(Selected) Topics in Anatolian Historical Grammar

Transcription

Universität zu Köln
Institut für Linguistik
Abteilung Historisch-Vergleichende Sprachwissenschaft
WS 2012/13
Materialien zum Blockseminar
(Selected) Topics in Anatolian Historical Grammar
von
Prof. Dr. H. Craig Melchert (UCLA)
Eine Übersicht über die einzelnen
Aufsätze finden Sie über die
Lesezeichenfunktion von Adobe Acrobat.
H. Craig Melchert
30th ECIEC
Harvard University
June 8, 2011
melchert@humnet.ucla.edu
[revised November 2012]
Hittite Auslautsgesetze Revisited
I. Absolute Final Vowels
A. Short Vowels
1. Unaccented *-i: generally preserved after both obstruents and sonorants.
a. In present active verb endings -mi, -si, -weni, -teni and adverbial kuwapi ‘where;
when’ < *kwó-bhi; also in D-LSg kēti ‘this’ < *ké-dhi (Szemerényi 1956: 63 &
Georgiev 1971: 65) and contra Melchert (1994: 183) also in indefinite -kki < *-ki
(NB distribution kuiški, kuinki, kuitki, kuedanikkki with particle -kki only when last
vowel of base is -i-!). Compare Latin ci-trā ‘on this side’, Lith šís ‘this’, etc.
b. Regularly lost in assibilated *-ts < *-ti (e.g. ḫanz(a) ‘in front’ < *h2énti = Latin ante
etc. as per Jasanoff 1976: 126), but preserved before -ya ‘also, and’ through late NH
(Abl -az but -azzi=ya) and also analogically restored in P3Sg and P3Pl active verbal
endings -zzi and -nzi (but NB rare regular outcome in ēšz(a) ‘is’ or išḫiyanz(a) ‘they
bind’; see GrHL §11.8). Also preserved in infinitives in -wanzi < Abl-Inst *-wenti,
probably due to close nexus with following finite verb (so also Kloekhorst 2008:
957, but analogical addition of -i after finite forms unlikely). Specially conditioned
apocope in Imv2Sg īt ‘go!’ < *í-dhi and -nut < *-nú-dhi (NB shifted accent, as per
Kimball 1999: 192), as in Latin dīc, fac etc. (see Weiss 2009: 147).
c. Indeterminate whether regularly lost in final *-ri: -i in medio-passive -ri secondary
after active present endings, so possible that Hittite šēr ‘above, over’ = CLuvian
šarri < *séri (but Hittite may equally well reflect endingless locative *sēr, while
Luvian may be trivial innovation with productive ending). Unlikely with anaptyxis <
*srí (Oettinger 1993: 155) in view of š(a)rā < *sr-ó.
2. Accented *-í: certainly preserved in monosyllabic kī < *kí; probably but not assuredly
in accented D-LSg -Cī < *-Cí (difficult to exclude entirely -Cī < DSg *-ei; see
below I.C.5).
3. Unaccented *-u: preserved in active third person imperative endings -tu and -ntu (and
spread from there to first person etc.). Also in conjunctions nu < *nu and šu <
asseverative *h1su (thus with Zimmer 1994: 165 & Dunkel 1998: 57, given clear
association of šu with preterite predicates, as per Weitenberg 1992: 327-8; notion of
šu remade from *só and comparison with OIrish se, as per Watkins 1963: 14,
extremely unlikely). Conjunctions unaccented, as per Kloekhorst (pers. comm.). For
addition of clitics (“postpositives”) to unaccented elements see Agbayani and
Golston (2010) 6-7.
2
4. Accented *-u: no certain examples.
5. Unaccented *-e: contra Melchert (1994: 183-4) clear regular result is Hittite -a.
a. Enclitic 2Sg D-A -tta < PA *-te with regular shortening of unaccented PIE *tē;
likewise with Josephson (1972: 419) Hittite -(a)šta by metanalysis < *-te = CLuvian
-tta and HLuvian /-ta/ (NB never rhotacized and never spelled with tà; see Rieken
2008a: 640-1); also in synchronic allative apēda < *-dhe parallel to kēti < *-dhi
above. Also almost certainly in indefinite -kka (kuēlka, kuēzka, kuiēška, kuiuška,
kuedaška) < *-ke, in view of OLatin -ce (conceivable but far less likely < *-ko, for
which there is no other evidence in use as a particle). See also I.A.10 below.
b. But lost either by apocope or assimilation in -kku < *-kwe (see for rare preserved sense
‘and’ Watkins 1985). NB that preserved -a in examples under 5.a could represent
[-ə], thus instead of apocope quite possible *-kwe > *-kwə > *-kwu > -kku.
6. Accented *-é: lengthened and raised to -ī, as in ImpfImv2Sg akkuškī ‘drink!’ and
azzikkī ‘eat!’ < *-ské. On adverb and prefix pē see I.C.2 below!
7. Unaccented final *-o appears as -a: certain in medio-passive endings with regular loss
of final *-r (see below III.A.3) 3Sg -tta < *-to and 3Pl -nta < *-nto; also in anda
‘in(to)’ < *éndo and geminating -a ‘also, and’ < *-h2o (*o required by Lycian -ke
‘also’).
8. Accented final *-ó appears as -ā certainly in p(a)rā < *pró (for Hittite p(a)rā as
locatival ‘before’ as well as allative ‘forth’ see p(a)rā ḫand(ant)ātar ‘providence’ as
per Puhvel 1991: 101 and duwān p(a)rā ‘long ago’ and renewed pēran parā ‘ahead
(of time)’ in Melchert 2008).
9. Unaccented final *-a results in -a in Pret1Sg -ḫḫa < *-h2e [Ha] and Pret2Sg -tta <
*-th2e [tHa].
10. Accented final *-a: no good examples.
B. Long Vowels
1. Long *ē > ē (probably shortened when unaccented). Given syntax (occurs with
indicative), prohibitive negative lē represents dissimilated *nē attested in other
Anatolian languages (perhaps starting from *nē=man ‘would that not’, as per
Oettinger 1994: 310). Derivation from Imv2Sg *leh1 (Kloekhorst 2008: 523 et al.)
leaves syntax unexplained. Pret3Pl -er < *-ēr.
2. Long *ō > a when unaccented, as in OH neuter n-stem N-APl šarāma ‘bread rations’
< *-mō (Gertz 1982: 298-9) and N-APl plurale tantum karza ‘swift’ < virtual *kértsō (Nussbaum, p.c., revising Rieken 1999: 391-2).
3. Long *ū in 2Sg subject *tū required by /u:/ in 1Sg subject ú-uk (NB spelling with ú
precludes preform in *-uh2, since latter would lead to *u-uk /o:g/; see Kloekhorst
2008: 59 on a-aš-šu-u /a:sso:/ ‘goods’ < *-uh2). But preform *tuh1 also possible (see
IV.C.2 below).
3
C. Diphthongs
1. Unaccented *-oi > -e securely in enclitic OH 3PDat -šše < *-soi.
2. Accented *-oi > -e securely in pronominal AnimNPl -e < *-oi in kē ‘these’, apē
‘those; they’ and also NeutN-APl kē and apē, as per Jasanoff 2008: 145-8. Also
likely in pē ‘hin-’ < *poi, as per Eichner (1973: 78) contra Melchert (1994: 133). NB
HLuvian †pa-si-ya- ‘throw away’ does not exist! Attested pa-sa-yá-tu-' (KAYSERİ
§18) highly unlikely to spell /passya-/. Preform *poi accounts for shape of pēḫute/a‘carry, bring’ and compatible with all other verbs with pē-, including pai- ‘go’ (but
all attested forms easily derivable from *pói-h1i- with accented preverb and zerograde root, contra Kloekhorst 2008: 616-8 with unbelievable spelling interpretations
and needless complex sound changes). New suggestion by Puhvel (2011: 191-2) that
Luvo-Hittite pašku(wa)- ‘forsake, abandon’ continues *pa-sk(e)u- (cf. Goth. Afskiuban ‘idem’) not nece. Incompatible: could show variant *po-. NB still no certain
Hittite-Luvian equations. In any case, no evidence for *pe-.
3. Unaccented *-ei probably results regularly in OH P3Sg -e in war(a)šše ‘plucks,
harvests’ < virtual *wórsei (see also OH/OS mazze ‘resists’), but evidence is less
than robust.
4. Accented *-ei: no sure examples.
5. Unaccented *-ai > Hittite -e certainly in OH Pres1Sg -ḫḫe < *-h2ei [-Hai].
6. Accented *-ai: no sure examples.
7. Accented *-ōi in amphikinetic neuter nouns in -āi with shifted accent, as in ḫaštāi
‘bone; bones’ < *h2esth2ōi etc.
8. Accented *-ēi > -ē(i) in kulēi ‘fallow land’ < virtual *kwlh1-ēi < *kwélh1-o- (see
Oettinger 1995: 211-14) and in utnē ‘land’ < *ud-nēi < *wédeno- (see Oettinger
2000: 185ff.). Derivation of NeutN-APl kue < *kwēi (Melchert 2004: 140) possible,
but *kwói as in kē and apē (see above I.C.2) more likely.
II. Interior Vowels in Final Syllables
A. Short Vowels
1. Short *i appears as i whether accented or unaccented: kuiš, kuin, kuit and likewise
šalliš, šallin, šalli etc.
2. Short *u appears as u whether accented or unaccented: g(a)nut ‘by the knee’ < virtual
*gnú-t and daššuš, daššun, daššu.
3. Accented short *e > ē in Inst kēt < *kéd (Melchert-Oettinger 2009: 60) and in Pret1Pl
and Pret2Pl -wen (NB exx. such as , -ten < *-wén, *-tén (Yoshida 1997: 191-2).
4. Unaccented short *-en# > -an in supine -škewan (Yoshida 1997: 192) and local
particle -an < cliticized *en. Unaccented *-es# > -iš in nēpiš ‘sky, heaven’ <
*nébhes (perh. better example is šaudišt- ‘new-born (animal)’ < *sóm-wetes-t- *‘of
the same year’.
5. Accented short *o > ā in AnimNSg kāš, apāš and NeutN-ASg apāt. Accented *-óm#
> /-o:n/ (spelled -u-un) in AnimASg kūn, apūn (Kloekhorst 2008: 99, revising
4
Melchert 1994: 186-7). NB makes very problematic derivation of mān ‘how; as; if,
when’ < *móm and equation with Lycian mẽ ‘as; thus’! Dubious derivation < Inst
*méh2m [mām] by Dunkel (1997: 72-4).
6. Unaccented short *o > a in NeutN-ASg -an < *-om, as in pēdan ‘place’ < *pédom
(NB iugan ‘yoke’ < *yugóm must then be analogical in either vocalism or accent)
and in Pret1Sg *-an < *-om, remade as -an-un.
7. Accented *-óms# > /-o:s/ (in ku-u-uš and a-pu-u-uš), as per Kloekhorst (2008: 99),
revising Melchert (1994: 186-7). Fate of unaccented *-oms (and *- s) remains less
than certain. Kloekhorst (2008: 56-7) assumes -ú-uš is regular spelling for AnimAPl
in OH, thus /-us/, replaced by /-os/ in NH, based on later spellings with -u-uš. But
NB that neither *-oms nor *- s should ever have been spelled plene! Of 10 exx. of
-ú-uš cited by Kloekhorst, 8 (including all OS and MS) are of type ta-lu-ga-ú-uš,
ḫar-ša-ú-uš, kap-pí-ú-uš, entirely parallel to ta-lu-ga-e-eš, ap-pé-ez-zi-e-eš (cf. ḫaan-te-ez-zi-ú-uš missed by Kloekhorst), argued convincingly by Kloekhorst
(forthcoming) to reflect use of -e- to mark [y] after -Ci- sign. Thus use of ú surely
parallel, to mark hiatus-filler ([Ɂ] or [w]), not vowel quality. Change to use of u
likely to align with real quality of following vowel, not OH to NH sound change, as
per Kloekhorst (see below III.B.1 for similar change with /-on/). Clinching evidence
wrongly dismissed by Kloekhorst (2008: 168): APl a-ku-u-uš-š=a ‘also sea-shells’
(OH/OS) to u-stem aku-! Shows even *-ums > *-ųs > *-ōs > /-o:s/ (shortened when
unaccented). Thus in all likelihood also same result for *-oms and *- s.
B. Long Vowels
1. Accented long *ē > ē in hysterokinetic collectives in *-ēr (still likely ḫašduēr
‘brushwood, twig; splinter’—see Rieken 1999: 346-7—and certain in *ḫappēr
‘transactions, business’ required by ḫappiriya- *‘market-place’ > ‘city’—Rieken
1999: 317). Also in AnimNSg ḫaštērz(a) ‘star’ < *h2stēr+s.
2. Accented long *ō > ā in amphikinetic plurals such as widār ‘waters’ < *wedōr (with
shifted accent that cannot be due to the “kwetwores-Regel”!). Shortened to a when
unaccented: OH/MH NSg antuwaḫḫaš ‘human’ < virtual *én-dhweh2-ōs.
3. Accented long *ā > ā in duwān parā ‘long before’ < *[dwa:m] (by Stang’s Law from
underlying ASg *dwéh2m ‘distance, long time’). See Melchert 2008.
4. Unaccented long *ō > a in NSg keššar(=šiš) < *ghésōr ‘hand’.
5. Unaccented long *ā > a probable in pēran ‘in front’ < *pérām ~ Grk. πέρᾱν ‘across,
opposite’ with Dunkel 1997: 72 contra Melchert 1994: 135.
C. Diphthongs
Long *-ōis with secondary accent > amphikinetic NSg -āiš (but ASg -āin unlikely to be
regular continuant of *-ōy- ). PIE status of rare matching -āuš < *-ōus uncertain.
III. Sonorants
A. Final Non-Syllabic
1. Final *-m merges with final *-n as -n (see various examples cited under II. above).
Date of merger (Common Anatolian or pre-Hittite) difficult to determine.
5
2. Final *-Vr# > -Vr: N-APl *wedōr with accent shift > widār and Pres3Sg of type *tukór > *tukkār → tuqqāri ‘is visible; matters’ (Yoshida 1990: 112-4).
3. Final *-Vr# > -V: N-APl *-Vr > -V in partawa ‘wings’, ḫattata ‘pieces of wisdom’ etc.
(Melchert 1988), Pres3Sg of type *kéi-tor > kitta ‘lies’ (Yoshida 1990: 112-4).
4. Final *-l: no sure examples. šuwī/ēl ‘thread’ could represent hysterokinetic
collective *suh1-ēl (after Eichner 1973: 54), but more likely substantivized
neuter*suh1-ílom with syncope, as is certain for similar cases (Rieken 2008b: 249).
See also Rieken 2008b: 250 for pronominal GSg in -ēl < syncopated *-élos/m, contra
Oettinger 1999: 264.
B. Final Syllabic
1. Final *- > -un, as in athematic Pret1Sg. Probably /-om/, based on sporadic NS
spellings pa-(a)-u-un and e-ep-pu-u-un (thus with Kloekhorst 2008: 609), but plene
spelling can hardly represent vowel length, and MH/MS shows rather pa-a-ú-un
(3x). Thus use of -ú-/-u- not true plene, but to mark transition sound in hiatus
/pa:_on/ < *pói-h1y- . Cf. II.A.7 above on AnimAPl ending.
2. Other final *- > -aR (via *-oR—see immediately below). NB development must be
ordered after loss of final *-r after unaccented vowel in III.A.3 above! Examples for
*- # in N-ASg wātar ‘water’, uttar ‘word’ etc. < *- and -ātar < *-t (types in -mar
and -eššar are secondary Hittite creations). Also in Pret3Pl -ar < *- such as
šapašiyar ‘spied’ (pace Kloekhorst 2008: 244). Result of post-vocalic *-w is
likewise -war, as in generalized N-ASg of verbal noun. Examples for *- in N-ASg
-man < *-m (lāman ‘name’ < *h1néh3-m etc.) and -an < secondary *- (lāgan
‘inclination’ etc.): for latter see Melchert 2007-2008[2010]. Probable *- in N-ASg of
deverbatives in -al (ard-al ‘saw’, išḫiyal- ‘binding’ etc.). Syncopated *-ólom less
likely in view of unsyncopated -āla- < *-ó-lo- in adjectives (then animate “agentnouns”) such as genzuwala- ‘merciful’ (see Rieken 2008b: 251).
3. Final *-(C)w > *-CwoR > -CuR at least for liquids. Result surely /-CoR/, with
regular lowering of *o > a as in regular *- > -aR blocked by preceding *w, which
is then lost by dissimilation. Not unmotivated resyllabification of *-w > -uR
(Melchert 1994: 132): e.g. ḫenkur ‘offering’. For interpretation as /-CoR/ NB
spellings of a-ni-u-ur with analogical postconsonantal treatment of *-Cw and
secondary accent, for which see examples like pít-tu-u-la- ‘snare’, hypostasized from
genitive *‘that of spreading out’ to virtual *péth2-w , *peth2-wén-s *‘spreading out’
(thus w/ Rieken 1999: 472 after Puhvel 1979: 211, pace Kloekhorst 2008: 681). NB
cannot belong to type of syncopated *-ú-los/m (Rieken 2008b: 252-3), since latter
show consistent expected /u:/, as in iš-ḫi-ú-ul ‘treaty’. Examples for *-w # lacking.
C. Interior Syllabic
For final *- s# in AnimAPl ending see II.A.7 above.
IV. Obstruents
A. Final Stops
1. Final *-t (after voiceless obstruent) preserved by addition of prop-vowel in miconjugation (where NB postvocalic ending was preserved as per preceding): Pret3Sg
6
ēšta ‘was’ < *h1és-t. Generalized after all obstruents: NB Pret3Sg e-ku-ut-ta ‘drank’
/e:ɣwta/. See Yoshida 1993: 23 with refs.
2. Final *-d (after V and R) preserved as -d: pronominal NeutN-ASg apāt, -at etc. < *-od,
endingless locative šiwat ‘day’, Inst < Abl *-d in kiššart(a) ‘hand’, wedand(a)
‘water’, kēt ‘this’ (NB for original ablatival sense kitpantalaz ‘from this time on’ and
suppletive enclitic possessives in -it for ablative). See Melchert-Oettinger 2009: 54.
Loss of final stop in NeutN-ASg of stems in *-nt probably already PIE. Final *-g in
ú-ug ← *ég after loss of final *-h2 also kept (see below IV.C.4). Contra Melchert
1994: 85 generalization of voiced stop here probably PIE: Goddard 2007: 122-3 w/ refs.
3. Whether ḫi-conjugation Pret3Sg -š continues *-s-t with regular loss of *-t or simply *-s
cannot be determined on phonological grounds.
4. No sure examples for original final labial stop. Preservation of final stops of all points
of articulation in Imv2Sg may be due to trivial analogical restoration/maintenance:
ēp ‘take!’, ḫark ‘hold!’, etc. NB eku ‘drink!’ with labiovelar /é:gw/ vs. lāḫ ‘pour!’ below.
B. Final *-s
1. Regularly preserved after vowel: see various nominal cases endings and OH/MS
Pret2Sg ending -Vš.
2. Also maintained in AnimNSg of stems in dental stop and *-nt-: karāz to karād‘innards’, šīwaz ‘day’ ← *dyéw-ot-s, -anz(a) < *-e/ont-s, etc.
3. Maintained after other stops and -ḫḫ- by anaptyxis, especially ḫi-conjugation
Pret3Sg in -iš (arguably after loss of final *-t in *-s-t—see IV.A.3 above): akkiš
‘died’, ḫuwappiš ‘hurled’, newaḫḫaš ‘renewed’, etc. and with analogical vocalism
also maniyaḫḫiš ‘handed over’ etc. See Melchert 1994: 174 with refs. Also surely
analogical in MH ḫatteš ‘pricked’ (Oettinger 1993: 155). Also after *h1 ([Ɂ] or [h]) in
N-ASg āiš ‘mouth’ < *(h1)óh1(-)s (Melchert 2010).
4. NB lost in secondary final sequences *-Car/ls < *-Cr/los with syncope and new
anaptyxis: AnimNSg ḫuppar ‘bowl’ < virtual *h2úpros, AnimNSg šarnikzil
‘restitution’ < *-tílos (see Melchert 1993 and significant revisions by Rieken 2008b:
246-50). Suggests Imv2Sg examples such as karš ‘cut!’ may show analogical
restoration/maintenance.
C. Final Laryngeals
1. At least *h2 regularly lost after vowel, via intermediate step of “lenition”: NB with
Eichner (1973: 59) miyaḫuwant- ‘mature’ (single -ḫ-!) reflecting *myaḫ < *m(h1)yéh2 ‘growth’. Geminate -ḫḫ- of mēnaḫḫanda ‘facing, opposite’ < allative *méneh2
plus anda ‘into (Nikolaev 2010: 68) due to initial *h1 in adverb?? I.e., virtual
*méneh2 h1endo? Next stage of loss of “lenited” laryngeal with compensatory
lengthening indirectly attested in -ātar < *-āt (NB with regular “lenition” after
accented long vowel, contra Melchert 1994: 69; see Melchert 2007: 2-3 after
Jasanoff) and -āwar < *-āw . Also directly in lexicalized āššū ‘goods (NB spelled aaš-šu-u, thus /a:sso:/ with lowering of /u:/ before *h2: see Rieken 2005: 537 with
refs. and Kloekhorst 2008: 59) and OH substantival mekkī ‘a lot’ < virtual *mégh2ih2.
Otherwise final *-Vh2 represented by short vowel (-a, OH and MH -i and -u, as per
7
Watkins 1982: 282-3, pace Prins 1997: 209-11 and passim, et al.). Absence of any
direct lexicalized -ā < *-eh2 more likely accidental than due to difference in
chronology of shortening of -ā vs. -ū and -ī (see Watkins 1982: 281).
2. Final *-Vh1 probably reflected in neuter i-stems reflecting old duals: e.g. GIŠēlzi
‘(balance) scales’ < *-tī < *-tih1 (see Puhvel 1984: 270 and Rieken 1994: 52, who
cites further plausible examples). Far less certain is -a (synchronically interpreted as
collective plural) < thematic *-oh1. Other examples cited by Puhvel (1969: 61) are
extremely dubious, but šākuwa ‘eyes’ may indirectly continue *-oh1 taken from
thematic examples (so cautiously Watkins 1986: 6133). Compare IV.C.4 below. 2Sg
subject *tī in zīk could represent *tu-h1 (Schrijver 1999: 157), but real *tū with
monosyllabic lengthening also possible. See I.B.3 above for impossibility of preform
*tuh2.
3. No clear examples for final *-Vh3. Imv2Sg dā probably reflects regularly *dóh3 (or
alternatively *déh3), but generalization of dominant allomorph dā- from rest of
paradigm cannot be excluded.
4. At least *h2 regularly lost after stop: OH NeutN-ASg mēk ‘much’ < *mégh2, OH/OS
CollPl aniat=šet ‘his regalia’ < *-ot-h2 (assumption of spelling error entirely
gratuitous), 1P subject ūg ‘I’ ← *ēg < *égh2.
5. Final *-h2 also lost after sonorant in ke/itkar ‘at the head’ < *kéd ‘here’ + endingless
locative *kérh2 (see Nussbaum 1986: 96ff.). Preservation of *h1 or *h3 in same
environment implausible. Imv2Sg walḫ ‘strike!’ < *welh3- shows trivial restoration/
preservation just like šāḫ, šanḫ (ša-an-ḫa/ša-na-aḫ) ‘seek, search!’< *senh2-.
8
References
Agbayani, Brian, and Chris Golston. 2010. Second-position is first-position:
Wackernagel’s Law and the role of clausal conjunction. IF 115.1-21.
Dunkel, George. 1997. B. Delbrück and the Instrumental-Ablative in *-m. In E. Crespo &
J. L. García Ramón (eds.), Berthold Delbrück y la sintaxis indoeuropea hoy, 63-83.
Wiesbaden: Reichert.
——. 2007. Chips from an Aptotologist’s Workshop I. In A. Nussbaum (ed.), Verba
Docenti. Studies in historical and Indo-European linguistics presented to Jay H.
Jasanoff by students, colleagues, and friends, 53-61. Ann Arbor/New York: Beech
Stave.
Eichner, Heiner. 1973. Die Etymologie von heth. mehur. MSS 31.53-107.
Georgiev, Vladimir. 1971. Die Herkunft der hethitisch-luwischen Dativ-Lokativendungen
des Singulars. IF 76.59-65.
Gertz, Janet. 1982. The Nominative-Accusative Neuter Plural in Anatolian. New Haven:
Yale University Ph.D. dissertation.
Goddard, Ives. 2007. Phonetically Unmotivated Sound Change. In A. Nussbaum (ed.),
Verba docenti. Studies in Historical and Indo-European Linguistics Presented to Jay
H. Jasanoff , 115-30. Ann Arbor/New York: Beech Stave.
GrHL = Hoffner, Harry A. Jr., and H. Craig Melchert. 2008. A Grammar of the Hittite
Language. Part I: Reference Grammar. Winona Lake: Eisenbrauns.
Jasanoff, Jay. 1976. Grk. ἄμφω, Lat.ambō et le mot indo-européen pour ‘l’un et l’autre’.
BSL 81/1.123-31.
——. 2008. *-bhi, *-bhis, *-ōis: following the trail of the PIE instrumental plural. In: J.
Rasmussen & T. Olander (eds.), Internal Reconstruction in Indo-European:
Methods, results, and problems. Section Papers from the XVI International
Conference on Historical Linguistics. University of Copenhagen, 11th-15th August,
2003, 137-149. Copenhagen: Museum Tusculanum.
Josephson, Folke. 1972. The Function of the Sentence Particles in Old and Middle
Hittite. Uppsala: Skriv Service AB.
Kimball, Sara. 1999. Hittite Historical Phonology. Innsbruck: Institut für
Sprachwissenschaft der Universität Innsbruck.
Kloekhorst, Alwin .2008. Etymological Dictionary of the Hittite Inherited Lexicon.
Leiden/Boston: Brill.
——. forthcoming. The Phonological Interpretation of Plene and Non-Plene Spelled e in
Hittite. In B. Nielsen et al. (eds.), The Sound of Indo-European. Selected papers from
the conference held in Copenhagen 16-19 April, 2009.
Melchert, H. Craig. 1988. Final -r in Hittite. In Y. Arbeitman (ed.), A Linguistic
Happening in Memory of Ben Schwarz, 215-234. Louvain-la-Neuve: Peeters.
——. 1993. A New Anatolian “Law of Finals”. Journal of Ancient Civilizations 8.105-13.
——. 1994. Anatolian Historical Phonology. Leiden/Atlanta: Rodopi.
——. 2004. Second Thoughts on *y and *h2 in Lydian. In M. Mazoyer & O. Casabonne
(eds.), Studia Anatolica et Varia. Mélanges offerts à Professeur René Lebrun.
Volume II, 139-50. Paris: L’Harmattan.
——. 2007. Luvian Evidence for PIE *h3eit- ‘take along; fetch’. Indo-European Studies
Bulletin, UCLA 12/1.1-3.
9
——. 2007-2008 [2010]. Neuter Stems with Suffix *-(e)n- in Anatolian and Proto-IndoEuropean. Die Sprache 47.163-181.
——. 2008. Hittite duwān (parā). In C. Bowern, B. Evans & L. Miceli (eds.),
Morphology and Language History in honour of Harold Koch, 201-9. Amsterdam/
Philadelphia: Benjamins.
——. 2010. The Word for ‘mouth’ in Hittite and Proto-Indo-European. International
Journal of Diachronic Linguistics 7.55-63.
Melchert, H. Craig, and Norbert Oettinger. 2009. Ablativ und Instrumental im
Hethitischen und Indogermanischen. Ein Beitrag zur relativen Chronologie. IncLing
32.53-73.
Nikolaev, Alexander. 2010. Hittite menaḫḫanda. JAOS 130.63-71.
Nussbaum, Alan. 1986. Head and Horn in Indo-European. Berlin/New York: de Gruyter.
Oettinger, Norbert. 1993. Mittelhethitisch hattes ‘sie schlachtete. Linguistica 32.153-5
(Bojan Čop septuagenario in honorem oblate).
——. 1994. Etymologisch unerwarteter Nasal im Hethitischen. In J. Rasmussen (ed.), In
honorem Holger Pedersen. Kolloquium der Indogermanischen Gesellschaft vom 26.
bis 28. März 1993 in Kopenhagen, 307-30. Wiesbaden: Reichert.
——. 1995. Griech. ὀστέον, heth. kulēi und ein neues Kollektivsuffix. In H. Hettrich et
al. (eds.), Verba et structurae. Festschrift für Klaus Strunk zum 65. Geburtstag, 21128. Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck.
1999. Zum nordwest-indogermanischen Lexikon (mit einer Bemerkung zum hethitischen
Genitiv auf -l). In P. Anreiter & E. Jerem (eds.), Studia Celtica et Indogermanica.
Festschrift für Wolfgang Meid zum 70. Geburtstag, 261-7.
——. 2000. Heth. udnē, armen. getin ‚Land‛ und lyk. wedre/i-. In A. Hintze et al. (ed.),
Anusantatyai. Festschrift für Johanna Narten, 181-7. Dettelbach: Röll.
Prins, Anna. 1997. Hittite Neuter Singular – Neuter Plural. Some Evidence for a
Connection. Leiden: Research School CNWS.
Puvhel, Jaan. 1979. Hittite words with initial pít/pát sign. In E. Neu & W. Meid (eds.),
Hethitisch und Indogermanisch,209-17. Innsbruck: Institut für Sprachwissenschaft
der Universität Innsbruck.
——. 1969. Hittite annaš šiwaz. KZ 83.59-63.
——. 1984. Hittite Etymological Dictionary.Vol. 1 Words beginning with A. Vol. 2
Words beginning with E and I. Berlin/New York/Amsterdam: Mouton.
——. 1991. Hittite Etymological Dictionary. Volume 3: Words beginning with H.
Berlin/New York: Mouton de Gruyter.
Rieken, Elisabeth. 1994. Der Wechsel -a-/-i- in der Stammbildung des hethitischen
Nomens. HS 107.42-53.
——. 1999. Untersuchungen zur nominal Stammbilding des Hethitischen (StBoT 44).
Wiesbaden: Harrassowitz.
——. 2005. Zur Wiedergabe von hethitisch /o/. In G. Meiser & O. Hackstein (eds.),
Sprachkontakt und Sprachwandel. Akten der XI. Fachtagung der Indogermanischen
Gesellschaft., 17.-23. September 2000, Halle an der Saale, 537-49. Wiesbaden:
Reichert.
——. 2008a. Die Zeichen <ta>, <tá>, und <tà> in den hieroglyphisch-luwischen
Inschriften der Nachgroßreichszeit. SMEA 50.637-47.
10
——. 2008. The Origin of the -l Genitive and the History of the Stems in -īl- and -ūl- in
Hittite. In: K. Jones-Bley et al. (eds.), Proceedings of the 19th UCLA Indo-European
Conference. Los Angeles, November 3-4, 2007, 239-56. Washington DC: Institute for
the Study of Man.
Schrijver, Peter. 1999. Vedic g bhṇāti, g bhāyáti and the semantics of *ye-derivatives of
nasal presents. MSS 59. 115-62.
Szemerényi, Oswald. 1956. Hittite pronominal inflection and the development of syllabic
liquids and nasals. KZ 73.57-80.
Watkins, Calvert. 1963. Preliminaries to a historical and comparative analysis of the
syntax of the Old Irish verb. Celtica 6.1-49.
——. 1982. Notes on the Plural Formations of Hittite Neuters. In E. Neu (ed.),
Investigationes Philologicae et Comparativae. Gedenkschrift H. Kronasser, 250-62.
——. 1985. Indo-European *-kwe ‘and’ in Hittite. In H. M. Ölberg & G. Schmidt (eds.),
Sprachwissenschaftliche Forschungen. Festschrift für Johann Knobloch, 491-7.
Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck.
——. 1986. The language of the Trojans. In M. Mellink (ed.), Troy and the Trojan War.
A Symposium Held at Bryn Mawr College, October 1984, 45-62. Bryn Mawr: Bryn
Mawr College.
Weitenberg, Jos. 1992. The Use of Asyndesis and Particles in Old Hittite Simple
Sentences. In O. Carruba (ed.), Per una grammatical ittita, 305-53. Pavia: Iuculano.
Yoshida, Kazuhiko. 1990. The Hittite Mediopassive Endings in -ri. Berlin/New York: de
Gruyter.
——. 1993. Notes on the Prehistory of Preterite Verbal Endings in Anatolian. HS 106.26-35.
——. 1997. A Further Remark on the Hittite Verbal Endings 1 pl. -wani and 2 pl. -tani.
In D. Q. Adams (ed.), Festschrift for Eric Hamp. Volume II, 198-93. Washington
DC: Institute for the Study of Man.
Zimmer, Stefan. 1994. Indogermanisch *h1su und *dus- im Kymrischen. ZCP 47.176200.
Luvo-Lycian Dorsal Stops Revisited
H. Craig Melchert
University of California, Los Angeles
1. Introduction
There is a broad consensus that at some prehistoric stage the Indo-European stop
system had only a two-way phonological contrast of labialized velars
(traditionally “labiovelars”) and non-labialized velars: thus schematically /Kw/
versus /K/. The latter had two conditioned allophones, palato-velars or front
velars (traditionally “palatals”) in fronting environments and non-front velars
(traditionally “velars”) elsewhere: [Kj] and [K]. Formulations of just what
constituted a “fronting environment” can differ markedly (one may compare
among many Meillet 1934: 94 and Lipp 2009: 1.7&53). Nevertheless, the highly
skewed—that is, non-random and largely complementary—distribution of
traditional “palatals” and “velars” strongly supports this fundamental premise.
The major point of dispute for more than a century has been whether or not a
partial loss of conditioning environment, analogy, and other factors led to a socalled phonemic split in PIE (that prehistoric stage that we reach by direct
comparative reconstruction). The result would have been a three-way contrast
/Kj/, /K/ and /Kw/, albeit very likely in a limited number of examples and with
much of the previous complementary distribution preserved, as is typical for
such splits. I contend that the answer to this question is yes, regardless of the
facts of Luvo-Lycian (likewise Kümmel 2007: 312).
The typological objection of Sihler (1995: 152-3) and Lipp (2009: 1.3165)
that the presupposed backing of palatal stops to velars in the “centum” languages
is unattested is a straw man. There is no compelling evidence that the PIE stops
conventionally labeled “palatals” were true palatal stops (IPA [c] and [ɟ]).
Evidence from all attested languages is fully compatible with front velars.1
There is in any case no necessity to assume that the phonological merger of /Kj/
and /K/ in the centum languages even took place by phonetic backing of front
velars. Since the place of articulation of /K/ in the older centum languages is
indeterminate, we may rather assume the more likely generalization of [Kj] in
fronting environments (along with other factors leading to an allophonic
distribution).
1
Compare Kümmel 2007: 318. I also regard as fully viable his alternative solution (2007:
318-27) of an original contrast of velar, uvular, and labialized velar stops.
2
No explanation is offered by Sihler, Lipp or others for the contrasting data
cited in the appendix below. Constructs such as *kreuh2- beside *kruh2- for
‘gore, bloody flesh’ (with a supposed front allophone before *r + front vowel) or
*kew-on- ‘dog’ (with wholly unattested root full grade!) as a variant of the root
generally cited as *(s)keuh1- ‘perceive, notice’ (Lipp 2009: 1.55&79) are
plausible enough, but they are totally unsupported for Proto-Indo-European as
reached by direct comparative reconstruction. They are rather the result of
internal reconstruction based on PIE and thus by definition represent pre-PIE.2
2. Luvo-Lycian Evidence I
The question at hand is thus not whether PIE had a three-way phonemic contrast
in its dorsal stops, but whether Luvian and Lycian truly offer further support for
that contrast. In Melchert 1987 I claimed that Luvian reflexes of voiceless dorsal
stops show a triple contrast before front vowels. Morpurgo Davies and Hawkins
1988 independently reached the same conclusion (as noted in the epilogue to my
article, Warren Cowgill had already made this analysis many years before, on
the basis of far less evidence). The issue was left open as to whether this contrast
was unconditioned or reflected a conditioned palatalization of *k but not *k.
However, in Melchert 1989: 23-31 (see especially page 30), I argued for an
unconditioned three-way contrast based on further putative evidence from
Lycian. This analysis had the important implication that Anatolian had merged
neither the PIE “labiovelars” and “velars” nor the “velars” and “palatals”. It was
thus neither “centum” nor “satem”, but was, as often, more archaic. This
formulation has met wide acceptance (e.g. Fortson 2004: 168, Kloekhorst 2008:
17-18), but occasional rejection (Sihler 1995: 154, Woodhouse 1998, Lipp 2009:
1.275-302).
The crucial Lycian evidence I cited then for *k > ts before *o (i.e., in a nonfronting environment) has not withstood scrutiny. Contra Melchert 1989: 29-30
the suffixes Lycian -(i)s- and Luvian -(i)zza- are not cognate, reflecting
*-i(s)ko-. Lycian -(i)s- forms abstracts and names for topographic features: e.g.,
wazzis- ‘__ship’ (an office) or Trmis- ‘Lycia’ (with Borchhardt-Eichner 199799: 82-3).3 The true equation is Luvian -zza- = Lycian -za- < *-tyeh2: e.g.,
2
Sihler (1995: 154) dismisses such evidence as “an artifact of the method”. This is
methodologically unacceptable: by definition all of PIE is an artifact of the comparative
method. One may always legitimately argue that the method has been incorrectly applied in a
given case (and that the results are thus false), but one cannot accept other results of the
comparative method on a large scale, as Sihler does, and airily dismiss those results that do
not happen to conform to one’s preconceptions of how some part of the grammar should look.
The data in the appendix is based on quite indisputable etymologies.
3
Lycian -(i)s- in my view continues *-sh2o- (cf. Luvo-Hittite -(a)šḫa- in tariyašḫa- ‘exertion,
exhaustion’, luliyašḫa- ‘marshland’, etc.), but this is an entirely separate issue. The key point
is that it does not match Luvian -(i)zza-, as I had mistakenly claimed.
3
CLuvian wašḫazza- ‘sanctified, holy’ = Lycian wasaza- (kind of priest),
HLuvian ku-ma-za- = Lycian kumaza- ‘priest’, as seen by Hajnal (2003: 193).
The HLuvian ethnic adjectives seen in /Karkamisa/itsa-/ ‘of Carchemish’ and
/nimu:witsa-/ ‘child’ (i.e. *‘of /nimu:wi-/ un-manhood’) likewise show the same
PIE suffix *-tyo- seen in Lycian Pttaraze/i- ‘of Patara’ and neleze/i- ‘of the
agora’ (for the Lycian suffix see Gusmani 1961 and Hajnal 1995: 82 with n. 96).
With removal of the supposed Lycian evidence showing *k > ts before back
vowel, my claim of an unconditioned three-way contrast in Luvo-Lycian
reflexes of voiceless dorsal stops cannot be upheld. The purported unconditioned
palatalized treatment of voiceless *k had always been suspect, due to the
appearance of voiced *g(h)- as Luvo-Lycian k- before back vowels (see below),
as per the apposite remarks of Woodhouse (1998: 40) and Lipp (2009: 1.297).
3. Luvo-Lycian Evidence II (voiceless stops)
We must reexamine all available evidence anew. Data for the voiceless stops is
summarized in Table 1:4
Table 1: Summary of Evidence for the Voiceless Series (C=CLuvian; H=HLuvan)
*k
*k
*kw
ziyari ‘lies’ (C) ~ Lyc. sijẽni
< *kéyo-
kišā(i)- (C) ‘to comb’
< *kes- OCS česati etc.
kui- ‘who’ (C) ~ Lyc. ti
< *kwi-
zār-(za)/zart- ‘heart’ (C/H)
*kēr/*kr̥d-
karš- (C) ‘to cut’
*k(e)rs- Toch kärst-/kärst-
kuar-/kur- ‘to cut’ (C)
*kw(e)r- Skt. kṛṇóti etc.
-za- (iter.) (C/H) ~ Lyc. -s< *-ske/o-
Lyc. tukedre/i ‘statue’
< *twek- Skt. tvác- ‘skin’
za/i- ‘this’ (C/H)
< *ko/i- Lith. šís etc.
kup(iya)- ‘to plot’ (C)
< *kup- Skt. kúpyati ‘desires’
Lyc. ti-se ‘some-/anyone’
< *-ke ~ Hitt. -kka
in kuēlka etc.
Lyc. xupa- ‘grave-house’
< virtual *kupeh2- ~ Gmc.
*xufa- ‘house; hill’ (Schaffner 2001: 143)
azu(wa)- ‘horse’ (H) ~ Lyc. esbeHLuv. kat- ‘fight’ in katunas
< *ékwo(Morpurgo Davies 1986: 132-3)
zuwan(i)- ‘dog’ (H)
~ Hitt. kattu- ‘weapon, talon’
< *kwon(thus contra Puhvel 1997: 138)
zurnid- ‘horn’ (H)
~ OIr. cath ‘fight’, Gmc. *haþu-,
< *kr̥ng-id- ~ Hitt. karkid-antOCS kotora ‘fight’, Germ. Hader
& Skt. śrnga< *kat(Starke 1990: 406-7; Puhvel 1997: 89)
wazi- ‘request’ (H)
< *wek-ye/o-5
4
Examples listed above the solid horizontal line in each column I regard as indisputable
etymologies, while those below it are merely probable or possible.
4
zarwaniya- ‘of horn’ (C)
< *kerw+
zalma- ‘shield, protection’ < *kelmn-o(Melchert 1988: 241-3; but cf. Lipp 2009:
1.27533 for an alternative Hurrian interpretation)
zalla- ‘trot’ (C), /zallal-/ ‘vehicle’ (H)
< *(s)kel- Lith. šúolis ‘jump’
(Rieken 1997)
/ziralamma/i-/ ‘fertile’ (H)
< *kērh1o- Lat. prō-cērus
(Rieken 2003: 45-50)
versus
kattawatnalli- ‘spiteful’ (C) & Hitt.
kadduwā(i)- ‘become aggrieved’
(Puhvel loc. cit.), + HLuv. kata- ‘enmity’
in ka-ti-i CRUS (L. Younger, pers. comm.
vs. Morpurgo Davies 1986: 132-3)
~ Grk. κότος ‘spite, anger’, Skt. śátru- ‘enemy’
< *kot-. Separate from kattu- ‘fight’ contra
Morpurgo Davies (1986: 144), Melchert (1987: 189-90),
Puhvel (1997: 140), Woudhuizen (2010: 44)6
Runtiya- < K(u)runtiya- < *kru-nt- ‘horned’
(Watkins 1999: 15-18, Woudhuizen 2010: 44)7
kumma- ‘sacralized’ (C/H) ~ Lyc. *kuma<*kunmo- < *kwn-mo- ~ Av. spënta- ‘holy’
(Schindler apud Watkins 1987: 401)8
?(477*)ku-tu-pi-li- (H) ‘fire offering’
< *ku- (Woudhuizen 2010: 44)
5
?kerut(i)- ‘Stag-god’ < *koru+
(thus revising Carruba 1978: 171-228 et al.)9
The stem also appears syncopated in the personal name element -uzzi- = IR, but contra
Zehnder (2010: 102) the HLuv. noun waza- ‘request’ is not a cognate of Skt. váśa- < *wóko-,
but a new creation among the productive deverbative action nouns in *-o-.
6
While one often fights one’s enemies, enmity or hatred is a mental state of existence, while
fighting is a concrete action, and the two cannot simply be conflated, as I and many others
have tried to do in this case.
7
I am much indebted to Heiner Eichner for calling my attention also to the name of Kubaba’s
consort in Carchemish, namely Karhuha, whose name is written in KARKAMIŠ A11+12,
§18 as (DEUS)CERVUS2+ra/i-hu-ha-ya and who surely also represents some form of a staggod (see Hawkins 2000: 106-107). The initial /k-/ would point to a preform *korh2u-.
8
As Lipp (2009: 1.293) correctly points out, my assumption of “Gutturalwechsel” (Melchert
1994: 252) was egregiously ad hoc.
9
The interpretation by Carruba and others of keruti as a sacrificial animal is contradicted by
the word order of the clause, which hardly allows keruti to be the direct object. I also am
unaware of any evidence for the use of cervids as sacrificial animals in IE-speaking Anatolia.
5
[Less likely but possible: kattu- ‘talon’ <*kat- based on Skt. śātaya- ‘cut; cut down’ (Jamison
1983: 90) (cf. Germ. Schlacht for sense ‘fight’).]
There are three critical points to be drawn from the preceding data. First, all
instances of *k ([kj]) > ts are either in exclusively fronting environments or in
paradigms with fronting environments Second, the number of examples cited for
*k > ts is immaterial; the crucial example is that of kišā(i)- ‘to comb’ that shows
*k > k in an exclusively fronting environment (more on this word below). Third,
a dispassionate review of all the relevant evidence includes some very credible
examples of Luvo-Lycian k < *k in non-fronting environments.
Current available evidence therefore points to a conditioned palatalization of
*k before front vowel (*i, *e, and *ē), yod, and *w (not u!). For *w as a fronting
environment note the virtual complete absence of PIE sequences in *Kw (versus
the well attested *Kw) and the strong tendency of labiovelars to be palatalized
(Hock 2009). HLuv. zurnid- ‘horn’ < *kr̥ng-id- suggests that also *kR- > *kjëR> zVR- (see below on zanta ‘down’ < *km/nt-).
Several points call for further comment. The existence of the stem kišš- ‘to
comb’ in Hittite (P3Sg kišzi) suggests that Anatolian inherited a root present
*kés-ti, whose weak stem *ks-(énti) might in principle have provided a nonfronting environment for Luvian kišā(i)-. However, whereas Hittite trivially
leveled the strong stem,10 one cannot possibly derive the Luvian verb in -ā(i)starting from the root present. Prehistoric *kés-/ks- could have led only to a
Luvian *kas+, leaving the i-vocalism of kišā(i)- wholly unexplained. One can
derive the latter with Čop (1970: 93) and Melchert (1994: 152) from a virtual
*kēséh2-ye/o- with lengthened grade like Latin cēlāre, and this is supported by
CLuvian ilḫā(i)- ‘to wash’, whose first vowel can only reflect a long *ē. Less
likely but not to be entirely excluded is *kes-éh2-ye/o- with a short root vowel (if
Luvian really does share with Hittite the change of pretonic short *e to i, for
which see Melchert 1994: 240, with due caution). In either case, the Luvian verb
My own attempt to take it as a personal name (2004a: 97) was an act of desperation also
effectively refuted by the context. The position of keruti in its clause (following ql=ebi, that
is, ql(a)=ebi ‘this here precinct’) suggests rather that it names a recipient of the sacrifice.
Given the prominence of the more common name for the Stag-god in western Anatolia
(reflexes of /Runtiya-/), such a recipient does not seem to me impossible. My hesitation rests
mainly on the less than clear morphology. In view of the tendency of ḫ to be lost in Luvian in
a sequence *-Rh2w- (Melchert 1994 : 258), it is not impossible that Lycian keru- reflects a
virtual *korh2wV-, i.e. that its base contained a laryngeal like that of Karhuha (see note 6).
10
Since the Hittite verb is attested all of twice, only in New Hittite manuscripts, its i-vocalism
before an s may easily be secondary and requires no special explanation (contra Kloekhorst
2008: 482). On the effective merger of /i/ and /e/ before /ss/ in NH verbs see already Melchert
1984: 147-9.
6
requires a preform with fixed mid-front vowel, and the non-affrication of the
initial *k retains its force.
Contra Lipp (2009: 1.273-5, 278 with note 36, 298) the reading zú for
hieroglyphic sign 448 is assured. Despite the protestations of Hawkins (2000:
36) et al., the alleged “interlocking” evidence for a value sù does not exist. The
fact that sign 108 (CORNU), the drawing of a horn, has a phonetic value sú
proves absolutely nothing, since as Hawkins himself concedes, we find in
KARATEPE the word “CORNU+RA/I”su-ra/i- meaning ‘plenty, abundance’
(where the use of ‘horn’ as the determinative reflects the notion of a
cornucopia). It is quite clear that the value sú for sign 108 is acrophonic to this
word, which is based on the verbal root su- ‘fill’ and has nothing to do with the
word for ‘horn’. If on the other hand the word for ‘horn’ truly began with /su-/, it
