informaciis struqturirebis ZiriTadi modelebi qarTvelur enebSi · (d) nawilakebi...

ivane javaxiSvilis saxelobis Tbilisis saxelmwifo universiteti

Ivane Javakhishvili Tbilisi State University

Teoriuli da SedarebiTi enaTmecnierebisa da fonetikis instituti

Institute of Theoretical and Comparative Linguistics and Phonetics

informaciis struqturirebis ZiriTadi modelebi

qarTvelur enebSi

MAIN MODELS OF INFORMATION STRUCTURING IN THE KARTVELIAN LANGUAGES

Tbilisi Tbilisi 2009

krebulSi warmodgenilia qarTvelologiis, humanitaruli da soci-aluri mecnierebebis rusTavelis fondis mier dafinansebuli proeqtis (№: 334, 2006-2009, `informaciis struqturirebis ZiriTadi modelebi qarTvelur enebSi~) farglebSi gamarTuli samecniero konferenciis masalebi. avtorebi arian proeqtis monawileni: r. asaTiani, m. ivani-Svili, iv. leJava, ir. meliqiSvili, eT. soselia, n. wereTeli, m. jiqia; da magistrantebi: q. grZeliZe, n. dundua, T. janeliZe. statiebSi aRwerilia qarTveluri enebisaTvis damaxasiaTebeli fonetikur-fonologiuri, morfosintaqsuri Tu leqsikur-semantikuri da pragma-tikuli modelebi, romlebic winadadebis formirebisas sainformacio nakadSi dawinaurebuli (resp. gamokveTili, xazgasmuli, TvalTaxedvis centrSi moqceuli, Tematuri da a.S.) wevris markirebis mizniT gamoiyeneba. kvleva efuZneba potsdam-berlinis saerTaSoriso proeqtis farglebSi SemuSavebuli saeqsperimento amocanebis gadawyvetis Sedegad savele pirobebSi mopovebuli masalis analizs, romelic anotirebulia saerTaSoriso standartebis Sesabamisad da xelmisa-wvdomia qarTvelologiuri sakiTxebiT dainteresebuli nebismieri (ara mxolod qarTulenovani) pirisaTvis. statiebSi ganviTarebul mosaz-rebebs Teoriuli Rirebulebis garda praqtikuli daniSnulebac aqvs: informaciis struqturirebis ZiriTadi formaluri modelebis gamov-lenis Semdeg muSavdeba garkveuli rekomendaciebi teqstis gamiz-nulobisda Sesabamisad mizanSewonili konstruqciebis SerCevisaTvis. amdenad, vfiqrobT, krebuli daainteresebs ara mxolod enaTmecnierTa wres, aramed farTo sazogadoebriobasac.

redaqtori: rusudan asaTiani

garekanis dizaini: sandro asaTiani

kompiuteruli dizaini: ivane leJava

teqnikuri redaqtori: luiza lorTqifaniZe

© qarTvelologiis, humanitaruli da socialuri

mecnierebebis rusTavelis fondi, 2009

© ivane javaxiSvilis saxelobis

Tbilisis saxelmwifo universiteti, 2009

ISBN 978-9941-404-72-6

gamomcemloba `nekeri~, 2009

3

rusudan asaTiani

intonaciis roli winadadebis sainformacio struqturis

formirebaSi

Sesavali

enis mTavari funqciaa komunikacia. komunikaciis dros xdeba informaciis

gadacema. informaciis TvalsazrisiT Setyobinebis yvela erTeuli erTnairi

Rirebulebis ar aris da, amdenad, lingvisturi interesebis sferoSi buneb-

rivad Semodis enis pragmatikuli aspeqti – yovel niSans formisa da

mniSvnelobis garda pragmatikuli Sinaarsic aqvs.

informaciis struqturirebisas, rogorc wesi, gamoiyofa komunikaciis

konkretuli aqtisaTvis pragmatikulad mTavari Rirebulebis mqone

centraluri elementi, romelsac SesaZlebelia vuwodoT aqtualizebuli. informaciis lingvistur formaSi moqceva (informaciis SefuTva) gulisxmobs

swored am aqtualizebuli nawilis formalur markirebas, rac sxvadasxvagvari

lingvisturi Tu paralingvisturi meqanizmebiT miiRweva. aqtualizeba –

informaciis centraluri elementis gamokveTa (xazgasma, fokusSi moqceva,

gaSuqeba) – sxvadasxva struqturis mqone enebSi gansxvavebuli formaluri

saSualebebiT xorcieldeba da nebismier enobriv doneze SeiZleba iyos

realizebuli. konceptualuri aqtualizeba

eqstralingvisturi situaciis lingvisturi struqturireba bunebriv enebSi

ZiriTadad sami mTavari konceptualuri ganzomilebis mixedviT mimdinareobs:

1. situaciis Semadgeneli semantikuri rolebis (agensi-paciensi-adresati)

markireba;

2. sainformacio nakadis pragmatikuli Rirebulebebis (aqtualizebuli-fonuri)

markireba;

3. sakomunikacio aqtis statusebis (molaparake/msmeneli-saubris sagani)

markireba.

konkretuli enebi irCeven maTTvis misaReb strategiebs da Sedegad vRebulobT

situaciis konceptualuri aqtualizebis gansxvavebul modelebs.

semantikuri rolebis markirebaze orientirebul enebSi informaciis

TvalsazrisiT aqtualizebul saxelad konceptualizdeba an agensi an paciensi,

ris Sedegadac warmoiqmneba nominatiuri an ergatiuli konstruqciebi. pirveli

afiqsirebs informaciis im nawilis aqtualizebas, romelic gamoxatavs, Tu vin moqmedebs, meore ki informaciis im nawilis aqtualizebas, romelic gamoxa-

tavs, Tu ra xdeba, ra iqmneba. konceptualurad aqtualizebuli saxeli

enobriv struqturebSi formalurad warmodgenilia aramarkirebuli, saxelo-

4

biTi brunviT: nominatiur konstruqciebSi es aris agensi, xolo ergatiul

konstruqciebSi – paciensi. enebi, romlebic, upirveles yovlisa, sainformacio nakadis pragmatikul

konceptualizacias axdenen, formaluri struqturebiT erTmaneTs upiris-

pireben informaciis aqtualizebul fonur monakveTebs. aseT enebSi, rogorc

wesi, aqtualizebuli saxeli markirebulia specialuri morfologiuri

afiqsiT.

rac Seexeba mesame tipis enebs, aq informaciis aqtualur wevrad konceptu-

alizdeba sakomunikacio aqtSi monawile pirebi da gramatikulad gamoixateba

piris I/II : III diqotomia. funqcionaluri aqtualizeba

funqcionarulad aqtualizebuli saxeli gramatikul tradiciaSi aRiniSneba

terminiT subieqti. es aris Setyobinebis is nawili, romlis Sesaxebac modis

garkveuli informacia da romelic, bunebrivia, aqtualizebulia sainformacio

nakadSi. konceptualurad ukve struqturirebuli konstruqciebis Semdgomi

aqtualizaciis procesi swored aseTi funqcionalurad dawinaurebuli elemen-

tebis formalur gamokveTas gulisxmobs. magaliTad, nominatiur enebSi, sadac

agensia konceptualurad aqtualizebuli, SeiZleba paciensi wamoiwios da

iqces gamonaTqvamis subieqtad; anu informacia gadalagdes ise, rom gamona-

Tqvami gvawodebdes garkveul informacias swored rom moqmedebis obieqtis,

paciensis Sesaxeb da centralur, mTavar elementad aqtualizdes ara agensi,

aramed paciensi. nominatiur enebSi aqtiuri konstruqciebi asaxaven agensis

aqtualizebas (resp. agensi asrulebs subieqtis funqcias) maSin, rodesac

pasiuri konstruqciebi warmoaCenen paciensis aqtualizebas (resp. paciensi aris subieqti). formalurad paciensis amgvari dawinaureba gamoixateba misTvis

nominatiuri formis miniWebiT. e.i. paciensi pasiur konstruqciaSi formalu-

rad imave brunviT aris markirebuli, romliTac aqtiur konstruqciaSi kon-

ceptualurad aqtualizebuli saxeli, agensia warmodgenili. (agensi pasiur

konstruqciaSi an sruliad gamqralia, an Tandebuliani formiT dasturdeba.)

diskursuli aqtualizeba

komunikaciisas Cveulebrivia informaciis gacvla-gamocvla: dialoguri formiT

xdeba axali informaciis mopoveba da informaciul nakadSi arsebuli garkve-

uli xarvezebis Sevseba. sainformacio nakadis ucnobi elementebi dgindeba

kiTxva-pasuxebis garkveuli wyvilebiT. kiTxviT moTxovnili informacia

moipoveba pasuxebSi, sadac moTxovnili informacia Cveulebriv aqtualize-

bulia. amgvar, diskursulad gamokveTil, aqtualizebul informacias uwodeben

fokuss. umravles SemTxvevebSi fokuss axasiaTebs specifikuri intonacia,

magram sxvadasxvagvar enebSi arsebobs, agreTve, sxvadasxvagvari formaluri

meqanizmebi fokusis markirebisaTvis.

5

pragmatikuli aqtualizeba

zogierT SemTxvevaSi situaciis asaxvisas saWiro xdeba informaciis dazusteba,

misi pragmatikuli Rirebulebebis gaxazva, movlenebs Soris arsebul (an

ararsebul) kavSirebze miniSneba, informaciis mosalodneloba-moulodnelobis

asaxva, axali-Zveli informaciis warmoCena, informaciaSi implikaciuri an

presupoziciuli mimarTebebis dafiqsireba da sxva amgvari. yoveli aseTi

specifikuri xasiaTis informacia vlindeba teqstis ufro farTo frangmentSi,

romelic metia vidre ubralod gamonaTqvami, winadadeba.

winadadeba, tradiciuli ganmartebiT, gamoxatavs dasrulebul azrs, magram is,

rogorc ufro maRali enobrivi donis, teqstis, Semadgeneli erTeuli,

rogorc wesi, ufro met pragmatikul informaciasac moicavs, vidre ubralod

konkretul leqsemaTa sintaqsuri wesebis mixedviT Sedgenili struqturuli

mTlianoba (resp. winadadeba). informaciis aseTi damatebiTi niuansebi, rogorc wesi, aqtualizebulia. informaciis amgvar pragmatikulad aqtualizebul

nawils ewodeba topiki. topikis markirebis ZiriTadi universaluri meqanizmia

specifikuri intonaciuri konturi, Tumca arsebobs enaTa mixedviT

gansxvavebuli sxvadasxvagvari formaluri modelebic. enebSi, romlebic

orientirebulia sainformacio nakadSi pragmatikulad aqtualizebuli

elementebis konceptualur markirebaze, rogorc wesi, topikis aRmniSvneli

specialuri morfologiuri mawarmoebelic gamoiyofa.

