*h2eb-
‘running water, river’
- Language
- Proto-Anatolian
- Class
- Base
- Grammar
- subst
- ID
- 352
1. Individual Anatolian Languages
1.1 Hieroglyphic Luwian
/hab(i)-/, FLUMEN-pi-
‘river’1.1.1Transmission
This lexeme is only attested in the Iron Age period. Its earliest attestation are found in IZGIN 1, §8, dated to 11th or 10th century BCE, and in MARAŞ 8, §8, dated to the reign of Larama (1000-950 BCE).
[D.S.]
1.1.2Forms
| acc. sg. | /habin/ | FLUMEN-pi-i-na | KARKAMIŠ A15b, §9 (Karkamiš) |
| FLUMEN-pi-na | KARKAMIŠ A15b, §7 (Karkamiš), KARKAMIŠ A15b, §8 (Karkamiš) | ||
| FLUMEN-na | IZGIN 1, §8 (Malatya) | ||
| FLUMEN(-)*311-pi-na | TELL AHMAR fragments 1-10, 6 | ||
| dat. sg. | /habi/ | FLUMEN-pi-i | TELL AHMAR 5, §9 (Tell Ahmar) |
| FLUMEN-pi | MARAŞ 8, §8 (Maraş) |
1.1.3Graphic Features
1.1.3.1Phonological Intrepretation
The phonological interpretation of this lexeme is based on its cognate in Cuneiform Luwian.
[D.S.]
1.1.4Meaning
The semantic assignment of this lexeme was facilitated by the logogram, which pictures flowing water (Gelb 1931a:14, 26; cf. Laroche 1960c:116f.). Laroche 1973a:181-184 connects HLuw. FLUMEN-pi- c. with CLuw. ḫāp(i)- c. ‘river’ and reads the word behind the HLuw. logogram correspondingly. The lexeme /hab(i)-/ c. refers to a topographic feature. Whenever it is said that its course can be modified (“PES2”)hi-nu- ‘to cause to flow’, KARKAMIŠ A15b), it refers to canals going through a city. In one case, it is the object of the action of settling (IZGIN 1, §8) and, thus, refers to the inhabitants nearby the river (Hawkins 2000a:317).
[D.S.]
1.1.5Compounds
- FLUMEN.DOMINS-ia-
1.1.6Derivatives
1.2 Cuneiform Luwian
ḫāpa/i-
‘river’1.2.1Transmission
Attested in Luwian rituals both in MS and NS.
[Zs.S.]
1.2.2Forms
| nom. sg. c. | ⸢ḫa⸣-a-pí-iš | KUB 35.108 rev. iv? 22’ | (CTH 765, MS) |
Sumerographic spelling:
| nom. sg. c. | ⸢íd⸣-iš | KUB 35.89, 18’ | (CTH 766, MS) |
| dat./loc. sg. | ⸢íd⸣-i | KUB 35.114, 8’ | (CTH 770, NS) |
| íd-⸢i⸣[ | KUB 35.89, 16’ | (CTH 766, MS) | |
| íd | KUB.35.89, 16’ | (CTH 766, MS) | |
| acc. pl. | ídḪI.A-in-za | KUB 35.107 rev. iii 15 | (CTH 764, MS) |
| dat./loc. pl. | ídMEŠ-an-za | KUB 35.45 ii 6 | (CTH 760, NS) |
| ⸢íd⸣[(MEŠ-an-za)] | KUB 35.49 i 11’ | (CTH 761, MS) | |
| abl./instr. | íd-⸢ti⸣ | KUB 35.54 rev. iii 17 | (CTH 758, MS) |
| ? | ⸢íd⸣ -⸢in?⸣(-) | KBo 29.46, 1’ | (CTH 770, unknown) |
Attestations following CLL:54, checked against the photos of the Mainzer Fotoarchiv
[A.B.], [Zs.S.]
