Zoomed Vinay Special · First-person editorial
Yoga Calls Them the Five Sheaths. The Upanishad They Come From Never Calls Them That — It Gives Each One a Tail, and Two Sides That Can Mean Wings
Taittiriya Upanishad 2.1 to 2.5, read with you line by line in the Sanskrit and in Max Müller's 1884 and Radhakrishnan's 1953 translations: five selves, each inside the last, from food to bliss, each given a head, two sides that can also be read as wings, a body and a tail. Then what Shankara did with the picture, how later Vedanta turned it into "sheaths", the argument over whether the fifth is a covering at all, and what the yoga-therapy trials built on it do and do not show.
The point of view
In February 1884, at Oxford, Max Müller signed off a volume of Upanishad translations. Deep inside it, on page 55, he translated a word as "arm" and then admitted in a footnote that the Persian translators had read the same passage as the limbs of a bird, "which is not unlikely." That passage is the source of what yoga classes now call the five koshas. Here is the strange part. The passage never calls them sheaths. It calls them selves, each one filling the last. So what does the text actually say, and how did a bird become a set of coverings?
Oxford, February 1884: a professor, a footnote, and a bird
Let me ask you something. If someone asked you, "What are you?", where would you point? Your body? Your breath? Your thoughts?
Hold on to your answer. A short Sanskrit passage, older than anyone can date, gives five answers to that question. One inside the other. And it draws each answer with a head, a right side, a left side, a body and a tail.
Now let's go to Oxford. February 1884. Friedrich Max Müller is finishing the second volume of his Upanishad translations for the series called the Sacred Books of the East. He signs the introduction "Oxford, February, 1884." (Müller, SBE XV, p. lii)
Inside that volume is the Taittiriya Upanishad. And in its second chapter Müller meets a word he has to make a choice about: pakṣa. It can mean the side of something. It can also mean a wing.
Müller chose the human reading. "This is his head, this his right arm, this his left arm, this his trunk (atman), this the seat (the support)." Then he added a footnote. The text, he admits, actually says "the tail, which is his support." And then: "In the Persian translation the different members are taken for members of a bird, which is not unlikely." (Müller, SBE XV, pp. 54–55 and p. 55 n. 1)
An Oxford professor, in print: I translated it as arms, but it might be wings.
Today yoga teachers and even medical trials call that passage the pañca-kośa, the "five sheaths". Layers to be peeled. So let's read the actual lines, check whether they are coverings at all, and then ask what the therapy trials built on them show. I'll tell you where the record is solid, and where it isn't.
Seven numbers to hold before we start
- 5 selves, 5 parts each. Food, breath, mind, understanding, bliss; each with head, two pakṣas, body and tail. (Taittiriya Upanishad 2.1–2.5, Sanskrit; Müller, SBE XV, pp. 54–57)
- 0 times the word kośa, "sheath", appears in those five passages, or in the Bhrigu chapter that repeats the ladder. The Upanishad's word for each layer is ātman, "self". (TU 2.1–2.5 and chapter 3, Sanskrit text)
- 5 trips. In chapter 3, Bhrigu asks his father "Sir, teach me Brahman" five times. (TU 3.1–3.6; Müller, pp. 64–66)
- 10 steps of a hundred. One human bliss, multiplied by a hundred ten times: a one followed by twenty zeros. (TU 2.8; Müller, pp. 59–61)
- Sixth century BCE, or eighth to seventh. Müller's date (1884) for the classical Upanishads in general, and the date Radhakrishnan (1953) reports as the scholarly consensus for the oldest prose Upanishads, the Taittiriya among them. Nobody has a year. (Müller, p. li; Radhakrishnan, The Principal Upanishads, p. 22)
- 15 trials, 1,048 people. Cochrane's review of yoga for asthma: small improvements in quality of life and symptoms, no clear effect on lung function. (Yang et al., Cochrane 2016)
- 120 adolescents, 8 weeks. A 2026 school trial whose yoga module was built on "the Pancha Kosha framework from the Taittiriya Upanishad." (Rao, Bhat and Sharma, Frontiers in Public Health 2026)
First, the words you need
Taittiriya Upanishad (Taittirīya Upaniṣad). A short text of the Taittiriya school of Vedic priests. Müller notes it sits inside a longer work, the Taittiriya Aranyaka, as its books seven to nine. (Müller, p. xxvii) It has three chapters, called vallīs.
