What is Mimamsa?

A fire is burning on a low earthen altar, and every gesture around it is exact. The wood is of a prescribed kind, laid in a prescribed order; the offering is poured at the appointed word; the priest’s hand moves as the text requires and not otherwise. Nothing here is improvised, and nothing is left to feeling. To the sensibility of the classical Vedic ritualist, this precision is not fussiness but the very shape of the sacred, and the school of philosophy that grew up to explain it asked a question no other darshana put first: not what is real, and not what is God, but what are we bound to do, and how do we know it. That school is Mimamsa, and of the six darshanas it is the one a modern reader is least prepared to take seriously — which says more about the modern reader than about the school.

It is the philosophy of ritual, and behind that homely description it built one of the most sophisticated theories of language, meaning and textual interpretation the ancient world produced. Its full name is Purva Mimamsa, the earlier inquiry, paired with the later inquiry, Vedanta. The word mimamsa means investigation or reflection, and the two schools divide the Veda between them: Purva Mimamsa takes the portion of action, the rituals and their commands, and Vedanta, the Uttara or “later” Mimamsa, takes the portion of knowledge, the Upanishads. What follows is the story of how a manual of sacrifice became a theory of everything a sacred text can mean.

Dharma is what the Veda commands

Mimamsa opens, in Jaimini’s Mimamsa Sutra, with the same austere turn that the Brahma Sutra of Vedanta will later echo word for word in its own domain:

athato dharma-jijnasa. Now, therefore, the inquiry into dharma. (Mimamsa Sutra 1.1.1)

And it answers its own question at once, in the second aphorism, with a definition that fixes the whole character of the school:

codana-lakshano ‘rtho dharmah. Dharma is that good which is indicated by the [Vedic] injunction. (Mimamsa Sutra 1.1.2)

This is a startling claim, and Mimamsa means it strictly. Dharma, duty, is not something we can read off from nature, or deduce by reason, or feel in the conscience. It is known only from codana, the command of the Veda. Shabara, the school’s great early commentator, glosses the sutra to make the point unmistakable: dharma is artha, a good or benefit, that has the distinguishing mark (lakshana) of being enjoined by a Vedic command. The Veda does not merely describe dharma or report it; it enjoins it, and the injunction is the thing itself. Where the texts merely praise, or tell a story, or state a fact, Mimamsa files that under arthavada, subordinate eulogy, whose whole job is to lend force to the commands that surround it. The command, the vidhi, is the living heart of scripture, and everything else is scaffolding around it.

That may sound like a narrowing, but it is the opposite. Because the injunction is central, and because injunctions in a vast and ancient text will seem to collide, contradict, repeat and qualify one another, Mimamsa was forced to become a science of reading. Its unit of analysis is the adhikarana, the topic — a fixed five-part structure in which a doubtful passage is stated, a prima facie wrong reading is set up, and then defeated, and a settled conclusion reached. Jaimini’s sutras are, in effect, some nine hundred such cases argued in sequence, and the discipline of weighing one textual claim against another under strict rules is Mimamsa’s true and lasting subject. The rituals are the occasion; interpretation is the achievement.

A scripture with no author

If dharma rests entirely on the Veda, then everything depends on the Veda being trustworthy beyond any possible doubt, and here Mimamsa makes its boldest move. The Veda, it holds, is apauruseya, authorless and beginningless. No person composed it, and no god either. It is eternal, heard rather than made, carried down unbroken through the seers who received it.

The arguments are precise, and they come in two orders. The first is historical: no author is remembered. For every human text we can trace a maker, or at least the living memory of one, but no tradition and no lineage preserves the name of anyone who composed the Veda; and Mimamsa reads that unbroken silence not as forgetting but as evidence that there was never a composer to remember. The second argument goes deeper. Any text with an author inherits that author’s limitations, the errors and biases of a finite mind, and could in principle be doubted wherever the author might have erred. A god-given scripture only pushes the problem back a step, for now the god’s existence and reliability can be questioned. But a Veda that no one ever composed, that simply is, as eternal as the truths it carries, stands wholly beyond that line of doubt. To secure dharma once and for all, Mimamsa removes the author entirely.

