*Hán-H-o-

The Lycian lexeme for ‘grandmother’ is an a-stem (< *-eh₂). The Hittite word ḫanna-, a synchronic a-stem, can go back to either a thematic o-stem, a stem in *-eh₂ or a root stem, which would have been transformed into an a-stem in analogy with other family terms such as ḫuḫḫa- ‘grandfather’, anna- ‘mother’ and atta- ‘father’. If one assumes that the o-stem is more likely to be original, then the stem in *-eh₂ found in Lycian might be analogical to xuga- ‘grandfather’. As Puhvel (HED H:85) points out, both matrimonial terms, ḫanna- and anna-, belong to the same group of words. Thus, it is probable that the term for ‘grandmother’, i.e. *Han-a-, geminated the nasal, resulting in *Hann-a-, in analogy to the term for ‘mother’. On the other hand, one could reconstruct a thematic o-stem derived from a stem in *-eh₂, *h₂en-h­₂-o- (similarly Kloekhorst 2008c:285). This process would also account for the gemination of the -nn-, as advanced by Kloekhorst 2008c:285, but for the position of the accent, i.e. *h₂én-h­₂-o-, see Melchert 1994a:79f. The nasal sequence in the Lycian root xñn- is not original, but it is due to its position after a consonant. This cluster resulted after the syncope of the root vowel, cf. Melchert 1994a:295f. Because of the syncope, one must assume an accent shift in Lycian, which would be secondary. Based on the evidence discussed in Part II Proto-Indo-European Etymology, the Proto-Anatolian lexeme for grandmother is reconstructed as follows: *Hán-H-o-.

A single attestation of the Hittite lexeme ḫanna- in the funerary ritual šalliš waštaiš (CTH 450) shows a plene spelling of the root vowel, i.e. dat. pl. [ḫu-]˹uḫ˺-ha-aš ḫa-a-an-n[a-aš] ‘to the grandfathers and grandmothers’ (HFAC 14 obv. 4, NS). This instance is probably due to Luwian ethnicity of the scribe. Although, the Luwian lexeme for ‘grandmother’ is not yet attested, the lexeme for ‘mother’ ānn(i)- shows plene spelling of the first syllable, as opposed to Hitt. anna- ‘mother’. This is supported by the analysis of Yakubovich 2010d:343f., who shows strong Luwian influence in the sequences ‘grandfather, grandmother, father, mother’ attested within the same text. Further evidence can be gathered from the use of the lexeme nata/i- ‘arrow’ (^GIna-ti-in KUB 39.14 + KBo 40.311 iii 10) as opposed to proper Hitt. nata- ‘arrow’ (Yakubovich 2010d:343f.) and ḫala(n)tuwa- ‘palace’ (e.g. ḫa-la-an-tu-wa KUB 30.25+ rev. 12, ḫa-la-an-tu-u-wa KUB 39.5+ rev. 350, ḫa-la-tu-u-wa KUB 30.23+ iii 44) as opposed to proper Hitt. ḫalentuwa- ‘palace’.

[D.S.] [E.R.]

Anatolian (PAnat. *HanHo-), Armenian (PArm. *Hano-) and Latin (PIt. *ano- → Lat. anu-) forms guarantee the reconstruction of PIE *h₂én‑h₂‑o- denoting females.The West Germanic n-stem, reflected in OHG ana ‘grandmother; old woman’, represents an original eh₂-stem which itself possibly is a hypercharacterized feminine o-stem PGerm. *ana-.

Two slightly different meanings, viz. ‘old woman’ (e.g., Lat., WGerm.) and ‘grandmother’ (e.g., Anat., Arm., WGerm.), are attested. Both semantic developments, viz. from ‘old ([wo]man)’ to ‘grandmother/-father’ and vice versa, are possible.Cf., on the one hand, Lith. senuõlis m. ‘grandfather’, senõlė f. ‘grandmother’, both via ‘the old one’ from Lith. adj. sẽnas ‘old’ (ALEW II:901f.). On the other hand, colloquial terms for ‘grandmother’ and ‘grandfather’ are used (partly pejoratively) for elder people.However, semantic narrowing of ‘old person’ to ‘elder relative’ is more likely.In most of the IE languages the meaning was connotated positively, although sometimes – mainly in later linguistic strata – meanings such as ‘sorceress’ (e.g., Latin, OHG) are encountered. Gamkrelidze-Ivanov 1995a I:668 reconstruct ‘husband’s/father’s mother’.

