The shadadhva: the map that is also the fetter

In Kashmir, in a rite whose instructions survive in a scripture called the Svacchanda, a teacher draws the universe onto his student. A picture would be the easy version. What goes on is the thing itself: one of six parallel versions of the universe, laid along the breath and installed in the body from the feet upward, with the other five folded inside it.

Then he burns it.

The six versions are the shadadhva, the six courses. They are what initiation in these traditions acts on, which is why it is not a welcome ceremony in the Shaiva scriptures. Something is supposed to be destroyed, and the six courses are a map of exactly what.

A note on the Sanskrit before going further. The verses below are given in transliteration and not in Devanagari. This site prints Devanagari only where three witnesses of independent provenance agree, and no verse in this article meets that bar, which is why every one of them is transliterated. For the Svacchanda this article holds one Devanagari witness and one romanised, and a later section gives reason to doubt that the two are independent of each other. Converting a transliteration back into Devanagari is how invented characters get into print, so it is not done here. Loci are given throughout, and three caveats attach to them. The Netra Tantra references follow Vrajavallabha Dwivedi’s edition in a scanned text, and the Kashmir Series printing could not be consulted. Kshemaraja quotes two verses of that sixteenth chapter while commenting on the Svacchanda, and the Kashmir Series edition prints 16.54 and 16.55 beside them; the numbers there are the modern editor’s, so they corroborate the edition rather than the manuscript tradition, and neither of the two verses used below is among them. The same apparatus is fallible: in the commentary on the fourth chapter it prints 4.158 against a quotation that is in fact 4.184 and 4.185, a transposition of the digits. The Svacchanda verse numbers follow the romanised e-text, whose numbering agrees with the printed Kashmir edition, though at least three of its readings do not: two are the subject of a later section, and at 4.200 it carries apane where the construction requires the past participle apanne, which the printed edition has and this article prints. When those same verses are quoted in the Tantraloka commentary the reference printed beside them runs one higher, 4.185 for 4.184, so a reader coming at a locus through the commentary should expect the offset. And the Tantrasara paragraph numbers used below belong to the electronic text, since those chapters are continuous prose.

What the six are

Three of the courses run on the side of extension, of things laid out:

Bhuvana, the worlds, also called pura, a word for a walled town used here for a world. Tattva, the principles or categories from which those worlds are made. Kala, the five great energies into which the whole hierarchy is gathered: nivritti, cessation; pratishtha, establishment; vidya, knowledge; shanta, the peaceful; and shantatita, that which is past the peaceful.

One warning about that last word, because this site writes Sanskrit without diacritical marks and here it costs something. Kala with a long final vowel, kalā, is the energy. Kala with a long first vowel, kāla, is time. They are different words, they are about to appear in the same two sentences, and only the Sanskrit distinguishes them.

Three run on the side of the word:

Pada, words. Mantra, the formulae built from those sounds. Varna, the phonemes.

Each course covers the entire universe. Each one is a complete map of the same territory, at different grain, and the tradition’s claim is that each is co-extensive with the others and contained in them.

The counts are what make that claim operable. The matching runs level to level, which is all the numbers permit: there are five energies and fifty phonemes. It is a nested partition. The Malinivijayottara walks the five energies one at a time, under a different set of names that Abhinavagupta notes it substitutes for the usual five, and gives each its own share of every course: the first takes a single principle, a single phoneme, a single word and a single mantra, with sixteen worlds; the second takes twenty-three principles, twenty-three phonemes beginning with ha, five words and fifty-six worlds; and so on to the fifth, which is the Shiva-principle with the sixteen vowels (2.49 to 2.57). The partition is not quite complete in either text: the Malinivijayottara gives its fifth energy no worlds at all, and adding up the shares it does give across the five is how this article arrives at the hundred and eighteen the text states elsewhere. The Svacchanda parcels out its own figures the same way, energy by energy, across five matching places from 4.102 to 4.196, and there the worlds of the fifth energy are called subtle and given no number, though the text’s own total later counts them in. Kshemaraja supplies the missing list in his commentary there. Every course is cut at the same five joints.

