Reconstruction talk:Proto-Slavic/divъ

Either *divъ is back-formed from *divьnъ or connection with dievas etc. is impossible (for semantic reasons), see Vasmer. Guldrelokk (talk) 07:13, 19 August 2018 (UTC)
 * All nouns ending in pIE -wos are in principle back-formations. This was originally an adjective suffix (from the Caland system). Anyhow, you're right that the derivation from *dyewos is not very convincing. In fact, the etymological section currently gives *dʰeyh₂- (attested also in Indo-Iranian) as the proto-source. Bezimenen (talk) 12:37, 9 January 2019 (UTC)
 * They are not back-formations, they are substantivised adjectives. Back-formation is a derivational process which produces words by substracting derivational morphemes (either etymological or not) from another word. Substantive adjectives like *divъ or *korva do not have to be back-formations. Guldrelokk (talk) 13:36, 9 January 2019 (UTC)
 * Note taken. Substantive adjective's what I meant... Yet, to call the reanalysis of Caland "adjectives" (some scholars insist on the term predicative adjectives) back-formation is not entirely wrong. For example, it often occurred via prosodic and even morphological innovations (e.g. < ). If we have to be more accurate, this type of evolution should be called Vṛddhi-derivation/reanalysis/back-formation/etc., but even this term is not exactly precise, because vṛddhi is a grammatical process. The evolution of the Caland system, on the other hand, is a diachronic process.

Adjective divъ(jь)?
Besides the noun *divъ, there is also the adjective *divъ(jь) (see ). Should we add it here or it's better to create a separate page? It should be pointed that -jь /in masculine gender/ is merely an extension conveying definiteness. It got fossilized in East Slavic and partially in West Slavic. That's the only reason why Trubachev distinguishes the two forms. In actuality, the dictionary form of *divъ(jь) should be also *divъ (at least in principle). Bezimenen (talk) 11:57, 9 January 2019 (UTC)
 * Wiktionary has short forms as lemmas. Guldrelokk (talk) 13:36, 9 January 2019 (UTC)
 * Precisely. I just guessed that the author of the original entry didn't add the second meaning, because it's not always given as a doublet of *divъ in the dictionaries. Bezimenen (talk) 14:36, 9 January 2019 (UTC)