The Ledger · sources

What is checked, what is not, and what was found.

A citation names a source. It does not say whether the author has a stake outside the evidence, whether several papers cited together came from one place, or whether the thing cited is settled or still a proposal. Those are four questions, and they are asked here separately.

The four questions.

asked of every audited source

Question	Meaning
Credentials	Trained in the field; the work passed review in a real venue.
Interest	A stake outside the evidence — funding from the subject, an institutional role in it, or a prior claim of one's own that a criticism happens to defend. Specialisation is not an interest. Long study of a subject is the reason to cite someone.
Independence	Whether papers cited alongside each other share authors, samples or a laboratory — and so count once rather than several times.
Claim status	Whether the specific thing cited is established or a live hypothesis. A property of the claim, not the author.

Sources carrying weight.

cited across several pages, or load-bearing on a central claim

Source	Interest	Independence	Claim	Finding
Asko Parpola Indus script · 9 pages	None found	One stream	Hypothesis	PhD 1968 Helsinki, Sanskrit and comparative Indo-European linguistics; Professor Emeritus of Indology; editor of the Corpus of Indus Seals and Inscriptions. His Dravidian reading of the script is a research conclusion. The script is undeciphered and the reading is unproven; every page invoking it says so.
Reich laboratory Narasimhan · Lazaridis · Chintalapati · Patterson	None found	Shared	Established	Harvard Medical School; methods and sample handling are the discipline's standard. These papers share authors, samples and a laboratory: cited together they are one line of evidence, not four. A 2019 criticism concerns monopolisation and under-representation of non-Western collaborators — a structural point about the field.
Iravatham Mahadevan Tamil-Brahmi; Indus epigraphy	None found	Same stream	Hypothesis	Not an academic by career — an IAS officer who worked on the script under a Jawaharlal Nehru Trust fellowship. His decipherment of Tamil-Brahmi is accepted; his Indus work advances the same Dravidian reading as Parpola, so the two are not independent support for each other. Ratnagar's assessment: he "put the whole thing squarely in a good framework," and she criticises Parpola for crediting him only in half a sentence.
Johannes Bronkhorst Greater Magadha	None found	Independent	Contested	Two doctorates, the Leiden one summa; trained in India with traditional Sanskrit scholars; ~230 papers. The chronology his thesis depends on is disputed among specialists. Full audit
Gwen Robbins Schug Harappan skeletal trauma	None found	Independent	Established	PhD biological anthropology; UNC Greensboro; editor, A Companion to South Asia in the Past. Subject specialist on the exact material. doi
Gregory L. Possehl Indus archaeology	None found	Independent	Established	University of Pennsylvania; eight books on the Indus. Sceptical of the Dravidian decipherment, and cited here partly for that.
Willerslev group Damgaard et al. 2018	None found	Independent	Established	Copenhagen. A separate laboratory whose findings on Anatolia contrast with Reich-lab framing. doi
Sudha Pai Dalit politics, UP	None found	Independent	Established	JNU; long-standing specialist. Regional limits stated in the paper. doi
Shah & Lerche caste and labour	None found	Independent	Established	LSE and SOAS; long-term ethnographic fieldwork. Their figures complicate the argument they are cited within. doi

Where an interest was found.

