It is a reasonable question with a straightforward-sounding answer, and Somchaitanya Shrivastava asks it directly.
The answer is that four Purāṇic accounts say so, they disagree about the mechanism completely, and he lays all four out without choosing.1
One. The Ānanda Rāmāyaṇa: Śiva, pleased by Añjanā's long austerity, promises her an eleventh Rudra as a son — and the blessing reaches her riding in the sacred kheer from Daśaratha's putreṣṭi sacrifice, after a kite snatches a portion and a sudden wind drops it into her waiting hands.1
Two. The Śiva Purāṇa's Śatarudra Saṃhitā: Śiva sees Viṣṇu's Mohinī form and loses his seed; the Saptarṣis convey it and place it into Añjanā's womb through her ear.1
Three. The Vāyu and Bhaviṣya Purāṇas: Śiva, in fierce form, enters the body of Añjanā's husband Kesarī and unites with her that way — and Vāyu immediately does the same, so that the child has, on this account, two fathers layered into one conception.1
Four. The Vinaya Patrikā: Śiva simply asked Rāma for the boon of serving him in a devotee's humility, Rāma granted it, and Śiva took Hanuman's form for exactly that purpose.1
The temptation is to treat these as regional colourings of a shared event. They are not, and the differences are categorical rather than decorative.
Account one is a transmission problem — a boon travelling through a food offering, an intercepting bird and a gust of wind. The mechanism is logistical and slightly comic, and Śiva is not present at the conception at all.
Account two is involuntary. Śiva loses his seed; he does not give it. Third parties handle the transfer. Nobody in the sequence intends a son.
Account three is possession, and it is the strangest. Śiva occupies the husband's body, and then Vāyu occupies it too, which produces a child with a layered paternity nobody in the story seems troubled by.
Account four has no conception in it whatsoever. It is a request, a grant, and an assumed form.
Different agents, different intentions, different degrees of Śiva's own participation, and in the fourth no biology at all. A reader looking for what happened has four incompatible answers; a reader looking for what the tradition needed has four different jobs being done.
Track the mother across the four and her agency swings from maximal to absent, which is a cleaner index of what each account is for than anything the accounts say about Śiva.
In the Ānanda Rāmāyaṇa she is the cause. Śiva is pleased by her prolonged austerity and the boon is her earning. Everything after — the kite, the wind, the kheer landing in her hands — is delivery of something she has already won.
In the Śiva Purāṇa she is a site. The seed is lost by Śiva, carried by the Saptarṣis, and placed into her. She does not act at any point in the sequence and nobody consults her.
In the Vāyu and Bhaviṣya account she is acted upon twice, by two beings occupying her husband's body in succession, and the text records no reaction from her at all.
In the Vinaya Patrikā she does not appear. The transaction is entirely between Śiva and Rāma.
So the account that makes her a practitioner earning a result is the same account that ties the birth to Daśaratha's sacrifice — the version with the most narrative machinery and the most human causation. The accounts that reduce her are the ones doing theological work between male principals.
Neither Shrivastava nor any other contributor remarks on the variation. But it means a reader who wants the tradition's account of Añjanā gets a different woman depending on which sonship story they arrive through, and the differences are not small.
The Śiva Purāṇa version has the Saptarṣis place the seed into Añjanā's womb through her ear, and the detail is specific enough that it is worth asking why that organ.
The ear is not an incidental route. In this tradition it is the organ of transmission — the entire Vedic apparatus is śruti, that which is heard, and the whole chain of authority runs through hearing rather than through reading or seeing. A teaching enters through the ear and that is how it acquires standing.
So a conception through the ear makes the child, structurally, a product of reception rather than of contact.
Which fits a figure whose defining characteristics are all reception-shaped. He is the one who hears the story, carries the message, receives the teaching from Sūrya while keeping pace, and whose ears — on Ramlal's anatomical reading — exist to drink in Rāma's story and are still doing it centuries later.4
Whether the Śiva Purāṇa intends any of that is unrecoverable and probably unlikely; aural conception is a widespread motif with its own history and it does not need a rationale specific to Hanuman.
