Hittite ‘fire’ and “proterokinesis” as epiphenomenon Anthony D. Yates University of California, Los Angeles adyates@ucla.edu §1 Introduction §1.1 Hittite ‘fire’ and “proterokinesis”: PIE word for ‘fire’ traditionally cited as an example of a “pro- terokinetic” (PK) nominal paradigm, with alternations between stressed/full-grade root and zero-grade ending in the strong cases vs. zero-grade root and stressed/full-grade suffix in the weak cases. • Hittite forms in (1) would continue reconstructed paradigm essentially intact, providing direct evi- dence for synchronic PK MOBILITY — i.e., alternating stress between root/strong cases and suffix/weak. (1) PIE HITTITE NOM/ACC. SG *péh 2 -wr ˚ > pah ˘ h ˘ ur [p´ aX: w or] ‘fire’ GEN. SG *p(e)h 2 -wén-(e/o)s > pah ˘ h ˘ wenaš [paX: w ´ e:nas] ‘of the fire’ LOC. SG *p(e)h 2 -wén-i > pah ˘ h ˘ weni [paX: w ´ e:ni] ‘in the fire’ · This reconstruction has been all but universally accepted since Kuiper (1942:18) and Schindler (1975a:9–10); see Rieken (1999:331–3), Fortson (2010:120), Meier-Brügger (2010:344–5), and Beekes and de Vaan (2011:206), i.a.. Hittite ‘water’ shows same stress alternation synchronically, but per Schindler (1975a:4–5) due to analogical influence of ‘fire’ (contra Kloekhorst to appear); see further §3.3 below. §1.2 PK mobility & PIE stress assignment: PK mobility difficult to reconcile with general principles of stress assignment reconstructible for PIE on the basis of other nominal and verbal evidence — in particular, Kiparsky and Halle’s (1977)’s BASIC ACCENTUATION PRINCIPLE (see §2 below). • Some proposed solutions to this problem: · PIE accentual system had a more complex set of properties — e.g., accentually “dominant” inflec- tional endings (Frazier 2006; Kim 2013). · Accentual grammar refers directly to [proterokinetic] — e.g., as a morphophonological template (Fellner and Grestenberger 2016). · PK mobility did not exist in PIE (Kiparsky 2010, Keydana 2013). §1.3 The status of “proterokinetic” in PIE: Primary question for today is thus: ◦ What are the implications of Hittite evidence in (1) for the reconstruction of PIE word prosody and the status of PK (mobility) in the grammar? • Major claims advanced here — in broad outline: (i) (1) supports the reconstruction of descriptively PK stress alternations for PIE. (ii) However, alternations in (1) confined to certain “heteroclite” stems, a consequence of their irreg- ular allomorphy and how they interact with regular principles of PIE stress assignment. (iii) (1) does not require reconstructing accentually “dominant” inflectional endings for PIE. (iv) (1) does not require the PIE grammar to reference “proterokinetic” (or any other templatic entity).