In the volume's longest chapter, a practitioner about to perform a rite stops and does something that looks, at first reading, like startling generosity.
He names four Śaiva sects — including the one that carries a skull — and accepts all four. He names the right-hand and left-hand paths and says there is no conflict. He names Sāṃkhya, Vaiśeṣika, Yoga and Nyāya and calls their followers "virtuous and distinguished." He names the followers of Sūrya, Brahmā, Rudra and Viṣṇu and says the same.
Read the verb, though. It is the same in every clause and it is not approve:
"Let all of them grant me permission that I should accomplish my objective."1
He is not endorsing anybody. He is asking every party who might have a claim to sign off, one constituency at a time, so that the rite he is about to perform cannot be challenged. And then he names the parties whose permission he is not seeking, and tells them to stay away.
This is not ecumenism. It is a title search.
The list itself is the most important sectarian self-location in three volumes, and it was missed entirely by an earlier pass because it sits in ch 453-8.2(31) — at 4,962 words the volume's longest chapter, twice the length of the next, and unread until a later sweep.2
*"There are those who follow the path of Shaiva Siddhantas. There are those who follow Shiva's Pashupata vow. There are those who follow Shiva's great vratas. There are others who are devoted to Shiva and follow the kapalika mode. They obey Shiva's instructions and as a result of Shiva's command, I have worshipped them too."*1
Four paths, named in one breath, with the kāpālika mode fourth and unqualified.
⚑ And the acceptance is not passive. "I have worshipped them too" — the speaker reports having done so, and gives Śiva's command as the reason.
⚑ The dakṣiṇa/vāma division follows immediately, and it is the flattest statement of its kind in the corpus:
*"There are those who are devoted to dakshina-jnana and those who follow paths beyond the dakshina one. There is no conflict with them."*1
⛔ Reported as what is on the page. This build does not adjudicate what these four terms denoted, when, or to whom — that is a chronology question requiring sources this build does not have, and it goes to the source map.
The four paths are one item in a much longer sequence, and reading the whole sequence is what changes the finding.
The same speech, in order, seeks permission from:1
⚑ That is a comprehensive enumeration of every constituency that could conceivably object, and it runs from scripture through lineage through rival philosophy through rival deity to rival sect. Nothing is omitted except by design.
⚑⚑ And it sits against ch 381-7(16), in the Kailāsa Saṃhitā, which does the opposite to one of the same groups: those "debating sacred texts composed by Kapila and others… in ancient times, they indulged in criticizing Shiva and Shivaa and were cursed by the six sages."3 Sāṃkhya's followers are virtuous and distinguished in one Saṃhitā and cursed in another. Unflagged, and neither passage knows about the other.
The whole page turns on one grammatical observation, and it is worth making carefully because the passage is easy to over-read in a warm direction.
Every clause has the same shape: there are those who [X]… let all of them grant me permission that this karma is successfully completed and is fruitful.1
That is a licensing formula, not a creed. What it asserts is not that the four paths are correct, or that Sāṃkhya is true, or that Viṣṇu's followers are right. It asserts that each of them has standing, and that the speaker wants their consent before proceeding.
⚑ Three consequences follow, and they are more interesting than a doctrine of tolerance would be.
Standing implies power. You do not seek permission from a party who cannot obstruct you. The list is therefore a map of who could plausibly spoil a rite — which is a much more concrete piece of evidence about the speaker's world than a statement of belief would be.
The subordination is intact and stated. "All of them are virtuous and distinguished, controlled by Shiva's commands", and the four paths "obey Shiva's instructions." ⚑ Everyone is included, and everyone is included as a subordinate. The generosity and the hierarchy are in the same sentence.
And it is instrumental. The stated purpose is "so that it is successful" — the permissions are sought because the rite requires them, not because the parties deserve recognition. ⛔ Which does not make the passage insincere; it makes it a different kind of document than a tolerance statement, and it should not be cited as one.
Immediately after the permissions, the same speech names who must stay away — and the criterion changes completely:
"However, let non-believers, deceivers, ingrates, those who follow darkness, heretics and great sinners maintain their distance from me."1
⚑ Not one of those six categories is a birth category. No varṇa, no jāti, no lineage. Belief and conduct, twice each: two about belief (non-believers, heretics), four about conduct (deceivers, ingrates, those who follow darkness, great sinners).
⚑⚑ Which gives the volume at least two independent axes of exclusion that never appear in one sentence. Elsewhere it excludes śūdras from worship rights, twice, in two Saṃhitās, and grants women rights only through a chain of male permission.4 Here it excludes by belief and behaviour and says nothing about birth at all.