This is bound up with a further doctrine, one of the most distinctive theses the school ever advanced: the eternity of the word itself, and of the bond between a word and its meaning. For most thinkers, ancient and modern, language is a human convention — a community once agreed that a certain sound would stand for a certain thing, and it could have agreed otherwise. Mimamsa denies this outright. Shabda, sound or word, is eternal, and so is the relation that ties it to what it names; the connection between the word “cow” and cow-ness was never established by anyone, because it always was. Jaimini calls this bond autpattika, inborn or primordial (Mimamsa Sutra 1.1.5). Language, on this view, is not invented but is part of the permanent furniture of reality — which is precisely why an eternal Veda can be made of it.

Here Mimamsa fought on a second front, against the Sanskrit grammarians. Bhartrhari and his heirs had developed the elegant doctrine of sphota, the “bursting forth” of meaning as a single indivisible flash that transcends the individual sounds. Mimamsa would have none of it. A word, Shabara argues in his commentary on Mimamsa Sutra 1.1.5, is nothing over and above its phonemes, its varnas, and it is the eternal phonemes themselves — not some further unitary entity hovering above them — that are permanent and that carry meaning. Understanding a sentence, Kumarila would later insist against the grammarians, needs no mysterious sphota; it can be explained by the memory of discrete, eternal sounds heard in order. The historian of ideas can only note the irony that the grammarians developed sphota in large part as a foil to the Mimamsakas’ hard doctrine of the eternal phoneme; two schools, both defending the sanctity of language, drove each other to opposite theories of how a word means.

Mimamsa braces the whole edifice with a third principle, epistemological this time: svatah pramanya, the intrinsic validity of knowledge. A cognition, on this doctrine, is valid by its own nature. It does not wait to be certified true by some further test; it arrives already carrying its own authority, and it stays true unless and until it is defeated by a later cognition that shows it to be mistaken. Truth is the default condition of awareness, and error the exception that must prove itself. Kumarila made this the signal contribution of his epistemology, and its locus classicus is a single verse of his Slokavarttika (in the section on codana): a means of knowledge has its reliability intrinsically, svatah, for a capacity that does not already belong to a thing by itself cannot be conferred on it from outside. The contrast is with those who hold validity to be paratah, established from without — a position the logicians of Nyaya defended, holding that a cognition is confirmed true only by something further, such as the successful action it prompts. Applied to scripture, the Mimamsa principle is decisive: the Veda’s statements are self-authenticating, and since nothing can ever defeat them, no external prop for their authority is needed, or even possible.

The rite that works itself

The strangest feature of Mimamsa, to a religious sensibility shaped elsewhere, is its near-indifference to the gods. The Vedic sacrifice is full of deities, invoked and offered to by name. Yet classical Mimamsa does not hold that the gods grant the fruits of the rite. The rite works by its own power.

The mechanism it proposes is apurva, literally “the unprecedented,” the not-before — a supersensible potency generated by the correct performance of the ritual. Here Mimamsa is solving a genuine puzzle, and solving it with real ingenuity. A sacrifice is performed today, and the Veda promises its fruit — heaven, perhaps — in a distant future, or even a future life. The act is over in an afternoon; the reward may lie decades or lifetimes away. What bridges the gap? Between the vanished deed and the far-off result, Mimamsa posits apurva, an unseen residue that the correct performance deposits and that persists, storing the act’s efficacy until the appointed season and then ripening into the promised fruit. The offering creates the apurva; the apurva, in its time, yields the reward. No divine bookkeeper is needed to remember the deed and repay it. Kumarila argued for the category at length in his Tantravarttika, tying it to his claim that what a command truly enjoins is something to be brought about — so that the very grammar of injunction points toward an unseen result the act produces.

The gods, meanwhile, are real enough as the addressees of the offerings, but they are, in the working of the sacrifice, almost incidental. This is why Mimamsa counts as orthodox yet non-theistic: it reveres the Veda absolutely while needing no presiding deity to make the Veda’s rituals effective. And it is why apurva, for all its strangeness, is in effect a rigorous theory of karma. The deed done now is not lost; it is stored, and it ripens on its own terms — a law of the world rather than the decision of a divine judge.

