The Pāśupatas were one of the oldest organized Shaiva ascetic orders — cremation-ground dwellers, deeply atimārga (the higher path of asceticism, distinct from later householder-oriented Tantra). The original Pāśupata position held a stark Buddhist-style account of liberation: liberation = duḥkha-nivṛtti, the cessation of suffering, with nothing positive added. Liberation was an absence of pain, not the attainment of anything.1
Lākulīśa, a historical reformer of the Pāśupata movement (probably 1st-2nd century CE — Selvalingam's source doesn't pin the date), pushed back. His Pāśupata-Sūtra (later commented on by Kauṇḍinya) argued that liberation must mean the attainment of something, not just removal of something. Specifically: the attainment of omniscience and omnipotence — the qualities that mark the liberated being as having actually arrived somewhere, not just escaped from where they were.2
This is a fundamental break inside Shaivism — and it sets up the architecture that Sadyojyoti will then use the Paramokṣa-nirāsa-kārikā to refute. Lākulīśa's Pāśupata is the first school to explicitly say: liberation has positive content, not just negative content. The negative-only Pāśupatas could not say what the liberated being had; Lākulīśa could.
A common misreading: Lākulīśa is grouped with the Pāśupata dualists. Selvalingam corrects this explicitly:
Many people think Lākulīśa is a dualist but he's not actually. That tradition is more of a Viśiṣṭādvaita kind of tradition. So we can study Lākulīśa as the aesthetic Pāśupata reform.3
The technical move that puts Lākulīśa in the qualified-non-dualism column rather than the dualism column: Lākulīśa starts with five tattvas — Kārya, Kāla, Yoga, Vidhi, and Duḥkha — and roots all five in Pati (the Lord). The five are real categories of reality, distinct from each other, but all of them inhere in or derive from the singular Lord. They are not eternally co-existent with the Lord as separate categories (which would be dualism); they are modes of the Lord (which makes it qualified non-dualism, viśiṣṭādvaita).4
When Sadyojyoti later builds Shaiva Siddhānta, he takes Lākulīśa's five tattvas and reduces them to three: Pati (kāraṇa, the cause), Pāśa (the world, the bond — taking kārya as effect and renaming it), and Paśu (the souls — distinguished from pāśa in Sadyojyoti's reduction, though Lākulīśa had collapsed them). Selvalingam: Sadyojyoti takes Pāśa, Paśu, and Pati and gets those from the five. The reduction is also a re-categorization: Sadyojyoti's three are eternally distinct categories (cleaner dualism) where Lākulīśa's five were modes of one reality with the Lord as the rooting principle (qualified non-dualism).5
The architectural lesson: the difference between Shaiva schools is not always doctrinal substance — sometimes it is categorial structure. Same fundamental ingredients (Lord, world, souls, time, action) get parsed into different category counts (3, 5, 36, etc.) and the parsing itself changes the metaphysical commitments.
Lākulīśa's Pāśupatas — like the Pāśupatas before them — were aghorī-style ascetics. Predominantly atimārgic cremation-ground dwelling movements, Selvalingam summarizes. The texts of this tradition include the Pāśupata-Sūtra itself plus texts Selvalingam mentions in passing: Aghora, Yāma, Pratāra, and other cremation-ground texts.6
The cremation-ground location is not incidental. It is an operative element of the practice. The atimārga (higher path) of Pāśupata asceticism aimed at transgression of social norms as a method of consciousness-transformation: deliberately doing what would horrify ordinary social-religious sensibility (eating from skull-bowls, wearing ash, dwelling among corpses) in order to break the conditioned ego-structure built by social norms. The cremation-ground was both the physical staging-ground for these practices and the symbol of the ego-death the practices were aimed at.
The Lākulīśa reform did not abandon the cremation-ground orientation. It refined the philosophical account of what the cremation-ground practices were doing — providing the theoretical structure (qualified non-dualism with the Lord as foundation; positive liberation as omniscience-and-omnipotence) that the older Pāśupatas had lacked. The lifestyle remained ascetic and transgressive; the theology became more sophisticated.
This is what Selvalingam means by Lākulīśa as the aesthetic Pāśupata reform — the reform that brought aesthetic and theoretical refinement to the older tradition's stark practices, without changing the practices themselves.
