The Sitting-Near

In a clearing somewhere in the Gangetic plain, more than two and a half thousand years ago, a man already old sits with his back to a fire that is no longer the centre of the scene. A younger man comes and sits down in front of him, lower, close enough to hear a voice kept deliberately low. No animal is tied for slaughter. No priest stands ready with the measured verses. The grand machinery of the sacrifice, the thing the previous volume of this work spent six chapters describing, has receded into the trees behind them. What passes between these two men is a question and an answer, and the answer is meant for one set of ears.

We do not know their names. In most cases we never will.

This is the scene the Upanishads return to again and again, and it is the scene their very name is supposed to preserve. The first volume of this treatise followed the Veda as sound: how four collections of hymns were kept word-perfect across a hundred generations without being written down, who the seers were, and how the memory held. This volume follows what happened at the far edge of that corpus, in its last and strangest layer, where the question stopped being how to perform the rite correctly and became what the rite was finally for.

The four layers

To see where the Upanishads sit, you have to see the shape of a single Veda. The tradition transmitted each one in four strata, laid down roughly in sequence.

First the Samhita, the collection of hymns itself, the verses addressed to the fire and the dawn and the warrior-god. Then the Brahmana, long prose manuals explaining how those verses are deployed in the sacrifice, what each act means, which formula belongs to which moment. Then the Aranyaka, the “forest text,” from aranya, forest, a body of material understood to be studied away from the settled village, often by those who had withdrawn from ordinary ritual life. And last the Upanishad.

The tradition draws a line through the middle of this sequence. The Samhitas and Brahmanas are the karma-kanda, the portion concerned with karma in its oldest sense, ritual action. The Aranyakas and Upanishads are the jnana-kanda, the portion concerned with jnana, knowledge. The first half tells you what to do. The second half asks what you are.

Because the Upanishads close the corpus, the philosophy that grew from them took its name from their position. Vedanta. Veda plus anta, end. The end of the Veda, in two senses at once: the last textual layer, and the culmination, the point the whole vast structure had been moving toward.

The boundary is not as clean as the scheme suggests. The oldest and largest of all the Upanishads, the Brihadaranyaka, is literally the closing section of the Shatapatha Brahmana of the White Yajurveda. Its name says so. Brihat-aranyaka, the great forest text. The Upanishad does not begin where the ritual book ends so much as grow out of its back pages. The turn from fire to question was gradual, and the texts themselves record it mid-motion.

The question of the name

Here the trouble starts, and it is worth slowing down for, because almost everything written about these texts begins by telling you what the word means, and the confident answer you usually meet is only one of three.

The popular explanation breaks the word into its parts. Upa, near. Ni, down. And the root sad, to sit. Put together: sitting down near. The picture is the one this chapter opened with, a student seated low and close before a teacher, receiving in person what could not be shouted across a courtyard. Max Müller, who produced the first great English edition in the nineteenth century, recorded that most European scholars of his day derived the word this way, “so that it would express the idea of session, or assembly of pupils sitting down near their teacher.”

It is a satisfying etymology. It matches the scenes inside the texts exactly. It may also be a later rationalization of a word whose first meaning was something else.

The Sanskritist Patrick Olivelle, whose modern edition is now the scholarly standard, argues that the oldest attested sense of upanishad in the texts is not sitting at all. It is connection, or hidden equivalence: a secret correspondence between one thing and another, between a syllable and a cosmic power, between the self in here and the reality out there. On this reading the Upanishads are named for what they do, which is to reveal hidden equations, and the schoolroom image of sitting near a teacher attached itself to the word afterward. We will keep that caution in front of us, because the discovery of hidden equivalence turns out to be the central act of the next chapter.

Then there is a third reading, older than the philologists and coming from inside the tradition. Shankara, the eighth-century commentator who shaped how most of Hindu India would read these texts, did not derive the word from sitting either. He took sad in its sense of loosening, wearing away, destroying, and read upanishad as the knowledge that loosens the bonds of ignorance and leads the seeker toward Brahman. For Shankara the word names a power, the saving knowledge itself, and neither a posture nor a correspondence.

Three derivations. A posture, a connection, a destroying knowledge. They are not reconciled in any source, and this treatise will not pretend to reconcile them. What can be said is that each captures something the texts actually do. They are taught in person. They reveal hidden equivalences. And they claim to end ignorance. The word holds all three claims, and the disagreement over which came first is itself a small lesson in how these texts have always been read in more than one register at once.

