9th February 1911. A man sits in a house in Pondicherry and looks at an ant carrying something across the floor. He doesn't move. He doesn't speak. He puts a wish on the ant — drop the object, go back — and he watches. The ant keeps going forward for a moment. Then it hesitates. Then it sets the object down and turns around.1
He writes it down.
That is the strangest thing about the whole episode. Not that an ant changed course — animals do unpredictable things — but that the man who willed it kept a notebook. He logged the will, logged the resistance ("after a short persistence in the forward movement"), logged the result, and dated the entry. The same morning he willed a sick disciple's diarrhea to stop and noted it "fulfilled as soon as made." He willed another disciple to wake, then to get up and bring tea, and noted that the tea came "after a slight resistance lasting 5 to 10 minutes."2 Three entries. One ant, two men, one morning.
This is the Record of Yoga: Sri Aurobindo's private diary of will-experiments, kept across the early Pondicherry years. It reads less like a mystic's journal and more like a chemist's lab book — date, intervention, conditions, outcome, anomaly.
The Record of Yoga is the empirical floor under the entire siddhi account in this hub. Everything else — the architecture of the eight powers, the theology of supramental versus vital, the safety rules — sits on top of this: a man for several years systematically exercised the three power-siddhis (aishwarya, ishita, vashita) on ants, dogs, disciples, strangers, and at the end of the period on world events, and recorded the data with the dispassion of an observer who genuinely did not know in advance whether each attempt would work.3
The structure of a typical entry is the structure of an experiment. There is the intervention named in the technical vocabulary — aishwarya (a will put, working on its own) or ishita (a need or intention without an exerted will) or both in combination. There is the target. There is the result, graded: immediate success, fulfilled as soon as made, succeeded after resistance, partly successful. And there is, crucially, the timing and the anomaly — "succeeded even the time coming correct," "fulfilled under adverse circumstances," "the tamas still struggles to remain."4
What makes it a record and not a boast is the grading. A man inventing miracles writes that everything worked. Aurobindo writes that the will for a thought to begin was "not begun in spite of struggle for many days," and only later adds, in a different ink of attention, "begun."5 He logs his failures and his slow successes alongside his instant ones. That is the signature of someone collecting data rather than building a legend.
Read enough entries and a measurement system emerges. Aurobindo isn't recording yes/no. He's recording a response curve, and three distinct shapes keep showing up.
Immediate. The will lands and the thing happens with no gap. Saurin's diarrhea: "fulfilled as soon as made."6 The nausea while eating: "immediately successful." When the target's nature offers no resistance — a body in distress that wants to settle — the will meets nothing and the result is instantaneous.
Delayed-with-resistance. The will lands, but something in the target pushes back, and the result arrives only after a measurable lag. Moni gets up to make tea "after a slight resistance lasting 5 to 10 minutes."7 Aurobindo names the thing pushing back: inertia, tamas, the dead weight in a nature that doesn't want to move. The five-to-ten-minute lag isn't the will being weak; it's the will grinding through a known obstacle. He's timing the friction.
Requires-renewal. The will lands, the result starts, then stalls, and a fresh act of will is needed to finish. The dog he wills to "shake off its heavy tamas and manifest the new soul": "rapidly successful but the tamas still struggles to remain and the old bhava persists." So the next day he puts "a renewed aishwarya," and that produces the full effect — "the dog is doing what it had never done before."8 One push isn't always one result. Sometimes the system relaxes back and you have to nudge it again.
He even states the rule out loud. Willing a boy, watched through a window, to roll up a mat: the boy looks at the mat, hesitates, walks off. Aurobindo sends "a fresh wave of the will." Now the boy stops, comes back, rolls the mat, puts it away. And the comment in the margin: this is typically what happens with the tamas signature — you have to nudge multiple times.9 That sentence is a finding, not a flourish. It's a generalization drawn from repeated trials about how inert nature responds to will: not in one clean step but in waves, against a backpressure he's learned to expect.
This page is where the siddhi material stops being theology and becomes evidence — or fails to. Every other page in the Siddhis Hub makes claims about powers. This is the one that says: here is the data, judge it yourself.
