*péd-o-

The comparison between Hitt. pēda- n. ‘place’ and Lyc. pddẽ(n)- n. ‘plain, ground’ allows for the reconstruction of a Proto-Anatolian noun *péd-o- n. (HEG P:589-591 with further literature, Melchert 1994a:262with outdated analysis on the position of the accent, EDHIL:672f.). Outdated is the analysis of HLuw. pi-ta-ha-li-ia- as containing the same word for ‘place’ (as in Melchert ibid.), due to the decipherment of the hieroglyphic sign as reflecting the fortis /-tt-/ as opposed to the lenis /d/ (see Rieken 2008b, which found immediate consensus). Whereas Hittite retains the thematic formation, Lycian seems to have transferred the original thematic neuter noun to the paradigm of neuter n-stems, i.e. *Pre-Lyc. *pdde- n. → pddẽ(n)- (Melchert 1992b:46, DLL:48). The syncope in the Lycian root implies an accent shift to the right, perhaps in relation to the paradigmatic transfer; but cf. under PIE lemma.

[D.S.]

The evidence of several branches (Anatolian, Greek, Italic, Celtic, perhaps Tocharian) permits the reconstruction of a Proto-Indo-European neuter noun *péd-o- ‘ground, place’. This is usually separated from the oxytone neuter *ped-ó- ‘footstep, track’ (Vedic padá- ‘footstep, trace; position’, Old Avestan pada-, Young Avestan paδa- ‘footprint, step; position’, Armenian het ‘trace’, Old Norse fet ‘step’, Old Irish ed ‘space, interval’; see NIL:528), at first glance perhaps a derivative of the word for foot, PIE *pód-, *péd- (see under Hitt. pāta-/patā- and its Anatolian cognates). Whatever the precise derivational origin of *péd-o- ‘ground, place’ (see below), the semantics are less easily understood as based on ‘foot’ than as based on the underlying root *ped- ‘fall, step’ (or “fall, descend, lie down; step, tread, go” as per Nussbaum 2017a:253). Vedic padá-, however, that means both ‘footstep, trace’ and ‘position, site, abode’ already in the Rigveda (cf. also Young Avestan paδa- ‘footprint, step’ and ‘position’ in compounds) shows one of two things: either that the meaning ‘place’ can develop naturally from ‘footstep, track’ (which would favor a derivational relationship between *ped-ó- and *péd-o-), or that *ped-ó- and *péd-o- have influenced each other’s meanings secondarily in the daugther languages. It is important to highlight that most languages (Old Avestan pada-, Young Avestan paδa-, Armenian het, Old Norse fet, Old Irish ed and ined, Latin Pedum, oppidum, (?) Tocharian B patsa, A päts) do not allow any judgement about the prehistoric position of the accent anyway and could perhaps equally well continue the barytone *péd-o- or result from a merger of the two nouns.

The coexistence of *péd-o- ‘ground, place’ and *ped-ó- ‘footstep, track’, and their relationship to *pód-, *péd- ‘foot’ has been a matter of debate for a long time (see the discussion in NIL:534 note 37 and 535 note 46), especially as thematic nouns with a full-grade root are quite rare. It seems tempting to analyze *ped-ó- ‘footstep, track’ as a substantivized denominal adjective *ped-ó- ‘relating to the foot’formally, this would most certainly be a ‘possessive’ adjective, though the meaning would have to be understood as more ‘relational’; compare *b^hā́g- f. ‘beech’ (Proto-Germanic *bōk- ‘book, beech’) → *bhāg-ó- ‘related to the beech, beech-’ (subst. as Greek φηγός f. ‘oak’, Latin fāgus f. ‘beech’) and *péd-o- ‘ground, place’ as an alternative substantivization of this adjective with accent retractionon this substantivization type see AiGr III:88; Schaffner 2001a:328ff.; Steer 2014a:397; Höfler 2015b:226–228. This is formally possible but doesn’t really account that well for the meaning of the latter (quasi ‘the area relating to (sc. the speaker’s) feet’ > ‘ground’?); see also above under Semantic Reconstruction. An alternative would be to see in *péd-o- n. ‘ground, place’ a primary formation derived from the root *ped- ‘fall, step’. As mentioned above, thematic neuters with a full-grade root are exceedingly rare, but they exist. Compare Latin serum ‘watery part of curdled milk, whey’ (from *ser- ‘flow’; accentuation unclear) or Greek ἔργον n. ‘work, deed’, Proto-Germanic *werka- n. > Old Norse verk, English work, German Werk (from *u̯erg̑- ‘make’). The semantics of these formations (viz. ‘what flows (away)’, ‘what is done’) might then imply a meaning ‘where one steps’ for péd-o- ‘ground, place’.