disclosed at every point of use

Source	The interest
Bhushan Patwardhan Ayurveda literature	Material. National Research Professor–AYUSH, WHO consultant, founding editor-in-chief of an Elsevier Ayurveda journal — a salaried and institutional stake in the system being assessed.
Alexander Wynne critic of Bronkhorst	Scoped to the criticism. Oxford DPhil; credentials sound. But his own book argues the Buddha was trained in a school whose ideology came from the early Upaniṣads, so his objection to Bronkhorst's dating of those texts defends his own foundation. A rival hypothesis rather than a correction.
Nicole Boivin ashmounds; Southern Neolithic dating	Disclosed, and it is serious. Cambridge PhD 2001; former director, Max Planck Institute for the Science of Human History. Removed from that directorship in October 2021 after an investigation found evidence of scientific misconduct and bullying — the published summary cites claiming credit for others' work and bullying of staff and junior researchers. A Berlin court reinstated her in December 2021; she was removed again in April 2022 and the case has been under appeal. What the finding concerns and does not: it concerns credit and conduct, not fabricated data, and it postdates the 2004 and 2007 ashmound work by fifteen years. The radiocarbon programme cited on this platform is co-authored with Dorian Fuller and Ravi Korisettar and its dates are independently checkable. Cited here for the dating, with this on the record.
Shereen Ratnagar Indus trade and urbanism	Disclosed, and minor. Deccan College Pune; Mesopotamian archaeology at the UCL Institute of Archaeology; Professor of Archaeology at JNU to 2000; fieldwork in Iraq. She gave expert evidence in the Ayodhya proceedings, arguing the archaeological case was inconclusive on disturbed stratigraphy — an entanglement in a politically charged matter, which is worth a reader knowing. Accusations of ideological motive were made by opponents in that dispute and are not treated here as findings. Her Indus work is unrelated to it, and she is notably willing to deflate claims — including rereading the Mohenjo-daro "dancing girl" as not depicting dance.
Michel Danino Harappan trauma	Fails on credentials and on the evidence. No advanced degree in history or archaeology; trained as an electrical engineer; a position adopted at twenty-one and held for fifty years. His reading of the injuries as "mostly clubbing, inconsistent with warfare" cannot account for a sword cut between the eye sockets, nor for a trauma rate of roughly 0–4% in the urban cemetery against ~50% in an excluded burial group. Retained as documented material, never as evidence. Full audit

Checked, single-paper sources.

audited, no interest found, cited once or twice

André Béteille (Delhi School of Economics; Padma Bhushan; Fellow of the British Academy; taught at Oxford, Cambridge, Chicago and the LSE) · Kanagarathinam & Lourdusamy (IIT Madras) · Zhang et al. (Max Planck / Jilin / CAS / SNU) · Badri Narayan Tiwari (G. B. Pant Institute) · Janaki Abraham (Sociological Bulletin) · V. K. Choudhary (CASTE, Brandeis) · Hugh B. Urban (Ohio State) · Vermeir & Margócsy (BJHS) · Amit Basole (Development and Change) · Canuto et al. (Tulane / PACUNAM) · Douglas et al. (Antiquity) · Anton Killin (Music & Science) · Ilona Kędzia-Warych (Cracow Indological Studies) · B. V. Subbarayappa (IUHPS)

One qualification within that list. Subbarayappa was eminent — first non-Westerner to lead the History of Science division of the IUHPS — but the statement drawn from him, that Siddha practitioners were secretive, is a two-clause attribution in a four-page overview rather than a study of practitioner behaviour. It is labelled that way where it appears.

Not yet checked.

roughly seventy-five citations

Of 99 source lines, around two dozen carry a resolvable link and a completed audit. The rest are publications in real venues whose authors have not been through the four questions: Blench, van Driem, Olivelle, Davis, Clarkson, Kenoyer, Chattopadhyaya, Kolipakam, Weerasekara and others. Unchecked is not suspect. The number is stated so that it can be watched.

Corpus.

measurements rather than sources

Asset	What it holds
gret_scan_v2.json	758 GRETIL files, 18,777,671 tokens; term counts with vṛddhi and orthographic variants.
dedr_roots.json	Dravidian etymological root sets for 26 languages.
Rigveda scan	Mention counts, metrical-stratum distributions, co-occurrence ratios.
Direct file checks	Śatapatha, Taittirīya, Bṛhadāraṇyaka, Tirukkuṟaḷ — read against the source text wherever a zero carries an argument.

These ship with the site. Every count states its corpus, its query and its gap, and can be re-run.

On criticism.

critics go through the same four questions

Recorded objections are easy to cite and hard to weigh, and repeating one uncritically demotes a source while appearing rigorous. An objection counts here only once its author has been assessed on the same terms. Where that has not been done — including the challenges to Parpola's decipherment programme, which argue the Indus signs are not writing at all — the dispute is recorded as unresolved among specialists, and no claim here depends on it.