Ask a tradition where its central figure came from and you expect one story, possibly with variants at the edges. Ask this one and you get three, and they do not disagree at the edges. They disagree about who the mother was praying to, how long she prayed, whether her father was a monkey or a demon, and whether a wandering fortune-teller had anything to do with it.
The ordinary reflex is to sort them. One is earliest and therefore real; the others are accretion. Or one is the mainstream and the others are regional. Somebody must be wrong.
The compendium's own closing note declines the sort entirely, and does it in a single move that is easy to read past. The tellings differ, it says, because they belong to different kalpas — different world-cycles — and within its own cycle each account is simply true. What is being asked of the reader is not adjudication. It is devotion.
That is not a dodge dressed as doctrine, or at least it is not only that, and the structure it proposes is worth taking seriously on its own terms before deciding what it costs.
The Wind's own son. Vālmīki's Bālakāṇḍa gives the shortest route. The gods petition for an end to Rāvaṇa; Brahmā answers that Viṣṇu will take human birth as Daśaratha's son, and orders every god to send a portion of himself down into the vānara and ṛkṣa races who will assist. Among those portions: a child conceived through the Wind in Kesarī's wife Añjanā.1
Mataṅga's instruction. The Skanda and Bhaviṣyottara Purāṇas give Añjanā a much harder road. Childless and grieving, she comes to the sage Mataṅga in tears. He sends her fifty yojanas east of Lake Pampā to Narasiṃha-āśrama, and from there south to a hidden spring, Ākāśagaṅgā tīrtha — bathe there, hold twelve years of austerity, and a worthy son will come. She does it exactly: circling the aśvattha, bowing to Varāha-deva, living on nothing, asking nothing. At the end of twelve years Vāyu himself, moved, grants the boon.2
The Brahmāṇḍa Purāṇa's demon. The strangest telling does not start with Añjanā at all. It starts with an asura named Kesarī who wants a son and wins Śiva's attention for it — five-syllable mantra, senses mastered, food refused. Śiva's answer is a refusal that redirects: I cannot give you that son; fate has already written that you will have none. But I can give you a daughter, and through her, in time, the son you are asking for will come. The daughter is Añjanā. Years later a vānara — also, unrelatedly, named Kesarī — asks for her hand and is given it. Still no child. Then dharma itself arrives disguised as a low-caste fortune-teller, cane in one hand and string in the other, is fed and seated, and prescribes seven thousand years of austerity on Veṅkaṭācala. Añjanā goes. The sky eventually speaks back with the same news the first account opened with: Rāvaṇa is coming, Viṣṇu will be born as Rāma, and a son of yours will be born to help him.3
The device doing the work is kalpa-plurality. A kalpa is a world-cycle, and the cosmology holds that cycles recur without being identical — the same great events happen again, cast differently. Divergent accounts are therefore not competing reports of one event but accurate reports of different instances of a recurring one.
Formally this is elegant. It converts contradiction into indexing. Two statements that cannot both be true of one world become straightforwardly true of two, and the incompatibility dissolves without either source being demoted, which is the outcome a devotional compendium wants for texts it reveres equally.
It also has a specific consequence for how the accounts may be used. Each is true within its cycle, which means details may not be mixed across them. The Brahmāṇḍa's demon-Kesarī is not a prequel to Vālmīki's vānara-Kesarī. A reader who splices — taking the twelve-year austerity from one and the fortune-teller from another to build a fuller biography — has violated the doctrine while believing they were honouring it, since a composite belongs to no cycle at all.
Notice what is conspicuously not offered. No account is marked as the authoritative one. No chronology of composition is proposed. No allegorical reading is used to make the tellings mean the same thing underneath, which is the harmonising move a modern sympathetic reader most expects and would find least objectionable.
The refusal is deliberate and it is doing something. Harmonisation would require a criterion for what counts as the real content beneath the variants, and any such criterion is imported from outside the texts. Kalpa-plurality keeps every text whole at the price of keeping them separate — the same instinct visible in Ardha-Kukkuṭī-Nyāya, where the objection to selective reading is precisely that it requires an unstated external criterion.
