9 prompts · 183 messages · 63 tool calls · 0 commits · 2 pages
In my ~/.notes/org-roam/ find an org-roam node "Proto-Indo European". That has my notes on PIE. I want you to convert that into a more serious defensible set of notes. For every claim in there, and for every example in there, search for academic papers or books or other reliable sources that support the claims/examples. If you are able to find it, include the link to the reference right there. If you are unable to find it, then clearly mark the claim as dubious; and if you find sources contradicting anything, or if there are errors, then clearly mention those. In the end create a short reading list of what someone interested in this area should read in what order. Put all of this in an html file here, and open it in my browser. Ask me questions if necessary before starting.
I've found and read the org-roam file (20260516080917-proto_indo_european.org). It's quite rich — covering reconstruction methodology, Grimm/Verner, laryngeal theory, Indo-Iranian split, Dravidian substrate influence on Sanskrit, dating evidence (Mitanni, wheel vocabulary), etc. There's a lot to verify.
Before I start, a few questions to make sure I produce what you want.
Include research references where it makes sense!
Opened /home/navin/dl/agents/pie/pie-verified.html in your browser.
What's in it (side-by-side, original on left, verification + clickable sources on right):
Errors flagged (red): - Hittite/laryngeal date conflation — Hittite was deciphered in 1915 (Hrozný), but Saussure's laryngeals were confirmed by Kuryłowicz 1927, not 1915. - Tamil minimal pair: "paṭu (lie down) vs patu (ten)" is wrong — Tamil for ten is pattu, not patu. - aṣṭa retroflex example: misattributed to Dravidian; it's a regular Indo-Iranian outcome (Avestan also has š). - "Echo reduplication: nowhere else except Turkish" — wrong; it's a broad Eurasian areal feature (Armenian, Mongolic, Bulgarian, Persian, etc.).
Debated (yellow): - Retroflex-from-Dravidian framing — three competing accounts (Kuiper/Emeneau, Hock, Witzel). - deva/daēva inversion — likely post-Indo-Iranian theological reform, not a PII split. - Sanskrit SOV rigidity from Dravidian — partially internal drift per Hock. - "Mama-papa universal" — common (~52%), not universal (Jakobson/Murdock). - "Population size → simplification" — real correlation but contested.
Nuances (blue): - Mitanni: god-names are from the treaty; horse-training numerals (aika, tera, …) are from Kikkuli's manual — two separate documents. - centum/satem is an areal isogloss, not a phylogenetic split. - "Sound laws have no exceptions" with directionality: majority-rules has limits.
Reading list: ordered 1–12, from Anthony (most accessible) through Watkins, Fortson, Beekes, Masica, Krishnamurti, Witzel, to LIV². Plus online standing references.
All claims have clickable links (Wikipedia, Britannica, Encyclopaedia Iranica, U Texas LRC, Witzel's PDF, Anthony excerpts, etc.) backed by named textbook citations.
Sources used: - Wikipedia: Filippo Sassetti - Wikipedia: William Jones (philologist) - Wikipedia: Laryngeal theory - Wikipedia: Indo-Aryan superstrate in Mitanni - Wikipedia: Centum and satem languages - Wikipedia: Indo-Hittite - Wikipedia: Substratum in Vedic Sanskrit - Wikipedia: Echo word - Wikipedia: Grimm's law - Wikipedia: Verner's law - Wikipedia: Brugmann's law - Wikipedia: Ruki sound law - Wikipedia: Tamil phonology - Britannica: Grimm's Law - Britannica: Sir William Jones - Britannica: Verner's Law - U Texas LRC: Laryngeal theory - Witzel 1999: Substrate Languages in Old Indo-Aryan (PDF) - Encyclopaedia Iranica: Eastern Iranian languages - Encyclopaedia Iranica: PIE poetic craft / prosody - Language Log: horse and wheel in early IE - Kloekhorst: Anatolian stop system & Indo-Hittite (PDF) - Hock 2013: PIE verb-finality (Benjamins) - Slade 2020: Verb concatenation in Asian linguistics (PDF) - Anthony, Horse Wheel Language, ch. 4 (online excerpt) - Narasimhan et al. 2019 (Science) — South/Central Asian population formation - Haak et al. 2015 (Nature) — Steppe migration - DSAL/Chicago — Burrow & Emeneau DED - Preply: Tamil minimal pairs - Sentence first: mamas & papas (Jakobson)
For the echo reduplication: what is a more defensible statement to make? And for the retroflex in Sanskrit, what are some good examples I can give?
"Echo / 'm-reduplication' belongs to a broad Eurasian areal feature running from Turkic and Armenian through Mongolic and Persian into the South Asian linguistic area, where it reaches its densest and most productive form. Inside South Asia it is shared across all four major language families — Indo-Aryan, Dravidian, Munda, Tibeto-Burman — and is one of Emeneau's classic 'India as a Linguistic Area' diagnostics. Outside that contact zone the pattern thins: English has only marginal Yiddish-borrowed schm- (fancy-schmancy), Turkish has m- (kitap-mitap), Armenian has m- (seġan-meġan). Full productive echo of the South Asian chāy-vāy / tēṉīr-kīṉīr type is rare in Western European languages."
