Source list — shopping list driven by open questions

Needed to close open questions #3 (dhyāna) and #7 (upgrade kavaca pin). Colophon lead from compiled Dakṣiṇa Kālī kavaca.

Needed to close open question #7 (compiled kavaca attributes Rudrayāmala / Śyāmārahasya).

Rights NOT established. GRETIL marks its text for reference and says source-edition terms apply; the separate IA scan lacks edition metadata. Rights-check the 1970 source before any redistribution.

URL is the publisher's own product page, not a text source. An IA upload of this edition resolves, but Nag Publishers still sells it, so the uploader CC0 label is not accepted. Treat as copyrighted unless permission is obtained.

Carries an IA Public Domain Mark. Scanned book with PDF, EPUB and derived text.

Good for cross-checking. Vol 2 available under similar archive.org path.

Digital searchable Sanskrit.

Start here — collected working links; use as source-finding hub.

Wikisource transcription is CC BY-SA, not public domain as a digital transcription. GRETIL version is marked unproofread / reference-only.

Public-domain-era translation; IA metadata says 'Copyright Permitted'. Navigable HTML transcription also at sacred-texts.com/tantra/maha. Ch. 1 vv. 50–52 is the substitute support for paśu/vīra/divya after the Kulārṇava locator failed verification.

Retained as a pointer only. A lumped entry cannot carry rights, because rights attach to an edition. The four texts are now recorded separately.

METADATA CORRECTION: IA describes item 69227 as Guhyakālī-khaṇḍa, but the scanned title page and publisher's note both say Kāmakalākhaṇḍa. Enter provisionally as Kāmakalākhaṇḍa unless folio-level examination says otherwise. Do not ingest the IA title field. The complete 2010 six-volume Chaukhamba set is copyrighted.

1922 text is public-domain-era but this transcription omits the Devanagari. The 1953 revised edition scan contains Devanagari plus commentary and its rights are NOT established — see karpuradi-stotra-1953.

Marked 'In Public Domain' by DLI/IA. Metadata language is Bengali — distinguish Sanskrit source passages from Bengali translation before extraction. Kṛṣṇānanda Āgamavāgīśa, active 16th century. Note for the Multiplicity chapter: the Kālī word-pictures here are quoted from earlier works, so this records transmission, not invention.

Year and press identifiable. No explicit public-domain declaration on the item page; verify for the intended jurisdiction.

Explicit Public Domain Mark. The 1923 Thompson & Spencer English volume is a selection only, not a complete translation — recorded separately as bengali-religious-lyrics-1923.

Imprint confirmed as University of California Press, not Chicago — the audit had flagged this as open. ISBN CORRECTION: the 1997 print is 9780520204997; earlier drafts used 9780520917729, which UC Press's own site shows is the Jul 2023 ePub+PDF. PAGE NUMBERS REMAIN NOT CONFIRMED: no paginated authorised copy has been opened, and the reachable full-text copies appear unauthorised with unreliable OCR pagination. Because this file requires a non-null locator for an [attested] claim, the Mahāvidyā dating and the Kāmākhyā spatial-distribution claims stay [open]. They are one confirmed page number away, not one source away. Load-bearing for two findings: the Mahāvidyās are known as a group only after the 10th century, with firmly attested origin accounts in Bengali upapurāṇas probably post-14th century; and three Mahāvidyās are stone yonis in the Kāmākhyā sanctum with the other seven in separate shrines across the hill — a spatial distribution, not an anatomical mapping.

Cite-only. Anchor for spanda metaphysics.

Was a stub with no named work and no URL. Oxford Academic's landing page is behind a Cloudflare bot check and was not reachable; the DOI's registered Crossref metadata, the OUP-licensed ebook preview front matter and the Open Library edition record were confirmed instead. The book's own stated scope — mid-eighteenth century onward — is what supports the 18th-century mass-devotion dating. The 'p. 46' locator for Kṛṣṇānanda 'at the latest, in the seventeenth century' is NOT CONFIRMED and must not be used; upenn-tantrasara-ms765 (colophon 1663) is the better witness for the 17th-century point.

Cite-only. Anchor for 64-Yogini roster-instability finding.

Cite-only. Some may be on archive.org — verify.

Free full text reported at journals.lww.com / PubMed, but no URL has been opened and confirmed. Downgraded from 'verified' to satisfy the file's own rule that verified requires a non-null url.

Explicitly marked 'In Public Domain'. This is the complete Mārkaṇḍeya Purāṇa, not a standalone Devī Māhātmya.

Marked unproofread / reference-only. Not corpus-grade.

Strongest readily downloadable candidate among the harder core-Tantra titles. Kāmākhyā-associated. Check scan completeness and title pages locally before use.

