The Upaniṣads say the self is brahman and they say the self is not brahman. Every other school takes one set of sentences as final and reads the other as qualified. Bhedābheda refuses the choice: both are literally true, difference and non-difference are equally real, and any philosophy that eliminates one of them has simplified the text rather than understood it.
It is the oldest position in the field, it is what Śaṅkara was arguing against, and it never had a single founder — which is most of the reason you have not heard of it.
Bādarāyaṇa reports, by name, three earlier teachers' accounts of the relation between the individual self and brahman — and then gives his verdict. This is the most direct evidence anywhere of what the pre-commentarial tradition looked like.
Bhartṛprapañca wrote a commentary on the Bṛhadāraṇyaka before Śaṅkara wrote his. It is lost. What survives is Śaṅkara's summary of his positions, given in order to reject them — roughly the situation of Cārvāka, and with the same methodological caution attached.
Śaṅkara names him more often than he names anyone else. That is a measure of how much of the field he occupied: the Bṛhadāraṇyaka-bhāṣya reads in places as a running argument with a specific predecessor rather than a commentary on a text.
Brahman is one and is also really many. It undergoes actual transformation into the world of selves and objects, and the plurality is not an appearance. Reality is described in eight modes or aspects — the details are not recoverable — and both unity and multiplicity are essential to it.
Liberation requires knowledge and ritual action together, since both portions of the Veda are authoritative and neither can be set aside. This is the samuccaya doctrine of §8, and Śaṅkara's refusal of it is one of the sharpest breaks he makes.
More than one might expect, for a reason particular to the case. Śaṅkara quotes Bhartṛprapañca's positions in order to argue against them in detail, and a misrepresented opponent's argument is a wasted refutation — the polemical incentive here runs toward accuracy on doctrine, though not on emphasis or merit. What is certainly lost is the reasoning: we have the conclusions Śaṅkara needed to defeat and not the case Bhartṛprapañca made for them.
Yādavaprakāśa's brahman is a real substance that actually exists in three modes: as cit, the conscious selves; as acit, unconscious matter; and as Īśvara, the Lord. All three are brahman, none is an appearance, and the differences among them are natural rather than adjunct-produced.
This is svābhāvika bhedābheda, and it avoids Bhāskara's difficulty: nothing has to be destroyed at liberation, because the distinctness was never a limitation imposed from outside.
Rāmānuja studied under him and broke with him, and the break is instructive. The objection was not to the realism — Rāmānuja keeps that — but to the claim that brahman itself takes on the imperfections of matter and the suffering of selves. If the states are all brahman's own, then brahman really suffers.
Rāmānuja's whole apparatus of body and embodied — see the next page but one — is designed to keep the realism while insulating brahman from what happens in the modes. It is a modification of his teacher's position, worked out in order to escape one objection to it.
Viśiṣṭādvaita is often introduced as though it arose in direct reaction to Advaita. It did not: it arose out of bhedābheda, as a repair to a bhedābheda problem, and the reaction to Advaita came afterwards and in a different register. Reading Rāmānuja without Yādavaprakāśa in view makes the body–soul doctrine look like a theological metaphor rather than what it is — a technical device for holding difference and non-difference together without letting brahman inherit the defects of the differentiated.
Six centuries after Bhāskara, and long after the field had been carved up by the Vaiṣṇava schools, a scholar in Banaras produced commentaries on the Brahma-sūtra, the Sāṅkhya-sūtra and the Yoga-sūtra, and argued that all three teach one doctrine.
