A liṅgam is a smooth vertical stone on a pedestal. There is not much to it, structurally: a base, a middle, an upper part, a top.
Ch 457-8.2(35) assigns each of those a syllable, a Veda, a guṇa, a deity, a cosmic function, a tattva, a kalā, a face of Śiva, a state of consciousness — and then says where on the stone each one sits.
*"'A' adhered itself to the southern side of the blazing lingam. 'U' was to the north and 'M' was in the centre. The sound of the half-matra was heard on top of the lingam."*
The object is a chart. Read from the pedestal upward, it delivers Ṛg / Yajur / Sāma / Atharva; rajas / sattva / tamas / nirguṇa; Brahmā / Viṣṇu / Rudra / Sadāśiva; creation / preservation / destruction / liberation; and waking / dream / sleep / turīya.
⚑ This is the most systematic passage in the volume, and it is worth having laid out because the correspondences are not decorative — the text derives one from another, states the position on the stone for each, and then breaks down in a way that is itself informative.
Each column is built the same way and the text says so: each Veda "established itself in ten aspects."1
| A | U | M | half-mātrā / nāda | |
|---|---|---|---|---|
| Veda | Ṛg | Yajur | Sāma | Atharva |
| Guṇa | rajas | sattva | tamas | nirguṇa |
| Deity | Brahmā | Viṣṇu | Rudra | Sadāśiva |
| Function | "creator of the worlds" | "preserver of the worlds" | "destroyer of the three worlds" | "the one who frees creatures" |
| Tattva | earth | space | "the supreme Shiva" | "beyond" |
| Kalā | Nivṛtti | Pratiṣṭhā | Vidyā | Śānti and Śāntyatīta |
| Face of Śiva | Sadyojāta | Vāmadeva | Aghora | Puruṣa and Īśāna |
| Position on the liṅgam | lower part | centre | just above the pedestal | the top |
| State | wakefulness | dream | sleep | "turiya and the state that is beyond turiya" |
12
⚑ Note that the position row is the one that makes this a diagram rather than a list. Every other correspondence is conceptual; the liṅgam row is spatial and physical, and it is what the other rows are hung on.
⛑ The fourth column does not behave like the other three, and the translator says so.
The first three each deliver a clean set. The Atharva column instead runs a chain of near-synonymous transcendences — "nirguna was its first aspect; next, Sadashiva… Shiva, the paramatman, devoid of kriyas in all kriyas; as compassion towards beings, the one who frees creatures; and as the world that cannot be reached through words and thoughts" — followed by two further worlds stacked above it, Unmanā and then Soma, "where he resides with Uma… a person who reaches it, does not return."1
And then it doubles the entries in the two rows that were single everywhere else: the kalās become "Shanti and Shantyatita", and the faces "Purusha and Ishana."1
Note 2085: *"Unlike the others, for Atharva Veda, it is impossible to precisely count which ten are meant."*3
⚑⚑ A four-column table in which three columns hold and the fourth cannot be counted. ⛔ And the translator's remark is a genuine admission rather than a hedge — the fourth of at least seven such admissions in this edition — which makes it evidence rather than caution.3
⚑ The interesting question is whether the breakdown is a defect or a design. A scheme whose top term is defined as beyond enumeration will fail to enumerate at the top, and this one names "the world that cannot be reached through words and thoughts" as an entry in a list of ten. The table is trying to count something it has just defined as uncountable. ⛔ That is a reading, and it is the charitable one; the uncharitable one is that a fourfold scheme was fitted onto a threefold structure and the join shows. Both are available and this page does not choose.
The scheme has a highest position — the top of the liṅgam, the half-mātrā, nāda — and then it keeps going.
*"There is the world above, known as Unmana. Beyond that, is the super world, where he resides with Uma. Ishvara always resides in the world known as Soma, along with Uma. This world is above Unmana and a person who reaches it, does not return."*1
⚑ Two further worlds stacked above the fourth column's own terminus, and the second of them is the one with the non-return clause.