For the practitioner pages, the Record supplies the texture the abstractions can't. Aishwarya, Ishita, Vashita — the three degrees of will is a clean taxonomy on paper; the Record shows it operating, with ishita and aishwarya combined on a single disciple in a single line. The tamas-resistance signature is named here from primary observation — the five-to-ten-minute lags, the renewal requirement, the margin note about nudging. And the boundary between making something happen and making someone believe — the heart of belief-injection versus will-compulsion — gets its cleanest worked example in the dog-and-bread sequence below.
Above all, this page hands the vault its sharpest case for yoga as evidence-based applied psychology. Aurobindo treats consciousness as a lab and himself as the instrument. The Record is the artifact that lets you take that claim literally instead of metaphorically — and then test whether the method survives the testing.
Take one entry apart all the way down. Mid-February 1911, the dog again.10
First, the simple version. Aurobindo puts an ishita — a bare intention, not even a forced will — for the dog to eat a kind of bread "which it had always refused." The dog, which has reliably refused this bread its whole life, "suddenly began eating with relish after first refusing." A standing habit, reversed by a wish.
Then the version that turns the entry into something else. He puts a will for the dog to refrain from one large piece of bread but to eat all the others. The dog refuses the large piece. So far, so consistent with the will. But now the person feeding it breaks the large piece into small pieces — physically, it's no longer "a large piece." And the dog still refuses. It refuses every fragment of the original large piece while happily eating the others. The will was bound not to the size or shape of the bread but to the bread's identity — this specific loaf, this object, the one the will was placed on. The dog is tracking identity, not geometry.
Here is the move. Aurobindo "projects the thought in the dog" — now this is not the same piece. And at that instant the dog eats the fragments "with relish."11
Sit with what just happened. The first will was a compulsion: refuse this object. The dog obeyed it so faithfully that even smashing the object couldn't fool it — the dog held the will's intent against the physical change. To release the dog, Aurobindo didn't lift the compulsion. He changed what the dog believed it was looking at. He injected a false premise — different piece — and the compulsion, still running, no longer applied, because its object had (in the dog's mind) ceased to exist.
That is two completely different operations on the same animal in the same minute. One forces behavior against the grain of belief. The other rewrites the belief and lets behavior follow. The precision is the point — and Aurobindo names it himself: "the level of intervention, precision, control, and the psychology that one is watching working inside the dog."12 He's not just moving the dog. He's reading the dog's internal model in real time and editing it. The compulsion proved the will was real; the "not the same piece" trick proved he could see why the will held and where its seam was. This single entry is the empirical seed of the whole belief-injection versus will-compulsion distinction.
It's afternoon and nothing in the room is happening. You decide to run one.
You pick the smallest possible target — not a person, an animal, or better, an insect, because an ant has no idea you exist and can't be flattered or shamed into anything. You hold the wish lightly: drop it, go back. Not strain, not teeth-clenched concentration. A clear, quiet intention placed on the ant, the way you'd rest a hand on something.
Then you watch, and here's the discipline: you watch honestly. The ant keeps going. Your impulse is to either give up ("nothing's happening") or to inflate ("well, it sort of slowed down"). You do neither. You note exactly what you see — persists in the forward movement — and you keep the wish steady without forcing it. A few seconds pass. The ant hesitates. Sets the object down. Turns.
Now the part that separates a practice from a daydream: you write it down before you can edit the memory. The date. What you willed. The lag. The anomaly. Aishwarya on ant to drop object — persisted forward briefly, then dropped and returned. Tomorrow you'll do another, and a disciple's slowness will take five minutes, and you'll write succeeded after resistance, 5–10 min, and a thought you've willed for a week still won't have come, and you'll write not begun in spite of struggle — and you'll leave that entry exactly as ugly as it is. The notebook only works if it can hold your failures. The instant you start logging only the wins, you've stopped experimenting and started preaching.
The failure mode here isn't that the will doesn't work. It's that the notebook stops being honest, and once that happens you can no longer tell whether anything worked at all.
