Thread:User talk:CodeCat/Latvian divi/reply (11)

I don't think there has been anything about it globally, no. But when others add reconstructed terms that seem doubtful to me, I do discuss them and sometimes they are moved. This has happened a lot with Proto-Germanic so far. Some of those entries are less than certain, while others are very solid and leave little doubt.

Proto-Germanic reconstructions rely on having a good body of descendants. This works well for Germanic because it is a single point of departure for many different languages, so if a single point can be found that fits with all of them, that strongly verifies the reconstruction. For Balto-Slavic it does not work so well because there are less branches, often only three (Latvian, Lithanian and Slavic).

There are sometimes words for which a PIE origin, or a derivational formation of PIE origin, can be identified. This is the case with for example, whose PIE ancestor is very solidly reconstructed by linguists and has many sources. In this case, even if there had only been a Slavic descendant, then I would still not doubt this reconstruction because it could hardly be anything else, given the regular sound changes involved. Said another way, we know the end point (Slavic) the start point (PIE) and the sound changes that happened in between, so we know that it must have existed in PBS as well. The body of known sound changes can then be used to reconstruct the middle point.