The volume uses the word constantly. Emancipation, liberation, moksha — it is the stated outcome of the lingam, the naivedya, the ash, the names, the counts.
And it never defines it.
The definition lives in an endnote:
Salokya (residing with), samipya (proximity), sarupya (same form), sarshti, sayujya (identification).1
Five kinds. And they are not five words for one thing — they are five degrees, and the volume awards them separately to different practices.
Which means every promise of "emancipation" in this book has been underspecified, and once you have the list you can go back and see which one was actually on offer each time.
Take them in order, because the order is a progression from geography to identity.
Salokya — sa-loka, same world. You reside in Shivaloka. Proximity of place.
Samipya — nearness. Closer than merely being in the same world.
Sarupya — same form. You look like him.
Sarshti — Debroy gives no gloss for this one, alone among the five. The standard sense is equality of power or splendour — sharing his majesty.
Sayujya — identification. Union. The end.
So the sequence is: same place → close → same shape → same power → same thing.
That is a very physical ladder for an abstract goal. Four of the five are described in terms a person could point at: where you are, how near, what you look like, what you can do. Only the last is identity.
And it means the tradition's language of liberation is, at four levels out of five, spatial and relational rather than ontological. You are not dissolving. You are getting closer and coming to resemble.
Chapter 25 does the thing that makes this list load-bearing: it maps the grades onto the release of specific bonds.
Therefore, the person resides in Shivaloka and is spoken of as having obtained salokya. When the tanmatras come under control, samipya with Samba is achieved. In weapons and in action, he achieves sayujya with Shiva. When he obtains his great favours, the intellect also comes under control. The intellect represents the working of Prakriti. Hence, this is known as sarshti. Thereafter, through his great favours, Prakriti itself comes under control... Those who are accomplished in the Vedas and the agama texts speak of this as sayujya.2
So:
| Bond released | Grade |
|---|---|
| The body (karma's product) | salokya |
| The five tanmatras | samipya |
| — | sayujya (stated here, out of order) |
| The intellect | sarshti |
| Prakriti itself | sayujya |
Which does not quite work, and the failure is worth stating rather than tidying.
Sayujya appears twice — once at "in weapons and in action," partway up, and once at the end where it belongs. Sarupya does not appear in the chapter's mapping at all.
Either the passage is corrupt, or sayujya is being used in two senses, or the mapping was assembled from two sources. The chapter does not distinguish and Debroy does not annotate the repetition.
Preserve it. The mapping is real and it is imperfect, and smoothing it would invent a tidiness the source does not have.
Step back and ask what it means for moksha to have five grades.
In the strong non-dual readings this tradition also contains, liberation is binary and absolute — there is nothing it could be more or less of, because it is the recognition of what was always the case. A graded moksha is close to a contradiction on that view.
Chapter 25 has grades, and it has them because of the structure it has already built: eight bonds, released one at a time, by successive favours.3
Once release is sequential, the grades follow automatically. If you can be free of the tanmatras and not yet free of the intellect, then there is a state that is more free than bound and less free than free — and it needs a name.
So the five grades are not a soft-headed compromise. They are the necessary consequence of an eight-bond model, and any tradition that decomposes bondage into parts will end up with graded liberation whether it wanted to or not.
Which explains why the Śaiva Siddhānta material in this vault runs the same shape from the other end — three orders of bound souls, graded by which malas still hold.4 Same commitment: bondage has parts, therefore freedom has degrees.
And it explains why the list lives in an endnote. A graded absolute is theologically awkward, and a text that uses the terms constantly while never defining them has found a way to keep both the vocabulary and the ambiguity.
You have been told, over the years, that most of what you do leads to emancipation.
The clay lingam does. The naivedya does. The ash does, and so do the names, and so does hearing the text at all.5
And you have always heard that as one thing, the way you hear "you will get to the city" as one thing.
Then somebody reads out the list and it turns out there are five cities.
So you go back through it, and the sorting is not what you expected.
The prasada — one mouthful, correctly taken — gives sayujya.6 The top of the scale, for eating rice.
The Parthiva lingam, worshipped without desire, gives sayujya.7
The morning japa, the daily discipline, the twenty years of it: the chapter says a Shivayogi "gradually achieves salokya and pure emancipation."8 The bottom grade, and gradually.
That is not the ordering you would have guessed. The cheap, quick, one-off acts are priced at the top. The long disciplines are priced at the bottom and given an adverb.
You do not know what to do with that and you keep doing both.
The distribution across the book is worth setting out, because it is the payoff of having the list at all.
Sayujya is awarded for: partaking of Shiva's naivedya; worshipping the Parthiva lingam without desire; offering the naivedya of a lifetime; constant lingam worship; wearing Tripundraka through to the end of the ascent; releasing the last bond.9
Salokya is awarded for: being a Shivayogi who uses the sixteen upacharas — "gradually achieves salokya and pure emancipation."10
Samipya, sarupya, sarshti appear almost nowhere in the body. Chapter 24 mentions samipya and sarupya once each in the panchakshara count sequence — samipya with Brahma at one threshold, sarupya at the next.11
So the volume overwhelmingly promises sayujya, the top grade, and almost never mentions the middle three.
Which tells you something about how the list is being used. It is not a working scale that practices get sorted into. It is a complete taxonomy held in reserve, from which the text takes the top item nearly every time it wants to promise something.
The grades exist. The book mostly ignores them and quotes the maximum.
That absence is the pattern this build has been tracking since Cluster A, and this is its clearest instance.
The body of the volume uses sayujya, salokya and samipya as though everyone knows them. It never says what they are.
Debroy's endnote 136 supplies the definitions, and without it a reader has five Sanskrit terms and no content.
