Ten names in one breath

Sometime in the eighth century of the Common Era, by the dating most scholars accept as conventional, a monk sat down to write commentaries on the texts he regarded as the foundation of everything. His name was Shankara. He wrote on the Brahma Sutras. He wrote on the Bhagavad Gita. And he wrote on the Upanishads, the texts at the end of the Veda where the tradition turned inward. Count the Upanishads on which commentaries under his name have come down to us, and you get a specific number.

Ten.

Centuries later, a text called the Muktika Upanishad staged a scene in Ayodhya. Rama sits with Hanuman, and Hanuman asks the only question that finally matters in this tradition: how is a person liberated? Rama answers with a curriculum, graded like a physician’s prescription. For the seeker after kaivalya, liberation in absolute aloneness, one text is enough: mandukyam ekam evalam mumukshunam vimuktaye, “the Mandukya alone suffices” (Muktika 1.26). If certainty does not come from that, the next dose: dashopanishadam patha. Read the ten Upanishads. If the knowledge is still not firm, thirty-two texts. And for the seeker of bodiless liberation, all one hundred and eight.

Then a verse names the ten, packed tight so a student could carry the whole canon in one breath:

ईशकेनकठप्रश्नमुण्डमाण्डूक्यतित्तिरिः । ऐतरेयं च छान्दोग्यं बृहदारण्यकं तथा ॥

isha-kena-katha-prashna-munda-mandukya-tittirih | aitareyam cha chandogyam brihadaranyakam tatha

The Isha, Kena, Katha, Prashna, Mundaka, Mandukya, Taittiriya, the Aitareya, the Chandogya, and likewise the Brihadaranyaka. (Muktika Upanishad 1.30)

Muktikā Upaniṣad 1.30 #

Even the compression is telling. Munda for Mundaka, tittiri for Taittiriya: the verse is a mnemonic, built to be memorised, and it was. The same ten as Shankara’s commentaries, confirmed across a gap of centuries from opposite directions. This treatise walks through those ten books, one door per chapter. This first chapter is about the doorframe: who made the list, who disputes its edges, and what can actually be said about when the books behind it were composed.

The word the Muktika never uses

The tradition and the textbooks call these the mukhya Upanishads, the principal ones. It is worth saying plainly that the Muktika never uses the word. It says dashopanishadah, “the ten Upanishads,” and moves on. The label mukhya is a convenience layered on later, by the tradition and by modern scholarship, and there is no harm in it so long as no one imagines an ancient council that ranked the texts. There was no such council. There is a late dialogue, a memorial verse, and a commentator’s shelf.

The Muktika itself is a late work. Paul Deussen, the German scholar who translated sixty Upanishads at the end of the nineteenth century, already grouped it with texts so late that they presuppose the whole corpus as a finished thing to be catalogued. How late, we cannot say with precision; no primary evidence fixes its century, and this treatise will not pretend otherwise. What can be said is that its frame is openly partisan. A text that declares the Mandukya alone sufficient for liberation is speaking from a settled Advaita position that took shape long after most of the ten were composed. The list is old testimony. The ranking is theology.

And the hundred and eight is a liturgical architecture of its own. The Muktika assigns every Upanishad to a Veda, with a peace-invocation for each family: ten to the Rig Veda, nineteen to the White Yajur Veda, thirty-two to the Black, sixteen to the Sama, thirty-one to the Atharva. The Shvetashvatara sits at number fourteen of the 108, the Kaushitaki at twenty-five. The Muktika, in other words, knew both texts well and still left them off the ten. Hold that thought.

Four Vedas, ten books

Every one of the ten belongs somewhere in the Vedic corpus; none of them floats free. The Isha is the strangest case: it is the fortieth and final chapter of the Vajasaneyi Samhita of the White Yajur Veda, which makes it, as Patrick Olivelle notes, the only Upanishad embedded in a samhita, a hymn collection proper, and explains why it opens the traditional order. The Brihadaranyaka closes the Shatapatha Brahmana of the same White Yajur Veda. The Taittiriya forms three chapters of the Taittiriya Aranyaka of the Black Yajur Veda, and the Katha belongs, at least formally, to the same Veda’s Kathaka school. The Chandogya and the Kena are Sama Veda texts, the first from the Tandya school, the second from the Jaiminiya, transmitted inside the Jaiminiya Upanishad Brahmana. The Aitareya sits inside the Rig Veda’s Aitareya Aranyaka. The last three, Prashna, Mundaka, and Mandukya, are ascribed to the Atharva Veda, with a caveat Olivelle states bluntly: Atharvan ascriptions are somewhat suspect, because tradition parked almost every late Upanishad in the Atharva Veda, the most hospitable of the four.

The affiliations matter because they are the plumbing of the whole system. An Upanishad is shruti, revealed scripture, precisely because it belongs to a Veda; its authority is inherited, and the schools of Vedanta that built systems on these texts were, formally, doing Vedic exegesis. The map of ten books under four Vedas is the first thing a reader should carry into the treatise.

The edges of the list

Stand at the boundary of the ten and two texts press against it from outside.