Lacks publisher/date/editor metadata for corpus-grade identification. Verify title, incipit, colophon and relationship to other works called Kālī Tantra before ingestion. A reliably catalogued printed edition is NOT VERIFIED and is an acquisition priority. Relevant to the Multiplicity chapter: the Tantrasāra labels its Dakṣiṇā Kālikā meditation verse 'Kālī Tantra, cited in the Tantrasāra'.

No printed or critical Sanskrit edition verified; needs a library catalogue search.

The Chaukhamba 2014 edition (Chaturvedi, Sanskrit with Hindi, 112pp) is a copyrighted commercial edition; no lawful free scan verified.

Ritual/mantra digest. Volumes recorded separately; repository rights fields are inconsistent between the three records.

Downloaded and inspected: contains Devanagari plus English translation and commentary, which the 1922 Sacred Texts transcription omits. Do not treat as unrestricted until revised-edition rights are checked.

A selection, not a complete translation of either padāvalī. Ramprasad supplies the greater part of it. A complete public-domain English padāvalī was NOT verified.

Jaideva Singh's modern translation is copyrighted; no lawful free full text verified. Use the PD edition for corpus work.

Supplies a usable old English translation, avoiding reliance on an unauthorised copy of the modern Singh edition.

Marked 'In Public Domain'. Sanskrit with Kṣemarāja's commentary — NOT a public-domain English translation.

Read directly: tantra = 'a loom … the warp' citing RV x.71.9 and AV x.7.42; tantu = 'a thread, cord, string, line, wire, warp (of a web), filament, fibre'. Supports the tan/tantu etymology at [attested] strength. Does NOT support the gloss that the thread refers to the channels of the body — that reading is separated out as [inference]. Currently the only Group A source carrying a page-level locator.

Documents Marshall's 1931 Mother Goddess reading and the sample and representational biases that produced it. Supports the [attested] half of the Indus claim (the figurines and their dating) and the [contested] status of the goddess identification.

Argues the female clay figurines were likely home-made representations kept to ward off harm rather than idols of a Mother Goddess, and notes the lack of securely dated archaeological contexts.

NOT VERIFIED — no link confirmed. Library acquisition. Title states the position directly.

NOT VERIFIED — no link confirmed. The foundational critique of Marshall's figurine classification.

NOT VERIFIED — no link confirmed. Recorded as the ORIGIN of the Mother Goddess reading and therefore as a superseded position, not as support. Cite only to attribute the claim being corrected.

NOT VERIFIED — no link confirmed. The classic caution against reading deep-time female figurines as goddesses; bears on the 'tens of thousands of years' sub-claim.

Third-party host — cite the Oxford Research Encyclopedia as the authoritative venue and check rights before relying on this copy. Supports the fifth-century emergence of distinctly tantric Śaiva traditions and quotes the 423 CE Gaṅgdhār inscription.

Third-party host; cite the original publication venue and check rights. Load-bearing for the Foundation F4 downgrade: documents Paippalāda Atharvavedin ritual manuals following Tantric rather than Vedic liturgical models, i.e. influence running INTO the Atharvavedic community. Directly concerns the Kālīkula, so relevant well beyond that one claim.

Records the oldest surviving Śaiva Tantra as a single palm-leaf manuscript in Nepalese 'Licchavi' script assigned to c. 850–900 AD. Supports the image-ancient / structure-late timeline.

Argues the 'magic' framing of the Atharvaveda persists as a colonial-era analytical category rather than a finding about the text. Underpins the F4 downgrade: it removes the premise that makes the Atharvaveda look like Tantra's natural ancestor.

NOT VERIFIED — no link confirmed. Encountered as a citation inside a retrieved INDOLOGY list discussion (list.indology.info/pipermail/indology/1997-April/007950.html). Consult the Gonda volume directly before citing. Bears on F5: the weaving metaphor in the ritual literature points to the framework of a rite, not to bodily channels.

Brill's own landing page could not be opened (HTTP 405, human verification); the DOI's registered Crossref metadata and the Open Library edition record were confirmed instead. Page numbers were read off printed running heads in an unauthorised IA scan, and p. 47 was cross-confirmed against the book's own printed index. Do not record or redistribute that scan. Bühnemann's introduction warns against the generalisation the chapter draws from her, so this supports the NAMED STRUCTURES only — the generalising sentence stays [inference].

This — not the yantra volume — is the work the kṣamā-prārthanā claim rests on. Distinct from buhnemann-2003-mandalas-yantras; do not merge. Resolves the audit's open question: the earlier scan was paginated independently of the 1988 print, and these page numbers come from the publisher's own table of contents. Scope caveat: the descriptive base is smārta domestic worship, largely Maharashtrian and Viṣṇu-directed, not a pan-Hindu survey, and the Tantric kṣamāpaṇa parallel is a footnote cross-reference. Narrow the word 'standard' accordingly.