The most this-worldly school

It is worth dwelling on how unusual this makes Mimamsa among the darshanas. Most of the six lift the gaze toward liberation and treat the ordinary world of duty and action as something to be transcended or escaped. Mimamsa begins exactly there, in the world of obligation, and takes it with complete seriousness. The daily and seasonal rites, the duties of one’s station, the offerings and observances that fill a traditional life — all of it Mimamsa treats as the very substance of dharma, not a lower rung to be climbed past on the way to something better.

This earthward focus is the deep root of the school’s quarrel with Vedanta. To the Vedantin, the rituals are preparatory, and the real goal is the knowledge of Brahman that ends rebirth altogether. To the classical Mimamsaka, the performance of dharma is itself the point, the lifelong work a person is here to do. Later Mimamsa would make more room for liberation, but the founding instinct of the school was to honour the world of action rather than to flee it. In a tradition often caricatured as world-denying, Mimamsa stands as the great counterweight — the school that took duty itself to be sacred.

Shabara, and the two who divided the school

Jaimini’s sutras are terse to the point of silence, and the school as we know it was made by its commentators. The dates are genuinely unsettled, and honesty requires ranges rather than false precision. Jaimini himself is placed anywhere from the third century BCE to the first century CE. The first commentary that survives whole is the Shabara-bhashya of Shabara (Shabarasvamin), the earliest of the completely extant commentaries on Jaimini’s sutras; its date is argued across the early centuries CE, often set around 400 CE, and it became the base text on which everything later was built. Every subsequent Mimamsaka wrote, in the first instance, as a commentator on Shabara.

Some centuries after Shabara the school was carried forward and split in two by a pair of thinkers who both flourished in the seventh century. Kumarila Bhatta flourished in the latter part of that century, around 660; Prabhakara a little later, around 700. Tradition makes Prabhakara the junior of the two, often his pupil, and reports that Kumarila nicknamed him Guru, “the teacher,” for the acuity of his objections — though some scholars, Ganganatha Jha among them, have argued that the Prabhakara school preserves an older reading of Shabara than the Bhatta line does. Their followers became the two rival schools, the Bhattas after Kumarila and the Prabhakaras after Prabhakara, and they would debate each other, and everyone else, for a thousand years.

Their founding works map neatly onto Shabara’s text. Kumarila wrote three, of which two are pillars of Indian philosophy. The Slokavarttika, the “critical commentary in verse,” treats the opening portion of Shabara’s first chapter, the part traditionally called the Tarkapada, the “section concerning reasoning” — and it is here, in the epistemological foundations for Vedic authority, that Kumarila mounts his defence of intrinsic validity, the eternal word, and the enduring self, and argues at length against the Buddhists. The Tantravarttika takes up the rest of Shabara’s first chapter (the remaining padas) and chapters two and three, and is the hermeneutical heart of the school. The Tuptika handles the balance of the commentary, the later chapters. Prabhakara, for his part, wrote the Brihati, a dense commentary directly on the Shabara-bhashya, from which his distinctive positions were drawn out by his own followers. Around these works the two schools took shape, and their disagreements — fine, technical, and consequential — are where Mimamsa did its sharpest philosophy.

Six ways of knowing, or five

The first great division between the two schools concerned the pramanas, the recognised means of valid knowledge — the question of exactly how many distinct doorways truth has into the mind. It is the same question that every Indian school had to answer, and the count itself is always a staked philosophical position.

Prabhakara accepted five. There is pratyaksha, perception; anumana, inference; upamana, comparison, the recognition of a thing from a likeness previously described; shabda, reliable verbal testimony, which for an orthodox school includes the Veda above all; and arthapatti, presumption or postulation — the move by which we supply what a fact silently requires in order to make sense. The stock example is old and vivid: a man is known to fast all day, yet he is growing stout; the only way to reconcile the two facts is to postulate that he eats by night. Presumption is not inference, Mimamsa insists, because it is driven not by a general rule but by an otherwise irreconcilable conflict between two established facts. Mimamsa was, in the wider Indian debate, the school most insistent that arthapatti is a means of knowledge in its own right.