In the etymological literature it is frequently assumed that not only PIE *an(n)a- [lemma link], but also PIE *h₂én‑h₂‑o- ‘grandmother/old woman’ is a nursery word.Cf., e.g., IEW:36, Ernout-Meillet 1985a:37, WH I:55, EDLIL:144, DGE II s.v. ἄννις, PKEŽ 2013:76f., Trautmann 1923a:9, Buck 1949a:109, 110, EIEC:238.However, at least PIE *h₂én‑h₂‑o- has to be viewed as a morphologically (cf. below) and semantically analyzable formation.

PIE *h₂én‑h₂‑o- is best analyzed as a thematic adjective of appurtenance of *h₂(e)n‑eh₂- ‘the elder people/generation’. Hence *h₂én‑h₂‑o- originally denoted a person ‘belonging to the elder generation, old woman’. In social structures consisting of several nuclear families this interpretation accounts for the relative homogeneity of the female members as opposed to the more individual male leaders.On the social structure of the PIE family, cf. EIEC:193. It is less likely to interpret the lemma word as directly derived from *h₂en- ‘old’, hence *h₂en‑o- ‘the old one’ According to the evidence given above, a *h₂en‑(h₂)‑o- is exclusively female and not merely an old person (m./f.). For the phonological and morphological reconstruction, cf. below.

Because in the Core IE branches the sequence *‑V́nHo- regularly loses the laryngeal without a trace,Cf. for potential pretonic post-nasal laryngeal hardening (*‑nh_2/3V́- > *‑ngV́-) Steer (fthc. 2017).it is not possible to determine whether the Core IE proto-form was *h₂én‑o- or *h₂én‑h₂‑o-. Only the Anatolian geminate (cf. Hitt. ḫanna-) points to the latter form. However, the geminate could also have arisen by analogy or expressivity, which weakens the evidence (cf. PAnat. Reconstruction above). The considerations regarding the semantic development (cf. above) as well as the phonological explanation for the Anatolian geminate suggest that PIE *h₂én‑h₂‑o- has to be reconstructed.

Morphologically, PIE *h₂én‑h₂‑o- is a substantivized adjective of appurtenance (‘the one belonging to the elder generation’), derived from a collective *h₂(e)n‑eh₂- ‘the elder generation’.

Arm. han, gen. sg. hanoy f. ‘grandmother’Cf. also Arm. loc. sg. i hanwoy, instr. pl. hanwovk‘ f. ‘grandmother’ (u̯o-stem byforms), cf. Martirosyan 2010:388 s.v. han.

Olsen 1999a:20, Martirosyan 2010:388

(?)Messap. anā- ‘(?)mother, mistress (epithet of Aphrodite)’According to MLM II:12, 13, Messap. anā- is a divine name. Attested are acc. sg. anan (MLM 2 Mu, 600–450 BCE) and dat. sg. ana (MLM 22 Cae, 3^rd c. BCE, 1 Ga, 3^rd c. BCE, (?)42 Ur, date unknown).
In view of the assumed meaning ‘mother’ it cannot be ruled out, that Messap. ana- rather belongs to PIE *an(n)a- [lemma link]. Both attributions remain possible, though.

MLM II:12, 13

(?)Lat. anus u‑stemCf. Meiser 1998:145, Weiss 2011:251 on the OLat. gen. sg. anuis (Ter.Heaut.287).f. ‘old woman; old (adj.)’ (Plaut., Enn.+)The Latin u-stem is considered to be an analogical formation, as, e.g., Ernout-Meillet 1985a:37, Olsen 1999a:20 and de Vaan (EDLIL:45) propose on the basis of other Latin (female) kinship terms found in the 4^th declension, cf. Lat. socrus, -ūs f. ‘mother-in-law’ (Naev., Enn.+), nurus, -ūs f. ‘daughter-in-law, young woman’ (Ter.+). Kinship terms are especially susceptible to mutual analogical reshaping, see Winter 1969a.

OLD:144 s.v. anus², EDLIL:45, WH I:55 s.v. 2. anus, Ernout-Meillet 1985a:37

OHG ana n-stem f. ‘grandmother, (Lat.) avia (gl., 9^th c.+), ancestress, old woman’

OHG ano n-stem m. ‘grandfather, ancestor (Otfried, 9^th c.+)’Cf. also MDutch aen m. ‘grandfather’.

EWAhd I:215–217, ChWdW9:137, 139, AhdWb I:410, Orel 2003:20, Kluge/Seebold 2011a:22, Falk-Torp 1960a I:28f. s.v. anerCf. also OHG altana n-stem f. ‘great-grandmother, (Lat.) proavia’ (gl., hapax), OHG altano n-stem m. ‘great-grandfather, (Lat.) proavus’ (gl., Otfried [hapax], 9th c.), cf. ChWdW9:139, EWAhd I:173. Furthermore, cf. (?)OHG hefi(h)anna f. ‘delivery nurse, (Lat.) obstetrix’ (gl., hapax, 9th c.), cf. ChWdW9:137, EWAhd IV:880 s.v. hevianna.
In all Germanic subbranches there are various personal names derived from PGerm. *anan-, cf., e.g., the male personal names Goth. Anila, Anala, ON Áli, Óli, Óláfr, OE Onela, OHG Analo etc., cf. EWAhd I:216, de Vries 1977a:6, 418. Some or all of the adduced names are diminutives like MHG enel ‘grandfather; grandson’ (IEW:36f.).
The masculine form is also the basis of the diminutive formation OHG eninchil ‘grandson’ < *‘little ancestor’, cf. EWAhd I:216.

