The standard instruction in deity practice is to build a figure in the space before you, or above your head. A colour, a posture, ornaments, implements, a face. You hold it until it's stable, then you dissolve it, then you become it.
Millions of hours a year go into that, and the verdict here is that most of them are wasted:
Many people think that in tantric Buddhist practice… you are supposed to visualize the body part clearly in your mind and generate yourself as some imaginary deity that is outside of your body. This is incorrect as that is only an exercise in concentration (visualization) rather than internal Qi transformation. These were instructions given ages ago to throw off practitioners who didn't have a teacher for the correct technique.1
Given to throw people off. Not a simplification, not a preliminary — a deliberate decoy for anyone without a teacher to correct it.
Qi is the body's live energy, used interchangeably with the Sanskrit prāṇa, and the correct target of every visualisation in this system is the practitioner's own flesh.2
The instruction that replaces it is stated once, plainly, with a price attached:
always link any mental visualizations with feelings/sensations in your physical body… otherwise you will lose years of practice effort.3
And the reason, in the flattest terms the book uses: "you can only cultivate the Qi of your physical body!"3 So a picture held in the mind is a picture held in the mind. It builds concentration, which is a real and useful capacity — but concentration is not the thing this practice is for, and the two get confused because they look identical from outside and feel similar from inside.
The correction is specific rather than general. Don't visualise an independent body in space; visualise your own body transforming. And if you're picturing channels opening in your cerebellum, feel the energy "in the anatomically correct position" rather than wherever your mental image happens to sit.3
That last point recurs elsewhere and is the practical heart of it: "In many cases you may visualize a muscle or organ in your head and then find that your visualization does not actually match the anatomical position of the desired target area. Therefore, you must always try to feel the target area being visualized."4
The failure comes in two shapes, and naming them separately is what makes the rule usable. Picture without feeling. "If practitioners just visualize the muscles in their head but don't actually feel them, then they are not really cultivating the Qi and channels of the muscles."4 Sound without feeling. "If they just recite a mantra and append it to a muscle without connecting with the muscle by feeling it, they are also not really cultivating."4 Both are complete-looking practices. Both have the visible components — the image, the syllable, the discipline of sitting down and doing it daily. Both are missing the one component nobody can see, which is contact with the tissue itself.
And the fix offered is almost embarrassingly practical: look at it. "If you look at it during practice while trying to move its Qi, that is even better," and "it is useful to actually look at body parts when doing visualization practices so that you end up actually cultivating that region's Qi channels rather than perfect a mental visualization."4 Eyes open, on your own forearm. Perfecting the mental picture is named as the thing you're not trying to do.
The rule comes with a list of specific practices declared to answer to nothing physical, and it is a long one for a book that teaches tantric method. Counting drops. Visualising drops melting within channels. Red and white bodhicitta. The fire element melting into the water element. All of it: "they do not correspond to any truly existent biophysical processes."3
The drops get a separate demolition and a replacement mechanism: melting substances inside the channels are "just nonsense… a false explanation invented ages ago to guide aspirants in visualization efforts, and the incorrect information has proceeded forwards until now." What actually produces the heat is friction as energy widens a passage.5 So a whole vocabulary — drops, melting, the elements dissolving into one another — is set aside as a teaching device that hardened into doctrine. What is left standing is the instruction to feel a region and get something moving in it.
The strongest version of the position concerns visions, and it's where the book is least comfortable to read:
The supernatural visions one receives during such practices are just illusions created and projected by Buddhas and devas into your brain, otherwise you wouldn't be able to see anything at all. They do not correspond to any real phenomena occurring but are just used as a motivational mechanism.6
And specifically about the drops: "If you ever see drops or bodhicitta in visions, it's just a visualization trick being played on you by devas."5 The framing offered for this is the Lotus Sutra's skilful means, glossed with unusual bluntness as "cheating people in order to lead them out of a burning house," and the same operation is said to run in Christianity, Judaism and Islam through their own angels.6 Which puts a practitioner in an odd position. A vision is not evidence of anything about the world, is not a milestone, and is not yours — it's a piece of encouragement delivered by somebody else into your visual cortex. So the most exciting experiences available in contemplative life carry no information at all, while the beings who produce them are kept.
There's a defence of the discarded method available in the book's own material, and it's worth having. Visualisation builds a real skill. Colouring the fingers red, blue, yellow, green and white during finger practice "not only helps you 'seize' the Qi of each finger, but also builds your powers of visualization which are useful for manipulating Qi when you finally attain the deva body."7
So holding a stable image is training for something later. A practitioner spending years on an external deity is developing a capacity they'll need — they're just not doing the physical work while they do it, and nobody tells them that's a separate job.
There's also the recruitment argument the book makes about everything else it demolishes: the correspondence lists mean nothing and are defended because "unusual correspondences motivate some people to practice who otherwise would not do so."8 A radiant figure in the air is a far better recruiter than a forearm.
The chapter that carries this rule is a set of instructions for Vajrayoginī practice, which makes it a live test rather than an abstraction. The practice is standard on its face. Twelve mantras, all beginning Om Vajrayogini Hum Phat and varying only in a final syllable — Soha, Ah, Oh, Eee, Lah, Hum, Hee, Hah, Hey, Bah, Too.9 Both men and women visualise themselves as the female Buddha; men do it specifically to work on their Yin energy.9
Then the correction lands on the same page. You are not building Vajrayoginī in the space in front of you. You are picturing your own body becoming her, and feeling the transformation in the tissue.3 The deity is a template for a change in your own flesh, not an object of contemplation.