Brahmananda Valli and Bhrigu Valli. Chapters 2 and 3: the five selves, then the same ladder told as a story. (Müller, pp. 54, 64)
Brahman. These texts' word for the ultimate reality. Not the god Brahmā. Chapter 2 opens by defining it: satyaṃ jñānam anantaṃ brahma: Brahman "the real, as knowledge and as the infinite." (TU 2.1; Radhakrishnan, pp. 541–542)
Ātman. "Self". Each layer is an ātman; so is the middle of each layer's body plan.
-maya. "Made of". Annamaya, made of food; ānandamaya, made of bliss.
Pakṣa. "Side" or "wing". Puccha. "Tail", paired with pratiṣṭhā, "foundation, support".
Kośa. "Sheath", like a sword's scabbard. The word the later tradition uses for these layers. Pañca-kośa: "the five sheaths".
The claim we are testing
Here is what you will hear in a yoga class:
"The Taittiriya Upanishad teaches that every human being has five sheaths, the pancha kosha: physical, energy, mental, intellectual and bliss. Peel them away and you find your true self. And yoga therapy heals all five."
I liked this sentence. That is exactly why I checked it.
- Five layers, one inside the other, from food to bliss. TRUE. Taittiriya 2.1 to 2.5 says exactly this.
- The Upanishad calls them "sheaths". NOT IN THE TEXT. It calls each layer a self and says the inner one fills the outer. Kośa for these five appears in Shankara's commentary and later handbooks.
- Each layer is drawn as a bird. PARTLY TRUE. The text gives each self a man's shape, then wings or sides and a tail. Müller made them arms; the Persian translators, a bird; a later teacher compared them to the bird-shaped fire altar.
- Bliss is the fifth sheath, to be peeled away. CONTESTED. That is Shankara's final reading. The Brahma Sutra on its face, and the commentator Ramanuja, say the self made of bliss is Brahman itself.
- Yoga therapy heals all five sheaths. UNSUPPORTED AS STATED. Trials test yoga programmes designed with the model. Some show real improvements. No trial measures a sheath.
And the true part is more interesting than the slogan.
Five selves, each filling the last
The passage starts with creation. From the Self comes space, then air, fire, water, earth, plants, food, and from food, the person. (TU 2.1; Radhakrishnan, p. 542) Then the first layer:
sa vā eṣa puruṣo 'nnarasamayaḥ. tasyedam eva śiraḥ. ayaṃ dakṣiṇaḥ pakṣaḥ. ayam uttaraḥ pakṣaḥ. ayam ātmā. idaṃ pucchaṃ pratiṣṭhā.
Radhakrishnan: "This, verily, is the person that consists of the essence of food. This, indeed, is his head, this the right side, this the left side, this the body, this the lower part, the foundation." (TU 2.1; Radhakrishnan, p. 542)
In plain words: here is you, made of food. Notice "this", again and again. The teacher is pointing at a real body.
Then comes the move that repeats four times. Müller:
"Different from this, which consists of the essence of food, is the other, the inner Self, which consists of breath. The former is filled by this. It also has the shape of man." (TU 2.2; Müller, p. 55)
Read it again. "The former is filled by this." In Sanskrit, tenaiṣa pūrṇaḥ, "by that, this one is full."
The text does not say the body covers the breath. The breath fills the body, like air in a balloon. Hold on to that. It is the biggest difference between the text and what we say about it.
The breath-self gets its own five parts. Prāṇa, which Müller calls up-breathing, is its head. Vyāna, back-breathing, its right side. Apāna, down-breathing, its left. Ākāśa, the ether or open space, its body. The earth, its tail and foundation. (TU 2.2; Müller, p. 55)
Now run the ladder.
