Workshop on East Caucasian preverbs University of Pavia - Sep. 11th 2014 Reanalysis of loan & compound verbs as bipartite stems in Lezgic Gilles Authier Ecole Pratique des Hautes Etudes, Paris 2 Lezgic 9-11 languages peripheric : Archi (-), Udi (+, opaque) // core : west : Tsakhur + Rutul South : Kryz + Budugh East : Lezgian + Aghul (2) + Tabasaran (2) Richest / most productive preverb systems of each branch Most productive preverb systems Budugh : all verb forms have a pre-agreement morpheme Aghul (2) + Tabasaran (2): richest and most complex (regular) systems Rutul: can derive preverbed verbs from non preverbed; complex, irregular combinations Rutul (west Lezgic) suˤ<buˤ>čvaˤ-s « dismantle, undo, demolish (A) » Tabasaran (east Lezgic) ži<b>k’ub « wash (N) » Budugh (south Lezgic) ħa<ba>raq’ar « talk (HPL) » « talk » « wash » « demolish » = ‘’basic’’ verbal concepts Rutul (west Lezgic) suˤ<buˤ>čvaˤ-s « take down, undo, demolish » Rutul suˤ<buˤ>čvaˤ-s « demolish » -č (v)- as a root : luˤ-čvaˤ-s « mount » suˤ-čvaˤ-a « dismount » cf. si-xhi-n ‘put down’ / li-xhi-n ‘put on’ +phonetic rule: all labialized palatals trigger pharyngealization. Rutul suˤ<buˤ>čvaˤ-s « demolish » / root -č-: « flow » ki<y>č-es « flow down » ağa<y>č-es « flow out » a<r>č-es « crumble » la-qh<yi>č-es « rush on » si-č-es « rot » Rutul suˤ<buˤ>čaˤ-s « demolish » č- as a preverb : « off » ča-<b>-qhva-s « scratch » ču<r>qhva-s « sneeze » Rutul suˤ<buˤ>čaˤs « demolish » Gender-number agreement interfix: Yi-c хal suˤ-<buˤ>-čuˤr 1-erg house pv-a-demolish.aor TAM sensitive: prs: suˤ-vuˤ-rčvaˤ-r PF: suˤ-buˤ-čuˤIPF: suˤ-vuˤ-r-čv-aˤFUT: suˤ-buˤ-čv-aˤ- Rutul suˤ<buˤ>čaˤs « demolish » … !!! Attested only in the southern dialect of Shin(/Borch), on the border with Azerbaijan Meaning both „take down a stone wall” & „undo a seam” = Azeri sök-, pronounced söč- in this region Transfer helped by the presence of similar looking preverb and root Bi-morphemic verbal structure as default ! Tabasaran (east Lezgic) ži<b>k’ub « wash » Tabasaran ži<b>k’ub « wash » Tamal-i battil ži<b>k’-ura. Tamal-ERG pot(N) PV<N>wash-PRS "Tamal washes the pot. » -ž- « expressive » PV, found in Rutul, Budugh, Lezgian, etc… Tabasaran 2d position preverbs t’Vc’Vč’VčVžV- hi-t’i-<b>-k’qh-c’a-<b>-xha-č’a-<b>-kkqh-ča-<b>-ğkka-ža-<b>-ğ- ‘to steal’ ‘to scratch’ ‘to reach for’ ‘to rinse’ ‘to run after’ = coronal Opaque: effort, energy… Can appear Tabasaran ži<b>k’ub « wash » = Aghul ‘to wash’ / ‘to find’ in Aghul ži-k’as ‘to wash (a corpse)’ / ‘to find’ not isolated : ži-rxvas ‘to comb’ ži-rħas ‘to scold; to throw up’ (kke-)ži-kas ‘to sweep (under)’ Aghul ‘to wash’ / ‘to find’ Mayit žik’i-naadava Corpse wash.PF-PERFNEG “The corpse has not been washed” kavan-is sakinvel da-žik’e-y shepherd-DAT rest NEG-find.IPF-PTCP