fokusi da topiki winadadebis sainformacio struqturis ganmsazRvreli

ZiriTadi terminebia. orive winadadebis aqtualizebul wevrs aRniSnavs.

gansxvaveba maT Soris is aris, rom fokusi aRniSnavs adre ucnob, axal

informacias, romelic mopovebul iqna Sesabamis kiTxvaze pasuxis amsaxvel

winadadebaSi aqtualizebuli wevris saxiT; xolo topikiT xdeba ukve cnobili

wevris, Zveli informaciis aqtualizeba masze informaciis mopovebis an

arsebuli informaciis gamokveTisa Tu gavrcoba-dazustebis mizniT. magaliTad,

informacia, romelic Seesabameba situacias, romlis drosac monadirem mokla

iremi, da, romelic struqturirdeba qarTuli enis konkretuli TxrobiTi

winadadebis – monadirem mokla iremi – saxiT, srulia da, amave dros,

neitraluri informaciis romelime Semadgeneli elementis aqtualizebis

TvalsazrisiT. Tu situaciis romelime wevri ucnobia, komunikaciisas Cndeba

kiTxva, mag.: vin mokla iremi? pasuxSi: monadirem mokla iremi –

aqtualizebulia monadire, e.i. is aris fokusi. an, vTqvaT: ra mokla monadirem? [iremi]F mokla monadirem. ra qna monadirem? [iremi mokla]F monadirem. ra uqna monadirem irems? [mokla]F monadirem iremi. Tu vsaubrobT monadireze da gvainteresebs mis Sesaxeb raime informaciis

mopoveba, es Sevsebuli informacia reprezentirebul iqneba winadadebiT, sadac monadire kvalificirdeba rogorc topiki:

6

ras ityviT monadiris Sesaxeb? [monadirem]T mokla iremi. formaluri TvalsazrisiT, topiki da fokusi SeiZleba markirebul iqnes

sxvadasxva enobriv doneze: 1. fonetika-fonologiis (intonacia, maxvili, toni, specifikuri fonetikur-

fonologiuri procesebi: gamJRereba-dayrueba, asimilacia-disimilacia da sxv.);

2. morfologia-sintaqsis (specialuri brunva, Serwymuli nawilaki an klitika;

sityvaTa rigis cvla, elifsisi, specifikuri konstruqciebi, frazebis

gagleja, tmesi da sxv.);

3. leqsika-semantikis (specialuri sityvebi, nawilakebi, artikli, kvantoruli

sityvebi: mxolod, marTlac, namdvilad, aseve, kidec da sxv.). yvela zemoT CamoTvlili SesaZlebloba xSirad Tanaarsebobs informaciis

“SefuTvisas” – warmoiqmneba rTuli, kompleqsuri SemTxvevebi da, amdenad,

rTuldeba formaluri analizic. aqtualizaciis ierarqiulad organizebul

rTul procesSi zogierTi informacia aucilebelia da ierarqiis umaRles

safexurzeve ganisazRvreba (mag., konceptualuri an funqcionaluri aqtuali-

zeba), zogi mxolod dialoguri sistemebisaTvisaa damaxasiaTebeli (mag.,

fokusi), zogic nebismieria da konkretuli komunikaciuri funqciiT ganisaz-

Rvreba (mag., topiki); formaluri meqanizmebi xSirad implikaciuri xasiaTisaa

(mag., sityvaTa rigis cvla xSirad intonaciur gamokveTasac iwvevs da garkve-

ul SemTxvevebSi sintaqsuri mimarTebebis cvlasac) da sxv. miuxedavad imisa,

rom es procesebi enaTa mixedviT specifikuria (yoveli ena Tavisi strategiis

mixedviT axorcielebs topik-fokusisa Tu sxva aqtualizebuli wevris forma-

lizebas), universalur maxasiaTeblebsa da niSnebze saubari mainc SesaZ-

lebelia. qarTuli enis monacemebi

qarTulSi naklebad aris Seswavlili winadadebis sainformacio struqturis

formirebisa da aqtualizebis procesebi, gansakuTrebiT, fokusisa da topikis

ganmsazRvreli formaluri modelebi. am TvalsazrisiT qarTuli enobrivi

masalis analizi saSualebas gvaZlevs warmovadginoT Semdegi debulebebi:

1. rogorc cnobilia, qarTulSi gvaqvs nominatiuri da ergatiuli konstruq-

ciebis ganawileba zmnis dro-kiloTa gansxvavebuli formebis mixedviT. amdenad,

konceptualuri aqtualizaciis modelebic gansxvavebulia dro-kiloTa sxva-

dasxvagvari kombinaciebis mixedviT: (a) I-seriis zmnuri formebi qmnian nominatiur konstruqciebs, rac gulisxmobs

agensis aqtualizebas konceptualur doneze. Semdgom safexurze (funqcio-

naluri done) SesaZlebelia paciensis rolis gazrda da aqtualizeba, rac

gulisxmobs pasiuri konstruqciebis formirebas, sadac paciensi iqceva

7

funqcionalurad aqtualizebul wevrad da kvalificirdeba subieqtad. miuxe-

davad amisa, konkretuli kiTxvarebis safuZvelze mopovebuli masala gviCve-

nebs, rom cdispirebi paciensis aqtualizebisaTvis iSviaTad iyeneben pasiur

konstruqciebs da upiratesobas aniWeben sityvaTa rigis cvlas, an ganusaz-

Rvrelnacvalsaxelian Tu ganusazRvrelsubieqtian konstruqciebs, gansakuT-

rebiT im SemTxvevebSi, roca paciensi sulieria. es Tavisebureba qarTuli enis

pasiuri konstruqciebis specifikurobidan gamomdinareobs: qarTulSi pasiuri

konstruqcia yovelTvis ar warmoadgens aqtiuris konversiul formas da

wminda semantikuri safuZveli aqvs. (b) II da III-seriis zmnuri formebi qmnian ergatiul konstruqciebs, rac

gulisxmobs konceptualur doneze paciensis aqtualizebas. ramdenadac paciensi

ukve aqtualizebulia konceptualurad, misi Semdgomi funqcionaluri aqtua-

lizeba analitikuri pasiuri konstruqciebis formirebiT sruliad gamoric-

xulia III-seriis formebisaTvis da sakmaod SezRudulia II-seriisaTvis. am

mizniT upiratesoba eniWeba ‘yofna’ zmniT Sedgenil aRweriT pasivebs.

2. fokusi qarTulSi xasiaTdeba specifikuri, maRali, talRiseburi intonaciiT, rac gansxvavdeba neitraluri, arafokusirebuli informaciisaTvis

damaxasiaTebeli rbilad damavali intonaciisgan. fokusirebuli wevri

ZiriTadad Tavsdeba winadadebis TavSi zmnis wina poziciaSi, rac sityvaTa

aramarkirebuli, neitraluri rigis gadalagebas gulisxmobs. ramdenadac

sityvaTa rigi qarTulSi sazogadod Tavisufalia, enobriv masalaSi SeiZleba

dadasturebul iqnes nebismieri wyoba. savaraudoa, rom yovel gadalagebas

Tavisi specifikuri pragmatikuli Sinaarsi aqvs, Tumca sakmaod rTuldeba

yvela niuansis garkveva da xSirad amgvari gadalagebebi subieqturi xasiaTisaa,

Tumca, Tu gadalagebas axlavs intonaciis markireba, viTareba Sinaarsebis

dadgenis TvalsazrisiT martivdeba. yovel SemTxvevaSi, gaanalizebuli masala

iZleva safuZvels davaskvnaT: fokusis gamoxatvis saukeTeso, naTeli saSua-

lebaa fokusirebuli wevris winadadebis TavSi zmnis wina poziciaSi moTavseba

saTanado intonaciuri xazgasmiT. fokusis formalizaciisaTvis pasiuri kon-

struqcia araefeqturi meqanizmia, ramdenadac pasuxi (sadac warmodgenilia

fokusi) ZiriTadad imeorebs kiTxvis konstruqcias: Tu kiTxva reprezentire-

bulia aqtiuri konstruqciiT, pasuxic, rogorc wesi, aqtiuri konstruqciis

saxiT formdeba da piriqiT, pasiuri konstruqciis saxiT gaformebuli kiTxva

pasiur konstruqciis saxiT gaformebul pasuxs gulisxmobs. fokusis nair-

saxeobebis sakvlevad gansakuTrebiT efeqturi Cans ganmeorebiTi kiTxvebi

(CakiTxva), sadac Cndeba sxvadasxvagvari nawilakebi da specifikuri sintaqsuri

konstruqciebi. magaliTebi:

ra xdeba? monadire klavs irems vin klavs irems? monadire1 klavs irems

1 magaliTebSi ‘Savi’ (boldi) aRniSnavs intonaciur gamokveTas.

8

ras klavs monadire? irems klavs monadire ras akeTebs monadire? monadire klavs irems klavs monadire irems vin ras klavs? monadire irems klavs monadire klavs irems irems monadire klavs irems klavs monadire monadire klavs irems? diax, monadire klavs irems diaxac, monadire klavs irems diax, es monadire-a (is), vinc klavs irems namdvilad (rom) monadire klavs irems swored (rom) monadire klavs irems marTlac (rom) monadire klavs irems marTlacda monadire klavs irems irems klavs monadire? diax, irems klavs monadire diaxac, irems klavs monadire diax, es iremi-a (is), visac monadire klavs namdvilad irems klavs monadire swored (rom) irems klavs monadire marTlac (rom) irems klavs monadire marTlacda irems klavs monadire monadire klavs irems? diax, klavs monadire irems diaxac, klavs monadire irems klavs monadire irems, aba ara? namdvilad klavs monadire irems swored (rom) klavs monadire irems marTlac (rom) klavs monadire irems marTlacda klavs monadire irems da a.S. Tu warmodgenil masalas SevajamebT, fokusis markirebis Semdegi modelebi

SeiZleba gamovyoT:

(a) markirebuli intonacia;

(b) sityvaTa rigis Secvla (fokusis win gadasma) (+intonacia);

(g) specifikuri sintaqsuri konstruqciebi (+intonacia);

(d) nawilakebi (+sintaqsi+intonacia).

3. topikis ZiriTadi formaluri niSania specifikuri intonaciuri konturi,

romelic gansxvavdeba fokusirebuli da neitraluri frazis intonaciisgan.