Abhinavagupta derives the six instead of listing them. There are two triads, and each triad has a gross, a subtle and a supreme form, so the whole path is spoken of as sixfold (Tantraloka 6.36). The Tantrasara states the pairing outright and ends on three words that carry the entire doctrine: ekasyaiva shadvidhatvam yuktam, the sixfoldness of one single thing is justified. That sentence is in its tenth chapter, on pages 111 to 112 of the Kashmir edition; the paragraph number 10.18 by which it is usually cited belongs to the electronic text, since those chapters of the Tantrasara are continuous prose and carry no verse numbers.

The word itself is the texts’ own. Adhvan is a road, and Abhinavagupta gives the etymology plainly: that by which one arrives, step by step, at the station to be reached (Tantraloka 6.30). The Svacchanda names all six in sequence from 4.242 onward, each with the word attached, and the Malinivijayottara says of its own account that thus the sixfold course has been declared in brief (2.57), a qualifier of the same force as the one it attaches to its count of the worlds.

There are two terms the scriptures do not use. The terms shabdadhvan and arthadhvan, which appear in modern expositions as the names of the two sides, occur in none of the four texts searched for them here: the Svacchanda, the Malinivijayottara, the Tantraloka and the Tantrasara. Not once. The Netra Tantra was not searched for these two terms. They appear to be a scholarly convenience that acquired the look of tradition.

Which side is time, and which is space

This is the point the popular accounts invert, and the inversion is easy to make.

The two triads are assigned to space and to time, and the assignment is the opposite of what intuition suggests. Abhinavagupta is explicit:

tatra kriyabhasanam yat so ‘dhva kalahva uchyate | varnamantrapadabhikhyam atraste ‘dhvatrayam sphutam

“That which is the manifesting of action is the course called time; the three courses named varna, mantra and pada plainly abide in it.”

— Tantraloka 6.34

yas tu murtyavabhasamshah sa deshadhva nigadyate | kalatattvapurabhikhyam antarbhutam iha trayam

“The portion that is the manifesting of form is called space; the triad named energy, principle and world is included here.”

— Tantraloka 6.35

So the word-side is the course of time, and the thing-side is the course of space. Sound unfolds serially, phoneme after phoneme, along the breath; worlds and principles are extension. Abhinavagupta does not spell that reasoning out anywhere this article could find, so it is offered as inference. The Tantrasara says the same in its sixth chapter, on page 47 of the Kashmir edition, putting varna, mantra and pada in the course of time and the principles, the worlds and the energies in the course of space.

The inversion is easy because a page lays words out in space and a journey through worlds takes time. The texts assign them the other way round.

The map is the fetter

The six courses are described in these texts in two registers at once, and the second one is what makes initiation make sense.

The Svacchanda says it twice, and the second time it needs no commentator. The two are told apart by their wording rather than by the printed numbers, which the Kashmir edition mislabels at exactly this pair. Both times the line lands at the end of a tally of what a single energy pervades. The first: mukhya hy ete smritah pashah, these are taught as the principal bonds (4.158). There Kshemaraja makes the reference explicit, introducing the half-verse with thus, those having the nature of the six courses, and explaining that they are principal because they are common to every bound soul. The second carries no gloss at all, because the root text puts both halves of the claim in a single line: mantrau dvau shadvidho ‘dhvaivam mukhyah pasha ime smritah, two mantras, thus the sixfold course, these are taught as the principal bonds (4.184). A little later the text says they are to be purified, and Kshemaraja again supplies the subject: all these very bonds are to be purified (4.197). And the payoff is stated without hedging:

etasmin shuddhim apanne muktim apnoti dikshitah

“When purification of this is attained, the initiated one attains liberation.”

— Svacchandatantra 4.200

Two verses before that, the text says the sixfold course is to be known as one (4.198). That line is also the only place in the whole Svacchanda where the bare word shadadhvan appears. Elsewhere the text says adhvashatka, the set of six, or joins shadvidha, sixfold, to adhvan, as it does in the half-line quoted just above.

The Netra Tantra puts it from the other end, and adds a condition the Svacchanda leaves implicit. By that very sixfold course, it says, bondage, which is impurity, is declared (16.60). Kshemaraja glosses that very: by the course not recognised as Paramashiva, the supreme Shiva. Two verses earlier the same chapter says the officiant who worships beyond the six courses is himself Shiva, a rescuer of all (16.58). The map binds whoever reads it as a separate thing.