xhu-naya. be.PF-PERf « The shepherd did not find rest. » furq’u-naya zas sa sanduq’ 1.DAT a trunk « I have found a trunk. » find.PF-PERF Tab. / Aghul ži<b>k’ub « wash » outside of Aghul-Tabasaran, no cognate for ži-k’- ‘to wash’ or ‘to find’ // reflexes of Agul furq’as ‘find’ are found in all branches // a root zi- ‘wash’ found in Kryz: zimayc But why two verbs for ‘find’ ? Why this homonymy ‘wash’ / ‘find’ Tab. ži<b>k’ub ‘wash’ (+ Aghul ‘find’) Cf. Azeri yaxala- 1) ‘to wash/rinse’ cf. Turkish yıka- ; 2) ‘to catch’ < yaxa ‘collar’, cf Turkish yaka-laOghuz y = Kipchak ž Kumyk: čayk- (to wash/rinse), čayqa- ’to swing/shake’, yağa ‘collar’ must be an Oghuz loan But Karachai-Balkar has čayqa- ‘to wash/rinse’ + ‘to swing/shake’, žaqa (collar) Uyghur has žaqa- ‘rinse’ and žaqa ‘collar’ Tab. ži<b>k’ub ‘wash’ (+ Aghul ‘find’) Conclusion: ži-k’- ‘wash’ (Aghul and Tabasaran) must be a loan from Kipchak copying its rare semantic properties in Aghul only : homonymy with ‘find’ Reanalysed as a bi-partite (agreement taking) stem in Tabasaran Budugh (south Lezgic) ħa<ba>raq’ar « talk » 1st position preverbs in Budugh ʕa-č’i ‘to go in’ qe-č’i ‘to go out’ ye-č’i ‘to cross’ ču-su ‘to put under’ o-su ‘to put down’ qu-su ‘to put on’ Most are related to case markers, with moderate productivity But a- in ħa-raq’ar ‘to talk’ is an hapax Budugh ħa-<ba>raq’ar ‘to talk’ causative forms: bi-partite stem: ‘to make her/me/you talk’ interfixed agreement: Front-vowel harmony (IPF) M: ħa-re-q’i M/F: ħa-ra-q’ar F: ħa-re-r-q’i A: ħo-ro-q’or A: ħö-rö-q’ü HPL: ħa-ba-raq’ar HPL: ħa-be-rq’i Budugh ħa-<ba>raq’ar ‘to talk’ 1st element : hapax But has a clear etymology: Aghul ħar + aq’ar ‘to let know, teach’ = Rutul xˁar ha-ʔa-s = Kryz hara v-aridž Tsakhur: a-xˁa-s ‘be able’ (neg. dya-xˁa-s) xˁa-r/b/d ‘large’ < ‘adult’ < ‘aware/able’ « Learn » in Rutul: shifting valency Old : agreeing with the experiencer Ğu хaˤr r-iš cvaˤr šiv-bı i-de! 2 knowing f-be.imp refl.hpl what-pl cop-q « Learn what sort of people they are! » Recent: agreeing with the stimulus Mi-bı ğvas haldıgı xaˤr ruˤq’uˤr ? prox-pl 2.dat who.el known (npl)become.pf « How have you learned these (things) ? Budugh ħa-<ba>raq’ar ‘to talk’ Resilience of the compound construction Rij Cümşüd-a’ dard ħar soxho-ci Girl C.-in worry aware be.a.pf-nar « The girl learned C.’s problem. » A-cber ħar si-b-Ɂir ki… 3-hpl aware pv-hpl-do.pf(imp) that… « Teach them that… » Budugh ħa-<ba>raq’ar ‘to talk’ Conclusion: In Budugh syncronically co-occurrence of old, frozen + reanalyzed compound a-raq’ar / ħereq’i « talk / make talk » new compound locution ar siɁi / yɨxha « teach/learn » In Tsakhur, a frozen action known har-ak’ « saying, proverb » likewise preserves the old form of the verb ‘to do’ beside the new one haɁ-