9

misTvis damaxasiaTebelia aRmavali (magram ara imdenad, rogoric gvaqvs

kiTxviT konstruqciebSi) intonacia frazis sasazRvro dabali intonaciiT

(LH*Lp). yvela sxva modeli, romelic dadasturda fokusisaTvis, SeiZleba

gamoyenebul iqnes topikisTvisac, mxolod topikisaTvis damaxasiaTebel into-

naciur konturTan erTad:

ras ityviT monadiris Sesaxeb? monadirem mokla iremi iciT rame irmis Sesaxeb? iremi mokla monadirem iremi moikla monadiris mier ras ityviT am monadiris Sesaxeb? am monadirem mokla iremi ai am monadirem mokla iremi namdvilad am monadirem mokla iremi swored (rom) am monadirem mokla iremi marTlac (rom) am monadirem mokla iremi marTlacda am monadirem mokla iremi es is monadirea, vinc iremi mokla iciT rame am irmis Sesaxeb? es iremi mokla monadirem ai es iremi mokla monadirem swored rom es iremi mokla monadirem marTlac (rom) es iremi mokla monadirem da a.S. (amave tipis konstruqciebi dasturdeba Sesabamisi pasiuri

konstruqciebisaTvis.

Tu SevajamebT warmodgenil masalas, topikis gamosaxatavad Semdegi modelebi

SeiZleba gamovyoT:

(a) markirebuli intonacia (gansxvavebuli fokusirebuli da

neitraluri intonaciisgan);

(b) sityvaTa rigis Secvla (topikis win gadasma da/an zmnis wina

poziciaSi dasma) (+intonacia);

(g) specifikuri sintaqsuri konstruqciebi (+intonacia);

(d) nawilakebi (+sintaqsi+intonacia).

Tu SevadarebT topikisa da fokusis modelebs, davinaxavT, rom gansxvaveba ZiriTadad

(Tu ar CavTvliT ai-nawilakian konstruqciebs, romlebic gvxvdeba implikaciuri

topikis gamosaxatavad da ar dasturdeba fokusis SemTxvevaSi) intonaciur

konturSia: fokuss axasiaTebs maRali talRiseburi intonacia, topiks ki LH*Lp. qvemoT warmodgenilia tipiuri magaliTebis fonetikuri Canawerebi:

10

(1) neitraluri TxrobiTi winadadeba

(2) TxrobiTi winadadeba topikalizebuli wevriT

meri–m i–qi’d–a bevr–i sk’am–i.

meri–ERG . SV–buy–AOR.Subj.3.Sg many–Nom. chair–Nom.Hp

L Hp L Hp L L Li

100

350

150

200

250

300

Time (s)0 2.06961

meri–m sk’am–i. ip’ova gushin sam–i

meri–ERG chair–Nom. found yesterdayL three–Nom

L H* L L* H* Lp L Hp L Hp L Li

100

400

200

300

Time (s)0 2.82941

11

(3)-(4) TxrobiTi winadadebebi fokusiT

.

4. sazogadod, gaanalizebuli masalis safuZvelze aSkarad ikveTeba intonaciis

gadamwyveti roli aqtualizebuli saxelis markirebaSi: intonacia ganasxvavebs

erTmaneTisgan neitralur, fokusirebul da topikalizebul nawils informa-

ciisa; misi saSualebiT xorcieldeba informaciis struqturireba; misi sxvada-

sxvagvari konturi ganapirobebs informaciis daSlas pragmatikulad Rirebul,

sxvadasxvagvarad Seferil nawilebad. intonacia, rogorc aqtualizebuli

saxelis markeri, gvexmareba sxvadasxva saxis omonimuri formebis gaxsnasa da

informaciis calsaxa interpretaciaSi. magaliTad, winadadeba: maRali gamyidvlis daxlTan dgas – omonimuria. am omonimiis moxsna SesaZlebelia

winadadebis sainformacio struqturis dadgeniT, risTvisac aucilebelia

ufro farTo teqstualuri konteqstis ganxilva. Tu es winadadeba aris pasu-

xi SekiTxvaze: sad dgas maRali? – maSin maRali gamyidveli iqneba fokusi da maRali, talRiseburi intonaciiT warmoiTqmis, xolo winadadeba iqneba uqvem-

debaro: [maRali gamyidvlis]FF daxlTan dgas; Tuki maRali aris

pragmatikulad gamoyofili wevri, romelic gulisxmobs, rom vinme maRali da ara, vTqvaT, dabali, dgas daxlTan, maSin igi warmoiTqmis aRmavali

intonaciiT, sadac intonaciuri maxvili meore marcvalzea, da topikad

kvalificirdeba: [maRali]T gamyidvlis daxlTan dgas. ase rom, winadadebis

waakiTxebs daTo wigns ninos

causes–to–read DATO book Nino

100

400

200

300

Time (s)0 3.20438

waakiTxebs dato ts’ign–s ninos

causes–to–read Dato BOOK Nino

100

400

200

300

Time (s)0 3.43864

12

sainformacio struqtura gamoricxavs omonimurobas da komunikaciur

orazrovnebas.

5. cdispirebis mier formirebuli masalis mixedviT topikalizaciis erT-erTi

saSualebaa agreTve sintaqsurad Warbi nacvalsaxelebis dafiqsireba; mag.: [me]T wavedi, [me]T viyide pamidvrebi, [me]T movutane dedas pamidvrebi... aq me Warbia (qarTulis enis specifikidan gamomdinare: ramdenadac pirveli piri morfo-

logiurad warmodgenilia zmnur formaSi pirveli subieqturi piris markeriT (v-), misi dublireba sintaqsur konstruqciaSi me nacvalsaxelis saxiT Warbia

da migvaniSnebs pirveli piris topikalurobaze: me da ara sxva vinmem, swored me da ara sxva vinmem, namdvilad me da ara sxva vinmem moutana dedas pamidvrebi. aseve, sintaqsurad Warbia refleqsuri nacvalsaxeli TavisTvis winadadebaSi ninom TavisTvis CainiSna davalebebi, ramdenadac zmnis saTaviso qcevis forma ukve migvaniSnebs, rom subieqti TavisTvis moqmedebs da misi

dafiqsireba sintaqsur konstruqciaSi gvawvdis damatebiT informacias imis

Sesaxeb, rom ninom [TavisTvis]T (da ara, vTqvaT, Tavis ZmisTvis) CainiSna davalebebi.

samomavlo perspeqtivebi

warmodgenili analizi, rasakvirvelia, srulad ver asaxavs qarTul enaSi

arsebul yvela specifikur models, romelic ki gamoiyeneba aqtualizaciis

sxvadasxva niuansebis gadmosacemad, magram, vfiqrobT, Semoxazavs im ZiriTadi

problemebis gadawyvetisa da modelebis moZiebis gzebs, romelic aucilebelia

qarTulSi informaciis struqturirebis meqanizmebis sruli suraTis

warmodgenisaTvis. amasTanave, sruliad aSkara xdeba am procesebSi intonaciis

gadamwyveti roli.

literatura asaTiani r., 2000, subieqtisa da obieqtis aqtualizaciis ZiriTadi

morfosintaqsuri saSualebebi arabulSi, tipologiuri Ziebani – IV, Tbilisi,

gv.77-87. 2002, aqtualizaciis ZiriTadi tipologiurad gansxvavebuli

modelebi, “aRmosavleTmcodneoba”, iv. javaxiSvilis saxelobis Tbilisis

saxelmwifo universitetis Sromebi, 341, Tbilisi, gv.19-26. Büring D., 1995, Topic, Cologne University, el-publ. Fery C., 1993, German Intonational Patterns, No. 285 in Linguistische Arbeiten, Tübingen: Max Niemeyer. van Hoof H., 1997, On Split Topicalization and Elipsis, Tech.Reo. 112, Arbeutspapiere des Sonderforschungsbereichs 340, Tübingen.

13

Rusudan Asatiani

Role of Intonation in the Formation of Sentence Information Structure

Abstract In the paper, it is shown that sentence prosody and intonation play an important role in the process of information structuring and some conclusions are suggested: (1) While speaking about prosody it seems better to speak about tendencies than about strict rules; (2) Canonical pitch tracks for unmarked affirmative sentences are with an unmarked word order (SOV/SV): LHp LHp …LLi; (3) The whole sentence is uttered in a single intonation phrase (i-phrase), and every p-phrase is downstepped relatively to the preceding one, that is, the high part of a p-phrase is slightly lower than that of the preceding p-phrase. So, the intonation contour of a whole affirmative sentence is falling; (4) A prosodic phrase has at least one pitch accent; (5) The intonation of wh-sentences can be falling or rising: These are two free alternates; (6) In Yes/No questions intonation is always verb-rising. Tag-questions are characterized by finally-rising intonation; (7) Focus mostly appears in preverbal position and it usually builds one p-phrase with the verb. The p-phrase has (L*)HL intonation followed by Hp (if the p-phrase is non-final) or Lp. Consequently, focus mostly is characterized by rising/high (or wave-like-rising when there are polysyllabic words in p-phrase) intonation; (8) If focus is sentence final intonation emphasis is fulfilled by L*; (9) For contrast or selection rising/high intonation is usual, with the highest pitch contrast (H* or L*) on conjunctions: tu ‘or’ (in selective questions) or k’i ‘and.but’ (in contrastive answers); (10) Any kind of topics are characterized by LH*Lp; The first or the second syllable of topic (according to the canonical position of stress) has a dynamic stress and the next, followed syllable is characterized by H*; (11) In elliptic, short answers focus has neutral or topic intonation; in repeated questions focus mostly has topic intonation; (12) Only intonation changes (without reordering or other syntactic devices) can form various information structures.

14

rusudan asaTiani

informaciis struqturirebis sintaqsuri modelebi qarTulSi

1. Sesavali: sakiTxis dasma

winadadeba, sazogadod, ufro met informacias moicavs da gadmogvcems, vidre

masSi Semaval sityvaTa mniSvnelobebisa da sintaqsuri mimarTebebis

struqturul-semantikuri mTlianoba iZleva. winadadebaSi informaciis struq-

turirebis sxvadasxvagvari saSualebebi cvlis winadadebis pragmatikul

Sinaarss da implikaciurad ufro farTo konteqstur Sinaarsebsac moicavs. am

fenomenis sakvlevad Tanamedrove enaTmecnierebaSi SemuSavda garkveuli Teori-

uli safuZvlebi da meTodologia; ganisazRvra ZiriTadi cnebebi, romelTa

saSualebiT, met-naklebi warmatebiT, SesaZlebelia winadadebis sainformacio

struqturis kvleva.

