A claim has been running through the whole episode — that Bhairava and Nārāyaṇa are not two things — and it has had nothing behind it but assertion. Then a place is named:
"…especially in the down[-]to[-]the[-]south there is one particular temple, especially in the Mangalore, called as Kadri ⚠️[Mañjunātha]. It is being considered as a combination of a Hari-Hara union of the Nātha sampradāya."1
Hari is Viṣṇu; Hara is Śiva. A sampradāya is a lineage.
And then the reason it is being brought up at all, which is a correction rather than an addition:
"You may be wondering [that the] Nātha sampradāya [is] only in the north. No — even in the Kadri … it is also connected to the Nātha sampradāya. Kadri Mañjunātha is the oldest Nātha ⚠️[associated] centre, because here it is very specially connected: Viṣṇu, Śiva and the Nātha — the yogic tradition — has been connected."1
🚩 That it is the oldest is asserted and is not checked here. No claim is made on this page about the site's dating, its history or its affiliation.
Then the detail the passage walks straight past, and it is the most substantial thing in the range:
"Viṣṇu — he is being … ārādhana has been done as ⚠️[Lokeśvara] Mañjunātha."2
Ārādhana is sustained worship.
Look at that name. Lokeśvara is Avalokiteśvara's short form — the bodhisattva of compassion, a Buddhist figure. Mañjunātha carries nātha, lord, the word the yogic order is named for.
So the deity at this site is worshipped under a name whose first half is Buddhist and whose second half is Nātha, is identified in this passage as Viṣṇu, and sits in a temple the passage calls a Śaiva and Nātha centre.
Four traditions with a claim on one image. The passage lists three of them and does not name the fourth. ⚠️ The Buddhist reading of the name is this page's observation, not the source's, and it is offered as an observation rather than as an identification.
What the merger produces gets a name:
"…all this getting merged — this particular svarūpa is called ⚠️[Harihara]-ika, the union of Śiva and ⚠️[Nārāyaṇa]."2
⚠️ The suffix is not recovered from the caption. Hari-Hara is not an unusual form in itself — a figure split down the middle, Viṣṇu on one side and Śiva on the other, is attested across the south and is old. What is being claimed here is not the iconography but the jurisdiction: that this particular Hari-Hara belongs to the Nātha lineage, and is the lineage's oldest seat. That is a considerably larger claim than a shared image, and it is made in the same breath.
The identification the whole episode has been building toward is finally stated at this site, and stated twice — first as non-separation:
"…here Bhairava and Nārāyaṇa is not separated. [It] is one particular consciousness."3
Then, a minute later, as something stronger than non-separation:
"…Śiva path — transformation, preservation — but Nātha path together, what we call it as Bhairava himself becomes Nārāyaṇa here."4
Those are not the same claim. Not separated says the distinction was never real. Becomes says there is a transition, and that it happens under a particular tradition's practice — Nātha path together.
The second is the one with practical content, because it locates the merger somewhere: not in the nature of things generally, but in what a specific lineage does at a specific place. And the setup for it comes a few minutes earlier, explaining why a Nātha site would be where this happens at all — in the Nātha view, Bhairava as first guru and as guardian of the ground is worshipped in both forms, Śiva and Viṣṇu, with no difference drawn between them.5
The claim is then extended to fire, and the extension is a functional one:
"Even [the] Himālaya[n] Nātha ⚠️[dhūnī]s, or even the aghorīs — it is also connected. Dhūnī fire, whatever it is called — it is not about only the triśūla and the Bhairava mantra … it is [an] invocation of Bhairava. It actually protects … the sādhakas."6
A dhūnī is the perpetual fire an ascetic keeps. Read what that says. The ordinary understanding is that a dhūnī is a hearth, and that a practitioner adds Bhairava to it — with a trident planted in it, with a mantra said over it. The claim here is that the fire itself is the invocation, and that the protection is the fire's presence rather than an effect produced by anything done at it.
Then, immediately: "Maybe I'll not try to disclose more about that."6 ⚠️ The withholding is recorded and not filled. What survives is the claim's shape, and it matches the other claims in the passage exactly — protection as a standing condition rather than as a response to anything.
