34. DID I SLEEP OR NOT? - An Insomniac's Epistemological Struggle

Another day feeling like a zombie. I pull into my parking space at work and can’t believe it’s only Wednesday. Two days after this still to go before the weekend comes, and with it the possibility of a good night’s sleep.

I have never slept well. Some of my earliest childhood memories are of sitting awake in my bed. I had to be six or younger because the bed is in the bedroom of my family’s first house and we moved from there when I was six. I remember cutting a deal with my mother, who previously had not allowed me to leave the bedroom and go downstairs until she or my father were up. On a Saturday morning, I could go downstairs early, at 6am, to watch the Batman cartoons that were showing before the rest of the family woke up at seven. My memories are of being up long before 6am, and for being wide awake in this way for weeks before the deal was made. This has always been the form my insomnia takes - not trouble getting to sleep, but trouble staying asleep. As I got older, the insomnia continued. I remember all kinds of superstitions forming around sleep - hoping new bedcovers would help me sleep a whole night through? Changing from blankets to a duvet? Moving the bed around my room? Better curtains? A better mattress? Listening to music? Talk radio? A white noise machine? The insomnia has been sporadic but permanent; on nights I do sleep well, I am all-too aware another sleepless night could be just around the corner. Eventually, it is.

These days I feel as if I haven’t slept a whole night through for over a year. Since around December of 2018 I keep finding myself waking up every three hours or so, with difficulty getting back to sleep. For someone whose natural circadian rhythms make it near impossible to go to bed, let alone be asleep, any time before 10pm, and whose job requires him to wake up by 6am, my best-case scenario has always been about seven hours of sleep, but these days I’m lucky in the week to get more than five. The weekends are better only because I don’t have to wake up so early - when I’m wide awake at 5:30am I can afford the time it takes to fall back asleep again and catch a few more hours of rest before I need to wake up. Of course I have tried new superstitions - meditation techniques, reading before bed, no screens - but while they may work for a few nights, soon enough it’s 3am and I’m wide awake again.

At least this is the story I keep telling myself.

I know that I lie awake at 3am most nights because I remember doing it. Stress is a major factor and I remember the triggers that start my mind racing when it should be sleeping (usually work things). I have clear, tangible memories of lying there, working through problems I don’t need to be working on in the middle of night but which my brain won’t let go of. I can see the time as I look at my watch or my phone.

But I also have another, contrasting, piece of information: my Fitbit.

As I sleep, a device on my wrist records my heart-rate and other data and tells me not only how much I slept, but what kind of sleep I got. Deep sleep, REM, light sleep. It also shows me when I was awake. Sure enough, some of those 3 o’clock awakenings are recorded on the daily graph which shows me my “sleep score”. A big red bump at 3am signifying the disruption to my slumber. And the accuracy can be uncanny. Nights when I am awakened mid-dream and lie in bed trying to recalibrate myself to reality after believing I was somewhere else mere seconds ago, the graph on my Fitbit app shows me shoot from REM sleep (where we dream) to awake. Nights without such dreams the line goes upward out of light, dreamless sleep. So what I’m saying is that I trust the Fitbit’s data. Sort of.

Because other nights I am sure I was lying awake for hours - have clear, lucid memories of the experience - and my sleep report tells me I am wrong. While I did wake up at the time I remember, the red line representing that fact is tiny. A brief, momentary awakening, not the hours I remember. In fact the report tells me that for the eight hours I was “asleep”, I was only awake for 38 minutes the whole night, achieving a pleasing seven hours and twenty-two minutes of rest. This, despite the fact that I am sure I didn’t sleep a wink. I tossed and turned all night, looking at the clock at three, four, five, six…

Other times, I wake up knowing I slept very little. Not because I woke up a lot, but because I went to bed too late. The alarm went off, it’s 6am, and I wasn’t asleep until 1am. Five hours only. But I feel great! Again, the Fitbit explains: for some reason my fatigued body managed to plunge into extensive deep sleep. The night before I only managed twenty-eight minutes of deep sleep; tonight it was one hour and twenty minutes of the stuff.

The more I awake to see reports from my Fitbit either conflicting with my personal experience of the sleep I got, or informing my personal experience with a level of understanding that was absent from it, the more I realise I don’t know how well I slept, or even if I did sleep, as well as this machine on my wrist.