Kumarila’s Bhatta school added a sixth: anupalabdhi, non-apprehension, the means by which we know an absence. The awareness that there is no jar on the table, Kumarila argued, cannot come from the senses, for there is simply nothing there for a sense to make contact with; the knowledge of the jar’s absence is a distinct mode of knowing, keyed to the very non-perception of a thing that would have been perceived had it been present. Prabhakara denied this. In such a case, he held, we merely perceive the bare table, and “absence” is not a further object requiring its own instrument. Behind this dry-looking dispute lies a real metaphysical stake — whether absences are genuine features of the world that must be known by a means of their own, or nothing over and above the presence of everything else. Kumarila’s realism about absence, and the sixth pramana it required, is one of the school’s most characteristic positions.

The shell, the silver, and two theories of error

The two schools divided just as sharply over the nature of error — over what exactly is happening in the mind when a person walking a beach at dusk mistakes a shell on the sand for a piece of silver. It looks like a small question. It is in fact the front line, because a school that has staked everything on the intrinsic validity of knowledge owes a very careful account of how knowledge ever goes wrong at all. If every cognition is true by its own nature, error is the standing embarrassment that has to be explained.

Kumarila’s answer is viparita-khyati, apprehension-as-otherwise, misjudgement. Error is a single, positive cognition that actively misrepresents: what is seen (“this”) and what is remembered (“silver,” glimpsed before in the market and stored in memory) are wrongly fused into one false judgement, “this is silver,” in which a real thing present to the eye is taken as another real thing that is not there. The error is a positive misconnection between two reals, and it is corrected — sublated — when the fingers close on shell and the misrelation is undone. Nothing wholly unreal ever appears; the mistake is a misplaced relation, not a window onto some third and ghostly kind of object.

Prabhakara’s answer is subtler and, in its way, more radical: akhyati, non-apprehension. Strictly speaking, he holds, there is no false cognition at all. There are only two perfectly valid ones — a present perception of the shining “this,” and a memory of silver seen elsewhere — and what the mind fails to do is notice that they are two. Error is never a positive misknowing but only an unnoticed gap, a failure to distinguish a perception from a memory; wrong action follows from the blur, not from any false object having been manufactured. The elegance is that this preserves intrinsic validity untouched: each of the two cognitions is true, and only their un-marked difference misleads. The Bhattas retorted that a merely negative failure cannot explain why the person actively reaches for silver; the Prabhakaras retorted that a positive false cognition cannot arise if every cognition is intrinsically valid. The two theories of one small mistake are, between them, a compressed statement of the whole Mimamsa problem of knowledge.

Their differences ran on, into the theory of the self and the theory of how a sentence conveys its meaning. Both schools, against the Buddhists, defended a real and enduring atman, a permanent self to carry the moral residue of deeds and to be the one who acts, remembers and reaps — Kumarila arguing in the Slokavarttika that the self is what the very notion “I” points to, secured by memory and by the genuine recognition, pratyabhijna, that the one who remembers is the one who earlier knew. But the schools parsed the knowing self differently, and they divided over language too: the Bhattas held abhihitanvaya, that a sentence’s meaning is built up by first grasping each word’s separate meaning and then combining them, while the Prabhakaras held anvitabhidhana, that words yield their meanings only as already-connected, already-in-construction, never in isolation. The dispute reads like grammar, but it is really a disagreement about whether meaning is assembled from parts or given whole — and it fed directly into the wider Indian philosophy of language.

A philosophy that needed no God

It is worth pausing on how far Mimamsa’s non-theism went, because it is easily softened into a vague “the gods don’t matter much.” In Kumarila’s hands it became something sharper: an argument that the Veda’s authority positively requires that it have no author, divine or human. The Nyaya theologians took the opposite path — thinkers such as Jayanta held that the Veda’s reliability is guaranteed precisely because God is its author, a trustworthy speaker behind a trustworthy text. Kumarila turned this on its head. An authored Veda would be only as reliable as its author, and would reopen every doubt that apauruseya was meant to close; a God posited to underwrite scripture is a God whose own existence and goodness must then be established, which cannot be done without begging the very authority at issue. Far better, Kumarila argues, that the Veda stand eternal and authorless, needing no one to vouch for it. His defence of the eternality of language thus carries within it a critique of theism — one that scholars have noted runs, ironically, close in form to the Buddhists’ own arguments against a creator.