⛔ Stated at that strength: two axes, in two places, never reconciled. No claim is made that either supersedes the other.
One clause in the permission-list is a cross-volume corroboration, and it is a useful one.
"There are the instructors, starting with Shveta and ending with Nakulisha, along with their shishyas."1
That is the Pāśupata roll of Volume 2 ch 233–234, named by its endpoints. That roll runs twenty-eight teachers: the first is Śveta — "I will be born as a great sage named Shveta" — and the last is Lakuli, who enters an abandoned corpse in a cremation ground at a site the translator locates as Kārvān in Vadodara district.5
⚑ Two volumes, one lineage, endpoints agreeing. That is worth recording, because the same lineage's arithmetic does not agree: Volume 2 announces 112 disciples and names 108; Volume 3 ch 431-8.2(9) announces 112 and numbers 106, with endnote 1948 diagnosing a lost shloka.5
⚑ The frame is stable across volumes and the contents are not — which is the corpus's characteristic failure mode in miniature, and it appears here in the one structure the vault had been treating as its control.
And one further item, which is a datum rather than an argument.
"There are other Shiva agama texts, divided into four categories, Kamika and the others. They are especially honoured by Shiva and Shivaa and have been worshipped by me."1
The Śaiva Āgama corpus, cited as an existing body of scripture, classified into four categories, with the first named.
⛔ This build does not adjudicate the chronology and no page in this vault may. What is on the page is a Purāṇa naming a non-Purāṇic scriptural corpus by category and by title, and treating it as an authority whose permission is worth seeking.
⚑ It belongs with the volume's other citations of named external authorities — the Śiva Sūtras quoted by number and expounded, the Śvetāśvatara, the Atharvaśiras, the Chāndogya, and Piṅgala's chandaḥ-sūtra, all in the Kailāsa Saṃhitā.6 Top item for the source map, alongside those.
You are reading a document that lists a lot of parties approvingly and you want to know what it is actually doing.
Read the verb, not the list. We recognise, we celebrate, we accept, we seek permission from — these are four different documents with the same roster. ⚑ Here the verb is permission, uniformly, and it converts a statement of tolerance into a clearance procedure.
Then treat the list as a map of obstruction. You do not ask consent from parties who cannot stop you. An enumeration of whose permission is sought is an enumeration of who has standing — which is far more informative about the real distribution of power than any claim about who is respected. Read it that way and this passage tells you which constituencies could spoil a rite in the compiler's world, in order.
Check whether the subordination is in the same sentence as the inclusion. It usually is, and it usually goes unnoticed because the inclusion is the surprising part. "All of them are virtuous and distinguished, controlled by Shiva's commands" is one clause doing both jobs.
And read the exclusion list for its criterion, not its contents. ⚑ Who is excluded matters less than what kind of thing gets you excluded. Here it is conduct and belief — and the same document elsewhere excludes by birth, on a completely different axis, without ever putting the two together. A body with two exclusion criteria that never meet has two constituencies it is talking to.
Evidence. Ch 453-8.2(31) read in full — the volume's longest chapter, 4,962 words — with the complete permission-sequence, the four paths, the dakṣiṇa clause, the rival-darśana and rival-deity clauses, the Āgama clause, the Śveta-to-Nakulīśa lineage clause and the exclusion list all quoted directly, and endnotes 2069–2071 in range.12 Ch 381-7(16)'s Kapila polemic read in place.3 The Volume 2 Pāśupata roll and its arithmetic taken from that page as corrected in this build.5
Tension 1 — the chapter was never opened in the first pass. ⚑ It is the volume's longest by a factor of two and an earlier inventory had nothing from it. The single largest coverage gap in the build, and it held the most important sectarian passage in three volumes.
Tension 2 — Sāṃkhya is virtuous here and cursed twelve chapters and one Saṃhitā away. ⛔ Preserved, not resolved. ⚑ And it is a case the corpus's own reconciliation machinery could not have handled anyway — kalpa-variance addresses rival narratives and has nothing to say about two assessments of a living school.
Tension 3 — a permission-formula is not a description of anybody's practice. The passage tells you which constituencies the compiler thought had standing. It does not tell you that four Śaiva sects existed in the form named, or in what relation, or where. ⛔ Nothing on this page is a claim about any community or period.
Open question. The permission-list names the Āgamas "divided into four categories, Kamika and the others" and the lineage "starting with Shveta and ending with Nakulisha." Both are pointers outward, and both are the kind of pointer that could be checked against the named corpora. Neither has been. Top source-map items.