is completely inexplicable why it was not spelled with the transparent 108 sign
for ‘horn’, but rather with the opaque 448: (“CORNU”)zú+ra/i-ni. Further
confirmation for the value zú is furnished by zú+ra/i-wa-ni-(URBS) ‘of Tyre’
(i.e. Ṣūr) (thus Lipiński 2004: 115 and Yakubovich 2010: 66-758), with a
predictable rendering of Semitic emphatic ṣ as an affricate ts (see the references
given by Lipp 2009: 1.310).
The three HLuvian words for ‘horse’, ‘dog’, and ‘horn’ thus cannot be
loanwords from Indic via Hurrian, since they show an affricate /ts/, not a sibilant
/s/. The morphological match of HLuvian /zurnid-/ and Hitt. karkid- in
karkidant- ‘horned’ also precludes borrowing. While the suffix -id- is very
productive in Luvian, including in loanwords, it is decidedly not so in Hittite,
and the word equation can only reflect a common preform *kng-id-, where a
further suffix has been added to the base seen in Sanskrit śṅga- (see for this
analysis of the Hittite already Puhvel 1997: 89). Hittite, which retained the stop
in what was now a sequence *-arngi-, deleted the nasal. Loss of the stop in
Luvian (see below) allowed the nasal to remain. As discomfiting as it may be,
sporadic uR < * beside aR in Anatolian is a fact (see Melchert 1994: 260),
comparable to the situation for uR and iR in Baltic as described by Lipp (2009:
1.24-7).11
New evidence presented by Kloekhorst (2008: 34-50) for u spelling /o/ in
Hittite versus ú for /u/ offers welcome support for my rather tentative claim
(Melchert 1994: 125) that the immediate result of * in Hittite was *oR: note
an-tu-u-ri-ya- ‘interior’ /ando:riya-/ < *andwóriya- < *en-dhw-. If the same
was true for Luvian, then the prehistoric result *oR would have normally
lowered to merge with /a/, but in some words rose instead to merge with /u/.
11
Obviously, however, my consistent citing there of the root of gurta- as beginning with a
front velar *g(h)- was wholly erroneous.
7
That Luvian did share the same development is suggested by kumma‘sacralized’. Rather than assume an unmotivated resyllabification as given in
Table 1 above (*kwn-mo- >*kunmo-), we should suppose rather a sequence
*kwn-mo- > *kwonmo- > *konmo- (with dissimilatory loss of the *w before *o)
and read the word as /kommo-/. Lack of any “plene spellings” in CLuvian
precludes proof of this analysis, but other use of u and ú in CLuvian confirms
that Kloekhorst’s basic generalization holds also for it. That is, we must assume
that Luvian also had both phonemic /u/ and /o/.
The need to assume that syllabic sonorants acted as a fronting environment
in a language where the anaptyctic vowel appears as either /a/ or /u/ is decidedly
unwanted. However, we may be dealing with the kind of reciprocal effects seen
in Greek κύκλα ‘set of wheels, chariot’ < *kwekwléh2, where the labiovelars
rounded the anaptyctic vowel, which in turn caused delabialization of the
preceding stop (see Eichner 1985: 139). Likewise, we are allowed to suppose
that the front velar stop led to a palatal onset of the anaptyctic vowel, which then
caused the palatalization (affrication) of the front velar, after which it was
absorbed in the affricate, leaving only the non-front nuclear vowel: *kjr̥ng-id- >
*kjjorng-id- > *tsjorng-id- > *tsorng-id- > zurnid- (again really /tsornid-/).
The recent demonstration by Goedegebuure (forthcoming) that CLuvian
zant/da means ‘down’ and equates to Hittite katta likewise requires that a
syllabic sonorant act as a fronting environment. Her analysis rests on
comparisons such as CLuvian zanta...kišamman āšdu with Hittite katta kišān
ēšdu ‘let it be combed down’ and likewise zanda dūpaimmi- with katta GULant- ‘struck down, afflicted’. She cites further the functional equivalence of
CLuvian ¬zantalanuna with Hittite tepnumanzi ‘belittle’, persuasively
interpreting the Luvian as ‘erniedrigen, humiliate’ < *zantala/i- *‘low’ (cf.
šarla/i- ‘high’ and šarlā(i)- ‘exalt’). As she correctly argues, there is no
connection with Hittite katta(n), katti in the meaning ‘with, beside’ (cognate
with Latin cum ‘with’, OCS kŭ ‘to’ etc.), so the only sure cognate is Greek
κατά/κάτω. This permits us likewise to start from front velar *kjm/ntV and
assume a development parallel to that sketched for ‘horn’ above: *kjm/ntV > *kj
jom/ntV > *tsjom/ntV > *tsontV > zanta.12 This “reciprocal effect” scenario is
manifestly hypothetical, and I do not insist upon it in the form presented. I offer
it only to show that such behavior of syllabic sonorants following a sound that
was itself already a front velar is by no means preposterous.
Palatalization of *k before short *ĕ is undeniable. The attempt by Lipp
(2009: 1.284-5) to eliminate short *ĕ conditioning is impossible: there simply is
12
Lipp (2009: 1.53) assumes that a syllabic nasal served as a fronting environment in pre-PIE,
in order to account for *dekm(t) ‘ten’ with consistent front velar, but does not explain how it
might have done so.
8
no conditioned change of posttonic *-es > Luvian -is, as he proposes. CLuvian
ta-piš-ša ‘sky, heaven’ is to be read as ta-pašx-ša, as shown by tap-pa-aš-ša (for
pašx see Neu-Rüster 1989: 216 w/refs.). There is thus no way to derive iter. -zafrom a non-existent preform *-ski(C). Lycian tise ‘someone, anyone’ < *kwis-ke
also directly contradicts any alleged rule of posttonic *ĕ > i. We thus confront
either a contrast of kišā(i)- < *kesāye/o- versus -za- < *-ske- with short *ĕ or of
kišā(i)- < *kēsāye/o- versus zār < *kēr (with vocalism from weak stem for *zīr)
with long *ē. Luvian and Lycian must therefore have inherited a three-way
contrast of */kj/, */k/ and */kw/, with conditioned palatalization of only*/kj/, but
not */k/, in fronting environments.
4. Luvo-Lycian Evidence III (voiced stops)
The picture for voiced dorsal stops (which include PIE voiced aspirated dorsal
stops) is unfortunately much less clear, for two reasons. One is simply the lack
of more than a handful of compelling etymologies for the velar set. Those
available do not cover enough different environments to determine with
confidence the conditioning for various reflexes. The second problem is that
there is clear evidence for widespread medial loss of voiced stops (see Kimball
1994 and Melchert 2004b). This means that, for example, in /zurnid-/ ‘horn’ <
*kr̥ng-id- (NB with velar *g) it is impossible to determine whether the loss is
due to a general loss of medial prevocalic *g(h) or results from palatalization to
yod before *i and then absorption of the glide (as certainly is the case in /imra/i-/
‘open country’ < *ghemro- cognate with Hittite gimra-).
The evidence that we have available appears in Table 2.
Table 2: Summary of Evidence for the Voiced Series (compare in general Kimball 1994)
*g(h)
*g(h)
*gw(h)
/kuttassra/i-/ ‘orthostat’ (C/H)
< *gh(o)ut- ‘wall’ < ‘pour’
gurta- ‘citadel’ (C)
< *ghr̥dho- Skt. grha-
wawa/i- ‘cow’ (H) ~ Lyc.
wawa- < *gwo/ew-
katmaršiya- ‘defecate’ (C)
< * ghodm(V)r
t(u)watra/i- ‘daughter’ (H)
< *dhugh2tr-
wāna- ‘woman’ (C)
< *gwón-
kallar(a)- ‘unfavorable’ < *ghalH- pa- ‘get a share’ (H)
ON galli ‘flaw’ OIr. galar ‘illness’ < *bhag- Skt. bhaj- etc.
Lith. žalà ‘damage, wound’
(Melchert, 2004b)
u- ‘to drink’ (C/H)
< *egwh- Lat. ēbrius etc.
t(i)yamm(i)- ‘earth’ (C)
< *dhghém(i)- (delocatival)
dakkuwa/i- ‘dark’ (C)
< *dhéngwo- Fris. diunk
‘dark’ Welsh dew ‘fog’
qãn- ‘slay’ < *gwen-(?)
īššra/i-, /istra/i-/ ‘hand’ (C/H)
< * g(h)esr-
zurnid- ‘horn’ (H)
< *kr̥ng-id- Skt. śrnga-
*kwanza- (REL-za-)
‘incise’ (H) < *gwhs(k)e/o-
9
/imra/i-/ ‘open country’ (C/H)
< *ghemro- (< ‘earth’)
nāna/i- ‘brother’ (C/H) ~ nẽne/i< *négno- ~ Hitt. negna- (?)
(“PES2”)waza- ‘drive’ (H)
< *wégh-ye/o-
(CORNU)ki-pu-tà/ra ‘horn’(?) (H)
< *geibh- ‘bent, curved’(?)
zūwa- ‘food’ (C)
< *gyóuHo- OHG kiuwan ‘chew’
parray(a)- ‘high’ (C)
< *bherghoi- Arm. barjr etc.
papparkuwa- ‘purify’ (C)
< *bhr̥h1gw- Av. brāzaiti ‘shines’
/taggam-/ ‘land’ (H)
< *dhéghōm Hitt. tēkan etc. (?)
ruwan ‘formerly’(H) < *ghrēw+
OIce. grýjandi ‘dawn’, Lith.
žėrėti ‘gleam’
Deocclusion of labiovelar *gw( h) > w except next to nasal seems assured.
The case of Lycian qãn- is still much debated. However, the recent
demonstration by Kloekhorst (2006: 97-101) that the one sure source of Lycian
q is *h2w effectively eliminates any comparison with Hittite ḫanna- ‘judge’ (thus
still Puhvel 1991: 82, with references to predecessors). The occurrences of the
Lycian verb in any case demand a sense ‘destroy, kill’. We would expect a
singular *gwénti to lead to *wẽti with deocclusion, but in the plural *gwnénti the
following n could have preserved the stop. What is much harder to account for is
the a-vocalism. I can only suggest that the immediate result of the third plural
*qnẽti led to a specifically Lycian anaptyxis, hence *qãnẽti (by chance
unattested), from which was back-formed a new singular qãñti. The number of
required unverifiable steps in this derivation leaves the word of uncertain status.
However, if one accepts the plausible though not absolutely compelling
etymology of HLuvian (CAPERE+SCALPRUM) REL-za- ‘incise’ from an
iterative of *gwhen- ‘strike’, then one must likewise here assume that occlusion
of the initial voiced labiovelar was blocked by the following nasal. Carruba
(1978: 16921) assumes a reading *kuenza-, but in an e-grade form we would
definitely predict deocclusion: compare CLuvian wi/enal- ‘stick’ (*‘striking
instrument) < virtual *gwhen- (thus with Starke 1990: 313-5). We should rather
assume *kwanza- (thus Hawkins 2000: 70 and passim), based on an iterative with
zero grade of the root *gwh- s(k)e/o-. It remains surprising to me that a syllabic
nasal would have such a blocking effect on deocclusion.
Examination of Table 2 shows that available assured etymologies permit
assumption of a conditioned treatment of just *g(h) > [j] >  /__Vfront, and *g(h)
> ts /__*y, with preservation initially before back vowels, but a crucial
10
contrasting example to prove that the first two changes did not apply also to
velars are lacking! Although the determinative obviously is quite suggestive, we
cannot be sure that HLuvian (CORNU)ki-pu-tà/ra- (/kibud(a)-/ or /kipud(a)-/)
refers to the crescent horn of the moon (see Hawkins 2000: 470 on the problem
and Puhvel 1997: 188 on the questionable Hittite-context SIkiputi-). Even if it
does, the derivation suggested above from a root meaning ‘bent, curved’ is a
mere possibility and is not the sort of etymology on which one bases a sound
law.
At least suggestive is the apparent contrast we find between palatalization of
*k in HLuvian zurnid- ‘horn’ < *kr̥ng-id- and the lack thereof in CLuvian gurta‘citadel’ < *ghr̥dho-. I must insist on the latter etymology, against that from
*gw(H)to- (e.g. Bader 1991: 127, Kimball 1999: 250). The problem is not
formal (against Melchert 1994: 260), since the labiovelar could have been
unrounded by the round anaptyctic vowel before being deoccluded: *gw(H)to- >
*gworto- > gurta- (read /gorta-/).13 The insurmountable obstacle is the
semantics: all reflexes of the root *gwerH- in question mean ‘mountain’
(Sanskrit giri-, Avestan gairi-, OCS gora) or refer to natural phenomena
putatively associated with mountains, such as Greek βορέᾱς ‘north wind’ <
‘wind off the mountains’. However, it is clear from the use of CLuvian gurta- in
the Deeds of Suppiluliuma that the word refers to a man-made fortified
enclosure, which is not necessarily on a high place: KBo 5.6 iii 33: šarāzzi gurti
‘into the upper citadel/walled city’. If gurta- meant inherently ‘acropolis, high
place’, it would be pointless to specify it as ‘upper’, which clearly implies the
existence of a similar lower enclosure. The problem for our present inquiry is
that the absence of palatalization in gurta- < *ghr̥dho- cannot prove that a
sequence of voiced front velar *g(h) plus syllabic * did palatalize. It is only
such a direct contrast within the voiced series that would be probative, because
even if there was conditioned palatalization of only front velar *g(h) and not
non-front velar *g(h), we cannot simply assume that the conditioning matched
precisely that of the voiceless series (especially in non-obvious environments
such as *).
As already noted above, the general treatment of medial voiced front velars
and non-front velars is obscured by uncertainty about the precise conditions on
the deletion or retention of both. Medial loss before vowel seems likely. See
Melchert 2004b: 377 for HLuvian pa- ‘allocate; receive a portion’ < *bhag-V-.
In the word for ‘daughter’ the unsyllabifiable *dhugh2tr- received an anaptyctic
vowel colored by the laryngeal, which was then lost regularly before stop. Loss
of the prevocalic *g and then glide insertion led regularly to HLuvian
13
This remains true whether one derives Hittite gulšš- and Luvian gulz- ‘inscribe’ from *gws‘prick’ (thus with Puhvel 1997: 244 following Carruba) or from *kws- (Oettinger 1979: 204 et
al.). As Puhvel rightly insists, that choice must be made on other grounds.
11
/tuwatra/i-/: *dhugh2tr- > *dhu.gah2.tr- > *du.gah.tr- > *du.ga.tr- > *du.a.tr- >
du.wa.tr-. Lycian syncope of the first vowel led further to kbatra-.14
Whether medial *g(h)/g(h) was lost medially before consonants is far less
clear.15 Čop (1970: 91) reads HLuvian dat.-loc. sg. ta-ka-mi ‘in the country’ as
/taggami/, reflecting generalization of the strong stem *dhégōm, with the medial
voiced *g being preserved due to its being geminated by Čop’s eponymous law
(followed in Melchert 1994: 232). However, ta-ka-mi may alternatively be read
/tagmi/ (Lipp 2009: 1.297 et al.) and taken as a reflex of *dheghm-(e)i, with
preserved voiced stop before nasal. Hittite negna- ‘brother’ is generally equated
with Luvian nāna/i- and Lycian nẽne/i- (most recently by Kloekhorst 2008:
601). However, a clear etymology for the word remains to be found, and we
cannot at present exclude that the Luvian and Lycian words are mere Lallwörter
that resemble the Hittite purely by chance (see Puhvel 2007: 108).16 I must not
fail to mention one last nagging complication. We have seen that *w acted as a
fronting environment and led to affrication of *k to ts. However, in the verb
papparkuwa- ‘purify’, most likely from reduplicated *bhr̥h1gw- to the root
*bhreh1g- seen in Av. brāzaiti ‘shines’, the *g is preserved as a velar stop. Is this
different treatment due to the difference in voicing of the stop, or did the
preceding non-syllabic r that developed in *bhr̥h1g- > *park- block the
palatalization? Strictly speaking, of course, we do not even control the outcome
of *g(h)w- more generally, for lack of evidence.
5. Conclusion
The unconditioned three-way contrast I claimed in 1989 for the voiceless dorsals
in Luvian and Lycian is to be rejected. Luvo-Lycian is rather the “mirror-image”
of Albanian: before the merger of the front and non-front velars, the voiceless
front velar, but not the non-front velar, underwent conditioned palatalization. A
similar conditioned palatalization of just voiced front velars is also possible, but
cannot at this point be affirmed. Thus Anatolian is in terms of the traditional
classification “centum”: it does show an eventual phonological merger of the
front and non-front velars.
14
Since we cannot know whether the HLuvian stem was /tuwatra/i-/ or /twatra/i-/, *duatrmay alternatively have led directly to *dwatr- with disyllabification of the *u, producing
/twatra/i-/ and kbatra- directly.
15
For further consideration of this problem see Ofitsch 1998: 426 with references.
16
Unfortunately, the very attractive etymology of Neumann (1991: 63-4), *ni-gnh1-ó‘inborn’, runs afoul of the already OS spelling ne-eg-na in KBo 20.31 Ro 6, whose e-vocalism
cannot be derived from *ni-. The Luvian and Lycian obviously also exclude *ni-, if they are in
fact cognate!
12
Appendix: Evidence for Three Sets of Dorsals in Contrasting Environments
Initial before *r:
*kremh2- ‘become slack’: Skt. śram- ‘become weary’, Grk. krémamai ‘hang down’
*kreuh2- ‘bloody flesh, gore’: Skt. kravís-, Grk. kréas ‘flesh’ etc.
*kwreih2- ‘obtain by exchange, buy’: Skt. krī∆āÂti, Grk. príamai ‘buy’ etc.
Initial before *l:
*klei- ‘lean, recline’ (intr.): Skt. śri- ‘recline’, Grk. klínō ‘lean’, etc.
*klep- ‘steal’: Lat. clepō, Grk. klépō, TochB kälypi- ‘steal’
Final after *i:
*weik- ‘enter’: Skt. viś- ‘enter’, Grk. oîkos ‘house’, Lat. uīcus ‘clan’ etc.
*weik- ‘separate, select’: Skt. vivékti ‘separate’, Lat. uictima ‘victim’, Gothic weihan
‘consecrate’
*leikw- ‘leave behind’: Skt. ri∆ákti, Grk. leípō, Lat. linquō ‘leave behind’ etc.
*leig- ‘bind’: Lat. ligāre, Alb. lidhem ‘bind’, MLG līk ‘bond’
*weig- ‘give away’: Skt. vij- ‘flee’, OE wīcan ‘yield’, TochAB wik- ‘disappear’
*neigw- ‘wash’: Skt. nij- ‘wash’, Grk. néniptai ‘is washed’ etc.
*leigh - ‘lick’: Arm. lizem, Grk. leíkhō, OIr. ligim ‘lick’ etc.
*steigh- ‘stride (upward)’: Skt. stigh-, Grk. steíkhō ‘advance’, Goth. steigan ‘climb’ etc.
*(s)neigwh- ‘to adhere’: Skt. sníhyati ‘love’, Grk. neíphei, Lith. sniẽga ‘snows’ etc.
Final after *r:
*derk- ‘look at’: Skt. dƒś- ‘see’, Grk. dérkomai ‘look at’ etc.
*perk- ‘fill up’: Skt. pƒk- ‘fill’, OIr. ercaid ‘fills’
*terkw- ‘turn’ (intr.): Hitt. tarku- ‘dance’, Lat. torqueō ‘turn’ (tr.), TochB tetarku ‘turned’
*werg- ‘make’: Av. vərəziieiti ‘does, makes’, Grk. éorge ‘has carried out’ etc.
*h2werg- ‘turn’: Skt. vƒj- ‘turn’, Hitt. ḫurki- ‘wheel’ etc.
*tergw- ‘threaten’: Skt. tarj- ‘threaten’, Grk. tárbos ‘fright’, Lat. toruus ‘glaring’ etc.
*bhergh - ‘high’: Av. bərəz- ‘height’, Hitt. parku- ‘high’ etc.
*bhergh- ‘guard, watch over’: Av. bərəjaiia- ‘honor’, Goth. bairgan ‘preserve’ etc.
Final after *l:
*HmelkÎ- ‘stroke’: Skt. mƒśáti ‘touches’, Lat. mulceō ‘stroke’
*selk- ‘pull, draw’: TochB sälk-, Grk. hélkō ‘pull, draw’
*h2melgÍ- ‘milk’: Grk. amélgō, Lith. mélžu, OE melcan ‘milk’ etc.
*welg- ‘roll’: Skt. válgati ‘looms up, leaps’, OE wealcan ‘roll, press’ etc.
13
Final after *n:
*denkÎ- ‘bite’: Skt. da(μ)ś-, Grk. dáknō ‘bite’
*kÎenk- ‘hang’ (intr.): Skt. śa©k- ‘be in doubt’, Hitt. kank- ‘hang’, Lat. cūnctor ‘hesitate’
*bhengÍh- ‘make firm, massive’: Skt. bahú- ‘thick, many’, Av. bązaiti ‘fixes’, Hitt. panku‘all’, Grk. pakhús ‘thick, massive’ etc.
*gÍhengh- ‘walk’: Lith. žengiù, Goth. gangan ‘walk’ etc.
*h1lengwh- ‘light, nimble’: Skt. ráμhate ‘hurries, runs’, Grk. elaphrós ‘light, quick’, OHG
gi-lingan ‘succeed’
14
References
Bader, Françoise. 1991. Problématique du génitif thématique sigmatique. BSL
86/1.89-157.
Borchhardt, Jürgen, Heiner Eichner, et al. 1997-99. Archäologischsprachwissenschaftliches Corpus der Denkmäler mit lykischer Schrift.
Anzeiger der philosophisch-historischen Klasse, Österreichische Akademie
der Wissenschaften, 134. Jahrgang, 11-96. Vienna: ÖAW.
Carruba, Onofrio. 1978. Il relativo e gli indefiniti in licio. Die Sprache 24.16379.
Čop, Bojan. 1970. Eine luwische orthographisch-phonetisce Regel. IF 75.85-96.
Fortson, Benjamin. 2004. Indo-European Language and Culture. An
Introduction. Malden MA: Blackwell.
Goedegebuure, Petra. forthcoming. The Cuneiform Luwian adverb zanda
“down, (along) with, together, jointly”. 7th International Congress of
Hittitology, Çorum, Turkey.
Gusmani, Roberto. 1961. Il suffisso -tjo- di aggettivi “locali” e la sua diffusione
nelle lingue indoeuropee. AION-L 3.41-58.
Hajnal, Ivo. 1995. Der lykische Vokalismus. Graz: Leykam.
——. 2003. “Jungluwisch” – eine Bestandsaufnahme. In M. Giorgieri et al.
(eds.): Licia e Lidia prima dell’ ellenizzazione, 187-205. Rome: CNR.
Hawkins, J. David. 2000. Corpus of Hieroglyphic Luwian Inscriptions of the
Iron Age. Berlin: de Gruyter.
Hock, Hans Henrich. 2009. Labiopalatalization in Indo-European Languages. In:
S. Jamison, H. C. Melchert and B. Vine (eds.), Proceedings of the 20th
Annual UCLA Indo-European Conference, 69-78. Bremen: Hempen.
Jamison, Stephanie. 1983. Function and Form in the -áya-Formations of the Rig
Veda and Atharva Veda. Göttingen: Vandenhoeck & Ruprecht.
Kimball, Sara. 1994. Loss and retention of voiced velars in Luvian: another
look. IF 99.75-85.
——. 1999. Hittite Historical Phonology. Innsbruck: Institut für
Kloekhorst, Alwin. 2006. Initial Laryngeals in Anatolian. HS 119.77-108.
——. 2008. Etymological Dictionary of the Hittite Inherited Lexicon.
Kümmel, Hans Martin. 2007. Konsonantenwandel. Bausteine zu einer Typologie
des Lautwandels und ihre Konsequenzen für die vergleichende
Rekonstruktion. Wiesbaden: Reichert.
Lipiński, Edward. 2004. Itineraria Phoenicia. Leuven: Peeters.
Lipp, Reiner. 2009. Die indogermanischen und einzelsprachlichen Palatale im
Indoarischen. Heidelberg: Winter.
Meillet, Antoine. 1934. Introduction à l’étude comparative des langues indoeuropéennes.7 Paris: Hachette.
15
Melchert, H. Craig. 1984. Studies in Hittite Historical Phonology. Göttingen:
Vandenhoeck and Ruprecht.
——. 1987. PIE velars in Luvian. In: C. Watkins (ed.), Studies in Memory of
Warren Cowgill, 182-204. Berlin/New York: de Gruyter.
——. 1988. Luvian Lexical Notes. HS 101.211-43.
——. 1989. New Luvo-Lycian Isoglosses. HS 102.23-45.
——. 1994. Anatolian Historical Phonology. Amsterdam/Atlanta: Rodopi.
——. 2004a. A Dictionary of the Lycian Language. Ann Arbor/New York:
Beech Stave.
——. 2004b. A Luwian Dedication. In J. Penney (ed.), Indo-European
Perspectives. Studies in Honour of Anna Morpurgo Davies, 370-9.
Oxford/New York: Oxford University Press.
Morpurgo Davies, Anna. 1986. Fighting, Ploughing and the Karkamiš Kings. In:
A. Etter (ed.), O-o-pe-ro-si. Festschrift für Ernst Risch zum 75. Geburtstag,
129-45. Berlin: de Gruyter.
Morpurgo Davies, Anna, and J. David Hawkins. 1988. A Luwian Heart. In: F.
Imparati (ed.), Studi di storia e di filologia anatolica dedicate a Giovanni
Pugliese Carratelli, 169-182. Firenze: Elite.
Neu, Erich, and Christel Rüster. 1989. Hethitisches Zeichenlexikon. Wiesbaden:
Harrassowitz.
Neumann, Günter. 1991. Hethitisch negna- “Bruder”. HS 104.63-6.
Oettinger, Norbert. 1979. Die Stammbildung des hethitischen Verbums.
Nürnberg: Carl.
Ofitsch, Michaela. 1998. Indogermanischer Grundwortschatz in den
anatolischen Sprachen. Probleme der Rekonstruktion anhand der
Verwandtschaftbezeichnungen. In: W. Meid (ed.), Sprache und Kultur der
Indogermanen. Akten der X. Fachtagung der Indogermanischen Gesellschaft
Innsbruck, 22.-28. September 1996, 421-36. Innsbruck: Institut für
Puhvel, Jaan. 1991. Hittite Etymological Dictionary. Volume 3. Words beginning
with H. Berlin/New York: Mouton de Gruyter.
——. 1997. Hittite Etymological Dictionary. Volume 4. Words beginning with
K. Berlin/New York: Mouton de Gruyter.
Rieken, Elisabeth. 1997. Zu den Reflexen von uridg. *(s)k'el- “springen, eilen”.
HS 110.167-175.
——.
2003.
Hieroglyphen-luwisch
zí+ra/i-la-mi-i
(„SCALPRUM.
ARGENTUM“)su-ha-pa-na-ti: ein Kompositum und eine neue luwischlateinische Isoglosse. HS 116.35-53.
Schaffner, Stefan. 2001. Das Vernersche Gesetz und der innerparadigmatische
grammatische Wechsel des Urgermanischen im Nominalbereich. Innsbruck:
Institut für Sprachen und Literaturen der Universität Innsbruck.
Sihler, Andrew L. 1995. New Comparative Grammar of Greek and Latin. New
York/Oxford: Oxford University Press.
16
Starke, Frank. 1990. Untersuchung zur Stammbildung des keilschrift-luwischen
Nomens. Wiesbaden: Harrassowitz.
Watkins, Calvert. 1987. Two Anatolian forms: Palaic aškumāua-, Cuneiform
Luvian wa-a-ar-ša. In: G. Cardona & N. Zide (eds.), Festschrift for Henry
Hoenigswald, 399-404. Tübingen: Narr.
——. 1999. A Celtic Miscellany. In: K. Jones-Bley et al. (eds.), Proceedings of
the Tenth Annual Indo-European Conference. Los Angeles. May 21-23,
1998, 3-25. Washington DC: Institute for the Study of Man.
Woodhouse, Robert. 1998. On PIE. tectals. IF 103.40-69.
Woudhuizen, Fred. 2010. Towards a Chronological Framework for Significant
Dialectal Tendencies in Indo-European. JIES 38.41-131.
Yakubovich, Ilya. 2010. Sociolinguistics of the Luvian Language.
Zehnder, Thomas. 2010. Die hethitischen Frauennamen. Wiesbaden:
Harrassowitz.
H. Craig Melchert
ECIEC 29
Cornell University
June 19, 2010
The Verbal Prefix “u-” and vs. <ú> Spellings in Anatolian Cuneiform
I. Hittite (and ostensibly Luvian) preverb u- ‘her-’ per Melchert (1994: 104) < *au- (~
Lat. au-ferō ‘carry away/off’) with semantic shift ‘away from’ > ‘towards’ (etymology
already in Hrozný 1917: 1222). Similarly Kloekhorst (2008: 909) < *h2ou with special
treatment in Anlaut vs. retention in eḫu. For semantic shift cf. Hackstein (2007: 138-42).
Per Oettinger (1979: 126 & passim) < *h1/3ou- after Eichner (1973: 54 & passim) without
explicit comparandum (PIE *au- scil. *h2au- seen rather in -ḫu of eḫu ‘come!’), but
Oettinger (2006: 43514) accepts derivation from *au-. Kimball (1999: 225) abandons
derivation < *(e)u- and equation with alleged Arcado-Cypriot ὐ-/εὐ- and views PIE
source as unclear. Kimball (2007:210) cites *(o)u- without comment.
II. Prefix spelled with in some verbs, with <ú> in others (see below). No problem
with assumption of equivalence of two signs (Melchert 1994: 26 w/ refs.; Kimball 1999:
79-80, et al.). But as per Rieken (2005: 538-9 after Eichner), invariance in spellings must
be given equal weight with variance. By this principle overwhelming consistency of 
vs. <ú> spellings in well-attested words argues for linguistically real contrast. E.g., 1x
a-pu-ú-un vs. 150+ a-pu-u-un argues for contrast, not equivalence. Similar arguments by
Kloekhorst (2008: 35-60). Therefore, contrast exists between /u/ (<ú>) and /o/ () in
Hittite. Details of both synchronic and diachronic distribution remain to be determined.
For very provisional formulation (as of 6/19/2010!) see Appendix 1. For evidence that
contrast exists also in Palaic and CLuvian see Appendix 2.
III. Attested spelling of directional prefix u- in Hittite problematic for all formulations:
u-un-ni/au-(i)-ya-
(ú-)uš-ši-yaú-daú-wa-te/aú-e-/ú-wa-
up-pi/a-
“Unmarked” result is clearly /u-/ (<ú>). Others require explanation. For u-un-ni/a- ‘drive
hither’ laxing/lowering before tautosyllabic nasal is plausible (with Rieken 2005: 540-1;
similar Kloekhorst 2008: 54, but with false generalization to position /_nV). No such
solution for u-(i)-ya- (allegedly ‘send here’). Not credible Kloekhorst (2008: 52):
lowering of /u/ to /o/ before “i” (actually yod, as per Kloekhorst 2008: 911). Rieken
ignores word entirely. Likewise problematic is up-pi/a- with no plene spellings.
Kloekhorst (2008: 36) assumes unattested OS spelling would be *ú-up-pi/a- (citing
restriction of plene ú-uš-ši-ya- to OS). Possible but ad hoc (NB total of OS ú-uš-ši-° 3x
vs. post-OS uš-ši-° 12x, but up-pí/a-° 100x, 22x in MS).
IV. Hittite uppi/a- ‘send; bring’
Since Hrozný (1917: 1222) near-universally assumed to contain prefix u- ‘her-’: Götze
(1933: 223 & 1938: 124); Friedrich (1952: 234 ‘(her)schicken’ [sic!]); Oettinger (1979:
489 ‘her-schicken’ [sic!]); Melchert (1994: 104&149); Yakubovich (2005: 245 [implicit
in gloss ‘to bring’]); Kloekhorst (2008: 921-2 ‘send (here)’ [sic!]). Objection of Pedersen
(1938: 116-7) wrongly ignored: actual usage shows no directionality whatsoever!
2
(1) HKM 18:18-20 (MH/MS)
nu=mu kā katti=mi ÉRIN.MEŠ KUR.UGU ÉRIN.MEŠ KUR URUIšḫūpitta kuiški
n=an=ta uppaḫḫi ‘There are some troops of the Upper Land and I. here with me. I’ll
send them to you.’
(2) KBo 5.4 Vo 22 (NH; Treaty with Targasnalli)
nu=mu ḫatrāši ÉRIN.MEŠ=wa=mu ANŠE.KUR.RA.MEŠ uppi nu=tta ÉRIN.ME.EŠ
ANŠE.KUR.RA.ME.EŠ uppaḫḫi ‘And you write me: “Send me troops and horses!” I will
send you infantry and horses.’
(3) KBo 3.6+ iii 77 (NH; Apology of Hattušili)
n=an ēppūn n=an=kan A.AB.BA tapuša uppaḫḫun
‘I seized him and sent him alongside the sea.’
See also HKM 25:22-23, HKM 30:8-10. For likely OH existence see KBo 20.67 iv 33
(invocation to the mountains, CTH 591, OH/NS).
(4) KBo 17.65+ Ro 38-39 (MH/MS; Birth Ritual)
(‘But how the festival of birth (is)—how they perform the festival when she gives birth,’)
[(n)=aš GIŠkurtaš iy]anza n=aš URUKizzuwatna nu=mu=kan EZEN KAxU-it [ŪL karta
n]=an apēz uppaḫḫi
‘[It is m]ade [(as) a kurta-tablet], and it is in Kizzuwatna. I [do not know] the festival
orally [by heart]. I will fetch it from there.’
Restorations based on parallel passage ibid. Vo 45-46, which has rather udanzi ‘they will
bring here’. See Beckman 1983: 136-7 and 161-4 with differing details.
(5) KUB 12.58 ii 36-42 (MH/NS; Ritual of Tunnawi)
(‘The ritual client goes to bathe,’) n=ašta MUNUSŠU.GI 9 GIŠGA.ZUM ŠA GIŠTÚG anda
upp[āi] GIŠšēn!an IM-aš anda uppāi
‘The “old woman” fetches in nine combs of boxwood. She fetches in a figurine of clay.’
Goetze (1938: 14-15) renders ‘brings in’, but the practitioner has been in the ritual space
in the entire preceding context.
(6) KBo 15.37 ii 49-59 (MH/NS; (ḫ)išuwa-Festival)
(‘When it dawns on the eighth day, the large loaf of cheese and the leavened bread of
three parīsu-measures of flour which (are) broken for the gods of the fathers and which
are placed back on the sacrificial tables,’) n=aš arḫa paršulānzi namma ANA PANI
DINGIR.MEŠ ḫūmandaš kuwapiya 1 paršullin GA.KIN.AG 1 paršullin NINDA=ya
uppiyanzi n=uš PANI DINGIR.MEŠ zikkanzi
‘They crumble them and then bring one morsel of cheese and one morsel of bread before
each of all the gods (lit. before all the gods in each place) and place them before the
gods.’
Contra Carruba (1966: 2020), Oettinger (1979: 489) and Nakamura (2002: 159) no basis
for separate verb ‘lift, raise’, but ‘send’ (Güterbock & Hoffner 1995: 193) also unlikely.
3
Given total lack of evidence for presence of prefix ‘her-’ but strong goal orientation (all
80+ examples with ‘send’ in sufficient context have explicit or implied goal), perhaps
deadverbial < *úpo as in Skt. úpa ‘up to’ (anticipated by Oettinger 1979: 489-90, but
with false semantics for non-existent homonym ‘raise, lift’). For process see Melchert
(2009) and compare rare transitive use of English ‘near’ = ‘bring near (to)’ (Ruskin &
Wilde). For unexpected stem uppi- compare opposite discrepancy in Hitt. šanna‘conceal’ < *s(h1)o vs. Latin sine and OIr. sain- < *s(h1)i.
V. CLuv. (u-)up-pa- ‘carry’ and HLuv. (CAPERE)u-pa- ‘carry (off)’
Pace Kloekhorst (2008: 922) basic sense ‘carry’ of CLuvian not in doubt, but his
rejection of equation with Hittite uppi/a- is probably correct:
(7) KBo 13.260 iii 13-15 (Birth Ritual)
ā=tta ādduwan=za pariyan adduwaliyan wattaniyan uppannandu
‘Let them carry the evil over to an evil land.’
Thus with Yakubovich (2010: 237).
(8) KBo 13.260 iii 16-20 (Birth Ritual)
zīla=pa=tta za-x-x-ti DUMU.LÚ.ÙLULU-ni ŠUM-ŠU ḫalzāiḫuidumar=ša ūppannandu
wayaḫi=ša ḫaddulaḫi=ša annarumāḫi=ša
‘Henceforth let them bring to this(?) human—one calls out his name—life, w., health
(and) vigor.’
(9) KUB 35.88 iii 11-12 (Birth Ritual)
upatta=pa=wa=du šarriyanin 2-šu 9-u[(n-)za] anta=wa=aš=ta walluna<š>šan wāni
uppanta
‘She furnished to her twice nine s., and they carried them in to the midwife (lit. woman of
lifting).’
See also KUB 35.107 iii 21 (anda uppanta) and KUB 25.39 iv 16 (ūppadda).
NB spelling with u-!
(10) KARKAMIŠ A 11b+c §13
(‘I devastated those countries,’) *a-wa/i-ta (SCALPRUM.CAPERE2)u-pa-ní-zi a-tá
(“CAPERE2)u-pa-ha ‘I brought in/collected trophies/spoils’ (‘and I came up from those
countries in glory.’)
(11) KARKAMIŠ A2+3 §7
*a-wa/i-ta *a-mi-ya-za LITUUS+AVIS(-)ta-ni-ya-za REGIO-ni-i a-tá (DEUS)BONUS-na
(DEUS)VITIS(-)ti-PRAE-ya-ha ARHA (CAPERE2)u-pa-ta
‘He (Tarhunza of Carchemish) in my days carried off the grain-god and the wine-god into
the country.’
(12) İSKENDERUN §§3-4
a-wa/i za-na-i (“*255”)ka-ru-na-na (“CAPERE”)u-pa-ha-i wa/i-tu-u-ta-i 4xMILLE
4xCENTUM a-ta (“CAPERE”)u-pa-ha zi-i-na (“*256”)zi-pa-ta-na-ti
‘I carried off this granary and I collected for it 4,400 with this z-measure.’
4
For (“CAPERE”)u-pa- as ‘bring’ vs. (PES)u-pa- ‘dedicate, furnish’ see Melchert 2004
contra Yakubovich 2005, but directionality implied there is dubious. Note further figura
etymologica /upani-/ /upa-/ ‘carry (off) spoils/trophy’. Contra Melchert (2004: 372)
‘spoils’ are not inherently ‘brought (in)’, but rather ‘carried off’. This figure and the
otherwise unexplained determinative CAPERE(2) ‘take’ argues for original meaning
‘carry off, remove’, bleached to simply ‘carry’.
Spelling of CLuvian cognate with u-up- points to /o-/, thus combined evidence argues
that this is the reflex of *au- ‘off, away’. Root *pa- is probably *(s)peh2- ‘set in (violent)
motion, draw’ > Grk. σπάω ‘tear/pull’, Arm. hanem ‘pull; take away; lead’: see García
Ramón 2009. /o:ppa-/ *‘take away, remove’ whence ‘carry’ < virtual *au-(s)p(e)h2-. NB
inflectional class unknown, but third plural *au-(s)ph2énti would give /o:ppanti/ in either
mi-or ḫi-verb.
VI. Hittite uye/a- ‘send, drive’
Universally assumed to be u-ye/a- ‘her-schicken’ vs. pe/i-ye/a- ‘hin-schicken’, but again
no evidence for speaker-oriented directionality of first. Pedersen (1938: 198); Friedrich
(1952: 232) ‘schicken’ (eigentlich ‘herschicken’); jagen’; Kloekhorst (2008: 910) ‘send
(here).
(13) KBo 3.40b:9-10 (OH/NS; narrative)
am[(mug=a dU-aš)] DINGIR pišeneš [LUG]AL-i uyēr īt=wa LÚ.MEŠGAL.GAL wemiy[(a)]
‘But me the male gods of the Storm-god sent to the king (saying): “go and find the great
men!”’
(14) KUB 33.5 ii 4 (OH/NS; Telipinu, 2nd version)
d
MAḪ-aš NIM.LÀL-an uyēt īt=za x[…]x dTelipinun zik šanḫa
‘Hannahanna sent the bee (saying): “Go [ ] and you search for Telipinu!”’
Duplicate KUB 33.4 ii 17 has piyēt! KUB 33.8 ii 22 in similar context has uyēt.
NB both uiye/a- and pe/iye/a- are securely OH: x-x-x[ ]iššāz LUGAL-i atti=mi u-i-x[ ]
(KUB 26.35:3) and [°-a]n arḫa imma piyēzzi (KUB 36.106 Ro' 5).
(15) KUB 23.72+ Vo 22-23 (MH/MS; Mida of Pahhuwa)
ŠA LÚ.KÚR=ma=za=kan LÚ ṬEMI INA KÁ=ŠUNU ŪL tarnanzi kuiš=a=šmaš LÚ.KÚR
LÚ ṬEMI uyēzzi a-pí-°[…n]=an ANA dUTU-ŠI uppianzi
‘They shall not let the messenger of an enemy in their gate. Whatever enemy sends them
a messenger, […] and they shall send him to His Majesty.’
(16) KUB 14.3 i 6-9 (NH; Tawagalawa letter)
nu=m[u U]N-an IGI-anda uiyat ARAD-anni=wa=mu dā nu=wa=mu LÚtuḫkantin uiya
nu=wa=mu ITTI dUTU-ŠI uwatezzi nu=šši LÚTARTENU uiyanun
‘He sent a person to meet me (saying): “Take me into servitude! Send the crown-prince
to me, and he shall bring me to Your Majesty.” So I sent the crown-prince to him.’
(17) KBo 4.8 ii 13-14 (NH; Tawananna Affair)
kī=ya=an 1-an dammešḫanunun IŠTU É.GAL-LIM=pat=kan kuit katta uiyanun
‘And I also did her this one harm, that I sent her down from the palace.’
5
(18) KUB 1.1+ iii 10-11 (NH; Apology of Hattušili)
URU
Ḫakpiššaš=ma kururiyaḫta [nu] LÚ.MEŠ GašgaḪI.A uiyanun n=an IŠTU NÍ.TE=YA
SIG5-aḫḫun
‘But Hakpis became hostile, [so] I drove out the Kaskeans and on my own put it (the city)
in order.’
Thus with Otten (1981: 17) contra van den Hout (2003: 199). Confirmed by parallel KUB
1.7 ii 6, which adds arḫa. NB: no space in KUB 1.1 iii 11 for any restoration beyond
[nu]. For sense ‘banish, drive away’ for arḫa uiya- see also KUB 14.8 Vo 17-18 and
Plague Prayers passim.
NB that sense ‘banish, drive (away)’ shows that original sense did not include control by
subject of object. Thus meaning ‘send’ is secondary. Preverb arḫa also not required for
sense ‘drive away, banish’. Argues for original *au-(Hi)Hyeh1- *‘cast away’. Bleached
generally to ‘drive, send’, hence creation of new univerbation with *pe- for ‘send (off)’.
VII. Reconsideration of prefix “u-”
1. Contrast of CLuvian /o:-/ in u-up-pa-‘carry (off)’ vs. /u:-/ in ú-pa- ‘dedicate, furnish’
(see Melchert 2004), also borrowed into pre-Hittite in ú-ba-ti- ‘land-grant’ and of Hittite
u-(i)-ya- ‘send, drive (away)’ vs. ú- ‘her-’ in ú-da- ‘bring’, ú-wa-te/a-‘bring’ (people
/animals), and ú-e-/ú-wa- ‘come’ argues for *au- ‘away’ in the former, but something
else in the latter.
2. Luvian awi- ‘come’ beside Hitt. ú-e-/ú-wa- seems to exclude zero-grade *u- and
spelling with ú- makes *ou- unlikely—assumption by Kloekhorst of different outcome of
*h2ou “in isolation” vs. usual *ou > /o/ () totally ad hoc. Thus *eu-, but from what?
A. Rehabilitate Kimball’s comparison with Arcado-Cypriot ὐ-/εὐ-? Difficult in view of
Strunk 1986. Compare very cautious remarks of Widmer 2008: 623 with note 17.
B. Derive from far-deictic *e/ow(o)- with same assumed semantic “reversal” as cited in I
above? NB vs. Av. auua-, OP ava- ‘that, jener’ mixed deixis in Slavic: Serbo-Croatian
óvâj, Macedonian ovoj ‘this’, but Polish ów ‘that’. Note especially USorbian
jow/LSorbian how ‘here, hither’. And/or the “particle” of Skt. asāú, amúm ‘yonder’ with
similar semantic development?
6
Appendix 1
Provisional Analysis of /u/ vs. /o/ as of 6/19/2010
1. For earlier literature see Rieken 2005: 537 with refs. to Weidner, Held & Schmalstieg,
Eichner, and Hart. The following represents blend of Rieken 2005 and Kloekhorst 2008:
35-60 with personal views. Note following premises: (1) if contrast is valid, also exists in
Palaic and CLuvian (see Appendix 2); (2) in poorly attested words with alternate and
<ú> spellings subjective treatment is unavoidable (each scholar will analyze in terms of
personal overall schema); (3) even in cases of secure root etymologies proving that given
word must be reflex of specific ablaut grade *eu, *ou, *u often difficult; (4) by any
analysis some data remains recalcitrant.
2. Prehistoric *u (NB *ū = PIE *ū/uh1, not *uh2/3) regularly > /u(:)/ spelled <ú>:
ḫa-tu-(ú)-ga/i- ‘frightful’ < *h2túgo- (but only 1x with -ú-)
i-ú-kán ‘yoke’ < *yugóm
ku-ú-na-aš ‘dog’ (GSg.) < *kúnos (regular contra Kloekhorst 2008: 54)
mi-(i)-ú- ‘soft, mild’ < virtual *mih1-ušu-ú-e- ‘push away, reject’ < *suh1-ye/ošu-u-ú ‘full’ (N-ASgNt) < *séuh3-u (last vowel)
ú-ug ‘I’ after *tū (NB: cannot reflect *túh2!)
ú-me-(e)-ni ‘we see’ and ú-wa-an-zi ‘they see’ < *u-wéni and *u-énti
ú-na-at-ta-al-la- ‘merchant’ pre-Hittite loanword < Luvian *us-no+
ú-nu- ‘adorn’ < *u-n(é)uú-ra-a-ni ‘burns’ < *ur-óri
aš-šu-ú-ul, im-mi-ú-ul, iš-ḫi-ú-ul, ták-šu-ú-ul (1x), wa-aš-du-ú-li (1x) syncopated from
*-úlom as per Rieken (2008: 252-3) contra Rieken (2005: 543-4)
3. *eu/ēu > /u:/ spelled <ú>:
i-ú-uk ‘brace, pair’ < *yéug (Kloekhorst 2008: 423 after Rieken 1999: 61f.)
ku-ú-ša- ‘daughter/son-in-law; bride’ < *géuso- *‘chosen one’ (Rieken 1999: 257)
ḫu-e-nu-ú-ut, wa-aḫ-nu-ú-mi < *-néuši-(i)-ú-(n)- ‘god’ < *dyēu(but u-up-zi ‘rises’ (of sun) then cannot reflect *(h1)éup-ti, as per Oettinger 1979: 233,
Melchert 1994: 104; ku-ú-uz-za ‘wall’ also problematic, since as animate t-stem, hardly <
*ghéut-s with Kloekhorst 2008: 57)
4. Prehistoric *u secondarily lowered to /o(:)/:
A. Before or after *h2 and *h3:
a-aš-šu-u ‘goods’ < *h1ósuh2
ḫu-u-ma-an-t° ‘all’ < *h2u-h1m-ent- (etym. with Kimball 2007: 10)
nu-u-wa ‘still, yet’ < *nú-h2o
šu-u- ‘full’ < *séuh3-
7
B. Before r except word-initially (see ú-ra-a-ni above!)
iš-nu-u-ra- ‘kneading pan’ < *is-nú-ro- (vs. Palaic ta-šu-ú-ra-!)
ú-e/i-šu-u-ri-ya- ‘twist’ < *weis-ú-roku-u-ur-ka- ‘foal’ (cf. Grk. κύρνος) and šu-u-ur-ka/i- ‘root’ (cf. Latin surculus)
5. *ou/au > /o(:)/ spelled 
mu-u-ga-i- ‘incite; goad’ < *móuko- (Melchert, FS Archi, revising Kloekhorst)
u-(i)-e-/ya- ‘drive (away), send’ < *au-Hyeh16. PIE *w between consonants and word boundary > prehistoric *woR > /o(:)R/ spelled
<(Cu-)u-uR>
a-ni-u-ur, a-ni-u-r° ‘ritual’ with postconsonantal /-o:r/ from ḫenkur etc.
an-tu-u-ri-ya- ‘interior’ < *en-dhwpíd-du-u-la- ‘loop’ hypostasized from *peth2w *‘spreading out’
u-ur-ki- ‘track’ < *wgiu-ur-re-er < *wh1ēr
[But sequence *wCC > warCC with two consonants as in d(u)warne/a- < *dhwnh1énti; Kloekhorst 2007: 456-7.]
7. In word-final *-óm(s)# the *o did not lower to merge with /a/ and appears preserved as
/-o:n/ in ku-u-un and a-pu-u-un and the plurals ku-u-uš and a-pu-u-uš (cf. Melchert 1994:
186-7 and Kloekhorst 2008: 54&56). The vocalism of uni has been remade after the other
two demonstratives, just like *ési Æ aši after kāš and apāš (Kloekhorst 2008: 220-21).
8. Open Issues: (1) quality of anaptyctic “u”; (2) further lowerings in NH before n and r;
(3) non-pronominal acc. pl. in -ú-uš; (4) given problems for *eu > /u:/ under 3. above,
rather give up direct equation of Hitt. ú-e- with Luvian awi- and assume zero-grade for
Hittite, thus returning to view that all *Vu > /o(:)/?
________________________________________________________________________
Appendix 2
Spellings with and <ú> in Palaic
1. <ú> (/u:/) < *u
nu-ú ‘now’ < *nu
tu-ú ‘to thee’ < *tŭ
mu-ú-ši ‘you are satiated’ (P2Sg with Yakubovich, JANER 5.117 contra Melchert, KZ
97.35) < *mús-si, to imitative root of Grk. μύω ‘be satiated’, ἀμυστί ‘in one gulp’
(Eichner)
šu-ú-na-at ‘has filled’, šu-ú-na ‘fill!’ < *sú-ne-h3(-t) with accent retraction
ta-šu-ú-ra- ‘offering table’ (loc. sg.) *dhh1s-ú-ro-; cf. Hittite išnura- ‘kneading pan’,
but latter regularly iš-nu-u-ra- (including OS!)
2. <ú> /u(:)/ < *eu (probably shortened when unaccented)
i-ú ‘come!’ per Melchert (1994: 212) ~ Hitt. eḫu, but required totally ad hoc loss of -ḫ-.
Rather with Oettinger (1979: 536) = reverse of Hitt. ú-e- and Luvian awi-, thus virtual
