Caroline Cherston, a physician training in reproductive endocrinology, opens her STAT First Opinion essay with a confession about where she first heard a whale song. Not at sea. In an intensive care unit, during her third year of medical school, when a man she calls Jake, who had collapsed playing soccer with his young daughter, lay in a coma and his mother's wails filled the unit. He died. The essay goes on to catalog the sounds she has heard since: quiet sobs after an early pregnancy loss, sharp howls after a full-term fetal demise, an elderly patient's crying after an ovarian cancer diagnosis. It argues that doctors must learn to soften rather than recoil, and that presence, as she puts it, is an earned skill.
The essay is right, and there is a stranger fact underneath it that its own frame does not quite reach. The sounds it describes are measurable physiology. Crying has a literature, with heart rates and respiratory traces and stress hormones attached. Whale song has a literature, built over half a century by scientists who treat sound as data. The wail in the ICU is one of the most information-dense signals in the room, and it is the only one that no instrument records. The chart has a column for the heart rate that the wail changed. It has no column for the wail.
The wail is physiology, and physiology gets measured
The belief that crying flushes stress hormones out through tears, popularized in the 1980s by the biochemist William Frey, has not survived testing. Careful experiments, including a 2020 study led by Leah Sharman, found no difference in cortisol between people who cried and people who did not. What the studies did find is more interesting. A person's heart rate decelerates just before crying begins and returns to baseline through the first bout of tears. Respiration, which spikes in distressed people who are not crying, stays steady in those who are. Crying, on the evidence, is a regulatory act: a way the body steadies its own breathing and its own pulse in the presence of overwhelming news.
That is worth sitting with. The sound that fills an ICU when a family learns the worst is not an interruption of physiology. It is physiology in action, as real as the tachycardia on the monitor, and a clinician standing in the room is watching two bodies regulate at once: the patient's, through the machines, and the family's, through the sound. Medicine has built a discipline around the first and almost nothing around the second. The stethoscope listens to the body. Nothing in the room is trained on the sound the body makes when the person it belongs to is losing someone.
Whale song is the study of sound that outruns its speaker
Whale scientists have spent decades taking exactly this kind of signal seriously, and their findings describe why the comparison the essay borrows feels so true. In 1971, Roger Payne and Douglas Webb showed that the low-frequency calls of baleen whales, built on frequencies between roughly 20 and 100 hertz, are adapted for transmission across extraordinary distances, traveling hundreds of miles through the deep ocean's sound channel. A 2024 paper in Nature, led by Coen Elemans, worked out the anatomy that produces the sound: a specialized larynx that lets a humpback sing without inhaling water, in a frequency band that overlaps, unhappily, with the rumble of shipping. The same physics that lets a whale's voice cross an ocean makes it vulnerable to being drowned out by one.
The grief sound and the whale's call share a property that is easy to miss and hard to forget: both are built to carry. The wail is not aimed at the person standing nearby. It is the acoustic form of a feeling too large for the room, and it travels down corridors and through walls in the same way, for the same reason, that low sounds travel through water. The ICU is an ocean in this limited sense: a place where sound is the oldest and furthest-reaching medium, where the news of a death arrives in the hallway before anyone speaks it. The difference is that the ocean's sounds have laboratories studying them. The ICU's sounds are left to the people standing in them.
The chart has no column for what everyone heard
The medical record is a list of what instruments transduce. Heart rate, blood pressure, oxygen saturation, respiratory rate, temperature: each has a number, a timestamp, a column. The record's power is exactly this, that it renders the body legible to anyone who opens it later. But the record is also a list of what medicine has decided to measure, and the wail is the most consequential thing in the room that appears nowhere in it. No one charts the mother's cry. No one notes its duration, its pitch, the way it changed the respiratory rates of the people around it. The information is in the air, and then it is gone, because the system has no port for the ear.