But the fit is there, and it is the sort of fit that makes a variant survive. An origin story that rhymes with the figure's function will outlast one that does not, regardless of which came first.
Read them by function and the incoherence organises.
The Ānanda Rāmāyaṇa account ties Hanuman's birth to Daśaratha's sacrifice, which is the same rite that produces Rāma and his brothers. That is a claim about simultaneity and design: the servant and the master come from one offering, and the alliance was arranged before either was born. Sarvānand Pāṭhak gives the same frame at greater cosmological scale — the gods petitioning Brahmā about Rāvaṇa, Viṣṇu agreeing to be born as Daśaratha's son, and Brahmā instructing the gods to father equally mighty sons in vanara form to support him.2
The Mohinī account connects Hanuman to the single most theologically loaded episode in the Śaiva-Vaiṣṇava relationship, and it makes him a product of that encounter. Its job is cross-sectarian.
The two-fathers account is the only one that preserves both the Vāyu and Śiva paternities without subordinating either. Its job is reconciliation, and it pays for it with a biologically incoherent conception nobody comments on.
And the Vinaya Patrikā account, which is Tulsīdās, has no interest in any of this. Its job is devotional: it says a god wanted to serve and arranged to. Shrivastava reports Tulsīdās elaborating the point elsewhere — whatever body lets you love and serve Rāma is the body worth having, and that is reason enough for a god to choose it.1
Four accounts, four problems solved. None of them the problem of what occurred.
The volume carries two more that Shrivastava does not include in his four.
Ramlal reports Giridhar's Gujarati Rāmāyaṇa, where Śiva grants Añjanā's tapas the boon of the eleventh Rudra as her son — and adds a transmission device of its own: Suvarcalā, a vulture-woman and former apsara cursed by Brahmā for a mistake in a dance, whose release was promised only on dropping a portion of Daśaratha's offering onto Añjanā's mountain.3
That is account one with a cursed intermediary substituted for the kite. Which is worth noticing as a type: the offering-transport story attracts a carrier, and the carrier accumulates a backstory.
And then Shrivastava's own Vedic account, which he treats separately and which is the most ambitious. Aniruddhācārya Veṅkaṭācārya's reading: Śiva as fire-breath, Śakti as moon-breath, their union producing the forty-nine Maruts — called in the Veda itself sons of Rudra. Ordinary wind descends from these Maruts, which is why Vāyu himself is called Marut. Hanuman is their subtlest member, seated in the sun, and therefore honestly a son of Rudra.1
That one does not compete with the other four. It operates on a different level entirely — deriving the category rather than narrating an event — and it is the only account that makes the Śiva-sonship true by definition rather than by incident.
Shrivastava's handling is the thing to study, because it is a deliberate methodological position and he states it.
He does not rank the four. He does not date them, or prefer the one with the best attestation, or note that the Vinaya Patrikā is a sixteenth-century devotional lyric and the Śiva Purāṇa is not. His closing move is that Hanuman answers to enough of these accounts, each carrying its own separate version of who he is, that no single label holds him — and that you call on whichever one the moment requires.1
Read as scholarship that is a failure to conclude.
Read as a working doctrine it is coherent, and it is the same doctrine the volume applies to the birth narratives generally through kalpa-plurality: incompatible accounts distributed rather than ranked.
But notice what is not being said here, and the absence matters. Kalpa-plurality is a published device with a stated justification — different world-cycles, so both are true of their own age. Shrivastava does not invoke it. He simply declines to choose, which is a different and weaker move: the plurality doctrine explains how the accounts can all be true, and the refusal to adjudicate only declines to say that any of them is false.
So the volume has an instrument for exactly this problem and its most thorough treatment of the problem does not use it.
