-plẽ

Eichner 1992a:62, 74 reconstructs Lyc. B tbiplẽ and trpplẽ as *du̯i-plom (or *du̯i-pelom) and *tri-plom (or *tri-pelom). Thus, the Lycian ending *-ẽ in -plẽ can be considered to be a frozen accusative case ending, i.e. *-om. The suggested variant *-pelom is not supported from a comparative perspective see under Proto-Indo-European Reconstruction. For the absence of the -i- in trpplẽ as opposed to expected *triplẽ; cf. Eichner 1992a:74, Kloekhorst 2014c:65.

[D.S.]

The second compound member of PIE *du̯i-pló- ‘thrice’, *tri-pló- ‘twice’ etc. is best connected with the root PIE *pel- ‘to fold, layer’, because “[t]he multiplicative adj[ective]s […] are mostly c[om]p[oun]ds of words for ‘fold’ or the like” (Buck 1949a:945). Thus, e.g., PIE *tri-pló- ‘twice’ literally means ‘three-fold’ albeit not ‘folded thrice’ (which would make four or eight and not three layers) but ‘folded so many times that the result consists of three layers’.In view of this analysis, PIE *sm̥‑pl‑ó- ‘once, one time, simple’ cannot be paraphrased by ‘with one fold’ (as per LIPP II:676).It should be clear that one has to seperate the number of repetitions of folding (two) from the number of resulting layers (three).
Because (at least) two different roots could theoretically underly PIE *-pló- (cf. below), the second compound member PIE *‑pl- originally means either ‘layer’ or ‘division, part’, depending on the semantics of the underlying root. Thus, the compounds should be interpreted either as ‘having X layers/parts’Cf. also LIPP II:676 n. 33 on Gk. δί-, τρί-πλαξ ‘two-, threefold’ < *“in zwei, drei Schichten”.or as ‘having an X-fold layer/division’, cf. below on the morphological analysis.

Even if PIE *-pló- could be understood as a suffix(oid) in the daughter languages, it obviously was an independent lexeme and the formation itself a compound. The compound member PIE *‑pl(‑ó)- ‘layer/division’ is the zero-grade variant of an unattested root noun *pel(H)- meaning either ‘layer’ or ‘division’ (cf. above). There are ultimately two options how to interpret the compounds of the type *[numeral]‑pló‑. Either *[numeral]-pl-ó‑ is the thematicized version of a bahuvrīhi *X‑p(E)l- ‘having X layers/parts’Similarly, PGerm. *-falþa- (cf. Goth. ain-falþs ‘one-fold’ etc.) and Gk. -πλάσιος ‘-fold’ are derivatives of such a compound.or the thematic vowel is of the possessive type and the compound endocentric, thus *X‑p(E)l- ‘X-fold layer/division’ → *[numeral]‑pl‑ó- ‘having an X-fold layer/division’. The second option (endocentric → possessive) is less likely insofar as the first compound member is the (zero-grade) cardinal and not the multiplicative numeral. Thus, it is opted here for the first option (bahuvrīhi → thematicization).

PGk. *-pló- ‘-fold’, cf. Gk. ἁ-πλός (An. Ox. 2.231), ἁ-πλό(ϝ)ος, ἁ-πλοῦς (Pi.+) ‘single, simple’The derivative Hom. ἁπλοΐς ‘simple, single (of garments)’ proves that its derivational base ἁπλό(ο)ς already existed in Homeric Greek.
According to EDG:117, the affiliation of Gk. ‑πλό(ϝ)ος “is problematic, as Gr. -πλος is late and rare compared to -πλόος”, but cf. Cret./Herac. adv. (loc.) διπλει, διπλη ‘twofold, twice’ (s.v. Lyc. B tbiplẽ) and Hom. adv. (dat.‑loc.) τριπλῇ (Il.1.128) (s.v. Lyc. B trpplẽ) for the antiquity of the variant Gk. -πλό- ‘-fold’. However, Gk. ‑πλόος could be influenced by or directly stem from homonymous πλόος ‘sailing, voyage’ (as per Kretschmer 1923a:218 and LSJ:191), cf. ἁπλῆ οἶμος ‘one-way road’ (A. fr.239), δεύτερος πλοῦς ‘(lit.) second voyage, second try, next best way/thing’ (Pl.+), SCr. jedan put ‘(lit.) one journey > once’ and Buck 1949a:943–945 with examples for phrases and univerbations for ‘three times’ originally meaning ‘three journeys’. Cf. differently Brugmann 1917a:133–135.(GEW I:121f., GEW III:35, DELG:97, EDG:117, LSJ:190f., Brugmann 1917a:128–135), for more cognates, cf. Lyc. B tbiplẽ, trpplẽ.It is not clear whether Gk. (δί‑, τρί‑)πλαξ ‘(two‑, three‑)fold’ (LSJ:435, 1821) belongs here or not, cf. Gk. πλάξ, πλακός ‘anything flat’ (LSJ:1411f.).

