Arthāpatti (Postulation): The Fat Devadatta, Dṛṣṭārthāpatti and Śrutārthāpatti, Why It Is Not Inference, and Its Critics

Mīmāṃsā: Theory of Knowledge.

Official Syllabus: Paper I, Section B, Unit 18

Arthāpatti — “postulation” or “presumption” — is one of the two pramāṇas that Mīmāṃsā added to the Nyāya list. It is the knowledge we get when two facts we are sure of seem to clash, and we are forced to suppose a third fact that alone can reconcile them. The classic case: Devadatta is growing fat, yet never eats during the day — so he must eat at night. Nobody saw him eat at night; no rule says “all fat people eat at night”; yet we know it. Mīmāṃsā insists this knowledge cannot be explained as perception, inference, comparison or testimony, so it needs a pramāṇa of its own. Nyāya replies that it is only a disguised inference. UPSC has asked about this debate four times, including the 2024 question about a topper never seen studying.

Context & Problem

• The word. Artha = “fact” or “thing”; āpatti = “supposition” or “falling into”. Arthāpatti is “the supposition of a fact” — knowledge that resolves the conflict between two facts by presuming a third.
• Who accepts it. Prabhākara, Kumārila and the Advaita Vedāntins accept arthāpatti as an independent pramāṇa. Nyāya, Vaiśeṣika, Sāṃkhya and the Buddhists reduce it to inference; the Cārvāka rejects it altogether.
• Why the question matters. If arthāpatti is only inference, Mīmāṃsā’s list shrinks; if it is a distinct pramāṇa, then there is a kind of knowledge — explanatory knowledge — that the syllogism cannot capture. For Mīmāṃsā much depends on it: some of its central doctrines (apūrva, the power of words) are themselves known by arthāpatti (section 8).

Core Doctrine

1. What Arthāpatti Is

• Definition. Arthāpatti is the necessary supposition of an unperceived fact which alone can explain a phenomenon that demands explanation.
• The structure.
  1. There is a fact to be explained — the explanandum (upapādya).
  2. It cannot be understood in any way without supposing some other fact — the explainer (upapādaka).
  3. Affirming this explainer is arthāpatti. Its typical form is: “this cannot otherwise be accounted for” (anyathā-anupapatti).
• Its source. It arises out of a demand for explanation — a search for the ground of something already known.
• The classic example — fat Devadatta.
  1. Fact 1: Devadatta is growing fat (observed).
  2. Fact 2: Devadatta fasts during the day (observed).
  3. Background: no one grows fat without eating.
  4. Conflict: fatness and fasting cannot both be true as they stand.
  5. Postulation: “Devadatta eats at night.” This alone explains the whole — fasting by day together with increasing fatness.

2. More Examples from Daily Life and Language

• The absent friend. We call on a friend and do not find him at home, though we are sure he is alive. We say: “He must be somewhere outside.” This alone explains how a living man is not at home.
• The village on the river. Told that a person comes from a village “on the Yamunā“, we understand that the village is on the bank of the Yamunā — a village cannot be in a river.
• Missing words. When some words are omitted from a sentence, we supply them. Hearing “Shut!” we supply “the door“; hearing “Shut up“, we supply “your mouth” — without them the meaning is incomplete.
• Figurative meaning. When the primary meaning of a word does not fit the context, we suppose a secondary meaning: told “Industry is the key to success“, we take “key” to mean “means“, not a metal key.

3. Two Kinds of Arthāpatti

• Dṛṣṭārthāpatti — postulation from what is seen. It explains something perceived: the fatness of a man who fasts by day; the absence from home of a man known to be alive.
• Śrutārthāpatti — postulation from what is heard. It explains the meaning of words heard: supplying the missing word in “Shut!”, taking “Yamunā” as “the bank of the Yamunā”, or — if Devadatta’s father tells us “Devadatta, who is alive, is not at home” — presuming that he is elsewhere. Here the conflicting facts come from testimony, not perception.
• Arising from any pramāṇa. Kumārila shows that the facts that call for postulation may be known by any pramāṇa. For example, from seeing that fire burns, we postulate in fire a power to burn (śakti); from hearing that a sacrifice performed now brings heaven later, we postulate an unseen potency (apūrva).

4. Why Arthāpatti Is Not Any Other Pramāṇa

• Not perception. We do not see Devadatta eat at night. No sense touches the postulated fact.
• Not comparison. It is not knowledge of similarity. We cannot compare Devadatta with others who fast by day — many who fast by day are thin.
• Not testimony. No one told us he eats at night. (In śrutārthāpatti the starting facts are heard, but the postulated fact is not stated by anyone.)
• Not inference. This is the real battleground (section 5). The short answer: there is no invariable concomitance (vyāpti) between fatness and eating at night — we cannot say “wherever there is fatness, there is eating at night“, as we can say “wherever there is smoke, there is fire”.