The sharpest test of the doctrine is a detail the Brahmāṇḍa account flags and then walks past: there are two Kesarīs. One is the asura whose austerity produces Añjanā. The other is the vānara who marries her. The essay's rendering notes the coincidence explicitly — also, unrelatedly, named Kesarī — and moves on.3
Under any harmonising reading this is a serious problem, and the obvious fixes are all available. They could be the same figure, with the asura reading being a hostile gloss. The duplication could be a scribal doubling. One could be the "real" Kesarī and the other an intrusion.
The kalpa doctrine forecloses all of it, and this is where the device shows its actual cost and its actual discipline. Within the Brahmāṇḍa's cycle both Kesarīs exist and the coincidence is simply a fact about that world, no more requiring explanation than any other coincidence. The doctrine will not let you tidy it, and it will not let you use the untidiness as evidence that the account is corrupt. A reader who wants to ask which Kesarī was Hanuman's grandfather in history has, by the doctrine's lights, asked a malformed question.
The closing instruction — pick a kalpa and get on with the devotion — is the part most likely to be misread as anti-intellectualism, and it is worth separating what it forecloses from what it does not.
It does not forbid analysis. The same compendium argues aggressively about grammar, etymology, Vedic attestation and sectarian identity, and does so at length. What it forbids is a specific inquiry: adjudicating between accounts of the birth, which it treats as a question with no available method and no devotional yield. Inquiry is redirected rather than shut down.
Whether that redirection is principled or merely convenient is not something the essay settles, and a reader is entitled to notice that the questions declared unanswerable are also the questions whose answers might embarrass.
You are three sources into a topic and they will not line up. The first gives a date, the second gives a different date and a different cast, the third omits the episode. Your instinct is to build the composite — take the date from one, the cast from another, note the third as a gap. Stop before you do. Ask the prior question: are these reports of one event, or instances of a recurring pattern that each source encountered separately? If the second, the composite you were about to build describes nothing that ever happened.
Now you are writing about the material and you feel the pull to rank. This version is earlier, so it is truer. Notice that "earlier" is doing two jobs at once — it is a claim about manuscripts and a claim about reality, and you have not shown they connect. Write the accounts side by side. Attribute each. Let the reader see three, which is what you actually have.
And later, drafting, you find that a question has been sitting open for two weeks and has stopped generating anything. Every pass adds qualification and no pass adds knowledge. This is the condition the doctrine is built for. You name the question as one you are declining, you say why — no available method, no yield — and you spend the attention on the part of the material that is still moving. The move is legitimate exactly to the extent that you declare it, and it becomes evasion the moment you leave the reader thinking you never noticed.
The device is powerful and it is not free.
It is unfalsifiable by construction. Any two incompatible accounts can be assigned to different cycles, which means the doctrine can absorb an arbitrary quantity of contradiction without ever registering strain — and a framework that cannot be strained cannot be tested.
It also does not scale downward. It handles narrative divergence well: different mothers, different austerities, different lengths. It handles doctrinal divergence badly. If one text says Hanuman is Rudra's portion and another says he is Vāyu's, assigning them to separate kalpas implies his fundamental nature varies by cycle, which is a much larger metaphysical claim than the device was built to carry — and one the compendium elsewhere would reject, since its Rudrāṃśa argument is presented as a truth about what Hanuman is, not about what he was in one cycle.
And it cannot absorb a date, which is where the whole apparatus meets something harder than itself. Janārdan Miśra "Pankaj" records two competing birth-date traditions: Chaitra Pūrṇimā, the full moon of Chaitra; and Narak Caturdaśī falling on a Tuesday evening in the dark fortnight of Kārtik.4 Kalpa-plurality could take these in its stride if a birth date were only ever a fact in a story. It is not. Hanumān Jayantī has to be observed on a day, and a community cannot distribute a festival across world-cycles and turn up on both.
So this is the one place in the corpus where incompatible accounts generate a decision that must actually be made, by people, annually. Plurality holds only while nothing depends on the answer.
⚑ And the resolution, when it came, came from outside the literature entirely. Miśra notes that the Kārtik dating derives from the almanac tradition established by Madan Mohan Malaviya, and that this is the dating behind the popularly observed Hanumān Jayantī.4 A twentieth-century pañcāṅg settled what the Purāṇas had left open — not by adjudicating between them, and not by claiming to. The device that refuses to choose was overtaken by a publishing decision, which is a much more ordinary mechanism than the one the volume spends its pages defending, and considerably more effective.