This is not a clerical omission. It is a decision about what counts, repeated millions of times a day, and it shapes the profession downstream. What is not written down cannot be studied, because researchers need data. What cannot be studied cannot be taught, because teachers need evidence. What cannot be taught cannot be reimbursed or staffed, because administrators need codes. The uncharted sound is invisible to every layer of the institution except the one where it actually happens, the clinician in the room, which means the skill of attending to it, the thing Cherston calls presence, is expected of physicians and given no formal support anywhere in the system that trains them.
The result is predictable, and it is not the clinicians' fault. A profession taught to trust the chart learns to attend to what the chart admits. The monitor beeps and everyone looks. The family wails and everyone, understandably, looks away, because looking toward it requires exactly the capacity that no one has taught, practiced, or protected. Recoil is not a moral failure. It is the trained response of a system that instruments everything except the moment when the instruments stop mattering.
The room is dense with sound, and none of it is the sound that matters
The ICU is one of the noisiest rooms in the building, and it is noisy by design. Monitors beep, pumps alarm, ventilators cycle, and the clinical literature on alarm fatigue exists because the soundscape has become a hazard in its own right: so many machine signals that the human brain learns to filter all of them, including the important ones. The irony is structural. Medicine has filled the room with sound in order to watch the patient, and the one sound that carries more information than any alarm, the family's voice breaking, has no sensor on it at all. The profession that spends engineering budgets on alarm thresholds has never instrumented the sound that tells everyone in the hallway what the monitors will eventually get around to saying.
There is also a physics detail in the whale research that belongs in the hospital for a different reason. Marine biologists studying whale song have found that the sound does not only travel as pressure waves. It also travels as particle motion, vibration moving through the water itself, and whales appear to feel it through bone as much as hear it through air, their jawbones conducting the vibration inward. Grief works the same way in a room. The wail is not only heard. It is felt, in the chest, in the floor, in the shared air of a room too small for it. The clinicians in Cherston's essay are not auditing a sound. They are standing inside a vibration, and their own bodies are the receiving instrument, whether or not anyone ever told them that.
The listener is the instrument that never gets calibrated
There is a practical argument hiding inside all of this, and it is the one the essay gestures at without quite landing on. If grief sounds are physiology, then attending to them is not a soft skill. It is a technical one. The clinician's own body is the instrument: the steadied breathing, the unhurried sentence, the willingness to stay in a room whose sound the chart cannot hold. Like any instrument, it can be taught and it can be calibrated, and it decays when it is never used. The research on crying suggests the family's wail is doing regulatory work. The clinician's presence does the same work in the other direction, and it does it better when the clinician knows that staying, breathing, and listening are not extras. They are the intervention.
None of this requires romanticizing grief or claiming that hospitals should become temples of sentiment. The case is narrower and harder to dismiss: the sound carries information, the information is currently discarded, and discarding it costs everyone in the room. The patient's family learns, in the worst hour of their lives, whether the person across from them can bear the sound they are making. What they learn shapes everything that follows, and nothing about it is written down.
Cherston ends her essay with a question: whose whale songs will she sing, and who will sing hers. It is the right question, and it has a clinical answer hiding in it. The sounds will keep arriving, in every unit of every hospital, as they have since hospitals began. The only thing left undecided is whether medicine will ever treat them as what they are: signals, from the oldest medium there is, carrying the most important information in the room, with no column in the chart to hold them.
Primary sources
- Caroline Cherston's STAT First Opinion essay for the ICU account of Jake and his mother, the catalog of grief sounds across clinical settings, the quoted phrase on presence, and the essay's framing.
- The 2024 Nature paper by Coen Elemans and colleagues for the larynx physiology and the overlap of whale song with shipping noise, and Undark's review of long-distance whale communication for the Payne and Webb 1971 transmission findings and the frequency ranges.
- The 2020 Emotion study by Leah Sharman and colleagues for the cortisol, heart-rate deceleration, and respiration findings, with the earlier Vingerhoets and Kirschbaum work providing the cortisol correlation.