eye, so the tears feel like they came from nowhere, but they came from the brain, on a system built to signal to other people rather than to clean anything.

The signal works, and across 41 countries and 7,007 people, adding tears to a face raised the intention to offer help. Tears are how a human asks for support without saying a word, which is why they leak out at the worst moment.

Under long pressure that signal fires early. You are not more upset than the situation deserves, but the threshold has moved, so the same amount of feeling now reaches it.

The build-up is measurable, and in 60 women watching films chosen to make people cry, heart rate climbed in the minute before tears started, then dropped quickly once crying began, with slower breathing along the way.

So a cry is two things in sequence, because arousal peaks on the way in, and the calming branch of the nervous system takes over on the way out. In that study every measured effect had settled within four minutes.

The lump in the throat and the tight face are part of the same package. Crying involves the breathing and trunk muscles, which is why it can feel physical and why fighting it takes work.

The loop closes socially. Tears at a desk or on a call feel exposing, so people brace against them, and bracing is more of the pressure that lowered the threshold in the first place.

Emotional tears come from the same small glands above each eye that keep the surface wet all day. What makes them different is the trigger. Nothing has irritated the eye, and the instruction comes down from the brain instead.

That is unusual, because plenty of animals produce tears to protect the eye, and as far as researchers can tell only people produce them in response to feeling. The best supported explanation is social rather than chemical.

Tears make distress visible at a distance. In a study spanning 41 countries and 7,007 participants, the same faces were rated as warmer, more helpless and more deserving of support when tears were added to the photograph.

So the honest way to read your own tears is as a signal that got sent, sometimes earlier than you wanted. It is a system doing its job in a setting it was not designed for, like an open-plan office or a video call, and that mismatch is most of what makes it feel humiliating rather than useful.

Two systems set the threshold. A fast nerve route changes heart rate and breathing in seconds, and a slower hormone route keeps the body switched on for much longer. A stressful week runs mostly on the slower hormone route.

You can watch the fast route work. Sixty women were shown neutral films and films chosen to bring on tears, with nothing physical done to them at all, and their heart rate rose in the run-up to crying.

The moment tears started that reversed, because heart rate fell, breathing slowed, and the calming branch of the nervous system showed up in the measurements. Everything had settled again about four minutes later.

A second study of 197 women found the same shape, with heart rate dipping just before tears and returning to baseline during the first minute of crying. Both were purely mental triggers, which matters, because it shows this whole sequence can run without a single thing happening to your body.

Most stress crying never becomes a full cry. It stops at a hot face, a tight throat and the effort of holding a normal voice, which people then dismiss as unimportant.

The diary evidence says otherwise, because across 1,004 crying episodes recorded by 97 women over about two months, the urge to cry tracked worse mood just as clearly as crying did, and it also went with bigger mood swings day to day.

The same records show how uneven this is. Crying showed up on about 16% of the days logged overall, but that ranged from 2% of days to 73% of days depending on the person, on the same measure.

So the useful reading is a threshold that moves rather than a switch that flips. If you are swallowing the urge several times a week, the load is already high, whether or not anything reached your eyes. That is worth acting on earlier than a full cry.

Sleep does a lot of the quiet work here, and pooled across 154 experiments and 5,717 people, cutting sleep reliably lowered positive mood and raised anxiety, which is the exact ground tears come from.

There is a brain-level version too, because after one night without sleep, 26 healthy adults showed a stronger alarm response to unpleasant images and weaker steadying input from the front of the brain.

The monthly cycle plays a part for some people. In one large study women reported feeling more like crying before and during a period than mid-cycle, though actual crying behaviour did not differ, and the pill made no difference either.

Medication belongs on the list as well, because antidepressants can flatten emotional response rather than sharpen it, reported by a wide range of users and usually reversible with a dose change, so a change in how easily you cry can follow a prescription change in either direction. Caffeine and short nights stack on top of all of this in a tense week.

The threshold is learned rather than fixed, and how easily a person cries tracks their habits of handling feelings, which is a skill set rather than a fixed trait, and skills respond to practice.

The strongest trial evidence comes from a clinical setting, where 432 inpatients with depression were randomly given standard therapy or the same therapy plus structured emotion-regulation training, the training group improved more, including on negative mood.

That population is not the same as a stressed office worker, so read it as direction rather than dose. What it establishes is that how strongly feelings arrive can be moved on purpose.

Context matters as much as skill, and across 35 countries, whether people felt better after crying depended heavily on who was with them and how they responded, which means telling one safe person is a real intervention rather than a nicety. If you have to hold it in all day, expect the threshold to stay low.

One popular belief deserves retiring, because tears do not flush stress hormones or toxins out of the body. That idea came from 1980s work on tear chemistry and it has not held up, and the relief people feel is not a drainage effect.

When it was tested directly, women who cried at an emotional film and women who did not showed no difference in their stress hormone afterwards. A larger study also found criers were no better at tolerating a later stressor.

The catharsis promise is shakier than the folklore. Of 1,004 diary episodes, most changed mood not at all, about three in ten improved it, and roughly one in eleven left the person feeling worse.

Timing may explain some of that, because in a study of 60 people, mood got worse immediately after crying, was back to baseline 20 minutes later, and was better than the starting point at 90 minutes. So a good cry may help, later, and anyone promising instant relief is ahead of the evidence.