The word the text uses is mukhya, principal, and it is picked because something else is subtler. The half-line that calls the courses the principal bonds goes straight on to say that the subtle ones are to be contemplated within them, and the next verse says which those are: the purification of the subtle elements and the sense-organs as other tantras establish it, the six sheaths, and the five objects, all to be taken as included (4.158 to 4.159). When the same formula returns at 4.184 and 4.185, Kshemaraja glosses the subtle bonds with two words, as before, sending the reader back to that list.

Separately, and fifty verses earlier, the chapter says the triple bond, the three impurities, is to be contemplated as situated in the course, and that initiation is performed for the purification of that (4.106). Nothing in the root text or the commentary runs the two statements together, and this article does not either.

The thing the teacher draws onto the student is also what is holding the student down. The rite is a controlled demolition of it.

That is one register, and it is not the only one. Abhinavagupta opens his treatment of the courses from the other end entirely: the whole of this course rests in consciousness, he says, and the course is the Great Lord’s own manifesting of form and action (Tantraloka 6.28 to 6.29). There the six are Shiva displaying himself, with no bondage in view at all. In a non-dual reading both are true of one reality differently known, which is the school’s whole position. The bond-register is the one the initiation liturgy runs on. The manifestation-register is the one Abhinavagupta opens with. Neither text treats the other as a correction.

The register this article leans on belongs to a contested text. The Svacchandatantra was claimed by both camps: Kshemaraja wrote his commentary to clear away dualist readings and argued against Sadyojyotih and against Bhullaka, whom Arraj can identify only as the presumed dualist author of a commentary on the Svacchanda itself, the dualist Siddhanta being the Shaiva school that held god and soul to be finally distinct, and William Arraj’s verdict after studying the whole scripture is that it is neither explicitly dualist nor clearly non-dualist, which is exactly what made it worth claiming.

Where the scheme does its work is the initiation liturgy. In the Svacchanda the six courses are set out inside the initiation rite. The text does have a cosmology chapter, a long tenth book of worlds, and Arraj judges that one to carry only marginal ritual justification. Inside the fourth chapter the courses are furniture for a rite. The division is less clean than that sounds, and Arraj is the one who complicates it: he also finds a purificatory initiation sitting inside the cosmology account, and a later section returns to it.

Reading the courses as the fetter also explains where the material sits in Abhinavagupta’s own architecture. The six courses occupy seven chapters of the Tantraloka, from the sixth to the twelfth, and Jayaratha says outright at the opening of the sixth that everything from there to the twelfth is a limb of the anavopaya, the lowest of Abhinavagupta’s four means, the one that works through breath, body and outward place. The most elaborate cosmology in the tradition is filed under the most mediated method. Whether that placement is a judgement about the scheme or only about where the material was useful, the text does not say.

And it explains the shape of the ritual action, which is otherwise peculiar. The Svacchanda has the officiant take the cord of bonds and hang it on the disciple’s body, and then join the course in the fire. The course stands in six places at once: the jug, the diagram, the fire, the cord, the officiant’s own self and the disciple, and all six are brought to a single junction (4.92 to 4.93). Kshemaraja counts them out as six substrates and is careful about one of them, noting that the self in question is the teacher’s. A map is being made several times over so that one copy can be destroyed.

And the text is literal about the destruction. At that junction the thing to be purified is resolved in the fire-pit and offered as oblation, a hundred and eight times, each offering made with the root mantra (4.94). When the pervasion has been contemplated and the course set in place, three oblations by the dhruva mantra follow, the fixed one, the syllable om, and from that, the text says, comes the purification of the course (4.99).

What the rite actually traverses

The formula repeated almost everywhere is that initiation traverses all six courses. The number is right and the mechanism it suggests is not.