2. Teoriuli midgoma

informaciis struqturireba, misi SefuTva, xorcieldeba opoziciis safuZ-

velze, sadac informaciis erTi nawili gamoikveTeba informaciis meore

nawilisgan. sakomunikacio, pragmatikuli TvalsazrisiT, es gamokveTili

nawili aris aqtualizebuli, xazgasmuli, mniSvnelovani da warmogvidgens

sainformacio nakadSi dawinaurebul informacias. nebismieri saxis

‘dawinaureba’ (igive: xazgasma, gaSuqeba, fokusSi moqceva, logikuri gamokveTa,

win wamoweva, TvalTaxedvis centrSi moqceva da sxva amgvari) SeiZleba

ganvixiloT rogorc erTi, saerTo fenomeni (asaTiani, 2007), romelic asaxavs

informaciis lingvisturi struqturirebis mTavar strategias. aqtualizeba

sxvadasxva struqturis mqone enebSi gansxvavebuli formaluri saSualebebiT

xorcieldeba da nebismier enobriv doneze SeiZleba iyos realizebuli.

2.1. konceptualuri aqtualizeba


ZiriTadad sami mTavari konceptualuri ganzomilebis mixedviT mimdinareobs

(Kibrik, 1997): 1. situaciis Semadgeneli semantikuri rolebis (agensi-paciensi-adresati)

markireba;

2. sainformacio nakadis pragmatikuli Rirebulebebis (topiki-komentari)

markireba;

3. sakomunikacio aqtis statusebis (molaparake/msmeneli-saubris sagani)

markireba.



15

semantikuri rolebis markirebaze orientirebul enebSi informaciis

TvalsazrisiT aqtualizebul saxelad konceptualizdeba an agensi an paciensi,

ris Sedegadac warmoiqmneba nominatiuri an ergatiuli konstruqciebi. pirveli

afiqsirebs informaciis im nawilis aqtualizebas, romelic gamoxatavs, Tu vin moqmedebs, meore ki informaciis im nawilis aqtualizebas, romelic gamo-

xatavs, Tu ra xdeba, ra iqmneba. konceptualurad aqtualizebuli saxeli

enobriv struqturebSi formalurad warmodgenilia aramarkirebuli, saxe-

lobiTi brunviT: nominatiur konstruqciebSi es aris agensi, xolo ergatiul

konstruqciebSi – paciensi. enebi, romlebic, upirveles yovlisa, sainformacio nakadis pragmatul kon-

ceptualizacias axdenen, formaluri struqturebiT erTmaneTs upirispireben

informaciis aqtualizebul:fonur monakveTebs. aseT enebSi, rogorc wesi,

aqtualizebuli saxeli markirebulia specialuri morfologiuri afiqsiT.

rac Seexeba mesame tipis enebs, aq informaciis aqtualur wevrad koncep-

tualizdeba sakomunikacio aqtSi monawile pirebi da gramatikulad gamoixa-

teba piris I/II : III diqotomia.

2.2. funqcionaluri aqtualizeba

funqcionalurad aqtualizebuli saxeli gramatikul tradiciaSi aRiniSneba






agensia konceptualurad aqtualizebuli, SeiZleba paciensi wamoiwios da

iqces gamonaTqvamis subieqtad; anu informacia gadalagdes ise, rom gamona-

Tqvami gvawodebdes garkveul informacias swored rom moqmedebis obieqtis,

paciensis Sesaxeb da centralur, mTavar elementad aqtualizdes ara agensi,

aramed paciensi. nominatiur enebSi aqtiuri konstruqciebi asaxaven agensis

aqtualizebas (resp. agensi asrulebs subieqtis funqcias) maSin, rodesac

pasiuri konstruqciebi warmoaCenen paciensis aqtualizebas (resp. paciensi aris subieqti). formalurad paciensis amgvari dawinaureba gamoixateba misTvis

nominatiuri formis miniWebiT. e.i. paciensi pasiur konstruqciaSi forma-

lurad imave brunviT aris markirebuli, romliTac aqtiur konstruqciaSi

konceptualurad aqtualizebuli saxeli, agensia warmodgenili (agensi pasiur

konstruqciaSi an sruliad gamqralia, an Tandebuliani formiT dasturdeba).

2.3. diskursuli aqtualizeba

komunikaciisas Cveulebrivia informaciis gacvla-gamocvla: dialoguri formiT

xdeba axali informaciis mopoveba da informaciul nakadSi arsebuli gar-

kveuli xarvezebis Sevseba. sainformacio nakadis ucnobi elementebi dgindeba

kiTxva-pasuxebis garkveuli wyvilebiT. kiTxviT moTxovnili informacia moi-

16

poveba pasuxebSi, sadac moTxovnili informacia Cveulebriv aqtualizebulia.

amgvar, diskursulad gamokveTil, aqtualizebul informacias uwodeben

fokuss. umravles SemTxvevebSi fokuss axasiaTebs specifikuri intonacia,

magram sxvadasxvagvar enebSi arsebobs, agreTve, sxvadasxvagvari formaluri

meqanizmebi fokusis markirebisaTvis. 2.4. pragmatikuli aqtualizeba

zogierT SemTxvevaSi situaciis asaxvisas saWiro xdeba informaciis dazusteba,

misi pragmatikuli Rirebulebebis gaxazva, movlenebs Soris arsebul (an

ararsebul) kavSirebze miniSneba, informaciis mosalodneloba-moulodnelobis

asaxva, axali-Zveli informaciis warmoCena, informaciaSi implikaciuri an presupoziciuli mimarTebebis dafiqsireba da sxva amgvari. yoveli aseTi speci-

fikuri xasiaTis informacia vlindeba teqstis ufro farTo frangmentSi,

romelic metia vidre ubralod gamonaTqvami, winadadeba.

winadadeba, tradiciuli ganmartebiT, gamoxatavs dasrulebul azrs, magram is,

rogorc ufro maRali enobrivi donis, teqstis, Semadgeneli erTeuli,

rogorc wesi, ufro met pragmatikul informaciasac moicavs. informaciis

aseTi damatebiTi niuansebi, rogorc wesi, winadadebis aqtualizebuli wevris

Sesaxeb grovdeba. informaciis swored amgvar, pragmatikulad aqtualizebul,

nawils ewodeba topiki. topikis markirebis ZiriTadi universaluri meqanizmia

specifikuri intonaciuri konturi, Tumca arsebobs enaTa mixedviT gansxva-

vebuli sxvadasxvagvari formaluri modelebic. enebSi, romlebic orien-

tirebulia sainformacio nakadSi pragmatikulad aqtualizebuli elementebis

konceptualur markirebaze, rogorc wesi, topikis aRmniSvneli specialuri

morfologiuri mawarmoebelic gamoiyofa.


ZiriTadi cnebebia. orive winadadebis aqtualizebul wevrs aRniSnavs.

gansxvaveba maT Soris is aris, rom fokusi aRniSnavs adre ucnob, axal

informacias, romelic mopovebul iqna Sesabamis kiTxvaze pasuxis amsaxvel

winadadebaSi aqtualizebuli wevris saxiT; xolo topikiT xdeba ukve cnobili

wevris, Zveli informaciis aqtualizeba masze informaciis mopovebis an

arsebuli informaciis gamokveTisa Tu gavrcoba-dazustebis mizniT. magaliTad,

informacia, romelic Seesabameba situacias, romlis drosac monadirem mokla

iremi, da romelic, magaliTad qarTulSi, struqturirdeba konkretuli

TxrobiTi winadadebis – monadirem mokla iremi – saxiT, srulia da, amave

dros, neitraluri informaciis romelime Semadgeneli elementis aqtualizebis

TvalsazrisiT. Tu situaciis romelime wevri ucnobia, komunikaciisas Cndeba

kiTxva, mag.: vin mokla iremi? pasuxSi: monadirem mokla iremi –

aqtualizebulia monadire, e.i. is aris fokusi; an, vTqvaT:

ra mokla monadirem? [iremi]F mokla monadirem. ra qna monadirem? [iremi mokla]F monadirem.

17

ra uqna monadirem irems? [mokla]F monadirem iremi.

Tu vsaubrobT monadireze da gvainteresebs mis Sesaxeb raime informaciis

mopoveba, es Sevsebuli informacia reprezentirebul iqneba winadadebiT, sadac monadire kvalificirdeba rogorc topiki:

ras ityviT monadiris Sesaxeb? [monadirem]T mokla iremi. formaluri TvalsazrisiT, topiki da fokusi SeiZleba markirebul iqnes

sxvadasxva enobriv doneze: 1. fonetika-fonologiis (intonacia, maxvili, toni, specifikuri fonetikur-

fonologiuri procesebi: gamJRereba-dayrueba, asimilacia-disimilacia da sxv.);

2. morfologia-sintaqsis (specialuri brunva, Serwymuli nawilaki an klitika;

sityvaTa rigis cvla, elifsisi, specifikuri konstruqciebi, frazebis gax-

leCa, tmesi da sxv.);

3. leqsika-semantikis (specialuri sityvebi, nawilakebi, artikli, kvantoruli

sityvebi: mxolod, marTlac, namdvilad, aseve, kidec da sxv.).

yvela zemoT CamoTvlili SesaZlebloba xSirad Tanaarsebobs informaciis

“SefuTvisas” – warmoiqmneba rTuli, kompleqsuri SemTxvevebi da, amdenad,

rTuldeba formaluri analizic. aqtualizaciis ierarqiulad organizebul

rTul procesSi zogierTi informacia aucilebelia da ierarqiis umaRles

safexurzeve ganisazRvreba (mag., konceptualuri an funqcionaluri

aqtualizeba), zogi mxolod dialoguri sistemebisaTvisaa damaxasiaTebeli

(mag., fokusi), zogic nebismieria da konkretuli komunikaciuri funqciiT

ganisazRvreba (mag., topiki); formaluri meqanizmebi xSirad implikaciuri

xasiaTisaa (mag., sityvaTa rigis cvla xSirad intonaciur gamokveTasac iwvevs

da garkveul SemTxvevebSi sintaqsuri mimarTebebis cvlasac) da sxv.

miuxedavad imisa, rom es procesebi enaTa mixedviT specifikuria (yoveli ena

Tavisi strategiis mixedviT axorcielebs topik-fokusisa Tu sxva aqtu-

alizebuli wevris formalizebas), universalur maxasiaTeblebsa da niSnebze

saubari mainc SesaZlebelia.

3. meTodologia informaciis struqturirebis formaluri modelebis moZiebisaTvis gansakuT-

rebiT Rirebulia bunebriv sametyvelo situaciebSi warmoqmnili winadadebebi.

aseTi situaciebis stimulireba SesaZlebelia garkveuli saeqsperimento amoca-

nebis SerCeviT. am amocanebis gadawyvetaze dafuZnebuli eqsperimentebis Sede-

gad mopovebuli masala1 warmoadgens spontanurad warmoqmnili sametyvelo

teqstebis korpuss, romlis analizsac eyrdnoba winamdebare kvleva. 1 eqsperimentebSi monawileobda 10 cdispiri – Tbilisis saxelmwifo universitetis

aTi studenti (gogonebi da biWebi).