The first is the Shvetashvatara, the great theistic Upanishad of the Black Yajur Veda. Shankara quotes it in his Brahma Sutra commentary as full scripture, calling it “the mantropanishad of the Shvetashvataras” (Brahma Sutra Bhashya 1.1.11, in George Thibaut’s translation), and a commentary on it is transmitted under his name. But that commentary is the most doubted item in the whole Shankara corpus. The doubts are old, going back to the nineteenth-century editors and to G. A. Jacob; Paul Hacker, the scholar who built the modern criteria for testing Shankara attributions, kept his reservations; and the most recent dedicated study, Ivan Andrijanić’s 2019 article in the Journal of Hindu Studies, surveys stylometric work that sides against authenticity. No one doubts the Upanishad. Many doubt the commentary.

The second is the Kaushitaki, of the Rig Veda. Shankara left no commentary on it at all, yet an entire section of his Brahma Sutra commentary (1.1.28 to 1.1.31) turns on its dialogue between Indra and Pratardana. Deussen counted the base passages behind the Brahma Sutras’ first chapter and found the Kaushitaki holding two of the twenty-eight, ahead of texts inside the ten. By the working rule of the man himself, the Kaushitaki was scripture.

This is why the counting never settles. Add the Shvetashvatara and you get eleven. Max Müller’s Sacred Books of the East selection of 1879 and 1884 printed twelve, and Olivelle’s 1998 standard edition also edits twelve, the ten plus Kaushitaki and Shvetashvatara. Robert Hume’s 1921 translation, for a century the most-read English version, added the Maitri and fixed “thirteen principal Upanishads” in its very title. S. Radhakrishnan’s 1953 edition printed eighteen, stating his rule in the preface: the Upanishads “commented on or mentioned by Shankara.” Every list is defensible. Every list is drawing the same circle with a different compass setting. The ten of the Muktika verse remain the stable core inside all of them, and that is the set this treatise walks.

One more honest edge. Within the ten themselves, one commentary attribution is contested: Hajime Nakamura, in his History of Early Vedanta Philosophy, argued that the commentary on the Mandukya and Gaudapada’s Karika transmitted under Shankara’s name is not his. The tradition retains it; many scholars do too. The chapter on the Mandukya will take this up where it belongs.

The honest chronology

When were the ten composed? Here the treatise adopts one sentence as its governing discipline. Olivelle, introducing the standard edition, writes that any dating of these documents “that attempts a precision closer than a few centuries is as stable as a house of cards.” Scholars have offered decade-level datings; Olivelle dismisses the exercise. What survives his scepticism is a relative sequence, and it is worth having.

The two prose giants come first. The Brihadaranyaka and the Chandogya are edited compilations with older material inside them, in all likelihood pre-Buddhist; the seventh to sixth centuries BCE, “give or take a century or so,” is Olivelle’s cautious placement. The Taittiriya, the Aitareya, and the Kaushitaki follow, probably still pre-Buddhist, in the sixth to fifth centuries. Then the verse Upanishads: the Kena oldest among them, then the Katha, the Isha, the Shvetashvatara, and the Mundaka, strongly theistic texts of the last few centuries BCE. Last come the Prashna and the Mandukya, later prose that “cannot be much older than the beginning of the common era.” Deussen’s four-period scheme, drawn up a century earlier on internal evidence, agrees in shape almost exactly, which is why the sequence commands broad confidence even while every absolute number stays soft. Even the anchor wobbles: the whole grid hangs on the Buddha’s dates, and Olivelle notes that on the later chronology of the Buddha’s death now favoured by many, every early Upanishad slides forward by about a century.

Sit with what the sequence means. Perhaps seven hundred years separate the Brihadaranyaka from the Mandukya, the distance between us and the age of plague and parchment. The tradition reads the ten as one voice, shruti speaking; the historian hears a conversation conducted across centuries, later texts answering earlier ones. Both descriptions are true to something, and this treatise will keep both in view without collapsing either.

Books with no authors

One absence should be named before the tour begins, because it will haunt every chapter. We do not know who composed any of these texts. Not one author’s name survives. This is a civilisation that remembered its grammarians, its ritualists, its genealogies of teachers in long formal chains; the Brihadaranyaka itself ends in lists of transmission running back dozens of names. Yet the composers of the books themselves chose, or accepted, anonymity. The texts name their characters, Yajnavalkya and Gargi, Uddalaka and Shvetaketu, Nachiketa and Death, and the tradition holds them to be historical seers. The record can neither confirm nor refute a single one of them. What the record shows is the texts, their Vedic homes, their recensions, and their afterlife in commentary. That is what this treatise has to work with, and it is enough.

The plan from here is simple. Volume I takes the six narrow doors in the traditional order, the short books a reader can hold in an evening: the Isha, the Kena, the Katha, the Prashna, the Mundaka, the Mandukya. Volume II enters the great forest, the four long prose Upanishads embedded deepest in the Vedic corpus: Taittiriya, Aitareya, Chandogya, Brihadaranyaka, and a closing chapter on what the ten amount to when read together. The overview of the whole corpus, for a reader who wants the wide shot first, is the cornerstone guide What are the Upanishads?, and the story of how the full canon of 108 was counted is told in the Unwritten Library’s chapter The Thirteen and the Hundred.

Ten names in one breath. Behind each name, a door.

The next chapter opens the first of them: eighteen verses at the end of a ritual book, beginning with the Lord and ending with a dying man’s prayer.