The texts of the forest and the secret

There is one more name to account for, and it carries the atmosphere of the whole layer.

The Aranyakas, the forest texts that house many of the early Upanishads, went by a second title in the tradition. Rahasya. Secret. These were not the public verses chanted at a royal sacrifice where the point was that everyone heard the fire being fed. These were teachings withdrawn from the settled world, restricted in their transmission, given under conditions.

The restriction is not something later readers projected onto the texts. The texts impose it on themselves. In the Chandogya Upanishad, after the honey doctrine of its third chapter, the text turns and bolts the door behind it: the doctrine, it says, is to be told to an eldest son or to a worthy resident pupil, and to no one who is not a son and no one who is not a pupil. The knowledge has a gate, and the gate has a keeper.

This is why the Katha Upanishad can say, in a verse the third chapter of this volume will come back to, that there are many who do not even hear of this teaching, and many who hear and do not grasp it, and that the one who can teach it is a wonder, and rare. The teaching is not hidden because it is small. It is hidden because the tradition held it to be the largest thing it had, and dangerous to hand over to a mind not made ready for it.

Hold the three things together and the layer comes into focus. A text studied in the forest, away from the village fire. A text whose alternative name is secret. A text transmitted mouth to ear, teacher to seated student, under a rule of fitness. The previous volume described a scripture distributed across millions of mouths so that there was no library to burn. This layer is the opposite motion. Here the corpus narrows to a whisper, on purpose.

Who taught, and where

There is a detail in the Upanishads that the later tradition tended to pass over quietly, and the historian should not. In several of the oldest dialogues, the one who holds the new knowledge is not a brahmin priest. He is a king.

In the Chandogya and again in the Brihadaranyaka, brahmins come to royal courts and are taught by kshatriyas, by men of the warrior and ruling class, doctrines they admit they had not heard before. Pravahana Jaivali instructs brahmins on the fate of the soul after death and remarks that this knowledge had never before passed to the priestly class. King Ajatashatru of Kashi teaches a learned brahmin named Gargya about the self in sleep, and the text notes the reversal openly, that here a brahmin approached a king as a pupil. Above all there is Janaka, king of Videha in the far east of the Vedic world, whose court is the stage for the most important debates in the whole corpus, and who more than holds his own with the greatest teacher of the age.

What to make of this is contested, and the careful answer is that we do not fully know. The scholar Michael Witzel has placed the milieu of the oldest prose Upanishads in exactly these eastern kingdoms, Videha and Panchala and Kashi, on the moving frontier of Vedic culture, in a period of new cities, new wealth, and new patrons. One reading is that the knowledge-teaching genuinely had roots outside the priesthood, in a warrior-class current of speculation, and that the texts preserve a memory of where some of it came from. Another reading is that this is a literary device, the brahmin authors dramatizing the supremacy of the new knowledge by having even kings possess it. The texts allow both readings and settle neither. What is certain is that the Upanishads record their own break with the purely priestly past, and they record it without embarrassment.

The same openness extends, in a tradition often assumed to have shut women out, to a handful of women who appear in these pages as serious minds. The third chapter of this volume will meet two of them. They are in the oldest text, arguing.

What the sitting-near is

So the Upanishads are the closing stratum of the Veda, the jnana-kanda, the place where a ritual civilization turned and interrogated its own ritual. They are called Vedanta because they stand at the end of the Veda and because the tradition held them to be its point. They are called rahasya because they were kept close. And they are called upanishad for a reason that the sources will not finally settle, whether for the sitting-near of student and teacher, or for the hidden connections the teaching reveals, or for the ignorance that knowledge destroys.

What the texts agree on, across every reading of their name, is the shape of the encounter. One who knows, one who does not, seated close, and a question that the fire could not answer being asked in a lowered voice.

The question itself has a content. It is not vague. From the Brihadaranyaka in the east to the Chandogya in the singing schools of the Sama Veda, the same astonishing claim keeps surfacing, stated as plainly as the texts ever state anything: that the reality the hymns had addressed out there, as fire and dawn and order, and the self seated here, asking, are one thing.

The next chapter takes up that claim, and the four sentences in which the tradition fixed it.