18

4. informaciis struqturirebis morfosintaqsuri modelebi qarTulSi qarTulSi naklebad aris Seswavlili winadadebis sainformacio struqturebis

formirebisa da aqtualizebis procesebi, gansakuTrebiT, fokusisa da topikis

ganmsazRvreli formaluri modelebi. masalis analizidan cxadia, rom qarTulSi ara gvaqvs topikis an fokusis

morfologiuri mawarmoeblebi, Tumca SeiZleba gamovyoT garkveuli morfo-

sintaqsuri da sintaqsuri konstruqciebi, romlebic gamoxataven aqtualizebis

sxvadasxvagvar modelebs. am procesebSi gansakutrebul rols TamaSobs

intonacia (asaTiani, 2006). 4.1. ergatiuli konstruqcia rogorc paciensis konceptualuri aqtualizebis

saSualeba qarTuli ar aris Tanmimdevrulad ergatiuli ena da gviCvenebs seriaTa

mixedviT gaxleCil ergatiulobas Dixon, 1979): I seria nominatiuria (e.i. aq agensia aqtualizebuli); II seria saxelur argumentTa mixedviT ergatiulia

(e.i. aq paciensia aqtualizebuli), zmnuri piris niSnebis mixedviT ki

nominatiuri (e.i. aq agensia aqtualizebuli); III seria micemiTbrunviani

subieqtiT – ergatiulia (e.i. aq paciensia aqtualizebuli). qarTulis es

monacemebi gvibiZgebs davaskvnaT, rom qarTulSi konceptualur aqtualizebas

ganapirobebs dro-kiloTa garkveuli semantikuri jgufebi: awmyosa da

imperfeqtis SemTxvevaSi kognitiurad wina planze iwevs subieqti da swored

is aqtualizdeba centralur, aramarkirebul erTeulad, maSin rodesac

perfeqtisaTvis informaciulad meti Rirebulebisaa paciensi da swored is

aqtualizdeba centralur, aramarkirebul saxelad. amgvari konceptualuri

aqtualizebis kognitiuri safuZvlebi met-naklebad naTelia: situaciis

aRqmisas awmyoSi swored rom subieqtia wina planze, paciensi ki, moqmedebis

msvlelobis gamo, faqtobrivad ganusazRvrelia, maSin rodesac perfeqtSi swored rom moqmedebis Sedegia saxeze (resp. paciensi), subieqti ki, mxolod

warsulis gamocdilebidan gamomdinare, SeiZleba iyos (wyvetili) an sulac ar

iyos (III seriis TurmeobiTebi) cnobili. amgvarad, ergatiuli konstruqcia gviCvenebs paciensis aqtualizebas,

nominatiuri ki agensis (ufro zustad, subieqtis). amgvar daskvnas mxars

uWers l. enuqiZis kvlevac (enuqiZe, 1981: 109-110). avtori garkveuli

teqstebis statistikur analizze dayrdnobiT askvnis: “statistikuri analizi

gviCvenebs, rom ergativis elifsi ufro xSiria, vidre nominativis” (Cven mier

mopovebuli sametyvelo teqstebis korpusis monacemebic adasturebs am

statistikas), ramdenadac aqtualizebuli saxelis elifsisi araa mosa-

lodneli (faqtiurad gamoricxulia), ergativis elifsisisken swrafvis

tendencia SeiZleba miviCnioT warmodgenili interpretaciis erT-erT Zlier

argumentad: ergatiul konstruqciaSi topikalizebulia ara ergativSi mdgari

agensi, aramed nominativSi mdgari paciensi.

19

4.2. specifikuri morfosintaqsuri konstruqciebi: nawilakebis funqcia

sainformacio struqturebis formirebaSi sazogadod, kargad aris cnobili nawilakebis roli winadadebis sainformacio

struqturis formirebaSi, Tumca qarTul specialur literaturaSi TiTqmis

araferia naTqvami nawilakebis am funqciaze. nawilakebi gaxazaven, gamokveTen

informaciis garkveul nawils; amasTanave, isini specifikur semantikur

elfersac mateben winadadebas da garkveuli TvalsazrisiT leqsikuri Sina-

arsiTac arian datvirTulni; amitomac maTi interpretacia topikis calsaxa

morfologiur markerebad ver CaiTvleba adekvaturad. nawilakebi, rogorc

wesi, qmnian specifikur morfosintaqsur konstruqciebs. magaliTisaTvis

SeiZleba ganvixiloT implikaciuri topikze mimaniSnebeli nawilakebi: ai, mxolod, marto, swored (rom), arc – arc, -c, -c ki, kidec. (wereTeli,

2008). 4.2.1. ai nawilaki cdispirebis mier gamoiyeneboda iseTi implikaciuri topikis

gamosaxatavad, romelic damatebiT, ZiriTadad sapirispiro informacias

gvawvdida rogorc zogadi (1), aseve konkretuli situaciis (2) meore,

nagulisxmev wevrze: (1) ai qalebi ar ewevian. (igulisxmeba, rom kacebi ZiriTadad ewevian) (2) ai biWs qudi ar axuravs. (situaciis mixedviT gogs qudi axuravs)

implikaciuri topikis gadmosacemad yvelaze xSirad swored es nawilaki

dafiqsirda. sazogadod, sainteresoa, ai nawilakis zogadi funqcia:

‘miTiTeba+substantivizacia+predikacia’, rac mas a. veJbickas mier navaraudebi

erT-erTi primitivis (demonstrativis, substantivisa da predikaciis

elementebiT) Sesabamis enobriv relizaciad warmogvidgens (Wierzbichka, 1996). swored am funqciebis gamo, is bavSvTa metyvelebis adreuli formebis

aucilebel Semadgenels warmoadgens, rac aseve TanxmobaSia a. veJbickas mier

‘winaswarmetyvelurad aRdgenil’ (ramdenadac inglisurSi msgavsi, sami

funqciis gamaerTianebeli, universaluri semantikuri primitivis enobrivi

realizacia ar dasturdeba, veJbicka ki primitivebis gamoyofisas ZiriTadad

inglisur enas iSveliebs) primitivTan. 4.2.2. mxolod nawilaki sixSiriT meorea implikaciuri topikis markirebis

procesSi. is gulisxmobs, rom topikTa SesaZlebel variantTagan erTi da

mxolod erTia dasaSvebi situaciisda Sesabamisad. es nawilaki mxolod aseT

konkretul SemTxvevebSia misaRebi: (5) mxolod biWia navSi. (igulisxmeba, rom gogo ar aris navSi)

20

4.2.3. EerT-erTma cdispirma mxolod nawilakis gamoyenebas marto nawilaki

amjobina. am nawilaks, rogorc Cans, igive funqciebi aqvs, rac mxolod nawilaks _ alternativebidan SearCios erTaderTi: (6) ara, biWi marto ar aris navSi. (igulisxmeba, rom sxvebic arian) Tumca maTi funqciebi msgavsia, isini Tavisuflad ar enacvlebian erTmaneTs:

marto uSvebs postpoziciur wyobas, mxolod ki ara. magaliTad, (6)

winadadebaSi marto-s mxolod-iT Canacvleba ver gvaZlevs gramatikulad

gamarTul winadadebas. 4.2.4. aseve erT SemTxvevaSi dafiqsirda ar+c nawilaki: 7) arc biWs ar axuravs qudi. (e.i. viRac sxvasac ar axuravs) am rTuli, nawarmoebi nawilakis funqciaa SesaZlebel alternatiul

topikTagan uaryofiTi niSniT gamoarCios erT-erTi imgvarad, rom

implikaciurad meore wevris igive mdgomareoba daafiqsiros. es nawilaki

informantma gamoiyena kiTxvaSi mocemuli uaryofiTi mniSvnelobis dasadas-

tureblad imgvarad, rom iplikaciurad meore wevris Sesaxebac msgavsi,

uaryofiTi informacia mogvawoda. 4.2.5. swored nawilaki dafiqsirda im SemTxvevaSi, rodesac informanti

implikaciurad adasturebda kiTxvaSi mocemuli informaciis sapirispiro

informacias situaciis meore wevris Sesaxeb. saiteresoa, rom ‘uaryofiTi

dadasturebis’ gasaZliereblad man gamoiyena kompleqsuri fraza swored rom. (8) swored rom kacs ar axuravs qudi. (e.i. qals axuravs) gaZlierebis mizezi SeiZleba iyos zogadi warmodgena (gansakuTrebiT Zlieri

qarTuli tradiciis farglebSi), rom, sazogadod, kacs ufro unda exuros qudi, vidre qals. swored rom frazis gamoyenebiT cdispirma gaxaza mis zogad warmodgenasTan Seusabamoba da eqsperimentis farglebSi SemoTavazebuli

konkretuli situaciis moulodneloba.

4.2.6. -c nawilaki Segvxvda, aseve, informaciis (mxolod dadebiTis)

dadasturebisas:

(9) qalic kiTxulobs wigns. (e.i. kidev viRac sxva kiTxulobs wigns)

21

rogorc Cans, swored da -c nawilakebis funqcia aris ‘daTanxmeba,

dadastureba’ situaciis meore wevris Sesaxeb damatebiTi (dadebiTi an

uaryofiTi) implikaciuri informaciis gadmocemis mizniT.

4.2.7. aRsaniSnavia isic, rom nawilakebi, rogorc mosalodnelic iyo, ar

dadasturda martivi implikaciuri topikis SemTxvevebSi:

(10) ara, biWs ar axuravs qudi. (11) ki, biWi navSia. (12) ara, qali ar kiTxulobs wigns. 5. informaciis struqturirebis sintaqsuri saSualebebi qarTulSi

5.1. funqcionaluri aqtualizeba: pasiuri konstruqciebi

qarTulSi pasiuri konstruqciebis saSualebiT paciensis aqtualizebis

SesaZlebloba SezRudulia dro-kiloTa seriebis mixedviT:

(1) I-seriis zmnuri formebi qmnian nominatiur konstruqciebs, rac gulisxmobs

agensis aqtualizebas konceptualur doneze. Semdgom safexurze (funqciona-

luri done) SesaZlebelia paciensis rolis gazrda, misi aqtualizeba, rac

xorcieldeba pasiuri konstruqciebis formirebiT, sadac paciensi iqceva funq-

cionalurad aqtualizebul wevrad da kvalificirdeba subieqtad.