A tradition with good machinery does not necessarily deploy it where it fits. The machinery gets used where somebody once had an argument, and the four-accounts passage is not an argument — it is a survey, and surveys do not reach for tools.
You are trying to establish how something actually came about — a decision, a policy, a piece of the culture at your organisation — and you have collected four accounts from four people who were there.
They do not disagree about details. They disagree about the kind of event it was. One says it was planned from the start. One says it happened by accident and got rationalised. One says two things happened at once and both count. One says it was somebody deciding they wanted to.
The instinct is to treat this as a reliability problem and start weighting the sources — who was closest, who has an interest, whose memory is better.
Try Shrivastava's move first, without adopting his conclusion. Ask what each account is load-bearing for.
Take each version and finish this sentence: if this one is true, then ___ follows. The planned-from-the-start account licenses a claim about the founders' foresight. The accident account licenses a claim that the current arrangement is arbitrary and can be changed. The two-things-at-once account keeps two factions both correct. The somebody-wanted-it account puts a person's agency at the centre.
You will usually find that the accounts sort by what each teller needs to be true now, and that the sorting is cleaner than any sorting by reliability.
That does not tell you what happened. It tells you what the disagreement is for, which is more actionable, because the argument you are actually in is about the consequence rather than the event.
Then decide whether you need the history at all. Shrivastava's answer is that you do not — call on whichever account the moment requires. That is available to you and it has a cost he does not mention: it works while nobody is checking, and it fails completely the first time two of your accounts are needed in the same room.
Four named texts with their mechanisms given in enough detail to be distinguished, plus a Gujarati regional variant and a Vedic derivation — six accounts across three contributors, which is the fullest treatment of any single question in this group.123
No chapter or verse for any of the four. The texts are named, the passages are not. [RENDERING — verify against Hindi], and the Śiva Purāṇa Śatarudra Saṃhitā account is the one most likely to be checkable, since another essay in the volume gives that Saṃhitā chapter-and-verse for other material.
The plurality device is available and unused, as above.
Account three has no comment attached anywhere. Two fathers layered into one conception is a considerable claim and no contributor in the corpus remarks on it, treats it as difficult, or explains what it means for the child's nature. Given how much of the volume is spent adjudicating whether he is Rudra's portion or Rudra entire, a source that makes him two beings' son simultaneously ought to have been engaged.
Open question with a clean falsification route. Does the Vinaya Patrikā account actually exclude a conception, or is Tulsīdās assuming one of the Purāṇic mechanisms and reporting only the motive? The four accounts are only incompatible if the fourth is a complete account. If it is a gloss on motive layered over any of the other three, the corpus has three accounts and a commentary, which is a much less interesting situation and a considerably more likely one.
⛑⛑ SOURCE-UPDATE, 2026-08-15 — TWO FURTHER ACCOUNTS WITH CHAPTER LOCATORS, AND A THIRD ESSAY SAYING THE MĀNAS NEVER STATES IT AT ALL.
Baldev Upādhyāy's Purāṇic survey supplies two more.5
⭐ Śiva Purāṇa, Śatarudra Saṃhitā 20: Viṣṇu takes Mohinī's form; Śiva's desire releases seed; the Saptarṣis convey it through the ear-canal of Añjanā — identified here as Gautama's daughter — into her womb. ⛑ Which makes Hanuman literally Śiva's begotten son, not an incarnation in a looser sense — the essay marks the distinction itself — and is a different claim from every version this page holds.
⭐ Bṛhaddharma Purāṇa 18, which the essay identifies as heavily shaped by Devī-tantra material: ⚠⚠ Śiva and Pārvatī were secretly dwelling in Laṅkā protecting Rāvaṇa, and Pārvatī's offence at Sītā's mistreatment prompted their departure; Śiva then agreed to become Hanuman, Brahmā to become Jāmbavān, and Dharma to become Vibhīṣaṇa. ⛑ Ch. 20 has Hanuman-as-Śiva reveal his viśvarūpa to Pārvatī in a Caṇḍikā shrine inside the Aśoka grove.