The self made of mind (manomaya). Its head is the Yajurveda, its right side the Rigveda, its left the Samaveda, its body ādeśa, "teaching", its tail the hymns of the Atharvans and Angirasas. (TU 2.3; Radhakrishnan, p. 544) For a Vedic student, the mind is where the memorised Vedas live.
The self made of understanding (vijñānamaya). tasya śraddhaiva śiraḥ. ṛtaṃ dakṣiṇaḥ pakṣaḥ. satyam uttaraḥ pakṣaḥ. yoga ātmā. mahaḥ pucchaṃ pratiṣṭhā. "Faith is its head, the right the right side, the true the left side, contemplation the body, the great one the lower part, the foundation." (TU 2.4; Radhakrishnan, p. 545) Müller gives yoga here as "absorption". (p. 57) So the word yoga sits in the middle of the fourth self. Remember that later.
The self made of bliss (ānandamaya). tasya priyam eva śiraḥ. modo dakṣiṇaḥ pakṣaḥ. pramoda uttaraḥ pakṣaḥ. ānanda ātmā. brahma pucchaṃ pratiṣṭhā. Müller: "Joy is its head. Satisfaction its right arm. Great satisfaction is its left arm. Bliss is its trunk. Brahman is the seat (the support)." (TU 2.5; Müller, p. 57)
Listen carefully to that last line. The tail, the foundation, of the fifth self is Brahman. The whole later argument hangs on it.
Between the layers sit short verses. The most famous: yato vāco nivartante, aprāpya manasā saha; ānandaṃ brahmaṇo vidvān, na bibheti kadācana. "He who knows the bliss of that Brahman, from whence all speech, with the mind, turns away unable to reach it, he never fears." (TU 2.4; Müller, p. 56)
A man with wings: where the bird comes from
So is it a man or a bird? The honest answer: the text's words allow both. The first self is the person himself. Every inner self, it says, is puruṣavidha eva, "of the shape of a man". And then it gives each two pakṣas and a puccha. A man with sides, or wings, and a tail.
Müller made the wings arms; Radhakrishnan, "sides". (Müller, p. 55; Radhakrishnan, p. 542) Both keep the bird out of the translation itself; Müller mentions it only in his footnote.
But Radhakrishnan's own note brings it back. "The bodily sheath is conceived in the form of a bird," he writes, quoting Sureshvara, a disciple of Shankara: "The sacrificial fire arranged in the form of a hawk or a heron or some other bird, has a head, two wings, a trunk and a tail. So also here every sheath is represented as having five parts." (Radhakrishnan, p. 543; Sureshvara as Shankara's disciple: Dhole, Panchadasi, p. 53 n.)
Go to the ritual text of the same Taittiriya school, the Taittiriya Samhita. Book 5 lists special altars. One line: śyenacitaṃ cinvīta suvargakāmaḥ; śyeno vai vayasāṃ patiṣṭhaḥ; śyena eva bhūtvā suvargaṃ lokaṃ patati. In plain words: the man who wants heaven should pile his altar as a falcon. The falcon is the best flier among birds; becoming a falcon, he flies to heaven. (Taittiriya Samhita 5.4.11.1, Sanskrit; my plain rendering) The very next line offers an altar shaped as a kaṅka, a heron, to the man who wants "a head" in the other world.
Hawk or heron. Exactly Sureshvara's two birds.
Students of a school whose priests built birds out of bricks heard this Upanishad. When their teacher said "this is his right pakṣa", I think they saw the altar. That is my reading, and only a reading. The Upanishad never says "bird".
And there is one more voice, the most famous commentator of all. Shankara, the great teacher of Advaita Vedanta, whom nineteenth-century scholars placed in the eighth or ninth century CE. (Jacob, Vedantasara, 1888, p. 43 and n. 4, citing Colebrooke, for the eighth; Müller, SBE XV, p. xii, for the ninth) He reads the first self as a man facing east: right hand the "southern side", left hand the "northern side". And the tail? "The puccha (here) is that which is comparable to a tail, on the analogy of hanging behind, as does the tail of a cow." (Shankara on TU 2.1, tr. Gambhirananda, Eight Upanishads vol. 1, p. 305)
A cow's tail. Not a falcon's.
