Bed Three
I’m a sleep technologist. Four nights a week I work in the sleep lab on the third floor of a hospital in Tulsa, from 7 p.m. to 7 a.m. We have four bedrooms that look like cheap hotel rooms and a control room full of monitors. Patients come in for overnight studies, mostly for sleep apnea, some for narcolepsy or strange behaviors at night. I glue about twenty electrodes to each of them, on the scalp, the face, the chest, the legs. I put a cannula under the nose and belts around the ribs. Then I turn out the lights and sit in the control room and watch them sleep on infrared cameras while their brain waves scroll across my screens.
It sounds dull. It’s actually lovely, most nights. There’s something about watching people sleep that makes you fond of them. You see them at their most helpless. You learn their rhythms. You know when somebody’s about to dream before they do.
The man in Bed Three was named Arnold Teague. He was fifty-four, a pharmacist, referred because his wife said he’d been talking in his sleep for six months. Not mumbling. Talking. In full sentences. In voices that weren’t his.
He was pleasant, a little embarrassed. He said he didn’t remember any of it. He said his wife had recorded him on her phone, and played it for me, and it was a woman’s voice, soft and clear, saying something that sounded like is anybody there. He laughed nervously. He said, “She thinks I’m possessed. I think I’ve got apnea.”
I hooked him up. Lights out at 10:40.
He fell asleep fast, which is common in people who sleep badly at home. I watched his EEG slow down through the stages, the waves getting longer and lazier the way they do, and at about 12:15 he went into his first REM period, his first real dream of the night. On the eye channels, the jittery spikes of rapid eye movement. On the chin muscle channel, flat. In REM your body is paralyzed so you don’t act out your dreams.
And then something happened on the EEG that I’d never seen in eleven years.
I’ll try to explain it simply. Your brain in REM sleep makes a particular pattern on the screen. Fast, low, a little jagged, with these distinctive sawtooth waves. Every person’s is a bit different, the way handwriting is different, but you learn to recognize it the way you’d recognize a word.
Arnold’s REM came up on the screen as normal. And then, on top of it, a second one did.
Not noise. Not an electrode artifact; I know what those look like. A second, complete REM pattern laid over the first, on all six brain channels, with its own sawtooth waves at a slightly different rhythm. It was as if two people were dreaming in the same skull, and the electrodes were picking up both of them. One a bit faster, one a bit slower, rubbing against each other on the screen like two radio stations on one frequency.
I checked everything. Impedances, the ground, the reference electrode. I went in and checked the leads. I checked the room for any electronics that could have been interfering. Arnold slept through it all, his eyes flicking under their lids.
When I sat back down, his lips were moving.
You’re supposed to be paralyzed in REM. You’re not supposed to be able to talk. People who talk in their sleep usually do it in lighter stages, mumbling. Arnold was in deep REM and his lips were moving clearly on the infrared, and on the audio channel, through the microphone over the bed, a woman’s voice was speaking.
Hello? Is anyone monitoring?
I stopped breathing.
Is someone there? This is Bed Three. Please. I can’t wake up.
The voice was calm, the way you’d be calm after a long time of being frightened. Patient. It sounded like somebody who’d been saying the same thing for a very long time and had stopped expecting an answer.
It’s very dark. I can see the red light on the camera. I know somebody’s watching. Please. I’ve been trying to wake up for so long.
Arnold’s chin channel was flat. His body was still. Only his lips moved. On the EEG, the second REM pattern spiked every time the voice spoke, and Arnold’s own pattern went quiet underneath it, like someone stepping back.
I pressed the intercom button. You talk to patients through it when you need to adjust something. My hands were shaking. I said, “This is the technician. Can you hear me?”
His lips stopped moving. There was a long pause. Then, very quietly: Oh. Oh, thank God. Oh, you’re there.
I’m going to give you the rest of the conversation as best I can. The recording exists. I have a copy.
I asked her name.
Delia. Delia Marchetti.
I asked where she was.
Bed Three. The sleep lab. I came in for a study. I think it was a Tuesday. They put the wires on and turned out the light, and I went to sleep, and then I couldn’t wake up. I tried. I can hear everyone. I can hear the technicians. People come and go. I’ve been calling for so long. Nobody ever answers. Why does nobody answer?
I asked her what year it was.
She said 2009.
I didn’t know the name. I’d started at the lab in 2015. The next morning, after Arnold went home, I asked our medical director, Dr. Haddad, who’s been there since the nineties, if she’d ever heard of a patient named Delia Marchetti.
She put down her coffee. She looked at me for a long time. Then she closed her office door.
Delia Marchetti had come in for an overnight study in March 2009. She was thirty-one. A routine study for insomnia. Sometime during the night, she’d gone into REM and never come out. No stroke, no seizure, no cause anyone could ever find. Her EEG simply stayed in REM, the brain dreaming and dreaming, and she would not wake up. She was transferred to the ICU and then, months later, to a long-term care facility. As far as Dr. Haddad knew, she was still there.
“Seventeen years,” Dr. Haddad said. “In a dream state. The neurologists wrote papers about it.” She rubbed her face. “Why are you asking?”
I played her the recording.
I called the care facility that afternoon. I said I was a technologist from the hospital following up on an old case, which was close enough. I asked about Delia Marchetti. The nurse on the phone said she was stable. That nothing had changed in years.
Then she said, “Except the talking.”
I asked her what she meant.
“She’s started talking in her sleep. About six months ago. Which shouldn’t be possible, in her condition. The neurologist came out and couldn’t explain it.” She hesitated. “It’s a man’s voice. It’s very strange. He asks for his wife. He keeps saying he’s a pharmacist and he needs to get to work. He asks what time it is. He seems confused.”
I asked her if he’d said a name.
“Arnold,” she said. “He keeps asking for someone to call his wife and tell her it’s Arnold.”
Arnold Teague’s follow-up appointment was two weeks later. He didn’t come. His wife called to cancel. She said he was doing much better. She said the talking had stopped. She said he was sleeping like a baby. Then she said, in a smaller voice, that he was different, somehow, since the study. Sweeter. He’d started humming songs she didn’t know. He’d started calling her “honey,” which he’d never done in twenty-six years. He’d taken up cooking Italian food.
I haven’t told anyone. I don’t know who I’d tell, or what they’d do.
But I’ve requested to work Bed Three every night. It’s in the corner, the farthest room from the door, and nobody else on staff likes it, so they let me. I hook up my patients, and turn out the lights, and sit in the control room, and watch the screen.
Most nights, nothing. Normal sleep. Normal dreams.
But sometimes, maybe once a month, when a patient in Bed Three goes into REM, a second pattern rises up under theirs on the screen, faint, a man’s slow and confused sawtooth, rubbing against their dream like a hand against a window. And I lean toward the microphone and listen to whoever’s sleeping there murmur, in a man’s tired voice, Hello? Is anybody monitoring? This is Bed Three. I can’t wake up.
I press the intercom button and tell him that I’m here. I tell him his wife’s name. I tell him I’m sorry. I don’t tell him that she’s happy. I can’t decide if that would be the kindest thing or the cruelest.