*h1ei-eu (but see remark under Appendix 1, 8.).
8
3. /o:/ < *uh2
wa-šu-u-(ḫa) ‘good ones’ < *wósuh2 (with shift of accent with added clitic)
4. /o:/ < *ou (?)
a-ru-u-na-am-pí per Melchert KZ 97.39-40 acc. sg. ‘high’ to aru-, thus probably
generalized weak stem *h1r-ów- (cf. CLuvian)
5. /o:/ < secondary ḫu < *gwhu
a-ḫu-u-na ‘to drink’
Spellings with and <ú> in CLuvian
1. <ú> (/u:/) < *u
a-ru-ú-da-ti abl.-inst. to arut(i)- ‘wing’ < *ar-ú-ti- *‘joint’ (Melchert, HS 101.224f.)
i-ú-na-ḫi-it- ‘ability to walk; mobile wealth’; built on infinitive i-ú-na ‘to go’ < *h1y-uneh2 [but also pa-aš-šu-u-na ‘to swallow’!]
ma-ad-du-ú-wa-ti ‘with wine’ < *médhudu-ú-pí-/du-ú-pa-i- ‘strike’ < *(s)tubhéye/o- (must assume zero-grade iterative type,
contra Melchert, FS Puhvel 135)
du-ú-ur, du-ú-n° ‘urine’ < virtual *sh2ú-n- and zero-grade generalized to nom.-acc.
NB *kwrV- leads to ku-ú-rV-: ku-ú-ru-na ‘to cut’, ku-ú-ra-am-mi ‘cutting’ (dat.-loc. sg.),
ku-ú-ri ‘cut!’ (to denominative kūri-/kūrai-)
2. <ú> (/u:/) < contracted *(u)wa
ḫi-(i)-ru-ú-un and ḫi-(i)-ru-ú-t° ‘oath’ (1x with -u- vs. several dozen with -ú-); long
vowel contracted from *-uwan, as per Melchert, GsForrer, after Watkins, FS Rix)
lu-ú-un-ni ‘we take’; -unni contracted from *-wanni < *-wéni by “Čop’s Law”
[but also pi-u-un-ni ‘we give’!]
ú-la-an-ti- ‘dead’ (2x) syncopated from walant(i)-; also in ú-la-an-ta-al-li-ya- (1x, but
latter also 3x with u-la-°!)
ú-li-ip-na/i- ‘wolf’ (2x) syncopated from wa-li-ip-na[but NB 1x u-na-at-ti-iš ‘woman’ syncopated from wa-na-at-ti-!]
3. Diphthong /aw/ spelled -Ca-(a)-ú- (cf. for Hittite Kloekhorst 2008: 42-44, 59)
i-ik-ku-ú-na-ú-na-aš-ši- ‘of anointing’ *īkkūnawar/*īkkūnaungul-za-a-ú-na ‘to draw’
kur-ša-ú-na-an-ti-in-zi ‘islands’ (erg. pl.)
pa-tal-ḫa-ú-na ‘to fetter’ [vs. unexpected la-la-u-na ‘to take’ in same ms.—interference
from putative NH change claimed by Kloekhorst 2008:44?]
also in secondary diphthongs /a(:)w/
ḫi-iz-za-ú-un-ni ‘we fetch’; -unni contracted from *-wanni < *-wéni by “Čop’s Law”
la-ú-na-i- ‘wash’ (cf. la-ḫu-ni-)
9
and unexplained Pret3Pl in /-a:wnta/
i-ik-ku-na-a-ú-un-ta ‘anointed’
na-ak-ku-uš-ša-a-ú-un-ta ‘furnished a scapegoat’
wa-a-ar-ma-a-ú-un-ta ‘?’
[vs. unexpected tar-ra-wa-u-un-ta ‘?’!]
4. <ú> /u:/ < *eu (or are first and third examples from zero-grade??)
nu-ú-(t°)- ‘assent, approval’ < *néu- (Rieken 2006: 294)
ú-ut-ti-iš ‘you (sg.) drink’ < *egwtis via secondary diphthong *eu
ú-pa- ‘dedicate’ (and ú-ba-ti-it- ‘land-grant’) < *eu-bhag- (Melchert 2004 contra
Yakubovich 2005)
5. /o:/ < *ou/au
*a-ru-u-wa ‘high’ (N-APlNt as adverb ‘highly, very’) (5x vs. 2x a-ru-ú-wa!) likely
from generalized weak stem *h1róumu-u-wa- ‘overpower’ < *móuh1/3- (hi-verb!) (and derivatives)
zu-u-wa- ‘food’ < *gyóuh1/3-o- (tómos noun to root of ‘chew’)
6. /o:/ < *u next to *h2/3
ḫu-u-t/da- ‘alacrity’ < *h2úh1-to- (with Eichner & Starke to root of ‘run’ *h2weh1-; with
accent retraction in substantivization)
ḫu-u-tar-la-a- ‘slave, servant’ (from preceding)
NB: no spellings †ḫu-ú- anywhere in CLuvian vs. 20+ lexemes with sequence ḫu-u.
7. /o:/ < *u before tautosyllabic nasal(?)
u-un-za-~u-za-aš ‘you (pl.)’ generally taken as *us- with secondary *uns- after first
plural *ans- < *s- (but NB very problematic hapax na-a-nu-ú-un(-pa) where plene
of second syllable is unexpected in view of na-a-!)
8. Open issues: popular derivation of *DUGšu-ú-wa-at-ra/šu-ú-wa-at-n° (vessel) from šuu-wa- ‘fill’ < *s(e)uh3- now dubious. NB CLuvian spelling of verb matches Hittite;
spelling of pu-ú-wa ‘formerly’ contrasts with pu-u-wa- ‘crush’!
References
Beckman, Gary. 1983. Hittite Birth Rituals (StBoT 29). Wiesbaden: Harrassowitz.
Carruba, Onofrio. 1966. Die Beschwörungsritual für die Göttin Wišurijanza (StBoT 2). Wiesbaden:
Harrassowitz.
Eichner, Heiner. 1973. Die Etymologie von heth. mehur. MSS 31.53-107.
Friedrich, Johannes. 1952. Kurzgefaßtes hethitisches Wörterbuch. Heidelberg: Winter.
Garcí-Ramón José Luis. 2009. Idg. *(s)peh2-, ‘in (heftige) Bewegung setzen, ziehen’: Ved. pā3, heth.
pipp(a)-ḫḫi und gr. σπάω, arm. hanem. In Rosemarie Lühr and Sabine Ziegler (eds.), Protolanguage
and Prehistory. Akten der XII. Fachtagung der Indogermanischen Gesellschaft, vom 11. bis 15.
Oktober 2004 in Krakau, 134-48. Wiesbaden: Reichert.
Götze, Albrecht. 1933. Über die Partikeln -za, -kan und -šan der hethitischen Satzverbindung. ArOr 5.1-38.
——. 1938. The Hittite Ritual of Tunnawi. New Haven: American Oriental Society.
Güterbock, Hans G. and Harry A. Hoffner Jr. 1995. The Hittite Dictionary of the Oriental Institute of the
University of Chicago. Volume P. Fascicle 2. Chicago: The Oriental Institute.
10
Hackstein, Olav. 2007. Ablative Formations. In Alan Nussbaum (ed.), Verba Docenti. Studies in historical
and Indo-European linguistics presented to Jay Jasanoff by students, colleagues, and friends, 131-53.
Ann Arbor/New York: Beech Stave.
van den Hout, Theo. 2003. Apology of Ḫattušili III (1.77). In William Hallo and K. Lawson Younger
(eds.), The Context of Scripture. Volume I. Canonical Compositions from the Biblical World, 199204. Leiden/Boston: Brill.
Hrozný, Bedřich. 1917. Die Sprache der Hethither. Leipzig: Hinrichs.
Kimball, Sara. 1999. Hittite Historical Phonology. Innsbruck: Institut für Sprach-wissenschaft der
Universität Innsbrück.
——. 2007. Hittite ḫūmant- ‘all, entire, each’. In Alan Nussbaum (ed.), Verba Docenti. Studies in historical
and Indo-European linguistics presented to Jay Jasanoff by students, colleagues, and friends, 201-12.
Ann Arbor/New York: Beech Stave.
Kloekhorst, Alwin. 2007. The Hittite Syllabification of PIE *CuR and *KR. In Detlev Groddek and
Marina Zorman (eds.), Tabularia Hethaeorum. Hethitologische Beiträge Silvin Košak zum 65.
Geburtstag. Wiesbaden: Harrassowitz.
——. 2008. Etymological Dictionary of the Hittite Inherited Lexicon. Leiden/Boston: Brill.
Melchert, H. Craig. 1994. Anatolian Historical Phonology. Amsterdam/Atlanta: Rodopi.
——. 2004. A Luwian Dedication. In John H. W. Penney (ed.), Indo-European Perspectives. Studies in
Honour of Anna Morpurgo Davies, 370-79. Oxford/New York: Oxford University Press.
——. 2009. Hittite ḫi-verbs from Adverbs. In Rosemarie Lühr and Sabine Ziegler (eds.), Protolanguage
and Prehistory. Akten der XII. Fachtagung der Indogermanischen Gesellschaft, vom 11. bis 15.
Oktober 2004 in Krakau, 335-9. Wiesbaden: Reichert.
Nakamura, Mitsuo. 2002. Das hethitische nuntarriyašḫa-Fest. Leiden: Nederlands Instituut voor het Nabije
Oosten.
Oettinger, Norbert. 1979. Die Stammbildung des hethitischen Verbums. Nürnberg: Carl.
——. 2006. Hethitisch watku- ‘(ent)springen’ und kymrisch godeb ‘Versteck’. Balkansko Ezikoznanie
45/3.433-6 (à la mémoire du Prof. Dr. Georgi T. Rikov).
Otten, Heinrich. 1981. Die Apologie Hattusilis III. Wiesbaden: Harrassowitz.
Pedersen, Holger. 1938. Hittitisch und die anderen indoeuropäischen Sprachen. Copenhagen: Munksgaard.
Rieken, Elisabeth. 1999. Untersuchungen zur nominalen Stammbildung des Hethitischen. Wiesbaden:
Harrassowitz.
——. 2005. Zur Wiedergabe von hethitisch /o/. In Gerhard Meiser and Olav Hackstein (eds.),
Sprachkontakt und Sprachwandel. Akten der XI. Fachtagung der Indogermanischen Gesellschaft. 17.23. September 2000, Halle an der Saale, 537-49. Wiesbaden: Reichert.
——. 2006. Zur Etymologie von luwisch nū(t)-. In Gerd Carling (ed.), GIŠ.ḪURgul-za-at-ta-ra. Festschrift
for Folke Josephson, 284-97. Göteborg.
——. 2008. The origin of the -l genitive and the history of the stems in -il- and -ul- in Hittite. In Kathleen
Jones-Bley et al. (eds.), Proceedings of the Nineteenth Annual Indo-European Conference, Los
Angeles, November 3–4, 2007, 239-256. Washington DC: Institute for the Study of Man.
Strunk, Klaus. 1986. Kypr. (ε)ὐ für ἐπί: eine vox nihili? In Annemarie Etter (ed.), o-o-pe-ro-si. Festschrift
für Ernst Risch zum 75. Geburtstag, 253-69. Berlin/New York: de Gruyter.
Widmer, Paul. 2008. Drei griechische *-ō-Stämme. In Brigitte Huber, Marianne Volkart und Paul Widmer
(eds.), Chomolangma, Demawend und Kasbek. Festschrift für Roland Bielmeier zu seinem 65.
Geburtstag, 615-30. Halle: International Institute for Tibetan and Buddhist Studies.
Yakubovich, Ilya. 2005. Carian monument. In Nikolai N. Kazansky et al. (eds.), HƒdāÂ mánasā: Сборник
статей к 70-летию со дня рождения профессора Леонарда Георгиевича Герценберга, 240-51.
St. Petersburg: “Nauka”.
——. 2010. Sociolinguistics of the Luvian Language. Leiden/Boston: Brill.
H. Craig Melchert
Università di Pavia
October 11, 2012
The Middle Voice in Hittite and PIE
I. Preliminaries
A. Middle voice widely defined as expressing “subject affectedness”. Typical is Lyons (1969:
373), cited by Kemmer (1993: 1): the middle voice indicates that “the ‘action’ or ‘state’ affects
the subject of the verb or his interests.” Similar Neu (1968b: 1-2), citing Brugmann and Indian
grammarians, Weiss (2009: 380) et al. But Kemmer (1993: 3 et passim) subsumes under
“relative elaboration of events” and degree of transitivity (one- and two-participant events).
B. Fundamental sense realized in various ways
1. Oppositional Middle
a. Reciprocal: NB distinction by Kemmer (1993: 96-108) between “naturally reciprocal” (e.g.
Skt. spárdhante ‘they quarrel/fight’) and “prototypically reciprocal” (e.g. zaḫḫanda ‘they hit
each other’ vs. zaḫḫanzi ‘they strike’).
b. Self-interest (“indirect middle/reflexive”): Skt. yájate ‘sacrifices for one’s own benefit’ vs.
yájati ‘sacrifices (for another)’.
c. Passive: universally recognized to be rare in PIE, fully developed only in individual languages.
Expression of agent with finite passive extremely rare even when passive becomes productive.
2. Media tantum (“deponents”): classes and labels after Kemmer 1993: 16-20
a. Body care/grooming: Lat. lauor ‘I bathe’ etc. (for arguments against being oppositional “direct
reflexives” see Kemmer 1993: 53-67 and Klaiman 1992: 45-6).
b. Non-translational motion: Hitt. ḫāliya ‘bows’ (intr.), Lat. (re)uertor ‘I turn’ (intr.).
c. Change of body posture: Hitt. eša ‘sits down’ etc.
d. Translational motion: Grk. ἔρχομαι ‘I come/go’ etc.
e. Mental states (emotion, emotional speech, perception, cognition): Lat. īrāscor ‘I am angry’,
Lat. queror ‘I complain’, Grk. δέρκομαι ‘I look at’, Skt. mányate ‘thinks’.
f. “Spontaneous” events (change of state without agent): Hitt. kiš- ‘become’ etc. (but NB
admission by Kemmer 1993: 147 used also for instances where external agent exists, but is
pragmatically deemphasized).
NB general problem that virtually all classes under media tantum also expressed by ordinary
actives!
II. Hittite Medio-Passive
A. Assume that -a(ri) vs. -t(t)a(ri) reflects renewal of PIE medio-passive in *-o by *-to and *-oto (Watkins 1969: 50-51, 84-5, Jasanoff 2003: 50-51, et al.), not separate voice “Stativ” vs.
medio-passive (Oettinger 1976, 1993, Rix 1988, et al.). But main problem is shared: description
of original function of forms in *-o and *-to is fundamentally false.
2
B. Evidence for basic types
1. Oppositional Middle
a. OH zaḫḫanta ‘they hit each other’ (NB no -za; not natural reciprocal). Also once ŠUaz=at=kan appant[at] ‘took each other by the hand’ without -za; usually renewed with reflexive
-za; thus no sure instance of “natural reciprocal” (NH zaḫḫiyawaštati ‘let us fight’ may be
merely detransitive form of transitive m-p ‘fight (someone)’, but likely existed. Example with
ḫanna- ‘quarrel, contest at law’ (Neu 1968a: 39) with -za probably renewal of true reciprocal
middle.
b. No examples of medio-passive to express self-interest. Always expressed with active
and reflexive -za (cf. Neu 1968b: 107). Most likely an archaism.
c. Passive: rare in OH (see e.g. Neu 1968b: 5&110-12 with rare OH example ḫullattati…
ḫuladaru ‘was struck…shall be struck’). Becomes very productive in later Hittite (over 50 exx.),
but NB almost always with -(a)tta(ri), very rarely with -a(ri) (5 exx.). Some cases of contrast:
OH ḫuettiyati ‘drew’ (trans.) vs. NH EGIR-pa ḫuettiyattat ‘(the legal case) was renewed’ (lit.
‘pulled (out) again’).
2. Media tantum
a. Body care/grooming: warp- ‘bathe’ (Imv2Sg warput w/o -za!), waššiya- ‘dress (oneself)’
(Imv3Sg waššiyattaru ‘let him put on’ w/o -za contra Neu 1968a: 193; Pret1Sg waššiyaḫḫaḫat ‘I
dressed’ w/ -za). Other verbs attested in m-p only renewed with -za: arra- ‘wash’, unu(wa)‘adorn oneself’. NB also same usage with “direct reflexives” (i.e. m-p plus -za): aniya- ‘ritually
treat oneself’, aršiya- ‘indulge oneself’, munnāi- ‘hide oneself’ (vs. m-p alone ‘disappear’!),
parkunu- ‘purify oneself’, daššanu- ‘strengthen oneself’.
b. Non-translational motion: ḫaliya- ‘bow’, lag- ‘stagger, fall’, nē(y)a- ‘turn’ (intr.). NB all
in -a(ri). Further in -(a)tta(ri): ḫa(i)nk- ‘bow’, mūriya- ‘crouch’ (or sim.), wēḫ- ‘spin, move
about’ (vs. wēḫta ‘pass/go away’).
c. Change of body posture: eša ‘sits down’ (OH!); NB also stative body posture arta(ri) ‘stands’
and kitta(ri) ‘lies’.
d. Translational motion: iya- ‘be walking’ (NB atelic!), ḫu(i)ya- ‘be running’ (vs. active ḫi-verb
originally telic ‘flee’), wēḫ- (wēḫta) ‘go/pass away’.
e. Mental states: lēlaniya- ‘become angry’, waršiya- ‘calm down’; walluške- ‘boast’ (but NB
always plus -za, so prob. direct reflexive *‘exalt oneself’). NB lack of verbs of perception or
cognition. My or may not be accidental gap.
f. “Spontaneous” events: very numerous for verbs in -a- (ā- ‘become warm’ vs. Neu 1968a: 1;
kiš- ‘become, happen’, kišt- ‘be extinguished’, zē(y)a- ‘be cooked done’ vs. Neu 1968a: 206-7)
and in -ta- (igāi- ‘grow cold’, kištanziya- ‘starve, become hungry’, lazziya- ‘become well, turn
out well’, mar-/mer- ‘disappear’): more than 25 total.
C. Major Problems
1. Hittite also shows transitive media tantum (in -a-!) with no sense of affectedness of subject.
In -a-: ark- ‘mount (sexually), climb’, ḫatta-‘prick, slit; sacrifice (by throat slitting)’, ḫuett(i)‘pull, pluck’, iškalla- ‘tear, slit’, pars(i)- ‘break in two’, paḫš- ‘protect’, tuḫš- ‘cut off’.
3
In -tta-: parḫ- ‘pursue, chase’, šanna- ‘conceal’, šarra- ‘transgress; break’, wešiya- ‘graze
(animals)’, zaḫḫiya- ‘fight’.
As conceded by Neu (1968b: 56), cannot be “indirect-reflexive” because never occur with -za.
Set in -a- argues for archaism, and every well attested example replaced in NH by active forms.
Not remotely credible as innovation (Oettinger 1993: 354&35933). As per Luraghi (2012), Hittite
is transitivizing language, and valency becomes more marked by diathesis from OH to NH: not
only elimination of transitive medio-passives, but rampant growth of passive. As in Latin (cf.
Neu 1968b: 177 and Kemmer 1993: 152), most intransitive middles also replaced by actives in
NH, but some new intransitive middles by hypercorrection. No real functional contrast
(‘lability’; see Luraghi 2012). Transitives also not explicable by alleged “oppositional actives”
created in response to new intransitive middles (Jasanoff 2003: 146, 151 et passim): would
predict old transitive ḫi-verbs †ḫattai etc. vs. intransitive ḫatta(ri) etc. Finally, secondary
transitive value of factitives in *-eh2 by opposition to new middles also not credible. Factitives
inherently transitive! Use of medio-passive forms in Hittite is sporadic and marginal. Reason
obvious: this role in PIE and Hittite filled mostly by statives and fientives in *-eh1ye/o- and
*-eh1s(k̑e/o)- (see Watkins 1971[1973], Jasanoff 2002-3[2004], and Luraghi 2012).
Finding 1: PIE middle (resp. “Stativ”) cannot originally have been inherently intransitive
category, with or without “subject affectedness” (contra Oettinger 1993: 359 “patientiver
Funktion”, Tichy 2000: 91 “statische oder patientive”, Jasanoff 2003: 159ff. “stativeintransitive”, et al.).
2. Widely held view that the stative middle is oldest type: Neu (1968b: 93) claims “…wird das
Zustandsmedium als das ursprünglichere bestimmt. Somit ist für das Medium die Entwicklung
“Zustand → Vorgang → Tätigkeit” anzusetzen” (see also ibid. 133&146). Clearly motivated by
derivation of middle from the perfect (ibid. pp. 186-8). Likewise Jasanoff (2003: 144ff.): “…the
middle endings…acquired their familiar formal and semantic properties through the renewal of
an earlier unitary perfect/middle set with broadly stative, processual, and ‘internal’ functions.”
Likewise, as per name, “Stativ” assumed to have originally stative function (Oettinger 1993:
359). See further Fritz apud Meier-Brügger (2000: 243): “Der Stativ drückt einen Zustand aus, in
dem sich das als Subjekt Bezeichnete befindet. Der Stativ fällt grundsprachlich einerseits mit
dem Medium zusammen, das neben seiner eigentlichen reflexiven Bedeutung zusätzlich die
Zustandsbedeutung des Stativs übernimmt.”
Facts not only of Hittite, but also other languages argue otherwise:
a. Of the seven Hittite transitive media tantum in -a-, six are change-of-state, only one refers to
an activity or state (see II.C.1 above, bottom of p. 2).
b. Of Hittite intransitive media tantum in -a-, 14 are change-of-state, 4 stative (to which add 3
statives from Luvian and Palaic); of those in -tta-, 29 are change-of-state, while 8 are stative.
c. In observed cases, historical development is change-of-state to stative: OH eša ‘sits down’, but
NH > ‘sits, be sitting’ (Neu 1968b: 29); arta means ‘stand’ in Hittite, but if connected with Grk.
ὦρτο ‘arose’ (Oettinger 1979: 524, Jasanoff 2003: 13), stative sense of the Hittite is innovative.
Change from change-of-state to stative is commonplace (see Jasanoff 2003: 158 with note 28).
d. As per Jasanoff (2003: 154-5), the Indo-Iranian “passive aorist” marks chiefly change-of-state:
ábodhi ‘awoke’, ároci ‘shone forth’, ádarśi ‘was seen, appeared’. Likewise, with Jasanoff (2003:
155-7)—despite the glaringly false label “stative-intransitive system”!—Tocharian class III/IV
4
presents with 3Sg -atär/-etär or -otär reflect change-of-state middles in *-o-(to)-: A pärkatär
‘ascends’, B wokotär ‘breaks open’, etc. NB included are stems in /-sk-/ and /-tk-/: B musketär
‘disappear’, A mloskatär ‘escape’, A sätkatär/B sätkentär ‘spread out’, B mlutketär ‘escape’. Cf.
Hitt. medio-passive -ške- verbs to active bases (Neu 1968b: 86-9). Synchronic sense generally
imperfective, but Tocharian points to original inceptive meaning. See Jasanoff (2003: 157-8) on
comparanda in Germanic and Balto-Slavic and characterization: “Functionally, the presents in 3
sg. *-o(r) denoted a non-perfective change of state.”
e. Finally, most ḫi-verbs are change-of-state: ānšš- ‘wipe off’, ar- ‘arrive’, ark- ‘cut off’, iškar‘stick, pierce’, išpar- ‘spread’, ištapp- ‘block, shut up’, kank- ‘hang’ (telic in Hittite!), k(a)rap‘devour’, lag- ‘knock down’, mark- ‘separate’, dā- ‘take’ (< *‘give/take’), wag- ‘bite’ (<
*‘break’). Only minority refer to activities: ārr(a)- ‘wash’, mall(a)- ‘grind’, māld- ‘speak
solemnly’, padda- ‘dig’, š(a)rap- ‘sip’, šipant- ‘libate’ (add stative šakk- ‘know’ < semelfactive
*‘make cuts in’). Similar ratio in ḫi-presents in -i-: arai- ‘arise’, ḫuwai- ‘flee’, išḫi- ‘bind’, išpai‘become sated’, mai- ‘grow’, nai- ‘turn’, piddai- ‘flee’, šai- ‘press’, dai- ‘put, place’, tarai‘grow weary’, zai- ‘cross’ vs. išḫamai- ‘sing’, parai- ‘blow, fan’, parip(a)rai- ‘blow’ (an
instrument).
Finding 2: true profile of early PIE middles argues for “change-of-state”, not “stative” as
most prominent meaning.
III. First Outline of a Solution
A. General Characteristics of Middle Voice and “Middle” Verbs
1. Kemmer (1993: 247) underscores that middle is fundamentally about transitivity (= valency,
as per Luraghi 2012): relative elaboration of events. As per Kemmer (1993: 7), meaning reflects
not directly real world, but real world as conceptualized by speakers.
2. Kemmer focuses heavily on relationship of middle and reflexivity (see chart on next page).
For her “spontaneous” activities are marginal, but concedes crucially (1993: 144-5) that these
actually include two-participant events with agent where speaker chooses not to express initiator.
Without using label, refers to “neuter”, “middle”, “labile” verbs, for which see Klaiman 1992.
3. Salient facts about “neuter” verbs: (1) widespread phenomenon; (2) overwhelmingly but not
exclusively have change-of-state semantics; (3) whether given verb is “neuter” or not is arbitrary
and not predictable (cf. remark of Jasanoff 2003: 2 on lack of semantic difference between miand ḫi-conjugation and Oettinger 1993: 347-8 on active vs. medium). English examples:
Neuter
Non-Neuter
I split the log. The log split.
I cut the apple. *The apple cut.
I broke the vase. The vase broke.
I crushed the vase. *The vase crushed.
I clogged the drain. The drain clogged.
I stuffed the bag. *The bag stuffed.
I ended the drama. The drama ended.
I completed the book. *The book completed.
I tore the cloth. The cloth tore.
I wounded the man. *The man wounded.
I closed the door. The door closed.
I blocked the door. *The door blocked.
I played the music. The music played.
The birds blackened the air. The air blackened.
Also neuter: rip, spread, open, melt, fill, finish. Not neuter: kill, injure, damage.
NB that status of verb can change: ‘perish’ now only transitive, but earlier neuter.
5
B. Provisional History of PIE Middle Voice
1. Pre-PIE/“Early PIE” had three classes of verbs based on valency: exclusively 2-participant
(prototypically agent-patient, but also experiencer-stimulus, aka “transitive”), exclusively 1participant (“intransitive”), and neuters (2-participant, but open to construal as 1-participant with
suppression of agent). First two inflected with endings *-m, *-s, *-t etc., thus active: *bheid‘split’, *spek̑- ‘look at’, *deik̑- ‘show, *h1es- ‘be’, *h2enh1- ‘breathe’, *h3erg- ‘perish’, etc. (NB
includes change-of-state verbs). Third inflected with endings *-h2e, *-th2e, *-e. Fundamentally 2participant, thus transitive, but also construable as 1-participant verbs: thus early *A B léghe ‘A
topples B’ or *B léghe ‘B topples’; also *A B skélHe ‘A splits B’ and *B skélHe ‘B splits’.
Likewise for handful of neuter activity verbs: *A B *wérte ‘A spins B’ or *B *wérte ‘B spins’.
Semantic Relations of Middle (after Kemmer 1993: 202 with Neuter Verbs added)
Two-Participant
Events
Indirect
Middle
Logophoric
Middle
Logophoric
Reflexive
Indirect
Reflexive
Natural
Reciprocal
Events
Reciprocal
Passive
Middle
Direct
Reflexive
Passive
Emotion
Middle
Non-Translational
Motion
Grooming
Cognition
Middle
Change in
Body Posture
Translational
Motion
One-Participant
Events
Spontaneous
Action or Process
Neuter
Verbs
6
2. Effects of change of state often lasting, hence motivation for category to express such a state:
*wéh2ge ‘A breaks B’/‘B breaks’, hence ‘is broken’ (stative). Furnished by secondarily derived
durative stem with iconic reduplication, o-grade of root (for which see DiGiovine 2012: 46-7)
and the *h2e endings: “perfect” *Ce-CóC-e; *we-wóh2g-e > Grk. ἔᾱγε‘is broken’ (with 2ary root
vocalism). For the few neuter activity verbs, resulting sense is iterative-durative: *we-wórt-e >
Skt. vavárta ‘spins, rolls’. NB salient achieved state is that of 2nd participant (patient), hence
perfect almost exclusively refers to state of grammatical subject.
3. Neuter verbs undergo functional and formal split: unitary neuter verbs construable as ‘A __s
B’ or ‘B __s’ with *-h2e, *-th2e, *-e endings differentiated into “active” (i.e. transitive, 2participant) type or “middle” (intransitive, 1-participant) type. Order of following steps not (yet?)
determinable:
(1) In intransitive function, *k̑éi-e *‘(lays)/lies’ → *k̑éi-or ‘lies’ (only!). Transitive function
retained by *-e inflection (see below on factitives and NB ḫi-verbs in -i-: *dhéh1-y-e(i) ‘puts’ >
Hitt. dāi; inflection as per Jasanoff 2003: 101-7).
(2) By internal derivation, “h2e-conjugation” type with acrostatic *ó/é (→ *ó/zero in TER(T)
roots) inflection (present or aorist depending on lexical semantics of root) takes on 2-participant
(“transitive”) function: *(h1)órH-e(i) > Hitt. ārri ‘washes’, *lóghe(i) > Hitt. lāgi ‘knocks down’.
Hence heavily transitive profile of ḫi-verbs (25 of 33 cited above, II.C.2.e, top of p. 4).
(3) Likewise, by internal derivation, type with zero-grade root and accented ending, initially
*legh-é ‘topples’ (tr.), whence per Jasanoff 2012 imperfect *leghét and eventually thematic aorist,
but mostly > *legh-ór ‘topples’ (intr.) > Hitt. lagāri ‘falls down’ (cf. Jasanoff 2003: 171). On
formal details of development of distinct middle inflection see Jasanoff (2003: 47-58). NB
illustrated instance of both developments in same verb uncommon. Typically, given neuter verb
replaced by one or the other.
(4) Speakers not automata: also a few intransitive h2e-verbs (e.g. Hitt. ari ‘arrives’, išpai‘become satiated’) and some transitive middles: Hitt. ḫatta(ri) ‘pricks’ < *h2étor and tuḫša(ri) <
*duh2s-ór etc. Again NB latter inexplicable from original exclusively intransitive middle by
“diathetic reversal” (Jasanoff 2003: 146&151).
(5) Denominative factitives in *-eh2 (3Sg *-éh2-e) have separate development: replaced in
intransitive ‘B __s’ (i.e. ‘enters state of __’) construal already in PIE by *-eh1-ye/o- and *-eh1s(k̑e/o)-. Rare Hittite exx. of medio-passives in -aḫḫ- may be remnant archaisms, but more likely
sporadic Hittite nonce creations. Transitive ‘A __s B’ function continued by *-éh2-e(i), which
become ordinary transitive ḫi-verbs and post-Anatolian replacements thereof. For Pres3Sg -aḫḫi
< *-éh2ei in factitives vs. -āḫi < *-óh2ei in new transitive *sóh2ei > Hitt. šāḫi ‘fills up, clogs’ see
Kloekhorst (2008: 65, 98 & 164; also 2006: 132).
4. Type of lagāri ‘falls down’ easily reinterpretable as true “spontaneous” 1-participant event
(see Kemmer 144-5 and Klaiman 1992: 46-7 for affinity of 1-participant version of neuter verbs
and 1-participant spontaneous events). Hence new pattern spread to verbs with 1-participant:
*geis- ‘turn’ (only intr.) > *géiso ‘turns’ > Hitt. kīša ‘happens, becomes’ (Kloekhorst 2008: 4801 after Eichner). NB means explicitly that some such verbs that originally inflected as mi-verbs
switched over (could not have been original neuter verbs since had only 1 participant).
7
5. From “spontaneous” actions and processes spread already in PIE to other 1-participant event
types (those underlined in diagram above): at least to grooming, translational and nontranslational movement, body posture, and emotion verbs. Probably also to reciprocals and
passives (NB general absence of agent in latter). Significance of absence in Anatolian of mediopassives of perception and cognition unclear. No positive evidence for existence in PIE of
indirect middles (i.e. self-benefactives). Almost certainly post-Anatolian innovation (shared
common in “Core IE” or independent in individual languages).
References
DiGiovine, Paolo. 2012. The Function of *o-Ablaut in the PIE Verbal System. In H. Craig
Melchert (ed.), The Indo-European Verb. Proceedings of the Conference of the Society for
Indo-European Studies, Los Angeles, 13-15 September 2010, 43-50. Wiesbaden: Reichert.
Jasanoff, Jay. 2002-03[2004]. ‘Stative’ *-ē- revisited. Die Sprache 43.127-70.
——. 2003. Hittite and the Indo-European Verb. Oxford/New York: OUP.
——. 2012. *u̯eid- ‘notice’ and the PIE thematic aorist. Paper presented at the 14th Fachtagung
of the Indogermanische Gesellschaft, Copenhagen, September 22, 2012.
Kemmer, Suzanne. 1993. The Middle Voice. Amsterdam/Philadelphia: Benjamins.
Klaiman, Miriam H. 1992. Middle Verbs, Reflexive Middle Constructions, and Middle Voice.
Studies in Language 16.35-61.
Kloekhorst, Alwin. 2006. Čop’s Law in Luwian Revisited. Die Sprache 46.2.131-6.
Luraghi, Silvia. 2012. Basic valency orientation and the middle voice in Hittite. Studies in
Language 36/1.1-32.
Lyons, John. 1969. Introduction to theoretical linguistics. Cambridge: CUP.
Meier-Brügger, Michael. Indogermanische Sprachwissenschaft.7., völlig neubearbeitete Auflage
unter Mitarbeit von Matthias Fritz und Manfred Mayrhofer. Berlin/New York: de Gruyter.
Neu, Erich. 1968a. Interpretation der hethitischen mediopassiven Verbalformen. Wiesbaden:
Harrassowitz.
——. 1968b. Das hethitische Mediopassiv und seine indogermanischen Grundlagen. Wiesbaden:
Harrassowitz.
Oettinger, Norbert. 1976. Der indogermanische Stativ. MSS 34.109-49.
——. 1979. Die Stammbildung des hethitischen Verbums. Nürnberg: Carl.
——. 1993. Zur Funktion des indogermanischen Stativs. In Gerhart Meiser (ed.), Indogermanica
et Italica. Festschrift für Helmut Rix, 347-61. Innsbruck: Institut für Sprachwissenschaft der
Universität Innsbruck.
Rix, Helmut. 1988. The Proto-Indo-European Middle. MSS 49.101-19.
Tichy, Eva. 2000. Indogermanisches Grundwissen. Bremen: Hempen.
Watkins, Calvert. 1969. Geschichte der Indogermanischen Verbalflexion [= J. Kuryłowicz,
Indogermanische Grammatik. Band III: Formenlehre]. Heidelberg: Winter.
——. 1971[1973]. Hittite and Indo-European Studies: the denominative statives in -ē-. TPS
1971.51-93.
Weiss, Michael. 2009. Outline of the Historical and Comparative Grammar of Latin. Ann
Arbor/New York: Beech Stave.
H. Craig Melchert
Università di Roma “La Sapienza”
October 15, 2012
The Function of the PIE Perfect
I. Standard View
A. The PIE perfect expressed an “attained state” (sometimes styled “naktostatisch”): see among
many Delbrück (1897: 177ff., with refs. to Buttmann and Kohlmann), Renou (1925: 7),
Chantraine (1926: 4), Hoffmann (1970: 39-41), Di Giovine (1996: 261&273), Kümmel (2000:
66), Meier-Brügger (2000: 155), Tichy (2000: 88), Weiss (2009: 379&409), Fortson (2010: 104-5).
B. Major debate has concerned definition as “aspect” or “Aktionsart”: for the first see MeierBrügger (2000: 155), Weiss (2009: 379); for the latter Di Giovine (1996: 273ff.), Kümmel
(2000: 66), Tichy (2000: 89).
C. Value of “attained state” is recessive already in Vedic Sanskrit and Homeric Greek and
appears elsewhere only in lexicalized relics (e.g. Latin meminī ‘I remember’). By what may be
termed a “metonymic” shift in focus, the perfect comes to indicate a past action whose effects
carry into the present (‘has arrived’) and by further developments into a narrative past tense (see
e.g. Kümmel 2000: 71-82 and Chantraine, passim).
1. Vedic examples for inherited sense “attained state” (translations Geldner):
RV 1,164,10ab: tisró māts trīn pitn bíbhrad éka ūdhvás tasthāu
‘Drei Mütter, drei Väter trägt der Eine und steht doch aufrecht da.’
RV 6,10,3ab:
pīpāya sá śrávasā mártyeṣu yó agnáye dadāśa vípra uktháiḥ
‘Der Redekundige schwillt an Ruhm unter den Sterblichen an, der gegen
Agni mit Lobliedern freigebig war.’
2. Homeric examples for inherited sense “attained state” (translations Murray):
Il. 1.37-38:
κλῦθί μευ, ἀργυρότοξ’, ὃς Χρυσήν ἀμφιβέβηκας Κίλλαν τε ζαθέην
‘Hear me, thou of the silver bow, who dost stand over Chryse and holy Cilla…’
Il. 11.470:
δείδω μή τι πάθῃσιν ἐνὶ Τρώεσσι μονωθείς
‘I fear lest some evil befall him, alone mid the Trojans.’
II. Dissent
A. Also a long history of interpreting the perfect as having an “intensive” value (wholly or in
part):
1. Bréal (1899-1900: 277) defines perfect as a “présent intensif”, citing Grk. ὄπωπα ‘je vois’,
ἀκήκοα ‘j’entends’ etc., allegedly expressing an “affirmation plus énergique”.
2. Hirt (1928: 279-80, citing Curtius and Bopp) rejects “attained state” analysis and insists on
“intensive Bedeutung”: τέθνηκα in the Iliad means ‘ist tot und ist tot’.
3. Meid (1983: 330-31) also stresses the intensive character of Skt. jujoṣa ‘ich genieße intensiv’
and Homeric δέδορκε used of a snake with a horrifying glare that ‘starrt und starrt und starrt’.
2
B. Perel’muter denies both the “attained state” and “intensive” readings (1967: 94-5) and
proposes based entirely on a small set of Homeric examples that the original function of the
perfect was to express psychic states and sensory perception (1967: 99-100).
III. Perfects with Descriptively Iterative Value
A. Most are to verba sonandi, including clearly onomatopoetic roots (see for Greek Di Giovine
2010: 189; for Sanskrit Kümmel 2000: 376ff., 486ff.):
Greek: βέβρῡχε ‘bellows, roars’, γέγωνε ‘shouts’, κέκρᾱγα ‘croak; shriek, cry’, μεμηκής
‘bleating’, μέμῡκε ‘bellows, roars’, etc.
Sanskrit: mimāya ‘moos’, vavāś-/vāvaś-‘low for’ (of a cow and her calf)
B. Account of Delbrück (1897: 202ff.) and Chantraine (1926: 16) by which these express an
attained state properly rejected as artificial (Perel’muter 1967: 94) or forced (Di Giovine 2010:
190, following Berrettoni).
C. Greek type explained as secondary from intensives by “metaplasm”: Di Giovine (1990: 83-86
and 2010:198ff.), following Tichy (1983: 70). Seen as part of general decline of Aktionsarten in
Greek. But such decline did not result in “metaplasm” in other categories, and true intensive
presents cited by Di Giovine (2010: 190) show no such alteration: e.g. κωκύω ‘shriek, wail’,
μορωῡρω ‘murmur, roar’, etc. Most long vowels in root syllable are function of onomatopoetic
origin and not evidence for an intensive origin (on γέγωνε see below). As per Schaefer (1994)
and Di Giovine himself (2010: 191-6), the intensive was a formally well defined category of PIE,
with heavy reduplication. Replacement of latter by Ce- pattern of PIE perfect not sufficiently
motivated. As per Di Giovine (2010: 190) both types in Greek (μέμῡκε and κωκύω) are
archaisms (for the former see also Perel’muter 1967: 95). NB: no claim that any specific example
is necessarily inherited, rather that model for creation of such must have existed.
D. Other Evidence for Iterative Perfects
1. *wewort- ‘turns’ (LIV2: 691-2): as often with verbs meaning ‘turn’, this root can be both telic
(‘turns’ = ‘changes one’s orientation’) and atelic (‘turns’ = ‘revolves, spins’). Sum of evidence
argues for both values already for PIE verb. From telic value, only resulting state predicted: ‘is
turned (= ‘ist gewandt’) (in a certain direction)’. But as per Kümmel (2000: 462ff.) vavart- also
well attested as ‘turns’: e.g. (translations Geldner)
RV 4,30,2: satrā te ánu kṣṭáyo víśvā cakréva vāvtuḥ
‘Hinter dir liefen alle Völker insgesamt her wie die Räder (hinter dem Roß).’
RV 1,166,9d: ákṣo vaś cakrā samáyā ví vāvte
‘Eure Achse läuft mitten durch die Räder.’
All telic examples taken by Geldner and Kümmel as present perfects ‘have/has turned
(towards)’. NB: does not exclude that some may represent rather attained state: ‘is/are turned
toward.’ If OCS vrъteti se ‘turns’ (intr.) is a remade perfect (thus Vaillant 1958: 387), then
iterative value is old.
2. Other examples obscured by dubious practice of assuming that roots with attested perfects
must on that account alone have had telic lexical semantics (effectively circular). Totally
unfounded for several verbs: e.g. *geh2dh- ‘rejoice’ > Grk. γέγηθε ‘rejoices’(cf. already
Perel’muter 1967: 97); *sr(e)iHg- ‘shiver, shudder’ > Grk. ἔρρῑγα ‘idem’ (cf. Hirt 1928: 279);
*h1lengwh- ‘hurry, hasten’ (LIV2: 247 ‘sich mühelos bewegen’) > Skt. rārahāṇá- ‘hastening’.
3
As to Grk. γέγωνε ‘shouts’, compelling comparison with TochA ken- ‘call’ points decisively to
an atelic root: “Beide teilen die Eigenschaft der Defektivität und bilden nur imperfektive
Stämme: im Osttocharischen den Präsensstamm, im Griechischen den Perfektstamm; perfektive
Stämme fehlen” (Hackstein 2002: 188). For differing attempts to account for long vowel in root
compare Hackstein (2002: 189-91) and Vine (2007: 352-4).
Evidence for original telic sense of Skt. kan- ‘rejoice’ (Kümmel 2000: 130-33, LIV2: 352) and
javi- ‘hasten’ (Kümmel 2000: 198-9, LIV2: 166) is also very weak (see Di Giovine 1990: 281 on
the latter).
3. Case of *leloígh- ‘licks’ more dubious: based on supposed equation of respectively isolated
Vedic ririhvāṃsaṃ ‘the licking one’ (RV 10,79,3d) and Greek λελιχμότες (for *λελιχϝότες)
‘licking’ (Hesiod, Theogony 826). Thus LIV2: 404, after de Lamberterie (1998: 375-9) and
Kümmel (2000: 429). If correct, contra de Lamberterie (1998: 379) the perfect does not express
“un état du corps”. Rather with Kümmel, loc. cit., one must concede: “Der vom Perfekt
bezeichnete Zustand ist folglich hier auch nicht naktostatisch, sondern eher durch andauernde
Iteration hervorgerufen bzw. habituell, d.h. es handelt sich um ein zeitlos-kontinuatives
Perfekt.” (emphasis mine—HCM). Note further with Kümmel that this root forms a PIE root
present, has iterative-durative Aktionsart, and the action results in no change in the state of the
subject. However, de Lamberterie’s analysis of Greek questionable: for alternative inner-Greek
analysis see Hackstein 2002: 195-6. Thus ririhvāṃsaṃ may be Vedic creation, merely modeled
on other cases cited.
IV. Compatibility of ‘achieved state’ and ‘iteration’
A. Realization of Aktionsarten dependent on “Verbalcharakter”/lexical semantics: see e.g.
García Ramón (2002).
B. System of Dowty-Vendler (see Dowty 1979) distinguishes states (‘shine’, ‘be’), activities
(‘weep’, ‘walk’)—from which one should distinguish semelfactives (‘blink’, ‘hit’)—, accomplishments (‘paint’, ‘bind’), and achievements (‘win’, ‘die’). For present purposes the crucial
distinction is statives and atelics (activities and semelfactives) versus telics (accomplishments
and achievements).
As per Di Giovine (1990), Kümmel (2000: 67), et al., stative verbs (e.g. *bheh2- ‘shine’) did not
form perfects in PIE. Verbs forming the perfect had to be processual. However, as noted by Meid
(1983: 330), it cannot be accidental that all well established examples of perfects expressing an
attained state are in fact telic, since only telic verbs lead to a result (‘bind’ > ‘hold bound’; ‘die’
> ‘be dead’). However, atelic verbs (‘turn=revolve’; ‘shiver’; ‘hasten’) cannot lead to a result
and an achieved state.
C. Hypothesis 1: the PIE perfect expressed a durative Aktionsart (cf. already in nuce Hirt 1928:
280 and Meid 1983: 330). In telic verbs duration could only be realized in the state resulting
from the action. For atelic verbs the effect was durative or iterative. That reduplication with
*CV- marked duration is not surprising (pace Di Giovine 1996: 117). Special value of durative
Aktionsart in the perfect (vs. varieties of the “present”) surely due to the other defining formal
characteristics: o-grade of root and special endings. For o-grade see Di Giovine (2012).
4
V. Similar Differentiation in Subsequent Development
As expected, perfects to atelic roots with iterative value do not undergo shift to “resultative”
perfects and then narrative preterites (see I.C above), but typically develop virtually synonymous
presents (see Perel’muter 1967: 95 and passim and concession by Chantraine 1926: 16ff.).
VI. Motivation for Creation of Durative to *-h2e Class
Hypothesis 2: orig. *h2e-inflection marked “neuter” verbs (English ‘break’: ‘I broke the vase’/
’The vase broke’). Such verbs overwhelmingly are change-of-state type. Need to express state
resulting from change of state: ‘is broken’ (vs. inherent statives like *h1es- or *bheh2-). Thus
creation of *Ce-CóC-, but also applied to handful of “neuter” verbs that were atelic like *wert(cf. English ‘I spin the bottle’/‘The bottle spins’/‘The bottle is spinning’).
References
Bréal, Michel. 1899-1900. Les commencements du verbe. MSL 11.268-284.
Chantraine, Pierre. 1926. Histoire du parfait grec. Paris: Champion.
Delbrück, Berthold. 1897. Vergleichende Syntax der indogermanischen Sprachen. Zweiter Teil. Strassburg: Trübner.
Di Giovine, Paolo. Studio sul perfetto indoeuropeo I. Rome: Dipartimento di studi glotto-antropoligici dell’
Università di Roma “La Sapienza”.
——. 1996. Studio sul perfetto indoeuropeo II. Rome: Il Calamo.
——. 2010. Declino di una categoria flessionale: l’intensivo in greco antico. In I. Putzu et al., La morphologia del
greco tra tipologia e diachronic, 189-203. Milan: Francoangeli.
Dowty, David. 1979. Word Meaning and Montague Grammar: The Semantics of Verbs and Times in Generative
Semantics and in Montague’s PTQ. Dordrecht: Reidel.
Fortson, Benjamin W. IV. 2009. Indo-European Language and Culture.2 Chichester/Malden: Wiley-Blackwell.
García Ramón, José Luis. 2002. Zu Verbalcharakter, morphologischer Aktionsart und Aspektin der
indogermanischen Rekonstruktion. In H. Hettrich, Indogermanische Syntax. Fragen und Perspektiven, 105136. Wiesbaden: Reichert.
Geldner, Karl F. 1951-57. Der Rig-Veda. Aus dem Sanskrit ins Deutsche übersetzt und mit einem laufenden
Kommentar versehen. Cambridge MA: Harvard University Press.
Hackstein, Olav. 2002. Die Sprachform der homerischen Epen. Wiesbaden: Reichert.
Hirt, Herman. 1928. Indogermanische Grammatik. Teil IV. Doppelung Zusammensetzung Verbum. Heidelberg:
Winter.
Hoffmann, Karl. 1970. Das Kategoriensystem des indogermanischen Verbums. MSS 28.19-41.
Kümmel, Martin J. 2000. Das Perfekt im Indoiranischen. Wiesbaden: Reichert.
de Lamberterie, Charles. 1998. Langues de feu (grec hésiodique λελιχμότες : sanskrit védique ririhvás-). In J.
Jasanoff et al. Mír Curad. Studies in Honor of Calvert Watkins, 373-390. Innsbruck: Institut für
Meid, Wolfgang. 1983. Bemerkungen zum indoeuropäischen Perfekt und zum germanischen starken Präteritum.