Three readers, three animals. The text holds all three, and nobody can tell you which the first teacher meant.
Pointing at a branch to show the moon
Here is what I love about Shankara. He tells you why the text uses a body plan. His problem is a student who thinks he is his body. You cannot hand that student the innermost Self with nothing to hold. His mind "cannot without the support of some distinct object, be suddenly made contentless". So the text does what you do when you show a child the new moon. You point at a branch first. "Therefore, on the analogy of the moon on the bough, ... the text takes the help of a fiction." (Shankara on TU 2.1, tr. Gambhirananda, pp. 304–305) The moon is far away, the translator's footnote explains, but we say it is "on the bough" because it looks near it. (p. 304 n. 1)
So head-wings-tail is the branch, a teaching device. Start with the body the student believes in; then the next has the same shape. How does each inner self get that shape? "Just as an image takes its shape from molten copper poured into a crucible." (p. 305) The body is the mould.
And then, at the second layer, Shankara uses the word that took over everything. The scripture wants to reveal Brahman "by following a process of eliminating the five sheaths, just as rice is extracted from the grain called kodrava that has many husks." (Shankara on TU 2.2, tr. Gambhirananda, pp. 308–309) In Sanskrit, pañcakośāpanayanena, "by removing the five sheaths".
There it is. Pancha kosha. Not in the Upanishad. In its commentary.
Kodrava is kodo millet, a small grain with a tough husk. If you have seen grain pounded in a village mortar, you have seen Shankara's picture.
But notice what has happened. On the same page, two pictures sit side by side. The breath-self fills the body, Shankara says, "just as bellows are filled with air." (p. 309) That is the Upanishad's picture: filling. And the five layers are husks to remove. That is the new picture: covering.
Same five layers. Opposite directions. In one you look in and find each layer full of the next. In the other you strip off what hides the kernel. The later tradition walks down the second road.
A son goes back to his father five times
Before we follow that road, go to chapter 3, the most human page in the text.
A young man, Bhrigu, son of Varuna, goes to his father. adhīhi bhagavo brahmeti. "Sir, teach me Brahman." (TU 3.1; Müller, p. 64)
The father does not give him a definition. He names a few things, "food, breath, the eye, the ear, mind, speech", and then gives him a test. "That from whence these beings are born, that by which, when born, they live, that into which they enter at their death, try to know that. That is Brahman." (Müller, p. 64)
So Bhrigu practises tapas, hard concentrated effort, and comes back with an answer: food. Everything is born from food, lives by food, returns to food. Makes sense, no?
He goes again to his father. "Sir, teach me Brahman." And the father says only: "Try to know Brahman by penance, for penance is (the means of knowing) Brahman." (TU 3.2; Müller, p. 64)
No "well done". No "wrong".
Breath. Back to the father. Same reply. Mind. Same reply. Understanding. Same reply. (TU 3.3–3.5; Müller, pp. 65–66)
Five times this boy walks to his father's door. Four times he brings his best answer, and four times the father sends him away with the same sentence. Every parent knows how hard that is. It is so much easier to just tell.
And then: ānando brahmeti vyajānāt. "He perceived that bliss is Brahman, for from bliss these beings are born; by bliss, when born, they live; into bliss they enter at their death." (TU 3.6; Müller, p. 66)
This time he doesn't go back. The text just says: this is the knowledge of Bhrigu and Varuna.
Nobody hands Bhrigu the five layers. He finds each one himself. Chapter 2 is the map; chapter 3 is the walk. And in this story bliss is not a husk he throws away. Bliss is Brahman, and he stays. Keep that in mind. It is going to cause a fight.
How happy is Brahman? The text does the maths
One more passage before the fight. Here the text does something I didn't expect. It measures.