PItal. *-plo- ‘-fold’, cf. Lat. sbst. sim-plum n., -a f. ‘the simple (sum)’ (Plaut.+), adj. sim-plus ‘single, simple’ (Cic.+) (WH II:540, EM:517, EDLIL:475, OLD²:1946), for more cognates, cf. Lyc. B tbiplẽ, trpplẽ.It is not clear whether Lat. (du‑, tri‑)plex ‘(two‑, three‑)fold’ (OLD²:637, 1531, 2178) and Umbr. acc. sg. n. (tu‑)plak ‘fork?’ (hapax, IgT III.14, Untermann 2000a:775) belong here or not.

On PGerm. *twīfla- ‘doubt’, cf. Lyc. B tbiplẽ.

PIE *pel- ‘to layer’
(?)Indo-Iranian
OInd. puṭa- ‘fold, pocket’ (Mn+)If from *pl̥‑to‑ ‘what is folded’. It could also belong to PIE ?*(s)pelt- ‘to split’ (LIV²:557), viz. *(s)pl̥t‑o- ‘slit < what is split’.
EWAia II:141, Böhtlingk 1855a IV:754, V:1602, VII:1771
(?)Armenian
Arm. haƚ ‘time’ (cf. haƚ k‘an zhaƚ ‘a little at a time’)If from *pol-o- ‘what is folded’, cf. IEW:803.
EDAIL:—, Olsen 1999a:531 with n. 670
Germanic
PGerm. *falþ- ‘(to) fold, ply’, cf. Goth.(+) falþan ‘to fold’, MHG valte ‘fold’ etc., Goth.(+) ain-falþs ‘simple (< *one-ply)’Another option is to affiliate this Germanic root with PIE ?*(s)pelt- ‘to split’ (LIV²:557) or to derive the strong verb from the verbal noun PGerm. *falþa- ‘fold(ed)’, cf. EWAhd III:26.19–31.
EDPG:126, de Vries 1977a:110, 116, GED:107f., EWAhd III:25.18–28.28, Orel 2003a:91
Albanian
Alb. palë ‘fold, bend, layer’
AED:309, Demiraj 1997a:308f.

PIE *(s)pelH- ‘to seperate, split off’
Indo-Iranian
OInd. phalati ‘to burst, shatter’ (Sū)
EWAia II:201, KEWA II:393, Böhtlingk 1855a IV:1199
Balto-Slavic
OCS(+) plěti ‘to weed’
EDSIL:394f., Vasmer II:394f., 398
BSl. *pelHu- ‘chaff’ (< ‘what is split off’), cf. OCS(+) plěvy (nom. pl.) ‘id.’, Lith. pẽlūs (nom. pl.) ‘id.’, Latv. pęlus (nom. pl.) ‘id.’, OPr. pelwo ‘id.’ (E 279)These nouns could also belong to a different root, viz. PIE *pelh₁- ‘to set in swing’ (LIV²:469f.), cf., e.g., EDBIL:350.
EDSIL:394f., 395, ALEW II:754f., EDBIL:350, SEJL:449, LEW I:568f., PKEŽ² s.v. pelwo