Zeitschrift für Phonetik, Sprachwissenschaft und Kommunikations-forschung 36.329-336 (also in Italian in
IncLing 4 (1978) 31-41).
Meier-Brügger, Michael. 2000. Indogermanische Sprachwissenschaft.7 Berlin/New York: de Gruyter.
Perel’muter, Ilya. 1977. О первоначальной функции индоевроейкого оерфекта. Voprosy Jazikoznanija 1.92-102.
Renou, Louis. 1925. La valeur du parfait dans les hymnes védiques. Paris: Champion.
Schaefer, Christiane. 1984. Das Intensivum im Vedischen. Göttingen: Vandenhoeck and Ruprecht.
Tichy, Eva. 1983. Onomatopoetische Verbalbildungen des Griechischen. Vienna: ÖAW.
——. 2000. Indogermanisches Grundwissen. Bremen: Hempen.
Vaillant, André. 1958. Grammaire comparée des langues slaves. Tome 3. Le verbe. Paris: Klincksieck.
Vine, Brent. 2007. Latin gemō ‘groan’, Greek γέγωνε ‘cry out’, and Tocharian A ken- ‘call’. In A. Nussbaum, Verba
Docenti. Studies in historical and Indo-European linguistics presented to Jay H. Jasanoff by students,
colleagues, and friends, 343-357. Ann Arbor/New York: Beech Stave.
Weiss, Michael. 2009. Outline of the Historical and Comparative Grammar of Latin. Ann Arbor/NewYork: Beech
Stave.
Hittite ḫi-Verbs of the Type -āC1i, -aC1C1anzi*
H. Craig Melchert
Of the many still unresolved problems regarding the Hittite ḫi-conjugation, one of the most
recalcitrant is the class of verbs showing a pattern of Pres3Sg in -āC1i vs. Pres3Pl in -aC1C1anzi.
For all of its notoriety the class is a small one: aki, akkanzi ‘die(s)’; ḫāši, ḫaššanzi ‘beget(s);
give(s) birth’; ḫāši, ḫaššanzi ‘open(s)’; ištāpi, ištappanzi ‘block(s), stop(s) up’; nāḫi, *naḫḫanzi
(ptc. naḫḫant-) ‘frighten; take(s) fright’; pāši, paššanzi ‘swallow(s)’; wāki, wakkanzi ‘bite(s)’;
zāḫi, *zaḫḫanzi (Pres1Pl zaḫḫueni) ‘strike(s), beat(s)’. There is also general agreement that we
should include šāḫi, *šaḫḫanzi ‘clog(s), fill(s) up’, although in this verb the strong stem has been
generalized (see Oettinger 1979: 512 and Kloekhorst 2008: 690-1). One should note that there is
a general tendency to spread the strong stem to positions where we would expect the weak: see
also Pres3Pl paša[nzi], zaḫanzi, VblNoun nāḫuwaš, etc. On a possible motivation for
generalization of šāḫ- in particular see below.
For most verbs of this class the derivatives argue that the stem with the geminate consonant
is basic: akkātar ‘death’, ḫāšša- ‘offspring, progeny’ and ḫaššatar ‘birth; family’, ištappeššar
‘dam’ and ištappulli- ‘lid, stopper’, UZUpap(p)aššala/i- ‘esophagus’ (or ‘gullet’), zaḫḫ(a)i- ‘fight,
battle’ (for ‘frighten’ one may adduce CLuvian naḫḫuwa- ‘be an object of concern for’). As
noted, šāḫ- has generalized the strong stem, whence šaḫeššar ‘fortification’ (based on the wellestablished Hittite use of the “Kastenmauer” type of construction, on which see now De
*
I am indebted to the editor Elisabeth Rieken and an anonymous reviewer for a number of
helpful suggestions. I am responsible for all views expressed not explicitly attributed to others.
2
Vincenzi 2008). The one exception is wāg- ‘bite’, where the derivatives NINDAwageššar ‘bread
morsel’ (or sim.), NINDAwagāta- (likewise a kind of bread) show that the single stop is basic.1
One factor that has made an account of this class particularly difficult is that many of the
verbs lack a sure etymology. However, the few etymologies that are clear confirm that the
geminate consonant is original for most, but not for wāg- ‘bite’. The verb naḫḫ- ‘to frighten; take
fright’ reflects PIE *neh2- seen also in OIr. nár ‘noble, modest; grievous’ < *neh2s-ro‘fearsome, awesome’ (also the base of Hitt. naḫšaratt- ‘fear’): see Kloekhorst 2008: 592 and
Puhvel 2007: 13 with reference to Pedersen. As per Kloekhorst (2008: 691), šāḫ- continues PIE
*seh2- ‘to fill up’, seen in the derived sense ‘to satiate’ in Latin satis ‘enough’, Grk. ἄμεναι ‘to
satiate oneself’ etc. Hitt. pašš- ‘to swallow’ represents *peh3-s-, an “s-present” or “s-enlarged”
form of *peh3 -, continued in “Core Indo-European” as ‘to drink’ (Kloekhorst 2008: 649 with
refs. pace Puhvel 2011: 183-4). Finally, whether one favors a preform *h2ens- (Melchert 1994a:
164) or *h2ems- (Kloekhorst 2008: 319-21) and whatever the ultimate etymology, it is certain
that ḫašš- with a geminate is the primary form of ‘to beget; give birth’, resulting from
assimilation of a nasal plus *s. On the other hand, wāg- ‘to bite’ is a reflex of a PIE verb ‘to
break’, either *weh2g- (thus Kimball 1988: 245, LIV2: 664, Kloekhorst 2008: 940, and adopted
here) or *wag- (Jasanoff 2003: 150).
The fact that for most of the class the geminate consonant (NB in all cases an obstruent) is
primary has naturally led to attempts to explain the single consonant of the strong stem (which
appears after a synchronically long vowel) as the result of “lenition” (or voicing). That is, one
1
Contra Rieken (1999: 196) and Kloekhorst (2008: 939) the form wagat/daš occurring in OH
texts in the context of lists does not prove an s-stem, but reflects merely the common use of the
nominative as the “default” case in lists (see Hoffner and Melchert 2008: 243, §16.9).
3
would like to attribute the alternation to the well-established Proto-Anatolian rule by which
voiceless stops and *h2 were lenited/voiced after a preceding accented long vowel (Eichner
1973: 79ff., Morpurgo Davies 1982/1983, Adiego 1993). Oettinger (1979: 447-50) posits a
sound change by which a sequence of accented short vowel in an open syllable followed by short
syllable lengthens the first, in time to “lenite” the following obstruent. Kloekhorst (2006a: 132
and 2008: 65 and 98) has attempted to revive this rule for accented short *ó (without the
specification of a following short syllable, but with restriction to initial and final syllables).
Unfortunately, the rule as stated by Oettinger and Kloekhorst cannot be correct, since there
are incontrovertible counterexamples: Hitt. ḫuwappi ‘throws, hurls’ reflects *h2wópei (with
initial ḫ- retained against the “Saussure-Hirt effect” after the weak stem ḫupp-). Contra
Kloekhorst (2008: 369), one cannot in this case arbitrarily invent a root-final first laryngeal *h1
to explain this example away, since the oldest participle of the Sanskrit cognate vap- is uptá-,
showing that the root is aniṭ (thus correctly LIV2: 684, but without recognition of the Hittite
evidence for the initial laryngeal). It is also methodologically illicit (effectively circular) to posit
a final *h1 on the root *dek- of Hitt. dakk- ‘match, resemble’ purely in order to avoid the
counterexample to the supposed lenition rule (again LIV2: 109 correctly reconstructs an aniṭ
root). Since the two verbs cited do have ablauting paradigms, one could try to explain the
geminate in the strong stem as taken from the weak, but then one would need to explain why this
did not happen in the case of the aki, akkanzi type.
Kloekhorst (2006a: 296 and 2008: 95) claims that ḫāppar ‘business; transaction’ reflects
*h3ép-r and that the *[o] resulting from *h3e did not fall together with apophonic *o and
therefore did not “lenite” the following *p. But the vowel of the word clearly is long, so he must
in ad hoc fashion assume that only after the lenition rule ceased to operate the short *o did finally
4
lengthen. On the contrary, the Hittite word shows that the accented *ó did in fact lengthen, but
did not lenite the following voiceless stop. There are two possibilities. First, one could assume an
original acrostatic r/n-stem *h3óp-r/h3ép-n-, whose weak stem was then modified to *h3ep-én(cf. Hitt. widen- ‘water’ < *wed-én-). The original weak stem is reflected in the derivative
ḫappena- ‘rich’ etc., while the base noun generalized the strong stem (contra Kloekhorst, loc. cit.
and Pinault 2012: 418, who uneconomically reconstructs an r-stem and and n-stem). By this
derivation, it was apophonic *ó that failed to lenite the following stop. Second, one could instead
reconstruct a proterokinetic paradigm *h3ép-r/h3p-én-, but also in this case the Hittite word
ḫāppar shows that the *[ó] lengthened, but failed to lenite the following stop. Once again, one
could appeal in this case to paradigm leveling, assuming that the unlenited /p/ comes from the
original weak stem.
Paradigm leveling will not explain other counterexamples. Hitt. āppa ‘back’ reflects a
remade *āppi (reflected in the derived verb āppai, āppianzi ‘be finished; step back’ and the
HLuvian cognate á-pi; Melchert 2009: 335-6). The preform is clearly *(h1)ópi matching Grk.
ὄπι (Puhvel 1984: 93-4). The analysis of Kloekhorst (2008: 193-4) starting from *h2op-o is quite
impossible, including the false claim that Hittite local adverbs are inherently unaccented,
contradicted by p(a)rā ‘out, forth’ with long vowel < *pró (see Kloekhorst 2008: 630!).2 Hitt.
2
Kloekhorst (2008: 730) likewise derives Hitt. š(a)rā from accented *sr-ó. Anatolian *āppi and
Grk. ὄπι cannot, pace Kloekhorst, be separated from Grk. ἔπι, so if the word had an initial
laryngeal, it must have been *h1. I also reject the claim of Kloekhorst (2006b: 83-4) that initial
*h2o- merges with *h1o- in Hittite. None of his putative examples are compelling, and the
development is contradicted by examples such as ḫāšš- ‘ash; soap’ < *h2óh1/3s- (see Rieken
1999: 22, with reference to Melchert 1994: 147f.).
5
wappu- ‘river-bank’ reflects *(h2)wópu-, cognate with Skt. vápra- ‘mound; rampart; high riverbank’ < vap- ‘throw, strew’ (thus with Catsanicos 1985: 125). Here, with generalization of the ograde strong stem, the “Saussure-Hirt effect” was not undone.3 In sum, accented short *ó
manifestly did not “lenite” a following voiceless labial or velar stop (compelling examples for
dental stops are lacking, but they surely behaved the same). We thus cannot explain aki etc. as
being the product of the Proto-Anatolian “lenition” rule after accented long vowel, which clearly
did affect following stops: *dhéh1-ti > *d di > Lyc. tadi ‘puts’,
-m > Hitt. wēkun ‘I
demanded’, etc.
Other attempts to account for the aki, akkanzi type must likewise be rejected. Whatever the
status of “long-vowel” perfects in PIE (see for a brief review Jasanoff 2003: 31), they cannot
help solve the current problem. The only verb of the class that might reflect such a category is
wāg- ‘bite’. But even if one accepts the view of Jasanoff (2003: 150) that the root is *waginstead of *weh2g-, and the further assumption (which he does not) that the ḫi-verb might reflect
a perfect, a preform *(we)wāg- would produce only a consistent wāg-. It could not possibly have
been the starting point for the alternation -āCi, -aCCanzi. The account of Melchert (1994a: 81)
was based on the false premise of a single verb išpar(r)- ‘to spread out (with the foot)’.
Kloekhorst (2008: 406-10) has demolished the entire basis for that scenario. The derivation of
aki, akkanzi < *ógei, égnti, (Melchert 1994b: 304) is contradicted by the fact that derivatives like
akkātar ‘death’ show that it is the allomorph akk- with geminate stop that is basic.
Although his rule of lenition after accented short *ó will not work as stated, Kloekhorst
makes a crucial new observation (2008: 164), which can serve as the basis for a solution: he
3
The word would have originally referred to walls/embankments resulting from throwing
down/piling up of earth and was then extended to natural formations of similar shape.
6
points out that factitives in *-eh2 (phonetically *[aħ]) appear as Hittite ḫi-verbs with Pres3Sg
-aḫḫi, never showing lenition of *h2 in direct contrast to nāḫi ‘affrights’ and šāḫi ‘fills up,
clogs’, which by any analysis must continue o-grade *nóh2ei and *sóh2ei. This striking
difference cannot be coincidental, and I see no way to avoid concluding that accented short *ó
did in fact “lenite” a following *h2. Such a restricted “lenition” of just *h2 after accented short *ó
was in fact proposed by Kimball (1999: 397). She made no explicit contrast with -aḫḫi < *-éh2ei
in the factitives, nor did she try to motivate the change phonetically, and in Melchert (2011: 128)
I dimissed the claim as ad hoc. The direct contrast cited by Kloekhorst between unlenited -aḫḫi
< *-éh2ei vs. lenited -āḫi < -óh2ei compels a reconsideration.
Before turning to the phonetic motivation for the change, I must deal with the putative
counterexample I cited (2011: 128): Hitt. lāḫḫa- ‘campaign’, which appears to be an action noun
of the τόμος-type reflecting a preform *lóh2o-. Kloekhorst (2008: 510-11) argues for an original
root noun instead, which would avoid the problem, since an ablauting paradigm *lóh2-/léh2could have generalized unlenited -ḫḫ- from the regular weak stem laḫḫ- < *léh2-. Unfortunately,
Kloekhorst’s argument in favor of a root noun is not entirely compelling: namely, that the
denominative verb laḫḫiya- ‘to campaign’ can hardly be derived from an a-stem. However, as I
argued in Melchert (2004: 376), Hieroglyphic Luvian hasi(ya)- means ‘to satiate’ (with reflexive
particle ‘to satiate oneself, enjoy to the fullest’), derived from the noun (LINGERE)hasa‘satiety, abundance’ (itself formed from the root seen in Palaic ḫaš- ‘be satiated’).4 So we cannot
4
This interpretation is now supported by an occurrence in Hittite context, KBo 20.107+ iii 22,
where we find the figura etymologica ḫāšiyamiš ḫāšiya ‘As a satiated one, satiate!’ (for the text
see Bawanypeck 2005: 112). Just how this derivational pattern came about is a separate issue.
One possibility is the existence in at least one case of an intervening adjective in *-iyo-, which
7
totally exclude laḫḫiya- < lāḫḫa-. An alternative account is that lāḫḫa- was only formed after the
rule leniting *h2 after *ó ceased to operate. We know that the formation of deverbative
action/result nouns remained very productive in Hittite, since some do not show the inherited ograde of the τόμος-type: e.g., gul(a)šša- ‘fate’ < gulš- ‘to draw, sketch, plan’ or kuera- ‘field’
(section of land) < kuer- ‘to cut’. It is thus likely that some examples like ḫāšša- ‘offspring’ that
could show old o-vocalism are likewise recent creations based directly on the synchronic verb
(ḫāšš- ‘to give birth’). We are permitted to suppose a similar origin for lāḫḫa- , even if the base
verb is missing, replaced by laḫḫiya-.
As to the phonetic motivation for “lenition” of *h2 after *ó, we must first review the status
of the better-known “lenition” process of Anatolian. According to the original conception, ProtoAnatolian had two separate “lenition” (or voicing) rules, affecting stops and *h2: one occurred
after a preceding accented long vowel (including long vowels resulting from loss of tautosyllabic
laryngeals) and the other between unaccented vowels (Eichner 1973: 79ff. and 10086; Morpurgo
Davies 1982/1983). However, as shown by Adiego (2001), Proto-Anatolian “lenition” (or
voicing) was actually a single rule which affected voiceless stops and *h2 between unaccented
morae, *V
. Thus Lyc. tadi ‘puts’ <
entirely parallel to
Lyc. esbedi ‘with horse’ < *ékwodi < *ékwoti. Adiego also adduces cross-linguistic evidence for
the effect being due to the low pitch of the surrounding unaccented vowels.
Since obstruents after an accented short *ó are not between unaccented morae, any lenition
or voicing in this environment must be attributed to an entirely different factor, which need not
served as the basis for the derived verb. As always, we need not and should not assume such a
link in every case. The pattern of forming a verb in -iya- from a-stem nouns may have become
moderately productive.
8
affect the same range of target sounds. I suggest that the different factor that lenited *h2 was the
well-known “stronger” or “longer” quality of what we call phonological “short” */o/ in PIE, on
which see most recently Kümmel (2012: 308-9) and Keydana (forthcoming). The most famous
effect of this phonetic quality of PIE *o is of course “Brugmann’s Law”, by which PIE *o in
open syllables yielded long ā in Sanskrit, while other PIE short vowels did not. Since I retain the
traditional conception of “Čop’s Law” in Luvian (contra Kloekhorst 2006a), it likewise suggests
that phonological “short” *o was in fact longer than phonological “short” *e. At some point
Luvian disallowed accented light syllables. The “solution” in the case of accented short *ó was
to lengthen the vowel: *dó.ru > tā.ru ‘wood’. But in the case of accented short *e, the following
consonant was geminated, producing a coda consonant for the accented syllable: *pé.rVm/n >
par.ran ‘in front’.
I therefore see no difficulty in assuming that due to this quality of the “short” *o in Hittite a
sequence *ó.h2 V resulted in
.ħV > āḫV, while *ó.pV led to *óp.pV (see Melchert 1994a: 18)
and *é.h2V (really [a.ħV]) became *aħ.ħV > aḫ.ḫV. I assume then that the “lenition” was
phonetically regular only in the roots ending in *-h2: *nóh2ei > nāḫi, *sóh2ei > šāḫi.5 If these
roots followed the productive ḫi-conjugation pattern of *ó/zero ablaut (whether this is viewed as
original or secondary), the weak stems would have been respectively *nh2- and *sh2-. The first
followed the pattern of *lógh- ‘incline’: phonetically regular, but paradigmatically aberrant *alg<*lgh- was renewed as *legh- > lag- beside strong lāg-. Likewise then *anḫ- < *nh2 gave way to
5
Whether zāḫi ‘strikes’ is phonologically regular or analogical after the other two verbs depends
on its etymology. By the suggestion of Schindler apud Oettinger (1979: 447) that zaḫḫ- reflects
*ds-eh2- < *das- seen Grk. δαΐ ‘in battle’, one would assume the latter, but this etymology is not
entirely assured (cf. Kloekhorst 2008: 1020).
9
*neh2- > naḫḫ-. However, as per Kloekhorst (2008: 691), *sh2-énti would have led to Hittite
*išḫanzi, and here the aberrant allomorphy was solved by generalizing the strong stem šāḫ-.
Unsurprisingly, the pattern -āḫi : -aḫḫanzi was extended analogically to roots with fixed avocalism and the other voiceless fricative s: hence also ḫāši, ḫaššanzi (for *ḫāšši, ḫaššanzi)
‘beget; give birth’ and ‘open’ and pāši, paššanzi ‘swallow’. It was also extended to just three
roots with fixed a-vocalism and a geminate stop. Notably, it was not extended to dākki, dakkanzi
‘match, resemble’. It would be desirable if we could account for this difference. The contrast of
dākki, dakkanzi and wāki, wakkanzi, whose etymologies are known, suggests a reason. The
former, whose immediate preforms were *dókei, *dekénti, was easily fitted into the dominant
*ó/zero ablaut pattern, but *wóh2gei, *uh2génti would have led to wāki, *ūganzi. In the face of
this very aberrant ablaut, the verb was remodeled after the existing -āCi, -aCCanzi type.6
We are thus led to suspect that ‘block, stop up’ and ‘die’ also joined this type because their
historically regular paradigms (i.e., in phonological terms) resulted in very irregular allomorphy.
I believe that a case can in fact be made that this applied to both verbs. For ‘block, stop up’
Kloekhorst (2008: 416) reasonably compares the Germanic family of English ‘stuff’, German
stopfen, etc., but concedes that these point to Proto-Germanic *stup-, which cannot be easily
reconciled with the Hittite. I suggest (see already the discussion by Puhvel 1984: 474) rather PIE
*stembhH- (LIV2 595), reflected in Skt.
6
‘‘prop, fasten, fix (in place)’, from which it is a
Kloekhorst (2008: 940) assumes that pretonic vowel+h2g results in assimilation to -kk- rather
than loss of the laryngeal and compensatory lengthening and that *ūkk- was avoided by
anaptyxis, but there is no support for the first assumption, and his alleged examples of anaptyxis
in a similar environment are false: there is no evidence for anything except /e/ in wek- ‘demand’,
and witen- in the paradigm of ‘water’ results from *wetén-.
10
short step to ‘block, stop up’.7 A paradigm *stómbhHei, stmbhHénti would result in *ištāmpi,
ištappanzi. Compare for the strong stem Hitt. dampu- ‘blunt’, cognate with OCS topъ ‘blunt’
(Kloekhorst 2008: 826 with refs.) and for the weak Hitt. kappi- ‘small’ < *kmb(h)i-, cognate with
Av. kamna- ‘small’, kambišta- ‘least’ (Kloekhorst 2008: 439, following Szemerényi, contra
Puhvel 1997: 63).8 It is important to note with Kloekhorst, loc. cit., that the geminate -pp- of
kappi- shows that *-mb(h)- results in -app-, with loss of the nasal, but fortition of the stop.
Likewise, then, in our verb there would have been an allomorphy between *ištāmpi with nasal
and lenis stop and ištappanzi without nasal and fortis stop. I believe that this discrepancy blocked
the more expected “repair” of the radical allomorphy by restoration of the nasal in the plural,
since this would still not have resulted in the normal pattern of išpānti, išpandanzi ‘libate(s)’,
where the stops matched in manner of articulation (/ispa:ndi/, /ispandantsi/). Instead, the
irregular *ištāmpi, ištappanzi was assimilated to the pattern of nāḫi, naḫḫanzi, which also had
the contrast of lenis vs. fortis in the strong and weak stems.
We come finally to aki, akkanzi ‘die(s)’. I believe it is fair to say that none of the
etymologies suggested for this verb have been remotely convincing. See for a summary of
7
LIV2sets up the PIE root with inherent *m and final laryngeal, but concedes that there are also
aniṭ forms and that separating reflexes of *stembhH- from those of *stebh- (LIV2 588) is difficult.
One could also suppose (as does Puhvel, loc. cit.) a single root *stebh- and a nasal infix verb
which was renewed in Sanskrit by a nasal suffix. The Hittite verb could just as easily reflect
*stómbhei, stmbhénti. Since the question of one or two roots is not decisive for our present
purposes, I leave the matter open here.
8
The western Anatolian word for ‘(grand)child’ attested in Greek inscriptions as καμβειν,
κομβος, etc. supports this derivation (Neumann 1961: 61).
11
attempts Tischler (1983: 8-9). Puhvel (1984: 22-3) and Kloekhorst (2008: 168) justifiably do not
even bother to mention all of the proposals. I suggest that as a “thought experiment” we start
with a PIE root that could be the source for ‘die’: *nek- reflected in TochB näk- (act.) ‘destroy’,
Latin nex ‘death’, Av. nasu- ‘corpse’, Grk. νέκῡς ‘dead; corpse’, OIr. éc ‘death’, etc. If we
reconstruct forward a ḫi-verb *nókei, *nkénti, what would be the expected result by normal
phonological changes? The third singular would lead to *nākki (for lack of “lenition” see above).
The outcome of the third plural is the vital question: what was the regular reflex of a syllabic
nasal before homorganic stop? Puhvel (1984: 22) explicitly rejects deriving akk- < *nk- (already
Hrozný 1917: 176!), claiming that the result would be *ank-.9 I contend that current evidence in
fact points rather to akk-. This requires a short excursus on the matter.
The example of kappi- < *kmb(h)i- suggests that the result was loss of nasalization in the
case of a labial sequence (with fortition in the case of an original voiced stop). We cannot be
certain where the accent was in this word, but a priori we would expect that it was on the suffix
(the word is too sparsely attested to draw any conclusions from the lack of plene spellings).10
Parallel treatment for the unaccented velar sequence *[ŋk] as akk- seems entirely justified.11 The
only putative counterexample is merely apparent. The Hittite suffix -anki/*-ankiš (for the latter
9
Hrozný did express serious reservations about his idea, presumably because he too already had
reason to expect ank instead.
10
An anonymous reviewer points out that the spelling καμβειν cited above in footnote 8 with ει,
representing a long and likely accented [i:], tends to support the idea of original oxytone accent
in *kmb(h)i-.
11
I follow here what I believe is a long-standing and widespread view that PIE */n/ had an
allophone [ŋ] before dorsal stops.
12
see Hoffner-Melchert 2008: 168) in 1-anki and 1-iš ‘once’ etc. was compared already by
Rosenkranz (1936: 249) with Greek -άκι(ς) in τετράκις ‘four times’, πολλάκι(ς) ‘many times’,
pointing to a preform *-ńki(s) (actually *[ŋ
.12 Since many of the Greek forms show a
syllable structure long-short-short, an original accent on the final syllable could have been
retracted onto the penult by “Wheeler’s Law”. However, the lack of plene in the final syllable of
the Hittite forms precludes accent there, so I believe that accent on the penult may be reasonably
inferred.13 Under the accent the nasalization was preserved, leading to -anki(s).14 I am unaware
of any probative examples for unaccented *[nt/d].15
12
The statement by Kloekhorst (2008: 181) that Hitt. -anki occurs only with 1-3 is false: 7-anki
KUB 33.105 i 5-8; 8-anki KBo 21.90 Ro 11, etc.
13
I am indebted to Michael Weiss for counsel on this point, but the interpretation offered here is
my own, not his.
14
The weak stem gank- for ‘hang’ is not probative, since a paradigm kānki, *kakkanzi < *kónkei,
*knkénti would certainly have been modified to the attested kānki, kankanzi after išpānti,
išpantanzi (I emphasize that in the case of ‘hang’ the final stop would have been consistently
fortis, contrary to the case of ‘stop up’ discussed above).
15
These facts are not contradicted by the apparently different treatment of sequences of non-
homorganic syllabic nasal and following stop, where we find no nasalization under the accent,
but nasalization when unaccented: *ń-mh2yent- > amiyant- ‘immature’ (see Kloekhorst 2008:
172 for this shape as the regular outcome), *kmta > katta ‘down’ (= Grk. κάτα and CLuvian
zanta; Goedegebuure 2010), *kmta/ō > katta ‘with, beside’ (the base must be *kom cognate with
Latin cum ‘with’ etc.), versus *mdhró- > antarā- ‘blue’ (see Kloekhorst 2008: 186, contra
13
We may thus assume that the prehistoric paradigm of ‘die’ was *nākki, akkanzi. It is hardly
surprising that this very irregular allomorphy was eliminated, and as in the case of wāki, *ūganzi
‘bite(s)’ and *ištāmpi, ištappanzi ‘stop(s) up’, the “repair” chosen was assimilation to the class of
nāḫi, naḫḫanzi. Depending on the relative chronology of the changes, it is possible that wāki,
wakkanzi served as the model for aki, akkanzi or vice-versa.
In sum, due to a “Brugmann” effect that “lenited” *h2 after accented short *ó, the pattern
Pres3Sg in -āC1i vs. Pres3Pl in -aC1C1anzi developed regularly in the case of naḫḫ- ‘to affright;
fear’ and *šaḫḫ- ‘to fill up’. It was analogically extended to ḫi-verbs in -ašš-: ḫašš- ‘to beget;
give birth’, ḫašš- ‘to open’, and pašš- ‘to swallow’ (also to zaḫḫ- ‘to strike’ if it was not regular
there). Finally, it also spread to verbs with final stop which for various reasons had developed
very irregular allomorphy: wāgi, *ūganzi ‘bite(s)’, *ištāmpi, ištappanzi ‘stop(s) up’, and *nākki,
akkanzi ‘die(s)’.16
References
Adiego Lajara, Ignacio-Javier. 2001. Lenición y acento en protoanatolico. In Onofrio Carruba
and Wolfgang Meid (eds.), Anatolisch und Indogermanisch. Anatolico e indoeuropeo. Akten
Melchert 1994a: 125). As per above (with note 2), the two adverbs katta cannot be derived from
unaccented forms (contra Kloekhorst 2008: 464).
16
The contrast of šāku(wa)- ‘eye’ < *sókwo- (Melchert 1994a: 61, Kloekhorst 2008: 704) or
more likely *sh3ókw- (cf. Rieken 1999: 59-60 with refs.) with nekku ‘nonne’ suggests that
“lenition” after short *ó may also have affected the voiceless labiovelar *kw, in contradistinction
to the other voiceless stops. However, to affirm this claim a full review of all the evidence is
required, which cannot be undertaken here.
14
des Kolloquiums der Indogermanischen Gesellschaft. Pavia, 22.-25. September 1998, 11-18.
Innsbruck: Institut für Sprachen und Literaturen der Universität Innsbruck.
Bawanypeck, Daliah. 2005. Die Rituale der Auguren. Heidelberg: Winter.
Catsanicos, Jean. 1985. Review of Jos Weitenberg, Die hethitischen U-Stämme (Amsterdam
1984). BSL 80/2.123-7.
De Vincenzi, Tommaso. 2008. Fortification Walls. Development and Conformation of Anatolian
“Saw-Tooth Wall”, “Kastenmauer”, “Casematte” Defence Systems, and Their Building
Techniques in the Bronze Age. In Hartmut Kühne, Rainer M. Czichon, and F. Janoscha
Kreppner, Proceedings of the 4th International Congress of the Archaeology of the Ancient
Near East, 309-20. Wiesbaden: Harrassowitz.
Eichner, Heiner. 1973. Die Etymologie von heth. mēhur. MSS 31.53-107.
Goedegebuure, Petra. 2010. The Luwian Adverbs zanta “Down” and *ānni “With, For, Against”.
In Aygül Süel (ed.), Acts of the VIIth International Congress of Hittitology. Çorum, August
25-31, 2008, 299-318. Ankara: T.C. Çorum Valiliği.
Hoffner, Harry A. Jr., and H. Craig Melchert. 2008. A Grammar of the Hittite Language. Part 1:
Reference Grammar. Winona Lake: Eisenbrauns.
Hrozný, Bedřich. 1917. Die Sprache der Hethither. Leipzig: Hinrichs.
Jasanoff, Jay. 2003. Hittite and the Indo-European Verb. Oxford/New York: Oxford University
Press.
Keydana, Götz. forthcoming. Brugmann’s Law and the role of perception in sound change. In
Roman Sukač and Václav Blažek (eds.), The Sound of Indo-European 2.
15
Kimball, Sara. 1988. Analogy, Secondary Ablaut and *oh2 in Common Greek. In Alfred
Bammesberger (ed.), Die Laryngaltheorie und die Rekonstruktion des indogermanischen
Laut- und Formensystems, 240-56. Heidelberg: Winter.
——. 1999. Hittite Historical Phonolgy. Innsbruck: Institut für Sprachwissenschaft der
Kloekhorst, Alwin. 2006. Čop’s Law in Luwian Revisited. Die Sprache 46/2. 131-6.
Kümmel, Martin Joachim. 2012. Typology and reconstruction. The consonants and vowels of
Proto-Indo-European. In Benedicte Nielsen Whitehead et al. (eds.), The Sound of IndoEuropean. Phonetics, Phonemics, and Morphophonemics, 291-329. Copenhagen: Museum
Tusculanum.
LIV2 = Rix, Helmut (ed.). 2001. Lexikon der indogermanischen Verben. Zweite, erweiterte und
verbesserte Auflage, Wiesbaden: Reichert.
Melchert, H. Craig. 1994a. Anatolian Historical Phonology. Amsterdam/Atlanta: Rodopi.
——. 1994b. ‘Čop’s Law’ in Common Anatolian. In Jens E. Rasmussen (ed.), In honorem
Holger Pedersen. Kolloquium der Indogermanischen Gesellschaft vom 26. bis 28. März
1993 in Kopenhagen, 297-306. Wiesbaden: Reichert.
——. 2009. Hittite ḫi-verbs from Adverbs. In Rosemarie Lühr and Sabine Ziegler (eds.),
Protolanguage and Prehistory. Akten der XII. Fachtagung der Indogermanischen
Gesellschaft, Krakau, 11. bis 15. Oktober 2004, 335-9. Wiesbaden: Reichert.
——. 2011. The PIE Verb for ‘to pour’ and Medial *h3 in Anatolian. In Stephanie W. Jamison,
H. Craig Melchert, and Brent Vine (eds.), Proceedings of the 22nd Annual UCLA IndoEuropean Conference, 127-32. Bremen: Hempen.
16
Morpurgo Davies, Anna. 1982/1983. Dentals, rhotacism, and verbal endings in the Luwian
languages. KZ 96.245-70.
Neumann, Günter. 1961. Untersuchungen zum Weiterleben hethitischen und luwischen
Sprachgutes in hellenistischer und römischer Zeit. Wiesbaden: Harrassowitz.
Oettinger, Norbert. 1979. Die Stammbildung des hethitischen Verbums. Nürnberg: Carl.
Pinault, Georges-Jean. 2012. Remarks on PIE amphikinetic and hysterokinetic nouns. In
Benedicte Nielsen Whitehead et al. (eds.), The Sound of Indo-European. Phonetics,
Phonemics, and Morphophonemics, 399-424. Copenhagen: Museum Tusculanum.
Puhvel, Jaan. 1984. Hittite Etymological Dictionary. Vol. 1. Words beginning with A. Vol. 2.
Words beginning with E and I. Berlin/New York/Amsterdam: Mouton.
——. 1997. Hittite Etymological Dictionary. Vol. 4. Words beginning with K. Berlin/New
York/Amsterdam: Mouton.
——. 2007. Hittite Etymological Dictionary. Vol. 7. Words beginning with N. Berlin/New York:
Mouton de Gruyter.
——. 2011. Hittite Etymological Dictionary. Vol. 8. Words beginning with PA. Berlin/New
York: Mouton de Gruyter.
Rieken, Elisabeth. 1999. Untersuchungen zur nominalen Stammbildung des Hethitischen.
Rosenkranz, Bernhard. 1936. Das griechische Adverbium -ως. KZ 63.241-9.
Tischler, Johann. 1983. Hethitisches Etymologisches Glossar. Teil 1 a-k. Innsbruck: Institut für
H. Craig Melchert
University of Siena
October 16, 2012
Newest Thoughts on The ḫi-Conjugation
I. Jasanoff (1979) and (2003: 71&89) posits PIE “h2e-conjugation” with orig. acrostatic
paradigm (see already Meillet (1916):
Present
1Sg
2Sg
3Sg
Imperfect/Injunctive
?
*mólh2-h2ei 1Pl *mélh2-meH
*mólh2-th2ei 2Pl *mélh2-(H)e?
*mólh2-e(sic!) 3Pl *mélh2-ti/?
*mólh2-h2e
*mélh2-meH?
*mólh2-th2e *mélh2-(H)e?
*mólh2-e(t)?? *mélh2-(s)
Paradigm reflected in verb for ‘crushes, grinds’: Hittite malli, mallanzi (1st for mālli*), Lat.
molō, OIr. melid, OCS meljo, Goth. malan, Arm. malem. Similarly išpānti, išpantanzi ‘libates’;
šākki, šekkanzi ‘knows’ etc.
For unpersuasive attempts to derive such forms from PIE perfect or preform thereof see inter al.
Eichner (1975), Risch (1975), Cowgill (1979), Kuryłowicz (1979) et al.
II. Kloekhorst (2012) makes new attempt to show that e-vocalism in Hittite weak stem is entirely
secondary.
A. Verbs k(a)rāp-/k(a/i)rēp- ‘devour’ and š(a)rap-/š(a)rep- ‘sip’ show anaptyctic vowel [ɨ] (see
already Kloekhorst 2008: 122-3 and 142-3). Supposedly based on similar pattern in mi-verbs
teripp- ‘plow’ (/trep-/, /trɨp-/) and wek- ‘ask for’ (/wek-/, /wɨk-/). Non-existent! As shown by
parallel case of Hitt. teri- = CLuv. tarri- ‘three’ (< *téri- < *trí-), Hitt. stem is fixed underlying
/téripp-/ (strong stem generalized to eliminate allomorphy /téripp-/, */tarp-/). No basis for any
alternation in wēk- ‘ask for’ (always ú-e-): thus simply acrostatic *wēkti, wékti (pace
Kloekhorst 2008: 996-7). In any case, putative original *ghróbh-, *ghbh- > k(a)rāpi, *karpanzi.
No motivation for anaptyxis. Given dominant pattern of descriptive *ó, zero (see below),
obvious “repair” could only have been k(a)rāpi, *k(a)rapanzi. No secondary source credible for
attested pattern. Must reflect *ghróbhei, *ghrébhti; *sróbhei, *srébhti. NB attested spellings
for weak stem are only ka-re-e-pé-er and ka/gi-re-p°. In absence of -ri-i-p° no basis for any
reading of re/i sign except /kreb-/. Likewise for ša-re-p°.
B. Per Kloekhorst (2012: 153ff.), other e-vocalism in weak stem of ḫi-conjugation verbs began
in preterite third plural in analogy to mi-conjugation and then spread subsequently to other parts
of the paradigm. On basis of Pres3Pl adanzi : Pret3Pl eter, ašanzi : ešer etc. :: Pres3Pl akkanzi :
Pret3Pl eker (replacing aker). Impossible as stated: proportion could only lead to *ekker. Use of
other example (e.g. arer → erer) as primary model does not solve basic problem: Pret3Pl of ḫiconjugation originally had strong stem in Hittite: aker ‘died’ (OS), ḫāner ‘drew (liquid) (NS),
ḫāšer ‘opened’ (MS) (see Kloekhorst 2012: 15414), ḫāšer ‘gave birth’ (NS); a-re-er ‘arrived’
ambiguous (but cf. Imv2Pl ārten). Part of more general pattern: dai- ‘put’ with older Pret3Pl
dāer, dāier, later tiyēr; also naier, nāer to nai- ‘turn’.
III. New Account of ḫi-Conjugation Ablaut Pattern
A. With Jasanoff (2003: 151) reconstruct also ḫi-conjugation aorist (to telic roots vs. presents to
atelic roots, entirely parallel to mi-conjugation). Jasanoff assumes original é-grade throughout
plural, just as in the corresponding present above:
2
Aorist
1Sg
2Sg
3Sg
*lógh-h2e
*lógh-th2e
*lógh-e
1Pl *légh-me2Pl *légh-(t)e3Pl *légh-(s)
B. But, in the mi-conjugation, aorist plural had strong stem in 1st/2nd persons (Jasanoff 2003:
82, citing Hoffmann 1968: 7f.): Skt. (á)karma, (á)karta, (á)kran etc. Generalized in Hittite to 3rd
person (Jasanoff 2003: 83; thus already Hart 1980 and Barton 1985): hence ešer ‘they were’,
ēpper ‘they took’, etc. Should thus assume same for ḫi-conjugation, hence likewise generalized
strong stem (i.e. *ó-grade!) in entire aorist plural, including third person.
C. Per Jasanoff (2003: 85), zero-grade in newly created Hittite presents of *h2e-aorists
analogical to mi-conjugation: *éswen, *ésten, *éser : *aswéni, *asténi, *asánti :: *sékwen,
*sékten, *séker :: x (→ *sakwéni, šaktēni, *sakkanti). But assumed e-grade plural of *h2e-aorists
patently false. And no model in mi-conjugation for spread of weak stem to Pret3Pl.
D. Rather, strong tendency to introduce zero grade into weak stem of acrostatic paradigm due to
established process in acrostatic nominal inflection (see Schindler 1972): TER(T) roots tend to
substitute weak T(T)- for TéR(T) and TeT roots T(e)T- for TéT. NB: such substitution must
have taken place in both paradigms of original presents (i.e. present and imperfect), but only in
the new presents from old aorists):
Aorists (not always as per Jasanoff 2003!): ār-/ar- ‘arrive’, ārk-/ark- ‘divide’, iškār-/iškar‘pierce, stick’, išpār-/išpar- ‘spread, strew’, kānk-/kank- ‘hang’ (telic in Hittite!), ?mārk-/mark‘dissect’.
Presents: māld-/mald- ‘make a vow, speak solemnly’ (likely, though in principle weak stem
could reflect *méldh-).
Further likely candidates for both lists, but lack of plene spelling in CVC(C) syllables leaves
uncertain. Also more of each outside Hittite (Jasanoff 2003: 74-6).
Likewise in roots TwE(R)T:
Aorist: ḫ(u)wapp-/ḫupp- ‘throw’, ḫ(u)wart-/ḫurt- ‘curse’ (possibly present instead).
Also in roots CET:
Aorists: lāg-/lag- ‘incline, knock over’ (lag- 2ary zero grade for *alg-).
Presents: šākk-/šakk- ‘know’ (contra Jasanoff 2003: 150, since per him 2003: 80, sense calls for a
present; no certain evidence for atelic root, pace LIV2 524; semelfactive sense ‘make cuts in’
compatible with all evidence); dākk-/dakk- ‘resemble, match (NB regular Hittite outcome of
renewed weak stem *sek-, *dek- would have been aberrant *sikk-, *dikk- with fixed i-vocalism).
Two possible solutions: (1) restore e after /sreb-/; (2) replace with “morphological” zero-grade a
after TER(T) type. For ‘know’ we find mostly the former, throughout the paradigm: šekkueni,
šekteni, šekkanzi, šekkuen, šekker, šeggallu, šekten, šekkandu, šekkant- (with some intrusions
even into the singular: šekti, šekta). Find also once the second solution: šaktēni. Fact that latter
by chance is only OH/OS example does not prove that this was the older pattern and that šekkwas a later replacement. Rather two alternatives alongside. For ‘resemble’ find only second
solution dakk-.
3
E. Ergo not coincidental that only survival of original weak *é-grade stem is in two roots of
shape TRET (R ≠ *y, *w): *ghrébh- ‘seize’ > ‘devour’ and *srébh- ‘sip’ and in CeT root *sekHanalogical to *srebh-. Latter is likely present, with original weak e-grade. *ghrebh- is surely
aorist, thus with newly created present also modeled after /sreb-/. These influenced to limited
extent some of verbs with *ó/zero > ā/a ablaut. NB contra Kloekhorst no basis for privileging
preterite third plural. Verb ar- ‘arrive’ shows Pres1Pl ērweni, Pres2Pl ērteni, Pret1Pl ērwen,
VblN erwar beside Pret3Pl erer; ḫašš- ‘open’ shows Pres1Pl ḫēšweni, Pres3Pl ḫēš(š)anzi,
Pret1Pl ḫēšwen, Imv2Sg ḫēš, Imv3Sg ḫēšdu, Imv2Pl ḫēšten, Imv3Pl ḫēšandu, VblN ḫēšuwar
beside ḫēš(š)er; šakk- ‘know’ shows Pres2Sg šekti, Pres1Pl šekkweni, Pres2Pl šekteni, Pres3Pl
šekkanzi, Pret1Pl šekkwen, Imv1Sg šeggallu, Imv2Pl šekten, Imv3Pl šekkandu, Ptc. šekkant(beside šakkant-!) beside Pret3Pl šekker. Purely by chance in akk- ‘die’ and ḫan- ‘draw (liquid)’
only Pret3Pl eker and ḫēner. Note further that Hittite likely inherited more aorists with only ograde than presents with o/e alternation. Those with shape TER(T) naturally assumed ā/a
inflection, not ā/e. With loss of overt *e by regular sound change in *mélh2-n̥ti >mallanzi and
*spénd-n̥ti > išpandanzi, preserved /e/ extremely sparse. Nevertheless undeniable.
F. NB that Hittite ḫi-conjugation (from original aorists) thus does show o-grade plural (cf. Kim
2012 on Tocharian Class I preterites).
III. Origin of Type (C)āC1i : (C)aC1C1anzi
A. Hittite shows small subvariety of ḫi-verbs with alternation Pres3Sg -āCi vs. Pres3Pl -aCCanzi
(and elsewhere in weak stem): aki, akkanzi ‘die’, ḫāši, ḫaššanzi ‘give birth’, ḫāši, ḫaššanzi
‘open’, ištāpi, ištappanzi ‘plug up, block’, nāḫi, naḫḫant- ‘be afraid’, pāši, *paššanzi (inf.
pāššuwanzi) ‘swallow’, šāḫi, šāḫanzi ‘clog, fill up’ (strong stem generalized), wāgi, wakkanzi
‘bite’, zāḫi, zaḫḫweni ‘strike’. Derivatives show that base stem is with one exception that with
the geminate: akkātar ‘death’ etc., ḫāšša- ‘offspring, progeny’ etc., ištappeššar ‘dam’,
pap(p)aššal- ‘esophagus’, zaḫḫai- ‘battle’, but wageššar ‘bread morsel’, wagāta- (sim.). NB that
wakkanzi 1x is only basis for membership in class.
B. Per Kloekhorst (2008: 65&98 also 2006: 132) caused by lenition by preceding *ó, which had
already become *ō in Proto-Anatolian. But quite impossible rule: cf. ḫ(u)wappi ‘throws’ <
*h2wópei (cannot be a root *h2weph1-; Vedic ptc. of vap- ‘throw, strew’ is uptá-); likewise no
evidence for root-final laryngeal in *dek- (*dókei > dākki). Likewise Hitt. āppa ‘back’ ← *āppi
(attested in CLuvian āppi) < *(h1)ópi. But contrast of nāḫi and šāḫi < *-óh2ei vs. no lenition ever
in factitives in -aḫḫi < *-éh2ei is striking.
C. NB that PA “lenition” rule affects stops and *h2 between unaccented vowels (in moraic
terms): see for correct formulation and motivation Adiego 2001 (due to low pitch). No reason to
regard “lenition” after accented short *o as same process. Due rather to well-known fact that