DOES NOT STATE A 9TH–13TH CENTURY BRACKET. Read in full. Kiss gives a nine-phase semantic history of the term yantra: the geometric-diagram sense appears in the Netratantra (dated AD 700/800–850 after Sanderson 2004: 293), is still absent from Kṣemarāja's early-11th-century reading of yantracakra as 'mantras in a particular pattern', and only later dominates, with 12th–13th-century texts assuming it. He closes by calling the question open. Indexing Kiss does not license the round numbers. Separately: Kiss dates the Niśvāsatattvasaṃhitā TEXT to the 5th–6th century, which is a different question from the c. 850–900 MANUSCRIPT date in nisvasa-mow-nomination — the two do not conflict.

ISBN CORRECTION: the 1984 first edition is 0836408675. Earlier audit drafts paired 1984 with 8120805577, which belongs to the 1988 second edition. The Motilal product page displays a current reprint (ISBN 9789394201576) with no year, so it verifies the work but not the 1984 edition; only the Open Library record does that. The WorldCat record cited in the Tool Chapters audit could not be opened and is not recorded here. A controlled-lending copy of the 1985 printing exists on IA — borrow-only, not a free text.

Full thesis freely downloadable from the ERC 609823 archive. Dates the Devī Māhātmya to the second half of the eighth century, with the early ninth 'also a possibility' (n. 42, p. 23), against the fifth-or-sixth-century dating she attributes to Coburn 1984: 1 (n. 16, p. 8). ALSO CONTRADICTS THE 608 CE DADHIMATĪ MĀTĀ TERMINUS used in the Tool Chapters audit: n. 42 reports Mirashi's re-reading of the year as 189 rather than 289, yielding AD 812–813 or 795–796, and doubts the verse derives from the Devī Māhātmya at all. Her conclusion is that the inscription cannot provide a definite terminus ante quem. Readable but not cleared for redistribution (Taverne notice). Groningen Pure dates the same work 2005; the literature cites 2004.

A distinct witness from the 1932 `tantrasara` print — separate entry, do not merge. The Kislak Center applies a Creative Commons Public Domain Mark to the images and content; this is an INSTITUTIONAL mark from the holding library, categorically different from the uploader-applied CC0 the standard rejects. Metadata is CC-BY-4.0, so attribution is owed if the metadata is reused. The colophon names Kṛṣṇānanda Āgamavāgīśa Bhaṭṭācārya and dates the copy to 1663. THIS IS A MANUSCRIPT DATE, NOT A COMPOSITION DATE — it evidences 17th-century circulation, which is exactly what the Tool Chapters claim needs, and nothing more. Cataloguer notes f. 39 probably missing. Full per-folio images and TEI XML available. Gives the Tantrasāra material its first non-null locator.

Edition identified from the scanned TITLE PAGE, not from IA metadata, which names no translator, publisher or date. The 1999 date is new information; no audit recorded it. RIGHTS: the IA item carries no licence field of any kind — not even an uploader CC0 — while the underlying edition is a 1999 commercial translation from a working publisher. This is a step below the uploader-cc0-only case. Do not ingest or redistribute; it is usable to check the claim and nothing else. Devotional-press translation, not a critical edition — the Avalon / Tantrik Texts edition (Vidyāratna, 1965 rev.) is the scholarly witness and is an acquisition item. Page_number_confidence is 12 and the OCR verse numbers are garbled: cite the section heading and printed page, NEVER a verse number. The Kulārṇava's own date is unsettled, so this text cannot carry a claim about what 'the tradition' held across time.

Read in full from the APA's own PDF; page numbers recovered from the document's printed running heads. Supports three things and refutes two. SUPPORTS: prolonged exposure is one of three co-equal first-line psychological interventions, not 'the' evidence-based treatment (p. 7); between-session extinction of distress repeatedly predicts better outcomes (p. 41); dropout is not always a consequence of symptoms worsening (p. 38). REFUTES BY ABSENCE: the words 'pacing', 'titration' and 'graded' do not occur anywhere in the document — an earlier draft of the chapter attributed a clinician-titration rule and a list of slow-down markers to this guideline, and no such rule is in it. ALSO CUTS AGAINST THE CHAPTER'S OWN MOVE: the guideline warns it is 'easy to assume' prolonged exposure works through extinction, and that mediators of outcome may not be obvious from a treatment's components (pp. 40–41). Within-session habituation is a much weaker predictor than between-session; reductions in negative post-traumatic cognitions also predict relief. The mechanism question is open, so the chapter may not name a mechanism and call it attested.