The rite takes one course as primary and manifest. The other five are present in it, pervaded by it, as latent power. Kshemaraja states the mechanism in his Netra Tantra commentary, glossing that text at 4.4: taking one single course to be purified, as manifest and pervading and primary, one should contemplate the five courses contained within it as pervaded, in the form of power. The Netra Tantra itself offers both: initiation may be performed by any one of the courses (4.3), or by all six together (4.4). Kshemaraja’s gloss is attached to the second of those. The verse division here follows the scanned edition described at the head of this article, and that numbering has no second witness. Even where the scripture says all six, what the officiant does is take one as manifest and hold the rest inside it. Abhinavagupta installs the sixfold course in the body of the disciple by taking one as primary (Tantraloka 16.150 to 151).

And in the Svacchanda, the course the liturgy actually uses is the kala, the five energies. Five members, not six. Kshemaraja says so directly in the gloss on 4.91 to 4.92: here the course at issue is the course of energies, and in an initiation run on the principles, or on any of the other four, one should substitute accordingly.

What licenses the shortcut is mutual pervasion, and Kshemaraja gives it two images in the Svacchanda commentary, in the long gloss attached to 4.96 and 4.97. Each earlier course stands in the later as pervading it, as clay stands in the pot. Each later course stands in the earlier as power, as the tree stands in its own seed. Therefore everything consists of everything.

This is not a Kashmiri exegete’s convenience either. Commenting on the Netra Tantra he grounds the move in another scripture, closing that gloss with as it is stated in the Svacchanda and quoting the line about surveying the course by what pervades and what is pervaded (ad Netra Tantra 4.4, citing Svacchanda 4.95). Nina Mirnig reports the same doctrine in the Jayadrathayamala, from Alexis Sanderson’s unpublished edition (2020, page 267), which is a chain this article could not follow to its end.

The worlds that will not be counted

Ask how many worlds the course of worlds contains and the sources give different answers.

The Malinivijayottara says a hundred and eighteen, and then says, in the same breath, that this is by abridgement and not in full (5.33). The Svacchandatantra counts two hundred and twenty-four (4.199). A passage of the Svacchanda tallies two hundred and seven worlds for one stretch of the hierarchy and then adds that further worlds are held inside those, a tally Arraj works through in his tenth chapter. Two hundred and twenty-six circulates in modern expositions. Subhash Kak’s account of the worlds attributes that figure to the Svacchandatantra and the Tantraloka and cites no verse in either, and no verse for it could be found in either text for this article.

Abhinavagupta’s handling is easy to read past. At Tantraloka 8.428 to 8.433 he sets out a much larger reckoning, and it is not his: it is the Matangaparameshvara’s, the Matanga for short. He says two verses later why he has set it out at all, which is to promulgate that scripture (8.435). Then at 8.436 he turns to the hundred and eighteen and names its source, in the Shripurva’s teaching, meaning the Malinivijayottara. And then he chooses it. That, he says in the next half-verse, is pradhana here, the principal one; the surplus worlds he will state only in brief. Jayaratha explains why: because this whole treatise proceeds on that scripture’s authority.

So he does not float a figure neutrally. He canvasses a rival reckoning, credits it to the text it belongs to, and then adopts one count by name and builds on it. The Tantrasara uses that count too: its seventh chapter enumerates the worlds energy by energy, and the printed gloss totals them at a hundred and eighteen, quoting the Malinivijayottara verse that states the figure without naming where it comes from. Jayaratha quotes the same verse in the Tantraloka commentary and says only that is stated there; the number 5.33 is the modern editor’s, printed beside the quotation.

Kshemaraja is asked, in effect, to choose, and refuses. Commenting on the Netra Tantra at 4.2 he says the worlds are a hundred and eighteen by the Malinivijayottara’s procedure and two hundred and twenty-four in the Svacchanda’s manner, and that since this scripture is a compendium of all the streams there is no contradiction in the order of initiation, because the division of the six courses is constructed as each respective scripture states it.

That is a refusal to harmonise, and it comes with a reason. The number of worlds is relative to the scripture you are working in, and the rite still functions, because what the rite needs is a complete course, whatever its census comes to.

The list inside the list

The same instability runs one level down, inside the course of principles.

The thirty-six tattvas came to the Shaivas already assembled. They took the twenty-five principles of Samkhya, the old enumerating school, entire, and capped them with more at the top. The capping is not a simple addition, since the imported terms change office. In the Shaiva list purusha is a bound soul, with the sheaths above it supplying what binds, and maya stands above prakriti as a principle in its own right rather than as the English word illusion. The list itself is the authority for that, and it can be read off the enumerations cited above; what the change meant doctrinally is not something this article went to the commentators for.