It opens with a thought experiment. Müller: "Let there be a noble young man, who is well read (in the Veda), very swift, firm, and strong, and let the whole world be full of wealth for him, that is one measure of human bliss." (TU 2.8; Müller, p. 59)
Multiply by a hundred: the bliss of the human gandharvas, Müller's "genii". By a hundred again: the divine gandharvas. Again: the ancestors in their long-lasting world. Again: gods born in the ājāna heaven. Again: gods by their works. Again: the gods. Again: Indra. Again: Brihaspati. Again: Prajapati. And once more: Brahman. (TU 2.8, Sanskrit; Müller, pp. 59–61)
Count with me. Ten multiplications by a hundred. A one followed by twenty zeros. One hundred quintillion times the joy of a young man who owns the world.
And the twist. From the second rung up, the text adds: "and likewise of a great sage who is free from desires." (Müller, pp. 60–61) In Sanskrit, śrotriyasya cākāmahatasya.
The young man needed the whole earth for one unit. The sage who wants nothing already has every rung. You climb this ladder by wanting less.
The fight over the fifth layer
Now the fight. Is the self made of bliss the last husk, or the kernel itself?
The Brahma Sutra is a set of compressed aphorisms every Vedanta school must explain. One reads, in George Thibaut's 1890 translation: "(The Self) consisting of bliss (is the highest Self) on account of the repetition (of the word 'bliss,' as denoting the highest Self)." (Brahma Sutra 1.1.12, tr. Thibaut, SBE XXXIV, p. 64) On its face: the bliss-self is Brahman.
Shankara, commenting, first explains it exactly that way, with a beautiful image. "Just as when a man, desirous of pointing out the star Arundhati to another man, at first points to several stars which are not Arundhati as being Arundhati, while only the star pointed out in the end is the real Arundhati; so here also the Self consisting of bliss is the real Self on account of its being the innermost." (Shankara on BS 1.1.12, tr. Thibaut, p. 66)
Then he turns around. "But, in reality," the bliss-self cannot be Brahman, because 2.5 says "Brahman is its tail, its support." (Thibaut, pp. 71–72) If Brahman is the bliss-self's tail, the bliss-self is not Brahman. It is the fifth husk.
Thibaut did not hide his view. Shankara first explains the Sutras the natural way, "and the text of the Sutras is certainly in favour of that interpretation," but finally prefers "a different and exceedingly forced explanation". And Ramanuja, the other great commentator, "is at one with" the first reading "in the chief point, viz. that the anandamaya is Brahman." (Thibaut, Introduction, p. xxxiii)
Even the suffix -maya is fought over: "made of", or "abounding in"? (Thibaut, p. 67)
And remember Bhrigu: "He perceived that bliss is Brahman." Radhakrishnan notes that Shankara, commenting on that very line, argues Bhrigu identified the ultimate reality with bliss, though in chapter 2 he puts the real self beyond the bliss-self. (Radhakrishnan, p. 547)
So "bliss is the fifth sheath, peel it away" is one school's reading, opposed by another with the Sutra's plain words on its side. I'm not going to settle a debate this old. I'm telling you it exists.
From five selves to five sheaths: the cocoon and the grass
Now follow the road Shankara's commentary takes. Over the following centuries, the "sheath" picture wins completely.
The later Paingala Upanishad lists them flatly: "the five sheaths made of food, vital air, mind, understanding and bliss." The food sheath "is the gross body"; the sheaths of life, mind and intelligence "form the subtle body"; the bliss sheath "is also the causal body." (Paingala Upanishad II.5, tr. Radhakrishnan, p. 910) The Vedantasara, a handbook its 1888 translator called "the most satisfactory summary of the modern phases" of Vedanta, agrees: "These three sheaths together constitute the subtile frame." (tr. G.A. Jacob, 1888, Preface p. v; p. 61)
Then the Panchadashi, a Vedanta classic attributed to Vidyaranya. His 1899 translator, Nandalal Dhole, places him in the fourteenth century as minister of Bukka I at Vijayanagara, while warning that "so much of fiction has been transplanted on" the life story that he nearly gave up writing it. (Dhole, Panchadasi, 1899, Preface p. 2)
How it explains kośa: "They cover the Atma like a sheath, hence they are designated kosha (sheaths). As the silk-worm after having woven its sac becomes confined and is subjected to much inconvenience, so these five sheaths cover the Atma, render him forgetful of his real nature." (Panchadasi 1.33, tr. Dhole, p. 8)
A silkworm trapped in its cocoon. A long way from "filled".
