*ēlō

Traditionally analyzed as descending from Proto-Indo-European *h₁éh₁l-eh₂- (‘awl’) from a theoretical root *h₁el-. Cognate with Sanskrit आरा (ā́rā) (ā́rā, ‘awl’) (Kroonen 2013: 19; Mallory & Adams 1997: 37). It is generally connected with *alaz (‘awl’), but the vowel alternation *a ~ *ē is not straightforward: (Kroonen 2013: 19, 117) alternatively derives *ēlō and *alaz from different stem grades of an inherited ablauting paradigm, with *ēlō generalized from the nominative *h₁éh₁l-eh₂ and *alaz from the oblique stem *h₁h̥₁l-h₂-, or treats *alaz as an independent formation *h₁oh₁-ló-, with pretonic shortening of earlier *ō. Mallory and (Adams 1997: 37) give the more general reconstruction *hₓólehₐ- (‘awl’, in standard laryngeal notation: *Hóleh₂-). However, I believe an analysis of *ēlō as an inherited instrument noun *h₁éh₁-tléh₂- is formally and semantically attractive in light of Kroonen’s account of Proto-Germanic *-l- from PIE *-tl-/*-dʰl-(Kroonen 2017). It competes with the conventional reconstruction with the root itself ending in *-l-, but that analysis is not based on an independently established root; rather, it reflects the traditional segmentation of Sanskrit आरा (ā́rā) (ā́rā, ‘awl’) and the PGmc forms. The relation of *alaz could then involve a parallel masculine instrument formation, perhaps *h₁oh₁-tló-, though the precise vocalism and PGmc consonant development would still require explanation, and this could sever the relationship with the Sanskrit noun as a direct cognate. (Liberman 2020) contends that the length of the root vowel in PGmc is essentially unknown and considers the word a likely Wanderwort.