The first sign is the disappearance of the response curve. A real record shows immediates, delays, renewals, and outright failures. A corrupted one shows only success — every will fulfilled, every time, instantly. The moment your log reads like that, you're not recording, you're confabulating: writing the will after the result and calling it cause. Aurobindo's notebook is credible precisely because it's full of friction — five-minute tamas lags, things "not begun in spite of struggle," effects that "still struggle to remain." Friction is the texture of truth here.
The second sign is the loss of the anomaly. Aurobindo notes the odd details — that the timing "came correct," that the dog tracked identity not size, that the boy needed a second wave. Those anomalies are what a real observer can't help recording, because they surprised him. When a log goes smooth and anomaly-free, the observer has stopped being surprised — which means he's stopped actually looking and started writing the story he expected.
The third and deepest failure is forgetting that you are the instrument and the recorder, and that the same ego that wants the power wants the result in the notebook. Aurobindo could keep this honest because, on the larger architecture of his yoga, the will isn't his — it's a supramental power moving through a vehicle he's trying to keep empty. A practitioner running these experiments to confirm that he is powerful will produce a flawless, worthless log every time. The notebook stays truthful only as long as the noter doesn't need it to flatter him. The discipline of recording the failure is, structurally, the discipline of ego-dissolution — applied to a diary.
The strongest thing in the Record's favor is the world-scale entries Sraddhalu flags at the end of the early period. The will-experiments graduate from ants and disciples to "circumstances much more complex, on specific individuals at a distance, verified subsequently through the newspaper."13 The Viceroy is ill; Aurobindo puts a will; the man heals; and the next day's newspaper reports a "miraculous healing."14 Here the result is logged and then independently corroborated by a source Aurobindo didn't control. That's the one move that lifts a private diary toward evidence — an outcome predicted/willed in the notebook and confirmed in print the morning after.
And yet the corroboration only goes so far, and the honest reader has to say so. The newspaper confirms the Viceroy recovered. It does not confirm why. The recovery is public; the causal claim — that Aurobindo's will did it — remains entirely first-person. Every link in the chain except the final fact is unverifiable from the outside. This is the rigor-versus-replicability problem in its purest form: the method is rigorous (dated, graded, honest about failure) but irreproducible (one subject, one observer, no controls, the experimenter and the instrument and the recorder all the same man). A single-subject log can be scrupulously kept and still establish nothing transmissible, because no one else can run it.
The deepest open question is whether the "results" need the will at all. An ant changing course, a sick man recovering, a dog's appetite shifting, a boy tidying a mat he might have tidied anyway — each of these has an ordinary explanation, and a believer's attention will find a will-cause behind exactly the events that would have happened regardless. The Record cannot, from inside itself, rule this out. What it can do — and what makes it more interesting than a credulous journal — is that Aurobindo logs the failures and the lags, which a pure self-deception machine would have no reason to invent. That doesn't prove the powers. It proves he was, at minimum, watching honestly. Whether honest watching of a sample of one ever amounts to evidence is the question this page hands forward and does not close.
Follow Patañjali and the powers are an output — siddhis arise when you fix saṃyama (the fused triad of dhāraṇā, dhyāna, samādhi) on an object, and the Yoga Sūtra lists them almost as a warning catalog: here is what comes, here is why you must let it pass. You'd experience them as the smoke off a fire you're trying not to be distracted by. Follow the Tantric and Bhairava lineages already in the vault and the powers are substance — fluids, goddesses, the body as a site where divinity is concentrated and transacted; you'd experience them as something you ingest and become, not something you observe.
Aurobindo's Record sits oddly between both and matches neither. He doesn't treat the powers as smoke to ignore — he sits down and uses them on an ant, deliberately, and writes the result. But he also doesn't treat them as substance to consume; he treats them as natural capacities to be measured, the way you'd measure any faculty whose limits you don't yet know. Following Aurobindo, you'd experience the siddhi as neither temptation nor sacrament but as instrument — a power you turn on, point at a target, and read off the dial. The convergence with Patañjali is that both insist the power must never be the goal; the divergence is that Patañjali won't let you study it without danger, and Aurobindo's whole Record is the study. Where the Tantric account makes the body the laboratory, Aurobindo makes consciousness itself the laboratory and the diary the proof. The tension the vault should keep live: Patañjali's caution exists precisely because the experiment Aurobindo runs is the one that hooks the ego — and Aurobindo's safeguard against that hook is not abstention but relentless honesty in the log.