(2) II da III-seriis zmnuri formebi qmnian ergatiul konstruqciebs, rac

gulisxmobs konceptualur doneze paciensis aqtualizebas. ramdenadac paciensi

ukve aqtualizebulia konceptualurad, misi Semdgomi funqcionaluri aqtua-

lizeba sinTezuri pasiuri konstruqciebis formirebiT sruliad gamoric-

xulia III-seriis formebisaTvis da sakmaod SezRudulia II-seriisaTvis. am

mizniT upiratesoba eniWeba ‘yofna’ zmniT Sedgenil aRweriT pasivebs.

miuxedavad imisa, rom pasiuri konstruqciebis formireba ar aris SezRuduli

I seriaSi (da nawilobriv II-Sic), konkretuli eqsperimentebis safuZvelze

mopovebuli masala gviCvenebs, rom cdispirebi paciensis aqtualizebisaTvis

iSviaTad iyeneben pasiur konstruqciebs da upiratesobas aniWeben sityvaTa

rigis cvlas, an ganusazRvrelnacvalsaxelian Tu ganusazRvrelsubieqtian kon-

struqciebs, gansakuTrebiT im SemTxvevebSi, roca paciensi sulieria. es Tavi-

sebureba qarTuli enis pasiuri konstruqciebis specifikurobidan gamom-

dinareobs. (pasiuri konstruqciebis Taviseburebebis analizisaTvis informaciis

struqturirebis TvalsazrisiT, rac saSualebas iZleva vnebiTi gvaris

axleburi semantikur-kognitiuri interpretaciisa (ix. statia amave krebulSi

r. asaTiani, ~vnebiTi gvaris konstruqciaTa pragmatikuli Tavisebureba qarTve-

lur enebSi” _ gv. 30-44).

22

5.2. sityvaTa rigi

5.2.1. sityvaTa rigi: qarTuli gramatikuli tradicia rogorc cnobilia, qarTuls aqvs metad ganviTarebuli morfologia mdidari

afiqsaciiT. amdenad, sityvaTa rigis simkacre qarTulSi ar aris aucilebeli

piroba sintaqsuri mimarTebebis gamosaxatavad, rogorc es gvaqvs enebSi Raribi

afiqsaciiT. swored mdidari afiqsaciis, kerZod brunvis niSnebis funqciuri

gansazRvrulobis gamo, qarTulSi sityvaTa rigi Tavisufalia da sityvaTa

faqtiurad nebismieri gadalageba dasaSvebia (SaniZe 1948; foCxua 1962;

Ciqobava 1968; daviTiani 1973; afridoniZe 1986; kvaWaZe 1996; da sxv.) Tumca

bevri mkvlevari aRniSnavs, rom sityvaTa rigi ar aris absoluturad

Tavisufali, magaliTad: Tanamedrove qarTulSi zedsarTavi Cveulebriv uswrebs

arsebiTs; aseve, zmnizedebi da damatebebi iswrafvian daikavon saxelzmnis wina

pozicia; kiTxviTi sityva yovelTvis zmniswina pozicias ikavebs; da sxv. da

garkveuli SezRudvebi tendenciebi ganapirobeben sityvaTa aramarkirebul rigs.

statistikuri kvleva naTels hfens amgvar tendenciebs (foCxua 1962;

afridoniZe 1986; Harris 2000) da saSualebas gvaZlevs ganvsazRvroT sityvaTa

kanonikuri, aramarkirebuli rigi: SOV / SVO. pirveli damaxasiaTebelia

literaturuli, samwerlobo enisaTvis, meore ki ufro xSirad sametyvelo,

gansakuTrebiT modernuli sametyvelo stilisaTvis (javaxiSvili 1929), aris

bunebrivi. sityvaTa am aramarkirebuli rigis nebismieri gadalageba ukavSirdeba

winadadebis sainformacio struqturis cvlas. sazogadod, qarTul specialur

literaturaSi sityvaTa rigis cvla am kuTxiT sistemurad ar gaanalizebula;

ZiriTadad ganixileboda aqtualizacia maTesiusis gagebiT (Mathesius, 1967), da daskvnebi mxolod zogadi xasiaTis SeniSvnebiT Semoifargleboda (javaxiSvili

1929; sergia 1972; daviTiani 1973; zardiaSvili 1978; enuqiZe 1981 da sxv).

Tanamedrove lingvisturi Teoriebi da meTodologia, rac faqtiurad

maTesiusis ideebis Semdgom ganviTarebas warmoadgens, ukeTes pirobebs iZleva

sainformacio struqturebis sakvlevad. QUIS-eqsperimentebi orientirebulia

swored am tipis sakiTxebis SeswavlisaTvis Rirebuli spontanuri sametyvelo

teqstebis mopovebaze.

5.2.2. sityvaTa rigi da axali/Zveli informaciis kodireba

enebSi Tavisufali sityvaTa rigiT ‘sityvaTa gadalageba’ mniSvnelovan rols

TamaSobs Zveli/axali informaciis aqtualizaciisaTvis (Skopeteas and Fanselow, 2007). winadadebis sawyisi pozicia yvelaze Rirebulia sainformacio nakadis

garkveuli monakveTebis xazgasasmelad. eqsperimentebis Sedegad mopovebuli

masalis nawili damuSavda statistikurad: davTvaleT Agent-new, Agent-given, Object-new, Object-given erTeulebis gamoCena winadadebis sawyis poziciaSi;

calke Sefasda [Ag-new-invisible] da [O-given, Ag-new-invisible] SemTxvevebi. ramdenadac masalaSi dafiqsirda ramdenime magaliTi lokativebiT, es SemTxve-

vebic aisaxa statistikuri odenobebis amsaxvel cxrilebSi:

23

1. Ag/P winadadebis sawyis poziciaSi

New Given Ag-new-invisible

Total

Ag 11 18 3 32P 1 14 +8 23Loc 2 2

2. axali/Zveli informacia winadadebis sawyis poziciaSi

am statistikuri monacemebidan gamomdinare gamoikveTa Semdegi tendenciebi: 1. agensi (gansakuTrebiT Zveli informaciis matarebeli) miiswrafvis daikavos

winadadebaSi sawyisi pozicia;

2. ucnobi agensi winadadebis sawyis poziciaSi warmodgenilia ganusazRvreli

nacvalsaxeliT viRac; 3. axali obieqti winadadebis sawyis poziciaSia iSviaTia; SeiZleba iTqvas,

gamonaklisia (faqtobrivad mxolod erTi aseTi SemTxveva dafiqsirda);

4. winadadebaSi sawyisi poziciis dakavebis TvalsazrisiT gamoikveTa

semantikur rolebs Soris arsebuli Semdegi ierarqia: Ag>O; 5. Zveli obieqti (gansakuTrebiT konteqstSi ucnob agensTan, romelic

gamoxatulia zmnaSi S.3.PL-sufiqsiT) aseve miiswrafvis daikavos winadadebaSi sawyisi pozicia;

6. Zveli informacia sjobnis axals winadadebaSi sawyisi poziciis dakavebis

TvalsazrisiT: Given>New; 7. gamovlenil or ierarqias (Ag>O,Given>New) Soris konfliqturi

situaciis warmoqmnis SemTxvevaSi moqmedebs ufro konkretuli ierarqia: Ag-given>O-given>Ag-new>O-new; 8. damatebiTi argumentebs (Given>New)>(Ag>O) ierarqiis sasargeblod iZleva

konstruqcia Zveli-lokativiT sawyiss poziciaSi. sazogadod, lokativebis sakiTxi (bolo, (8) SemTxveva) Semdgom kvlevas

moiTxovs.

5.2.3. Llokativebi

qarTulSi sityvaTa Tavisufali rigi xelsayrel pirobebs qmnis lokatiuri

frazebis winadadebis sawyis poziciaSi moTavsebis TvalsazrisiT: isini Tavisi

Cveuli, aramarkirebuli poziciidan (winadadebis bolo) gadmoinacvleben axal markirebul poziciaze, rac mniSvnelovan rols TamaSobs informaciis

Ag P Loc TotalNew 11+3 1 17

Given 18 14+8 2 40

24

struqturirebisas. Cvens masalaSi dadasturebuli 134 SemTxvevidan 61

SemTxvevaSi isini win uswreben saxels, romlis lokalizebac xdeba. masalis

analizi gviCvenebs, rom gadanacvlebis kriteriumebi icvleba cdispirebis

mixedviT, Tumca garkveuli ganzogadebebi mainc SesaZlebelia:

1. meoTxe sesiaSi (sul Catarda cdispirebis mixedviT gansxvavebuli 4 sesia)

monawile cdispirebi lokativebis pozicias gansazRvraven lokalizebuli

saxelis semantikuri niSnebiT: mxolod adamianebi da didi cxovelebi uswreben

lokativebs;

2. mesame sesiis cdispirebisTvis Rirebuli Cans lokatiur frazaSi Semavali

saxelisa da lokalizebuli saxelis zomebi: zomiT (moculobiT) ufro didi

saxeli win uswrebs ufro pataras;

3. es ori kriteriumi met-naklebad Rirebulia danarCeni, pirveli da meore,

sesiis cdispirebisaTvisac, Tumca Zveli/axali informaciac TamaSobs garkveul

rols: Zveli lokativi an axali saxeli miiswrafvis daikavos pirveli adgili.

rogorc masalidan Cans, lokativebis Semcveli informaciis struqturireba

ZiriTadad mosaubris (resp. cdispiris) individualuri kognitiuri

warmodgenebis safuZvelze xorcieldeba.

5.3. sintaqsurad Warbi erTeulebi rogorc topikalizaciis markerebi

5.3.1. qarTuli zmnis polipersonaloba da S/O piris niSnebis siWarbe qarTuli zmnis polipersonalobis gamo, ramdenadac zmnis forma aRniSnavs

erTdroulad subieqtsac da obieqtsac, winadadebaSi piris nacvalsaxelTa

dafiqsireba ar aris aucilebeli (Warbia); da Tu isini mainc warmodgenilni

arian, es umravles SemTxvevaSi niSnavs, rom Sesabamisi subieqti (an obieqti)

topikalizebulia; magaliTad:

[me]T vipove gza, [me]T viyide pamidvrebi, [me]T movutane dedas…

aq me Warbia (qarTulis enis specifikidan gamomdinare: ramdenadac pirveli

piri morfologiurad ukve warmodgenilia zmnur formaSi pirveli subieqturi piris markeriT (v-), misi dublireba sintaqsur konstruqciaSi me nacval-

saxelis saxiT Warbia da migvaniSnebs pirveli piris topikalurobaze:

me da ara sxva vinmem, swored me da ara sxva vinmem, namdvilad me da ara sxva vinmem moutana dedas pamidvrebi.…

5.3.2. qcevis kategoria da Tandebuliani frazebis siWarbe

qarTulSi qcevis kategoria gamoxatavs moqmedebis kuTvnileba-daniSnulebas da

xmovani prefiqsebi aRniSnaven im saxels, Tu visTvis (risTvis) xdeba moqmedeba, amitom Tandebuliani saxeli, romelic moqmedebis mizans aRniSnavs, Warbia da

misi dafiqsireba migvaniSnebs am saxelis topikurobaze:

25

ninom [TavisTvis]T CainiSna davalebebi am winadadebaSi Tandebuliani refleqsuri nacvalsaxeli TavisTvis Warbia,

ramdenadac saTaviso qcevis maCvenebeli xmovanprefiqsi -i- ukve aRniSnavs

moqmedebis samizne saxels, romelic subieqtis identuria (anu aRniSnavs

refleqsurobas); da Tu is mainc warmodgenilia, es migvaniSnebs, rom is aris

implikaciuri topiki:

ninom TavisTvis (da ara, vTqvaT, ZmisTvis) CainiSna davalebebi. amgvarad, winadadebebi sintaqsurad Warbi erTeulebis elifsisis gareSe aris

implikaciuri topikis arsebobaze mimaniSnebeli: sintaqsurad Warbi erTeulebi

sainformacio struqturis topikalizebul wevrebs warmogvidgenen.