The number of additions was not fixed. It ranges from five in the Rauravasutrasamgraha to fourteen in the Mrigendra, with eleven appearing in the texts where the figure eventually settled, and becoming standard only in the post-scriptural Siddhanta. That range is reported from Christopher Wallis, citing Dominic Goodall’s edition of the Parakhyatantra. It reaches this article at second hand, through a research summary. Wallis’s own text could not be consulted. Goodall’s edition could, and the pages the summary names are about something else entirely, so the range is printed here uncorroborated by the authority given for it. Mrinal Kaul, in a 2018 study of the hierarchy in the Tantraloka, adds that even the agreed eleven were put in different orders text to text, with the Sarvajnanottara omitting niyati, necessity, altogether and the Svayambhuva, the Kirana and the Matangaparameshvara each sequencing them differently, and that the Trika, the three-goddess school Abhinavagupta belonged to, then pushed beyond thirty-six to a thirty-seventh and a thirty-eighth.

Abhinavagupta entertains both of those himself in the Tantrasara and then stops the series. Contemplate the supreme and you name a thirty-seventh to stand under it; contemplate that in turn and you name a thirty-eighth. No further rung follows, he says, because once the contemplated one is made an object of knowledge the series terminates in the thirty-seventh alone (Tantrasara 10.15). Jayaratha puts the reason the other way up, as a knower-and-known relation running both directions between the two upper rungs, so that no regress gets started. Jayaratha is blunter about what the extra rungs are. No such principle really exists, he writes; the supreme reality that all the scriptures sing as the thirty-sixth is undivided awareness and nothing else, and the thirty-seventh is that same reality separated out under a description (Tantraloka 11.21 onward, with Jayaratha’s commentary, the Viveka). The Netra Tantra and the Tantraloka both present thirty-six, eighteen, nine, five, three and one as deliberate alternatives, each available for use (Netra Tantra 4.1 to 4.2; Tantraloka 11.34 to 11.40).

So thirty-six is the standard reckoning over an unsettled list, and even the reckoning has alternatives beside it. And the course of principles is only one course of six, which means the scheme’s claim that six parallel series pervade one another is doing a great deal of work with a term whose membership the scriptures never agreed on.

Where the arithmetic stops mattering

Having set out the five-principle, three-principle and single-principle reckonings, Abhinavagupta writes: tattvadhvaivayam ittham cha na shadadhvasthiteh kshatih, which is to say that this is just the course of principles taken in that way, and there is no damage to the sixfoldness of the path (Tantraloka 11.36).

Jayaratha frames it with the standard commentarial objection, the nanu, but surely, that introduces passage after passage in his commentary. Whether it registers a real worry or only the form of one is not something the text settles. The objection itself is narrow and taxonomic. If the principles can be reckoned as three, or as one, is that a further course of its own? Abhinavagupta answers that it is still the course of principles at another grain, so the sixfoldness is untouched.

The six holds because the courses are grains of one thing, which is the mutual-pervasion doctrine set out earlier, applied to a case that looked like a counter-example. He is showing that the arithmetic was never what carried the scheme.

If the contents move and the count of six does not, what is the six doing? What the texts support is that the sixfoldness is derived. It falls out of two sides, each taken at a gross, a subtle and a supreme grade, as Tantraloka 6.36 says. It is a structural necessity of the derivation. Nobody counted the world and arrived at six.

Where the sixfold may have come from

There is a case, and it has a named holder, that the sixfold is a later envelope around something simpler.

William Arraj, whose 1988 Chicago dissertation on the Svacchandatantra remains the fullest free study of the text, argues that the initiation conducted through any of the six courses appears to have developed through the superimposition of this cosmology on an earlier rite that severed the triple impurity. He notes that the purificatory initiation sitting in the middle of the cosmology account may be a survival of that simpler rite.

One thing in the text presses the same way: the working liturgy uses the course of energies alone.