To the lab notebook — Record When Hot — Study Hot Streaks (business). Hormozi's elite closers do something structurally identical to what Aurobindo does: they record everything, then study the recordings of their hot streaks with surgical attention to find the specific micro-behaviors that produced the lift — the two-beat pause, the early reframe — and engineer them into a daily protocol. The discipline turns intuition into protocol; the unconscious becomes conscious, the conscious becomes automated. That is the Record of Yoga exactly. Aurobindo is running gametape on his own will. He notices that one push isn't always one result, that tamas takes five-to-ten minutes, that the boy needs a second wave — and these are findings about his own peak performance that he's converting into a reliable method. The structural parallel is precise: a single operator, recording his own outputs obsessively, mining the variation to systematize what works. But the handshake exposes a tension neither field sees alone. The closer's gametape is replicable — the recording is external, anyone can watch it, the micro-behavior is visible to a coach. Aurobindo's gametape is the inside of his own consciousness; the "recording" is his own honest report of what he willed and what happened. The business case and the yogic case use the same method — log, study the wins and losses, systematize — but only one of them produces a transferable artifact. The insight that neither domain generates alone: self-experimentation scales to a method only when the recording medium is external to the experimenter. The closer's camera makes the practice teachable; Aurobindo's diary, however scrupulous, locks the practice inside the one nervous system that ran it. The very honesty that makes a private log credible is also what makes it unrepeatable — you can trust Aurobindo's notebook precisely because no one but Aurobindo could ever keep it.
To expectation and the body — Placebo: When Belief Rewires Neurobiology (psychology). This is the handshake that lets us read the Record skeptically without dismissing it, which is the only honest way to read it. The placebo page establishes that expectation is not a trick that fools the body — it is a biological pathway where symbolic meaning becomes neurochemical reality: a sugar pill labeled "painkiller" releases endogenous opioids and actually silences pain at the anterior insula before it reaches consciousness. Belief becomes chemistry becomes absence of symptom. Now look back at the Record. Saurin's diarrhea stops "as soon as the will is made." The Viceroy recovers. The disciple who's expecting Aurobindo's spiritual power, who wants the shift and is "open and eager and yielding consciously," responds far more readily than the inert dog. From the placebo frame, this is not a refutation of Aurobindo — it's a mechanism for some of his hits. A disciple who reveres the man and senses his attention is primed by expectation exactly the way the placebo subject is primed by the labeled pill; the body settles, the symptom lifts, and the will-account and the expectation-account predict the same observable result. The two readings are not enemies. They converge on the data and diverge only on the cause. But the handshake cuts both ways, and here is what neither page sees alone. Placebo requires a believing subject — the symbol must be received. The Record's hardest entries are the ones with no believer in them: the ant, the crow that changes direction at the instant the will touches, the boy watched silently through a window who never knew he was a target. An ant cannot be primed by expectation; it has no model of Aurobindo to assimilate. So the placebo mechanism elegantly explains the disciple-and-dog half of the Record and fails completely on the insect-and-stranger half — which is exactly where the will-claim is strongest and exactly where ordinary explanation (coincidence, projection, the believer finding causes for events that would have happened anyway) is thinnest but never quite zero. The insight the collision produces: the cases an expectation-effect can explain and the cases a will-claim most needs are nearly disjoint. Where Aurobindo has a primed believer, you don't need the siddhi to explain the result; where you'd need the siddhi, there's no believer for placebo to work through — only an ant, a crow, and the possibility you're reading agency into noise. That's not a verdict. It's the precise shape of what would have to be ruled out, and the Record, kept by one man watching himself, can't rule it out from the inside.
Sharpest implication: The Record of Yoga is the most rigorous thing in the siddhi corpus and the least transmissible — and those two facts are the same fact. Its credibility comes entirely from being a private, honest, single-observer log; its evidential weakness comes entirely from being a private, single-observer log. You cannot fix the second without destroying the first.
Generative questions: