Talk:āᵤw

RFD discussion: April 2019–March 2020
The source I gave for this word, both when I listed it as a descendant of PIE *h₂ówis and when I made the actual entry on Wiktionary, uses the singular form of the word as an entry. It is not reconstructed; it is entirely attested. The person who originally listed awi as a descendant of *h₂ówis gave no source for their entry and as to why it should be in the plural, and I can't see any reason why there should even be an entry for the plural, as that's not the lemma form for Tocharian B. GabeMoore (talk) 14:43, 10 April 2019 (UTC)
 * āᵤw, or ā₍ᵤ₎w, as R:txb:Adams:2013 fashions it, is a reconstructed transliteration. There is no <ᵤ> character in Tocharian alphabet, and, as far as I know, that word is only attested in plural. -- 15:26, 10 April 2019 (UTC)
 * But it is common across Wiktionary to infer the lemma form from the attested forms. We do it for Gothic, we do it for Latin. I don't see why this is any different. —Rua (mew) 16:11, 10 April 2019 (UTC)
 * We also don't do so for a lot of languages, like Old Persian and Oscan, and for poorly attested languages, where scholars still debate the declension these attestations, this should be the norm. --  16:27, 10 April 2019 (UTC)
 * I emailed Dr. Adams about it, who wrote the source I referenced. He confirmed that the nominative singular form is attested, and specifically that he did not reconstruct it. GabeMoore (talk) 03:22, 12 April 2019 (UTC)
 * Really? Funny because I wrote him and he said it isn't. =P Hearsay external communications are inadmissible. Please show me a source text. -- 03:30, 12 April 2019 (UTC)
 * The source text is the dictionary itself. If the term was reconstructed, it would have an asterisk by it, as reconstructed terms typically do; Adams confirmed this explicitly, both in our email and in the foreword of the book. Would submitting a screenshot of the email count as sufficient proof to back this claim? (Genuine question; I'm not completely sure what's considered "admissible" in that regard.) GabeMoore (talk) 03:57, 12 April 2019 (UTC)
 * I'm aware where it's from -- what I'm asking for a citable text with the word in the nominative singular, ex. PK LC I.4: orotsana awi as cited for the nominative plural, or at least another source that cites the existence of a nom.sg. And no, a screenshot wouldn't help. -- 04:10, 12 April 2019 (UTC)
 * Why would the locus of the term be necessary though? There's nothing at all in the dictionary indicating that the term is unattested apart from the lack of an example of the attestation, which isn't given for many of the entries therein. GabeMoore (talk) 15:15, 12 April 2019 (UTC)
 * Because I doubt it's existence, thus the RFD. Published works aren't infallible, which is why we look for multiple agreeing sources. -- 19:45, 12 April 2019 (UTC)
 * per the principle of placing entries at the lemma. —Rua (mew) 20:27, 12 April 2019 (UTC)
 * Delete unless it can be cited elsewhere as attested because we don't lemmatize most poorly attested languages. -- 00:23, 31 May 2019 (UTC)


 * Keep per Rua and GabeMoore. Tocharian B is not nearly as poorly attested as Oscan. And while ᵤ isn't a separate letter of the Tocharian alphabet, it is the conventional transcription of a specific way of writing vowels in the Tocharian script. Krause & Thomas (p. 40) write: "Eine Eigenart der tocharischen Schirft ist die gelegentliche Verwendung von nichtvollsilbischen Vokalen in offener Silbe und in der unsilbischen Komponente eines Diphthongs, vor allem des u, aber auch des ä, i und selten des o, z.B. A ᵤ͡pādhyā, k͡ᵤpre, k͡ᵤli; ᵢ͡śe, w͡ᵢnāṣlne; ₒ͡kät. — B ᵤ͡pādhyāye, k͡ᵤse, kr͡ᵤi; n͡ₐ̈no, s͡ₐ̈suwa; p͡ᵢśākka, w͡ᵢnāssī. — tāko͡ᵢ; ke͡ᵤ, käsko͡ᵤ; sā͡ᵤ. In der tocharischen Schrift erscheinen in solchen Fällen beide Vokalzeichen in einem Akṣara." So the presence of a ᵤ in our entry (we can dispense with the tie mark as unnecessary) does not in any way suggest that this must be a reconstructed form. —Mahāgaja · talk 09:25, 24 August 2019 (UTC)
 * RFD-kept. —Μετάknowledge discuss/deeds 06:40, 23 March 2020 (UTC)