The way out? A plant again. "As in the plant saccharum munja the tender and new fibres covered by the firmer covering of older fibres can plainly be separated mentally by argument and reasoning," so the Self is separated from its five sheaths. (Panchadasi 1.42, tr. Dhole, p. 11)
Its third chapter even reads the "secret place" of Taittiriya 2.1, Dhole's "cell", as the five sheaths, with "the blissful sac" innermost. (Panchadasi 3.1–2, tr. Dhole, p. 37)
A man with wings. A husk. A cocoon. Fibres in a grass stalk. By the time the word reaches a yoga class, nobody remembers the bird.
The five layers go to hospital
In 1985 the British Medical Journal published a study by R. Nagarathna and H.R. Nagendra. Fifty-three people with asthma trained for two weeks in "an integrated set of yoga exercises", from postures and breathing to meditation and "a devotional session", and were told to practise 65 minutes a day. Compared with 53 matched patients on their usual drugs, the yoga group improved significantly more in weekly attacks, drug scores and peak flow. (Nagarathna and Nagendra, BMJ 1985;291:1077–9, abstract)
Notice "integrated". The same two names are on a 2004 book, Integrated Approach of Yoga Therapy for Positive Health, which researchers at Swami Vivekananda Yoga Anusandhana Samsthana (SVYASA), Bengaluru, cite as the source of their model. In 2014 two of them wrote that the Integrated Approach of Yoga Therapy (IAYT) was "developed by" their institute, "based on the Pancha Kosa concept", combining postures, diet, breathing, meditation, devotional sessions and counselling. And they quoted, with approval, a line worth remembering: "no component singularly can claim to be the IAYT, nor could possibly have the same effects as the whole model." (Jagannathan and Bishenchandra, International Journal of Yoga 2014;7:166–7)
The same letter reports before-and-after improvements in 560 people with diabetes at the institute's residential centre, 2008–2010. But there was no control group, and diet, medication and a calm campus were all part of the stay. (Jagannathan and Bishenchandra, 2014) Numbers like that cannot tell you which part did the work.
Then this year, a proper randomised trial with the Taittiriya in its methods. One school, 120 Grade 8 students with internet gaming disorder: 60 did eight weeks of yoga, 60 kept their normal routine. The module "was conceptualized using the Pancha Kosha framework from the Taittiriya Upanishad," and "practices were selected to target each sheath." The yoga group improved more on gaming-disorder symptoms, anxiety, depression, stress, loneliness and quality of life. (Rao, Bhat and Sharma, Frontiers in Public Health 2026)
That's a real result. But read the authors' own honest limits. Everyone knew who was doing yoga. Outcomes were mostly questionnaires. The control group did nothing special, which "limits the ability to isolate yoga-specific effects from non-specific factors such as social interaction, structured activity, or time away from gaming." One urban school. No long-term follow-up. (Rao et al. 2026, section 4.1)
And the wider evidence? Take asthma. In 2016 Cochrane pooled 15 randomised trials with 1,048 people. Its conclusion: "moderate-quality evidence that yoga probably leads to small improvements in quality of life and symptoms in people with asthma." The change in lung function, the air you can force out in one second, was "not statistically significant." Only one trial was at low risk of bias in every area. (Yang et al., Cochrane Database of Systematic Reviews 2016, CD010346, abstract)
So what do these studies show? Structured yoga programmes, some explicitly designed around the five-layer model, have produced measurable improvements in how people feel and in some symptoms, over weeks.
And here is what they do not show. They do not show that a "breath sheath" or "bliss sheath" exists as something you can measure; no trial measured one. They do not show which layer's practices did the work, and researchers at the institute that developed the model say no single component could have the same effects as the whole. They do not show that yoga replaces treatment; the 2026 authors call it "not a substitute for established psychological treatments." And for asthma, they do not show better lung function.
In these trials the five layers are a design checklist. The evidence is about the programme, not the metaphysics.
