Thread:User talk:CodeCat/Appendix:Proto-Indo-European/ātr-/reply (6)

Yes, they indicate that the laryngeal is unknown. Some authors tried to solve it with labialization ʷ, but I have also never seen it in the combination with Hʷ before, except for labialized voiceless pharyngeal fricative in the Abkhaz language, labialized voiceless glottal fricative  in Akan, Tlingit and Tsakhur languages, and labialized voiceless velar fricative. The latter one seem to be the source for Iranian Avestan ătarš (-> Labialization#Fricatives). It's a very soft w coming from the throat being combined with the vocals a or o/u. Correct me if I'm wrong, but for now I prefer *hₓehₓtr-, since Serbo-Croatian vātra, Romanian vatră, and Albanian votre can be only explained by Hʷ/hₓ. This would make sense since all three vātra/vatră/votre are thought to have an Iranian Avestan origin. If you are not fine with *hₓehₓtr-, we can also use either *ăt(e)r- (from IEW: 69) or *ăter- (from Wat: 4). They use ă instead of the long ā.