Nothing in this requires the denial of gods, and Mimamsa did not trouble to deny them; the deities named in the rites are real enough. What Mimamsa denied was that any deity is needed to make the world go, to author the Veda, or to hand out the fruits of ritual — the offices a creator God is usually kept on to fill. The world for the Mimamsaka has no beginning to be created at and no maker to create it; the moral order runs on apurva, an impersonal law, not on divine decree. Only much later, and under the pull of a devotional age, did Mimamsakas make room for God: the medieval figure Venkatanatha would fashion a sesvara, “theistic,” Mimamsa, folding Mimamsa’s ritual world into a Vedantic theism. But that was an accommodation to later sensibilities, not the school’s own founding voice. In its classical form, Mimamsa remains the great proof that a philosophy can be wholly orthodox, wholly committed to the Veda, and yet have no essential use for God at all.

The long shadow: law, Vedanta, and the end of an age

The strangest thing about this school of sacrifice is where its influence finally landed. Its rituals faded with the ritual culture that sustained them; its non-theism was overtaken by the devotional tide. And yet Mimamsa cast the longest interpretive shadow of any darshana, because what it had really built was not a theory of the fire-altar but a general science of how to read an authoritative text.

Its rules of interpretation — the structure of the adhikarana; the great maxims for resolving conflicts, such as tantra, where one performance serves several purposes at once, and prasanga, where an act done for one end incidentally serves another; the principles for telling a binding injunction from mere praise, for handling apparent contradictions, for deciding when a passage is primary and when subordinate — became the standard method far beyond ritual. The whole tradition of Dharmashastra, the vast literature of Hindu law and custom, runs on Mimamsa’s rules of reading; when a later jurist had to decide what a text truly required, he reasoned like a Mimamsaka, and the scholar P. V. Kane devoted much of his monumental History of Dharmasastra to tracing exactly these maxims through the law. The distinction between shruti and smriti, and the rules for ranking them, is Mimamsa’s legacy as much as anyone’s, and it is treated more fully in shruti and smriti.

The deepest irony is reserved for Vedanta. The two schools agreed on almost nothing of substance — Mimamsa exalting action where Vedanta exalts knowledge, Mimamsa needing no God where much of Vedanta is drenched in the divine. And yet Vedanta, the Uttara Mimamsa, took over its elder sister’s entire exegetical machinery to interpret its own scriptures. When a Vedantin argues from an injunction, weighs a eulogy, reconciles two passages of the Upanishads, or lays out a topic as an adhikarana, he is using tools that Mimamsa forged. The younger inherited from the elder its very manner of reading, even while rejecting nearly everything the elder read.

And there is a last, darker thread the tradition insists on remembering. Kumarila’s long polemic against the Buddhists — his defence of the eternal Veda, the enduring self, and the reality of the external world, and his assault on Buddhist epistemology that the scholar John Taber has translated and reconstructed at length — was cast by later memory as a turning of the tide. Traditional accounts, still echoed in some Indian scholarship, credit Kumarila’s dialectical victories with the decline of Buddhism in India, and hagiography wove legends around him: the study of Buddhist doctrine at Nalanda, the great debates, the reversals. Sober historians treat these stories with caution — the decline of Indian Buddhism had many causes, and one philosopher’s arguments were surely not chief among them. But the legend endures because it captures something real about how the age remembered Kumarila: as the man who argued the eternal, authorless Veda back to the centre of Indian thought.

A final note on the uncertain. The dates of Jaimini and Shabara are unsettled across centuries, and this article has given ranges rather than false precision throughout; the pupil-relationship of Prabhakara to Kumarila is traditional, and contested; and the classical non-theism was, in time, softened by thinkers who found room for God within the system. What did not change was the core conviction — so alien to the modern ear, and so central to the tradition it shaped — that there is a real order of things we are obliged to uphold, and that the knowledge of it comes to us, eternal and authorless, in the commands of the Veda.

Vedanta would ask what is ultimately real. Mimamsa asked, first, what we are bound to do. The tradition thought both questions worth a school of their own — and then, quietly, let the school of duty teach the school of knowledge how to read.