Miśra also carries the Ānanda Rāmāyaṇa account, in which Hanuman originates from a portion of Daśaratha's putreṣṭi offering carried to Añjanā by a kite on Añjana-giri — a fourth origin narrative, and one to which no date attaches at all.4
The strongest evidence for reading this as genuine doctrine rather than post-hoc patching is that the essay states it as a general principle about Purāṇic literature, not as a special plea for this case, and that the principle is old and widely deployed rather than invented here.
The tension the compendium does not close. Kalpa-plurality and the Rudrāṃśa argument pull in opposite directions. The first says accounts may differ irreconcilably and all be true. The second argues, at length and against opponents, that one identification is correct and its rivals are mistaken. Both appear in the same volume. No essay reconciles them, and the compendium shows no sign of noticing the strain — it reaches for plurality on origins and for adjudication on identity, and the choice tracks, uncomfortably, which questions have a sectarian stake.
A second tension: the doctrine's own instruction is pick a kalpa, which asks the reader to make a selection the doctrine simultaneously says has no basis. On what grounds does one pick?
A third, added on source-update: the curse that made Añjanā a vanara has no settled author either. Rai Bahadur Lala Sītārām, reporting Jāmbavān's recitation at Kiṣkindhā 66, has her as the apsara Puñjikasthalā cursed into vanara form by the sage Atri — an attribution that is not the common one and that the essay states without argument.5 The same essay's Brahma Purāṇa material then leaves the curse authorless entirely, and puts a different sage, Agastya, in the story as the one who grants the sons.5 Recorded here as a variant rather than reconciled, and worth noting that it is a fourth axis of divergence — after mother's austerity, father's identity and date, the curse is also unfixed. [RENDERING — verify against Hindi] on the Atri attribution specifically, which is unusual enough that a rendering slip is a live possibility.
Open questions. Is kalpa-plurality doing metaphysical work here or literary work — a real cosmology, or a way of holding an anthology together? Does the doctrine license any selection principle, and if not, what constrains it? And is there a case where a tradition using this device has ever declared two accounts genuinely incompatible?
Shridharacharya's plurality converges strikingly with The Three Darśanas Simultaneously True, which holds that Hanuman instantiates three mutually exclusive philosophical positions at once without collapsing into any. That page's device is simultaneity within one world; this one's is distribution across many. Same refusal to adjudicate, two different pieces of machinery — and the two pages together suggest the refusal is a feature of how this figure is handled generally rather than a local fix for a birth-narrative problem.
It stands against Five Flagged Contradictions and One Denied, where contradictions are held open as contradictions. Kalpa-plurality does not hold them open; it dissolves them, and the vault should not let the dissolution pass as though it were the same intellectual act.
One handshake only. The connection below is load-bearing; a second would be manufactured to fill the section.
To Ta'wīl: The Return Hermeneutic. Both are techniques for keeping a revered text whole in the face of material that resists literal reading, and their difference is the useful part. Ta'wīl goes downward: the surface is retained and a deeper register is opened beneath it, so that one text carries two levels and the reader ascends. Kalpa-plurality goes sideways: the surface is retained at full literal strength and multiplied, so that incompatible surfaces are distributed across worlds rather than stacked in one.
The insight neither generates alone concerns what each technique protects. Ta'wīl protects the text's unity at the cost of its plainness — there is one revelation, but you can no longer read it plainly. Kalpa-plurality protects plainness at the cost of unity — every account may be read exactly as written, but there is no longer a single story to read. A tradition's choice between them therefore reveals what it is least willing to give up, and this compendium's choice tells you something specific: it will surrender a single coherent biography of Hanuman before it will surrender the literal readability of any Purāṇa. That ordering is itself a doctrinal commitment, made silently, in the shape of a device rather than in a statement.
Sharpest implication. Kalpa-plurality is not a claim about cosmology so much as a claim about which questions are worth a scholar's attention, smuggled in as one. A framework that makes contradiction costless also makes it uninformative — and the tradition pays that price knowingly, because what it wants from these texts is not a reconstruction of what happened but a form of attention that survives their disagreement.
Generative questions.