You've watched the video. Good lighting, clean audio, a person reciting words they clearly memorized an hour ago — and something about it reads as dead on arrival. No expression moves across their face at the right moment. Their hands don't do anything. The voice has one setting. You believe every word is technically accurate and you believe none of it.
Prathamesh Krisang's diagnosis of that dead feeling is specific: the standard advice for on-camera content runs the process backward.1 Write a script. Memorize the script. Perform the script. Each of those steps happens in the wrong order relative to how authentic expression actually gets produced — and the fix isn't better acting classes. It's reversing the sequence back to how speech naturally moves from the inside out.
The framework he's applying — Parā, Paśyantī, Madhyamā, Vaikharī — already has a page in this vault describing it as a four-level cosmology of sound, descending from silent source to audible word.2 Krisang isn't contesting that map. He's running it forward as a production sequence for on-camera speech, and naming what happens at each rung in plain, applied terms.
Parā — feeling. First, before anything else, you feel something. If you don't feel it, Krisang says, it doesn't come out — or if it does, it isn't authentic.3
Paśyantī — a visual forms. Once you feel something, an image starts assembling in the mind. It doesn't have to make sense to the listener. It only has to be connected to the feeling.4
Madhyamā — the thought articulates. The visual and the feeling become words in the mind — the mental sentence, formed but not yet spoken.5
Vaikharī — expression happens. The words leave the mouth.
Krisang is careful about the timescale here, and it's worth taking the claim exactly as stated rather than rounding it up into something grander: this whole sequence runs in a split second for a fluent, in-the-moment speaker.6 It's not four separate deliberate steps you consciously perform one at a time. It's a real, ordered cascade that happens fast enough to feel instantaneous — which is exactly why skipping the first link (feeling) is so easy to do without noticing, and exactly why the audience can tell when you did.
Set the two sequences side by side and the mismatch is stark.
Krisang's sequence: feel → see → think → speak. Standard content-creation advice: write a script (the end product of thinking) → memorize it (forcing Madhyamā, the mental-word stage, to run in reverse — from spoken words back into memory) → perform it on camera (attempting to fake your way back to Vaikharī without ever passing through Parā or Paśyantī at all).7
Nothing in that second sequence touches feeling until, maybe, an acting class gets bolted on afterward to compensate — because a script-first process produces flat delivery, and the standard fix for flat delivery is teaching people to act like they feel something, rather than arranging things so they actually do.8 Krisang's claim is that this compensating layer is unnecessary once the sequence runs the right direction. When feeling comes first, the rest follows without being separately managed: "my voice gets modulated automatically, my hands are moving automatically, my face expression is changing automatically. I don't have to worry about anything."9 Happy, the face shows it. Sad, the face shows it. No acting course required, because nothing is being performed on top of an absent feeling — the feeling is present and doing the work the acting course would otherwise be hired to fake.
There's a specific failure mode Krisang names that only makes sense once you take the sequence seriously as a real cascade rather than a metaphor: ideas that arrive fully formed and then vanish before they're captured. You've had this happen — a genuinely good idea shows up, you don't write it down, and later, no matter how hard you try, it's gone.
His explanation: the idea got lost in Parā and Paśyantī — it never made it down to Madhyamā or Vaikharī, never became a mental sentence or a spoken one, and so it left no trace to retrieve.10 Expression, in this framework, isn't just communication after the fact. It's the action that makes something concrete enough to exist as a retrievable object at all. This is also, on his account, why watching your own recorded content back functions as genuine self-discovery rather than mere review — seeing yourself speak is watching the concretized result of a Parā-to-Vaikharī cascade you couldn't otherwise access, feedback on who you actually are that thinking alone never generates.11
You sit down to record. The old instinct says: open a doc, write out what you're going to say, read it until it's memorized, hit record. Skip that.
Instead: before the camera is even on, ask yourself what you actually feel about the thing you're about to talk about. Not what you think about it — what you feel. Sit with that for a second, long enough for an image to show up unbidden. It doesn't need to be a sensible image. It just needs to be there, connected to the feeling.
Then hit record and let the words assemble themselves in real time, one sentence arriving a half-beat before it leaves your mouth — which is, per Krisang, literally what's happening even in fluent, unscripted speech, just too fast to notice consciously.12 Don't reach backward to check the sentence against a plan. If a thought doesn't fully form until it's already halfway out of your mouth, that's the sequence working, not failing. The face, hands, and vocal tone are not separate tasks to manage during this. They are downstream of the feeling you started with, and per Krisang's account they'll move on their own if the feeling at the top of the chain is real.