Same as the three kinds of bhakti, the three hardships, the eight bonds, the shudra sub-classes, the transgressive-women taxonomy, jati versus varna.12 Every taxonomy this volume invokes as common knowledge is defined only in the annotation layer.
Which is not a defect of the Purana. It is evidence about its audience: these were terms you did not need explained, in a culture where the categories were ambient. The text is written for people who already have the vocabulary and want the application.
And it means a modern reader without the endnotes is not reading a difficult text. They are reading a text with its content removed — promises whose denominations they cannot see.
The closing structural point.
If liberation has five degrees, three things follow that the volume never states.
Liberation is not an event. It has stages, and a person can be in one of them. There is no moment at which it happens.
It is therefore also, in principle, losable — or at least the lower grades are, since they are described as bonds "coming under control" through favours, and the volume elsewhere makes gained states revocable (the second upanayana, the forced re-entry to the forest).13 The chapter does not raise this and it is the obvious question.
And the top grade is not more of the lower ones. Sayujya is identification; salokya is a postcode. You do not reach identity by accumulating proximity, any more than you reach Kalatita by climbing worlds.14
That last is the volume's recurring discovery and it surfaces here for the fourth time in this cluster: the top of every ladder in this book is made of a different material than the rungs. Ripe karma resists any quantity of merit. The single-syllable Pranava is terminal rather than introductory. Kalachakra is a process above a hundred and twelve places. And sayujya is identity above four degrees of nearness.
Four instances, four chapters, no cross-reference.
Strongest support is chapter 25's explicit mapping of grades onto specific bond-releases, which shows the five functioning as a real scale rather than as synonyms.
Tension one. The mapping is imperfect: sayujya appears twice at different stages, and sarupya is absent from it entirely.
Tension two. A graded moksha sits uneasily with the non-dual readings the same tradition holds, where liberation is recognition and admits no degrees.
Tension three. The volume promises sayujya for cheap one-off acts (a mouthful of prasada) and salokya, gradually for the long disciplines. The pricing runs opposite to the effort.
Tension four. Four of the five grades are spatial or relational — where you are, how near, what you look like, what you can do — which makes the tradition's liberation language overwhelmingly proximity language, not dissolution language.
Open question. Is sarshti genuinely a power-sharing grade? Debroy defines four of the five and leaves this one bare, which is unusual for him and may indicate the term was not settled in his sources either. The chapter attaches it to the release of intellect, which fits a splendour/capacity reading better than an equality-of-status one, but nothing in Volume 1 settles it.
The vault holds the dualist Śaiva position on this and it agrees with the Purana more than the non-dual pages do.
Śivatva: Identity of Essence Despite Difference holds the Siddhānta view that the soul attains Śiva-nature but does not merge — it becomes a teaching siddha, not a dissolved one.
That is close to sarupya and sarshti and stops short of sayujya. So the dualist school has effectively selected a grade from this list and made it the terminus.
Three Orders of Bound Souls grades from the bondage side — vijñānakala, pralayākala, sakala. Read with this page, the tradition has two graded scales, one of bondage and one of freedom, and they are the same scale read from opposite ends.
Against both, Practice Can't Reveal Śiva (Anupāya) holds the non-dual position that no practice reveals Śiva at all. On that view a ladder of grades is a description of a journey nobody is taking.
The vault should hold all three. The Purana is squarely in the graded camp and does not know the argument is happening.
Development vs Spiritual Transcendence (psychology). The developmental material distinguishes stage claims — where each level includes and reorganizes the last — from state claims, where an experience is available at any point and does not accumulate. And it holds that conflating the two is the commonest error in transpersonal accounts.
The five grades are stage claims: each is defined by which bond has released, and they are sequential. But the volume awards its top grade for one-off states — a mouthful of prasada, a night of Śivarātri.
What the pair produces: a tradition holding both a stage ladder and state-triggered maxima will always look incoherent, and the incoherence is not sloppiness — it is two different things being called by one word. Salokya-through-sayujya is a developmental sequence. Sayujya-for-eating-prasada is a state claim. The list in endnote 136 describes the first; most of the volume's promises are the second. Which is why the grades are defined in an annotation and ignored in the body: the book is mostly selling states and only occasionally describing stages, and it has one vocabulary for both.
Brand Equity Beyond Purchase (business). The operator material observes that where a category has graded tiers but a single popular name, the name migrates upward — everyone claims the top tier, the middle tiers go unused in marketing, and the grading survives only in technical documentation.
That is exactly the distribution here. Sayujya is promised everywhere; samipya, sarupya and sarshti appear a handful of times; the full scale survives only in the annotation layer.
The insight neither gives alone: a graded scale whose top term is freely available as a promise will collapse into that term, and the collapse is a language effect rather than a doctrinal decision — nobody voted to stop using sarshti. It simply lost, the way every middle tier loses when the top one costs nothing to say. Which predicts something checkable across religious literature: middle grades survive in commentaries and disappear from sermons, and their presence in a text is a signal about genre rather than about sophistication.
Sharpest implication. The volume promises "emancipation" on almost every page and never defines it — the five grades survive only in Debroy's endnote 136: salokya (same world), samipya (nearness), sarupya (same form), sarshti, sayujya (identification). Four of the five are spatial and relational, which makes this tradition's liberation language proximity language rather than dissolution language. Chapter 25 maps them onto the release of specific bonds, imperfectly — sayujya appears twice and sarupya not at all — and the grades exist at all only because bondage was decomposed into eight parts that release one at a time. But the book overwhelmingly quotes the maximum: sayujya for a mouthful of prasāda, and salokya, gradually for a lifetime of discipline. The pricing runs opposite to the effort, and the middle three grades are effectively unused.
Generative questions.