phonological “short o” in PIE longer than “short e” (and surely also “short a”). Cf. “Brugmann’s
Law” and “Čop’s Law”. Therefore assume that rule (Hittite only or earlier?) affects phonetically
only *h2 (question of voiceless labiovelar left open for now). Then extended analogically to cases
with geminate -šš-: naḫḫanzi: nāḫi :: ḫaššanzi : x (→ ḫāši for *ḫāšši). NB that “lenition” can
hardly be regular in ḫāšš- ‘give birth’ < *h2óns-ei or in *pāss- ‘swallow’ < *póh3s-ei. Finally,
extended to examples with geminate stop and originally very irregular stem alternations: for
*ištāmpi, ištappanzi ‘block’ < *stómbh-ei, *stm̥bh-énti (root of Skt. stabhnā́ ti ‘fixes, fastens’);
for *wāgi, *ūganzi ‘break’ < *wóh2g-ei, *uh2g-énti; for *nākki, akkanzi ‘die’ < *nók̑-ei, *n̥k̑-énti
(details on last example forthcoming).
4
References
Adiego, Ignacio-J. 2001. lenición y acento en protoanatolico. In O. Carruba and W. Meid (eds.),
Anatolisch und Indogermanisch, 11-18. Innsbruck: Institut für Sprachwissenschaft der
Barton, Charles. 1985. Hittite me-ri-ir, epp- and a note on the ablaut of root verbs. KZ 98.13-19.
Cowgill, Warren. 1979. Anatolian ḫi-conjugation and Indo-European perfect: Installment II. In
E. Neu and W. Meid (eds.), Hethitisch und Indogermanisch, 25-39. Innsbruck: Institut für
Eichner, Heiner. 1975. Die Vorgeschichte des hethitischen Verbalsystems. In H. Rix (ed.),
Flexion und Wortbildung, 71-103. Wiesbaden: Reichert.
Hart, Gillian. 1980. The ablaut of present and preterite in Hittite radical verbs. Anatolian Studies
30.51-61.
Hoffmann, Karl. 1968. Zum Optativ des indogermanischen Wurzelaorist. In J. Hesterman et al.
(eds.), Pratidānam. Indian, Iranian and Indo-European Studies Presented to Franciscus
Bernardus Jacobus Kuiper on his Sixtieth Birthday, 3-8. The Hague: Mouton.
Jasanoff, Jay. 1979. The position of the ḫi-conjugation. In E. Neu and W. Meid (eds.), Hethitisch
und Indogermanisch, 79-90. Innsbruck: Institut für Sprachwissenschaft der Universität
Innsbruck.
——. 2003. Hittite and the Indo-European Verb. Oxford/New York: Oxford University Press.
Kim, Ronald I. 2012. Unus testis, unicus testis? The Ablaut of Root Aorists in Tocharian and
Indo-European. In H. Craig Melchert (ed.), The Indo-European Verb. Proceedings of the
Conference of the Society for Indo-European Studies, Los Angeles 13–15 September 2010,
137-49. Wiesbaden: Reichert.
Kloekhorst, Alwin. 2006. Čop’s Law in Luwian Revisited. Die Sprache 46.2.131-6.
——. 2012. Hittite “ā/e”-ablauting Verbs. In H. Craig Melchert (ed.), The Indo-European Verb.
Proceedings of the Conference of the Society for Indo-European Studies, Los Angeles 13–15
September 2010, 151-60. Wiesbaden: Reichert.
Kuryłowicz, Jerzy. 1979. Die hethitische ḫi-conjugation. In E. Neu and W. Meid (eds.), Hethitisch und Indogermanisch, 143-6. Innsbruck: Institut für Sprachwissenschaft der Universität
Innsbruck.
Meillet, Antoine. 1916. De quelques presents athématiques à vocalisme radical o. MSL 19.18190.
Risch, Ernst. 1975. Zur Entstehung des hethitischen Verbalparadigmas. In H. Rix (ed.), Flexion
und Wortbildung, 247-58. Wiesbaden: Reichert.
On Wh-(Non)-Movement and Internal Structures
of the Hittite Preposed Relative Clause
MATTYAS HUGGARD
Relative constructions have received much attention from linguists over the past
few decades as they are interesting on the level of their syntax, typology, and semantics. As a clause type, relative clauses (RCs) have a high frequency of usage,
independent of text or register type, and in many languages they exhibit features
such as movement or non-movement of the relativized Determiner Phrase (DP),
the presence or absence of a resumptive constituent, and restrictive versus appositive semantics, all of which provide access to basic structural properties of a
given language. Of the various syntactic strategies employed dealing with relativization (postnominal, prenominal, circumnominal, and correlative), Hittite, the
oldest attested Indo-European language, predominantly exhibits the correlative
construction.
It has now been over sixty years since Held’s description (1957) of the Hittite
RC, and fifteen years since the last structural account offered by Garrett (1994) in
the course of discussing RC syntax in Hittite and Lycian. During this time, advances have been made both in the realm of syntactic theory and within the field
of Anatolian studies. Until recently it was assumed that the correlative strategy
was the only relative construction available in Hittite. However, Probert (2006)
provided evidence for the existence of embedded relatives in Old Hittite, which
exploits the same morpheme ku- for both relative and interrogative function. This
dual strategy of embedded-correlative is similar to what occurs in Latin, Ancient
Greek, Hindi, and Old English (Bianchi 2000:54), which thus further encourages
a revision of the relativization strategy proposed for Hittite in the literature.
1. The raising analysis
In this paper I would like to discuss the properties of relative clauses in Hittite in
light of the head-raising analysis proposed by Kayne (1994:85–97). For the past
decade, a number of reanalyses of the syntax of relativization have been proposed
by scholars using the raising analysis instead of the Principles and Parameters
adjunction proposal, which takes relative clauses to be right-adjoined to a nominal constituent that they modify.
Stephanie W. Jamison, H. Craig Melchert, and Brent Vine (eds.). 2011.
Proceedings of the 22nd Annual UCLA Indo-European Conference. Bremen: Hempen. ###–###.
2
Mattyas Huggard
In the head-raising analysis, the head noun of the relative clause (claim in
(1)) is taken to originate inside the relative clause (2):
(1)
the claim which John made
(2)
[DP the [CP Co [TP John To [vP made [DP which [NP claim ] ] ] ] ] ]
The relative construction in (1) is derived from (2) by raising the relativized Determiner Phrase (DPrel) to the specifier position of the Complementizer Phrase
(SpecCP) (3a), followed by the raising of the Noun Phrase (NP) claim to spec
DPrel (3b). This paper discusses in more detail the syntax of the elements appearing in the left periphery of the relative clause in Hittite. Following Rizzi (1997:
257) and Bianchi (2000:72), I would like to propose that Hittite relativization
provides evidence for such an articulated structure as illustrated in (4). Note that
this articulated structure includes two Topic Phrases (TopP), a Focus Phrase
(FocP), and a Finite Phrase (FinP) between the Complementizer Phrase (CP) and
the Tense Phrase (TP). The specifier position of CP is the predicted landing site
of interrogatives and relatives in wh-move languages, and the Finite Phrase’s
head hosts subordinating connectors in many languages. The following discussion is exploratory and in no way intends to present a full description for all types
of relatives encountered in the Hittite corpus. Its objective is to provide a possible
account for the variety of orderings encountered in Hittite preposed correlatives
in terms of the head raising analysis, and make sense of the relationship between
their syntactic and semantic properties.
On the Hittite Preposed Relative Clause
(3) a.
[DP the [CP [DP j which [NP claim] ]j Co [TP John To [vP made tj ] ] ] ]
b.
[DP the [CP [DP j [NP k claim ]k which tk ]j Co [TP John To [vP made tj ] ] ] ]
3
4
Mattyas Huggard
(4)
2. “To move or not to move,” a question about Hittite questions
It is noteworthy that most research on the syntax and semantics of correlatives
has been based on Hindi (Dayal 1995; de Vries 2002:145–50; Mahajan 2000;
Srivastav 1991). With regard to the internal syntax of the correlative clause, de
Vries (2002:147–8) predicts that the DPrel should occupy the specifier of the adjoined CPrel, but notes that the wh-operator in Hindi correlatives may or may not
move overtly to SpecCPrel, which would correspond to the wh-in-situ property of
Hindi interrogatives. One salient and well known feature of Hittite preposed RCs
is that they are divided into types: one where the relative pronoun occupies the
first position (indeterminate RCs), and another where it does not (determinate
RCs). Since relativization is sensitive to the Wh-Island condition, the Complex
DP condition, the Subject Island condition, and the Adjunct Island condition,
then at some level relativization involves wh-movement. Hence before examining
RCs, one should first investigate the mechanics of wh-movement, the theoretical
predictions made by the hypothesis, and their implications for the status of Hittite
interrogatives.
5
2.1. Movement driven by features
Wh-movement is an operation induced by morphosyntactic features. Crosslinguistically, question words generally display certain morphological regularities; these expressions make up a limited class of words that are morphologically
related, share a semantic commonality, and display the same syntactic behavior
(Adger 2003:285–7):
(5) a.
b.
c.
d.
e.
English: who, what, where, when
French: qui, quoi, où, quand
Kiowa: hatel, honde, haya
Hebrew: mi, mah, ayin, matay
Hittite: kui, kuit, kuwapi
In languages which exhibit wh-movement, the wh-phrases must obligatorily
move to the highest specifier position of the CP, whether they be a subject, an
object, a prepositional or an adverbial wh-expression. This syntactic behavior and
the common semantics are easily demonstrated in the case of English, a well
documented wh-move language (modified from Adger 2003:288):
(6)
Who read the book?
For which X, X a human, is it the case that X read the book.
(7)
When did you read the book?
For which X, X a time, is it the case that you read the book at time X.
(8)
Which student read the book?
For which X, X a student, is it the case that X read the book.
(9)
What did she read?
For which X, X a thing, is it the case that she read X.
The same applies to English why, where, and how. The mechanics of whmovement are just as in (3a) above, and repeated here (10).1 Most variability
cross-linguistically is linked to the strength properties of features of functional
heads: strong features trigger movement, and if a feature is weak, the element
remains in situ, as illustrated in the following Mandarin Chinese example (11).
Now let us examine a few interrogatives in the Hittite corpus to determine
whether the language is of the wh-move or wh-in-situ type.
1
For more on feature checking see Adger 2003:291.
6
Mattyas Huggard
[CP [DP j what [NP claim ] ]j Co did [TP Puduepa To [vP make tj ] ] ] ?
(10)
(11)
a.
Hufei mai-le yi-ben-shu
Hufei bought one-CL-book
“Hufei bought one book.”
b.
Hufei mai-le shenme?
Hufei bought what
“What did Hufei buy?”
2.2. Hittite interrogatives
Previously, Hittite has been described as a wh-move language (Garrett 1994:45,
following Hale 1987; Hoffner and Melchert 2008:350). However, questions with
the wh-phrase initial in the clause are generally short and contain only the whphrase and the verb. Superfically they seem to support wh-movement, but since
Hittite is an SOV language and given the brevity of the question clauses, these
types of sentences, as seen in (12a) and (12b), may not be used as evidence for
either option:
(12)
a.
7
kui=war=a=kan
kuenta
who-nom.sg=quote=them-acc=ptc kill-3sg.pret
“Who killed them?” (KBo III 67 ii 4)
b.
kuit=ta
memai
what-acc=you-sg.dat say-1sg.pres
“What can I say to you?” (KBo XXVI 65 iv 23, 25)
However, in longer sentences the Hittite wh-phrase is in fact non-initial, as shown
in the following examples:
(13)
d
URU
U
Nerik TUKU-an
ZI-an
kuzza
Stormgod of Nerik angry-acc.sg soul-acc.sg what-abl
KASKAL-ami
appease-1sg.pres
“By what means can I appease the angry soul of the Stormgod of Nerik?”
(KUB 5.1 i 92–3, oracular inquiry)
(14)
“The seeds about which you, imuili, said to me: ‘Seeds have now been
sown in Dapikka, and some in Anziliya, some in ariya and some in
anikka,’
nu ap
NUMUNI.A kuwapi pai
conj those-n.acc.pl seed-pl
where to give-2sg.past
where have you given those seeds (instead)?” (HKM 55 obv. 16–17,
MH/MS letter)
(15)
“The King of Isuwa will no longer come to His Majesty. Know thus, Queen
my Lady.
nu MUNU.LUGAL GAAN=YA kuit
iyai
conj Queen
Lady=my
what-n.acc do-2sg.non-past
What will you the Queen, my Lady, do?” (Kbo 8.23: 13–17, letter of an official to the Queen)
(16)
EME-a
EN-a
kuwapi psi
tongue-gen.sg lord-nom.sg where to go-2sg.non-past
“Where are you, lord of the tongue, going?” (KUB XII 62 obv. 10, cf. rev.3)
In her study on Focus operations in Hittite interrogatives and declaratives,
Goedegebuure (2009:947) observes that from a corpus sample size of one hundred interrogatives, forty-six clauses only had two constituents. The remaining
8
Mattyas Huggard
fifty-four ku- words show the following distribution as presented by
Goedegebuure (2009:947), and repeated here in table 1:
Table 1. Relationship between linear order and grammatical role of Hittite Q-word
questions
Position
Initial
Preverbal
Total
S
O
6
1
7
Obl
0
5
5
1
2
3
Adv Total
12
19
27
35
39
54
Notice that in (13) the wh-phrase kuzza is in the ablative case and appears in
the predicted position for an oblique object. In (14) and (16) kuwapi is an adverbial wh-phrase and appears in the normal position adjoined to the verb phrase.
In (15) kuit is in the accusative case and again appears in the predicted basegenerated position for an accusative object in an SOV language. According to
the cross-linguistic evidence and analysis cited in §2.1, Hittite patterns with
Mandarin Chinese: it is a wh-in-situ language in which the question words remain
in their base-generated position in unmarked interrogatives.
However, although wh-expressions do not undergo wh-movement in an interrogative clause type, this does not prevent other types of movement from taking
place. Comparing Focus structure in Hittite declaratives and interrogatives,
Goedegebuure (2009) provides strong evidence that Hittite wh-phrases occur in a
pattern parallel to focused constituents in declarative sentences: the wh-word may
remain in situ or be focused, either in initial position (the Focus position, as
shown above in (4), which is lower than the C position) or preverbal position,
depending on the type of focus. Her results are summarized below in table 2:
Table 2. Focus scope and focus typology of Hittite Q-word questions
Position in the clause
Identification focus
Information
Replacing/selecting/rejecting
Comment focus
Information, subject Q-word
Information, object Q-word
Reporting focus
Information, subject Q-word
Template
Initial position
Preverbal
nu ku- (…) V
nu (…) ku- V
in situ S(O)V
in situ (S)OV
(nu) ku- (…) V
(nu) (…) ku- V
in situ S(O)V
(nu) ku- (…) V
9
Thus whether a clause has a Declarative Force or an Interrogative Force, both
types of clauses are subject to Topicalization and Focus movements, which supersede other types of syntactic operations in Hittite, resulting in the observed
variations in word order as laid out by Goedegebuure (2009).
3. Hittite preposed correlatives
3.1. Syntactic preliminaries
Descriptively, the correlative structure is composed of two clauses, the main
clause (i.e., matrix clause) and a dependent clause (i.e., relative clause) which
may be preposed or postposed with respect to the matrix clause. The relative
clause and the matrix clause contain two constituents which are interpreted
roughly as coreferent, often termed the relative and the correlative element, respectively. De Vries (2002:145) defines a correlative construction as one in
which “the relative CP is left adjoined to the matrix clause, the head noun is internal to the relative clause, and the matrix contains a demonstrative correlate,”
schematically represented in (17):
(17)
[IP [CP wh Ni …] [IP … Demi …] ]
Typologically, correlatives predominantly occur in SOV languages (de Vries
2002). Thus it is no great surprise that the correlative structure constitutes the
most common relativization strategy across old Indo-European languages, including Latin, Sanskrit, Ancient Greek, Medieval Russian, Old English, and (of primary relevance here) Hittite.2
(18)
a.
Latin
liberatae sunt
[CP [DPQuibus [NPdiebus ] ]i Cumae
REL-abl
days-abl Cumae-nom released are
obsidione [ [ isdem
diebus ]i … Ti. Sempronius
…
siege-abl
same-abl days-abl
Tiberius Sempronius
2
Examples from Bianchi 2000.
10
Mattyas Huggard
prospere
pugnat ]
successfully fights
“Tiberius Sempronius won a victory in the same days in which Cumae
was released from the siege.” (Liv. 23.37.10)
b.
Sanskrit
] [ yáje
támi
[ [CP [DP yásmd yóner ]i udrith
]]
REL-abl womb arose-2sg.perf honor-1sg.pres that-acc
“I honor the womb from which you arose.” (RV 2.9.3c)
c.
Medieval Russian
I
[CP [DP kotoruju [NP zvezdu ] ]i potrebno bylo nam videt ]‚
and
which-acc
star-acc necessary was for us to see
[ [ tu zvezdu]i zaslonilo
tueju ].
that star
was covered by cloud
“The star we needed to see was covered by a cloud.”
De Vries’ observation (2002:148) that the wh-operator in Hindi correlatives may
or may not move overtly to the initial position of the clause is related to the whin-situ property of Hindi interrogatives. Such a phenomenon may also be observed in Hittite correlatives, as exemplified in (19):
(16)
[CP nu=kan mAparrun [DP kuedani
INA URULakku
conj=ptc Aparru-acc
REL-dat/loc in Lakku-loc
URU
kuenner
] nu mTarini
Lakkun pta
kill-3pl.pret conj Tarini-nom Lakku-acc seize-3sg.pret
“Tarini seized (the city of) Lakku in which they killed Aparru.” (KBo 2.5
iv 7–9)
The relative element (kuedani INA URULakku in (19)) appears to sit in base position in the RC just as the correlative demonstrative appears to be in base position
in the main clause. The two constituents are introduced by specific determiners
and their NP complement may be realized in both clauses as in (19), or it may be
deleted in either one of them. Thus following de Vries (2002:145–53), the schematic representation of the correlative structure (19) is (20):
11
(20)
3.2. Previous accounts of Hittite preposed relatives
Predating generative syntax, Held’s survey of Hittite relative clauses (1957) divides Hittite relatives into preposed and postposed relatives. With respect to preposed relatives, Held (1957:4–13) notices two distinct patterns: wh-phrase initial
relatives, which he terms indeterminate, and relative clauses with at least one accented constituent preceding the wh-element, which he terms determinate relative
clauses.
Expanding on Hale (1987) and Held (1957), Garrett’s proposal (1994) has the
advantage of formalizing the semantics of Hittite correlatives based on typological evidence following Srivastav (1991), who postulates that preposed relative
clauses have the semantic property of being quantificational, whereas postposed
relatives are non-quantificational.
Concerning preposed relatives, according to Garrett (1994:44), determinate
relative clauses denote objects or individuals whose existence is presupposed in
the discourse and in which the relative NP is definite; whereas for indeterminate
relative clauses, the relative NP is indefinite and non-specific, denoting an
indefinite entity whose existence may be in doubt. To illustrate, the following
minimal pair of relatives can be found in an omen text. Example (21) occurs in
the text before the precise nature of the votive offering has been determined,
whereas, when example (22) occurs further along in the text, more information
has been given about the nature of the goddess’ anger and which votive offerings
are to be made:
(21)
kui
IKRIBU
arninkuwa
n=an
REL-nom votive offering compensation-gen.sg conj=it-acc
arninkanzi
give compensation-3pl.pres
“Any votive offering which is to be made in restitution, they will make it in
restitution.” (KBo 2.2 iii 33f)
12
Mattyas Huggard
(22)
nu IKRIBUI.A=ma
kuie
arninkue
conj votive offering-pl=Top REL-nom.pl compensatory
n=a
arninkanzi
conj=it-acc.pl give compensation-3pl.pres
“But the votive offerings which are to be made in restitution, they will make
in restitution.” (KBo 2.2 iv 7f)
In both instances, Garrett (1994:46–7) proposes that the wh-phrase undergoes whmovement in the relative clause, followed by a further “fronting” process of the
relative pronoun in the case of indeterminate relatives, and of any accented constituent in the case of determinate relatives. In the case of the determinate RC,
this Front position must be obligatorily filled by a single syntactic constituent as
in (23):
(23)
Hittite determinate RC syntax (= (22))
In the Hittite indeterminate RC, this Front position is occupied by the relative
pronoun itself, as in (24). Although this proposal permits retaining Held’s
classification of Hittite RCs, it nevertheless falls short of accounting for the variety of internal orderings of constituents observed within preposed RCs in the corpus (in some cases with more than one constituent preceding the relative
element), and needs to be updated with respect to current syntactic theory. Closer
examination of Hittite determinate RCs shows the possible orderings in (25).
(24)
13
Hittite indeterminate RC syntax (= (21))
(25)
a.
b.
c.
d.
[ [ N + REL ] …
[ XP [ REL + N ] …
[ XP [ N + REL ] …
[ N XP [ REL ] …
I shall here propose a modified version of the approach of Garrett (1994) to
Hittite relativization, taking into account Goedegebuure’s (2009) treatment of
Focus operations in Hittite interrogatives combined with the “raising” analysis
proposed by Kayne (1994:85–97). This raising analysis has been adopted by
Mahajan (2000) to provide a unified account for the various possible outcomes of
relativization in Hindi, and also by Bianchi (2000), who developed a diachronic
proposal to relate the passage from the correlative to embedded relative strategy
in Latin. In the following section, I shall propose that the internal structure of
preposed RCs in Hittite parallels that of Hittite interrogatives with respect to focus-topic configurations, which in turn will affect the internal ordering of constituents within the preposed RCs.
4. Consequences of Focus and head raising interaction in Hittite
By incorporating a head-movement approach like that of Kayne (1994:85–97)
and Bianchi (2000) to the analysis of Hittite correlatives, along with the model of
Goedegebuure (2009) for Focus processes in interrogatives, one can suggest a
14
Mattyas Huggard
unified analysis for the variety of ordering of the consituents in Hittite preposed
relatives, whether indeterminate or determinate. Let us examine the following
examples of Hittite preposed RCs:
(26)
Preposed RC with ordering [ REL + N … ], traditionally termed indeterminate
nu=i=an
kuit
aan
LUGAL-u d
i
conj=3sg.dat=ptc REL-n.acc service-n.acc King-nom put-3sg.pres
nu ap
t
ai
conj that-acc do-3sg
“Any service which the King imposes on him, he shall do that.” (KBo 6.4 iv
15–16)
(27)
Preposed RC with ordering [ NP + REL … ], traditionally termed determinate
nu=za=kan zaimus
kuie
ukezzi n=a
conj=refl=ptc dream-acc.pl REL-acc.pl see-3sg conj=3pl.acc
memikezzi
tell-3sg
“The dreams which he sees, he tells them.” (KUB 7.5 iv 5–7)
(28)
Preposed RC with ordering [ XP [ REL + NP ] … ], traditionally termed determinate
“I fought against Mt. Arinnanda. The Sun Goddess of Arinna, my Lady, the
Stormgod, my Lord, Mezzulla and all the gods ran before me. I conquered
Mt. Arinnanda.
d
UTU-I
kuin
NAM.RA
INA
nu=za
conj=refl Majesty=My REL-acc deportee-coll.acc in
É.LUGAL uwatenun
n=a
1 SIG7 LIM 5 ME
Palace
bring-1sg.pret conj=3sg.nom 15‚000
500
NAM.RA ta
deportee be-3sg.pret
The deportees whom (I) My Majesty brought into the palace, they were
15,500.” (KBo 3.4 ii 38–42, Ten Year Annals of Murili, year 3)
(29)
15
Preposed RC with ordering [ XP [ NP + REL … ], traditionally termed determinate
“When the captives prostrated themselves at my feet, I brought the captives
down from Mt. Arinnanda, and I brought into my house 15,500 captives.
URU
attua=ma=za
ÉRIN.ME ANE.KUR.RA.ME
attua-gen.sg=Top=ptc infantry
cavalry
ÉRIN.ME
arikua=a
NAM.RA
kuin
uwatet
arikuwa troops=conj deportee-coll.acc REL-acc bring-3pl.past
nu=an
kappwauwar NU.GÁL ta
conj=of them counting
none
be-3sg.pret
The deportees whom the infantry and cavalry and arikuwa troops of
attua brought, there was no counting of them.” (KUB 14.16 iii 15–22,
Ten Year Extended Annals of Murili, year 3)
(30)
Preposed RC with ordering [ NP XP REL … ], traditionally termed determinate
“‘Oh, our lord, transfer us for settling, lead us to attua! Make us into
infantry and chariotry!’ Then I, My Sun, did not conquer the city of
Dukkama for plunder, I transported the captives.
URU
Duggama
3 LIM NAM.RA
INA
NAM.RA
deportee-coll.acc Dukkama-gen.sg 3‚000 deportee-coll.acc into
É=YA
house=my
kuin
uwatenun
n=an=za=an
REL-acc bring-1sg.past conj=them=ptc=ptc
ÉRIN.ME ANE.KUR.RA.I.A iyanun
infantry
cavalry
make-3pl.pret
The deportees of Dukkama, 3,000 deportees, whom I brought into my
house, I made them infantry and cavalry.” (KBo 4.4 iv 22–7, Ten Year Extended Annals of Murili, year 10)
4.1. Indeterminates
A first observation which needs to be accounted for is that preposed RCs of the
indeterminate type only exhibit the type of ordering of constituents as shown in
(26), i.e., [REL + N ]. Secondly, as observed by Held (1957), the discourse environment in which indeterminates and determinates occur is regular and consistent, as discussed above in §3.2 for (21) and (22). Semantically, preposed RCs
such as (21) and (26) correspond to Goedegebuure’s (2009) IdentificationInformation focus, where new information is introduced in the discourse (i.e.,
16
Mattyas Huggard
indeterminate RCs) and whose existence is thus not presupposed. I would like to
propose SpecCP as the landing site for these types of relatives, as illustrated in
(31):
(31)
= (26)
Hence in parallel with Hittite interrogatives, indeterminate relatives can be
redefined as having the typology of Identifying-Information focus, and the morphologically identical interrogative/relative occupies the inital position of the
clause (see again table 2, first type). As such, indeterminate RCs have the unique
feature that the full focus is on the DPrel, which moves as a unit.
4.2. Determinates
Examples (27)–(30) are traditionally classified as determinates, but I would like
to suggest that each structure in fact carries slightly different semantics, resulting
in different internal structures. For these types of clauses, the relative pronoun
along with the relative DP remains in the unmarked position, but the head NP
may be focused or topicalized according to the given context in the discourse.
The semantics of preposed relatives such as (27) and (30) are compatible with an
Identifying-Selecting focus. For these types of relatives, the relative DP remains
preverbal, and the head noun is extracted out of the relative DP according to what
focus role is attributed to the head NP in the discourse. In the case of (30), the
shift of topic goes from Murili’s actions to those deportees which were brought
back to the palace, and are thus topicalized:
17
(32)
= (30)
As a side note, the postposition and separation of the relative pronoun from its
head NP by a topicalized constituent is also found in Latin embedded relatives
(Bianchi 2000:72):
(33)
istos
captivos
duos‚ / heri
quos
emi
…
those-acc prisoners-acc two-acc yesterday REL-acc I bought
a
quaestoribus
from commissioners-abl
“those two prisoners who I bought from the comissioners yesterday.” (Pl.
Capt. 110–11)
18
Mattyas Huggard
The context and semantics of the relative clause in (28) corresponds to a Reporting focus: the king is simply reporting how many deportees he brought back,
and thus the object DPrel remains in situ with all of its constituents (34).
The internal structure in (29) results from more than one syntactic operation.
First, there is a change in Topic: in the previous clause, the king is reporting his
own actions, and now the topic switches to the actions of the infantry, cavalry,
and arikuwa-troops of attua. Thus the subject DP is topicalized and marked
with the contrastive enclitic particle =ma indicating the contrast in agents from
the previous context. Secondly, the context in (29) gives the semantics of a comment focus. To expand on this type of focus, in the context The books on the desk
are gone! [Who took]Foc [them ]Top?, the question is a comment about the previous statement. In the context of a declarative, as in I put the books on the desk.
[You should read]Foc [them]Top., the focus assigned on [You should read] is that
of a comment. In the Hittite example the comment focus has the relative pronoun
19
remaining in situ in a preverbal position, but the head noun NAM.RA receives
additional focus and is fronted (35).
(34)
= (28)
It would thus appear that clauses that are marked +RELATIVE are also subject to
the same Focus/Topic patterns as for declaratives and interrogatives, which
would account for the various possible permutations of the relative pronoun with
respect to its head NP and other clausal constituents observed in the corpus. As
shown in §2.2, Hittite is a wh-in-situ language like Mandarin Chinese and does
not show wh-movement. The variety in word order in these clause types is driven
rather by the focus and topicalization phenomena motivated by the discourse.
5. Concluding remarks
In this paper I have examined some data from preposed headed relative clauses in
Hittite that appeared problematic for Garrett’s (1994) account. New proposals
which have become prominent since then, notably Rizzi’s proposal of the split
20
Mattyas Huggard
(35)
= (29)
CP system (1997) and Kayne’s analysis of relativization in terms of head movement (1994:85–97), contribute to providing a framework which accounts for the
data which was problematic for previous descriptions of the internal structure of
preposed relatives. The observation made by Goedegebuure (2009) that focus
assignment rules are the same for interrogatives and declaratives may be extended to relatives as well. These rules explain the different possible permuta-
21
tions observed for the preposed relative-clause type in Hittite, while maintaining
the semantic asymmetries observed by Held (1957) and formalized by Garrett
(1994).
The next step would be to examine preposed free relatives, postposed relatives, and Old Hittite embedded relatives, in order to verify whether the present
proposal holds for them as well. Furthermore, comparative work based on Hittite,
Ancient Greek, Latin, and Sanskrit may also provide support for a more extensive
use of the left periphery in other daughter languages of PIE, as well as providing
further clues for the role of pragmatics in PIE syntax and surface word order.
References
Adger, David. 2003. Core Syntax: A Minimalist Approach. New York: Oxford University
Press.
Alexiadou, Artemis, Paul Law, André Meinunger, and Chris Wilder (eds.). 2000.
The Syntax of Relative Clauses (Linguistik Aktuell 32). Amsterdam/Philadelphia:
Benjamins.
Bianchi, Valentina. 2000. Some Issues in the Syntax of Relative Determiners. In
Alexiadou et al. 2000, 53–81.
Dayal, Veneeta. 1995. Quantification in Correlatives. In Emmon Bach, Eloise Jelinek,
Angelika Kratzer, and Barbara H. Partee (eds.), Quantification in Natural Languages, 179–205. Dordrecht/Boston/London: Kluwer.
Garrett, Andrew. 1994. Relative Clause Syntax in Lycian and Hittite. Die Sprache 36.29–
69.
Goedegebuure, Petra. 2009. Focus Structure and Q-word Questions in Hittite. Linguistics
47/4.945–69.
Hale, Mark R. 1987. Studies in the Comparative Syntax of the Oldest Indo-European
Languages. Ph.D. diss., Harvard University.
Held, Warren H., Jr. 1957. The Hittite Relative Sentence. Language 33/4 II.3, 7–52.
Hoffner, Harry A., Jr., and H. Craig Melchert. 2008. A Grammar of the Hittite Language
I: Reference Grammar (Languages of the Ancient Near East). Winona Lake, IN:
Eisenbrauns.
Kayne, Richard S. 1994. The Antisymmetry of Syntax (Linguistic Inquiry Monograph 25).
Cambridge, MA: MIT Press.
Mahajan, Anoop. 2000. Relative Asymmetries and Hindi correlatives. In Alexiadou et al.
2000, 201–30.
Probert, Philomen. 2006. Clause Boundaries in Old Hittite Relative Sentences. Transactions of the Philological Society 104.17–83.
Rizzi, Luigi. 1997. The Fine Structure of the Left Periphery. In Liliane Haegeman (ed.),
Elements of Grammar, 281–337. Dordrecht/Boston/London: Kluwer.
22
Mattyas Huggard
Srivastav, Veneeta. 1991. The Syntax and Semantics of Correlatives. Natural Language
and Linguistic Theory 9.637–86.
de Vries, Mark. 2002. The Syntax of Relativization. Ph.D. diss., University of
Amsterdam.
Focus structure and Q-word questions
in Hittite1
PETRA GOEDEGEBUURE
Abstract
The relationship between question words (Q-words) and focus is far more
complex than generally assumed, in FDG or elsewhere. Q-words are not always contrastive, the focus of a Q-word clause can have more elements than
the Q-word in its scope, and Q-word clauses may contain another, separate
focus besides the Q-word. This becomes apparent when studying languages
in which the Q-word seems to be freely distributed across the clause. A
clear example is Hittite, an extinct Indo-European SOV language in which
the placement of Q-words is not restricted to the preverbal focus position
expected for SOV languages. In this article I will show that Hittite uses
multiple strategies for the placement of Q-words, depending on the degree
of counterexpectancy of the Q-word question. This motivation for the placement of the (always focused) Q-constituent is shared with the placement of
focus constituents in Hittite declaratives. We should thus abandon the notion of a special question focus type, and describe the focus structure at the
interpersonal level in FDG irrespective of the type of illocutionary act. As a
corollary, the existential presupposition of a Q-word question, often taken
to be its only pragmatic presupposition, is detached from its focus structure.
In this article the existential presupposition is considered an inherent property of the Q-word, not of the question. Another consequence is that an FGtype focus typology as already established for declaratives can be set up for
Q-word questions as well.
1.
Introduction
The typological literature on SOV languages does not provide an adequate descriptive model for the placement of Q-words. The reason is
probably that the relationship between Q-words and focus is more
complex than is generally assumed. Typically, Q-words are taken to be
Linguistics 47–4 (2009), 945–969
DOI 10.1515/LING.2009.033
0024–3949/09/0047–0945
6 Walter de Gruyter
946
P. Goedegebuure
intrinsically focused. Because Q-words generally pattern with identificational (or contrastive) focus elements, it is often claimed that they occur
in preverbal position in SOV languages (Givón 2001: 306; Kim 1988;
Kiss 1998a: 772, 1998b). On the other hand, Dik (1997b: 285) finds that
Q-words in SOV languages prefer pattern position or initial position. If
these observations are to be reconciled, we must initially dissolve the ties
between the placement of Q-words and special focus positions and refine
our typology of focus with respect to interrogatives.
An important step towards the explanation of these contradicting typological observations is the recognition that Q-word questions have
fully articulated focus structures (cf. Drubig 2003: 7 on declaratives and
Erteschik-Shir 1986: 126–129, 1997: 134; Hajičová 1983; Jacobs 1991:
201–202; Lambrecht 1994: 55, 239; and Rosengren 1991 on other illocution types), including di¤erent types of focus that prefer di¤erent positions in the clause. The implication is that Q-words are not always in contrastive focus, and there may be other constituents in focus besides the
Q-word.
Hittite, an SOV language, has multiple Q-word and focus positions,
making it an ideal candidate to investigate the nature of the relationship
between focus and Q-words in SOV languages. At present, Hittite Qwords are believed to occur in ‘‘a special position’’ (Ho¤ner 1995; Hackstein 2004: 351 n. 7). Ho¤ner, for instance, notes that
the interrogative word sometimes precedes the finite verb as closely as possible.
And since the finite verb is usually clause final, the interrogative word gravitates
to the end of the clause. In very short clauses, it is usually initial. (1995: 101)
In this article I show that these observations need to be modified: Hittite
Q-words, like focus constituents in DECLs, occur in pattern position, initial position or preverbal position. I will suggest a framework to account
for the relationship between focus and Q-words at the interpersonal level
of FDG (Sections 3 and 4), after providing a description of Hittite Qword questions (Section 2).
2.
Hittite Q-word questions
Hittite has virtually no means to distinguish a yes-no or Q-word question
from a DECL. There are no question particles, there is no word order inversion, and although questions are expected to receive special phonological treatment, this is only rarely indicated in the script.2 Only in Old Hittite, could a yes-no question be marked by means of an otherwise highly
unusual plene writing3 of the final syllable of the final word (1). This is
947
di¤erent for the Q-word question in (2), where the final syllable of the
nonfinal Q-word is singled out for plene writing (this is the only case I
know of where lengthening of the final syllable of a Q-word is indicated
in the text):4
(1)
nu
kissan
AWAT
ABI¼YA
arhān
˘
conn this way command father¼my keep:ptcp.N.acc.sg
har-tenı̄
˘have-2pl.npst
‘Is this the way you have kept my father’s command?’
(OH/OS5 instruction, a56)
(2) The War God says:
takku natta¼ma kuwatka pai-mi
ug¼a
if
not¼but
perhaps go-1sg.npst I:nom.sg¼but
kuı̄-t
dā-hhi
˘˘
what-N.acc.sg take-1sg.npst
‘ ‘‘But if I perhaps do not go, what will I take?’’ ’
(OH/OS myth, a 50b)
Even the presence of Q-words in a sentence does not always unambiguously indicate the illocution type. The only genuine Q-word is kuwat
‘why’, cognate with Latin quod. The other Q-words are formally similar
to several subordinating conjunctions (mahhan ‘how’, mān ‘how’) and rel˘ ˘ The Q-words studied here
ative pronouns (kui- ‘who’, kuwapi ‘where’).
are all formed on the interrogative/indefinite/relative stem ku-, and will
henceforth be referred to as ku-words.
My corpus consists of 100 clauses, of which 93 are from Ho¤ner
(1995). I will only discuss direct questions with a verbal predicate. Of the
100 ku-word clauses, 46 contain only two constituents. These need to be
excluded because it is impossible to determine whether the ku-word occurs in preverbal or (modified) initial7 position. The remaining 54 kuwords show the following distribution (see Table 1):
Table 1. Relation between linear position and grammatical role of ku-words
position #
S
O
Obl
Adv
Total
initial
preverbal
Total
6
1
7
0
5
5
1
2
3
12
27
39
19
35
54
Table 1 suggests that the position of the Q-word depends on its grammatical role: the subject usually occurs in initial position (3), the object occurs
in preverbal position ([2] and [4]) and the oblique forms (5) and the
948
P. Goedegebuure
ku-adverbs (6) are at first sight distributed freely across the clause. This
corresponds to the unmarked word order in DECLs, which is SOV, suggesting pattern position for the ku-words.
(3)
(4)
(5)
(6)
The Divine Queen said to the companions of the deity Pirwa:
d
kui-s¼war¼an
hara-n
Pirwa[-i ]
˘
Who-nom.sg¼quot¼him:acc.sg eagle-acc.sg deity Pirwa[-dat.sg]
uru
Hassuw-aza uwate-z[zi ]
city ˘
Hassu-abl
bring-3sg.npst
‘ ‘‘Who will bring the eagle from the city of Hassu [to] Pirwa?’’ ’
(OH/NS myth, a 38a)
O my father, you built a house for yourself, and made it as high as
a door (?). In width you made it 9 ‘bones’ (¼ c. 3 m.).
arha¼ma¼kan
kui-t
da-tti
˘
away¼but¼ptcl
what-N.acc.sg take-2sg.npst
‘But what will you take away (from it)8?’ §
(NH proverbs, a52)
‘‘Go (and) search for the Stormgod!’’ . . . Thus (said) the father of
the Stormgod:
nu¼war¼an
kuwapi pai-mi
sanh-mi
˘
conn¼quot¼him:acc.sg where go-1sg.npst search-1sg.npst
‘ ‘‘Where shall I go (and) search for him?’’ ’
(OH/NS myth, KUB 33.24 þ 28 i 47)
[t]uk¼ma
ŠEŠ-tar
kuwatta ser [h]atrā-mi
˘
you:dat.sg¼but brother-hood why
write-1sg.npst
‘But why should I write about ‘brotherhood’ to you?’ (Are you and
I born to the same mother?)
(NH letter, a73)
One of the INTERs in the corpus shows a marked OSV word order.
In Hittite DECLs this always points at replacing/selecting focus on the
subject NP (see Section 4.2). This is indeed also the case in our deviant
example:
(7)
The one you conducted to the city Arinna, that contestor of mine,
is he not my donkey? I will sit on him! You shall lead me (there
instead)!
KUR.KUR.HI.A[¼m]a
hūman
kui-s
˘all:N.acc.sg. who-nom.sg
˘
Country.country.PL¼but
har-zi
˘hold-3sg.npst
‘Which one (of us) holds all countries? (Do not I fix the rivers, the
mountains and the seas in position?)’
(OH/NS legend, a13)
949
This question occurs in a context in which the audience has chosen not