This page's entire evidentiary base is one auto-generated transcript, all claims [PARAPHRASED]; the Sanskrit terms are phonetically mangled in the source audio ("parapasanti madama vari") and reconstructed here against the vault's existing four-levels-of-speech page.13 Krisang's claim that expert delivery genuinely completes the full Parā-to-Vaikharī cascade in a literal split second is a strong, specific, testable-sounding claim that the source doesn't actually test — it's offered as introspective report, not measured. Treat the sequence (feeling before image before thought before word) as the load-bearing claim, and the split-second timing detail as a vivid but unverified elaboration on top of it.
An open tension worth naming: Krisang teaches specific meditations in his paid courses to make this process "more powerful," which he declines to describe publicly in the source material.14 That's a legitimate business boundary, but it also means the page can describe the sequence and the diagnosis without being able to independently evaluate the practice he claims closes the gap between knowing the sequence and reliably producing it on demand.
Set next to the existing classical-grounding page on the four levels of speech, the convergence is close to exact and the divergence is a difference of altitude, not content. The classical page treats Parā as a genuine metaphysical destination — silent, transcendent, the source speech descends from — and frames the practitioner's task as a contemplative ascent back toward that silence, a project with no obvious finish line.15 Krisang collapses the same four-rung structure into a production pipeline for a fifteen-second video and treats it as something a person runs, start to finish, most days, without characterizing Parā as unreachable or infinite. Where the classical page's central image is a waterfall you climb, Krisang's is a signal you route correctly. Neither reading invalidates the other — a cosmology and an applied technique can share a structure and serve different purposes — but a reader moving between the two pages should notice that Krisang's version quietly answers the classical page's own open question (is Parā a level you actually reach, or a limit you only approach?) with a practical "reach it constantly, in fractions of a second, every time you speak authentically" — an answer the more contemplative framing never commits to.
Psychology — Felt Sense and Somatic Awareness. Gendlin's felt sense is a pre-verbal, pre-cognitive, whole-body knowing that precedes both emotion-as-category and thought-as-language — the thing you're actually consulting in the pause before you answer "how are you, really?" Set beside Krisang's Parā stage, the overlap is close to total: both name a real, accessible layer of knowing that exists before words form, both treat that layer as more trustworthy than the narrated version that comes later, and both treat the practice of attending to that layer (rather than skipping straight to articulate language) as the actual skill. The difference is application: Gendlin built his concept for therapeutic work, tracking the felt sense to complete an interrupted process; Krisang is routing the identical pre-verbal layer into performance, using it as the first domino in a chain that ends in spoken content. What the two accounts produce side by side: a scripted, memorized video isn't just aesthetically flat — by the felt-sense model, it's structurally cut off from the one layer of information (the pre-verbal, holistic sense of what's actually true right now) that both a therapist and a content creator are independently trying to access, which means "write it down first" isn't a neutral production choice. It's a specific, well-documented way of losing access to the layer that makes delivery feel alive.
Business — The Accusation You Can't Afford to Let Slide. This page documents why being called "fake" is a categorically worse threat to a public creator than being called untalented — talent criticism is survivable, authenticity accusations are corrosive to the entire relationship an audience has with a creator. Read against the script-memorize-perform sequence Krisang is criticizing, the connection sharpens into something close to a mechanism: scripted, over-rehearsed delivery is a structural risk factor for exactly the "fake" accusation this page treats as uniquely dangerous, because an audience's instinct for detecting rehearsed-versus-live delivery is close to the same instinct that flags manufactured authenticity elsewhere (auto-tune, staged spontaneity). The Parā-first sequence isn't only a production technique in this light — it's a defense. A creator who genuinely feels first and speaks from that has nothing to be caught faking, because there's no gap between the felt input and the spoken output for an audience's authenticity-detector to find. The insight neither page states alone: "sound authentic" isn't a delivery style layered on top of content — it's what happens by default when the Parā-to-Vaikharī sequence runs in its real order, and its absence is what an audience is actually detecting when they call something fake, whether or not they could name the mechanism.
Sharpest Implication. The advice to "write a script, then memorize it, then perform it" isn't neutral, low-risk guidance for nervous first-time creators — by Krisang's framework it's a structural guarantee of the flat, robotic delivery it's supposedly protecting against, because it forces production to run in the reverse of the order feeling actually generates speech. The fix isn't more confidence or better acting. It's refusing to write the script at all, and trusting that a felt sequence too fast to consciously manage will produce better delivery than a managed one ever could.
Generative Questions.