Most bedtime advice for a racing mind asks you to make your thinking calmer. Breathe slowly. Picture somewhere peaceful. Let the thoughts pass like clouds. Cognitive shuffling asks for something almost opposite: make your thinking incoherent on purpose. Picture a boat, then a bell, then a basket on a doorstep, holding each for a second or two and dropping it before it can become anything.
It sounds like a trick, and in a sense it is one. The idea is that a mind still producing organised, connected, forward-moving thought is a mind that isn’t done with the day yet, and that scattering the thought is a more achievable goal than silencing it. What follows is the actual protocol, the reasoning underneath it, where the evidence genuinely stands, and how to tell whether this is the technique your particular kind of sleeplessness needs.
What the technique actually is
Cognitive shuffling was developed by Luc Beaudoin, a cognitive scientist at Simon Fraser University, who named it serial diverse imagining. The name is more descriptive than the popular one: a series of images, deliberately diverse, imagined rather than merely listed.
The version most people use runs on a seed word. Choose something neutral with no repeated letters, bedtime being the one Beaudoin often uses. Take the first letter and start generating words that begin with it. B: boat. Picture a boat, briefly, in whatever detail arrives without effort, and then let it go. Bell. Basket. Bridge. When that letter stops producing easily, move to the next one. E: envelope, elbow, empty road. Then D, then T.
Three details do most of the work. The images should be concrete objects rather than concepts, because justice gives attention nothing to rest on and a shoelace does. Each one gets a second or two and no more, so nothing has time to turn into a scene with a plot. And they must not connect. If you notice the boat and the bridge assembling themselves into a story, you’ve slipped back into narrative thinking, which is the thing you were leaving.
You’re also not supposed to keep score. Losing track of which letter you’re on isn’t a mistake; it’s roughly the point at which the technique starts working. Most people who fall asleep this way can’t recall where they were.
Why scrambling works when clearing doesn’t
The instruction to clear your mind fails for a well-documented reason. Suppression is effortful, and monitoring whether you’ve stopped thinking about something requires holding that something in mind. Daniel Wegner and colleagues explored this in the late 1980s under the heading of ironic process: trying not to think about a thing reliably increases how often it turns up. Anyone who has lain in the dark trying to stop thinking has run the experiment personally.
So the family of techniques that actually helps at lights-out doesn’t aim at silence. It gives attention somewhere else to be. That’s what the breath count in 4-7-8 breathing does, what the slow sweep of a body scan does, and what the single blank image at the end of the military sleep method does. Shuffling belongs to the same family and differs in one respect: instead of one steady thing to hold, it demands a small new act of generation every couple of seconds.
That difference matters more than it sounds. A monotonous task leaves capacity spare, and rumination is very good at using spare capacity, which is why counting can be done on autopilot while the argument you didn’t win replays underneath it. Having to produce a fresh unrelated image on a short cycle occupies more of the working memory the loop needs.
Beaudoin’s own account goes further than distraction. His proposal is that the brain reads its own thought stream as evidence about whether it’s safe to let go, and that organised, purposeful thinking is a signal to stay awake, since a mind still working on something presumably has a reason. Disconnected imagery resembles what the mind produces on its own as it crosses into sleep, the drifting fragments of the hypnagogic state. On this account you’re not distracting the sleep system. You’re imitating the state it’s looking for.
That’s a theory rather than an established mechanism, and it should be held loosely. What’s not in doubt is the observation it started from: the mentation of sleep onset really is fragmentary and strange, and a mind producing tidy sentences is not near sleep.
Hearing the pace once is more useful than reading the instruction twice. Most people run the shuffle far too quickly on their first try.
Where the evidence actually stands
The honest position is that cognitive shuffling is promising and under-tested.
Most of the direct research on serial diverse imagining comes from Beaudoin’s own group and has been presented in small studies at sleep conferences rather than replicated widely by independent labs. That isn’t a reason to dismiss it, particularly for a low-risk behavioural technique with no cost and no side effects, but it is a reason to be sceptical of any article that describes it as proven. The developer also built an app around the method, which is ordinary and disclosed, and which is exactly the kind of thing to know when you’re weighing the evidence.
The stronger support is one step to the side. In 2002, Allison Harvey and Suzanna Payne tested what people should do with unwanted thoughts before sleep, comparing a specific imagery instruction against general distraction and against no instruction at all. The imagery group fell asleep measurably sooner and reported less intrusive pre-sleep thinking. The general-distraction group, the closest thing in that study to counting sheep, did no better than nothing. The finding that carries over is not that any one script is best, but that attention given a definite task at lights-out beats attention left to fend for itself.
