Before affirmations were affirmations, they were autosuggestion. The word belongs to Émile Coué, a French pharmacist who spent the first quarter of the twentieth century teaching people one sentence — every day, in every way, I’m getting better and better — along with a set of instructions for saying it that were far more precise than anything the modern genre inherited.
The instructions are the interesting part, and they are mostly gone. Coué specified when to say it, how fast, in what tone, with how much feeling, and with what kind of claim. Nearly every one of those specifications runs against what a current affirmations article will tell you. This page covers what he actually prescribed, why he was so insistent about effort, what survived scrutiny and what didn’t, and which of his constraints are worth putting back.
The pharmacist who noticed the placebo
Coué (1857–1926) ran a pharmacy in Troyes, and by his own account the thing that started him off was mundane. Patients seemed to do better when he said something encouraging about the medicine as he handed it over than when he handed over the same medicine in silence. That observation is now unremarkable. At the time it was a loose thread worth pulling.
He pulled it toward hypnosis, studying with the Nancy school, whose figures — Ambroise-Auguste Liébeault and Hippolyte Bernheim — held that hypnotic phenomena were the work of ordinary suggestion rather than of any special force passing from operator to subject. Coué took that one step further, and the step is the whole idea. If suggestion is doing the work, then the operator is optional, because no suggestion has ever taken hold without the person receiving it accepting it first. Every suggestion, on this reading, is really a self-suggestion that happened to arrive from outside. Hence auto.
That is a cleaner statement of the case than most of what came after it, and it’s the same logic underneath self-hypnosis for sleep: the guide is a convenience, not a requirement. His book, Self-Mastery Through Conscious Autosuggestion, reached English readers in 1922, and for a few years he was a genuine phenomenon, touring to large audiences in Britain and the United States.
The method as he actually prescribed it
The instruction was short enough to fit on a card.
Say tous les jours, à tous points de vue, je vais de mieux en mieux — every day, in every way, I’m getting better and better — about twenty times, twice a day. Once in the morning, before you’re properly awake. Once at night, as you’re going to sleep. Say it quietly but out loud enough to hear yourself. Say it in a monotone, mechanically, like a child reciting. Do not try to feel it. Do not concentrate on the meaning. To keep count without spending attention on counting, use a piece of string with twenty knots in it and move your fingers along.
Four constraints are doing the work there, and each one is a live disagreement with modern practice.
The timing is drowsiness, not alertness. Both prescribed windows sit at the threshold where deliberate thought is thinning out, which is roughly the territory the hypnagogic state covers, and roughly what the rest of this cluster means when it talks about the receptive hour rather than the motivated one.
The delivery is flat. Not passionate, not convinced. Coué thought that trying to feel the sentence introduced exactly the effort that spoils it.
The counting device exists to remove decisions. Twenty knots means no part of you is tracking a number, and nothing about the practice requires a choice once it’s started.
And the claim is comparative. Not I am well. Not I am successful. Getting better — a direction with no destination attached, which is a sentence almost nobody is in a position to contradict on any given evening. That last one is the most quietly sophisticated thing in the method, and it’s the first thing the genre threw away.
The law of reversed effort
Coué’s central theoretical claim was that will and imagination are not allies, and that when they conflict, imagination wins. Worse, the harder the will pushes, the more strongly imagination pulls the other way. He illustrated it with a plank: laid on the ground, anyone can walk its length without a thought; suspended between two rooftops, the same plank becomes impossible, not because your legs changed but because you can now vividly picture falling off. Resolving to be steady doesn’t help. It adds a second thing to fail at.
This is why he banned effort from the practice, and it’s where his method and its descendants part company hardest. A great deal of contemporary affirmation advice tells you to say the words with conviction, to force the feeling, to repeat them until you believe them. Coué would have identified that as the failure mode rather than the technique. Straining to believe a sentence keeps your attention fixed on the fact that you currently don’t.
Anyone who has lain awake will recognise the shape of it. Trying to fall asleep is the most reliable way to stay up, which is why the practical guidance in how to fall asleep fast is about lowering arousal rather than pursuing sleep, and why the conditioned monitoring described in sleep anxiety is so self-sustaining. Coué reached the same conclusion from the other direction a century earlier, and it is the most durable thing he said.
The flatness is deliberate and it sounds wrong the first time. That reaction is worth noticing, because it’s usually the modern instruction arguing with the old one.
What the modern genre kept, and what it dropped
What carried forward was the outer shape: a short first-person sentence, said to yourself, repeatedly, on purpose. That’s the whole inheritance.
What went missing was almost everything that made it careful.