⛔ Which is structurally odd in a way that is easy to miss. The fourth column was already defined as the beyond — nirguṇa, past the three guṇas, "the world that cannot be reached through words and thoughts." And then there is somewhere above it, and somewhere above that.
⚑ It is also the only place in the table where Umā appears at all. Every other cell is a syllable, a Veda, a guṇa, a god, a function, a tattva, a kalā or a face. The top world is specified by who is there with him — "where he resides with Uma" — which is a different kind of coordinate from everything below it.
⚑⚑ And it is the same move the volume makes elsewhere, on smaller scales: a final term, and then a clause after it. The eight-tier siddhi cascade ends at sixty-four and the yoga chapters continue; the mendicant reaches Śiva's own world and residual desire is still burned off there. The pattern here is that this text does not comfortably write a last line.
The table is not an inventory. It is an argument, and the conclusion is stated twice in the same passage.
*"Hence, it is said that Shiva is Pranava and Pranava is Shiva. There is no great difference between the expression and what is expressed."*2
*"Since he is the cause behind the three causes, Shiva is the supreme cause."*2
⚑ The first is an identity between a syllable and a god. The second is a ranking. The table is what connects them — because if the syllable's parts map onto the three deities and their three functions, and the syllable as a whole is Śiva, then Śiva contains the three by construction.
⚑⚑ The argument runs entirely on the correspondence and not on any claim about Śiva. Nothing in the passage asserts that Śiva is greater; it asserts that A, U and M are Brahmā, Viṣṇu and Rudra, that AUM is Śiva, and that a whole contains its parts. The disputed conclusion is derived rather than stated, which is why the fifteen uncontroversial rows are load-bearing rather than decorative.
⛔ And the identity clause is doing more work than it looks. "There is no great difference between the expression and what is expressed" is the premise that lets the mapping transfer — without it, AUM is a name for Śiva and the parts of a name are not parts of a thing. It is stated once, in passing, and everything rests on it.
The most striking feature of the passage is not the table. It is that each Veda states its own column in the first person, one after another.
"Rig Veda also stated, 'I have spoken of the state of wakefulness. Through this, I always ordain aham atmatattva.' Yajur Veda also stated, 'I have spoken of the state of dreaming. It is through this, with Vidya or lack of Vidya in me, that objects of pleasure are known.' Sama Veda also stated, 'I have spoken of the state of sleep as pervading everything. This means Shiva and the nature of what I have spoken about is tamas.' Atharva Veda also stated, 'I have spoken about turiya and the state that is beyond turiya.'"2
⚑ The correspondence table is delivered as testimony. Not the Ṛg Veda corresponds to waking but "I have spoken of the state of wakefulness" — each Veda claiming its own cell.
⚑⚑ Which does two things at once. It makes the scheme self-attesting: the authority for the correspondence is the party being corresponded. And it makes the whole structure an argument for AUM's sufficiency — *"Hence, it is said that Shiva is Pranava and Pranava is Shiva. There is no great difference between the expression and what is expressed."*2
⛔ And it is also why this passage belongs beside the volume's finding about speech. The four Vedas deliver all of this in their own voices, and the two listeners do not believe a word of it — "those two felt that this was a dream and still did not comprehend" — until it is written on the stone.4 The most systematic doctrinal passage in the volume fails as speech and lands as an inscription, which is the sharpest possible version of that finding and the reason the two pages have to be read together.
⚑ The passage has an occasion, and it explains the shape of the scheme.
Brahmā and Viṣṇu are quarrelling about precedence and cannot find the ends of the pillar of fire. Praṇava manifests to settle a rank dispute — and the table's first three columns are exactly the three claimants, in order, at three heights on the stone.
⛔ Which is worth naming because it constrains what the diagram is doing. This is not a neutral cosmological chart that happens to include Brahmā and Viṣṇu. It is a chart on which they occupy the lower part and the centre, Rudra sits just above the pedestal, and the fourth position — the top, the half-mātrā, Sadāśiva, nirguṇa — is above all three.