The specific theory of how liberation works that Lākulīśa proposed is what Sadyojyoti targets in the Paramokṣa-nirāsa-kārikā under the name saṃsthāna-śānti. The Lākulīśan account: liberation is like fragrance — by getting close enough to Shiva through ascetic practice and devotion, Shiva's qualities of omniscience and omnipotence transfer to the practitioner, the way fragrance transfers from a fragrant object to anything near it.7
This is a transmission model of liberation, distinct from both:
Sadyojyoti's three-fold refutation hits the saṃsthāna-śānti model in the second slot: contagion can be lost as easily as it can be gained; the metaphor is incoherent; the heat-transmission analogy implies Shiva would lose what he gives (which is impossible). See Sadyojyoti's Three-Fold Critique for the full rebuttal.
The Lākulīśan response (not directly recorded in this source but implicit in the surviving Pāśupata literature): the fragrance metaphor is a teaching device not a literal mechanism; the actual transmission is by Shiva's free choice, not by mechanical contagion; and the asymmetry between Shiva (who never loses) and the practitioner (who gains) is a feature of Shiva's nature, not a refutation of the model.
Whether the response holds against Sadyojyoti's logic is a live question. Selvalingam doesn't push it.
Lākulīśa's importance for the rest of the vault:
Architectural: Lākulīśa is the bridge from the older Pāśupata negative-liberation account to the later Shaiva positive-liberation tradition. Without Lākulīśa's reform, Sadyojyoti's positive doctrine (innate divinity, recognition not attainment) would have nothing to push off from.
Doxographical: Lākulīśa's qualified-non-dualism is the second rung on Abhinavagupta's Tantrāloka 29.51 hierarchy, above Vaiṣṇava Pāñcarātra and below Shaiva Siddhānta. The graded sequence Veda → Pāñcarātra → Pāśupata-Lākulīśa → Siddhānta encodes a movement from less-Shaiva to more-Shaiva and from more-ascetic to more-householder. (See Doxographical Hierarchies as Institutional Strategy.)
Practical: Lākulīśan cremation-ground ascetic transgression is a precursor of the later transgressive moves of Bhairava-āgama and Krama practice — the structural pattern of deliberate norm-violation as method survives across the Shaiva traditions even as the institutional setting moves from cremation-ground to householder ritual.
Selvalingam's relationship to the Lākulīśan tradition is honestly mixed. He places Lākulīśa structurally above Vaiṣṇava Pāñcarātra in his survey but well below Shaiva Siddhānta and Trika; he uses Sadyojyoti's refutation of the saṃsthāna-śānti model as a key point in his account of how Shaiva Siddhānta got built; he gives the Lākulīśan tradition relatively little space in the lecture compared to the schools he cares more about.
But the convergence at the level of practice attitude is real. The cremation-ground orientation — deliberate transgression of social norms as a method of ego-death — is the operative substrate for the entire later Tantric corpus that Selvalingam teaches. The Krama mantras Selvalingam describes as universe-destroying are operating on the same logical principle as Lākulīśa's cremation-ground transgressions: deliberate exposure to the unbearable as a method of dismantling the ego that cannot bear it.
The Lākulīśan tradition is older than the Trika and the Krama; the structural move it pioneered (transgression-as-method) is still operative in the schools that long-since superseded it doctrinally.
The qualified-non-dualism reading depends on a specific parsing of Lākulīśa's five tattvas. Other historians read Lākulīśa as a dualist (the five tattvas as five eternally distinct categories), citing different passages. The "more of a Viśiṣṭādvaita kind of tradition" reading is Selvalingam's lineage's reading; the dualist alternative is also defensible.
The fragrance theory may not actually be Lākulīśa's. Sadyojyoti's Paramokṣa-nirāsa-kārikā is Selvalingam's source for attributing the saṃsthāna-śānti view to the Lākulīśan tradition. But Sadyojyoti is a Shaiva-Siddhāntin polemicist; his characterization of the rival school may be uncharitable. Modern Pāśupata scholarship (Hara, Bisschop) reads Lākulīśa's actual surviving texts more cautiously.
Lākulīśa's reform raises a question Selvalingam doesn't fully answer: what is the relationship between cremation-ground transgressive practice and the doxographical position the school comes to occupy? The Lākulīśan tradition's transgressive practices are more extreme than the Shaiva Siddhānta householder practices that succeed it on the doxographical ladder. But Shaiva Siddhānta is higher on the ladder. The pattern repeats: at every level, the more-householder school is higher than the more-ascetic school it succeeds. Why?
Selvalingam's implied answer: householder practice is harder than ascetic practice — Tantra glorifies the householder; asceticism is kind of embarrassing in a sense — because integrating the practice into social-economic-relational complexity is more demanding than removing oneself from it. This is the inversion the Tantric tradition makes against the older Shramanic assumption that ascetic withdrawal is the high path. But the inversion is contestable, and the older Pāśupatas would not have accepted it.