Young Avestan

(?)Av. *niiā- < *h₂n‑(i)i̯eh₂- in YAv. niiākā- f. ‘grandmother’ (V.12.9, 11), niiāka- m. ‘grandfather’Cf. also OPers. nii̯āka- ‘grandfather’. Differently, pre-PBalt. *h₂en‑i(e)h₂- and pre-PArm. *h₂en‑ih₂- reflect full-grade of the root, cf. on those below.
Szemerényi 1950a:235f. traces OPers. nii̯āka- back to *ni‑āu̯aka- which ultimately would belong to PAnat. *Héu̯h‑eh₂‑/*HuH‑éh₂- ‘grandfather’. This view is called into question by Andrés-Toledo 2009a:398.(EIEC:238)

Bartholomae 1979a:1094

Greek

Boiot. ἀννίς f. ‘(Gk.) μητρός ἢ πατρός μήτηρ, grandmother’ (Hsch. α 5283 (Latte), acc. sg. αννιν in IG VII:3380, 2^nd c. BCE) ?< *h₂en‑(n)i(h₂/d)-

Thess. acc. sg. f. αννων ‘(?)grandmother’ (IG IX,2:877, Roman period) < *h₂en‑(n)‑oi-?

GEW I:112, DÉLG 2009:91, EDG:107, LSJ:145, DGE II s.vv. ἄννις, ἄννως

Armenian

Arm. hani f. ‘grandmother’ ?< *h₂en‑ih₂-If this reconstruction is valid, Lith. aný‑ta f. ‘husband’s mother’ (cf. below and OPr. ane) can be directly compared.

Martirosyan 2010:288.

Arm. aner, dial. hanεr ‘wife’s father (Bible+); wife’s brother’ < *h₂en‑(t)ero- ‘someone like a grandfather (Martirosyan 2010:82); opposed to the grandmother (*h₂en-)’Olsen 1999a:32f. reconstructs *h₂en(‑En)‑teró-.
Since frequently the derivatives of *h₂en- designate females, it is likely, that Arm. aner as a male relative is derived from a word meaning ‘grandmother, old woman’, not from ‘grandfather’. On the contrastive suffixes *‑ero- and *‑tero- cf., e.g., LIPP I:201 and LIPP I:211f. resp.

Martirosyan 2010:81f., Olsen 1999a:32f.

Celtic

(?)OIr. Ana ‘mother of the gods’

eDIL s.v. Ane

Baltic

OPr. ane f. ‘(MHG) altmuter, grandmother’ (EV 172) < PBalt. *ani̯ā- < *h₂en‑i̯eh₂-The alternation between *h₂en‑i̯eh₂- (> OPr. ane) and *h₂en‑ih₂- (> EastBalt. *anī- → Lith. anýta, cf. also Arm. hani above) could reflect a paradigm split of *h₂én‑ih₂-, *h₂n̥‑i̯éh₂- (→ *h₂en‑i̯eh₂-).

PKEŽ 2013:76f., Smoczyński 2000a:203Smoczyński (Smoczyński 2000a:203, Smoczyński 2007a:15f.) suggests a loan from MHG ane f. ‘grandmother, ancestress’.

Lith. anýta f. ‘husband’s mother’ ← *anė < PBalt. *anī- < *h₂en‑ih₂-

Trautmann 1923a:9, Smoczyński 2007a:15f., Fraenkel 1962a:10f., Derksen 2014a:55, ALEW I:41f., ALEW 2.0 s.v. anýta

Slavic

ORuss. vŭnukŭ, Cze. vnuk, SCr. ùnuk (m.) ‘grandson’ < PSlav. *ononko- ?< *h₂n̥‑h₂‑en‑ko- ‘little grandfather’It is also possible, however, to reconstruct a reduplicated formation *h₂en‑h₂en‑ko-, cf. EIEC:238. For the semantic shift cf. OHG eniklīn (m.) ‘grandson’, which is a diminutive of OHG ano ‘grandfather, ancestor’, cf. EWAhd II:1075.

Derksen 2008:530f., Vasmer 1953a I:211

[A.O.] [C.S.]