the speaker but the speaker’s enemy as the leader. The notion that a leader typically controls the country can be assumed to be part of general
background information, shared by speaker and addressees. The only informationally salient part in this question is ‘which one (of us is the one
who holds all the countries)’. The ku-phrase therefore constitutes the focus and is moreover selective, one of the contrastive focus types in functional grammar (see Section 3).
The position of the subject kuis in (7) might be a statistically irrelevant
deviation (2%). The observation however that this Q-word question shows
selective focus on the Q-word, combined with the fact that in DECLs this
type of focus leads to a preverbal position of the focused constituent, warrants further research into the focus assignment rules of Q-word questions. In Section 3 I will present an overview of the focus model used
here, which I then will apply to the Hittite material in Section 4.
3.
The focus model, terminology and definitions
The model to be presented here is based on the focus typology in FG and
the description of the pragmatic functions indicating saliency in FDG (Dik
1997a: 330–335; Hengeveld and Mackenzie 2008; Keizer and van Staden
this issue), with some modifications. Focus is understood there as signaling
the speaker’s strategic selection of new information. However, given information, that is, contextually or situationally available information, can be
(contrastively) focused as well, as in (8) (Lambrecht 1994: 211, Ex 5.1 0 ):
(8)
Q:
A:
Where did you go last night, to the movies or to the restaurant?
We went to [the restaurant]ContrFoc .
Therefore, as already argued by Lambrecht (1994: 46–50, 209), the notion of ‘‘newness’’ should only be applied to the status of the relationship
between the information contained in the subacts, not to the discourse informational status of a certain subact. These considerations lead to the
following definitions of focus and presupposition in this article:
(9)
a.
b.
A referential or ascriptive subact stands in a focus relation to
the rest of the communicated content when that subact stands
in a pragmatically new relation to the remaining subacts of the
communicated content (for the addressee, as estimated by the
speaker), irrespective of contextual or situational availability of
the subacts themselves.
When the speaker assumes that the addressee is already aware
of some relation between subacts, that is, the relation between
950
P. Goedegebuure
subacts is contextually or situationally given, focus is not
assigned (though topic may be assigned). Such a given relation
may not only hold within a communicated content but also
between a subact in the communicated content and the
preceding discourse or extra-linguistic world. The pragmatically
given relations belong to the set of pragmatic presuppositions.
Additionally, focus can be specified for contrast. Hengeveld and Mackenzie (2008: 99–100) (see also Keizer and van Staden this issue) do not treat
noncontrastive focus and contrastive focus as subtypes of focus but derive
the latter by adding the independent function contrast to an as yet undifferentiated focus. Contrast indicates a pragmatic relation between subacts
of the same category within di¤erent acts (whereas focus indicates a pragmatic relation between subacts within one act).
The most important functional distinction within focus typology is between noncontrastive or information focus and contrastive focus.9 Information focus signals that a subact stands in a new pragmatic relation
with another subact in a single communicated content, whereas contrastive focus signals that the subact is also related to information contained
in the contextual component. The latter type of focus covers replacing
(‘‘[Not X but] Y’’), restricting (‘‘[Not X and Y,] only Y’’), selecting (‘‘Y
[out of X or Y]’’), rejecting (‘‘Not X’’) and expanding focus (‘‘[Not just
X,] also/even Y’’) (Dik 1989: 281–284, 1997a: 331–334).
It should be noted that ‘‘contrast’’ in the FDG sense (Hengeveld and
Mackenzie 2008; Keizer and van Staden this issue), as employed here, is
di¤erent from other uses of this term. Erteschik-Shir (1997) and Kiss
(1998a, 1998b), for example, define contrast as a notion that only applies
when an element in focus is selected from a closed set of alternative entities provided in the context (see 4.5 for further discussion). Chafe (1976:
34), on the other hand, assigns contrast when the focused element is contradicting an item or selecting the correct item from a set of limited alternatives that is available in the addressee’s mind according to the speaker.
These notions of contrast only capture replacing, restricting and selecting
focus, because those three types all exhaustively identify a subset for
which the proposition holds true, to the exclusion of the complementary
subset. In FDG, however, contrast means that there are di¤erences or
similarities between a communicated content and contextually available
information, without reference to the notion of exclusion. This broad
view on contrast allows the inclusion of additive focus adverbs such as
also or even (both expanding focus) in a contrastive focus typology.
Another factor which needs to be incorporated in a focus typology is
the degree of counterexpectancy. This notion is necessary in order to dis-
951
Figure 1. Three focusing strategies
tinguish between the two types of expanding focus also and even, which
are both contrastive and nonexhaustive: the focus adverb even has a high
degree of counterexpectancy whereas also has none.
Not every language will need these three factors — exhaustiveness,
contrast and counterexpectation — to describe its focus typology. Hittite
however makes use of all of them. As I will show in Section 4, only the
degree of counterexpectancy governs the position of the focused NP,
while exhaustiveness and nonexhaustive contrast receive optional marking by means of clitics. Still, I will use contrast as the cover term, since it
captures all noninformation focus.
Contrastive and information focus alike can have scope10 over a referential subact (R), a combination of subacts which typically includes a
nonembedded ascriptive subact (e.g., T, R), or the communicated content
itself (C).11 These domains correlate with three communicative goals,
identifying a referent, commenting on a topic and reporting an event or
presenting a new discourse referent (Lambrecht 1994; Van Valin and LaPolla 1997: 206). In this article I will use identifying focus, comment focus,
and reporting focus to capture these di¤erent communicative goals and
scope domains.
The three major focusing strategies are graphically represented in Figure 1. The lines indicate the pragmatic relations. The subset of subacts
maintaining a focus relation with the subset of subacts that are part of
the presupposition, and the focus relation itself are indicated in bold. All
presuppositional material including the given relations is represented by
thin lines (I have excluded discourse relations).
These graphs can be represented as follows in FDG notation (Table 2).
The braces mark the subsets. As with the graphs, boldface indicates that
Table 2. Representation of pragmatic relations at the interpersonal level in FDG
Communicative Identifying
goal
Commenting
Reporting
Focus type:
Identifying focus
Comment focus
Reporting focus
FDG
C:[{(R1 )}Foc {(R2 )(T1 )}](C) C:[{(R1 )}{(R2 )(T1 )}Foc ](C) C:[{(R1 )(R2 )
(T1 )}](C)Foc
Representation:
952
P. Goedegebuure
the subset stands in a focus relation to the rest of the C, roman type that
the subset is part of the presupposition.
4.
Focus and Q-word questions in Hittite
I will use Hittite to illustrate how the focus model presented above is capable of including Q-word questions besides DECLs. For ease of comparison I will use the following set of English Q-word questions with di¤erent elements in the scope of the focus:12
(10)
a.
b.
c.
4.1.
(RF) Someone notices that the cookies are gone, and says:
‘‘[Who ate the cookies?]RFoc ’’
(CF) With verbalized context: ‘‘The cookies are gone!
[Who ate]CFoc [them]Top ?’’
(IdF) speaker: ‘‘Someone ate the cookies’’ — addressee:
‘‘[Who]IdFoc ate them?’’
The presuppositions of Q-word questions
In order to find the focus structure of an INTER we need to divert from
the common practice of automatically equating the existential presupposition entertained by just the speaker with the presupposition of an identifying focus structure, and the Q-word with the identifying focus. Instead,
we need to concentrate on which presuppositions are shared between
speaker and addressee. This ultimately leads to the recognition that other
parts of the clause may be in focus besides the Q-word.13
Nevertheless, the existential presupposition entertained by the speaker
is considered an important property of Q-word questions (Comorovski
1996: 23–24; Dik 1997b: 264; and many others). This existential presupposition is captured in FDG as follows (Hengeveld and Mackenzie 2008:
122). The Q-word is treated as a nonspecific (s) and identifiable (þid)
referential subact (R) at the interpersonal level, in FDG notation: þid
s R (see Keizer and van Staden this issue). This type of act, also called
attributive (Dik 1997a: 189; further references in Lyons 1999: 171–172),
does not allow cancellation of the existential presupposition it contains.14
Thus, the existential presupposition is evoked by the representation of the
Q-word itself. The presupposition of a focus structure on the other hand
requires separate marking at the interpersonal level (see Figure 1 and
Table 2).
Because of its communicative saliency the Q-word is always in focus.15
The full representation of a Q-word is therefore Bid Cs R.
4.2.
953
Identifying focus and Q-word questions
Goedegebuure (2003) has shown for Hittite DECLs that di¤erent types of
identifying focus, such as expanding and replacing focus, are correlated
with di¤erent word order patterns. Additionally, focus can be marked by
means of the enclitic -ya ‘and, also, even’ or the enclitic of specification,
limitation and identity -pat ‘only, the same, even, etc.’ (Güterbock and
Ho¤ner 1995: 212–230).
The driving force behind the placement of identifying focus elements in
Hittite is the degree of counterexpectancy, and not exhaustiveness or contrast in the FDG sense. Absence of counterexpectancy is connected with
initial position, presence of counterexpectancy is connected with preverbal position. However, exhaustiveness and contrast still play a subordinate role: only exhaustive focus can and sometimes must be marked by
means of the enclitic particle -pat (this marking is optional for replacing/
selecting focus but obligatory for restricting focus). In that case there do
not seem to be any restrictions on placement in the clause (but further
study is needed). Nonexhaustive contrastive focus is always marked by
means of the enclitic particle -ya ‘also, even’.
Table 3 shows how each combination of particle and position uniquely
determines a type of focus, with its own template. The only exceptions are
replacing and selecting focus, which receive the same expression. Contrastiveness in the FDG sense is indicated by means of eC, exhaustiveness by
means of eE.
In the remainder of this article I will test whether the placement of Qwords conforms to the patterns established for the di¤erent types of focus
constituents in DECLs.
Although the three instances of oblique ku-words seem to be freely
distributed through the clause (see Table 1), whether functioning as the
indirect object in ditransitives or as the complement of a directional or
Table 3. Identifying focus in Hittite declaratives
Non-counterexpectant
1. Information (C, E)
2. Expanding ‘also’ (þC, E)
Counterexpectant
3. Expanding ‘even’ (þC, E)
4. Replacing/Selecting (þC, þE)
5. Restricting (þC, þE)
Particle Position in
the clause
Template
ø
-ya
initial
initial
(nu) Foc . . . . . .
(nu) Foc¼ya . . . . . .
-ya
ø
-pat
-pat
preverbal
preverbal
unrestricted
unrestricted
X ...
X ...
no template
no template
V
V
Foc¼ya V
Foc
V
954
P. Goedegebuure
locatival verb, they nicely follow the patterns established for identifying
focus constituents in DECLs.
The focus expression kuwapi ‘where(to)’ in (5) is found in initial position, even though its position before the verb pai- might imply otherwise.
Usually the phraseological verbs pai- ‘to go’ and uwa- ‘to come’ occur in
initial position, followed by other constituents and finally the main verb
(Van den Hout 2003: 184–188). Here, however, paimi ‘I go’ is in preverbal position, pushed aside by kuwapi in initial position. According to
Table 3, a1 this could point at identifying information focus for kuwapi,
and this is indeed the case. (5) is a rare example of a reactive move with
more than two constituents. As a result the rest of the clause, ‘the going
and searching for him’ is part of the common background; because there
is only one element in focus, kuwapi, we have an identifying focus structure. Additionally, kuwapi does not occur in a counterexpectant context,
which means that we have identifying information focus.
The representation of (5), an identifying information focus structure, is:
(11)
(AI : [(FI : INTER (FI )) (PI )S (PJ )A (CI : [{(Bid Cs RI ) Inf IdFoc }{(RJ )
(RK )(TI )}] (CI ))](AI ))
The other instances of kuwapi ‘where(to)’ are found in identifying replacing focus, without -pat. This adverb is accordingly found in preverbal
position (Table 3, a4), as (12) illustrates:16
(12)
§ Pulli just wrote me from the city of Kasipura: ‘‘It is Himuili who
does not give the seed for the plowed fields of Dapikka and
Tahasara: there is no seed!’’ § The seeds about which you Himuili
said to me: ‘‘Seeds have now been sown in Dapikka, and some in
Anziliya, some in Hariya, and some in Hanikka’’,
nu
ap-ē
numun H˘ I:A kuwapi p[ai-]s
conn dem-N.acc.pl seedpl
whereto give-2sg.pst
‘where to have you g[iven] those seeds instead?’
(MH/MS letter, HKM 55 obv. 16–17).
The representation of the identifying replacing focus structure of (12) is
given in (13); (14) represents the structure of (7) with selecting focus on a
Q-word:
(13)
(14)
(CI : [{(RI )Top? (RJ )(TI )}{(Bid Cs RK )Rep/S IdFoc }] (CI ))
(CI : [{(Bid Cs RI )Rep/S IdFoc }{(RJ ) (TI )}] (CI ))
In (13) the ‘þid s RK ’ act is the Q-word, the RI -act stands for ‘you’, and
the RJ -act represents ‘those seeds’. The combination {(RI )(RJ )(TI )} ‘you
giving those seeds’ forms the presupposition, in this case because it is part
of the contextual component (the shared background).17 In (14) {(RJ )
955
(TI )} ‘controlling the countries’ forms the presupposition (see the discussion at [7]).
Whereas Hittite subacts in identifying replacing focus are marked by
means of word order, English adds for example the intonationally prominent instead to di¤erentiate between identifying replacing focus ([15a] and
[15b]) and identifying information focus (10c), repeated here as (16a):
(15)
a.
(16)
b.
a.
b.
‘‘Mary ate the cookies’’ — ‘‘No she didn’t’’ — ‘‘Then [who]foc
ate the cookies/them [instead ]foc ?’’
(CI : [{(Bid Cs RI )Repl IdFoc }{(RJ ) (TI )}] (CI ))
Speaker: ‘‘Someone ate the cookies’’ — addressee: ‘‘[Who]foc
ate them/did?’’
(CI : [{(Bid Cs RI ) Inf IdFoc }{(RJ ) (TI )}] (CI ))
The remaining ku-words with seemingly free distribution are the two kuadverbs kuit and kuwat, both ‘why’. They both occur in initial or preverbal position, although preverbal position occurs twice as often as initial
position (see Table 1, the adv column). In view of the patterns for kuword questions established above, it is not unreasonable to assume that
the position of these ku-adverbs is governed by focus assignment as well.
The state-of-a¤airs (SoA) questioned by a why-clause often coincides with
the shared background. The SoA is therefore part of the presupposition,
and only the Q-adverb is informationally salient: therefore only the Qword is in focus. If this is the case, there are two options. The SoA is either neutral regarding speaker expectations (the [a] examples in [17] and
[18] below) or it is counterexpectational (the [b] examples):
(17)
(18)
a.
The Council nominated them (the traitors) for the death
penalty. But I, the King, said:
[kuwa]t¼war¼e
akk-anzi
why¼quot¼they:3.pl.nom die-3pl.npst
‘[Wh]y must they die?’
(OH/NS narration, a82)
b. Do scribes perform feudal duties?
apiya¼ma¼at
kuwat issa-i
there¼but¼it:N.acc.sg why
do:iter-3sg.npst
‘Then why is he performing them there (with you)?!’
(MH/MS letter, a101)
Mary ate the cookies
a. Why (did she eat them)? Was she hungry?
b. Why (did she eat them)?! She was not supposed to!
In the (a) examples the speaker accepts the fact that SoA took place, having no prior assumptions. The focus is informational and has only the
956
P. Goedegebuure
why-phrase in its scope because Mary ate the cookies is presupposed. In
the (b) examples, on the other hand, the presuppositions behind the expressions seem to clash with the speaker’s prior assumptions. The speaker
cannot endorse them, and the focus type seems therefore to be counterexpectational, i.e., a contrastive focus type. In Hittite the speaker sometimes supports a counterexpectational ku-question by adding a rhetorical
question emphasizing the normal situation, as in (6) and (17b). The
speaker assumes that the SoA of the INTER was not supposed to obtain:
X is the case whereas ‘‘everyone’’ knows X should not happen at all, so
the speaker wants to know the reason for the rejection of not-X (‘‘why
not [not-X]’’ ¼ ‘‘why X’’). On the other hand, in (19) the SoA, ‘sending
the envoys back to me’ (¼ X), was supposed to happen. After all, it is
common practice to return one’s envoys. Now the speaker asks, ‘‘why
not X?’’:
(19)
:MES
LU
TEMI¼YA¼mu
kuwat UL
˙
messengers¼my¼me:1.sg.dat
why
not
uie-ski-ttani
send-iter-2pl.npst
‘Why are you not sending my envoys (back) to me?!’
(MH/MS letter, a95)
man:pl
In view of my analysis of why-questions as either ‘‘why not X’’ or ‘‘why
not (not-X) ¼ why X’’, I tentatively classify the focus type of the counterexpectational why-phrase as rejecting. The why-phrase has scope over the
negation, whether expressed or not.
To summarize, in Hittite all counterexpectational why-phrases,18 kuwat
with or without negation, occur in preverbal position. But without counterexpectation the ku-adverb appears in initial position, as in (17a). Clearly,
the position of ku-adverbs is governed by the identifying focus assignment
rules of Table 3. In English the di¤erence in communicative point between (18a) (information focus) and (18b) (counterexpectational focus) is
expressed by di¤erences in pitch accent.19
Although the SoA of a ku-adverb clause typically coincides with the
shared background, which is nonfocused, in certain contexts focus assignment is not restricted to the ku-phrase. In those cases the speaker expresses that only parts of the shared background clash with some prior
assumptions. The following two examples show preverbal constituents in
replacing focus, while kuwat is in initial position. In (20) the princess observes that her own, implicit assumption that she should receive a large
amount of goods in view of her status as daughter of the king, is changed
for some reason into ‘‘receiving this little (ki tepu)’’. She requests an explanation for the replacement of much by this little by asking a question
957
with information focus on kuwat and replacing focus on ki tepu (‘‘Not X,
but Y’’). In (21) the angry deity should have been seeking to do good for
mankind given the honours received by him, as the speaker reminds the
angry deity. The noun idalauwani ‘for harm’ replaces the expected phrase
assuli ‘for good’. Initial kuwat, in information focus, has scope over the
preverbal constituent in replacing focus.20
(20)
(21)
She (the princess) said:
[ku]wat¼mu
kı̄
tepu
pai-tta
[W]hy¼me:dat.sg this:N.acc.sg little:N.acc.sg give-2sg.pst
‘[‘‘W]hy have you given me this little bit’’.’
(OH/NS narrative, a 81)
[E]a, King of Wisdom, said to Kumarbi:
kuwat¼pat21¼wa
[ z]ik d Kumarbis dumu.lú.u19 .lu-UTTI
why¼foc.ptcl¼quot
you:nom deity K. mankind
idalauwani sanh-is[ki-si ]
˘
evil:dat.sg search-iter-2sg.npst
‘ ‘‘Why are you seeking to harm mankind, O Kumarbi? (Does not
mankind fetch a heap of grain to promptly honour you,
Kumarbi?’’)’
(NH myth, KUB 33.103 ii 9–10)
In English one could think of the following INTERs with why in information focus and with main accent on the constituent in replacing focus.
(22)
a.
b.
Why (did) SHE (eat them)? JOHN said he would eat them.
Why (did she eat) the COOKIES? I thought she wanted to eat
the CAKE.
The representations of (17a) with only information focus on the whyphrase, (17b) with counterexpectational focus on the why-phrase and
(20) with information focus on the why-phrase and replacing focus on
the preverbal constituent are (23a), (b) and (c), respectively:
(23)
a.
b.
c.
(CI : [{(RI )(TI )}{(Bid Cs RJ ) Inf IdFoc }] (CI ))
(CI : [{(RI )(RJ )(RK )(TI )}{(Bid Cs RL )Rej IdFoc }] (CI ))22
(CI : [{(RI )(RK )(TI )}{(RJ )Repl IdFoc } {(Bid Cs RL ) Inf IdFoc }]
(CI ))23
4.3.
Comment focus and Q-word questions
In comment focus clauses one of the subacts is what the other subacts
‘‘are about’’. This subact is assigned topic function. To find the sentence
958
P. Goedegebuure
topic of an INTER one can apply the ‘‘about’’-test. If the question, ‘‘Ask
(something) about X’’, can be properly answered by the question, X is the
topic of the INTER.24 As in DECLs, at least one other subact is assigned
focus function. The di¤erence with identifying focus questions is basically
one of focal scope. Whereas an identifying focus question has only one of
its referential subacts (R) in focus, the commenting Q-word question at
least holds the ascriptive subact (T) in focus.
The word order in Hittite DECL Comment focus clauses is (S)(O)V.
Again, this also applies to Q-word clauses with comment focus. There
are two Q-word questions in the corpus with contrastive subject-topics in
comment focus clauses (a 47 below, and 50b ¼ (2)):
(24)
As for my sons: if a man has been killed, can they restore him to
life? And if a cow (or) a sheep has been killed, can they restore it
to life?
[tue]ll¼a
DUMU.MEŠ¼KA kui-n
you:sg.gen¼and son.pl¼your
what-acc.sg
sagai-n
iya-nzi
miracle-acc.sg make-3pl.npst
‘And what kind of miracle can your sons perform?’
(OH/NS myth, a 47)
(24) is an instance of comment information focus, not of identifying
replacing/selecting focus on the object-ku word, as one might conclude
from the preverbal position of the ku-word phrase. Had the latter been
the case, the common background would have been: ‘‘Your (i.e., addressee’s) sons are performing some miracle, against expectations’’. Instead, it
is the actions of the sons of the speaker himself that are the topic of the
previous discussion, so speaker and addressee do not yet share information concerning addressee’s sons. In (24) the addressee’s sons are the new
sentence topic, in contrast with the preceding sentence topic, the speaker’s
sons. The rest of the clause, ‘performing some kind of miracle’ presents
the salient information. The interpersonal level representation of (24) is:
(25)
(AI : [(FI : INTER (FI )) (PI )S (PJ )A (CI : [{(Bid Cs
RI )(TI )}Inf Foc (RJ )ContrTop ] (CI ))](AI ))
In four of the comment-focus clauses in my corpus (a52, 63, 64, 120) the
referent of the subject, which is usually25 only marked on the verb, is
what the INTER is about. This referent is therefore the topic of the
clause. To show that these clauses are indeed comment focus clauses, we
may approach the analysis from a di¤erent angle with the help of
(4 ¼ a52, arha¼ma¼kan kuit datti ‘‘But what will you take away (from
˘
it)?’’). Reporting
focus in (4) is excluded because that type of focus re-
959
quires a lexically expressed subject. Additionally, the verb is not part of
the common background: the ‘taking away by father’ is not contextually
present or inferable. This excludes identifying focus for the object-ku
word. Furthermore, regarding communicative point, there is no contrast
or counterexpectancy. The FDG representation of Hittite (4) is therefore:
(26)
(AI : [(FI : INTER (FI )) (PI )S (PJ )A (CI : [(RI )Top
{(Bid Cs RJ )(TI )}Inf Foc ] (CI ))](AI ))
The topic subact usually involves the referent of the subject but this need
not be the case (Lambrecht 1994: 146, ex. 3.29):
(27)
Pat said they called her twice.
Since the clause is about ‘‘Pat’’, the referent of Pat and her, the object
pronoun her qualifies as the topic subact. This is also the case in the remaining comment-focus INTERs (a38b, 68a, 111a and KUB 33.87þ i
31 0 , 33 0 ). In each INTER the subject Q-word together with the predicate
provides the comment for the object-topic, as in (28):
(28)
(One of the gods of the elder generation, Kumarbi, is creating
gods and monsters to dethrone the Storm God Tessub, the
supreme deity. Several battles have already taken place, and this
time Kumarbi has fathered a Basalt monster. The monster, still a
child, is presented to Enlil.)
And Enlil began to say to himself: ‘‘Who is he, this child, whom
they raised again, the Fate-Goddesses and the Powerful
Goddesses?’’
kuis¼war¼as
[namma] u-ski-zzi
who-nom.sg¼quot¼them:acc.pl longer
watch-iter-3sg.npst
sallay-as
DINGIR.MEŠ-as dassaw-es
zahha-u[s]
˘˘
great-gen.pl god.pl-gen.pl
strong-acc.pl battle-acc.pl
‘Who can [any longer] watch them, the intense wars of the great
gods?’
(NH myth, a 38b)
The battles of the gods are not only part of the pragmatic presupposition,
they are also the sentence topic, marked by means of the enclitic pronoun
-as ‘them’. However, the activation state of the battles in the addressee’s
mind is very low at this point in the discourse, even though they are at the
center of attention for the speaker. The addition of a full coreferential NP
in Afterthought position (Luraghi 1990: 106–107) is the regular Hittite
‘‘repair’’ mechanism to guide the addressee in correctly identifying the
referent of an unaccented topic pronoun in Direct Discourse.
960
P. Goedegebuure
In the following example on the other hand, the referent of the objecttopic expression is discourse active for both speech participants, allowing
easy identifiability:
(29)
‘‘To whom shall I give him, this child?’’
kuis¼war¼an¼za¼an
who-nom.sg¼quot¼him:acc.sg¼refl¼him:acc.sg
[dā-i ]
take-3sg.npst
‘Who [shall take] him (and treat him as a gift?)’
(NH myth, a 68)
In the English comment focus clause (10b: ‘‘The cookies are gone! [Who
ate]foc [them]top ?’’), the cookies are discourse active when the question is
uttered. The focus of the question is the string ‘‘Who ate’’26 and is intended to request additional information about the mental representation
of the cookies. The question is ‘‘about’’ the cookies, which means that
they stand in a topic relation to the rest of the clause. Besides being a request for information about the identity of the culprit, the communicative
function of this act is to comment on the discourse and sentence topic the
cookies. The FDG representation of Hittite (28) versus (29) and English
(10b) is (30) versus (31), respectively:
(30)
(AI : [(FI : INTER (FI )) (PI )S (PJ )A (CI : [{(Bid Cs RI )(TI )}Inf Foc
{(RJ )(RK )}Top ] (CI ))](AI ))
(31) (AI : [(FI : INTER (FI )) (PI )S (PJ )A (CI : [{(Bid Cs RI )(TI )}Inf Foc
(RJ )Top ] (CI ))](AI ))
In English the ascriptive subact is intonationally prominent, and the referential topic subact is expressed by means of an unaccented pronoun. In
Hittite we also find an unaccented pronoun for the topic subact.
4.4.
Reporting focus and Q-word questions
When the speaker assumes that none of the information in the INTER is
shared with the addressee, the question is out-of-the-blue. There is no
shared pragmatic presupposition and the focus has therefore the whole
communicated content in its scope (reporting focus). INTERs of this
type could be a response to the question ‘‘What would you like to ask’’.
Such an out-of-the-blue INTER is the equivalent of an event-reporting
DECL answering ‘‘What happened’’ (Lambrecht 1994: 121, 124). The
word order in Hittite event reporting DECLs is SOV/SXV (Luraghi
1990: 85).27 The same word order is observed for Hittite out-of-the-blue
Q-word questions.
961
In (3) the speaker holds the existential presupposition that someone
should bring the eagle, but she cannot assume that the addressees hold
this presupposition as well. We are therefore dealing with reporting focus.28 Without a contrast that can be derived from the context, the focus
type is informational.29 In the few Hittite reporting information focus INTERs we find the Q-word in pattern position, whereas in English this
type of question is expressed with an accent on both the initial Q-word
and the last accentable constituent (see [10a]). The FDG representations
of Hittite (3) and English (10a) are respectively:
(CI : [{(Bid Cs RI )(RJ )(RK )(RL )(RM )(TI )30}] (CI ) Inf RFoc )
(CI : [{(Bid Cs RI )(RJ )(TI )}](CI ) Inf RFoc )
(32)
a.
b.
4.5.
Set notion of contrast versus counterexpectancy31
In Haan’s view (2002: 187–188) the Q-word always carries a focus of contrast. She ties the term focus of contrast to the notion ‘‘correct item from
a subset’’: the speaker brings the comment, i.e., the Q-word, in focus to
check whether it is the correct comment vis-a-vis other comments (Haan
2002: 184, also see Erteschik-Shir 1997; Kiss 1998a, 1998b).
However, Q-words do not universally involve a correct member or even
just a member from a subset and therefore do not always carry a focus of
contrast by this definition. This claim is borne out by the fact that set
membership does not play a role in the focus assignment rules for Hittite
INTERs. This also means that the dichotomy of identificational focus
and information focus as presented by Kiss (1998b), with preverbal position marking identificational focus, cannot be applied to Hittite.
The set notion of contrast does not explain why the ku-words in (3) and
(7) are treated di¤erently. Both involve a closed set of contextually (7) or
situationally (3) available candidates and should therefore qualify for
both exhaustively identifying and contrastive focus as defined by Kiss
(1998b: 249, 267) and Haan (2002: 188). In (3) the restricted set consists
of the servant deities of the deity Pirwa and in (7) of the contester and the
speaker. However, only the ku-word in (7) is placed in preverbal position,
the locus for counterexpectational focused constituents. In (5) and (12) we
have on the other hand two contextually and situationally unrestricted
sets of possible answers from which to choose. This excludes contrastive
focus, but the focus type is still exhaustively identifying, which would require preverbal position for both (5) and (12) within the framework of
Kiss (1998b). This is not the case, and again the di¤erence in position of
the ku-word is explained by the degree of counterexpectancy of the INTERs, without any recourse to (closed) set membership.
962
P. Goedegebuure
4.6.
Evaluation
We may conclude that Hittite ku-word questions are governed by the
same focus assignment rules as DECLs, as has been noted for other languages (see Section 1). The correspondence between focus types and patterns discussed above are collected in Table 4 (compare Table 3 for identifying focus in Hittite DECLs):
Table 4. Focus scopes and focus typology of Hittite Q-word questions
Identifying focus
Information
Replacing/Selecting/
Rejecting
Comment focus
Information, SubjQ-word
Information, ObjQ-word
Reporting focus
Information, SubjQ-word
Position in the clause
Template
Initial position
(nu) ku- . . .
Preverbal position
...
ku- V
Pattern position S(O)V
Pattern position (S)OV
(nu) ku- . . .
...
V
ku- V
Pattern position SOV
(nu) ku- . . .
V
V
(ex. 5, 17a,
20, 21)
(ex. 6, 7, 12,
17b, 19)
(ex. 28, 29)
(ex. 2, 4, 24)
(ex. 3)
Besides the fact that INTERs and DECLs have the same focus scope possibilities, the Hittite material has shown that INTERs and DECLs may
also share the same focus typology. It should therefore be possible to set
up a focus typology for non-DECLs comparable to the focus typology
as described in FG. The INTER column in Table 5 is a modification of
the table for DECLs (Dik 1997a: 332, Figure 7).32 The most important
distinction between the DECL and INTER columns is that the roles are
Table 5. Focus types in declaratives and interrogatives
DECL
Type Of Focus
INTER
Expression Type
PS ! ðPA ÞS
ðPA ÞS ! PS
X
???
X
???
What did Mary buy?
X
Y
X
Y
Restricting
X
X and Y
X
X and Y
Selecting
Expanding
X
X or Y
X and Y Y
(If not Y, then) what did
Mary buy instead?
(If not X and Y, then)
which one did Mary buy?
Which one did Mary buy?
What else did Mary buy?
Noncontrastive
Information
Contrastive
Replacing
X
X or Y
X and Y Y
963
reversed for the latter. Whereas in the DECL situation the speaker has
some information, contrastive or not, to add to the pragmatic information of the addressee (indicated by the arrow: transfer information from
(PS ) to (PA )S ), in the INTER situation the speaker assumes that the addressee has some information to add to the speaker’s pragmatic information (that is, transfer information from (PA )S to (PS )). Again this information may be contrastive or not.
Since the Q-word questions discussed in this study only showed
replacing/selecting, rejecting and information focus, Table 5, which lists
more focus and expression types, can only be considered provisional.
5.
Conclusion
Hittite belongs to those SOV languages in which the placement of Qwords is not restricted to the preverbal focus position. Instead, Hittite
uses multiple strategies for the placement of Q-words, depending on the
degree of counterexpectancy of the Q-word question. If the counterexpectancy of the Q-word question is high, the Q-word appears in preverbal
position. But if the counterexpectancy of the Q-word question is low or
zero, i.e., if there is only a request for completive information, then the
Q-word appears in initial position or in pattern position. This motivation
for the placement of the (always focused) Q-constituent is shared with the
placement of focus constituents in Hittite DECLs.
The observation that the focus assignment rules are the same in INTERs and DECLs naturally raises the issue of the status of the existential
presupposition usually connected with Q-word questions. It turned out
that the existential presupposition could easily be accounted for in the
representation of Q-words at the interpersonal level of FDG. Thus, the
existential presupposition was separated from the focus-presupposition
structure of the INTER.
Once the existential presupposition is separated from the focuspresupposition structure of the INTER, one should be able to set up a focus typology for INTERs. The focus typology of DECLs as used in FG
has been extended, with a few modifications, to a provisional focus typology of INTERs in FDG.
A necessary requirement for the retrieval of the INTER focus structures is to study INTERs in context. Because most studies on INTERs
fail to do this, it is often impossible to further qualify INTERs as out-ofthe-blue or identificational, as informational or contrastive. This omission
could be the cause for the contradictory statements on the placement of
Q-words in the typological literature, clouding the fact that Q-words are
964
P. Goedegebuure
not necessarily the only element in focus, and that the focus of the
Q-word is not necessarily contrastive.
Received 4 September 2006
Revised version received
12 June 2007
University of Chicago
Notes
1. Hittite belongs to the extinct Anatolian branch of Indo-European. It is documented in
cuneiform script on clay tablets and incidentally on metal from c.1700–c.1200 BC, and
is the oldest Indo-European language available to us through contemporaneous documents. Currently 30,000 pieces of clay tablets have been excavated, constituting about
10,000 texts. Most tablets were found in the palace and temple archives of the ancient
capital Hattusa, modern Boghazkale in the province of Çorum, Central Anatolia. See
Ho¤ner and Melchert (2008) for the standard reference work on Hittite grammar. Correspondence address: The Oriental Institute, 1155 East 58th Street, Chicago, IL 60637.
E-mail: pgoedegebuure@uchicago.edu
2. In three of the 145 questions collected by Ho¤ner (1995), the first and only large collection of Hittite questions. The Hittite interrogatives are also discussed by Mascheroni
(1980, 1981, 1983).
3. Plene writing is the addition of a vowel sign to a syllable of the type CV, that is, CV-V.
Plene writing is transcribed with a lengthened vowel.
4. In the Hittite transcriptions CAPS represent Sumerian lexemes or grammatical morphemes, ITALIC CAPS represent Akkadian lexemes or grammatical morphemes, and
italics represent Hittite. Square brackets demarcate sections that are not preserved on
the tablet. The glosses follow the Leipzig glossing rules (http://www.eva.mpg.de/
lingua/files/morpheme.html). In addition to the list of standard abbreviations I use the
following glosses: conn ¼ sentence connective, iter ¼ iterative and ptcl ¼ particle.
5. Following the conventions of the Chicago Hittite Dictionary, the sigla OH, MH, and
NH indicate that the date of a composition is Old, Middle, or Neo-Hittite. The sigla
OS, MS, and NS indicate that the date of a manuscript belongs to one of these periods.
The combination OH/NS therefore indicates that the text is palaeographically dated to
Neo-Hittite, but was composed in Old Hittite.
6. The numbers refer to the numbered questions in Ho¤ner (1995).
7. Initial position is the position of the first accented word, including the basically untranslatable sentence connectives nu, ta, su. Modified initial position is the position of
the first accented word which is only preceded by a few sentence adverbs or the sentence connectives, both optionally followed by pronominal and/or local enclitics (Luraghi 1990: 12–13). These structurally di¤erent positions are generally considered to be
functionally similar. One of the anonymous referees believes that these two positions
are actually functionally di¤erent, and I do agree with him/her to a certain degree, as
defended in my proposal for NWO project 275-70-017. However, I have not (yet) been
able to detect a di¤erence between modified initial position and initial position in the
corpus studied here, and neither in the extended INTER corpus containing only two
constituents. To cite just one example: ZI-anza¼wa¼kan uris ZI-anza¼wa¼kan uris
— (Question 1) kuel¼wa¼kan ZI-anza uris — (Answer 1) dandukēs¼wa¼kan ZI-anza
8.
9.
10.
11.
12.
13.
965
uris — (Question 2) nu kuin KASKAL-an harzi — (Answer 2) uran KASKAL-an harzi
˘ is great, the soul is great!’’ (Question
˘ 1)
marnuwalan KASKAL-an harzi ‘‘The soul
˘
‘‘Whose soul is great?’’ (Answer 1) ‘‘The mortal soul is great!’’ — (Question 2) ‘‘And
what road does it travel?’’ — (Answer 2) ‘‘It travels the Great Road, it travels the Invisible Road!’’ (OH/NS myth, a66). The Q-word kuel in question 1 occurs in initial
position, whereas the Q-word kuin in question 2 follows the sentence connective nu,
and is therefore in modified initial position. Alternatively, the constituents to which
these Q-words belong could be in preverbal position, but I will disregard this possibility
for the present discussion because this position always indicates counterexpectant focus
(see Tables 3 and 4). Both Q-words are in identifying information focus (see Section 3
for the terminology). As the cotext shows, each Q-word is the only part of its clause
which is not part of the presupposition, and which therefore is the only element in focus (hence identifying focus). Both questions are also merely asking for information
about the person whose soul is mentioned, and the road that soul will travel (hence information focus). Therefore, from an information structural point of view, the presence
of nu in the second question does not seem to matter: the Q-words in initial position
and modified initial position have the same focus scope and meaning.
Because Hittite does not have ablative pronominal clitics, unaccented from it cannot be
expressed.
To mention a few other terms for the tandem ‘information focus versus contrastive focus’: focus of assertion versus contrastive focus (Givon 2001a: 343–344, 2001b: 221–
223), presentational focus versus operator focus (Drubig 2003), presentational focus
versus contrastive focus (Rochemont 1986: 52–53), information focus versus identificational focus (Kiss 1998a, 1998b), plain focus versus contrastive focus (Ertheshik-Shir
1997), new information focus versus exhaustive listing focus (Green and Jaggar 2003).
Due to di¤erences in definition and theoretical framework a one-to-one mapping of
these terms often cannot be achieved.
Information focus often has scope over a wide domain, whereas contrastive focus generally has scope over a narrow domain. Although this correlation between domain and
