/naba/, NEG2-pa
‘or’
- Language
- Hieroglyphic Luwian
- Grammar
- conj
- ID
- 3155
1. Individual Anatolian Languages
1.1 Hieroglyphic Luwian
/naba/, NEG2-pa
‘or’1.1.1Transmission
This lexeme is attested 26 times in inscriptions from Aleppo, Tell Ahmar, Karkamiš, Maraş, Commagene and the Amuq, dating from the late 10th to the 8th century BCE.
[A.P.]
1.1.2Forms
The attestation at KARKAMIŠ A27g, l.3 could not be verified for lack of a photograph.
[A.P.] [AH.B.]
1.1.3Graphic Features
1.1.3.1Phonological Interpretation
The phonological interpretation of this word can be viewed as certain, because the parts of this compound are well-known. Its elucidation is due to Hawkins 1975a:120, who teased apart the signs REL (sign *329, 〈kwa/i〉), NEG2 (*332b, 〈ná〉) and NEG3 (*332c), which until then had erroneously been conflated and analysed as forms of the relative pronoun, cf. also the discussions at /na/ ‘not’ and /ni/ ‘not (prohibitive)’.
Although the present lexeme with its partly logographic spelling was misanalysed for a long time as a relative, the phonological form /naba/ had been established by scholars as extant (see Hawkins 1975a:124), albeit connected to the etymologically related particle /niba/ ‘or’, whose spelling was read /naba/ until Mittelberger 1962a:280, followed by Hawkins, Morpurgo Davies & Neumann 1973a:151, established the signs *411 and *214 to represent 〈ni〉 and 〈ní〉 respectively instead of NA.
[AH.B.]
1.1.4Meaning
The meaning of this word as a disjunctive could only be elucidated once its spelling had been understood as containing NEG2 rather than REL, see Hawkins 1975a:120-124, followed by Morpurgo Davies 1975a:160f., and fully updated by Hawkins & Morpurgo Davies 2010a:110-115. The particle is a mainstay in curse formulae, often found linking together the parts of lengthy protases, e.g. BABYLON 1, §14: NEG2-pa-wa/i-tà ARHA MALLEUS-i, approx. /naba ahha wallai/, meaning ‘(He who … and removes this stele from its place), or smashes it, (the storm-god of Tarhunt of Aleppo shall then not give him food and drink to destroy!)’ (cf. Hawkins 2000a:392).
The particle /naba/ mostly occurs clause-initially and hosts a string of clitics, but it can also appear clause-internally, e.g. at MARAŞ 14, §5. In contrast to /niba/, /naba/ is not necessarily followed by the particle /=wa/ when used clause-internally, cf. Hawkins & Morpurgo Davies 2010a:113f.
Noticeably, /naba/, like its synonym /niba/ ‘or’, is only attested in 1st millennium BCE texts, whereas the older inscriptions from the 2nd millennium BCE do not seem to feature it – possibly by chance (cf. Hawkins & Morpurgo Davies 2010a:110). There is no discernible semantic difference between /naba/ and /niba/. Rather, Hawkins & Morpurgo Davies 2010a:112 note that /naba/ appears slowly to be replaced by /niba/ throughout the course of the 1st millennium BCE, although there is considerable temporal overlap between the two forms.
[AH.B.]
1.1.5Stem
The particle is a compound of the factual negation /na/ ‘not’ and the connective particle /=ba/ ‘and, but’, see Morpurgo Davies 2011a:211 also for details on the grammaticalisation process.
For previous, outdated morphological analyses cf. e.g. the synopsis given by Morpurgo Davies 1975a:160f.
[AH.B.]
2. Proto-Indo-European
2.1Morphological Reconstruction
In combination with the clitic /=ba/, three of these four particles (/na=ba/, /ni=ba/ and /nis=ba/) acquired the disjunctive meaning ‘or’. HLuw. /niba/ and /nisba/ function indiscriminately as negative disjunctive conjunctions meaning ‘or not’. The phonological reconstruction is ambiguous, allowing for PIE *ne-, *nē-, *nei̯- and *-pe, or -po.Lycian nepe might point to *no-po; cf. Hajnal 1995a:102 n. 68. Formal and functional cognates are found in other Indo-European languages: PIE *ne-pe, *nē-pe (Osc. ni-p ‘or not’), *nē-poi̯/po- (East Tocharian mā pe, mā pat ‘or not’) alongside PIE *nei̯-pe (Osc. nei-p ‘or not’).
The reconstruction of Luwian /ni-/ remains uncertain because PIE *ne, *nē and *nei̯ are likewise used indiscriminately in the following forms:
- PIE *ne=ku̯os/ku̯is (OLat. nequis ‘no one’, Ved. na-kih) alongside PIE *nei̯=ku̯os/ku̯is (OCS ni-kŭto, Av. naē-čiš), PIE *nei̯ (*nei̯ ‘if not’, 1st line of Duenos inscription; Old Latin quippi-nī , quid-nī ‘why not?’; the reconstruction as PIE *néi̯d as suggested by Fries 2020a is false); PIE *noi̯ (*noi̯-si ‘if not’, 2nd line of Duenos inscription, Harðarson 2011a:157). For its part, HLuw. /naha/ as in /naha… naha …/ ‘neither … nor …’ points to PIE *ne-h2o … ne-h2o … ‘neither … nor …’ (see *=h2o); cf. Lat. neque … neque ….
2.2Syntactic Reconstruction and Phraseology
HLuw. /niba … niba…/ is used to encode an alternative interrogative complementizer ‘whether … or …’. Furthermore, Luw. /=ba/ is an adversative and disjunctive particle; cf. adversative *=pe in Lat. is-pe ‘he (himself) however’. The overlap of adversative and disjunctive meaning is documented by East Tocharian pat ‘or, however’ (from PIE *po-).
Morphological and syntactic parallels (Hackstein 2016a:179f.) include Russian ne to … ne to … ‘whether … or ….’, and East Tocharian mā te … mā te …, e.g., mā te nātäk cam brā[maṃ] epe mā te was entsaträ ‘Whether the lord keeps this Brahmin or whether he keeps us’ (A342b2-4).
HLuw. /naba/ and /niba/ exhibits a functional shift from negation to disjunctive conjunction (Hawkins & Morpurgo Davies 2010a:110–127; Morpurgo Davies 2011a:209–211); cf. PIE *ku̯id=pe?. This functional shift (interrogative-negation-to-disjunction shift) is well attested across the various Indo-European branches; see Hackstein 2016a:176–180.
[O.H.]