⚑ And the argument is stated: *"Since he is the cause behind the three causes, Shiva is the supreme cause."* The trimūrti are not denied; they are assigned rows. ⚑⚑ A sectarian claim delivered as a correspondence table is much harder to argue with than the same claim delivered as an assertion, because the disputed conclusion arrives as one cell in a scheme whose other cells are uncontroversial.
You are reading a correspondence scheme and want to know what it is doing.
Find the row that is physical. ⚑ Most rows in a table like this are conceptual and could be reordered without anyone noticing. The liṅgam row cannot — lower part, centre, just above the pedestal, top — because it is a claim about an object somebody can walk up to. The physical row is the anchor, and every other row inherits its ordering.
Then check what the scheme was built to settle. ⛔ This one manifests during a precedence quarrel between two of the parties it ranks. ⚑ A taxonomy produced during a dispute will have the dispute's shape, and the tell here is that the three claimants occupy three consecutive positions and a fourth position exists above them.
Look at the column that does not behave. ⚑ Three columns here deliver a clean ten and the fourth cannot be counted, by the translator's own admission. ⛔ That is where the scheme's seams are, and the two candidate explanations — a term defined as beyond enumeration, or a fourfold overlay on a threefold structure — make very different predictions about the rest of the corpus.
And notice who is speaking. ⚑ The correspondences are delivered by the corresponded parties in the first person. A scheme that has its own components testify to it cannot be checked against them, which is a rhetorical property rather than a doctrinal one — and it is doing real work here, because there is no other authority in the scene.
Evidence. Ch 457-8.2(35) (Countering Hari and Vidhatri's Delusion), lines 1551 and 1553, read in full on the source file: the fourfold division of AUM and its placement on the liṅgam; each Veda's ten aspects in turn, with the guṇa, deity, function, tattva, kalā, brahman-face and liṅgam position for each; the Atharva column's chain and the Unmanā and Soma worlds; the doubled kalās and faces; and the four first-person statements of the avasthā series.12 Endnotes 2082–2086 read in full.3
⛔ Cited in the Saṃhitā-internal form — the running chapter numbers in this edition are not unique, and note numbers restart per section; neither is disclosed in the edition.5
Tension 1 — the rows are not all stated with equal explicitness. ⚑ The Veda, guṇa, deity, function, kalā, face and liṅgam-position rows are given item by item for each column. ⛔ The tattva row is thinner — earth, space, "the supreme Shiva", "beyond" — and reads more like four remarks than one series. The table above presents them in a grid; the text does not present them as a grid, and a reader should know the tabulation is this page's arrangement of a running prose passage.
Tension 2 — the fourth column's breakdown has two readings and neither is testable here. ⛔ Stated above and left open. ⚑ The translator's note establishes the breakdown as real rather than as a counting error by this build, which is the part that matters.
Tension 3 — the passage sits in the Vāyavīya, and the corpus's other Praṇava material does not agree with it. ⚑ The Kailāsa says Praṇava has five components, twice; this passage says four, and ch 428 says four as well, citing "the Upanishads."6 ⛔ So this diagram is one Saṃhitā's scheme, not the volume's — and the neighbouring Saṃhitā would have to split the top cell in two to draw it.
Open question. The scheme manifests to settle a precedence dispute and assigns its two disputants the two lowest positions. Does any other correspondence table in the corpus arrive during a quarrel — and where one does, is the outcome always encoded in the ordering?
Against Texts Are Objects; Speech Fails: the two pages are one scene read for opposite things. ⚑ This page holds what the four Vedas said; that one holds that saying it did not work. ⛔ The most systematic doctrinal passage in the volume is also its clearest case of spoken authority failing, and neither page is complete without that fact.
Against Two Hells, Two Shadows, Two Tattva Counts: this diagram is one side of that page's Praṇava pairing. ⚑ Four components here, five in the Kailāsa, each stated twice on its own side. ⛔ Which means the table on this page cannot be drawn at all in the other Saṃhitā's scheme without splitting the fourth column — and that is exactly what the Kailāsa does, into bindu and nāda.
Against The Apparatus Is the Translator's: note 2085 is a clean specimen of the good kind. ⚑ "it is impossible to precisely count which ten are meant" is an admission of non-comprehension rather than a silent repair — the fourth of at least seven in this edition — and it is doing the reader a service the twenty disclosed emendations do not.