communicative function is quite strong, these parameters should still be kept separate.
For example, a noncontrastive answer to a Q-word question often contains a focus
with narrow scope. The di¤erence between the parameters scope of the focus and the
communicative function of the focus has been observed by e.g., Rochemont (1986: 68–
74); Comrie (1989: 62); Dik (1989: 281–285, 1997a: 331–332); Erteschik-Shir (1997:
11–12); and Green and Jaggar (2003). The necessity to distinguish between scope and
communicative function is confirmed by experimental studies (see, for example, Toepel
and Alter 2004).
See Dik (1997a: 330–331) for more focusing strategies.
Note that ‘‘HE did/ate them/ate the cookies’’ is an appropriate answer to each question.
This shows that the focus structure of a Q-word question does not necessarily correspond with the focus structure of the answer. Also see Kiss (1998b: 249) for Hungarian.
For a similar view, see Rosengren (1991: 196, 198), who refers to the cotext and context availability of information in interrogatives. Also see Jacobs (1991: 201–202) and
the studies mentioned in Section 1. A di¤erent course is taken by e.g., Haan (2002).
Haan (2002: 189) always treats the existential speaker presupposition as part of the
common background. The presupposition is therefore the topic of the INTER, while
the Q-phrase is the comment and thus the only part of the clause in focus. Still, according to Haan, one may doubt whether the topic is part of the addressee’s background.
The speaker can therefore accentuate the topic for (re-)introduction into the discourse.
For a similar approach see Lambrecht (1994: 282–283).
966
P. Goedegebuure
14. By separating the representation of the Q-word from the representational level, which
deals with the semantics of a speech act, the inferences connected with Q-words are
also separated from semantics. This excludes the possibility that we are dealing with a
conversational implicature.
15. This is the case even when the Q-word applies to a closed set of salient discourseentities (‘‘which of . . .’’). This type of Q-word is used to instruct the addressee to select
the appropriate subset from a contextually defined set of entities. In some languages the
Q-word phrase that contains or refers to the contextually defined set may be treated as
a discourse topic. For example, ‘‘which of . . .’’ Q-word phrases receive topic marking
in Chinese (topic position, Jianxin Wu 1996), Japanese (-wa marking, Kees Hengeveld
p.c.) and Bulgarian (clitic doubling of object Q-words, Jaeger 2004). In Hittite on the
other hand, ‘‘which of . . .’’ Q-words receive focus marking (see [7] and [14]). To conclude, although the closed set may be treated as a discourse topic, the Q-word may not.
16. The remaining instance is, with context: The Storm God sent for the Sun God: ‘‘Go
(and) bring the Sun God!’’. Thereupon they searched for the Sun God, but they did
not find him. The Storm God says: ‘‘So why have you not found him?! My own limbs
are warm! [ap]ās¼a¼wa kuwapi harkta ‘‘Where else could he have gotten lost?’’ (OH/
˘ presupposition of this focus structure is that, first of
NS myth, VBoT 58 obv. 25). The
all, the Sun God is lost somewhere (leading to identifying focus in the INTER), and
secondly, that he is not found — and therefore was lost — in any of the places they
were looking for him, hence the replacing focus (‘‘where else could he be?’’).
17. I have tentatively assigned topic function to those seeds since they are what the INTER
is about. For topics in INTERs, see Section 4.3. This example is actually an attestation
of a proposition with multiple focus articulations, as described by Lambrecht (1994:
236–237), in this case an identifying focus and comment focus articulation.
18. The other instances of this type are a 70, 73, 76, 80a, 96–98a, b, 99, 100, 102–110,
111c, 113, 114a, 114b, 115.
19. Although I assume that both why-s have a H*-L% accent, (18b) probably has a higher
pitch rise and fall than (18a). Alternatively, (18b) could have a high-rise (H*-H%) to
express disbelief.
20. The other attestations are a61b, 83–85, 87, 88, 90, 92–94. The only exception known
to me is: ‘I call for mercy! My calf should have four legs! kun¼ma¼wa¼za ŠA 2
GÌR.MEŠ¼ŠU kuwat hasun ‘Why have I borne this two-legged thing?’ (NH myth,
˘
a113). Not the SoA ‘giving
birth to something’ is counterexpectant, only the fact that
the newborn has two legs. This would call for identifying replacing focus on the noun
phrase kun . . . ŠA 2 GÌR.MEŠ¼ŠU ‘‘this of two legs’’, in preverbal position, and with
kuwat in initial position.
21. None of the meanings of -pat listed in Güterbock and Ho¤ner (1995: 212–230) seems
to fit (a selection: ‘‘the same, the aforementioned, likewise, -self, own, only, also,
rather, even, surely, certainly’’).
22. RI ¼ he, RJ ¼ them, RK ¼ there, RL ¼ why.
23. RI ¼ you, RJ ¼ this little, RK ¼ me, RL ¼ why.
24. This discourse-contextual approach to recognizing topics in INTERs disregards the requirement in FG and FDG that topic function is only assigned when the subact receives special marking (Keizer and van Staden this issue). Several linguists within FG
have argued against such a formal approach to topic function assignment (see Cornish
2004: 132–133, with further references).
25. a120 contains a postverbal subject NP. For discussion see fn. 28.
26. It is important to note that the topic-comment partition of the clause does not automatically coincide with the subject NP-VP partition. Especially in INTERs the comment
may include subject and verb.
967
27. In presentational DECLs with the verbs es- ‘‘to be’’ or ki- ‘‘to lie’’ the preferred word
order is LocSV (Luraghi 1990: 105–106).
28. As one anonymous referee suggests, the eagle is apparently thematized via the presence
of a coreferential pronoun, suggesting a topic-comment structure rather than reporting
focus, with the eagle as object-topic and the rest of the clause in comment focus. An
appropriate translation would thus be ‘‘Who will bring him, the eagle, from the city
of Hassu [to] Pirwa?’’. The only reason for not following this otherwise plausible suggestion is that coreferential NPs occur in postverbal, nonpattern position if the referent
of a pronoun is the topic of the INTER (exx. a38b, 68a, 111, KUB 33.87þ i 31 0 , 33 0 ).
See a120 for a topic subject pronoun with postverbal coreferential subject NP, and
Section 4.3 for further discussion of postverbal coreferential NPs in topic-comment
clauses. It should also be noted that event-reporting or thetic DECLs allow topical
nonsubject constituents (Lambrecht 1994: 145; Rosengren 1997: 469–470). I assume
that event-reporting INTERs allow them as well.
29. The other reporting information focus examples in the corpus are a41 and KUB 8.65 i
27 (both with the subject Q-word in initial position).
30. RI ¼ who, RJ ¼ -an ‘‘him’’, RK ¼ the eagle, RL ¼ deity Pirwa, RM ¼ city Hassu.
31. One anonymous referee addresses the relationship between the adversative connective
-ma ‘and, but, while’ and the notion of counterexpectation. Although it is true that -ma
occurs in contexts in which some kind of contrast can be perceived, I do not find any
connection between -ma and counterexpectant focus. Among the 27 INTERs with
counterexpectant focus in the corpus used here, 11 contain -ma (a73, 98b, 99–102,
104, 107, 111c, 113, 115). Of these 11, 7 show an accented pronoun (usually a Subject)
in initial position followed by -ma (a73, 99, 100, 104, 107, 113, 115). This strongly suggests that -ma may be used as a contrastive Topic marker, as was already shown for
Old Hittite DECLs by Rieken (2000). Additional evidence that further weakens a possible connection between -ma and counterexpectant focus is that -ma may also occur
with an information focus in the clause, as is illustrated by (2) and (4). Finally, of the
66 examples of counterexpectant focus in DECLs with 3rd person accented pronouns
in Goedegebuure (2003: 283–319), only one example contained -ma.
32. Questioning focus as a separate category should be deleted from the focus typology in
Dik (1997a: 330–333) since INTERs may also contain the other types of focus mentioned there.
References
Chafe, Wallace. 1976. Givenness, contrastiveness, definiteness, subjects, topics and point of
view. In Charles Li (ed.), Subject and topic, 25–56. New York: Academic Press.
Comorovski, Ileana. 1996. Interrogative phrases and the syntax-semantics interface. Dordrecht & London: Kluwer.
Comrie, Bernard. 1989. Language universals and linguistic typology: syntax and morphology,
2nd revised edn. Oxford & Chicago: Blackwell & University of Chicago Press.
Cornish, Francis. 2004. Focus of attention in discourse. In J. Lachlan Mackenzie & Marı́a
de los Ángeles Gómez-González (eds.), A new architecture for functional grammar, 117–
150. Berlin & New York: Mouton de Gruyter.
Dik, Simon C. 1989. The theory of functional grammar, part 1: The structure of the clause
(Functional Grammar Series 9). Dordrecht: Foris.
Dik, Simon C. 1997a. The theory of functional grammar, part I: The structure of the clause,
edited by Kees Hengeveld (Functional Grammar Series 20). 2nd rev. edn. Berlin & New
968
P. Goedegebuure
Dik, Simon C. 1997b. The theory of functional grammar, part II: Complex and derived constructions, edited by Kees Hengeveld (Functional Grammar Series 20). Berlin & New
Drubig, H. Bernhard. 2003. Toward a typology of focus and focus constructions. Linguistics
41(1). 1–50.
Erteschik-Shir, Nomi. 1986. WH-questions and focus. Linguistics and Philosophy 9. 117–149.
Erteschik-Shir, Nomi. 1997. The dynamics of focus structure (Cambridge Studies in Linguistics 84). Cambridge: Cambridge University Press.
Givón, Talmy. 2001. Syntax. An introduction. 2 vols. Amsterdam: John Benjamins.
Goedegebuure, Petra M. 2003. Reference, deixis and focus in Hittite: The demonstratives ka‘‘this’’, apa- ‘‘that’’ and asi ‘‘yon’’. Amsterdam: University of Amsterdam dissertation.
Green, Melanie & Philip J. Jaggar. 2003. Ex-situ and in-situ focus in Hausa: syntax, semantics and discourse. In Lecarme, J. (ed.), Research in Afroasiatic Grammar II, 187–213.
Amsterdam: John Benjamins.
Güterbock, Hans G. & Harry A. Ho¤ner (eds.). 1995. The Hittite dictionary of the Oriental
Institute of the University of Chicago. Volume P/2. Chicago: The Oriental Institute of the
University of Chicago.
Haan, Judith. 2002. Speaking of questions: An exploration of Dutch question intonation (LOT
Dissertation Series 52). Utrecht: LOT.
Hackstein, Olav. 2004. Von der Diskurssyntax zur Satzsyntax: Hethitisch kı̄ kuit. In Detlev
Groddek & Sylvester Rößle (eds.), Šarnikzel. Hethitologische Studien zum Gedenken an
Emil Orgetorix Forrer (19.02.1894–10.01.1986) (Dresdner Beiträge zur Hethitologie 10),
345–359. Dresden: Dresden University of Technology.
Hajičová, Eva. 1983. On some aspects of presuppositions in questions. In Ferenc Kiefer
(ed.), Questions and answers, 85–96. Dordrecht: Reidel.
Hengeveld, Kees & J. Lachlan Mackenzie. 2008. Functional discourse grammar: A
typologically-based approach to language structure. Oxford: Oxford University Press.
Ho¤ner, Harry A. 1995. About questions. In Theo P. J. van den Hout & Johan de Roos
(eds.), Studio historiae ardens: Ancient Near Eastern studies presented to Philo H. J. Houwink ten Cate on the occasion of his 65th birthday, 87–104. Leiden: Nederlands HistorischArchaeologisch Instituut te Istanbul.
Ho¤ner, Harry A. & H. Craig Melchert. 2008. A grammar of the Hittite language: Part I:
Reference Grammar. (Languages of the Ancient Near East 1). Winona Lake: Eisenbrauns.
Hout, Theo van den. 2003. Studies in the Hittite phraseological construction I: Its syntactic
and semantic properties. In Gary Beckman, Richard Beal & Gregory McMahon (eds.),
Hittite studies in honor of Harry A. Ho¤ner Jr. on the occasion of his 65 th Birthday, 177–
204. Winona Lake: Eisenbrauns.
Jacobs, Joachim. 1991. Implikaturen und ‘alte Information’ in w-Fragen. In Marga Reis &
Inger Rosengren (eds.), Fragesätze und Fragen: Referate anläßlich der 12. Jahrestagung der
Deutschen Gesellschaft für Sprachwissenschaft, Saarbrücken 1990, 201–221. Tübingen:
Niemeyer.
Jaeger, T. Florian. 2004. Topicality and superiority in Bulgarian wh-questions. In Olga Arnaudova, Wayles Browne, Maria-Luisa Rivero & Danijela Stojanovic (eds.), Formal approaches to Slavic linguistics 12: The Ottawa Meeting 2003, 207–228. Ann Arbor, MI:
Michigan Slavic Publications.
Kim, Alan Hyun-Oak. 1988. Preverbal focus position in type XIII languages. In Michael
Hammond, Edith Moravcsik & Jessica Wirth (eds.), Studies in syntactic typology, 148–
171. Amsterdam: John Benjamins.
É. Kiss, Katalin. 1998a. Discourse-configurationality in the languages of Europe. In Anna
Siewierska (ed.), Constituent order in the languages of Europe, 681–727. Berlin & New
969
É. Kiss, Katalin. 1998b. Identificational focus versus information focus. Language 74. 245–
273.
Lambrecht, Knud. 1994. Information structure and sentence form: Topic, focus and the mental representations of discourse referents. Cambridge: Cambridge University Press.
Luraghi, Silvia. 1990. Old Hittite sentence structure. Routledge: London & New York.
Lyons, Christopher. 1999. Definiteness. Cambridge: Cambridge University Press.
Mascheroni, Lorenza M. 1980. Il modulo interrogativo in eteo I: Note sintattiche. Studi Micenei ed Egeo Anatolici 22. 53–62.
Mascheroni, Lorenza M. 1981. Il modulo interrogativo in eteo II: Impieghi stilistici. Oriens
Antiquus. Studia Biblica et Orientalia 20. 99–105.
Mascheroni, Lorenza M. 1983. Il modulo interrogativo in eteo III: Usi argomentativi. In O.
Carruba, M. Liverani & C. Zaccagnini (eds.), Studi orientalistici in ricordo di Franco Pintore (Studia Mediterranea 4), 123–140. Pavia: GJES.
Rieken, Elisabeth. 2000. Die Partikeln -a, -ia, -ma im Althehtitischen und das Akkadogˆ Zinko (eds.), 125 Jahre Indogermanistik in
ramm Ù. In Michaela Ofitsch & Christian
Graz, 411–419. Graz: Leykam.
Rochemont, Michael S. 1986. Focus in Generative Grammar. Amsterdam: John Benjamins.
Rosengren, Inger. 1991. Zur Fokus-Hintergrund-Gliederung im Deklarativsatz und im wInterrogativsatz. In Marga Reis & Inger Rosengren (eds.), Fragesätze und Fragen: Referate anläßlich der 12. Jahrestagung der Deutschen Gesellschaft für Sprachwissenschaft, Saarbrücken 1990, 175–200. Tübingen: Niemeyer.
Rosengren, Inger. 1997. The thetic/categorical distinction revisited once more. Linguistics
35. 439–479.
Toepel, Ulrike & Kai Alter. 2004. On the independence of information structural processing
from prosody. In Anita Steube (ed.), Information Structure, theoretical and empirical aspects, 227–240. Berlin & New York: Mouton de Gruyter.
Van Valin, Robert D., Jr. & Randy J. LaPolla. 1997. Syntax: Structure, meaning and function. Cambridge: Cambridge University Press.
Wu, Jianxin. 1996. WH-topic, WH-focus and WH-in-situ. University of Maryland Working
Papers in Linguistics 4. 173–192.
H. Craig Melchert
The Aegeo-Anatolian Interface
ISAW-New York University
April 13, 2012
Lycian and Greek Linguistic Contact
A Two-way Street
I. Premises
A. Speakers of Lycian and Carian already in SE Anatolia by 2nd millennium: the YALBURT
inscription of the Hittite king Tuthaliya IV assures that the “Lukka land(s)” of Hittite texts were
centered on classical Lycia, with some possible northward extension (see editions by Poetto 1993
and Hawkins 1995: 68-85); the geography of Western Anatolia as established by Starke (1997)
and Hawkins (1998) leaves no realistic location for Karkiša/Karkiya except in the vicinity of
classical Caria. Contra Bryce (2003: 102) et al., Lycian is not a lineal descendant of 2ndmillennium Luvian. As per Adiego (2007: 345-7), Carian shares crucial grammatical features
with Luvian and Lycian, not Lydian (see further below). Ergo, Lycian and Carian likely open to
Greek influence from 2nd millennium onward, decidedly more so than Lydian in the interior
river valleys of western Anatolia (cf. Yakubovich 2010: 86-96 and 157-60).
B. Need to distinguish “micro” from “macro” contact phenomena (possible interference in
bilingual texts vs. structural effects on grammar of language).
II. Bilingual Texts Need Not Reflect Translation of “Primary” Text in Language A into
Language B, as e.g. generally assumed for the Lycian-Greek of the “Létôon Trilingual”
(Blomquist 1982, Rutherford 2002: 207-8, et al.). Rather parallel renderings in respective
languages by commissioned authors. NB differences in Lycian and Greek of N320:
eõke: trmõmisñ: xssahrapazate: pigesere: katamlah: tideimi: s=eõ=ñne=ñte=pddeõ=hadeõ: trmõmile:
pddeõnehmõmis: ijeru: se=natrbbijeõmi: se(j)=arñna: asaxlazu: erttimeli: me=hñti=tubedeõ: arus:
se(j)=epeweõtlmõmeõi: arñnãi: mõmaiteõ: kumezijeõ: hheõ: xñtawati: xbideõñni: se(j)=arKKazuma:
xñtawati:
‘When Pixodaros, son of Hekatomnos, ruled Lycia as satrap and installed as commissioners for
the Lycians Hieron and Natrbbijemi and as asaxlaza for Xanthos Artemelis, the citizenry and the
perioikoi of Xanthos agreed (that) they built a sacred cult stone for the King of Kaunos and
Arkesima.’
μEpeiå LukÌas Òadr·pgs e•cÈneto PiÒfidaros ºEkatÛmnz ¤Ûs, katÈfltgfle ¿rxontas
LukÌas ºIÈrzna kaiå μApollÛdoton kaiå —·nthou e•pimelgtgån μArtemglin, »doÒe dgú
—anhÌois kaiå toÓs perioÌkois i¨dr˙flaflhai bzmÚn BaflileÓ KaunÌzi kaiå μArkeflimai
‘When Pixodaros, son of Hekatomnos, became satrap of Lycia, he appointed as rulers of Lycia
Hieron and Apollodotos and as governor of Xanthos Artemelis, and it pleased the Xanthians and
the perioikoi to establish an altar for the King of Kaunos and Arkesima.’
s=eõ=ñn=aiteõ: kumazu: mahãna: ebette: eseimiju: qñturahahñ: tideimi: se=de: eseimijaje:
xuwati=ti:
‘And they made priest to these gods Eseimija, son of Qñturaha, and whoever xuwa’s Eseimija.’
kaiå eœlonto iërÈa SimÌan Kondoraflios ¤Ún kaiå ˆs ≈n SimÌai e•cc˙tatos g™i tÚn ƒpanta
xrÛnon
‘And they chose as priest Simias, son of K., and whoever is close to Simias, for all time.’
se=i pijeõteõ: arawã: ehbijeõ: esi=ti: ‘And they gave to him freedom (of) that which is his.’
kaiå »doflan a⁄t˛i ¡tÈleian t˛n “ntzn ‘And they gave to him freedom of (his) property.’
s=ed=eli=ñtãteõ: teteri: se(j)=epeweõtlmõmeõi: hrmõmada: ttaraha: me=xbaiteõ: zã: ese=xesñtedi:
qñtati: se=pigreõi:
‘The city and the perioikoi xxx-ed land parcels of the city. Khesñtedi, the qñtati(?) and Pigreõi
xxx-ed the za.’
kaiå »dzkan g¨ pÛlis ¡crÚn ˆc Keflindglis kaiå Picrgs katgrc·flato
‘And the city gave a field which K. and P. worked.’
seõ=ñte=ñte=kmõmeõ: se(j)=eõti: hheõ: sttati=teli: se=t=ahñtãi xñtawatehi: xbideõñnehi:
se(j)=arKKazumahi:
‘And as much as (is) therein and (is) where the cult stone is set down also (shall be) of the
property of the King of Kaunos and Arkesima.’
kaiå ÷flon prÚs t˛i ¡cr˛i kaiå t„ oÕkg¢mata eÏnai BaflilÈzs KaunÌou kaiå μArkeflima
‘And as much as (is) from the field (is) also to be the property of the King of Kaunos and
Arkesima.’
…
se=i(j)=ehbij(a)=aiteõ: tasa: mere: ebette: teteri: arñnas: se(j)=epeweõtlmõmeõi: arñnãi: me=t=epi
tuweõti: mara: ebeija: eõti: sttali: ppuweõti=meõ: ebehi:
‘And the city of Xanthos and the perioikoi have made their oaths to these laws, (that) they shall
carry out these laws as written on this stele.’
kaiå e•poig¢flanto ÷rkous —·nhioi kaiå oi¨ perÌoikoi ÷fla e•n tg§i fltg¢lgi e•ccÈcraptai
poig¢flein e•ntelg§ toÓs heoÓs to˙tois kaiå t˛i iëreÓ
‘And the Xanthians and the perioikoi have made oaths to carry out as much as is written on the
stele for these gods and the priest.’
…
pigesereje: me=i(j)=eseri=hhati: me=hriqla: asñne: pzziti=ti
‘They shall defer to Pixodaros. It is for the supreme religious authority to do what he xx’s.’
PiÒfitaros dË k˙rios »fltz
Virtually no clauses in the two versions are structured the same. The Greek is largely idiomatic,
while the Lycian shows clear signs of strain to render the desire content: me=hñti=tubedeõ:…
mõmaiteõ: ‘agreed…they built’ (unique asyndeton!); circumlocutions (xuwati=ti, ehbijeõ esi=ti,
eseri=hhati, epi tuweõti; last is elsewhere only of physically erecting a standing object).
References
Adiego Lajara, Ignacio-Javier. 2007. The Carian Language. Leiden/Boston: Brill.
Blomqvist, Jerker 1982. Translation Greek in the trilingual inscription of Xanthos. Opuscula
Atheniensia 14.2.11-20.
Bryce, Trevor. 2003. Chapter Three: History. In H. C. Melchert (ed.), The Luwians, 27-127.
Hawkins, J. David. 1995. The Hieroglyphic Inscription of the Sacred Pool Complex at Hattusa
(SÜDBURG). Wiesbaden: Harrassowitz.
——. 1998. Tarkasnawa King of Mira. ‘Tarkondemos’, Boğazköy sealings and Karabel.
Anatolian Studies 48.1-31.
Poetto, Massimo. 1993. L’iscrizione luvio-geroglifica di Yalburt. Nuove acquisizioni relative alla
geografia dell’Anatolia sud-occidentale. Pavia: Iuculano.
Rutherford, Ian. 2002. Interference or Translationese? Some Patterns in Lycian–Greek
Bilingualism. In J. N. Adams et al. (eds.), Bilingualism in Ancient Society: Language
Contact and the Written Text, 197-219. Oxford: Oxford University Press.
Starke, Frank. 1997. Troia im Kontext des historisch-politischen und sprachlichen Umfeldes
Kleinasiens im 2. Jahrtausend. Studia Troica 7.446-87.
Yakubovich, Ilya. 2010. Sociolinguistics of the Luvian Language. Leiden/Boston: Brill.
H. Craig Melchert
57th R.A.I.
Università “Sapienza” Rome
July 7, 2011
Reciprocity and Commerce in Bronze and Iron Age Anatolia
I. The Problem
A. Cuneiform & Hieroglyphic Luvian /washaiya-/ means ‘consecrated, dedicated’:
1. Of ritual objects and cult personnel:
(1) KBo 7.68 ii 2-12 (text Starke 1985: 361–2, but restorations mine—HCM)
GIŠ
tīrana wašḫa[iya …]KI.MIN GIŠtalān=za K[I.MIN šarra] wašḫa ānnan=ḫa wa[šḫa] §
pa=ti(y)=aš aduna a-ša-x[
] GIŠwaššanza wašḫaiyan=za NINDAḫa-x[
] wašḫaīš
NINDA
tannaša wašḫaiya [--] NINDA-iš wašḫaīš GIŠzappalālla KI.MIN DUGu-x[ ]x-iš
KI.MIN paršul=za KI.MIN gu[r?- ]x-ašu KI.MIN maddu KI.MIN [
-]iš KI.MIN
LÚ
NINDA.DÙ.DÙ-aš KI.MIN [LÚSAG]I-iš KI.MIN LÚŠÀ.TAM-aš KI.MIN
‘The t. are consecrated. [The __ ] likewise. The t. likewise. [Above] are consecrated
things, and also below are conse[crated things.] § He/She s[its down??] to eat. The table
is consecrated. The h-bread is consecrated. The t-breads are consecrated. The bread is
consecrated. The z. likewise…The morsel likewise…The wine likewise…The baker
likewise. The drink-server likewise. The chamberlain likewise.’
(2) BABYLON 2 (text Hawkins 2000: 395–6; interpretation mine—HCM)
á-mu-pa-wa/i-tu (“*419”)wa/i-sa-ha-i-za ku+ra/i-i-sà( )ka-tara/i-hi-ha i-zi-i-ha
‘And for him (Tarhunt) I made (a) consecrated k. (and) k.’
2. Of a tutelary deity ‘patron’ (i.e., dedicated to an individual person):
KBo 34.186:5' […AN]A dLAMMA wašḫai ‘to the patron tutelary deity’ and KBo 12.100
Ro 13 wašḫaī[š dLAMMA] (see Starke 1985: 211&244 and below)
B. Likewise CLuvian wašḫazza- means ‘patron’, referring to a tutelary deity:
(3) KBo 35.107 (+) 108 iii 10 (text and reading with Starke 1985: 238)
URU
Taurišizzaš wašḫazzaš dLAMMA-aš dUTU-tī dātī tarkumī[ta]
‘The patron tutelary deity of Taurisa reported to the father Sun-god.’
Similarly KUB 25.37 i 7 1 UDU ANA dLAMMA wašḫazza ‘one sheep for the patron
tutelary deity’ and KBo 21.54:9,10,21.
The direct Lycian cognate wasaza- (TL 38,7) is a professional title and thus refers to a
kind of priest < ‘sacralized, consecrated’ (Melchert 2004a: 78).
2
C. The CLuvian noun wašḫa- (clearly base of preceding adjectives) also means ‘things
consecrated, sacra’. See in positive sense example (1) above and in negative:
(4) KUB 35.54 ii 27-41 (text after Starke 1985: 67 except for reading NU[MUN]. [I.A])
[
]KÙ.BABBAR GUŠKIN NU[MUN]. [I.A?] ḫūman [GIŠḫaḫ]ran GIŠmuwilan
GIŠ
intaluzi [
]x GIŠtiddutri katt[a] ḫikzi n=ašta anda kiššan memai § zāwi ziyar
NUMUN. I.A-na [p]ūnāta inzagān wašḫa a=(a)ta [BE]L SÍSKUR GIŠḫattarāti
ḫatta[r]itta GIŠtūrāti=pa=(a)ta tūr[ā]tta a=(a)ta imrašša<n> dIŠKUR-u[nt]i pari
tarāwītta § a=(a)ta piyatta imma[r]aššan dIŠKUR-ti a=(a)ta zappatta attu[w]al=za
utar=ša [ḫal]liš=ša a=(a)ta ā[pp]a DINGIR.MEŠ-anza ŠA EN SÍSKUR parran ni[š]
§ awiti
(Hittite) “He offers downward [ ] silver, gold, all the seeds, a rake, a m., a shovel, […],
(and) a t. and interjects as follows (Luvian): ‘Here are laid down all the seeds, the i., the
sacra. The ritual client has hoed them with the hoe, while he has speared them with the
spear. He has handed them over to the Storm-god of the open country. He has given them
to the Storm-god of the open country. He has z-ed them—the evil word and the illness.
Let them not come back before the gods of the ritual client’.”
For the negative sense of wašḫa ‘sacra’ referring to the seeds symbolizing the evils that
are to be relegated to the divine sphere see Benveniste 1969: 2.187ff. on Latin sacer and
Watkins 1975a on Hittite šuppi-.
D. But HLuvian (*419/*420)washa- refers to commercial transactions:
(5) KARKAMIŠ A4a §§1-4 and §11 (text Hawkins 2000: 152, including restorations)
§1 [za-ya]-wa/i [DOMUS]-na [... ...]x-sa-´ [(INFANS)]ní-za-a-sa ka-ma-ni-ya REX-ti
CUM-ni ARHA (CONTRACTUS)DARE-ta
§2 ka-ma-ni-sa-pa-wa/i+ra/i PRAE-ri+i-SARMA-ma-ya-´ FRATER-la-sa-na
|(INFANS)ni-za-´ pa-pi-tà-ti-sà-na-´ INFANS.NEPOS CUM-ni ARHA
(CONTRACTUS)DARE-ta
§3 |wa/i-tu-u 20+2[+?] (“SCALPRUM[”)ma-na-zi ARGENTUM-za DARE-mi-na]
§4 [wa/i-tu-u PANIS.PI]THOS-[li]-za i-zi-ya-mi-na …
§11 (“PANIS.PITHOS”)á-za-li-sa-pa-wa/i DOMINUS-na-ní “*419”-sa-ha-sá-´ DAREmi-na
‘[These hou]ses [X, Y’s] son, sold to Kamani the King, and Kamani sold them to
Parisarma (his) brother’s son, grandson of Papidati. To him 22(?) m[inas of silver are to
be given(?)], and [for him a] a feast is to be made…A meal is to be given to the owner for
the washa.’ (less likely: ‘to the lord of the washa’).
On the terminology for buying and selling see Hawkins & Morpurgo Davies 1982, but
for the verbal forms in -mi-na as gerundives see Melchert 2004b. For restorations and the
provision of a feast for the seller see CEKKE §§6-11 (Hawkins 2000: 145).
(6) ASSUR letter f+g §§27-28 (text after Hawkins 2000: 537)
§27 |á-pi-ha-wa/i-za |(*420)wa/i-sa-ha-sa |REL-za |VIA-wa/i-ni-ta
§28 |wa/i-za |á-pi 4-zi-i |ka-max+ra/i-zi |i-sa-u-ta
‘Furthermore, what did they send us of the washa-? They bought another(?) four k’s for
us.’ (lit. ‘again four’)
3
(7) TÜNP 1 (text after Hawkins 2000: 155)
§1 [... ...]ara/i-FRATER-la-ya CUM-ni sà-ta-ti-wa/i+ra/i-sa-na (“TERRA”)ta-sàREL+ra/i-na CUM-ni (“CONTRACTUS”)i-ya-sá-ta
§2 a-wa/i (“SCALPRUM”)á-su-sa ARHA (“CAPUT+SCALPRUM”)ku-sà-mi-na…
§5 (“PES2”)tara/i-pi+ra/i-pa-wa/i REL-sà
§6 |wa/i-´ 1 “ARGENTUM”-sa 1 (“SCALPRUM”)ma-na-sa |1 (“SCALPRUM”)ma-nasa-ha-na (“*419”)wa/i-sa-ha-sa
‘[X ] bought the land from Aralani(?), son of Santatiwari. The (boundary) stone is to be
effaced…He who shall oppose, one ARGENTUM-unit, one mina, and one manas(a)haunit is the washa.’
In addition to Hawkins 2000: 155–6 see Yakubovich 2010 and Giusfredi 2010: 264–6.
II. Solution
1. No connection between the HLuvian and CLuvian words wašḫa- (thus Giusfredi 2010:
232) very unlikely. Nor is a sense ‘fine, financial punishment’ (Giusfredi loc. cit.)
plausible for examples (5) and (6). These call for ‘purchase, sale’ or the like and
connection with the root of Hittite waš- ‘buy’ and ušne-/ušniya- ‘sell’, as per Hawkins
and Morpurgo Davies (1982: 1009), who also allow for a sense ‘fine’. Giusfredi’s formal
analysis, an animate action noun in -šḫa-, is surely correct (CLuvian wašḫa can easily be
the collective nom.-acc. plural to this animate stem).
2. Terms for ‘buying’ and ‘selling’ renewed in HLuvian: see Hawkins and Morpurgo
Davies 1982 for (CUM-ni arha) piya- ‘sell’ (specialized use of piya- ‘give’) and CUM-ni
ijasa- ‘buy (from)’ (but for source of ija(sa)- and Lycian ije- ‘buy’ < PIE *ai‘give/receive’ see Melchert 1989: 42–5 contra Hawkins and Morpurgo Davies 1982:
101–4). But root *wes- in commercial sense reflected in unattalla- ‘merchant’ (hapax in
the Hittite Laws §5 for ‘merchant’, spelled ú-na-at-tal-la-an). Derivation from Hittite uun-ni/a- ‘to drive’ semantically dubious and phonologically impossible (see correct
remarks of Kloekhorst 2008: 917). Rather loanword from Luvian with regular outcome of
*usna- < *us-no- or syncopated *we/os-no-. For loss of *s before *n in Luvian compare
Luvian /ta:na-/ ‘sanctified, inviolable’ < *dh(e)h1s-no- (Melchert 1997: 49). Therefore
commercial sense of HLuvian /washa-/ (<*was-sha-) perfectly in order.
3. PIE *wes- clearly referred to both buying and selling: see especially Benveniste 1969:
1.125–8 and Watkins 1975b: 100–7, the latter elucidating the prehistory of Latin uēnum
dare ‘sell’ and uēnum īre ‘be sold’ and cognate phrases in Greek and Sanskrit. This
duality clearly continued in Hittite waš- ‘buy’ vs. ušne-/ušniya- ‘offer for sale’. Contra
Szemerényi (1979: 122) Hittite and Luvian reflexes show that root *wes- was not
restricted to sense ‘to buy’ already in PIE (also belied by Iranian reflexes he cites 1979:
120–1). See further Mallory and Adams 1997: 185–6 and 224–5 with refs.
4. As per Benveniste (1951), in PIE society giving and receiving regarded as part of
single process of reciprocal exchange. Notion permeated PIE society, including
relationship of humans and gods (principle of do ut des). No hesitation on part of Hittites
to invoke mutual dependence of humans and gods (see Singer 2002: 40 on prayer of
Arnuwanda and Ašmunikal and ibid. 48 on plague prayers of Mursili).
4
5. Therefore, propose that PIE *wes- meant ‘pledge in exchange’, referring to both parts
of any reciprocal exchange: (1) of commodities between people, (2) of people and things
dedicated to gods and of a deity dedicated to a person, (3) of a fine or penalty given in
exchange for release from debt. All three senses persisted into pre-Luvian, with addition
of new mode of exchange of commodity for recognized medium of exchange (precious
metal). Luvian adjective /washaiya-/ specialized in religious sphere, while base noun
/washa-/ apparently preserved in same sense in CLuvian, but evolved to commercial
sense in HLuvian (recall prehistoric presence of such sense for *wes- in unattalla-).
6. Sense of ‘pledged to’ used of persons likely also reflected in female personal names
with second members -wašḫa/i- and -wašti- (see Zehnder 2010: 97–8). Clearest in latter
with inflected dative first members (thus contra Zehnder): fTāti-wašti ‘a pledge/gift to the
father’, HLuvian /Tarhunti-wasti/ ‘a pledge/gift to Tarhunt’ (MARAŞ 2,§1), fḪani-wašti
‘a pledge/gift to Hani’. Simple fWašti either simply ‘gift’ or a Kurzname. Noun stem
wašti- < action noun *wés-ti- ‘pledging’ > ‘thing pledged’. Likewise then also fĀššuiwašḫa- ‘a pledge to the a-stone’ (like Hittite ḫuwaši-, Luvian /a:ssu-/ likely meant ‘cult
stone’ as well as ‘boundary stone’), fLalanti-wašḫa- ‘a pledge to Lalanta (city-name)’,
f
Malia-wašḫi- ‘a pledge to Maliya’ (rather than ‘gift from M.’), etc.
Restriction of wašḫa- and *wašti- to female names may not be coincidental. Per new
suggestion of Pinault (2011), PIE *wes- also used for mutual exchange of marriage:
Greek ὄαρ ‘wife’ (neuter!) < *wós- , *wés-n- (source of *wé/ós-n-o- above) against
semantically implausible derivation from *swós- by Janda (1999). The wife would be
the thing pledged in exchange (HCM). For IE marriage as exchange see Sanskrit passages
with ména-*‘exchange’ and ‘concubine’ in Hoffmann 1960 cited by Pinault.
7. Also traces of older non-commercial sense ‘pledge, match’ (on a balance scale) for
Hittite ušneške-:
(8) KUB 21.27 iv 38-42 (Prayer of Puduhepa, NH)
Ḫattušili[šš]=a ARAD=KA ANA ZI DINGIR-LIM šer dariyat nu=za apē[l] SAG.DU-an
m
ZI=ŠU=ya uššanišket kuitman [ŠA DINGI]R-LIM EN=YA URUNeriqqan āššiyant[an
UR]U-an EGIR-pa wetet ‘Also Hattusili, your servant, has exerted himself for the god’s
will and pledged his body and soul, until he rebuilt Nerik, beloved city of the god, my
lord.’
Singer (2002: 105) renders ‘engaged’. Also possible is ‘put in the balance’ in the sense of
‘put at risk’ (Friedrich 1952: iv 40 ‘aufs Spiel setzen’).
(9) KUB 21.19+KBo 52.17 iii 19'-21' (Prayer of Hattusili and Puduhepa, NH)
nu KUR URUNerik EGIR-pa ašešanunun URUN[eriqqan URU-an] EGIR-pa wedaḫḫun
nu=za ANA KUR URUNer[ik šer?] SAG.DU=YA ZI=YA uššaniškenun ‘I resettled the land
of Nerik. I rebuilt the city Nerik. For(?) the land of Nerik I pledged my body and soul.’
(10) KBo 21.22 Ro 14-21 (blessings for the king, OH/MS)
nu=za kuit labarnaš LUGAL-uš ištanzanaš=šaš [x-x-x]-x-aš ilāliškezzi n=at=ši anda
arān ēštu [nu=za kui]t MUNUStawanannaš MUNUS.LUGAL ŠA ZI=ŠU ŠA UZU=ŠU
ilališkezi [n=a]t=ši anda aran ēštu § kāša GIŠRÍN karpiyemi nu labarnaš taluqauš
5
MU.ḪI.A-uš ušneškemi kāša GIŠ.RÍN karpiyemi n=ašta
MU.ḪI.A-uš ušneškemi AWAT GIŠ.RÍN araḫza QATI
MUNUS
tawanannaš taluqauš
‘Whatever the labarna the king desires with his soul and flesh(?), let it come to him.
Whatever the tawananna the queen desires with her soul and flesh(?), let it come to her. I
hereby lift up the scales and weigh out/match long years for the labarna. I hereby lift up
the scales and weigh out/match long years for the tawananna. The word of the scales
outside is finished.’
Contra Neu (1980: 81–82) and Watkins (1987: 293), ušneškemi is very unlikely to mean
“put up for sale”. As per Otten (1958: 132), Carruba (1964: 415), Kellerman (1978: 201),
Archi (1979: 47) and Puhvel (1983: 52), rather the technical term for ‘weigh’ on scales,
since such weighing requires putting something in one pan as a pledge/payment against
what is put in the other (as per Otten 1958: 132 “einsetzen”). However, contra Puhvel it
does not mean here that the king’s and queen’s lifespan “is symbolically hung in the
balance,” nor contra Kellerman (1978: 204) that “if the ‘long years’ weigh much, the king
will live a long time.” Rather with Otten (1958: 132) and Archi (1979: 44) the ritual
action is to assure long years for the royal couple, as indicated by the preceding context,
the similar sentiment of KUB 29.1 ii 7-10 cited by Archi (GIŠḫūšuš šūwaduš ḫarkanzi nu
LUGAL-waš MU.KAM.ḪI.A-uš malkiyanzi wittann=a kudreš=šmit kappūwuwar=šamet
ŪL duqqāri ‘They (the fate goddesses) are holding full spindles. They spin the years of
the king, and of the years no reckoning (nor) counting is to be seen’), and other evidence
that long life was among the most wished for things for the Hittite king and queen. Thus
not plausible that ‘long years’ would merely be offered to the king and queen as
interested buyers. Rather, symbols for long years are placed in the one pan of the scales,
and enough weight (surely metal) is put in the other to balance them and thus assure the
desired long life of the royal couple (for such symbols of long years compare the hunting
bag/cornupcopia of Telipinu, KUB 17.10 iv 31). Compare with Otten (1958: 131–2)
Hurro-Hittite ritual KBo 17.95 iii 5-11, where the king puts a lead ingot onto a set of
scales that is then held out to the Sun-god. One would expect ušne- to be used either of
object A pledged to match B, or of object B that is matched by object A.
8. The reference to some kind of meal or feast being furnished to the seller by the buyer
appears to be an innovation. For a ceremonial meal furnished by the buyer to conclude a
land sale in second-millennium Mesopotamia see Gelb, Steinkeller and Whiting 1991:
243-4. For a similar practice in Late Bronze Age Emar see Beckman 1997: 99 and
Scurlock 1993.
References
Archi, Alfonso. 1979. Auguri per il Labarna. In O. Carruba (ed.), Studia Mediterranea Piero Meriggi
dicata, 27–51. Pavia: Aurora.
Beckman, Gary. 1997. Real Property Sales at Emar. In G. D. Young, M. W. Chavalas, and R.E. Averbeck
(eds.), Crossing Boundaries and Linking Horizons: Studies in Honor of Michael C. Astour on His
80th Birthday, 95–120. Bethesda: CDL Press.
Benveniste, Émile. 1951. Don et échange dans le vocabulaire indo-européen. L’Année sociologique 2.7–20.
[Reprinted in Problèmes de linguistique générale (1966), 315–26. Paris: Gallimard.]
——. 1969. Le vocabulaire des institutions indo-européennes. Paris: Éditions de Minuit.
6
Carruba, Onofrio. 1964. Hethitisch -(a)šta, -(a)pa und die anderen “Ortsbezugpartikeln.” Orientalia NS
33.405–36.
Friedrich, Johannes. 1952. Hethitisches Wörterbuch. Heidelberg: Winter.
Gelb, Ignace J., Piotr Steinkeller, and Robert M. Whiting Jr. 1991. Earliest Land Tenure Systems in the
Near East: Ancient Kudurrus (Oriental Institute Publications 104). Chicago: The Oriental Institute of
the University of Chicago.
Giusfredi, Federico. 2010. Sources for a Socio-Economic History of the Neo-Hittite States (THeth 28).
Heidelberg: Winter.
Hawkins, J. David. 2000. Corpus of Hieroglyphic Luwian Inscriptions. Volume 1. Inscriptions of the Iron
Age. Berlin/New York: de Gruyter.
Hawkins, J. David, and Anna Morpurgo Davies. 1982. Buying and selling in Hieroglyphic Luwian. In
Johann Tischler (ed.), Serta Indogermanica. Festschrift für Günter Neumann zum 60. Geburtstag
(IBS ), 91–105. Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck.
Hoffmann, Karl. 1960. Der vedische Typus menāmenam. Zeitschrift für Vergleichende Sprachforschung
76.242–8.
Janda, Michael. 1999. Zur Herkunft von homerisch ὄαρ. In Sigrid Deger-Jalkotzy et al. (eds.), Floreant
Studia Mycenaea. Akten des X. Internationalen Colloquiums in Salzburg vom 1.-5. Mai 1995, 315324. Wien: Österreichische Akademie der Wissenschaften.
Kellerman, Galina. 1978. The King and the Sun-god in the Old Hittite Period. Tel Aviv 5.199–208.
Kloekhorst, Alwin. 2008. Etymological Dictionary of the Hittite Inherited Lexicon. Leiden/Boston: Brill.
Mallory, James P, and Douglas Q. Adams. 1997. Encyclopeda of Indo-European Culture. London/Chicago:
Fitzroy Dearborn.
Melchert, H. Craig. 1989. New Luvo-Lycian Isoglosses. Historische Sprachforschung 102.23–45.
——. 1997. Luvian /tāna-/ ‘sanctified, inviolable’. Historische Sprachforschung 110.47–51.
——. 2004a. A Dictionary of the Lycian Language. Ann Arbor/New York: Beech Stave.
——. 2004b. Hieroglyphic Luvian Verbs in -min(a). In Adam Hyllested et al. (eds.), Per Aspera ad
Asteriscos. Studia Indogermanica in honorem Jens Elmegård Rasmussen sexagenarii Idibus Martiis
anno MMIV (IBS 60)355–62. Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck.
Neu, Erich. 1980. Die hethitischen Verben des Kaufens und Verkaufens. Welt des Orients 11.76–89.
Otten, Heinrich. 1958. Hethitische Totenrituale. Berlin: Akademie-Verlag.
Pinault, Georges. 2011. Indogermanische Verwandtschaftswörter und das Bild der Gesellschaft. Handout
from Arbeitstagung der Indogermanische Gesellschaft, „Menschenbild“, Halle, 31. März 2011.
Puhvel, Jaan. 1983. Homeric questions and Hittite answers. American Journal of Philology 104.221–7.
Scurlock, Joann. 1993. Once more ku-bu-ru. Nouvelles Assyriologique Brèves et Utilitaires 1993/21.
Singer, Itamar. 2002. Hittite Prayers (Writings from the Ancient World 11). Atlanta: Society for Biblical
Literature.
Starke, Frank. 1985. Die keilschrift-luwischen Texte in Umschrift (StBoT 30). Wiesbaden: Harrassowitz.
Szemerényi, Oswald. 1979. Germanica I (1-5). Zeitschrift für Vergleichende Sprachforschung 93.103–25.
Watkins, Calvert. 1975a. La designation indo-européennes du ‘tabou’. In J. C. Milner et al. (eds.), Langues,
discours, société. Pour Émile Benveniste, 208–14. Paris: Éditions du Seuil.
——. 1975b. Some Indo-European verb phrases and their transformations. Münchener Studien zur
Sprachwissenschaft 33.89–109.
——. 1987. Linguistic and archeological light on some Homeric formulas. In S. Skomal and E. Polomé
(eds.), Proto-Indo-European: the Archaeology of a Linguistic Problem. Studies in Honor of Marija
Gimbutas, 286–98. Washington DC: Institute for the Study of Man.
Yakubovich, Ilya. 2010. The West Semitic God El in Anatolian Hieroglyphic Transmission. In Yoram
Cohen et al. (eds.), Pax Hethitica. Studies on the Hittites and their Neighbours in Honour of Itamar
Singer (StBoT 51), 385–98. Wiesbaden: Harrassowitz.
Zehnder, Thomas. 2010. Die hethitischen Frauennamen. Katalog und Interpretation (Dresdner Beiträge zur
Hethitologie 29). Wiesbaden: Harrassowitz.
H. Craig Melchert
Naming in Anatolia
Oxford, Wadham College
July 11, 2011
‘Naming Practices in 2nd and 1st Millennium Western Anatolia’
I. Limitations/Caveats
1. All figures in section III are mere approximations to give order of magnitude of
available data:
a. Reading of word as personal name vs. appellative not always assured (nor even
segmentation in case of scriptio continua).
b. Constant publication of new texts adds new examples not yet in handbooks.
2. Assignment of indirect attestations to specific Anatolian Indo-European languages
varies in degree of assurance: hence designations “in Lycia” or “in Isauria”, etc. rather
than “Lycian”.
a. Higher degree of certainty for Lydian and Carian, much lower for the others: due to
relative dialectal isolation of Lydian and unique phonological developments in Carian.
b. But some Luvian or Carian names in Lydian context (see IV.2 (7) & IV.3 (13) below).
c. Other evidence for spread of individual names: NB case of Greek-influenced