Read the actual document
Here is the passage on the second self, as Müller printed it in 1884. Read it yourself first:
"Different from this, which consists of the essence of food, is the other, the inner Self, which consists of breath. The former is filled by this. It also has the shape of man. Like the human shape of the former is the human shape of the latter. Prâna (up-breathing) is its head. Vyâna (back-breathing) is its right arm. Apâna (down-breathing) is its left arm. Ether is its trunk. The earth the seat (the support)." — Taittiriya Upanishad 2.2, tr. F. Max Müller, Sacred Books of the East vol. XV (Oxford, 1884), p. 55
In plain words: inside the body you can see is another you, made of breath. Not wrapped around the body; filling it, the way air fills a bag. It takes your shape. Then the text does this three more times, each time further in. That's the whole teaching in the text's own grammar. Everything else is commentary.
The strongest case for the other side
First, kośa is not a modern invention. Shankara uses pañcakośa in his commentary on this very passage. (Gambhirananda, pp. 308–309) And the Taittiriya itself uses kośa in its first chapter, in a student's prayer: "Thou art the sheath of Brahman, veiled by intelligence", addressed, on Radhakrishnan's reading, to the syllable Om. (TU 1.4.1; Radhakrishnan, p. 530) So "sheath" is in the same text and in the most famous commentary on it. Calling the five sheaths follows a reading more than a thousand years old.
Second, the covering picture may be the text's real purpose. Each layer is "different from this" and "inner". Keep saying "not this, the one inside", and you are removing layers. Shankara's kodrava and the Panchadashi's munja grass are honest pictures of a method the text really uses.
Third, on the bird: Müller called it "not unlikely", and the falcon altar is in the same school's ritual text. A teacher who says "each layer is a bird" is closer to the text than Müller's arms.
Fourth, the 2026 trial was properly randomised, with sealed envelopes and no dropouts, and students' and parents' reports agreed. (Rao et al. 2026) A model that helps design programmes that help people is doing useful work, whether or not anyone can weigh a sheath.
I accept all four. My quarrel is only with "teaches". The Upanishad teaches five selves, each filling the last. The sheaths are a reading, powerful and ancient. You should know it is a reading.
What we could not prove
- A date. The Taittiriya names no king and no year. Müller (1884) put the classical Upanishads in general around the sixth century BCE; Radhakrishnan (1953) reported a consensus of the eighth or seventh for the oldest prose Upanishads, the Taittiriya among them. I could not open Patrick Olivelle's 1996 translation for a modern date, so I give you neither as fact.
- What the first teacher meant by pakṣa. The falcon altar makes "wing" likely to me. Nothing proves it.
- The Persian translators' bird. I did not open that translation.
- Sureshvara's own words. I quote him through Radhakrishnan's note.
- Dates for Shankara, Vidyaranya and the Vedantasara. The first two are estimates from nineteenth-century translators (Müller 1884 and Jacob 1888 for Shankara, Dhole 1899 for Vidyaranya), and they disagree on Shankara by a century; I found none for the Vedantasara.
- The full 1985 BMJ paper and the 2004 IAYT book. I read the paper's abstract (controls "matched"; randomisation and any five-layer design not stated) and cite the book only from the 2014 letter.
- A measured sheath. No study I found claims to have measured one.
Why it still matters today
So why should this matter to you?
Because "what are you?" is still the question. The Taittiriya's answer is not "your body" and not "your thoughts". It is a direction. Go in. Then further in. Each time you think you've reached the bottom, ask again, like Bhrigu.
Because "filled" versus "covered" is not a quibble. One says your body is full of breath, mind and, at the centre, joy. The other says your body hides you. Both are in the tradition. You are allowed to know you have a choice.
And because five short paragraphs from a Vedic classroom have reached hospitals and schools. The honest way to honour that is to keep the claims the size of the evidence.
Verdict
So let's put it together.
The documents prove five selves in Taittiriya 2.1 to 2.5, from food to bliss, each with a head, two pakṣas, a body and a tail, and each filling the one outside it. They prove Bhrigu's five trips and a bliss ladder ten hundredfold steps tall. They prove kośa is not in those passages, that Shankara's commentary brings it in with a grain and its husks, and that the later tradition made the selves a cocoon. And they prove the great commentators disagreed about whether the fifth layer is a covering at all.