1.2.3Meaning ‘river’
Communis opinio since Laroche 1968d:89 (yet without meaning in Laroche 1959a:41, cf. HED H:115; CLL:54; Tischler 2008a:44, Tischler 2016a:68; EDHIL:294; but without meaning in Ünal 2016a:956). The meaning is assured by the phrase ḫāpiš Mālaš ‘the river Euphrates’ in KUB 35.108, 22 (and supported by the general context, since the previous line mentions mountains).
[Zs.S.]
1.2.4Stem
The currently known attestations allow an a-stem with i-mutation, a consonant stem with i-mutation, and an i-stem (only the last one is assumed by Laroche 1968d:89 and HED H:115).
[Zs.S.]
1.2.5Derivatives
1.3 Carian
*kbo-
‘river (?)’1.3.1Transmission
Reconstructed from the Carian name of the city of Keramos attested (abbreviated?) on the local coins (c. 420-400 BCE or c. 400 BCE, Konuk 2000a, Konuk 2007a:479, resp.) as well as from its presumed ethnic adjective on a false-door stele from Memphis (E.Me 24, last third of the 6th c. BCE, Kammerzell 1993a:169). It may be also present in Kbokt, a personal name attested in Thebes (a graffito in the tomb of Montuemhat, during the rule of Psammetichus I, i.e. around the mid-7th c. BCE).
[Zs.S.]
1.3.2Forms
On coins:
| kbo | M22, M23 | Keramos |
In ethnic adjectives from the toponym:
| nom. sg. | kbos | E.Me 24 |
In personal names:
| gen. sg. | Kboktś | E.Th 2 |
For the attestations see Adiego 2007a:372 and Konuk 2007a:479, checked against the photos of the publications (except the inscription from Thebes, where no photos are available).
[Zs.S.]
1.3.3Meaning ‘river (?)’
The toponym Kbo ‘Keramos’ may be a regular Carian reflex of the Luwic word for ‘river’, as assumed by many scholars (Adiego 2007a:334, 372: “possible”; followed by Simon 2011b:540; Brosch 2016a:7). Schürr 2001c:64 supports this with the observation that Keramos is situated at the delta of the only river that breaks through the Kiran Dağ along the Ceramic Gulf. Although this suggestion is possible both semantically (cf. the toponym Fiume) and phonologically (while the exact Carian reflexes of the laryngeals are debated, there is a general agreement that Carian initial continues initial <ḫ>, Schürr 2001d:97–99; Adiego 2007a:260; Kloekhorst 2008a:145 [implicitly]; Simon 2011b:540–542; Brosch 2016a:10), it cannot be verified.
Following Schürr 2003c:69 n. 1 Adiego 2007a:278, 372 (with question mark) suggests that kbos is an ethnic adjective with the regular -s- suffix from Kbo (according to Konuk 2000a:162f. this connection is too early). This suggestion fits if Kbo is not an abbreviated form of the toponym, as pointed out by Adiego as well (which is unclear at the moment).
The personal name Kboktś was also added by Adiego 1994b:252, Adiego 2007a:372 to the root kbo-, who analyzed this name as a compound of kbo- and kt, which, however, depends on the existence of kt, for which see s.v. (in view of Konuk 2000a:162f. this is a premature connection). Adiego’s further cautious connection, the personal name Taqbo- (E.xx 6), is contradicted by the different consonants.
[Zs.S.]
1.3.4Derivatives
1.4 Lydian
kof(i)-
‘water’1.4.1Transmission
This noun occurs in the curse formula of the Aramaic and Lydian bilingual inscription that was found at the cemetery of Sardis and is datable to the early 4th century BCE.
[I.Y.]