5. The Debate with Nyāya: Can Arthāpatti Be Reduced to Inference?

• Nyāya’s reduction. Every arthāpatti, says Nyāya, can be rewritten as an inference.
  1. Disjunctive form. “Fat Devadatta eats either by day or by night; he does not eat by day; therefore he eats by night.” Likewise: “Living Devadatta is either at home or elsewhere; he is not at home; therefore he is elsewhere.”
  2. Categorical form. “All fat persons who do not eat by day eat at night; Devadatta is a fat person who does not eat by day; therefore Devadatta eats at night.” And: “All living persons who are not at home are elsewhere; Devadatta is a living person not at home; therefore he is elsewhere.”
  3. Negative-only form (kevala-vyatirekī anumāna). “Whoever does not eat at night is not fat while fasting by day — like Maitra, who eats by day; Devadatta is fat while fasting by day; therefore Devadatta is not one who does not eat at night”, i.e. he eats at night. The rule is known from negative cases alone.

So “arthāpatti” is just inference that Mīmāṃsakas have not recognised.

• The Mīmāṃsā refutation.
  1. Where would the universal premise come from? The rule “all fat people who fast by day eat at night” could never be learned from observation, because the night-eating is exactly what is never seen. The only way to reach such a rule is by postulating night-eating in each case — that is, by arthāpatti itself. So the “inference” presupposes arthāpatti; it cannot replace it.
  2. No proper middle term. In inference the middle term (hetu) must be a sign present in the subject and connected with the conclusion by vyāpti. Here neither fact alone can serve as the middle term — fatness alone does not point to night-eating, nor does fasting alone. Both together might be taken as the middle term — but the combination “fat-yet-fasting-by-day” already contains the problem whose solution is the conclusion; a middle term should not include the conclusion.
  3. A different direction. Inference starts from an established fact and concludes to what follows from it — a search for consequents. Arthāpatti starts from a conflict among facts and looks for what would explain them — a search for grounds. We use syllogism to draw conclusions from given facts, not to explain given facts.
  4. A different awareness. In inference we are aware “wherever there is A, there is B“. In arthāpatti we are aware “A cannot be so unless B” — the felt inexplicability of the facts drives the knowledge.
  5. Negative-only inference is really arthāpatti. Mīmāṃsā (and Advaita) do not accept a purely negative (kevala-vyatirekī) inference as inference at all. The cases Nyāya puts under it are exactly the cases of arthāpatti.
• Comparison table.

6. Prabhākara and Kumārila: Doubt or Inconsistency?

Both accept arthāpatti, but they explain what distinguishes it from inference differently.

• Prabhākara — doubt.
  1. There is an element of doubt in arthāpatti, absent from inference. From the undoubted perception of smoke we infer fire.
  2. In arthāpatti, one of the two perceived facts is made doubtful by the other. When we see that Rāma is not in his house and know that he is alive, the presumption that he is outside arises only when his “being alive” is put in doubt by his absence.
  3. Arthāpatti removes this doubt and reconciles the two apparently inconsistent facts. Since inference contains no doubt, arthāpatti is not inference.
• Kumārila — inconsistency, not doubt.
  1. There is no doubt in arthāpatti. We perceive Rāma’s absence and we know for certain that he is alive.
  2. If our knowledge of his being alive were doubtful, it could not be a sound basis for presumption.
  3. What arthāpatti removes is the mutual inconsistency of two well-ascertained facts, by presuming a third. In inference there is no such inconsistency.
• What both agree on. In arthāpatti there is no middle term at all — so no vyāpti, and so no inference.

7. Arthāpatti and Hypothesis

• A resemblance. Arthāpatti resembles an explanatory hypothesis in Western logic — a supposition made to explain facts.
• The difference. It lacks the tentative, provisional character of a hypothesis. What is known by arthāpatti is not merely entertained “to be tested”, but believed as the only possible explanation.
• The objection. Critics say that this is just where arthāpatti is weak: a supposition is only probable; there may be other explanations. If so, it cannot be a pramāṇa, which must give certainty.
• The reply. Arthāpatti operates only when there is no other explanation; where several explanations are possible, there is no arthāpatti, only doubt.

8. The Logical Necessity of Arthāpatti for Mīmāṃsā

The 2016 paper asked why it was logically necessary for Mīmāṃsā to treat arthāpatti as an independent pramāṇa.

• (1) Some knowledge has no vyāpti behind it. In explanatory cases the universal rule cannot be observed (the explaining fact is never seen). If only inference were allowed, such everyday knowledge would be unaccounted for.
• (2) Apūrva. The Veda says a sacrifice brings heaven; the sacrifice ends at once; heaven comes later. These facts conflict unless something endures in between. So Mīmāṃsā postulates an unseen potency — apūrva — by śrutārthāpatti. Without arthāpatti, its whole theory of ritual efficacy would have no pramāṇa.
• (3) Śakti. The power to burn in fire, the power to sprout in a seed, the power of a word to convey meaning — none is perceived; each is postulated to explain observed effects. Mīmāṃsā’s doctrine of potency rests on arthāpatti.
• (4) Interpreting the Veda. Vedic sentences are often elliptical; filling them out, and choosing secondary meanings when the primary meaning fails, is śrutārthāpatti. A school of interpretation needs it constantly.
• (5) The eternity of words. Śabara argues that a word must endure beyond its utterance, since speech is for another: if the word perished the moment it was spoken, the hearer could never grasp its meaning. This, too, is an argument from the impossibility of explaining communication otherwise.
• (6) Keeping inference honest. Treating explanation as inference would force Mīmāṃsā to accept rules (like “all fat day-fasters eat at night”) that are unknowable. Keeping arthāpatti separate preserves the strictness of inference.

