*seP-/*sP-i̯é/ó-

The etymological relationship between Luw. šapiya(i)-^(di) ‘to scrub, to cleanse’ and Hitt. šappā(i)-^(mi) ‘to peel’ or šapp-^(mi) ‘to churn (milk)’ suggested by Oettinger 1979a:383f., 535 must be left aside due to the different semantics and the different spellings of the labial stop (Melchert 2003c:149f.). Considering the Hittite-Luwian word group of šapiyā(i)-^(di) and the Palaic cognate šapawina-⁽ⁱ⁾ ‘to scrub, cleanse’ (Carruba 1970a:69, Eichner 2010a:44), one can reconstruct a verbal root with two kinds of primary stem formations, i.e. a root verb and a deradical verb in *-i̯e/o-. The latter, i.e. *šapi(ya)-, must be reconstructed as the derivational base of the abstract noun Luw. šapiya- ‘washing vessel’ and Hitt. ša/epiya- c. ‘id.’ (obtained by hypostasis from *šapiya- c. ‘scrubbing, cleansing’), from which the denominative verb šapiyā(i)-^(di) is derived. Under ša/epiya- c., it is argued that the vacillation in the initial cluster of the Hittite word is to be read as /sp-/. Thus, one could reconstruct a verbal formation in *-i̯é/ó- with zero-grade of the root, i.e. *sP-i̯é/ó-, which would go back to a root *seP-. The Palaic cognate allows for both readings, /spawina-/ or /sabawina-/. For new evidence on the treatment of the Palaic initial cluster /sp-/, see Sasseville 2019a:27 n. 34.

[D.S.] [E.R.]