*sh₁-i̯é/ó-

The comparison of Hittite šiye/a-^(mi) ‘to impress (with a seal), shoot’ with Palaic šī- ‘to press, urge’ allows for the reconstruction of a Proto-Anatolian formation in *sh₁-i̯é/ó- of the mi-conjugation (thus Oettinger 1979a:474). A root accented *séh₁-i̯e/o- with loss of *h₁ and subsequent monophthongisation is not impossible (see the lenition of the ending in the 3sg. med. šītari), but cf. CLuw. ti(ya)-^(di) for the morphological development of verbs with similar roots. The proposal of Kimball 1987b:163f. (followed by EDHIL:670f, 694f.) that the root was *h₁s- on account for the univerbated pēššiye/a-^(mi) and ūššiye/a-^(mi) is not valid; cf. Melchert 1994a:154. The geminate -šš- comes simply from the sequence *pe-sh₁-i̯é/ó- (with retraction of the accent onto the preverb). Hittite shows an additional formation built on the same root, i.e. šāi-/-i(ya)-^(ḫi) (cf. CHD Š:15-21). According to Sasseville 2021a:429-438, the transfer of an original root verb of the ḫi-conjugation to the class of -āi-/-i(ya)-^(ḫi) occurs in roots ending in *-Vh₁ (but not in *-Vh₂ or *-Vh₃). Therefore, the assignment of the root as *seh₁- would be supported by the Hittite evidence, i.e. PAnat. *sóh₁-/*sh₁´⁽ⁱ⁾ >> *sói̯h₁-/*sih₁-´ > Hitt. šāi-/-i(ya)-^(ḫi); see under PAnat. *sóh₁-/*sh₁´⁽ⁱ⁾. Contra HEG S:703, the Luwian verb šā-⁽ⁱ⁾ ‘to release’ must belong to a different root, see under lemma.

[D.S.]

The meaning of the continuants of the root *seh₁- exhibits a striking distribution. In Anatolian, the verbal forms mean ‘to impress, press, seal’ whereas in all non-Anatolian branches the corresponding verbs refer to the act of sowing. It is easy to see how ‘to impress, insert, press sth. in(to) sth.’ could develop a specialized meaning ‘to press seeds into the soil’, whence ‘to sow’ (cf. LIV²:517 s.v. *seh₁- note 1, EDHIL:695). The interesting thing is that this semantic change probably happened at a common stage in the prehistory of Italic, Celtic, Germanic, and Balto-Slavic (the root is not securely attested in Indo-Iranic, Greek, Armenian, Albanian, Tocharian), and thus seems to reflect the early split-off of Anatolian (cf. Oettinger 2013-2014a:168, Kloekhorst & Pronk 2019a:3). It has been claimed, however (cf. LIV²:518 s.v. *seh₁- note 4), that the original meaning of *seh₁- ‘to insert’ is still preserved in Latin īn-serō ‘to graft on (a scion); to graft a scion on to (a tree); to implant, root, fix (in)’ (cf. OLD:922 s.v. inserō¹).

A zero-grade *-i̯e/o-present *sh₁-i̯e/o- in Anatolian (Hittite šiye/a-^(mi), pēššiye/a-^(mi), ūššiye/a-^(mi), Palaic šī-, perhaps Lydian ši-) corresponds to a full-grade *-i̯e/o-present *seh₁-i̯e/o- in Germanic and Balto-Slavic (Proto-Germanic *sēana-, Lithuanian sė́ju, Proto-Slavic *sějǫ), suggesting that the latter acquired the full grade secondarily, perhaps in analogy to the root aorist (cf. Latin perf. sēvī < *seh₁-u̯°, OCS aor. vъ-sě (?); see LIV²:517f. s.v. *seh₁- note 3). The second primary stem formation attested in Anatolian is the ḫi-conjugation verb PAnat. *sóh₁-/*sh₁´⁽ⁱ⁾ (Hitt. šāi-/-i(ya)-^(ḫi), PLuw. *sói̯-/sii̯-´⁽ⁱ⁾, perhaps Lydian ši-) which does not have a formal counterpart in the other branches. The reduplicated verb Hittite šišša-ⁱ / šišš- is reminiscent of Latin serō, seris, serere (if from *si-sh₁-e/o-; see EDLIL:557 s.v. serō, -ere 1).