Against Why a Swan and a Boar: the same episode, and this is its doctrinal payload. ⚑ That page holds the liṅgodbhava as a story about two gods failing to find an end; this one holds what the pillar turned out to be teaching. The search fails and the stone is a chart.
Against The Strata This Purāṇa Carries: the kalā series — Nivṛtti, Pratiṣṭhā, Vidyā, Śānti, Śāntyatīta — is Śaiva Siddhānta's five kalās, arriving here without a school name attached. ⚑ Another item for that page's list, in a passage that presents itself as pure Vedic correspondence.
Manufactured Legitimacy — that page's territory is how a claim acquires the appearance of a right to be believed.
The Purāṇic passage runs a specific version of that which is worth isolating: the disputed claim delivered as one cell of an undisputed table.
⚑ The sectarian conclusion here is Śiva is above Brahmā and Viṣṇu. Stated flatly, it is exactly the assertion the two of them are in the scene to dispute. Stated as a correspondence table, it arrives with fifteen uncontroversial neighbours — A/U/M, the four Vedas, the three guṇas, creation/preservation/destruction, waking/dream/sleep. Every one of those is common property, and each one carries the ordering.
⚑⚑ By the time the disputed row is reached, disputing it means disputing the table — and the table is mostly things nobody disagrees with. ⛔ That is a much stronger position than an assertion, because an assertion invites a counter-assertion and a scheme invites the objector to explain which of the other rows they also reject.
And the passage adds two reinforcements that the general framing does not usually get to see so plainly. The components testify for themselves — each Veda states its own cell in the first person, so the scheme's authority is the scheme's own contents. And the whole thing is anchored to an object: lower part, centre, just above the pedestal, top. ⚑ A correspondence you can point at on a stone is not obviously a proposition at all, which is the deepest version of the move — the claim stops looking like something that could be false.
⛔ Running back the other way, the corpus supplies the failure mode. The table did not work. Delivered in full, by unanimous authority, to the two people it was built for, and "those two felt that this was a dream and still did not comprehend." ⚑ A structure this elaborate persuaded nobody in the room until it was physically written down — which suggests the legibility of the anchor matters more than the elegance of the scheme, and that the fifteen uncontroversial rows bought nothing at all until the disputed one was inscribed on the object it described.
Sharpest implication. A liṅgam is a stone with a base, a middle and a top. Ch 457 makes it a chart: *"'A' adhered itself to the southern side of the blazing lingam. 'U' was to the north and 'M' was in the centre. The sound of the half-matra was heard on top of the lingam."* Then each Veda "established itself in ten aspects" — *Ṛg with rajas, Brahmā, creator, earth, Nivṛtti, Sadyojāta, "the lower part of the lingam"; Yajur with sattva, Viṣṇu, preserver, space, Pratiṣṭhā, Vāmadeva, "the centre"; Sāma with tamas, Rudra, destroyer, supreme Śiva, Vidyā, Aghora, "just above the pedestal"* — and each states its own row in the first person: *"Rig Veda also stated, 'I have spoken of the state of wakefulness'… Yajur Veda… 'the state of dreaming'… Sama Veda… 'the state of sleep'."* ⚑⚑ The scheme manifests during a precedence quarrel and assigns the two quarrellers the two lowest positions, with a fourth above all three: nirguṇa, Sadāśiva, "turiya and the state that is beyond turiya", at the top of the stone. ⛑ And the fourth column is the one that will not count. Three deliver a clean ten; the fourth stacks Unmanā above itself and Soma above that, doubles the kalās and doubles the faces, and the translator concedes *"Unlike the others, for Atharva Veda, it is impossible to precisely count which ten are meant."* A table whose top term is defined as beyond words does not tabulate at the top — which is either the scheme working exactly as designed or the seam where a fourfold overlay meets a threefold structure. ⛔ And none of it persuaded anyone. The Vedas said all of this aloud, and "those two felt that this was a dream."
Generative questions.