| Work | What it does |
|---|---|
| Vijñānāmṛta-bhāṣya | His Brahma-sūtra commentary. Bhedābheda in substance, with the relation described as avibhāga — non-separation rather than identity. |
| Sāṅkhya-pravacana-bhāṣya | His Sāṅkhya commentary, and the most read of his works. Argues that Sāṅkhya's atheism is provisional and its dualism compatible with a supreme self. |
| Yoga-vārttika | On the Yoga-sūtra with Vyāsa's commentary; the same synthesis applied. |
Vijñānabhikṣu is a synthesiser, and syntheses of this kind are always vulnerable to the charge that they are achieved by softening what each system actually says. Sāṅkhya really is atheist — see the Sāṅkhya page, §12 — and reading its īśvarāsiddheḥ as a provisional teaching is a stretch that later scholarship has generally declined to make. What he does supply, and what makes him worth reading, is the fullest late statement of the bhedābheda relation as non-separation: a formulation designed to be weaker than identity and stronger than mere dependence, and the last serious attempt to state the school's central relation in new terms.
He does not name Śaṅkara, and nobody has ever doubted whom he means. The commentators who have gone before, he says, have concealed the sense of the sūtras and set out a doctrine of illusion that is not in the text. And the doctrine is not even theirs: it is the Mahāyāna's, taken over and given Vedic dress.
1 · Māyā is an importation. No Upaniṣad teaches an indeterminable principle that is neither real nor unreal. The word appears; the doctrine does not.
2 · It empties scripture. If the world is sublatable, so are the Veda, the teacher and the injunctions — the whole apparatus by which the doctrine is supposed to be learnt. A ladder that dissolves as you climb it.
3 · It cannot account for difference at all. Calling difference an appearance does not explain it; it renames the problem. Something must appear, to someone, and both terms are exactly what the doctrine denies.
4 · It makes liberation empty. If bondage never occurred, nothing is achieved by its removal, and the entire teaching is addressed to nobody about nothing.
Bhāskara's case is not merely negative. Take the sūtras at face value, he argues, and they teach real causation, real selves, real obligation and a real liberation — which is what a plain reader finds and what the Upaniṣads' difference-sentences require. The burden of proof lies on the commentator who must explain away half his own canon.
| Objection to Bhedābheda | From | The reply available |
|---|---|---|
| Contradiction. One thing cannot be both identical with and different from another | Advaita; and every logician | The respects differ — substance and adjunct for Bhāskara, whole and state for Yādavaprakāśa. And a sea genuinely is and is not its waves |
| Then state the respects, and you have conceded that identity and difference are not in the same respect after all | Advaita | Conceded, and it costs nothing: the claim was never that a single respect is both, only that both are equally real and neither derived from the other |
| Brahman inherits the defects. If the selves are brahman really transformed, brahman really suffers | Rāmānuja, against Yādavaprakāśa | Bhāskara: the suffering belongs to the adjunct-conditioned state, not to the substance. The reply is available and it is thin |
| Liberation destroys the self. Remove the adjunct and there is no one left | Rāmānuja, against Bhāskara | What ceases is limitation, not being. But since the remainder is brahman, there is nobody for whom the outcome is good |
| Difference is not a relation. If it were, it would need a further relation to relate it, and so on | Advaita, following the Buddhists | Difference is the nature of the differing things, not a further entity between them — the standard realist answer, shared with Nyāya |
The second row is the one worth dwelling on. Bhedābheda is often dismissed on the bare charge of contradiction, and that charge is answerable — nobody in the school ever claimed that a single thing is both P and not-P in one respect. What the school is genuinely committed to is something subtler and harder: that neither identity nor difference is more basic than the other, so that neither can be defined in terms of the other or derived from it. Advaita and Dvaita each achieve economy by making one fundamental. Bhedābheda accepts two irreducible relations, and pays in parsimony for what it gains in fidelity to the text.
Bhedābheda is a thesis several people held, not a school one person founded. There is no ācārya to venerate, no maṭha, no succession and no initiation. The Vaiṣṇava sampradāyas that displaced it all have those things, and Advaita has them in the most durable form of all.
Its two live descendants took its best material with them. Nimbārka's Dvaitādvaita is a svābhāvika bhedābheda under a different name, and Caitanya's Acintya-bhedābheda is the same relation with an extra qualification. The position survived; the label did not.