One datum belongs beside the thesis. The Tantrasara names a station beyond the five energies, kalatita, that which is past all energy, and then closes the count in the very next clause: thus there are just five energies and thirty-six principles (Tantrasara 10.10 to 10.11). The five are the steadiest list in the scheme, and what a reader makes of that against a thesis about later envelopes is left here where the texts leave it.

No answer to Arraj could be found to report. So far as this article could establish, nobody has taken the thesis up in print in either direction, and no argument for the historical integrity of the sixfold was found. Three of the places such an argument would sit, Sanderson’s papers, Padoux’s sixth chapter and Bhaskarakantha’s ninth, were not read here. The Tantraloka sentence from the previous section does not do that work: Abhinavagupta is settling a question about grain there, and the layers are a modern question.

A related question has to be left open. Whether the scheme began in the dualist Shaiva Siddhanta and was taken over by the non-dualists is a question this article could find nobody answering. The same structure turns up outside the Shaiva non-dualists altogether. It is in the dualist Siddhanta, and it is in the Pancharatra, the Vaishnava tantric tradition, where the Paushkarasamhita lists a shadadhvamandala among its twenty-five diagrams and the Lakshmitantra treats the courses in its twenty-second chapter. Andre Padoux is reported as noting the shared possession. No scholarship devoted to the Shaiva and Pancharatra comparison could be reached at all, so there is no literature here to report, only an open question. Those two loci come from this article’s research notes and neither Vaishnava text was read directly.

Where the exegetes and the logicians stop meeting

What is attributed to Andre Padoux in this section comes at two removes, through later work that quotes him. His Vac is reported to be the standard treatment of exactly the question below, and the section on what could not be asked says what that absence costs.

The boldest part of the scheme is the claim that the word-side and the thing-side correspond. Phonemes to energies, mantras to principles, words to worlds. The whole of Kashmir Shaivism’s metaphysics of speech leans on it.

The exegetes assert it, and Kshemaraja builds his whole account of the courses on it. Padoux is reported as reaching the six courses only after his chapters on the levels of the Word, so that in his architecture a doctrine of language ends in a cosmology. That is a claim about the shape of a book this article has not opened.

But Somananda, who founded the school this article has been quoting, refused the move that would ground it.

Somananda, in the Shivadrishti, attacked Bhartrihari’s doctrine of the levels of speech and specifically declined to identify pashyanti, the seeing word, with the absolute. Evgeniya Desnitskaya sets herself against the usual scholarly verdict that Somananda simply misread Bhartrihari. She finds no trace of a fourth level of speech in Bhartrihari at all, and suggests that what Somananda was targeting was a real ambiguity in the ontological status of speech. Bhartrihari’s three levels run from vaikhari, speech as uttered, through madhyama, speech in the middle, to pashyanti, speech as seen. The later Pratyabhijna, the Recognition school, took those three over and set a fourth, para, the supreme, above them. Its epistemologists went another way with the material: Jesse Berger shows Utpaladeva, the school’s founding logician, working out the relation of signifier to signified without turning it into a cosmos at all.

So the parallelism is asserted by the ritual and exegetical lineage, and the philosophers of language in the same school are not found underwriting it. That is a silence, and a silence is weaker evidence than a denial. What the exegetes assert and what the epistemologists discuss do not meet, and no way of making them meet was found here.

An eighth course in a modern e-text

The last instability is in the present tense. GRETIL, the standard online archive of Sanskrit e-texts, carries the Svacchandatantra. At 4.223 in its own numbering its text reads adhvashtaka, which looks like a compound for an eightfold course and is not one: the compound would lengthen the vowel. At 4.241 it reads adhvashtka, the same transposition with the vowel gone too, which is no word at all.

The printed Kashmir edition reads adhvashatka, sixfold, in both places, and Kshemaraja’s commentary glosses the sixfold division. The word for an eightfold course does not occur in the volume at all. The word for the sixfold course occurs five times in the printed volume, three in the root text and twice in the commentary.

The GRETIL file contradicts itself, reading sixfold at 4.103, so nobody searching it could hold an eighth course for long. And the file is no independent witness: chapter four of it was typed in from the printed Kashmir edition in 2004, by the Shaiva reading group at the EFEO, the Ecole francaise d’Extreme-Orient, in Pondicherry. The header says as much in terms, naming the KSTS edition as the base for that chapter and S. A. S. Sarma as the person who entered it. What circulates now is a conversion of that file dated 2020, and the header does not record at which of the two steps the slip entered.

