Still Someone in There

She had never spoken a word. Then she sang.

Anna Katharina Ehmer spent her whole life — 1895 to 1922 — in a German asylum called Hephata. The staff counted her among the most disabled people the place had ever held. The chaplain who knew her, Friedrich Happich, wrote that she never learned to speak, never seemed to notice the room, soiled herself day and night. Then she caught tuberculosis, and one afternoon the doctor called Happich to her bed. She was singing. Dying hymns, in tune, one verse after another, clear enough to make out the words. She sang, she stopped, she died. The chaplain and the doctor had both watched her for years, and for the rest of their lives they each told it the same way.

Happich said it made him rethink her whole life.

There’s a name for this now: terminal lucidity. The mind coming back — clear speech, real memory — in someone whose brain we had given up on. Two researchers, Michael Nahm and Bruce Greyson, went back through old medical records in 2009 and dug out the cases. Most of the ones they could still find had been written down before 1849, back when a bedside and a good memory were the only tools anyone had.

I’ll be honest about my first reaction, because I think it’s everyone’s. It was warmth. The last clear goodbye. Hold on to it.

That’s the reading I want to throw out.

Because the sweet version quietly skips the hard part. If she came back, then in some way she had been there all along — through the years of silence, behind the broken brain, someone was home. Or she was never there, and a dying body faked a person who was already gone. Those are two completely different worlds. And the singing, on its own, doesn’t tell you which one you’re standing in.

So which is it?

You used to be able to dodge that by calling these stories folk tales. Not anymore. In 2018 the U.S. National Institute on Aging ran a workshop on lucidity in dementia. The paper that came out of it used the phrase paradigm shift and said, in a serious journal, that some of what late dementia takes away might come back for a while. Then the Institute put money on the table to study it for real — and mentioned, almost in passing, that we may have to rethink what personhood means late in the disease. When a government funding body starts reopening the word person, you know the thing has left the ghost story and walked into the lab.

It isn’t common. The best guess is that it happens in about two to six of every hundred deaths. But ask the people who actually sit with the dying — hospice nurses, dementia-ward staff — and most of them say they’ve seen it at least once. Rare for any one patient, normal across a career. That’s exactly how a thing hides from medicine for a hundred years.

Here are the two honest ways to read the return. Both get their best shot.

The first is the last flare. As the body shuts down, its chemistry swings hard — swelling drops, hormones spike, the usual electrical order falls apart — and for a few hours the cells that are left fire together in a way they hadn’t in years. We’ve actually caught dying brains throwing off short bursts of fast, organized activity. On this reading the “return” is a bulb burning brightest the second before it dies. Nothing came back. The light just jumped. This is the easier story for science, and it deserves the benefit of the doubt, because everything else we know says the brain builds the mind.

The second is the filter. What if the brain doesn’t make the mind so much as narrow it down — a reducing valve, as Aldous Huxley put it, or the “transmission” idea William James floated in a lecture back in 1898, later picked up by Bergson and kept alive by a stubborn group of researchers in a book called Irreducible Mind. On this view, dementia is interference, not deletion, and lucidity is the noise dropping away for a moment. It sounds like mysticism. It’s a minority view. And it holds one genuinely awkward card: nobody has cracked the hard problem. We’ve found loads of brain activity that comes with being conscious. We have never found the step that turns three pounds of wired-up fat into the feeling of hearing a hymn. Until someone does, you can’t fully prove the brain isn’t a receiver.

Now guess how you’d actually tell them apart.

You’d need content. If the mind is only being passed through, a clear-headed dying patient might know something their broken senses could never have picked up. If the mind is made by the brain, the clarity should line up exactly with the biology — restored function sitting on restored wiring, nothing more. Nobody has shown the first under real controls. Nobody can yet scan the second in a dying, demented brain. So both stories walk away standing, which is an infuriating place for a question this big to sit.

For what it’s worth, my money is on the last flare. I think the brain makes the mind. I think Ehmer’s song was living tissue firing one last time together, and I’d bet real money on it. I also think anyone who’s certain here is selling something — me included.

Look at what my own bet costs, though. Take the old idea that a person just is a thread of memory and intention running through time. Ehmer’s thread was cut for years. Then, for an hour, it wasn’t. Was that the same woman, briefly stitched back — and if so, where had she been all that time? Or a stranger using her voice? You can answer, but only by picking one, and neither answer feels clean.

And the bedside keeps giving us reasons not to be smug. In 2010, a team led by Adrian Owen and Martin Monti slid patients labeled “vegetative” into a brain scanner and asked them to imagine playing tennis. Some of them lit up the same movement areas a healthy person does — on cue, steadily enough to answer yes-or-no questions with their thoughts alone. These were people we’d said had no awareness at all. The tools we use to declare a mind gone are worse at the job than we admit.

Which is where this stops being philosophy and starts costing money and choices.

“Giving up on someone” is really three questions we keep jamming together. When is it right to stop treatment that only prolongs the body? When is it okay to stop talking to the patient as a person? And when do you stop watching for any sign that someone is still in there? The first is hard but has rules. The other two are where terminal lucidity cuts — because if you treat a maybe-present person as an empty room, you might be building the emptiness you assumed. The living will that a sharp, healthy person signed years ago was written by someone who may not be the one now surfacing, once in a while, to sing.

I don’t have a neat ending here, and I don’t trust anyone who does.

So here it is, plainly. A woman who had never spoken sang on the day she died, and two careful men swore to it for the rest of their lives. Either she came back from somewhere, or a dying body made her up for an hour and then shut off for good. We can’t yet tell which. And every moment we can’t — every quiet ward, every long slow decline, every case we’ve decided is over — we are guessing about who’s home.

Where was she, all those silent years?