The closing definition arrives by three refusals:
"⚠️[Nārāyaṇa] is not about [the] giver of … not the promises of the miracles — but nothing[,] far greater than that[,] is: who guardians … your destiny, who protects your fortune. The tantra of the prosperity that never leaves you."7
Not a giver. Not a miracle-worker. A guardian of what is already yours.
That is the same shift the whole passage keeps making. The Gorakhnāth-pīṭha material has the same form protecting before the danger even appears rather than defeating anything.8 The dhūnī protects by being lit rather than by being worked. And here the office is keeping rather than getting.
⚠️ A scarcity note is attached: "in this particular svarūpa of the Bhairava … very rarely, or very less information … is available in today's ārādhana."4 Unverified, and consistent with how this material treats its other rare forms.
Stand in front of the deity at this site and count what is being claimed about it. It is called Lokeśvara — a bodhisattva's name. It is called Mañjunātha — a nātha name. It is worshipped, on this account, as Viṣṇu. It stands in what the passage calls a Śaiva and Nātha centre. And the merged form is named Hari-Hara, which is a Vaiṣṇava-Śaiva compound with no Buddhist and no Nātha content in it at all.
Five names, four traditions, one object — and not one of the five is retracted in favour of another. That is the phenomenon this whole episode has been asserting, standing in a building. Everywhere else the non-separation is a claim somebody makes: Bhairava and Nārāyaṇa are one consciousness, said by a teacher, with no text behind it. Here it is a fact about an object that people go and worship, and the object has been holding the claim for longer than anybody arguing about it.
And notice what the site does not do, because this is where the case is sharpest. It does not resolve the claims. There is no story in the passage about one tradition winning, or about the others being reinterpreted as versions of the winner. The names are simply all still attached.
Which makes the site a different kind of evidence from the doctrine it is cited for. The doctrine says the distinction was never real. The site says the distinction was never enforced — and those are not the same finding. A thing that four traditions never got round to fighting over is not, for that reason, proof that they were all saying one thing.
⚠️ The passage does not draw this distinction, and treats the site as straightforwardly confirming the doctrine.
Nothing here is a practice, and the passage says so about the one thing that looked like one: the dhūnī material is cut off before it becomes usable. What the site supports is a way of reading claims of this kind, and one caution.
Step one — separate what a place shows from what a teacher says. The doctrine and the site are being offered as one piece of evidence and they are two. A site with four names on it establishes that four traditions coexisted around one object. It does not establish what any of them thought they were doing.
Step two — notice which claims here are checkable and which are not. Kadri Mañjunātha exists and holds a Hari-Hara form is checkable. It is the oldest Nātha-associated centre is checkable and is not checked. Bhairava becomes Nārāyaṇa is not the sort of thing a building can settle either way. They are delivered in one paragraph.
Step three — take the guardian office literally if you take it at all. Not a giver, not a worker of miracles — a guardian of what you already have. That is a claim about which practices are appropriate to this form, and it rules out most of what people bring to a temple.
Step four — leave the dhūnī alone. The one functional claim in the range is followed immediately by a refusal to say more. Anything built on top of the fire is the invocation would be built past the point the source chose to stop.
⚠️ No mantra, no count, no duration, no rite and no marker is given anywhere in this range, and the passage states elsewhere that this form has a tantra and a yantra of its own without supplying either.
WW 33:59–37:45, continuous. [PARAPHRASED] — auto-generated captions.
Unreconstructed: Mañjunātha throughout (as mjunata, manchunata, read from the Kadri temple's known name and flagged); Lokeśvara at 35:53 (as loceshwara); Harihara at 35:53 (as hariharayika, suffix not recovered); dhūnī at 36:31 (as duni, dunis); lineages at 34:44 (as "the lion ages"); Nārāyaṇa throughout (as narana, narhanna, naranda).
