Reconstruction talk:Proto-Celtic/swexstos

Form
Could this be instead? The evidence does seem to point in that direction. —CodeCat 17:57, 25 September 2016 (UTC)
 * I'm not 100% sure, but I believe that without a Celtiberian reflex, we can't know which it was. A word-internal st cluster became /ts/ in Gaulish and /ss/ in Insular Celtic, and it's eminently plausible that /xts/ and /xss/ would develop exactly like original /xs/ in both branches, unless there is some other word with -xst- where we know that not to be the case. And given, , and , is more to be expected than , and it seems that the former is not inconsistent with the reflexes. —Aɴɢʀ (talk) 18:36, 25 September 2016 (UTC)