5.4.4. sainformacio struqturis sxva sintaqsuri niSnebi

5.4.1. omonimuri sintaqsuri konstruqciebi da winadadebis sainformacio

struqtura

sxvadasxva saxis sintaqsuri omonimuri formebis gaxsnasa da informaciis

calsaxa interpretaciaSi xSirad gvexmareba winadadebis sainformacio

struqturis dadgena; magaliTad, winadadeba: maRali gamyidvlis daxlTan dgas – omonimuria. am omonimiis moxsna SesaZlebelia winadadebis sainformacio

struqturis garkveviT, risTvisac aucilebelia ufro farTo teqstualuri

konteqstis ganxilva. Tu es winadadeba aris pasuxi SekiTxvaze: sad dgas myidveli? – maSin maRali gamyidveli iqneba fokusi da fokusisaTvis

damaxasiaTebeli aRmavali, talRiseburi intonaciiT warmoiTqmis, xolo

winadadeba iqneba uqvemdebaro (myidvelis elifsisiT): [maRali gamyidvlis]FF daxlTan dgas; Tuki maRali aris pragmatikulad gamoyofili wevri, romelic

gulisxmobs, rom vinme maRali da ara, vTqvaT, dabali, dgas daxlTan, maSin igi

warmoiTqmis aRmavali intonaciiT, sadac intonaciuri maxvili meore

marcvalzea, da topikad kvalificirdeba: [maRali]T gamyidvlis daxlTan dgas. ase rom, winadadebis sainformacio struqtura gamoricxavs omonimurobas da

komunikaciur orazrovnebas. 5.4.2. Nsaxeluri frazebis gaxleCa, rogorc topikalizaciis erT-erTi

meqanizmi

saxeluri frazebis gaxleCa damaxasiaTebelia qarTuli enisaTvis: nebismieri saxis NP, garda Tandebuliani frazisa, SeiZleba gaixliCos da es aris

fokusirebuli saxelis warmodgenis erT-erTi formaluri meqanizmi. aseTi

gaxleCili konstruqciebi iSviaTia saliteraturo enisaTvis, Tumca bunebrivia

metyvelebasa da poeziaSi:

qalis xeli uWiravs bavSvs? ara, [fexi]F uWiravs qalisi

26

patara Wiqas iRebs biWi? diax, [pataras]F iRebs is Wiqas 5.4.3. Cleft –konstruqciebi e.w. Cleft –konstruqciebi qarTulSic dasturdeba, Tumca isini arc Tu ise

produqtiulia: cdispirebi mxolod Zalian specifikur SemTxvevebSi iyenebdnen

informaciis struqturirebis am models topikis gamosayofad:

es biWia, visac qali urtyams. es konstruqciebi aigeba yofna-meSvelzmniTa da daqvemdebarebuli winadadebis

formirebiT: topikalizebuli saxeli Tavsdeba mTavar winadadebaSi meSveli

zmnis klitikasTan erTad da qmnis Sedgenil Semasmenels; topikis Sesaxeb

mopovebuli axali informacia (komentari) gadmocemulia damokidebul

winadadebaSi.

yofna zmnis III piris klitikuri elementi -a topiks aRniSnavs ara mxolod Cleft-konstruqciebSi, aramed, sazogadod, nebismier winadadebaSi.

5.4.4. meSvelzmniani klitikuri konstruqciebi rogorc topikis da/an fokusis

markerebi

meSveli zmna ‘yofna’ subieqtur mesame pirTan iqceva klitikad da erTvis

Sedgenili Semasmenlis saxelad nawils: is aris qali > is qalia. is SeiZleba

daerTos winadadebis nebismier sityvas da aqcios is Sedgenili Semasmenlis

saxelad nawilad. es klitika topik-markerad gvevlineba: sityva, romelic

warmodgenilia -a klitikiT, misi wyalobiT topikalizebulia. magaliTisaTvis

ganvixiloT, vTqvaT, winadadeba:

lamazi gogo dRes aris saxlSi. topikalizaciis yvela logikurad dasaSvebi SesaZlebloba realizdeba -a klitikuri elementis darTviT:

(1) saxlSia dRes lamazi gogo. (da ara, vTqvaT, kinoSi)

(2) dResaa lamazi gogo saxlSi. (da ara, vTqvaT, erTi kviris mere) (3) lamazi gogoa dRes saxlSi. (da ara, uSno gogo) (4) lamazia gogo dRes saxlSi. (1) da (2) magaliTebi gamoxatavs CarCo-topikebs droisa da sivrcis sxva

SesaZleblobebis implikaciuri gamoricxviT. igive winadadebebi intonaciuri

konturis cvliT SeiZleba gamoxatavdnen droisa da sivrcis viwro fokusebs.

mTeli saxeluri frazis topikalizaciis nimuSia (3) winadadeba, xolo rac

27

Seexeba (4)-winadadebas, is Zalian specifikur situacias aRwers, romelic

perifrazulad niSnavs: dRes gogo lamazia, imitom rom aris saxlSi.

5.4.5. mimarTvis formebi, rogorc filerebi da maTi funqcia winadadebis

sainformacio struqturis formirebaSi

mimarTvis formebi xSirad asruleben Txrobisas pauzis Semavsebeli

elementebis (e.w. filerebis) funqcias. aseTi filerebi SeiZleba gaCndes

winadadebis nebismieri wevris Semdeg: (1) me, Cemo kargo, didi xania, aq vcxovrob. (2) wigns, Cemo batono, mwerali siyvaruliT wers. (3) wer-s, Cemo kargo, aba ara?! (4) lamazi, Cemo daTo, qali sando ar aris. (5) kargad, RmerTo Cemo, aba vin aris?! aseTi filerebi mniSvnelovan rols TamaSoben informaciis struqturirebaSi.

isini Cveulebriv Cndebian sxvadasxva saxis fokusebisa Tu topikebis

momdevnod; magaliTad, zemoT mocemul (1), (2), (5) winadadebebSi isini

gamoxataven topikebs, maSin rodesac (3) da (4) winadadebebSi – fokusebs.

5.4.6. topikis ‘dasadgeni’ konstruqciebi qarTulSi

topikis gamosacnobad (dasadgenad) Cveulebriv iyeneben Semdeg konstruqciebs: ‘As for X’, ‘As far as X is concerned’, ‘Concerning X’, ‘As regard X’. qarTulSi am

konstruqciebs Seesatyviseba rTuli qvewyobili mimarTebiTi winadadeba: ras ityviT am monadiris Sesaxeb? rac Seexeba monadires, [swored rom is]T klavs irems.

6. daskvnebi

informaciis struqturirebis ganxiluli sintaqsuri saSualebebis Sejamebis

Sedegad SeiZleba davaskvnaT:

qarTulSi informaciis struqturirebis ZiriTadi sintaqsuri modelebia:

1. markirebuli intonacia//markirebuli sityvaTa rigi;

2. gadalageba – informaciulad gamokveTili wevris winadadebis

TavSi dasma (+intonacia);

3. specifikuri sintaqsuri konstruqciebi (+intonacia);

4. nawilakebi (+sintasuri cvlilebebi + intonacia).

am modelebs Soris unda arsebobdes garkveuli ierarqiuli mimarTebebi.

vvaraudobT, rom 1>2>3>4 ierarqia asaxavs aqtualizebis xarisxobriv zrdas:

rac meti formaluri saSualebaa gamoyenebuli informaciis garkveuli nawilis

gamosakveTad, miT maRalia aqtualizaciis xarisxi.

28

literatura afridoniZe, S., 1986. sityvaTganlageba axal qarTulSi. Tbilisi: mecniereba.

asaTiani, r., 1982. martivi winadadebis tipologiuri analizi. Tanamedrove

qarTuli saliteraturo enis masalaze. Tbilisi: mecniereba.

daviTiani, ak., 1973. qarTuli enis sintaqsi. I. martivi winadadeba. Tbilisi:

ganaTleba. E

enuqiZe, l., 1981. winadadebis aqtualuri danawevreba da misi mimarTeba

sintaqsuri da semantikuri analizis Tanamedrove meTodebTan

Tanamedrove zogadi enaTmecnierebis sakiTxebi. VI. Tbilisi:

enaTmecnierebis instituti. 96-110.

foCxua, b., 1962. sityvaTa rigi qarTulSi. ike. XIII. Tbilisi: mecn. akad.

gamom. 109-125.

kvaWaZe, l.,. 1996. Tanamedrove qarTuli enis sintaqsi. Tbilisi: rubikoni.

SaniZe, ak., 1948a. qarTuli enis gramatika. II. sintaqsi. Tbilisi: Tsu

gamomcemloba.

1973b. qarTuli enis gramatikis safuZvlebi. Tbilisi: Tsu

gamomcemloba.

Ciqobava, arn., 1968. Mmartivi winadadebis problema qarTulSi. I.

qvemdebare-damatebis sakiTxi Zvels qarTulSi. Tbilisi: Mmecniereba.

wereTeli, n., 2007. winadadebis sainformacio struqtura: implikaciuri

topiki qarTulSi.samagistro naSromi. Tbilisi. Tsu.

javaxiSvili, iv., 1929. Tanamedrove qarTuli saliteraturo enisaTvis. M mnaTobi. 1. Tbilisi.

Asatiani, R., 2007. The Main Devices of Foregrounding in the Information Structure of Georgian Sentences. Proceedings of Tbilisi Symposium on language, Logic and Computation 1005. Amsterdam: Spriger. 21-31.

Chafe, W. L., 1971. Meaning and the structure of language. Chicago and London: Chicago Un. Press.

Dixon, R. M. W., 1979 Ergativity. Cambridge: Cambridge Un. press. Harris, Al. C., 1998a, Georgian Syntax: A Study in Relational Grammar. Cambridge

etc.: Cambridge University Press. 2000b, Word Order Harmonies and Word Order Change In Georgian. In

Sornicola R, Poppe E., Haley A. (eds), Stability, Variation and Change of Word-Order Patterns over time. 133-163. Amsterdam/Philadelphia: Benjamins

Hewitt, B. G., 1995, Georgian: A Structural Reference Grammar. Amsterdam, Philadelphia: Benjamins.