They do not prove that the Upanishad "teaches five sheaths". They do not settle whether pakṣa meant wing or side. And no trial I could open shows that a sheath can be healed, because none measures one. What the trials show is that structured yoga programmes, some built on this map, have been followed by modest improvements in how people feel and in some symptoms, in trials with real limits.
If you ask me, the true text beats the slogan. The slogan gives you five wrappers. The text gives you a young man walking back to his father five times, priests building a falcon of bricks, and words that turn back, unable to reach it.
Go back to my first question. Where would you point?
The Taittiriya's answer: point at your head. Now point inside it. Now further in. Keep going. The rest, you decide.
Source Library
Taittiriya Upanishad, Sanskrit text with Shankara's bhāṣya, Brahmānanda-vallī and Bhṛgu-vallī, Sanskrit Wikisource (Taittirīyopaniṣadat/Brahmānandavallī and /Bhṛguvallī), and
sa.wikisource.org
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.
F. Max Müller (tr.), The Upanishads, Part II, Sacred Books of the East vol. XV (Oxford: Clarendon Press, 1884), Introduction pp. xii, xxvii, li–lii; Taittiriyaka-upanishad…
archive.org
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.
S. Radhakrishnan (ed. and tr.), The Principal Upanishads (London: George Allen & Unwin, 1953), p. 22, p. 530, pp. 541–547, p. 910, scan
archive.org
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.
Swami Gambhirananda (tr.), Eight Upanisads, with the Commentary of Śaṅkarācārya, vol. 1 (Calcutta: Advaita Ashrama; first edition 1957), pp. 304–305, 308–309, scan
archive.org
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.
George Thibaut (tr.), The Vedânta-Sûtras with the Commentary by Śaṅkarâkârya, Part I, SBE XXXIV (Oxford, 1890), Introduction p. xxxiii; pp. 64–72, scan
archive.org
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.
Taittiriya Samhita 5.4.11.1, Sanskrit, Sanskrit Wikisource (Taittirīyasaṃhitā (visvaraḥ)/Kāṇḍam 5/Prapāṭhakaḥ 4),)/काण्डम्_५/प्रपाठकः_४
sa.wikisource.org
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.
G.A. Jacob (tr.), A Manual of Hindu Pantheism: The Vedantasara, 2nd ed. (London: Trübner, 1888), Preface p. v; p. 43 and n. 4; p. 61, scan
archive.org
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.
Nandalal Dhole (tr.), A Hand-Book of Hindu Pantheism: The Panchadasi of Sreemut Vidyaranya Swami, vol. I, 2nd ed. (Calcutta, 1899), Preface p. 2; pp. 8, 11, 37, 53 n., scan
archive.org
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.
Śaṅkara, Taittirīya-upaniṣad-bhāṣya, palm-leaf manuscript, Whish collection, Royal Asiatic Society of Great Britain and Ireland, catalogue record and scan
archive.org
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.
R. Nagarathna and H.R. Nagendra, "Yoga for bronchial asthma: a controlled study", British Medical Journal 291 (1985), pp. 1077–1079, abstract
pubmed.ncbi.nlm.nih.gov
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.
Z.Y. Yang et al., "Yoga for asthma", Cochrane Database of Systematic Reviews 2016, CD010346, abstract
pubmed.ncbi.nlm.nih.gov
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.
Aarti Jagannathan and Yuman Bishenchandra, "Decoding the integrated approach to yoga therapy" (letter), International Journal of Yoga 7.2 (2014), pp. 166–167
pmc.ncbi.nlm.nih.gov
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.
Nagalakshmi S. Rao, Raghavendra Bhat and Manoj Kumar Sharma, "Effect of an eight-week yoga program on adolescents with Internet Gaming Disorder in an Indian school setting: a…
pmc.ncbi.nlm.nih.gov
Original citation linked to this published story. Its downloadable Source record is still completing editorial publication.