That reframes the mantras too. Each of the twelve "requests Qi assistance from a different family of enlightened Buddhas," and the recommended procedure is to try each one across a day, watch how your energy responds, keep the one that works, and stay with it at least thirty days.9 Which is testing by felt effect — the same criterion the rule demands everywhere.
And the families get a line that's hard to read straight: they "are actually a bit competitive in trying to outdo one another in helping you make progress."9 What the case shows is that the rule doesn't abolish deity practice. It relocates the target. Same mantras, same figure, same thirty days — and the question changes from is my image stable to did anything move.
One further claim sits in the same passage and is asserted rather than argued: six named figures are said to have worn women's clothes while imagining they were women, in order to cultivate the cooling half of their energy.10 No mechanism is offered for it, and it is recorded here as a named claim rather than developed.
Take whatever visualisation you already do and move it inside. If you build a figure in front of you, stop. Build the same figure as your own body — your arms becoming those arms, your colour changing, your posture becoming that posture — and check for sensation in the tissue as you go.3
Open your eyes for part of it. Look at your own forearm, your own instep, your own hand, while you work it.4 The picture in your head can be anatomically wrong; the thing in front of you cannot.
Run the two-part check on every session. Did I picture it? Did I feel it? If only the first, you did concentration training — worth something, not what you came for.4
When a region resists and you want to skip it, that's the moment to stay: "double your efforts and change your tactic by trying to sense the area fully as if from inside it."4 And if you see something spectacular, note it and carry on. On this account a vision is encouragement rather than information, and treating it as a milestone is how practitioners stall.6
Strongest support. The rule is falsifiable in the small: a practitioner can run a month of external-deity practice and a month of own-body practice and compare what happens in their tissue. And the two named half-practices — picture without feeling, sound without feeling — are recognisable to anyone who has done either.
The central tension. Visions are declared informationally worthless and are also said to be deliberately produced by enlightened beings as motivation.6 So a practitioner is told to ignore the most compelling experiences available to them, while being told those experiences are being sent by the very authorities the system tells them to trust. A second tension. The external-deity instruction is called a deliberate decoy for people without teachers.1 Elsewhere the same book releases practices from any need for a teacher at all — no empowerment required, design your own sequence, judge by felt effect. If the decoys exist because unsupervised practitioners are dangerous to themselves, the permissions undercut the diagnosis.
Source notes. 🚩 SINGLE SOURCE — no Vajrayāna teacher or text is cited for the claim that the external-deity instruction is a decoy. 🚩 MOTIVATED REASONING — the argument concludes that a large, well-established body of practice is a blind, and that this author's emphasis is the real one. [POPULAR SOURCE].
Absence note. Drawn from printed pages 66–71, 78, 133 and 144. Nothing here claims what the book says elsewhere.
Open questions. If visualisation trains a capacity needed later, is there a point at which external-deity practice becomes the right thing to do rather than the wrong one? And has anybody compared the two targets deliberately, on the same practitioner, over a comparable period?
Two accounts of what an image is for run through this material and they'd produce different sessions. On the first, the image is a tool aimed at tissue. Its only job is to help you find and move something in a specific region, it's judged entirely by whether the region responds, and a beautiful stable image with no sensation attached is a failure. On that reading a rough picture that produces warmth beats a perfect one that doesn't.
On the second, the image is a skill being built. Stability, colour and detail are the point, the capacity transfers to later stages, and the practitioner who can hold a complex figure has something the one who can't lacks.7 Both are here, and the book teaches the first while conceding the second in a subordinate clause. Nothing says how to split the time.
The convergence with Visualize Then Release is close and instructive. That page has the image as a provisional construction, built in order to be dropped. This one adds the requirement that while it stands it must be anchored — and the two together give the full shape: build it, anchor it in tissue, use it, drop it.
It runs against The Visions Are Fictitious in an interesting way rather than a contradictory one. That page has visions assembled from the practitioner's own cultural stock. This one has them projected in by other beings as motivation. Both deny that a vision reports on anything real, and they disagree completely about who is doing the assembling.
Felt Sense and Somatic Awareness draws a line most people never notice: between thinking about a part of the body and sensing it. They're different acts, they recruit different capacities, and one of them is easy while the other is initially blank. Somebody asked to attend to their own kidney will almost always produce an idea of a kidney's location rather than any sensation from it, and will not be able to tell the difference without being taught to.
That distinction is exactly the one this rule is built on, arriving from a completely different direction. Picture without feeling is the easy act. Feeling is the one that's initially blank and gets mistaken for nothing.
What the pairing gives that neither side supplies alone is an explanation for why the decoy would work as a decoy at all. If the two acts were obviously different, an instruction to build an external figure would fail immediately and practitioners would notice. Because they're easily confused, a person can run the wrong one for years with full sincerity and no signal — which is the precise condition under which a misleading instruction could survive being copied for centuries.
It also explains the eyes-open remedy, which reads as a beginner's crutch and is offered as an upgrade. Visual attention on a real limb constrains the mental image to something anatomically true and gives interoception a target to lock onto. That's using a strong faculty to scaffold a weak one, which is standard practice in somatic training and completely absent from the contemplative instruction as it's usually transmitted.
One handshake. A second candidate — visions as motivational projections — belongs with the vault's existing pages on visionary content.
Sharpest implication. If the external-deity instruction is a decoy for the unsupervised, then the practice that reaches the most people through books and public teachings is the version designed not to work. The correction requires a teacher, and the people most likely to be reading are the ones least likely to have one.
Generative questions.