Sources: Stolz, Stroh & Urdze, Total Reduplication (Akademie 2011); Abbi, Reduplication in South Asian Languages (1992); Wikipedia: Echo word; Emeneau (1956).
Use these three buckets together. They show the actual structure of the argument better than a single list.
Bucket 1 — Retroflex from a regular Sanskrit-internal rule (not substrate evidence). Anyone arguing for substrate influence needs to exclude these first.
*s after i → ṣ), then sd → ḍ. Pure internal IE.References: Mayrhofer, EWAia; Wikipedia: RUKI.
Bucket 2 — Retroflexes where no internal IE rule predicts the retroflex, and a clean Dravidian cognate exists. These are the real candidates for substrate evidence.
| Sanskrit | gloss | Dravidian / Tamil cognate | Notes |
|---|---|---|---|
| kuṭa, kuṭī | hut, dwelling | Tamil kuṭi "dwelling" | One of the strongest candidates (DEDR 1655). |
| daṇḍa | stick, staff | Tamil taṇṭu "stalk, staff" | Standard substrate example (DEDR 3056). |
| mayūra | peacock | Tamil mayil | No PIE etymology; classic "culturally Indian" loan. |
| naḷa / naḍa | reed | Tamil naḷ | Already in Rigveda. |
| kuṇḍa | pot, bowl | Tamil kuṇṭu "depression" | Cluster of pot/cavity words. |
| piṇḍa | lump (of food) | non-IE | No PIE cognate. |
| karpāsa | cotton | Tamil paruttī / Munda? | Cultural/agricultural loan. |
| aṇu | small, atomic | Tamil aṇu "small" | Already Rigvedic. |
| kāṇa | one-eyed | Tamil kāṇ "to see" cluster | Often cited (Kuiper). |
References: Burrow & Emeneau, Dravidian Etymological Dictionary (DED², 1984) — searchable at DSAL/Chicago; Kuiper, Aryans in the Rigveda (1991); Krishnamurti, The Dravidian Languages (2003), §1.6.
Bucket 3 — Debated cases (use sparingly).
Reference: Witzel, "Substrate Languages in Old Indo-Aryan", §§4–5.
The presentation trick: show Bucket 1 first to establish that some Sanskrit retroflexes are demonstrably internal — that grants credibility — then introduce Bucket 2 as the residue that needs an external explanation, and acknowledge Bucket 3 honestly. That's the structure Krishnamurti and Hock both use, and it's much harder to attack than a flat "retroflexes = Dravidian" claim.
Want me to fold these into the HTML?
I have made some changes in the original org-roam note. Update the html accordingly, and after that is done, spawn a subagent to check your work...
I see the changes: Hittite date corrected to 1927, Dravidian retroflex section completely rewritten with my three-bucket structure (internal: nīḍa, iṣṭa, aṣṭā(u); Dravidian loans: kuṭa, daṇḍa, naḷa, aṇu), and echo reduplication rewritten with the correct cross-linguistic comparison. Updating the HTML now.
HTML updated and audit complete. Summary of the subagent's findings:
Clean: - All three updated sections (laryngeal, Dravidian retroflexes, echo reduplication) faithfully reflect your edited org. - No stale ERROR badges remain. - Load-bearing citations check out: Kuryłowicz 1927 paper title verified exactly; Abbi 1992 reduplication monograph verified; nīḍa derivation correct. - All URLs spot-checked work.
Four minor refinements worth knowing about (none are errors):
*h₃eḱtō for "eight" — defensible (Beekes/LIV style) but other handbooks give *oḱtṓw (no laryngeal) or *h₁oḱtṓw. Worth a NUANCE flag if you want maximal defensibility.*Hi-Hs-tó- for iṣṭa — unusual notation; current handbooks (Mayrhofer, Wiktionary) prefer *h₂is-tós from root *h₂eys- "wish." Both are defensible; yours is the older reduplicated zero-grade form.aṣṭā(u) "palatalization of ḱt" — strictly it's RUKI-like behavior of the satem palatovelar before t, not palatalization per se. Substantively the conclusion is the same.Want me to (a) add a NUANCE flag for #1, (b) update my prose for #3, or (c) leave it as-is?
Add a nuance flag for #1, update your prose for #3, I have changed #2 and #4. Check and make the changes.
All three changes done:
*h₂is-tós for iṣṭa, "RUKI-like substitution of ḱt" for aṣṭā(u).zd → ḍ); iṣṭa cites the root *h₂eys- "wish"; aṣṭā(u) prose now says "RUKI-like behavior of the satem palatovelar before t" instead of "palatalization." Added a Wiktionary link for iṣṭa.*h₃eḱtō, *oḱtṓw, and *h₁oḱtṓw all appear in standard handbooks, with citations to Beekes, Mallory & Adams, Fortson, and a Wiktionary link.Refresh the browser tab to see the updates.