1.4.2Forms
| dat. pl. | kofuλ | LW 1.9 (Sardis) |
1.4.3Meaning
The correspondence between the Aramaic phrase ṭyn w-myn ‘soil and water’ and its Lydian counterpart kλidaλ kofuλ=k was already established in (Littmann 1916a:36) even though Littmann did not commit himself to a specific correspondence between the elements of the Lydian and Aramaic formulae. But all the subsequent treatments of this pair, including Gusmani 1964a:153, opted for the match in word order between the two languages within this collocation and assigned the meaning ‘water’ to kofuλ. Such a solution derives additional support from the etymology of kof(i)-, cf. immediately below.
[I.Y.]
1.4.4Stem
The form kofuλ was traditionally assigned to the u-stem, see e.g. Gusmani 1964a:153. In synchronic terms, this solution is the most straightforward, but not the only possible one. Given that the personal pronoun p(i)- ‘he, she, it’ has the alternating dative forms pλ and puλ, one can hypothesize that kof(i)- likewise had the dative form kofuλ. Presumably, the insertion of -u- after a labial consonant and before the dative ending -λ was an optional feature of semi-vocalic stems. In this particular case, the preference for the variant kofuλ may have been triggered by its juxtaposition with another disyllabic form kλidaλ within the same collocation. The main argument in favour of the stem kof(i)- as opposed to kofu- is the possibility of comparing it with Hitt. ḫapa- and the related Anatolian words for ‘river’, e.g. CLuw. ḫāp(i)-. In addition, the semi-vocalic stems are generally more frequent than u-stems in Lydian.
[I.Y.]
1.5 Hittite
ḫapa-
‘river’1.5.1Literature / Comments
For the attestations, see HW (2nd ed.) H:197-205, HED H:114f.
2. Reconstruction
2.1Reconstruction ‘running water, river’
A formation without Hoffman’s suffix must be reconstructed in addition to *Habon- for the Proto-Anatolian word for ´river´. Since n-stems were productive in Anatolian, Hitt. ḫapa- and CLuw. ḫāp(i)- cannot be explained convincingly as secondary thematic stems originating from *Habon- (pace McCone 1992a:109, Stüber 1998a:99) through an elimination of the n-stem in the oblique cases based on the nominative singular *Habo-. Therefore, it is preferable from an inner-Anatolian perspective that Hittite ḫapa- (c.) reflects a Proto-Anatolian root noun *Hab- (< *h2eb-), which was secondarily thematized (Rieken 1999a:19).
Kloekhorst 2014b:460 argues that Hittite ḫapa- is oxytone on the premise that the logographic spelling of the dat./loc. sg. forms reflects accentuation of the ending. On the other hand, Luwian ḫāp(i)- appears to be accented on the first syllable because of the plene spelling. If the Hittite lexeme is oxytone, then it is probably due to the thematization and it is thus secondary.
A collective in *-eh2-, i.e. *HabaH-, should also be posited based on Proto-Luwic *Habā́d- (c.) ‘river-valley’ and Proto-Anatolian *h2eb-eh2i̯e/o- ‘to water’.
The Hittite attestations ÍD-ni (KUB 17.8 iv 23, NS) and ÍD-an-na (KUB 35.14 iii 14´, MS) cannot be attributed to the same stem as ḫapa- (c.). If ÍD-ni contains the same root as the one of the latter, one could assume that ÍD-ni reflects the same stem formation as the one continued in Palaic ḫāpna-. However, the geminate -nn- in ÍD-an-na does not support this view. On the other hand, they could reflect instead a secondary derivative such as the one found in Luw. ḫāpinn(i)- c. ´stream, creek´; cf. the discussion on Hitt. ḫāppan- under Pal. ḫāpna-.
Cf. HEG A-K:159f., HED H:114f., Kloekhorst 2006b:132 n. 5, EDHIL:294f.
[D.S.]