A different account of *péd-o- ‘ground, place’ and *ped-ó- ‘footstep, track’ in particular, and of thematic neuters with a full-grade root in general was proposed by Nussbaum 2017a:253–255 who attributes both formations to the root *ped- “fall, descend, lie down; step, tread, go” (Nussbaum 2017a:253). According to him, the root formed a very predictable R(ó)-o- nominal *pód-o- ‘a step, tread’ (perhaps attested in Lithuanian pãdas m. ‘sole, flooring’, Old Russian podъ ‘floor, bottom’, Serbo-Croatian pȍd m. ‘floor, ground’, though the absence of Winter’s lengthening is a problem). From this was derived a R(e)-ó- possessive derivative *ped-ó- ‘stepped, trodden’, which is attested (according to Nussbaum) in its substantivized form *pedóm ‘trodden’ in Vedic padá- ‘piece of ground’and in Lycian pddẽ ‘place’, for which, however, see above, and in the meaning ‘a step’ in Vedic padá- ‘a step’ (and perhaps in a subset of the continuants in the other languages; see above under Semantic Reconstruction), for which he compares examples such as Vedic krāntá- ‘stepped’ and krāntá- n. ‘a step’. A different substantivization process, namely with accent retraction (as above), then led to *péd-o- ‘trodden’ > ‘place, ground’ as attested in Hittite pēda- ‘place’, Greek πέδον ‘ground, earth’ (and perhaps in a subset of the continuants in the other languages; see above under Semantic Reconstruction).

Note that while this scenario is an elegant solution for *péd-o-, it might still be preferable to analyze *ped-ó- ‘footstep’ as a denominal formation (as above).

[S.H.]

Hittite pēda- n. ‘place’, Lycian A pddẽ(n)- n. ‘plain, ground (vel sim.)’, pddãt(i)- c. ‘place, locality’, pddẽ ‘in front of’.See above and EDHIL:672f. s.v. peda-.

Ancient Greek πέδον n. ‘ground, earth’ (Il.πέδονδε ‘to the ground, earthwards’+), πεδίον n. ‘plain’ (Il.+).Cf., e.g., GEW II:485f. s.v. πέδον, EDG:1160f. s.v. πέδον.

Latin Pedum (place name in Latium; Liv. 2.39.5 etc.), oppidum n. ‘town, city’ (Plaut.+), ‘the barriers of the circus’ (Naev. apud Varro), oppidō adv. ‘very much, completely’ (Plaut.+)*‘to the ground’; Nussbaum apud EDLIL:431, Umbrian peřum, perso n. ‘ground, place, location’, + *-en peřume, persome.Cf., e.g., WH II:214f. s.v. oppidum, 215 s.v. oppidō; EDLIL:430f. s.v. oppidum; Untermann 2000a:524f.

Old Irish ined, inad o, m. ‘place, spot, position’ (*eni-pedos), perhaps the place name Taruedum (in Great Britain) ‘Bull Place (vel sim.)’, and the word/place name Gaulish bouido (Châteaubleau).Cf., e.g., EDPC:116 s.v. *eni-fedo-; NIL:528; Delamarre 2003a:80 and 292.

(?) Tocharian B patsa, A päts ‘bottom (vel sim.)’ ( < *pedā).Cf., e.g., Adams 1999a: s.v. patsa, but retracted in Adams 2013a:387 s.v. patsa (rather *bʰudʰi̯eh₂-).