[S.H.]

Hitt. šāi-/-i(ya)-^(ḫi) ‘to impress (with a seal), shoot’, pēššiye/a-^(mi) ‘to throw away, to cast, to shove; to abandon; to cast off; to ignore’, ūššiye/a-^(mi) ‘to draw open (of curtains)’, šišša-ⁱ / šišš- ‘to impress’, Hieroglyphic Luwian /sas(a)-/, (“SIGILLUM”)sa-sa- n. ‘seal’,  Palaic šī- ‘to press’, Lydian ši- ‘to impress (with a seal), make official’, šadmẽ- (c.) ‘seal, inscription’.See EDHIL:694f. s.v. šai-ⁱ / ši-; šii̯e/a-^zi, 670f. s.v. pēššii̯e/a-^zi, 930 s.v. ūššii̯e/a-^zi, 757 s.v. šišša-ⁱ / šišš-, LIV²:517f. s.v. *seh₁-.

Latin serō, seris, serere (*si-sh₁-e/o- or *se-soh₁-) ‘to sow, plant’, perf. sēvī (*seh₁-u̯°), ptcpl. satum (*sh₁-tó-), sator, -tōris m. ‘sower’, sēmen, -inis n. ‘seed’Sēmō, -ōnis, Oscan seemuneí dat. sg., Paelignian semunu gen. pl. (?),  the name of a deity, and the derivative Umbrian sehmeníar gen. sg., name of a feast, do not belong here, see Weiss 2017b for a reconstruction of a Proto-Italo-Celtic *seĝ^homōn- ‘strong one, strongman’ that also accounts for Gaulish Segomoni and Ogham Irish SEGAMANAS.Cf., e.g., WH I:522 s.v. 1. serō, EDLIL:557 s.v. serō, -ere 1, LIV²:517f. s.v. *seh₁-.

Proto-Celtic *sīlom ‘seed’ (Old Irish síl, Welsh, Cornish, Breton hil), *satV- ‘seed; offspring’ (Welsh had, Middle Cornish has, Middle Breton had).Cf., e.g., EDPC:336 s.v. *sīlo-, Zair 2012a:57, 109.

Proto-Germanic *sēana- ‘to sow’ (Gothic saian, Old English sāwan, Old High German sāwen, German säen, Dutch zaaien, Old Norse sá, Swedish, Danish, Norwegian så), *sēdiz f. ‘seed’ (Gothic °sēþs in manasēþs ‘world’, Old High German sāt, German Saat), *sēda- n. ‘seed’ (Old English sǣd, Old Frisian sēd, Dutch zaad, Old Norse sáð, Swedish säd, Danish, Norwegian sæd), Proto-West-Germanic *sāmō m. ‘seed’ (Old Saxon sāmo, Old High German sāmo, German Samen, Same).Cf., e.g., Seebold 1970a:386f., EDPG:428 s.v. sēan-, 429 s.v. *sēda-, 430 s.v. *sēdi-, LIV²:517f. s.v. *seh₁-.

Lithuanian sė́ju, sė́ti ‘to sow’, sėklà, sė́kla f. ‘seed’, sėmuõ m. ‘seed’, Latvian sẽju, sẽt ‘to sow’, sȩ̃kla f. ‘seed’, Old Prussian semen ‘seed’.Cf., e.g., ALEW II:1042f. s.v. sė́ti, LIV²:517f. s.v. *seh₁-.

Proto-Slavic *sějǫ, *sě̀ti, *sě̀jati ‘to sow’ (Old Church Slavonic sějǫ, sěti, Serbo-Croatian sȉjēm, sȉjati, Russian séju, séjat’, Czech seju, sít), OCS aor. vъ-sě (?) ‘sowed’, PSlav.*sě̀mę n. ‘seed’ (OCS sěmę, SCr. sjȅme, Russ. sémja, Polish siemię).Cf., e.g., ALEW II:1042f. s.v. sė́ti, EDSIL:447f. s.v. *sě̀ti I; *sě̀jati I, 446 s.v. *sě̀mę, LIV²:517f. s.v. *seh₁-.