⚠ Śiva protecting Rāvaṇa is a theodicy problem of a different order from the eleventh-Rudra device, which exists to let Śiva move against his devotee — here he is on the other side entirely until his wife objects.
⛑⛑ AND THE COUNTER-FINDING IS THE SHARPEST OF THE THREE. Bhikṣu Ānand, the volume's Tulsīdās specialist, says the Mānas narrates gods descending in vanara form and never specifies which god took which form — that the familiar assignments are editorial speculation — and that the whole textual base for Hanuman-as-Śiva is one epithet, Āśutoṣ, at one scene.5 ⛑ So the identification is most explicit in the Purāṇas, with chapter numbers, and absent from the poem that founded the cult. ⚠ Upādhyāy supplies no dating and no hierarchy of authority anywhere. Full treatments: tulsidas-never-says-it, a-sin-an-expiation-a-shrine.
Shrivastava and Śrīdharācārya face the same structural problem and reach for different instruments, which is the clearest illustration in the volume of how its machinery gets distributed.
Śrīdharācārya, handling three incompatible birth accounts, invokes kalpa-plurality explicitly. The doctrine does real work: it converts a contradiction into a distribution, and it comes with a justification a reader can accept or reject.
Shrivastava, handling four incompatible sonship accounts, invokes nothing. He lays them out and moves to his conclusion about names.
The difference is not carelessness. Their essays have different subjects — Śrīdharācārya is writing about the birth, so the contradiction is his topic and he needs a device; Shrivastava is writing about names, and the four accounts are material he passes through on the way to Śaṅkarsuvan. A contradiction encountered in transit does not get the same treatment as one you sat down to address.
Which produces a general observation about this corpus worth carrying: the same disagreement gets a doctrine in the essay where it is the subject and a shrug in the essay where it is context. A reader assembling the volume's position on any question has to notice which kind of passage they are reading, and the volume gives no signal.
The Founder Problem in Historical Perspective. The historiographic difficulty is that origin accounts are produced by communities with present needs, so the founding is narrated in whatever shape the current arrangement requires, and multiple incompatible foundings survive because each serves a constituency. The four sonship accounts are that pattern with unusual clarity, because the constituencies are legible — a Rāma-tradition account tying the birth to Daśaratha's sacrifice, a cross-sectarian account routing through Mohinī, a reconciling account preserving both paternities, and a devotional account with no biology at all. What the pairing produces is a test the historiographic material can use: sort origin accounts by what each licenses rather than by attestation, and check whether the sort is cleaner than the evidential one. Here it is dramatically cleaner. When the functional sort outperforms the evidential sort, you are looking at a founding narrated by need rather than transmitted from an event — and the number of surviving versions is a measure of how many constituencies the figure had to serve.
Keep It Vague, Keep It Simple. The influence principle is that ambiguity travels further than specificity because it lets each listener supply their own content, and that a claim precise enough to be checked is a claim precise enough to be rejected. Shrivastava's refusal to adjudicate produces exactly that effect at the level of a doctrine: a Hanuman who answers to four sonship accounts is available to four audiences, none of whom has to give anything up. But the case also exposes the limit of the principle, which the influence literature rarely states — the strategy works only while the audiences stay separate. A compendium is precisely the artefact that puts them in one room. This volume prints all four accounts within a hundred and fifty pages, which converts a portfolio of audience-specific ambiguities into a visible set of contradictions, and the tradition then needs a second instrument — the plurality doctrine — to handle the collision its own success created.
Sharpest implication. Four accounts of one event, and the cleanest way to sort them is by what each makes possible rather than by what each attests. Push that and it becomes uncomfortable for any origin story, including the ones you hold about yourself: the version you tell is the one that licenses the conclusion you need, and the presence of several versions is not confusion but coverage. The tradition is not embarrassed by this. It calls on whichever the moment requires, and the only thing that breaks the arrangement is somebody laying all of them out on one page.
Generative questions.