3. Proto-Indo-European
3.1Semantic Reconstruction
PIE *h2óp-/*h2ep- f. means ‘(running, living) water, river’ (NIL:311, EIEC:636, Watkins 2011a:5). Unlike PIE *u̯ód-r̥/*u̯ed-(e)n- n. ‘water’ (inanimate), it is an animate word for ‘water’. The opposition animate vs. inanimate can also be observed with words for ‘fire’, cf. PIE *h1óg(u̯)ni-/*h1eg(u̯)ni- (vel sim.) m. ‘fire’ (animate; Ved. agní-, Lat. ignis, etc.) vs. *péh2u̯r̥/*ph2(u)u̯én- n. ‘fire’ (inanimate; Hitt. paḫḫur, Gk. πῦρ, etc.). According to Watkins 2011a:5, “the neuter-gender terms referred to water or fire as a substance, while the animate (gendered) terms were used when water or fire was conceived of as an active, living force.” For other PIE words for ‘water’ (e.g., *h2eku̯eh2- or *u̯ó/éh1r-), cf. EIEC:636 with lit. PIE *h2ábon- < *h2ép-h3on-, which was derived from PIE *h2óp-/*h2ep-, means ‘river, stream’ (< *‘having running water’); for the derivational details, cf. below sub Morphological Reconstruction.
3.2Morphological Reconstruction
A. Inflection
The reconstructed o/e-ablaut in the root syllable of *h2óp-/*h2ep- is reflected by the ā/a-ablaut of the Indo-Iranian forms: PIIr. nom. sg. *Hā́p-s (YAv. āfš), acc. sg. *Hā́p-am (YAv. āpǝm), instr. sg. *Hap-ā́ (Ved. apā́, YAv. apā-ca), dat. sg. *Hap-ái̯ (YAv. ape), gen./abl. sg. *Hap-ás (Ved. apáḥ, YAv. ā̆pō), etc.; nom. pl. *Hā́p-as (Ved. ā́paḥ), gen. pl. *Hap-ā́m (Ved. apā́m, YAv. apąm), etc. < PIE nom. sg. *h2ṓp-s, acc. sg. *h2óp-m̥ (with PIE *o > PIIr. *ā in open syllables by Brugmann’s Law), instr. sg. *h2ep-éh1, dat. sg. *h2ep-éi̯, gen./abl. sg. *h2ep-és, etc.; nom. pl. *h2óp-es (with PIE *o > PIIr. *ā by Brugmann’s Law), gen. pl. *h2ep-ō̆́m, etc.
The amphikinetic derivative PIE *h2ép-h3on- > *h2ábon- ‘having running water; river, stream’ had the following forms (cf. Hamp 1972a, McCone 1992a:110, Stüber 1998a:100, NIL:311–313): nom. sg. *h2áb-ō̃ (< *h2ép-h3ōn), acc. sg. *h2áb-on-m̥ (< *h2ép-h3on-m̥), dat. sg. *h2(a)b-n-éi̯ (< *h2[e]p-h3n-éi̯), instr. sg. *h2(a)b-n-éh1 (< *h2[e]p-h3n-éh1), gen./abl. sg. *h2(a)b-n-és (< *h2[e]p-h3n-és), loc. sg. *h2(a)b-ón(-i) (< *h2[e]p-h3én[-i]); nom. pl. *h2áb-on-es (< *h2ép-h3on-es), gen. pl. *h2(a)b-n-óm (< *h2[e]p-h3n-óm), loc. pl. *h2(a)b-n̥-sú (< *h2[e]p-h3n̥-sú).
B. Derivational Morphology
On the basis of the weak stem of PIE *h2óp-/*h2ep- ‘running water’, a possessive derivative *h2ép-h3on- (f.) ‘having running water; river, stream’ was formed by adding the “Hoffmann suffix” *-h3on-. For the reconstruction of this suffix with *h3, cf. Schaffner 2015a:147f. n. 7. For the voicing *p > *b before *h3 in *h2ép-h3on- > *h2ábon-, cf. the reduplicated present PIE *pí-ph3-e/o- ‘to drink’ (root *peh3- ‘to drink’) > *píbe/o- (Ved. píbati, OIr. ibid, etc.)Cf. Hamp 1972a:36, Mayrhofer 1986:143, McCone 1992a:108, Stüber 1998a:99. Hamp’s proposal was disputed by Watkins 1972a:80f., because forms containing voiced *-b- are allegedly also found outside the n-stem (cf. also Mayrhofer 1986:144). However, as shown by McCone 1992a:108f. and Stüber 1998a:98f., all those forms can be traced back to an n-stem formation..