A position that concedes something to each side must fight on two fronts and can be attacked with either side's weapons. Advaita presses the contradiction charge; Rāmānuja presses the defect-inheritance charge. Neither is fatal, and answering both at once takes more machinery than the school ever assembled.
Three reasons. It is the position the Brahma-sūtra reports by name (§3) and arguably holds. It is what Śaṅkara's arguments were actually aimed at, so Advaita's polemic is unreadable without it — a great deal of the Bṛhadāraṇyaka-bhāṣya is addressed to an opponent most readers have never met. And its central methodological claim is a serious one that no other school makes: that a commentary which must demote half its canon has already conceded the interpretive argument, whatever it wins on metaphysics.
1 · What is Bhedābheda's central methodological claim?
And the price it pays is metaphysical rather than interpretive: it must make sense of two irreducible relations holding at once.
2 · What is the difference between aupādhika and svābhāvika bhedābheda?
The attributions are the other way round: Bhāskara holds the adjunct version, Yādavaprakāśa and Nimbārka the natural one.
3 · What is the significance of Brahma-sūtra 1.4.20?
That last is Auḍulomi at 1.4.21. Kāśakṛtsna follows at 1.4.22, and Śaṅkara takes him as the conclusion.
4 · Why does Rāmānuja break with Yādavaprakāśa?
Viśiṣṭādvaita is a repair to a bhedābheda problem, not a reaction to Advaita — which came later and in a different register.
5 · Why does Bhāskara hold that works cannot be dropped?
Every other difference from Advaita follows from this: whether bondage is a condition or an error.
The case is real and is not decisive. In its favour: Āśmarathya's view is reported without evident disapproval; 2.1.14's language of non-difference sits beside 2.3.43's description of the self as a part of brahman, and a reading that takes both at face value needs no machinery; and the sūtra nowhere states a doctrine of illusion. Against: the sūtra is terse enough that “face value” is doing a great deal of work, and the sequence at 1.4.20–22 does look like an ascent to Kāśakṛtsna. The honest position is that bhedābheda is the least strained reading of the text and that the text is not determinate enough to settle the matter — which is why eight schools could grow in it.
This is the question underneath the contradiction charge, and it is a live one in metaphysics generally. Bhedābheda says identity and difference are equally basic; neither is analysable into the other, and neither is a mere absence of the other. The objection is not that this is contradictory — the respects differ — but that it is unexplanatory: to say that a thing stands in two irreducible relations to another is to describe the situation rather than account for it. The school's reply is that every philosophy stops somewhere, that Advaita stops at an indeterminable ignorance and Dvaita at an unanalysable particular, and that stopping at two relations is no worse than either.
Less than the polemics suggest, and more than nothing. Both are realist, both reject māyā, both make the world a real effect of brahman, and both hold selves eternally distinct in some respect. The genuine difference is structural: Rāmānuja's body–soul relation is asymmetric — the modes depend on the substrate and never the reverse — whereas bhedābheda's relation is symmetric, with identity and difference on equal footing. That asymmetry is what lets Rāmānuja keep brahman free of defect, and it is bought at the cost of the pure literal reading that was bhedābheda's whole point. Whether the trade is worth it is exactly the argument between them.
| IAST | Devanāgarī | Sense |
|---|---|---|
| aupādhika | औपाधिक | Due to a limiting adjunct; Bhāskara's account of how the difference arises. |
| avibhāga | अविभाग | Non-separation; Vijñānabhikṣu's term for the relation, weaker than identity and stronger than dependence. |
| jñāna-karma-samuccaya | ज्ञानकर्मसमुच्चय | The combination of knowledge and works as joint means to liberation. |
| pariṇāma | परिणाम | Real transformation; brahman actually becomes the world, as milk becomes curd. |
| svābhāvika | स्वाभाविक | Natural, belonging to the thing itself; the version of the difference held by Yādavaprakāśa and Nimbārka. |
| upādhi | उपाधि | A limiting adjunct — and for Bhāskara, unlike for Śaṅkara, a real one. |