Shuffling and Harvey’s engaging-scene imagery pull in slightly different directions, which is worth noticing. One asks for something interesting enough to absorb you; the other asks for a stream too fragmented to hold. Both beat suppression, and there’s no good evidence ranking them against each other. Which suits you is worth finding out empirically rather than from an article.
Which problem this actually solves
Sleeplessness at lights-out usually has two components, and matching the technique to the right one saves a lot of wasted nights.
There’s the body: heart rate up, a general feeling of being switched on, the sense of lying at attention. And there’s the mind: the replay, the rehearsal, the list. They feed each other, but they respond to different levers. Cognitive shuffling is a mind technique and does almost nothing for physiological arousal. Attempting it while your body is still keyed up tends to produce fast, jumpy switching between images rather than any settling, which then reads as failure.
The practical sequence is to settle the body first. A minute or two of extended exhale from the breathing exercises for sleep guide, or a pass of progressive muscle relaxation if you’re holding tension physically, then shuffle once you’re lying still and there’s nothing left to do but think. That ordering is the same one how to fall asleep fast recommends, and it’s the single most common fix when a technique that works for other people isn’t working for you.
There’s one obstacle shuffling doesn’t touch. If the problem is that being in bed has itself become the trigger, and the thing running is a low-grade alarm about whether you’ll sleep at all, you’re in sleep anxiety territory, where the fix is structural rather than attentional. Deploying a technique while quietly monitoring whether it’s working converts the technique into one more thing to be evaluated, and monitoring is arousal.
The mistakes that stop it working
Four things account for most of the reports that it does nothing.
The images are too interesting. A holiday you’re planning, a person you’re arguing with, the room you want to renovate. Any image with a stake attached recruits the exact machinery you were trying to stand down. Neutral is not a stylistic preference here, it’s the entire mechanism. Boring objects are correct.
The pace is too fast. Rattling through a word list without seeing anything is a verbal task, not an imaginal one, and it keeps the language system running. The instruction is to picture the thing, however loosely, and a second or two per image is faster than most people think and slower than most people do.
The letter rule gets treated as homework. There’s nothing sacred about alphabetical seeding; it exists to spare you the work of choosing, because choosing is a decision and decisions are alerting. If the letters become something you’re trying to get right, drop them and take whatever unrelated object arrives next.
And the shuffle is being used at the wrong hour. A technique applied at lights-out cannot undo an evening spent adding to the pile. If your mind arrives in bed with unprocessed material every night, the leverage is earlier: a short written pass over the loose ends in the early evening, which is what journaling and the constructive-worry approach in the worry guide are for, inside a bedtime routine that gives the day somewhere to end.
The second sample sits at the handover, where chosen images start being replaced by ones that arrive on their own. Noticing that shift is usually the last thing people remember.
The smallest version of this practice
Choose your seed word now, while you’re awake. Bedtime, shoulder, lantern, anything without a repeated letter. Deciding at midnight is a decision, and decisions wake you up.
Tonight, once you’re lying still and the breathing has slowed, start on the first letter. One object, seen for a moment, let go. Then another. Then the next letter when the first stops coming easily. If a real thought interrupts, and it will, don’t argue with it. Note it and return to the next image, the same way returning to the breath is the practice in every attention technique rather than an interruption to it.
Expect to lose the thread. That’s the intended failure. And expect the first two or three nights to feel like effortful admin, because a new attentional habit is clumsy before it’s automatic; give it a week before deciding whether it’s yours. Some people find it immediately natural and some find the generation itself too effortful and do better with a single held image or a voice to follow. Both outcomes are information.
How Murmora fits
We publish this as a sleep affirmations company, so it’s fair to say plainly where our interest lies and where it doesn’t. Shuffling needs nothing from us. It’s free, it takes no equipment, and if it works for you, use it.
What it doesn’t do is put anything in. It’s a scrambler, and it’s built to leave nothing behind, which is the right design for its job and a limitation if you also want the sleep-onset window to carry something. That’s the part Murmora is built for: from what you describe about your own situation, it writes specific lines and paces them for that window in one of a small set of guide voices, later regenerable in a clone of your own. The format is sparse on purpose, a whisper every few minutes rather than a continuous track, which leaves the quiet between them free for exactly the kind of drifting the shuffle is trying to produce. Some people run both: shuffle until the images start arriving by themselves, and let the audio hold the gaps. If your interest is what to put in that window rather than how to clear it, sleep affirmations is the place to start.