Trust calibration. The reasoning here is argument from existing practice, which is the strongest footing available to this material and is worth crediting. It has no text for the Nārāyaṇa-Bhairava doctrine anywhere, so it reaches instead for a functioning temple where the worship does not distinguish the two — and a temple is at least something a listener could go and see. The speaker also states his own limit: "I'm not trying to tell more from a Nātha sampradāya perspective, but I'm trying to tell you how exactly these are connected."3 🚩 The weakest link is that the site is being asked to prove more than a site can. That people worship one image under several names is a fact about practice; that the underlying realities are one is a claim about what is there, and no building settles it.
🚩 NOT VERIFIED — no claim is made here about the site's dating, history or affiliation. 🚩 NO LOCATOR anywhere in the passage. 🚩 SINGLE SOURCE.
Density dropped from high to medium on 2026-08-25. The cited range was read in full, and a substantial part of it — the four predicates and the protects before the danger claim at 34:25 to 34:44 — belongs to the companion page rather than to this one.8 What is left is the site itself, the Hari-Hara merger and the dhūnī, which does not carry a high tier honestly. Restored from inside the range: the both-forms setup at 33:59, the Bhairava himself becomes Nārāyaṇa formulation at 36:57, and the scarcity note attached to it. The previous version carried the site and the dhūnī and not the sentence the site is cited to support.
Open. The suffix in harihara-ika is not recovered, and the passage does not remark on the deity carrying a bodhisattva's name.
Two claims about the relation between the two deities sit ninety seconds apart and they are not compatible. Not separated. At the site, Bhairava and Nārāyaṇa is not separated; it is one particular consciousness. Follow that and there was never a distinction to be crossed — the appearance of two is the error, and the site simply shows what was always the case.
Becomes. A minute later, Nātha path together, what we call it as Bhairava himself becomes Nārāyaṇa here. Follow that and there is a transition, it takes place under a particular lineage's practice, and it happens here rather than everywhere.
The second has a location and a mechanism in it. The first has neither and does not need them. One is a claim about how things are; the other is a claim about what a tradition does, and the passage moves between them as though the site established both.
A second strain runs between the site and the argument it is recruited for. The whole episode is about a Bhairava form and its Viṣṇu-side fusion. The site is presented as a Nātha centre, and its claim to significance is that it corrects a northern assumption about the Nātha order.
So the evidence and the conclusion are about different things. A Nātha site holding a Hari-Hara form supports a claim about the Nātha order's geographical reach. It supports a claim about Bhairava and Nārāyaṇa only if the Nātha identification of the two is already granted — which is what the site was brought in to establish.
Vaudeville's Sufi-Jogi Conversion Thesis — where the structural finding is about what incoming Sufis actually met in mediaeval north India. Not a clean doctrinal field with a boundary to be crossed, but an existing popular religion in which the jogi was already the recognisable figure of religious power — so the encounter looks far less like conversion and far more like one more layer arriving on ground that was already layered.
That is this temple, at the other end of the subcontinent, as an object rather than as a thesis. A Buddhist name, a Vaiṣṇava identification, a Śaiva building and a Nātha claim, all attached to one image and none of them removed. Whatever arrived at that site last did not arrive at an empty field. It arrived at something that already had names on it, and it added one.
Which reframes what the passage is doing when it says the oldest Nātha-associated centre. The phrasing is doing more work than the speaker seems to notice — associated, not founded, not established. On the layered-ground account that is the accurate word, and it is the word a tradition uses about a place it inherited rather than built.
And the pairing supplies what neither side has alone. The historical account explains the superimposition — why a site would carry names from four traditions with no story of any of them displacing the others. What this material supplies is the thing the historical account cannot reach: what the people there now say the merged form is. Not a compromise, not a coexistence, not four claims held in an uneasy truce. One particular consciousness. That is the layered ground speaking about itself, and it is a different sort of evidence from a scholar reconstructing what the layers must have been.
Sharpest implication. The site is cited to prove that a distinction is unreal and what it actually demonstrates is that a distinction was never enforced. Four traditions have a claim on one image, five names are attached to it, and not one of them has been withdrawn in favour of another — which is exactly what you would expect of a place nobody ever fought over, and is not evidence that the parties agreed. The doctrine wants the temple to say the two were always one. The temple says only that nobody at Kadri ever needed to decide.
Generative questions.