9. The Case of the Topper Never Seen Studying

The 2024 paper put the doctrine to work: “A candidate who is never seen to be studying during the day-time secures a high position in a competitive exam.” How would the Bhāṭṭa Mīmāṃsakas and the Naiyāyikas explain the success?

• The facts.
  1. The candidate is never seen studying by day (observed).
  2. The candidate secures a high rank (observed).
  3. Background: a high rank in a competitive exam is not achieved without study.
• The Bhāṭṭa explanation — arthāpatti.
  1. The two observed facts appear inconsistent.
  2. The inconsistency is removed only by postulating that the candidate studies when unseen — at night, or in private.
  3. This knowledge is not perception (no one saw the study), not inference (there is no observed rule “those not seen studying by day study at night”), not comparison, not testimony. It is arthāpatti — the Devadatta pattern applied to a topper.
  4. For Kumārila the starting facts are certain (no doubt about the rank or the day-time idleness); what drives the knowledge is their inconsistency.
• The Nyāya explanation — inference.
  1. Disjunctive: “This candidate studied either by day or at some other time; not by day; therefore at some other time.”
  2. Negative-only: “Whoever does not study at all does not secure a high rank — like an idle candidate who failed; this candidate secured a high rank; therefore this candidate is not one who does not study.”
  3. The knowledge is a valid inference based on a negative concomitance; no new pramāṇa is needed.
• Evaluating the two.
  1. For Nyāya: the reasoning clearly can be set out as a syllogism.
  2. For Mīmāṃsā: the “universal” premises are themselves reached only because we first postulate unseen study; and the mind’s actual movement is from a puzzle to its solution.
  3. A warning. Real life offers other explanations — exceptional memory, earlier preparation, an easy paper, even unfair means. This shows the danger of calling any postulate “the only explanation”, and it pushes arthāpatti towards a hypothesis that needs testing — a point in Nyāya’s favour.

Critical Evaluation

Objections to Arthāpatti

• Nyāya: it is disjunctive or negative-only inference; the “search for grounds” versus “search for consequents” is a psychological difference, not a logical one.
• Buddhists: the Devadatta case is inference supported by memory (of the rule that eating causes fatness). Multiplying pramāṇas beyond necessity is epistemologically redundant.
• Cārvāka: presumption is mere guesswork without empirical foundation — it posits what no one perceives.
• Hypothesis objection: a supposition is at best probable; a pramāṇa must be certain. “Only possible explanation” is rarely provable.
• Against Prabhākara: if arthāpatti begins in doubt, its conclusion inherits doubt and cannot be valid knowledge (Kumārila’s own objection).
• Even Advaitins who accept it admit that critics have some force in saying it overlaps with inference.

Mīmāṃsā’s Replies

• The universal is borrowed. Any syllogism that “reduces” arthāpatti uses a premise known only through arthāpatti — so the reduction is circular.
• Logic follows knowledge. That a piece of knowledge can be rewritten as a syllogism does not show that it arose from one; pramāṇas are distinguished by how knowledge actually arises.
• Certainty. When the facts are certain and no alternative explanation exists, the postulated fact is as certain as they are.
• Against redundancy. A pramāṇa is “redundant” only if another pramāṇa actually gives the same knowledge in the same way — which, here, none does.

Critics and Defenders

Cross-Tradition Comparison

Arthāpatti and Western Thought

Arthāpatti across Indian Schools

Synthesis (Siddhānta)

• Arthāpatti is the necessary supposition of an unperceived fact that alone reconciles apparently conflicting facts — fat Devadatta who fasts by day must eat at night.
• It is of two kinds: dṛṣṭārthāpatti (from what is seen) and śrutārthāpatti (from what is heard — missing words, secondary meanings, apūrva).
• It is not perception (the fact is unseen), comparison (no similarity), testimony (no one says it) or inference (no vyāpti, no proper middle term; the “universal” is itself reached by postulation; it searches for grounds, not consequents).
• Prabhākara finds its mark in doubt; Kumārila in the inconsistency of certain facts.
• Nyāya rewrites it as disjunctive, categorical or negative-only inference; Mīmāṃsā replies that the rewriting is circular and misdescribes how the knowledge arises.
• It is logically necessary for Mīmāṃsā because its doctrines of apūrva, śakti and Vedic interpretation rest on it.
• Contemporary relevance.
  1. Medicine: a doctor reasons from conflicting symptoms to the one disease that explains them.
  2. Science: postulating unseen entities (dark matter, to reconcile observed galaxy rotation with visible mass) is a modern arthāpatti — though science keeps it tentative.
  3. Law: circumstantial evidence — “the accused was seen entering, the jewels are gone, no one else had access” — works by postulation.

“Arthāpatti is the presumption of a thing, because something seen or heard cannot otherwise be explained.” — Śabara