But at the same time, this machine is not clever. I can fool it easily when it tries to count my footsteps. I can sit at my classroom desk and gesticulate wildly as I speak for a few lessons from the comfort of my chair only to be told by watch face that I have walked thousands of steps. I’ve lost count of the number of times I have sat clapping in an assemblies, or at the theatre, and watched the step-count go up, up, up as I sit in place, my feet going nowhere as the arm on which my Fitbit rests flaps wildly.

Which leads to a further problem: if I can’t trust this machine to count something as simple as steps, why do I trust its data when it comes to sleep? Because I most definitely shouldn’t. I’ve lain awake some mornings and not picked up my phone for half an hour or so only to see my Fitbit proudly tell me I was only now waking up. Likewise, I have fallen asleep in front of the television downstairs hours before going to bed and the machine on my wrist appears to be oblivious to the fact I was asleep at all.

Which means I can’t trust the Fitbit either.

So I can’t completely trust myself on the question of how much sleep I got, or even if I did sleep, nor can I completely trust the Fitbit. We both provide best guesses based on inferences - mine on memories of being conscious when I should be unconscious, the numbers I recall from clocks, the frequency of remembered disruptions; the Fitbit’s on heart rate, my physical stillness, and the general hope that the device is properly attached to my wrist and not on too loose or too tight. And inferences merely suggest a truth, they can’t guarantee them. It is a good rule of thumb, as Descartes famously taught, not to completely believe something which has proven itself to get it wrong in the past. Both the Fitbit and myself have been spectacularly wrong about our data on numerous occasions, so can our inferences even be trusted?

Just last night, I fell asleep watching RuPaul’s Drag Race with my wife. My Fitbit recorded this sleep, as it happens, but my own brain struggled with it. My wife had noticed I’d fallen asleep and so paused the episode of Drag Race and flicked through the channels instead, stopping on the movie The Green Mile. At some point, I woke up and saw the movie on the screen, only my brain could not compute what was going on. It thought I was still awake and watching Drag Race, so as I watched the screen, half awake, half dreaming, my sleepy brain attached the faces of the Drag Race competitors to the actors in The Green Mile, making me believing I was watching a drag parody of the Stephen King film until I finally woke up completely and remarked to my wife that this long scene with Tom Hanks complaining about a urinary tract infection didn’t seem very funny and seemed to go on too long to have been included between the parody sketches. She had no idea what I was talking about, and as I continued I realised I had mistaken the tail-end of a dream for reality and performed the Drag Race equivalence of incorporating the sound of an alarm clock as the sound of a fire alarm into my dream.

I hadn’t realised I was sleeping, even as the device on my wrist did. This morning though, I woke up early to watch AEW: Revolution, a wrestling pay-per-view, on my iPad in bed. I watched an hour of it before getting up to make coffee. My Fitbit recorded that hour that I was merrily watching Dustin Rhodes versus Jake Hagar and Darby Allin versus Sammy Guevara as sleep.

For over a year I have felt like a zombie, dragging my insomniac carcass across the working week and building up my sleep debt until each weekend or holiday allows me to replenish a little. But that feeling, it turns out, may be fabrication. And unless I submit to extensive sleep tests, we shall never know how well, or how poorly, I actually sleep. And even then, in the alien environs of a sleep-lab, connected to untypical machines as I try to re-enact an everyday bedtime experience, I would not know what my sleeping is like on a “normal” night, only what it was like in that completely non-typical study.

We often, as philosophers, ask the question to our students: how do we know we are not dreaming right now? I am no longer concerned with our knowledge of dream-states - we know they can be confusing - but with our knowledge of sleep itself. I am not troubled by the prospect that I might be dreaming right now when I think I am awake, but by the idea that I may never truly know how much sleep I have got on any given night. That I may lie in my bed for hours and believe myself to be awake when I am, in fact, sleeping. Or to be asleep but wake up feeling unrested because I got the wrong kind of sleep. Or to get the right kind of sleep, but not enough of it.

When we are, and when we are not, conscious seems to be a fairly fundamental piece of self-knowledge every human being should have access to. The more I worry about my insomnia, however, the more I realise how little about our own unconsciousness we actually know.

Author: D.McKee