So this looks like a copying slip, and no manuscript was consulted here that could say otherwise. Nobody is known to have been misled by it. It bears on something larger. A file typed from a printed edition is not a second witness to that edition, whatever its divergences, and this article could not establish how far the two stand apart. That is one reason nothing here is offered as three-witness verified. And it shows how quietly a number can move when the copy is faster to search than the page it came from.

How one number became the number in English

The hundred and eighteen is the figure an English reader meets, and its route into English is traceable.

Swami Lakshman Joo taught this material in Kashmir across the middle of the twentieth century. John Hughes edited the teaching into Kashmir Shaivism: The Secret Supreme, published in 1985, and the lineage’s academy publishes the relevant chapter online. It sets the division of the courses out in the usual way. It states that the worlds are said to number a hundred and eighteen. And its warrant for the number is the experience of yogins, practitioners, in samadhi, contemplative absorption. No verse is cited, and no text beyond a general reference to the Tantraloka.

That is the channel. Calling it a distortion would be too strong, since a hundred and eighteen is a real scriptural figure. It is a narrowing: one text’s abridged count, presented without the other counts beside it, and grounded in attainment, with no locus given.

Both Alexis Sanderson and Mark Dyczkowski are widely reported to have studied with Lakshman Joo, though this article traced the claim no further than that report. If it holds, the scholarly and devotional streams share a root. And Dyczkowski’s own completed translation of the Tantraloka, finished in 2023, is reported as giving two hundred and twenty-four for the eighth chapter. There is less conflict there than it looks. The enumeration that runs through most of the eighth chapter, before the hundred and eighteen is adopted at 8.436, is the Svacchanda’s own: Jayaratha marks the seam, saying that having stated the division by the Svacchanda’s procedure, Abhinavagupta now states it by another scripture’s method. The chapter never totals that enumeration, so where a translator’s two hundred and twenty-four comes from could not be established here. The same lineage produced both numbers.

What could not be asked

Several things bearing directly on this were not read for this article. Alexis Sanderson’s published work on Shaiva ritual and on the Shaiva exegesis of Kashmir is the obvious place to find a position on the six courses, and none of it could be read for this article. Andre Padoux’s Vac is reported to devote its sixth chapter to exactly this scheme and to be the standard treatment of it; not a page of it was read for this article, which is why nothing above cites a page of it, and why what is attributed to him here comes at two removes, through a 2014 study by McCarter and through the notes to the Cidvilasastava edition, both of which quote him. Bhaskarakantha’s Chittanubodhashastra, whose ninth chapter is the one dedicated pre-modern Kashmirian exposition of the six, has a critical edition of its first three chapters only, by Stanislav Jager in 2018. The ninth exists in Sushma Pandey’s 1990 Benares edition of the whole work and in no translation. It was not read for this article. Navjivan Rastogi’s substantive treatment of the hierarchy, most likely his Introduction to the Tantraloka of 1987, could not be obtained either.

Nor could this article establish which text first states the hundred and eighteen, or how that figure relates to the two hundred and twenty-four, the two hundred and seven and the two hundred and twenty-six.

What the texts show the scheme doing

The scheme is usually introduced as Shaiva cosmology. That is true, and it is not what the texts spend their pages on. Six complete maps of the universe sit in these books, and the liturgy lays one of them over a jug, a diagram, a fire, a cord, the officiant and the disciple, and then takes it apart in order. The worlds get counted because somebody has to walk through them. Why the phonemes should answer to the energies at all is the question this article could not answer, and the exegetes assert it without argument.

And the tradition was candid about the seams. Abhinavagupta sets out a rival reckoning, says whose it is, and then names the count he is going to work with. Jayaratha supplies the reason it is that one: the treatise proceeds on that scripture’s authority. Kshemaraja names two incompatible figures and says the rite does not depend on choosing between them. Neither man pretends the numbers agree. Each does in public what a careful person does with an inherited system that does not quite close.

The six courses map everything there is.

They were drawn in order to be burned.