Against The Pāśupata Charter: the direct corroboration and the direct complication. The endpoints match across volumes — Śveta first, Lakuli/Nakulīśa last — and the disciple arithmetic does not, running 108 in Volume 2 and 106 in Volume 3 against an announced 112 both times. ⚑ A lineage whose frame is stable and whose contents are damaged in two different places is a better argument for its antiquity than a clean one would be, and this page states that as a reading rather than a finding.
Against The Sectarian Exclusions: the same corpus, the opposite register, and the pages must be read together. That page's material is exclusion issued as curse and heresy manufactured to order. This passage excludes six categories and not one of them is a birth or a named rival tradition — they are non-believers, deceivers, ingrates, followers of darkness, heretics and great sinners. ⚑ The corpus's loud exclusions are of named groups; its quiet ones are of dispositions.
Against The Corpus of the Disqualified: the second axis, stated. That page finds worship rights restricted by varṇa in two Saṃhitās and women's rights routed through male permission. Here the gate is conduct and belief. ⚑ Two criteria, both absolute, never in one sentence — which is the volume's standing behaviour on every question it holds two positions about.
Against The Kailāsa Is a Different Book: the Kapila contradiction falls exactly on that page's boundary. Kailāsa curses Sāṃkhya's followers; the Vāyavīya calls them virtuous and distinguished. Add it to that page's list of things the two Saṃhitās disagree about — and note that this one is not a technical count but an assessment of a living school, which is a different species of divergence.
Institutional Override — that page's territory is how an institution's claim is made to prevail over competing claims without the competitors having to be defeated.
The permission-sequence is an unusually complete specimen, because it does the whole operation in one direction and in one paragraph, and the mechanism is visible in the grammar rather than inferred from outcomes.
Nobody is refuted here. Sāṃkhya is not argued with; the kāpālikas are not disciplined; Viṣṇu's followers are not ranked below anybody explicitly. Every constituency is granted standing — and every grant arrives with the same subordinating clause attached: "controlled by Shiva's commands", "they obey Shiva's instructions." ⚑ The override is not performed against the rivals. It is performed by describing the frame inside which their validity holds.
That is a distinct and cheaper technique than the ones usually catalogued. Defeating a rival requires an argument, keeps the rival in play, and invites a reply. Granting a rival full validity within your frame requires nothing, produces no opponent, and cannot easily be objected to — because the rival's own adherents get everything they wanted except the frame, which was never the thing they were defending.
⚑ And the diagnostic that falls out is precise and portable: when an institution is conspicuously generous, read the subordinate clause. The generosity is real and the clause is the point. All of them are virtuous and distinguished is doing no work; controlled by Shiva's commands is doing all of it. The tell is that the two arrive together, in the same sentence, and that the second never appears alone — because alone it would be a claim, and claims get disputed.
Running back the other way, this passage supplies the override's cost: the speaker has to ask. He enumerates every party whose permission he needs, which means the frame he is asserting is not yet secure enough to be assumed. A tradition that had actually overridden its rivals would not seek clearance from them.
Sharpest implication. In the volume's longest chapter — 4,962 words, twice the next longest, and never opened until a late pass — a practitioner about to perform a rite names four Śaiva paths and accepts all of them: "the path of Shaiva Siddhantas… Shiva's Pashupata vow… Shiva's great vratas… the kapalika mode. They obey Shiva's instructions and as a result of Shiva's command, I have worshipped them too." He then names dakṣiṇa and the paths "beyond the dakshina one" — "There is no conflict with them" — and Sāṃkhya, Vaiśeṣika, Yoga and Nyāya, and the followers of Sūrya, Brahmā, Rudra and Viṣṇu, all "virtuous and distinguished." ⚑ But the verb in every clause is permission, not approval: "Let all of them grant me permission that this karma is successfully completed." This is not a tolerance statement. It is a clearance procedure — an enumeration of every constituency that could obstruct a rite — and each grant of standing arrives welded to the same subordinating clause, "controlled by Shiva's commands." Then the exclusion list, and not one of its six categories is a birth category: non-believers, deceivers, ingrates, followers of darkness, heretics, great sinners. A volume that elsewhere bars śūdras from worship rights in two Saṃhitās here gates on conduct and belief and says nothing about birth — two absolute criteria, never in one sentence. And twelve chapters away in another Saṃhitā, the followers of Kapila are cursed by six sages for criticising Śiva.
Generative questions.