In PAnat., the PIE paradigm nom. sg. *h2áb-ō͂, gen. sg. *h2áb-n-os, dat. sg. *h2áb-n-ei̯, etc. first gave nom. sg. *Hábō̆s (< *h2ábō͂+s; cf. Hitt. ḫāraš c. ‘eagle’ < PAnat. *Hárō̆s < PIE *h2/3érō͂+s), gen. *Hábnos. The nom. sg. PAnat. *Hábō̆s was then reanalyzed as the nom. sg. of a thematic stem PAnat. *Hábo-, whence the secondary thematic stem Hitt. ḫāpa- [ḫā́ba-]; cf. Hitt. ḫāra- ‘eagle’ (e.g., acc. sg. ḫāran) : nom. sg. ḫāraš, with original stem ḫāran- < PIE *h2/3éron-, cf. PGerm. *aran- ‘eagle’ (cf. McCone 1992a:109, Stüber 1998a:99). The ideographically spelt dat. sg. ÍD-ni, however, seems to be a relic form of the old n-stem paradigm. For further potential n-stem forms, cf. EDHIL:295. A secondary thematic stem PAnat. *Hábo- was also the basis of CLuw. ḫāpa/i- c. ‘river’, HLuw. /hab(i)-/ c. ‘id.’. In the prehistory of Palaic, a new thematic stem form ḫābna- < PAnat. *Hábno- was formed on the basis of the gen. sg. PAnat. *Hábnos (< PIE *h2áb-n-os), which was interpreted as the gen. sg. of a no-stem PAnat. *Hábno- (cf. Melchert 1989a:98 n. 4, Stüber 1998a:99).
As an alternative, Watkins 1972a:81 n. 3 (followed, e.g., by HED III:115, EDHIL:295) opted for a root *h2eb(h)- ‘river’ different from *h2ep- ‘water’ to explain the forms attested in Anatolian, Latin, and Celtic. He concludes that there is no reason to invoke the possessive suffix *-h3on-. There is, however, no clear evidence in favor of a PIE root *h2ebh- with bh. A root *h2eb- with b, on the other hand, could have developed easily as a variant of the root *h2ep-, namely in the bh-case forms of the root noun PIE *h2óp-/*h2ep- – i.e., *h2ep-bh° > *h2eb-bh°. Cf., e.g., instr. pl. *h2ep-bhis > *h2eb-bhis > OPers. abiš (simplified from *abbiš), Ved. adbhíḥ (dissimilated from *abbhíḥ). The root variant *h2eb- could have been generalized in Proto-Anatolian, with PIE *h2óp-/*h2ep- ‘running water’ (root noun) → PAnat. *Hab- → *Hab-o- > Hitt. ḫāpa-, CLuw. ḫāpa/i-, HLuw. /hab(i)-/. From an inner-Anatolian perspective, it might be more plausible to derive the Hittite and Luwian words for ‘river’ from a thematized root noun than from a secondary thematic stem based on an n-stem noun (see PAnat. Reconstruction). Nevertheless, beside a thematic stem PAnat. *Hab-o- based on the root noun *Hab- ← PIE *h2óp-/*h2ep-, there must also have been an n-stem noun PAnat. *Hab(o)n-, as indicated by dat. sg. Hitt. ÍD-ni and, with thematization, Pal. ḫābna- ‘river’. This n-stem, which is also presupposed by Lat. amn(i)- ‘river’ and PCelt. *abon- ‘id.’, goes back to PIE *h2eb(o)n-, which was either a primary n-stem derived from the secondary root *h2eb- or regularly developed from a possessive formation *h2ép-h3(o)n-, as argued above.