Μουσατης in Lydia, Carian Músat-, and Pisidian Μουσητα- (see 5.1.2 (22) below).
3. Some differences cited in typology and percentages of borrowed names reflect special
circumstances of our corpora:
a. Overwhelming majority of Carian texts are from Egypt, not Caria.
b. Hieroglyphic Luvian texts include Southern Anatolia and Syria.
4. With single exception (“Satznamen”), presentation is confined to (synchronic)
typology with no attempt to determine origins of various types.
5. Do assume with Goetze 1954, Houwink ten Cate 1965, Adiego 2007 et al. strong
continuity from 2nd millennium to 1st.
II. Morphological Preliminaries
1. Both forms of Luvian, Lycian, and to a lesser extent Carian and Lydian all show the
phenomenon dubbed “i-mutation” (first established by Starke 1990: 54-85; for the correct
diachronic account see Rieken 2005): animate stems insert an -i- between stem and
ending just in the nominative and accusative cases. A stem-final -a- is deleted before the
inserted -i-. This addition does not alter the underlying stem. Effect for personal names is
widespread existence of doublets: thus Cuneiform Luvian mZida-/Ziti- ‘man’ or
piyam(m)a- ‘given’ (as 1st cpd. member) vs. Lycian -pijemi- ‘given’ (2nd cpd. member).
2. At least Lycian also attests the PIE “individualizing” suffix *-on- often used in
personal names (type of Latin Catō ‘sharpy’, Greek Στράβων ‘the squinty-eyed one’),
with or without “i-mutation”: thus Xudalijẽ- ‘the nimble one’ beside Xudali- or Mutlẽi‘the mighty one’ (cf. Luvian Muwattalli- ‘mighty’).
2
III. Data Base (omitting indirect Hellenistic evidence)
1. Luvian (NB: the three sources include significant overlap, so total is not their sum!)
a. H(ieroglyphic) Luvian texts ±210: see Hawkins 2000.
b. In 2nd millennium cuneiform texts (Hittite Empire period) ±280: see Laroche
1966&1981 and Trémouille 2006.
c. On personal seals: ±200: see Güterbock 1940&1942, Herbordt 2005 (esp. 359-70).
2. Lycian ±340: see Melchert 2004: 91-111, Cau 2005a & 2005b (non uidi).
3. Carian ±200 (including ±5 < Greek, ±25 < Egyptian): see Adiego 2007 (but I ignore
his idiosyncratic transliteration of signs involving /i/ and /u/!).
4. Lydian ±50 (including <5 < Iranian): see Gusmani 1964 and 1988.
5. Pisidian ±30: see Brixhe 1988.
6. Sidetic ±25 (including ±9 < Greek): see Nollé 2001: 630-46, but for readings Pérez
Orozco 2003, 2005 and 2008.
IV. Formal Types of Personal Names
1. “Elemental”/Unanalyzable (including but not restricted to “Lallwörter”)
(1) /Ta-/: Da-a-a (cuneiform, karûm period; Laroche 1966: 169), mTā- (Hittite Empire
[henceforth HE]; Laroche 1981: no. 119a), Tá-a- (hiero[glyphic] seals HE; Hawkins apud
Herbordt 2005: 272), Ta- (HLuvian, KULULU lead strip 1, §3.9; Hawkins 2000: 507),
Ta- (Lycian, TL 65,72; 74c,2; Melchert 2004: 104), Τα- (Pisidian; Brixhe 1988: 149).
(2) /Titti(ya)-/: mTittiya- (karûm period & HE; Laroche 1966: 186), Titi- (Lydian, 30,1;
Gusmani 1964: 213), Τιττι- (fem.!) in Isauria, Pisidia-Lycia; Zgusta 1964: 516.
2. Appellatives, all Luvian in origin (many also in compounds; see IV.5-6 below)
(3) /mo:wa-/ ‘might’ (including but not restricted to sexual potency): mMu-u-wa- (HE;
Laroche 1966: 122), Mu-wa/i-á- (hiero. seals HE; Hawkins apud Herbordt 2005: 265),
Mu-wa/i- (HLuvian, KARKAMIŠ A4a, §9; Hawkins 2000: 152), = Μοας, Μουας, Μως
in Pisidia, Pamphylia, Cilicia, Isauria; Zgusta 1964: 320, 334, 343. NB: in some instances
could represent “Kurzname” < possessive compounds in IV.6.1.1 below. Plus derivatives:
(a) /mo:wata/i-/ ‘might, potency’ (via denominative verb /mo:wa:-/ ‘be mighty, potent’):
Muwatti- (HE; Laroche 1966: 124), Mu(wa)-ti- (hiero. seals HE; Hawkins apud Herbordt
2005: 265-6), mute- (masc. Carian; Adiego 2007: 386), = Μο(υ)τας in Cilicia, Phrygia;
Zgusta 1964: 334, 339.
f
(b) /mo:watalli-/ ‘mighty’ (adj. < preceding): mMuwattalli- (Hittite royal name et al.;
Laroche 1966: 123), Mu-wa/i-ta-li- (HLuvian, dynasty of Maraş; see Hawkins 2000:
251), Mutlẽi- (Lycian, ‘the mighty one’ with *-on+“i-mutation”, TL 150,2 & M210;
Melchert 2004: 100), = Μότυλος in Caria; Zgusta 1964: 334 (perhaps Μοταλις in
Phrygia, per Houwink ten Cate 1965: 103, but see Zgusta 1964: 334300).
(4) /pi:ha-/ ‘luminescence’ < *bhēh2-o-, but by association with Storm-god (of lightning)
> ‘might, power’: Pi-ha-á- (hiero. seal HE; Hawkins apud Herbordt 2005: 267, refuting
other alleged exx. in Laroche 1966: 139). Plus derivatives:
3
(a) /pi:hamma/i-/ ‘resplendent; mighty’ (denom. adj. < preceding): Pi-ha-mi (hiero. seal
HE; Hawkins apud Herbordt 2005: 267; also HLuvian, ASSUR letter e, §1; Hawkins
2000: 535), Pixma- (Lycian, TL 116,1; Melchert, 2004: 101), = Πιγομος in Lycia;
Zgusta 1964: 427.
(b) */pi:hra-/ ‘resplendent; mighty’ < virtual *bhēh2-ro- (contra Houwink ten Cate 1965:
157, Luvian not directly attested): Pixre-, Pigrẽi- (Lycian, TL 55,1 & N320,15; Melchert
2004: 101), Pixre-, Pixra- (Carian; Adiego 2007: 397—one of two may have *-on- suffix
like Pigrẽi-).
(c) */pi:hramma/i-/ ‘resplendent, mighty’ (blend of preceding two or denom. adj. <
intervening noun < */pi:hra-/): Pik(a)rm- (Carian; Adiego 397), = Πιγραμις, Πιγραμος in
Lycia; Zgusta 1964: 427-8.
(d) /pi:hassa/i-/ ‘resplendent; mighty’ (also substantivized to ‘lightning’): Pigśe- (Sidetic,
S9,5; Pérez Orozco 2008: 127), = Π(ε)ιγασις in Lycia; Zgusta 1964: 427, and Πιγασσως
in Caria; Blümel 1992: 23 & 1994: 74. See also -p/biks- in Carian compound names.
(5) /hant-/ ‘front’ < *h2ent- with derivatives ‘foremost’ whence ‘first’ and ‘ruling/ruler’
(see Houwink ten Cate 1965: 149-50):
(a) /hantil(i)-/ ‘first’: mḪantili- (early Hittite king et al.; Laroche 1966: 58), Xñtla(Lycian, N324,14; Melchert 2004: 109—with suffix *-ileh2).
(b) */hantawa-/ ‘foremost, ruling’ < virtual *h2(e)ntowo- (assured base of Lycian denom.
verb xñtawa- ‘to rule’ etc.): probably attested in Γδαβα- (Pisidian; Brixhe 1988: 14950—with regular voicing of t > d after nasal, loss of nasal, shift of w > v/b, and
assimilation of *Kd- > Gd- ).
(c) /hantawat(i)-/ ‘ruler, ruling’: probably in Γδβετι- (Pisidian; Brixhe 1988: 150—exact
word equation with Lycian xñtawat(i)- ‘ruler, king’).
(6) /immra-/ ‘open country’ < virtual *dhghemro-: probably directly in Ιμβρα/ης in Lycia,
Zgusta, 1964: 199. Or “Kurzname” from compounds? Also in various derivatives:
(a) /immrassa/i-/ ‘of the open country’: Iβ(a)rsi-/βrsi- (Carian, Adiego 2007: 335, contra
Schürr 1991-93: 171 & 2001: 104-5), = Ιμβρασ(σ)ις/Ιμβαρσις in Caria; Zgusta 1964:
198-9 & Blümel 1994: passim. But Lycian para- does not belong here (see Schürr
1991-93: 164 et alibi, contra Houwink ten Cate 1965: 103).
(b) /immralla/i-/ ‘of the open country’: Ιμβαρηλδος in Caria (rendering with -λδ- assures
the name is Carian; Adiego 2007: 335).
(7) /huh(h)a-/ ‘grandfather, forefather’ < *h2éuh2o- (= Latin auus): Quq- (Carian; Adiego
2007: 334), = Γυγος in Caria. As per Adiego, either Carian or Luvian is also source of
“Lydian” Γύγης, since *h2 is not preserved in native Lydian words. Also frequent second
member of compound names.
4
(8) /ariyamma/i-/ ‘lifted, exalted’, participle to verb ariya- ‘to lift’: *Arimmi-, i.e.
PUGNUS-ri+i-mi- (HLuvian, KULULU lead strip 1, §4.15&25, §9.58; Hawkins 2000:
506&508—for verb see Hawkins 2000: 380), probably also in Aríom- (Carian; Adiego
2007: 354, but analysis mine—HCM).
(9) /hwiyamma/i-/ ‘running, runner, attendant’: possibly in Kbíom- (Carian; Adiego 2007:
371, but analysis mine—HCM), = Κεβιωμος in Caria. See also Carian compound name
Šar-kbiom- (Adiego 2007: 416), probably ‘running above, super-attending’, similar but
not identical to Lycian hri-xuwama- ‘super-attending’ or ‘super-attendance’ (see
Melchert 2004: 86 and in extenso García Ramón 2011 and forthcoming).
3. Theophores (here only direct or with appurtenance suffix; determinative and
possessive compounds treated separately below in IV.5-6, Satznamen in IV.7-8)
(10) /Iya-/ (i.e., the Luvo-Hittite form of Mesopotamian Ea-): fIya- (HE, Kuşaklı, KuT 49
Ro 3,12; Trémouille 2006), I-ia- (HLuvian, BEIRUT §1, prob. masc.; Hawkins 2000:
559), also in Ια- (fem.) in Sidé (text 102; Nollé 2001: 392), Bithynia, Phrygia, Pisidia,
Isauria, Cilicia; Zgusta 1964: 188-9.
(11) /Kruntiya-/ ~ /Runtiya-/, the Luvian Stag-god and tutelary deity (see Houwink ten
Cate 1965: 129-31): m.dLAMMA-a- (HE, King of Tarhuntassa, Laroche 1966: 223—on
the reading /Kruntiya-/ for the first see Hawkins apud Herbordt 2005: 290 on spellings
CERVUS1-3-ti- on hiero. seals HE), *Runtiya-, i.e. CERVUS and CERVUS2-ti-ia(HLuvian, GÜRÜN §1b and BABYLON 3, Hawkins 2000: 297&397), in Ρωνδας in
Cilicia and Ρω(ι)ζις in Pisidia (latter < assibiliated *Runza-); Zgusta 1964: 446-7. NB also
double theophoric name Αρμαρωνζας (Moon- and Stag-god) in Cilicia; Zgusta 1964: 93.
(12) /Sanda-/, Anatolian war-god (see Houwink ten Cate 1965: 136-7): mŠanta (HE;
Laroche 1966: 156), in Σανδα(ς), Σανδης, Σανδος in Lycaonia, Pisidia, Caria, Cilicia;
Zgusta 1964: 454. NB also double theophoric names ISà-ta-ti-wa/i+ra/i- /Sanda-tiwara-/
(Sanda and Sun-god) and ISà-ta-SARMA-max- /Sanda-sarma-/ (Sanda and Šarruma)
(HLuvian, TÜNP 1, §1 and NIMRUD; Hawkins 2000: 155&570).
(13) /Tiwad(a)-/, the Luvian Sun-god: Tivda- (“Lydian”; Gusmani 1964: 213), also in
*Tiwarama/i- (HLuvian, CEKKE §17i; Hawkins 2000: 146—rhotacized variant of
appurtenance adjective /Tiwad-ama/i-/ ‘of the Sun-god’). NB also in double theophoric
name /Sanda-tiwara-/ in (12). But no traces whatsoever in late Cilicia, Pisidia, Lycia,
Caria, not even in compounds.
(14) /Yarra/i-/, southern Anatolian war-god: mYarri- (HE; Laroche 1966: 77), II-a+ra/iri+i-i- (HLuvian, KARKAMIŠ A6, A7, A15b, regent—read /Yarri-/ contra Hawkins
2000: 124 and passim, with otherwise unattested †/Yariri-/). NB also as personal name
likely appurtenance adjective /Yarrissa-/, i.e. i-ia+ra/i-sà-, ‘of Yarri’ in KULULU lead
strip 1, §5.31. No clear traces in late first millennium.
(15) Artemis (only in first millennium): Artimu- (Lydian; Gusmani 1964: 64-5 and
1988: 192), probably also Artima- (Lydian, thus with Zgusta 1964: 101 contra Gusmani
1988: 192), Artmi- and Rtim- (Carian; Adiego 2007: 356-7 with refs. & 410), also in
Αρτ(ε)ιμας, Αρτειμος, Αρτ(ε)ιμης in Lycia, Pisidia, Caria, Lydia; Zgusta 1964: 99-101.
NB also derived Erttimeli- = Αρτεμηλιν (Lycian, N320,5; Melchert 2004: 94). All
examples are masculine.
5
4. Ethnica (in contrast with Hittite names in -uman- from karûm period a very rare type in
western and southern Anatolia)
(16) Hittite Empire Luvian ethnica are limited to mAlauwanni- and mUrawanni- (Laroche
1966: 27&198) and mLuggawanni- (Trémouille 2006).
(17) HLuvian only assured ILa-ka-wa/i-ni- /Lakawann(i)-/ ‘the Lakean’ in ADIYAMAN
2, §3 (Hawkins 2000: 351-2) and possible I*447-nu-wa/i-ia- /Ninuwiya-/ ‘one of
Nineveh’ in KARKAMIŠ 11b+c, §2 (Hawkins 2000: 103-4).
(18) Lycian only assured Pttaraze/i- ‘one of Pttara’ in TL 113,1 and probably Xa/erigaand Xerẽi-, which cannot reflect an appellative ‘eagle’, but more likely (NB variant
Xariga- M129) represent ethnica ‘(the) Carian’ (see Melchert 2004: 108).
5. Determinative Compounds (including with divine names as first or second members)
5.1 Noun+Noun
5.1.1 First member in dative function: N2 for N1 (function assured by some examples with
inflected first member, underscored below):
(19) Exclusively female names in -wašḫa/i- and -wašti- ‘pledge/gift to X’ (X = deity,
toponym, or person): fAla-wašḫi- ‘a pledge to Ala (city-name)’, fĀššui-wašḫa- ‘a pledge to
the a-stone’, fTāti-wašti ‘a pledge/gift to the father’ (HE; Laroche 1966: 27,47,182),
/Tarhunti-wasti/ ‘a pledge/gift to Tarhunt’ (HLuvian, MARAŞ 2, §1; Hawkins 2000:
273), perhaps also in Κρεσαουεστις in Isauria-Cilicia and Παπουστις ‘pledge to the
father’(?) in Phrygia-Lycia; Zgusta 1964: 257&416. See for further 2nd mil. examples
Zehnder 2010: 97-8 but with false analysis; for sense ‘pledge, gift’ see Melchert 2011.
(20) Name with 2nd member /-warra/i-/ ‘help, aid to X’ (X = deity, person, entity):
/Parni-warra/i-/, i.e. DOMUS-ni-wa/i+ra/i- ‘help to house(hold), /Parna-warra/i-/, i.e.
(“*69”)pa+ra/i-na-wa/i-ri+i- ‘idem’ (HLuvian, ASSUR letter b, §1 and f+g, §33;
KARKAMIŠ 17a, §6; Hawkins 2000: 534,537,192), /Tarhu-warra/i-/, i.e. TONITRUShu-wa/i+ra/i-i- ‘help to Tarhunt-’ (EĞREK §1; Hawkins 2000: 492—NB abbreviated 1st
member!), FRATER-la-wa/i+ra/i- ‘help to the brother and /Huhha-warra/i-/, i.e. AVUSha-wa/i+ra/i- ‘help to the grandfather’ (CEKKE §17d and §17j; Hawkins 2000: 145-6),
/Kumma-warra/i-/, i.e. ku-ma-wa/i+ra/i- ‘help to the sacred’ (KARKAMIŠ A4a, §10;
Hawkins 2000: 152), probably also Ubat-woro- ‘help to the demesne’ (Sidetic S5,1-2;
Nollé 2001: 639—reading Pérez Orozco 2008: 128-9, but analysis mine—HCM).
5.1.2 First member in genitival function: N2 belongs to N1 or N2 has quality of N1:
(21) Second member is term of relationship, first is deity or other: /Arma-nana/i-/
‘brother of the Moon(-god), mMI.ŠEŠ (HE; Laroche 1966: 39), LUNA.FRATER2 (hiero.
seals HE; Hawkins apud Herbordt 2005: 249-50), also HLuvian, KARAHÖYÜK §1;
Hawkins 2000: 289, Erme-nẽne/i- (Lycian, TL 121; Melchert 2004: 93), = Ερμενηννις
in Lycia; Zgusta 1964: 172 (see Houwink ten Cate 1965: 133 for likely further reflexes),
*Tarhu(nta)-nana/i-, i.e. TONITRUS-FRATER2- ‘brother of Tarhunt-’ (hiero. seals HE;
Hawkins apud Herbordt 2005: 272-3); mḪila-nani- ‘brother of the courtyard’ (HE, Maşat
HKM 113 Vo 16; Trémouille 2006), Dquq- (Carian) ‘grandfather of Ida’, also in Ιδαγυγος
in Caria (thus with Adiego 2007: 362—further equation with Lycian Ddxug[a]- possible,
but not assured).
6
NB special subtype: mTati-ŠEŠ- ‘brother of the (same) father’ (HE; Laroche 1966: 182),
= Τεδε/ι-νηνις in Lycia and Cilicia (Zgusta 1964: 508), contra Houwink ten Cate 1965:
144-5—see on type Watkins 1995: 360-1 and cf. Enẽhi-nere/i under 5.3 (28) below.
(22) Second member is /tsida/i-/ ‘man’ (usu. written LÚ in cuneiform and VIR in
hieroglyphs), first is deity, toponym, or other (not always identifiable)—rampantly
productive in 2nd millennium (40+ examples): deities include mArma-LÚ-i-, mŠanta-ziti-,
m
Tarḫunda-ziti-, *Tiwada-ziti- (m.dUTU.LÚ-i-), mYarra-LÚ-i-; city names mAnkuwa-LÚ-,
m
Ḫalpa-ziti- (also TONITRUS.HALPA-VIR.zi/a-), mNinuwa-LÚ-; various appellatives
m
Ḫarwa-LÚ-ti- ‘man of the road’, Imm(a)ra-ziti- ‘man of the open country’ (see below!),
m
Kuwalana-LÚ- (also EXERCITUS.VIR.zi/a-) ‘man of the army’, mMuwa-LÚ-i- (also
Mu(wa)-VIR(.zi/a)-) ‘man of might/potency’ (see below!), mPiḫa-ziti- (also Pi-haVIR.zi/a-) ‘man of resplendence; might’, mŪḫḫa-LÚ-i- ‘man of the grandfather’ (probably
thus with Yakubovich 2010: 91), mUba-LÚ-i- ‘man of the dedication’, mWalwa-LÚ-i(also LEO-VIR.zi/a-) ‘lion-man’, mḪattagga-LÚ- probably ‘bear-man’ (all Hittite
Empire; Laroche 1966: passim and Hawkins apud Herbordt 2005: 254,261,266,268,278);
Ipre-side/a- (Lycian; Melchert 2004:96—as per Carruba, = Imm(a)ra-ziti-, also =
Ιμβρασιδης in Lycia, as per Schürr 1991-93: 165, but with false historical analysis);
/Mo:wa-zida/i-/ ‘man of might’ attested in modified form in Carian Músat- (Adiego
2007: 386), Pisidian Μο(υ)σητα- (Brixhe 1988: 143), = Μουσατης in Lydia and
Μωσητας in Cilicia (Zgusta 1964: 337&343). Attested forms cannot show regular reflex
(NB real Lycian -sida- in Ipre-sida-). Neither voiceless stop nor low vowel explicable
(contra Brixhe 1988: 139, eta cannot represent [i]). Rather with loss of sense that name is
compound remodeled after Greek masc. names in -ατη/ας, for which see Leukart 1994:
173ff. NB 1st mil. attested simplex Μο(υ)ας, Μως, but no simplex *Σιδα/ης.
(23) Parallel examples in feminine names are rare: fAli-wanatti- ‘woman of Ali’ (HE,
Maşat, HKM 113 Ro 11; Trémouille 2006), fḪurma-wanatti- ‘woman of Hurma’ (HE;
Laroche 1966: 73—thus far unique exx. with ‘woman’ parallel to ‘man’ in preceding
type!), f.dSIN.IR-i- (Arma-) ‘servant of the Moon(-god)’, fDINGIR.MEŠ.IR-i- (Maššana-)
‘servant of the gods’ (Laroche 1966: 39&115). NB: both examples with ‘servant’ are
feminine names.
(24) Second member is /walwa/i-/ ‘lion’ (usu. written URMAḪ in cuneiform and LEO in
hieroglyphs), first varies: *Arma-walwi- i.e. LUNA-LEO1-2- ‘Lion of the Moon(-god)’
(hiero. seals HE; Hawkins apud Herbordt 2005: 250), mMūwa-URMAḪ- ‘lion of might’,
m
Piḫa-walwi- ‘lion of resplendence; might’ (HE; Laroche 1966: 124&141), also as Pi-haLEO- and variant Pi-ha-sà-LEO- (hiero. seals HE; Hawkins apud Herbordt 2005: 267-8).
(25) Second member is deity: mKuwalana-dLAMMA- ‘tutelary deity of the army’ (HE;
Laroche 1966: 102), /Halpa-tiwara-/, i.e. TONITRUS.HALPA-pa-SOL-wa/i+ra/i- ‘Sun-god
of Aleppo’, /Wasu-sarma-/, i.e. wa/i4-su-SARMA-ma- ‘Šarruma of good/favor’, and
/Huhha-sarma-/, i.e. AVUS-ha-SARMA-ma- ‘Šarruma of the grandfather’ (HLuvian,
respectively CEKKE, §17o, TOPADA, §1-2, ERKİLET 1, §1; Hawkins 2000: 146,452,494).
(26) Second member is abstract/result noun: Wašḫ-uba- ‘pledge-donation’ (karûm period;
Laroche 1966: 206), surely = Was-ube- ‘idem’ (Lycian, TL 32u,1; Melchert 2004: 107—
identification by Laroche); Hlmi-dewe- ‘extra gift’ or similar (Lycian, TL 139,5;
Melchert 2004: 95).
7
5.2. Adverb+Noun (with deity or term of relationship as 2nd member; those with abstract
2nd member probably instead possessive compounds, listed below under IV.6.2):
(27) /P(a)ri-Sarma-/, i.e. PRAE-ri+i-SARMA-ma- ‘forward (= most prominent) Šarruma’
(HLuvian, KARKAMIŠ 4a, §2; Hawkins 2000: 152); Epñ-xuxa- ‘behind-grandfather =
great-grandfather’(?) (Lycian, TL 127,1; Melchert 2004: 93), Šr-quq- ‘super-/hypergrandfather (idem)’ (Carian; Adiego 2007: 419). See also complex Dd-epñ-newe- ‘Afterdescendant of Ida’ (Lycian, TL 98,1 et alibi; Melchert 2004: 93—analysis of last two
elements with Carruba 1969: 275).
5.3 Adjective+Noun (relatively rare and of varied types)
(28) mKummaya-LÚ- (*Kummaya-ziti-)‘sanctified man’ (HE; Laroche 1966: 97, and also
on hiero. seal HE; Hawkins apud Herbordt 2005: 259&299), mPiyama-(a)radu- ‘given
devotee’ (HE; Laroche 1966: 141—for analysis see Melchert apud Yakubovich 2010: 93
and compare with genitival first member m.dU-na-ra-du- = *Tarḫunna-aradu- and
m
Tarḫunda-(a)radu- ‘devotee of Tarhunt-’; Laroche 1966: 176-7), /Tiyamma-aradu-/, i.e.
Ti-ia-ma+ra/i-tu- ‘__ed devotee’ (HLuvian, İVRİZ 1, §4; Hawkins 2000: 516); also
m
SUM-ma-dLAMMA- ‘the given tutelary deity’ (HE; Laroche 1966: 141—for this
interpretation see Melchert 2011), mUra-ḫattuša- ‘Great Hattusha’ and also hybrid LuvoHittite mUra-walkui- ‘great lion’ (HE; Laroche 1966: 197) beside pure Luvian *Urawalwi-i.e. MAGNUS-LEO- (Hawkins apud Herbordt 2005: 277), mUra-dU- ‘great Stormgod’ (HE; Laroche 1966: 198) and mGAL.dIŠTAR-a- ‘great Šaušga’ (HE; Trémouille
2006); Enẽhi-nere/i- masc. ‘(of) sister of the same mother’ (Lycian, TL 137,1, thus per
Oettinger 1994: 318 with nasal perseveration for *ẽnehi-, effectively equivalent to
Noun+Noun * Ene-nere/i- and type cited in 5.1.2 [21] end).
5.4 Noun+Adjective:
5.4.1 Ordinary adjective:
(29) mMaššana-ura- ‘great (one) of the gods’ (Hittite Empire; Laroche 1966: 115—cf.
titles like tuppa(la)n-uri- ‘chief of the t.’); D-b(i)krm- ‘Ida-mighty’ (Carian; Adiego
2007: 361—see Pik(a)rm- above under IV.2 [4c]); X-uśoλ- ‘X-blessed’ or similar,
including Iduśol-/Duśoλ- = Ιδυσσωλλος ‘Ida-blessed’, Pnuśoλ and variants =
Пονυσσωλλος ‘All-blessed’, Μαυσσωλλος likely ‘Much-blessed’(Carian; Adiego 2007:
passim, but analysis of last with ma- = HLuvian ma ‘much’ < *mégh2, nom.-acc. sg.
neut. is mine—HCM; for uśoλ- see 6.1.4 below).
5.4.2 Past participle (probably with sense ‘__ed by X’, but cf. 8.5.3 below):
(30) Mahane-pijeme/i- ‘given by the gods’, Natrbbijẽme/i- ‘given by Natr’, “translated”
by Ἀπολλόδοτος (Lycian, respectively N302,2 and N320,4; Melchert 2004: 98&100),
latter = Νετερβιμος in Caria. NB also *Wese-pijeme/i- ‘given by/with W.’ = Οσαβαιμις in
Lycia—for last example see refs. in Melchert 2004: 107 and discussion in 8.5.3 below—
and examples in first-millennium Lycia and Pisidia in -π(ε)ιμις (see Houwink ten Cate
1965: 177); Masa-uwẽt(i)- ‘favored/regarded by the gods’ (Lycian context N314a,4 with
Milyan shape; cf. Schürr 2009: 102 and Unuwẽme/i- below in 5.5 [31]).
8
5.5 Adverb+Adjective (usually past participle):
(31) Un-uwẽme/i- ‘well regarded’ (Lycian, TL 62,1; see Schürr 2009, esp. 104); Šaruśoλ- ‘super-/hyper-blessed’ (Carian; Adiego 2007: 418). Possible also is Σροημις in
Cibyratis, virtual *Ser-uwammi- ‘highly-regarded’ (tentatively thus Schürr 2009: 100).
6. Possessive Compounds (“Bahuvrihis”)
6.1 Noun+Noun (all clear examples with genitival relationship ‘having the N2 of N1’):
6.1.1 Massively productive and persisting type /X+mo:wa-/ ‘having the might/potency
of X’ with 1st member deity, toponym, person, or other noun, as already per Houwink ten
Cate 1965: 166-9 (see also below with adjective or adverb as 1st member):
(32) Deities: fḪepa-muwa- ‘having the might of Hebat’, m.dIŠTAR-mūwa- ‘having the
might of Šauška’ (HE; Laroche 1966: 68&171), = sà-US-ka-mu(wa)- (hiero. seals HE;
Hawkins apud Herbordt 270), */Sanda-mo:wa-/ ‘having the might of Sanda’ attested in
Kurzname /Sandamu-/, i.e. ISà-ta-mu- (HLuvian, CEKKE §17i; Hawkins 2000: 146—for
claim of Kurzname see below in 8.1); first example probably also in first-millennium
Κβα-μοας in Lycia (Zgusta 1964: 220).
(33) Toponyms: mḪalpa-muwa- ‘having the might of Aleppo’, mMizra-A.A-a- ‘having
the might of Egypt’, etc. (Laroche 1966: 55&119 and passim), = mi-zi/a+ra/i-mu(wa)(hiero. seals HE; Hawkins apud Herbordt 2005: 264-5), /*Halpa-mo:wa-/ also in
Kurzname /Halpamu-/, i.e. TONITRUS.HALPA-pa-mu- beside */Harrana-mo:wa-/ in
Kurzname /Harrana-mu-/, i.e. Ihara/i-na-mu- ‘having the might of Harran’ (HLuvian,
CEKKE §17k and §17a; Hawkins 2000: 145-6), Kbd-mu- ‘having the might of Kaunos’
(Carian; analysis with Adiego 2007: 370).
(34) Various appellatives: mPūna-A.A- ‘having all might/potency’ (HE; Laroche 1966:
150), = Puna-muwe- (Lycian, TL 35,12 etc.; Melchert 2004: 102), mPiḫa-A.A- and
m
Piḫašša-A.A- ‘having the might of resplendence’ [but prob. actually ‘having the might
of the luminous Storm-god of lightning] (HE; Laroche 1966: 139-40), = pi-ha-mu(wa)and pi-ha-sà-mu(wa)- (hiero. seals HE; Hawkins apud Herbordt 267), mUḫḫa-mūwa‘having the might of the grandfather’ (prob. thus with Yakubovich 2010: 191—cf. Nanimuta- below), mIrḫa-A.A.- ‘having the might of the border territory’ (HE; Laroche 1966:
80), /Pidanti-mo:wa-/, i.e. ILOCUS-ta/i4-ti-mu-wa/i- ‘having the might of the place’
(HLuvian, BOYBEYPINARI 1 §11; Hawkins 2000: 336—NB full form of 2nd member);
first millennium examples with unclear 1st member, including u/úksmu- ‘having the
might of waksa-’ (Carian; Adiego 2007: 427) = Ουαξα-μοας/-μως in Isauria and Cilicia.
See further 6.1.4 below.
(35) Also with 2nd member /mo:wata/i-/ (see above IV.2 [3a]): Puna-muwati- ‘having all
might/potency’ (karûm period; Laroche 1966: 150), /Pana-mo:wati-/ fem. (HLuvian,
BOYBEYPINARI 1, §1 etc.; Hawkins 2000: 336 —NB first vowel and full form of 2nd
member!), /Nani-mo:(wa)ta-/, i.e. Ina-ni-mu-ta- ‘having the might/potency of the brother’
(HLuvian, KULULU lead strip 1, §7.41; Hawkins 2000: 508).
6.1.2 X+/tsalma-/ (with rhotacized variant /tsarma-/) ‘having the protection of X’ (thus
with Zehnder 2010: 52 contra Satzname as per Neumann 1976: 141; for historical
analysis of word see Melchert 1988: 241-3, pace Lipp 2009: 1.27533):
9
(36) */Tarhu(nta)-tsalma-/, i.e. dU-zalma- ‘having the protection of Tarhunt-’, mYarraPAP-/-zalma- ‘having the protection of Yarri’ (HE; Laroche 1966: 177&76), first also in
TONITRUS-hu-za+ra/i-ma- /Tarhu-tsarma-/ (HLuvian, EĞREK, §1; Hawkins 2000:
493); also as per Neumann 1976: 140-1 in 1st millennium Τροκο-ζαρμας in Cilicia, also
Ια-ζαρμας ‘having the protection of Ea’ and Ρω-ζαρμας (for *Ρωνζ-ζαρμας) ‘having the
protection of Runza-’ (contra Houwink ten Cate 1965: 136, who takes as divine name
Šarruma-).
6.1.3 X+/wa:su-/ ‘having the good (i.e., favor) of X’ with deity, toponym, and other less
certain first members:
(37) mAli-wašu- ‘having the favor of A.’, fAnni-wašu- ‘having the favor of the mother’(?)
(HE; Laroche 1966: 33), /Halpa-wasu-/, i.e. TONITRUS.HALPA-pa-wa/i-su- ‘having the
favor of Aleppo’ and /Tarhu-wasu(wa)-/, i.e. TONITRUS-hu-wa/i-su-wa/i- ‘having the
favor of Tarhunt-’ (HLuvian, respectively MARAŞ 11, §7 and CEKKE §17l; Hawkins
2000: 271&146), latter also in mTarḫu-wašu- (HE; Trémouille 2006).
6.1.4 Probably in X+/waksi-/ ‘having the waksa- of X’ with “i-mutation”. In the absence
of a clear 2nd-millennium Luvian source, the meaning of the word remains unclear, but
waxs(s)a- is well-attested as an appellative in Milyan (see Melchert 2004: 134 w/ refs.),
where the sense is likely positive. As per (34) above, it appears as the 1st member of
Carian uksmu- = Ουαξα-μοας/-μως ‘having the might of waksa-’. I interpret names from
Caria with 2nd member -υασσις, -υαξις, -υαƬις as containing the same element: Πανυασσις/-υαƬις ‘having all w.’, Ακτα-υασσις ‘having the w. of a.’ (NB also Βρυαξις/-υασσις Blümel 1992: 12). Likely source of further derived adjective uśoλ- ‘waksaed’ (‘blessed’ or similar) < virtual *waksy-alla-: see Carian determinative compounds in
5.4.1 (29) above. See also 6.2 (38) immediately below on Σαρυασσις. Cf. discussion by
Adiego 2007: 344 with note.
6.2 Adjective/Adverb+Noun:
(38) mPariya-mūwa- ‘having might beyond’, i.e. ‘having surpassing might’ and surely
also mPariya-watra- ‘having surpassing w.’ (HE; Laroche 1966: 137), mUppara-mūwa‘having superior might’ (HE; Laroche 1966: 197 —*uppara- ‘superior’ < PIE *upero- =
Avestan upara- and German ober ‘upper, situated above’), = Οπρα-μοας in Lycia and
Οπρα-μως in Cilicia-Pamphylia. See Houwink ten Cate 1965: 162-3, but with confusion
of Οπρα- ‘superior’ with distinct Οβρα- (see correctly Schürr 2009: 107). NB also
Pisidian Ουπερ-δοται- ‘having superior X’ (see Brixhe 1988: 146-7); /Ura-mo:wa-/, i.e.
I
MAGNUS+RA/I-mu-wa/i- ‘having great might’ (HLuvian, KULULU lead strip 1, §2.3;
Hawkins 2000: 506), Σαρυασσις ‘having super-/hyper-waksa-’ (cf. Adiego 2007: 340).
7. “Satznamen”
Not a typical Indo-European type and in Anatolia almost certainly created in imitation of
Hurrian and Akkadian models. Source betrayed by unexpected verb-initial order in some
examples, but eventually adapted further to (S)OV word order of Anatolian IE languages.
Some putative examples necessarily remain speculative!
10
7.1 Type with verb ‘to be’ in “Wunschnamen” either with imperative or indicative:
7.1.1 Copulative (‘shall be X’)
(39) Aruwãti(j)-esi- ‘he shall be high/exalted’ (Lycian, TL 44b,18 etc.; Melchert 2004: 92
—analysis with Neumann 1978: 127).
7.1.2 Possessive (‘X shall be [to him]’ = ‘he shall have X’)
(40) /A:stu-alamantsa-/, i.e. Iá-sa-tú-wa/i-la-ma-za- ‘let name/renown be (to him)’ and
variants, /A:sti-wa:su-/, i.e. Iá-sa-ti-wa/i-su- ‘good/favor shall be (to him)’, /A:stiTarhunza-/, i.e. Iá-sa-ti-TONITRUS-hu-za- ‘Tarhunza- shall be (to him)’ (HLuvian,
respectively KARKAMIŠ A27u, l. 2, MARAŞ 11, §7, KARKAMIŠ A7, §8; Hawkins
2000: 165, 266,129), Esi-tmãta- ‘renown shall be (to him)’ (Lycian, TL 35,18; Melchert
2004: 94—reading as personal name and analysis with Neumann 1983: 147).
7.2 Putative examples with other verbs
(41) mAnza-paḫḫadu- ‘let him protect us’ and m.dU-manaddu ‘let Tarhunt- see (him)’
(HE; Laroche 1966: 34 and 1981: no. 1279a respectively—analysis with Yakubovich
2010: 92-3); more speculatively perhaps (HCM) mMana-pa-dU- = *Mana-pa-tarḫunta‘Just see (him), Tarhunt! and mMušiši-pa-dU- ‘You shall be satiated, Tarhunt!’ (for
names see Laroche 1966: 112&121)—but presumed inflection of verb muš- ‘be satiated’
far from assured!
(42) /Atsa-tiwada-/, i.e. I(LITUUS)á-za-ti-i-wa/i-tà- ‘the Sun-god favors’ and /Piyatarhuntsa-/, i.e. pi-ya-TONITRUS-hu-zá- ‘Tarhunt- gave/has given (him)’ (HLuvian,
respectively KARATEPE passim and CEKKE §17h; Hawkins 2000: 48ff. and 146)
perhaps by haplology < *Azati-tiwada- and *Piyata-tarhunza- (for genuine example of
haplology see Kupa-piya- for *Kupapa-piya- ‘Kubaba gave/given by Kubaba’ (HLuvian,
SHEIZAR §1; Hawkins 2000: 417), or rather with uninflected stem based on Hurrian
models? Compare type of fTatu-ḫepa- (tād=o=Ḫeba(t)) ‘Heba(t) loved’ with transitiveergative stem marker -o- but no personal ending. Imperative reading ‘Favor (him), Sungod!’ also possible for first, but seems less likely for second (on Hurrian Satznamen see
Giorgieri 2000: 283ff.).
On putative examples with second member -piya- see immediately below in discussion
of Kurznamen.
8. “Kurznamen”
8.1 Reasonably certain examples in HLuvian for possessive componds in /-mo:wa-/:
/Halpa-mu-/, /Harrana-mu-/, /Sanda-mu-/, /Tsuna-mu-/ in CEKKE §17 (Hawkins 2000:
146&150). Mere syncope unlikely in view of contemporary examples with full form of
second member: /Ura-mo:wa-/ (see (38) above), /Pidanti-mo:wa-/ (see (34) above), and
others.
8.2 Also very likely in Luvian names in /-atsi-/ abbreviated from compounds with /tsidi-/
as 2nd member (see above 5.1.2 [22]): mTarḫunazi- < mTarḫunda-ziti- (HE; Laroche
1966: 176—NB also abbreviated form of 1st member, but attestation is indirect, in
Assyrian source), /Mo:watsi-/, i.e. IMu-wa/i-zi- < *Muwa-ziti-, i.e. mMuwa-LÚ-i(HLuvian, MARAŞ 1, § etc; Hawkins 2000: 262-3—his reading Muwizi- very unlikely,
since suffix -izza- never shows “i-mutation”).
11
8.3 mTarḫu(n)miya- (Maşat, HKM passim; HE), contra Alp (1988: 98) hardly with suffix
-mi- or progressive assimilation of -piya-. Rather Kurzname with productive
appurtenance suffix -iya- < *-iyo- from mTarḫu-mimma- cited in 8.5.2 below.
8.4 Some examples listed above under IV.2 as simple appellatives could (but need not!)
be extracted from compounds (e.g. mūwa- ‘might, potency’, pīḫa- ‘resplendence;
might’). Likewise Lycian Uwẽme/i- ‘regarded, favored’ (TL 109,2; Neumann 2007: 413
& Schürr 2009: 102) and HLuvian /atsamma/i-/ ‘favored’ (Iá-za-mi- in İZGİN 2, §9 etc.;
Hawkins 2000: 316). If Sidetic Zem- is equivalent, in absence of evidence for aphaeresis
in native Sidetic, it could be extracted from compounds like Ubat-zem- ‘favored
by/favorite of the demesne’ (cf. Pérez Orozco 2008: 128-9).
8.5 Most problematic are examples with descriptively bare verbal stems as second
member (see acknowledgement regarding -piya- by Laroche 1966: 318-9). NB no
evidence for nouns piya-, mimma-, or wiya- (claim of Neumann 1978: 126 for first based
on false older readings of HLuvian signs):
8.5.1 Numerous examples in X-piya-. Are these ‘X gave/has given’ or ‘given by/to X’?
(43) mIyara-SUM-ya- ‘Yarri-’, mMašna-piya- ‘god(s)-’, m.dU-SUM- ‘Tarhunt-’ (HE; Laroche
1966, passim), last example also in /Tarhu(nta)-piya-/, i.e. TONITRUS-hu-pi-ya(HLuvian, MARAŞ 9; Hawkins 2000: 275), also Wese-pije- ‘W.-’ (Lycian, TL 9,2) =
Οσσαπιας (see Schürr apud Melchert 2004: 107), and numerous indirect first-millennium
examples in-πις, -πιας, -πεας etc. (Houwink ten Cate 1965: 176-7; Zgusta 1964: passim).
NB mWašu-piya- (HE; Trémouille 2006) tends to favor ‘S/he (the deity) has given favor’,
but adverbial 1st member ‘well-given’ not impossible.
8.5.2 Also at least one example with X-mimma-:
(44) mTarḫu-mimma- ‘Tarhunt- (has) favored’ or ‘favored by Tarhunt-’ (HE; Laroche
1966: 176—for mimma- as ‘regard, favor’ see Melchert 1988: 218-20).
8.5.3 As per Yakubovich 2011, likewise X-wiya- in feminine names with deities,
toponyms, and other first members, either ‘X (has) sent’ or ‘sent by X’, or ‘sent to X’ (or
both of the latter?):
(45) *Arma-wiya- (f.dSIN/MI-wiya-) ‘the Moon-god (has) sent’ or ‘sent by/to the Moongod’, f Ḫalpa-wiya- ‘sent to/by Aleppo’, fParšana-wiya- ‘sent to/by the panther’,
f
Tirguta-wiya- ‘sent to/by T.’, and nearly twenty more (for attestations see Laroche 1966
and Trémouille 2006).
Two alternative accounts: first, can all be Kurznamen from determinative compounds
with participles as second members. NB co-existence of Lycian Wese-pije- = Οσσαπιας
and *Wese-pijeme/i- = Οσαβαιμις; likewise possible mTarḫu-mimma- < *Tarḫu(nta)mimmamma/i-, and Tunna-wiya- < *Tunna-wiyamma/i- ‘sent by/to Tunna’. But no
attested examples with second member *-wiyamma/i- and no 2nd millennium examples
for those in *-piyamma/i-. Second alternative is to assume Satznamen with descriptively
bare verbal stem “derived” as per above 7.2 (42), by haplology or after Hurrian models.
Favored by likely example with piya- as first member in HLuvian /Piya-tarhuntsa-/ with
verb initial as in Hurrian. mSUM.dU- in Maşat (HKM 63 Ro 1 & 64 Ro 3) more likely to
be matching *Piya-Tarḫunza- ‘Tarhunts (has) given’ than *Tarḫu-piyamma- (thus Alp
1991: 88 followed by Trémouille 2004): only tutelary deity is given/dedicated.
12
9. Hypocoristica
As per Zehnder (2010: 42-5) only likely candidates are some of names in -nna/i- that
likely contain variant of Luvian diminutive suffix -(a)nna/i- (e.g. armanna/i- ‘lunula’ <
arma- ‘moon’). Most suggestive examples include:
(46) *Tarḫunni-, i.e. dU-ni- (HE; Laroche 1966: 176—unlikely to be simple theophore
*Tarḫunni- because this form of the Storm-god’s name is always Tarḫunna-), also =
first-millennium Ταρκυννις in Cilicia (Zgusta, 1964: 488); mZidanna/i- < mZida/i- (HE;
Laroche 1966: 211). NB also fAnni-wiya-ni- (HE; Laroche 1966: 33)—thus also
Yakubovich 2011.
V. Tentative Summary by Language (based only on direct attestations!)
1. Luvian shows all types.
2. Lycian shows all types except assured Kurznamen.
3. Carian also lacks (native) Satznamen, but phonological truncation may have obscured
examples. Many names borrowed from Egyptian are Satznamen.
4. Lydian thus far lacks any (native) compounds or Satznamen.
5. Pisidian attests use of appellatives and compounds.
6. Sidetic attests uses of appellatives and compounds.
VI. Broader Summary
1. Use of determinative and possessive compounds is robust in 2nd and 1st millennia in
southern and southwestern Anatolia (Cilicia, Pisidia, Lycia, Caria). Absence in Lydian
probably significant (NB Μουσατης not only not native but probably not analyzed as
compound), but caution is in order.
2. “Arzawan” mAnza-paḫḫadu- may or may not attest penetration of Satznamen further
north than Cilicia and Lycia (cf. Yakubovich 2010: 92). Absence in Pisidian and Sidetic
texts likely due to extremely limited corpora.
3. Strong continuity in first millennium in theophores and theophoric compounds with
Arma- (Moon-god), Ea-, Runtiya-/Runtsa- (Stag-god), Sanda- (war god), and Tarhunt/Tarhunza- (Storm-god), but striking disappearance of Luvian Sun-god Tiwad(a)- (only
once in Lydian context) and war god Yarri-. Also no assured trace of Šarruma- (contra
Houwink ten Cate 1965: 136), but at least one example for Hebat (see (32) above).
References
Adiego, Ignacio J. 2007. The Carian Language (Handbook of Oriental Studies. Section
One. The Near and Middle East 86). Leiden and Boston: Brill.
Alp, Sedat. 1988. Hethitische Briefe aus Maşat–Höyük. Ankara: Türk Tarih Kurumu
Basımevi.
Blümel, Wolfgang. 1992. Einheimische Personennamen in griechischen Inschriften aus
Karien. Epigraphica Anatolica 16.29-43.
13
——. 1994. Über die chronologische und geographische Verteilung einheimischer
Personennamen in griechischen Inschriften aus Karien. In M. E. Gianotta et al.
(eds.), La decifrazione del cario. Atti del 1° Simposio Internazionale. Roma, 3-4
maggio 1993, 65-86. Rome: CNR.
Brixhe, Claude. 1988. La langue des inscriptions épichoriques de Pisidie. In Yoël
Arbeitman (ed.), A Linguistic Happening in Memory of Ben Schwarz, 131-55.
Louvain-la-Neuve: Peeters.
Carruba, Onofrio. 1969. Su alcuni nomi di parentela in licio e in nesico. Parola del
Passato 24.269-78.
Cau, Nicolà. 2005a. Onomastica licia. Studi Ellenistici 16.345-76.
——. 2005b. Nuovi antroponimi indigeni nelle iscrizioni greche della Licia di età
ellenistico-romana II. Studi Ellenistici 16.377-421.
García Ramón, José Luis. forthcoming. Anatolian, Greek, Indo-European: From ‘run’ to
‘assist, help’. Phraseology and syntax in the reconstruction of the Indo-European
lexicon. Handout from Thirtieth East Coast Indo-European Conference, Harvard
University June 9, 2011. To appear in Innsbrucker Beiträge zur Sprachwissenschaft.
Giorgieri, Mauro. 2000. Schizzo grammaticale della lingua hurrica. In La civiltà dei
hurriti (= La Parola del Passato 55), 171-295.
Goetze, Albrecht. 1954. The Linguistic Continuity of Anatolia as Shown by its Proper
Names. Journal of Cuneiform Studies 8.74-81.
Güterbock, Hans G. 1940. Siegel aus Boğazköy I (Archiv für Orientforschung Beiheft 5).
Berlin: Weidner.
——. 1942. Siegel aus Boğazköy II (Archiv für Orientforschung Beiheft 7). Berlin:
Weidner.
Gusmani, Robert. 1964. Lydisches Wörterbuch. Heidelberg: Winter.
——. 1988. Anthroponomie in den lydischen Inschriften. In Yoël Arbeitman (ed.), A
Linguistic Happening in Memory of Ben Schwarz, 179-96. Louvain-la-Neuve:
Peeters.
Hawkins, John David. 2000. Corpus of Hieroglyphic Luwian Inscriptions. Volume 1:
Inscriptions of the Iron Age. Berlin and New York: de Gruyter.
Herbordt, Suzanne. 2005. Die Prinzen- und Beamtensiegel der hethitischen Grossreichszeit auf Tonbullen aus dem Nişantepe-Archiv in Hattusa. Mainz: von Zabern.
Houwink ten Cate, Philo H. J. 1965. The Luwian Population Groups of Lycia and Cilicia
Aspera during the Hellenistic Period. Leiden: Brill.
Laroche, Emmanuel. 1966. Les noms des hittites. Paris: Klincksieck.
——. 1981. Les noms des hittites. Supplément. Hethitica 4.3-58.
Leukart, Alex. 1994. Die frühgriechischen Nomina auf -tās und -ās: Untersuchungen zu
ihrer Herkunft und Ausbreitung (unter Vergleich mit den Nomina auf -eús). Vienna:
Österreichische Akademie der Wissenschaften.
Lipp, Reiner. 2009. Die indogermanischen und einzelsprachlichen Palatale im
Indoarischen. Heidelberg: Winter.
Melchert, H. Craig. 1988. Luvian Lexical Notes. HS 101.211-43.
——. 2004. A Dictionary of the Lycian Language. Ann Arbor and New York: Beech
Stave.
——. 2011. Reciprocity and Commerce in Bronze and Iron Age Anatolia. Handout from
the 57th Rencontre Assyriologique Internationale. Rome, July 7, 2011.
14
Neumann, Günter. 1976. Zu einigen hethitisch-luwischen Personennamen. Zeitschrift für
Vergleichende Sprachforschung 90.139-44.
——. 1978. Spätluwische Namen. Zeitschrift für Vergleichende Sprachforschung 92.12631.
——. 1983. Zur Erschließung des Lykischen. In E. Vineis (ed.), Le lingue indoeuropee di
frammentaria attestazione, 135-51. Pisa: Giardini.
Istanbuler Forschungen 30.
——. 2007. Glossar des Lykischen. Überarbeitet und zum Druck gebracht von Johann
Tischler (Dresdner Beiträge zur Hethitologie 21). Wiesbaden: Harrassowitz.
Nollé, Johannes. 2001. Side im Altertum: Geschichte und Zeugnisse. Band II. Bonn:
Habelt.
Oettinger, Norbert. 1994. Etymologisch unerwarteter Nasal im Hethitischen. In Jens E.
Rasmussen (ed.), In honorem Holger Pedersen. Kolloquium der Indogermanischen
Gesellschaft vom 26. bis 28. März 1993 in Kopenhagen, 307-30. Wiesbaden:
Harrassowitz.
Pérez Orozco, Santiago. 2003. Propuesta de nuevos valores para algunos signos del
alfabeto sidético. Kadmos 42.104-8.
——. 2005. Los signos para consonante aspirada en sidético. Kadmos 44.75-7.
——. 2008. La lengua sidética. Ensayo de síntesis. Kadmos 46.125-42.
Rieken, Elisabeth. 2005. Neues zum Ursprung der anatolischen i-Mutation. Historische
Sprachforschung 118.48-74.
Schürr, Diether. 1991-93. Imbr- in lykischer und karischer Schrift. Die Sprache 35.163-73.
——. 2001. Karische und lykische Sibilanten. IF 106.94-121.
——. 2009. Lykisch und karisch un-. Historische Sprachforschung 122.100-110.
Starke, Frank. 1990. Untersuchung zur Stammbildung des keilschrift-luwischen Nomens
(Studien zu den Boğazköy-Texten 31). Wiesbaden: Harrassowitz.
Trémouille, Marie-Claude. 2006. Répertoire onomastique. Online at:
http://www.hethport.uni-wuerzburg.de/hetonom/
Watkins, Calvert. 1995. Some Anatolian words and forms: Hittite nega-, negna-, Luvian
*niya-, nani-. In Heinrich Hettrich et al. (eds.), Verba et Structurae. Festschrift für
Klaus Strunk. Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck.
Yakubovich, Ilya. 2010. Sociolinguistics of the Luvian Language (Brill’s Studies in IndoEuropean Languages & Linguistics 2). Leiden and Boston: Brill.
——. 2011. The Luwian Names in -wiya. Paper presented at Luwian Identities:
Culture, Language and Religion Between Anatolia and the Aegean, University of
Reading, June 10, 2011.
Zehnder, Thomas. 2010. Die hethitischen Frauennamen (Dresdner Beiträge zur
Hethitologie 29). Wiesbaden: Harrassowitz.
Zgusta, Ladislav. 1964. Kleinasiatische Personennamen. Prague: Czechoslovakian
Academy of Sciences.

(Selected) Topics in Anatolian Historical Grammar

Transcription

Similar documents

Hittite Empire and Society

WordGumbo Indo-European Comparative Indo

The Text of Hesiod`s Theogony and the Hittite Epic of Kumarbi

Anatolian Hieroglyphs: Logogram vs. Ideogram¹.

JOSOS project completed (PDF Document)

Yachting Monthly Turkey 2006

HAPPEN

palaeography and scribal practices in syro

BUONGiORNO!

know the rules…. - Saint Mary Gaels