The window moved. Coué’s practice lived at the two drowsy edges of the day; the genre relocated it to mirrors, commutes and morning routines, performed at full alertness in front of the part of you most equipped to object. Mirror work is the clearest example of that migration, and it’s a genuinely different technique with a different mechanism, not the same one relocated.
The calibration inflated. Getting better and better became I am abundant, I am a millionaire, I am completely healed. Those are not stronger versions of Coué’s sentence; they’re a different kind of sentence, and there’s reason to think the difference matters. Wood, Perunovic & Lee found in 2009 that repeating a flatly positive self-statement made participants with low self-esteem feel worse rather than better, presumably because a claim far outside what you can accept mostly succeeds at reminding you of the gap. A comparative claim has no gap to point at. This is the same “almost believable” calibration the guidance in sleep affirmations is built around, arrived at independently and about a hundred years later.
The effortlessness reversed outright, into conviction and feeling and intensity.
Honesty requires flagging one place where the modern instruction may be right and Coué may not be. Later practice, including ours, generally holds that some felt sense helps a line land, and flat recitation of words you’re not present for tends to do very little. That isn’t settled, and it may not even be a real disagreement. Coué was aiming at bodily complaints and general wellbeing, where expectancy does most of the lifting. The current genre is mostly aiming at self-concept, which is a claim about who you are rather than about how a symptom will go, and a claim about identity may well need more of you present than a claim about a headache. Different targets, plausibly different deliveries.
What the evidence actually supports
The honest summary is that his mechanism has aged better than his claims.
Coué’s protocol itself has never been through a controlled trial, and the enthusiasm around it in the 1920s collapsed within a few years, partly under the weight of what he and his followers claimed for it. Testimonials of organic disease resolving through repetition were the overreach, and they’re the reason the whole approach was easy to dismiss afterwards. Nothing in the current evidence supports that version, and the same caution applies to the modern restatement of it in the power of the subconscious mind.
What did hold up is the narrower thing. Expectancy genuinely alters how symptoms are experienced, and it is one of the more thoroughly documented effects in medicine. The most pointed result for Coué’s purposes is the open-label placebo work: Ted Kaptchuk and colleagues reported in 2010 that patients with irritable bowel syndrome who were told outright they were receiving an inert pill, and why it might still help, reported more improvement than those given nothing. Deception turned out not to be load-bearing. Something about the ritual, the attention and the stated expectation was.
That is close to what Coué was arguing, and it’s still narrower than what he wanted it to mean. The effects are typically modest, they show up most clearly on self-reported symptoms rather than on disease processes, and none of it licenses the stronger claims. The reasonable position is that a repeated, believable expectation can shift what you notice, what you attempt, and how you feel about both. That’s worth having. It is not what the 1920s posters promised, and it’s the same line the cluster draws in do subliminals work.
Putting the old constraints back
If you want to run the original rather than its descendants, it’s a two-week experiment and costs nothing.
Pick one sentence and keep the grammar. Comparative, not absolute. If your interest is identity rather than symptoms, I’m becoming someone who finishes what he starts keeps Coué’s structure while pointing where you actually want it; I am disciplined does not. The mechanics of that translation are covered in how to reprogram your mind, and the underlying reason it has to be believable is in what is the subconscious mind.
Say it at the two edges. Once as you surface in the morning, before the day has its shape. Once with your head already on the pillow. Twenty repetitions is about forty seconds and does not need to be exact.
Say it flatly, and let go of the outcome while you’re saying it. If you catch yourself checking whether it’s working, you’ve reintroduced the effort the method was designed around. Checking is the plank at rooftop height.
Then leave it alone for a few weeks before you judge. Nothing in this territory reports back quickly, and the expectations worth holding are laid out in how long affirmations take to work. The most common reason people conclude autosuggestion did nothing is that they ran it for four days with their jaw clenched.
The second sample is the night half of the protocol, at the point where the repetition should be carrying itself and you should not be supervising it.
How Murmora fits
We publish this as a sleep affirmations company, which makes Coué something like an ancestor, and it’s worth saying that we think he got two things right that the century after him mostly lost: the hour, and the size of the claim. Murmora is built on both. The lines it writes for you are drawn from what you describe about your own situation, which is what keeps them close enough to your position to be accepted rather than argued with, and they’re paced for the minutes before sleep rather than for a mirror at eight in the morning.
The format is the part where we’d have disagreed with him. He asked for twenty repetitions in a row; our audio is sparse on purpose, a whispered line every few minutes in one of a small set of guide voices, later regenerable in a clone of your own, with quiet between the lines rather than a queue of them. The reasoning is his, though. Nothing you have to work at gets past the part of you that objects, and at that hour there should be nothing left to work at.