Finally, a loc. sg. *h2ep-er(-i) > *h2ap-er(-i) ‘at the water’ of PIE *h2óp-/*h2ep- ‘running water’ probably served as derivational base of two vr̥ddhi adjectives, viz. *h2ā́p-er-o- ‘being at the water’ > PGerm. *ōfera- ‘shore’ (OE ōfer, MHG uover, ModHG Ufer) and *h2ā́p-er-i̯-o- ‘id.’ > Gk. ἤπειρος, Dor. ἄπειρος (ᾱ) f. ‘land(mass), mainland, coast’; cf. Zehnder 2012a:469–471). Differently Klingenschmitt 2005a:542, who reconstructs a preform *é-h2p-ero- “der am Wasser gelegene Teil des Landes” > PGerm. *ōfera- m./n. alongside *é-h2p-eri̯o- ‘id.’ > PGk. *āperi̯o-.
[S.S.], [T.S.]
4. Transmission
Ved. áp- f. ‘water’ (RV [instr. sg. apā́, gen./abl. sg. apáḥ; nom. pl. ā́paḥ, acc. pl. apáḥ, loc. pl. apsú, etc.] +); apsu-kṣít- adj. ‘living in the waters’ (RV), ap-túr- adj. ‘crossing the waters’ (RV+), ap-tū́rya- n. ‘crossing of the waters’ (RV 3.12.8; 3.51.9); dvīpá- n./m. ‘island, river island, sandbank in a river, holm’ (RV [1.169.3; 8.20.4] +); etc. – Av. ap-, OPers. ap- f. ‘water’ (cf. nom. sg. YAv. āfš, acc. sg. YAv. āpǝm, apǝm-ca, instr. sg. YAv. apa, apā-ca, dat. sg. YAv. ape, apaē-cit̰, abl. sg. YAv. apat̰, gen. sg. YAv. ā̆pō, loc. sg. OPers. api, apiy-ā, api-šim; nom./acc. du. YAv. āpa; nom. pl. YAv. āpō, acc. pl. OAv. apō, YAv. apō, OAv. apas-cā, YAv. apas-cā, YAv. also āpō [V 6.2], instr. pl. OPers. abiš, dat. pl. YAv. aiβiiō, gen. pl. YAv. apąm)Cf., e.g., EWAia I:81f. s.v. áp-, 769 s.v. dvīpá-, ALF:136f., NIL:311f., 313f. nn. 4–9.
OIr. ab, oub, aub f. ‘river’ (n-stem), MW afon ‘id.’, MBret. auon ‘id.’, OCorn. auon ‘id.’; OIr. abac m. ‘beaver’, MW afanc ‘beaver; water demon’, Bret. avank ‘water monster’Cf., e.g., McCone 1992a:100ff., Stüber 1998a:97–101, Schaffner 2005a:263f., NIL:311f., 315 nn. 19, 21, 317 n. 34, EDPC:23f.
OE ōfer, MHG uover n. ‘shore’, ModHG Ufer n. ‘bank, shore’Cf., e.g., EDPG:394 and see below sub Morphological Reconstruction.
OPr. ape f. ‘river’, api-sorx ‘kingfisher’, Lith. ùpė f. ‘stream, river’, Latv. upe f. ‘river, brook’Cf., e.g., NIL:311f., 314f. nn. 10–13, Larsson 2010a:15 n. 8, 22, 24, 39f., EDBIL:483, ALEW II:1155f.
TochA āp-, TochB āp- ‘water; river; flood’Cf., e.g., NIL:311, 315 n. 14, Adams 2013a:46f.
Alb. Gheg a͂m/ë, -a f. ‘riverbed, stretch of water, spring, well’, Central/South Gheg (h)ãm, -i m. ‘well, spring’Cf., e.g., Demiraj 1997a:75f., NIL:311, 315 n. 20.