Kibrik, A., 1997. Beyond Subject and Object: Toward a Comprehensive Relational Typology. Linguistic Typology.I.. Berlin-New York: Mouton de Gruyter. 279-346.

Mattesius, B. O., 1967. O sistemnom grammaticheskom analize. prajskij ligvisticheskij krujok. Moscow: Nauka. 226-238.

Shibatani, M., 1985. Passives and related constructions: A prototype Analysis. Language, Vol.61,No4. Kobe: Kobe Un. press.

Skopeteas, S., Fiedler I., Hellmuth R., Schwarz, A., Stoel, R., Fanselow, G., Féry, C., and Krifka M. 2006. Questionnaire on Information Structure (QUIS).

29

Interdisciplinary Studies on Information Structure 4. Working Papers of the SFB 632, Potsdam: Universitätsverlag Potsdam.

Skopeteas, S., Fanselow, G., 2007. Effects of givenness and constraints on free word order. In Information Structure from different perspectives. Zimmermann, Malte and Caroline (eds.), Oxford: Oxford Un. Press.

Tuite, K., 1998. Kartvelian morphosyntax. Munich: Lincom Europa. Wierzbicka, A., 1996. Semantics (Primes and Universals). Oxford-New York: Oxford

Un. press. Rusudan Asatiani

Syntactic Devices of Information Structuring in Georgian

Abstract Foregrounding of certain parts of the information can be realized on various linguistic levels and it is possible to distinguish: Conceptual, Functional, Discourse and Pragmatic devices, which can be represented by various formal means: Phonetic-Phonological, Morphological-Syntactic and Lexical-Pragmatic. In Georgian there is no morphological topic marker, but all other devices of foregrounding are possible. The paper examines the main models of syntactic devices, which are used for the formalization of various kinds of topics and/or focuses: passive constructions, reordering of unmarked word order, fronting of topics and/or focuses, cleft constructions, presentation of syntactically redundant elements, function of auxiliary verbs, fillers and particles in building of specific syntactic constructions showing foregrounding of certain parts of information.

30

rusudan asaTiani

vnebiTi gvaris konstruqciaTa pragmatikuli

Tavisebureba qarTvelur enebSi


pasiuri konstruqcia Teoriul enaTmecnierebaSi (funqcionalur mimarTebaTa

gramatikis Teoriis farglebSi) ganixileba Sesabamisi aqtiuri konstruqciis

konversiul formad, sadac paciensi dawinaurebulia sintaqsur mimarTebaTa

ierarqiulad organizebul rigSi (S>DO>IO) da qceulia subieqtad, agensi ki

transformirebulia windebulianFfrazad da, funqcionaluri TvalsazrisiT,

aRar warmoadgens zmnis pirianobiT gansazRvrul birTvul arguments; amdenad,

pasivi sintaqsur kategoriad moiazreba. mraval enaSi aqtiuri konstruqciis

konversiul (resp. pasiur) konstruqciaSi warmodgenili zmnuri forma morfo-

logiuradac markirebulia da, Sesabamisad, am enebSi gvaqvs vnebiTi gvaris

morfosintaqsuri kategoria.

qarTvelur enebSi pasiuri konstruqcia yovelTvis ar warmoadgens Sesabamisi

aqtiuri konstruqciis konversiul formas; e.i. is ar aris mkacrad sintaqsuri

kategoria, Tumca aqtiur-pasiur zmnur formaTa morfosintaqsuri

dapirispireba saxezea [1, 1990], miuxedavad imisa, rom aglutinaciuri tipis

(calsaxad mxolod vnebiTi gvaris gamomxatveli) morfologiuri markeri arc

erT qarTvelur enaSi ar dasturdeba [1, 2001].

2. Tanamedrove midgoma

Tanamedrove enaTmecnierebaSi miRebuli erT-erTi Teoriuli midgomis Tanaxmad

aqtiur-pasiur konstruqciebs Soris arsebuli funqcionaluri sxvaoba

aixsneba winadadebaSi informaciis struqturirebis sxvadasxvagvarobiT.

sazogadod, informaciis struqturireba, misi SefuTva, xorcieldeba opozi-

ciis safuZvelze, sadac informaciis erTi nawili gamoikveTeba informaciis

meore nawilisgan. sakomunikacio, pragmatikuli TvalsazrisiT, es gamokveTili

nawili aris aqtualizebuli _ xazgasmuli, mniSvnelovani _ da warmogvidgens

sainformacio nakadSi dawinaurebul informacias. nebismieri saxis

‘dawinaureba’ (igive: xazgasma, gaSuqeba, fokusSi moqceva, logikuri gamokveTa,

win wamoweva, TvalTaxedvis centrSi moqceva da sxva amgvari) SeiZleba

ganvixiloT rogorc erTi, saerTo fenomeni, romelic asaxavs informaciis

lingvisturi struqturirebis mTavar strategias. aqtualizeba sxvadasxva

struqturis mqone enebSi gansxvavebuli formaluri saSualebebiT xorciel-

deba da nebismier enobriv doneze SeiZleba iyos realizebuli [1, 2006].

31

2.1 konceptualuri aqtualizeba


ZiriTadad sami mTavari konceptualuri ganzomilebis mixedviT mimdinareobs

[12, 1997]:

1. situaciis Semadgeneli semantikuri rolebis (agensi : paciensi :

adresati) markireba;

2. sainformacio nakadis pragmatikuli Rirebulebebis (topiki :

komentari) markireba;

3. sakomunikacio aqtis statusebis (molaparake/msmeneli : saubris

sagani) markireba.



semantikuri rolebis markirebaze orientirebul enebSi informaciis Tvalsaz-

risiT aqtualizebul saxelad konceptualizdeba an agensi an paciensi, ris

Sedegadac warmoiqmneba an nominatiuri an ergatiuli konstruqciebi. pirveli

afiqsirebs informaciis im nawilis aqtualizebas, romelic gamoxatavs, Tu vin moqmedebs, meore ki informaciis im nawilis aqtualizebas, romelic gamoxa-

tavs, Tu ra xdeba, ra iqmneba. konceptualurad aqtualizebuli saxeli

enobriv struqturebSi formalurad warmodgenilia aramarkirebuli, saxelo-

biTi brunviT: nominatiur konstruqciebSi es aris agensi, xolo ergatiul

konstruqciebSi – paciensi.

2.2. funqcionaluri aqtualizeba

funqcionarulad aqtualizebuli saxeli gramatikul tradiciaSi aRiniSneba






agensia konceptualurad aqtualizebuli, SeiZleba paciensi wamoiwios da iq-

ces gamonaTqvamis subieqtad; anu informacia gadalagdes ise, rom gamonaTqvami

gvawodebdes garkveul informacias swored rom moqmedebis obieqtis, paciensis,

Sesaxeb da centralur, mTavar elementad aqtualizdes ara agensi, aramed

paciensi. nominatiur enebSi aqtiuri konstruqciebi asaxaven agensis aqtua-

lizebas (resp. agensi asrulebs subieqtis funqcias) maSin, rodesac pasiuri

konstruqciebi warmoaCenen paciensis aqtualizebas (resp. paciensi aris subi-eqti). formalurad paciensis amgvari dawinaureba gamoixateba misTvis nomi-

natiuri formis miniWebiT. e.i. paciensi pasiur konstruqciaSi formalurad

imave brunviT aris markirebuli, romliTac aqtiur konstruqciaSi koncep-

tualurad aqtualizebuli saxeli _ agensia warmodgenili. (agensi pasiur

konstruqciaSi an sruliad gamqralia, an Tandebuliani formiT dasturdeba.)

32

amdenad, sintaqsurad gansxvavebuli aqtiuri-pasiuri konstruqciebis dapiris-

pirebas gansazRvravs informaciis struqturebis gansxvavebuli modelebi.

3. kvlevis meTodologia

informaciis struqturirebis formaluri modelebis moZiebisaTvis gansakuT-

rebiT Rirebulia bunebriv sametyvelo situaciebSi warmoqmnili winadadebebi.

aseTi situaciebis stimulireba SesaZlebelia garkveuli saeqsperimento amoca-

nebis SerCeviT. am amocanebis gadawyvetaze dafuZnebuli eqsperimentebis Sede-

gad mopovebuli masala warmoadgens spontanurad warmoqmnili sametyvelo

teqstebis korpuss, romlis analizsac eyrdnoba winamdebare kvleva.

3.1. eqsperimentebis aRwera

cdispirebs eZleodaT baraTebi ori erTmaneTTan dakavSirebuli situaciis

amsaxveli suraTiT, sadac mTavari participantebi icvlebodnenNsulierobisa

da xilvadobis (gansazRvruloba-ganusazRvrelobis) niSnebis mixedviT:

sulierobis cvla: A sulieri paciensi B usulo paciensi

gansazRvrulobis cvla: C gansazRvrebadi agensi D aragansazRvrebadi agensi

suraTebi imgvarad iyo SerCeuli, rom asaxavdnen am niSanTa logikurad

SesaZlebel yvela kombinacias: {A,C}, {A,D}, {B,C}, {B,D}.

eqsperimentebis Sedegad movelodiT, rom ucnobi, aragansazRvrebadi agensis

da/an sulieri paciensis SemTxvevaSi informantebi upiratesobas mianiWebdnen

pasiur konstruqciebs.

informantebs unda exelmZRvanelaT Semdegi instruqciiT:

Tqven geZlevaT ori scenis amsaxveli suraTebi, romlebic erTmaneTTan dakavSirebulia. warmoidgineT, rom jer pirvel suraTze aRbeWdil scenas SeeswariT, uSualod mis momdevnod (vTqvaT, ramdenime wuTis Semdeg) ki _ meores. gTxovT aRweroT, ra xdeba yovel scenaze?

33

magaliTisaTvis warmogidgenT ramdenime baraTs:

AC

BC BD

BD

3.2. informantebi

qarTvelur enaTa masala mopovebul iqna savele pirobebSi: yoveli enisaTvis

Catarda rva-rva sesia TbilisSi, anakliasa da gardabnis raionSi (iq

Casaxlebul svan informantebTan). qarTuli, megruli da svanuri masalis

Sesagroveblad gamokiTxul iqna, Sesabamisad, rva-rva wyvili informantebisa,

ZiriTadad, studenti axalgazrdebi.

4. masalis analizi

QUIS-esperimentebis [14,2006] safuZvelze mopovebul spontanur sametyvelo

gamonaTqvamebSi, rodesac agensi situaciurad ucnobi, gaurkveveli partici-

pantia da pasiuri konstruqcia (gauCinarebuli agensiT) yvelaze ufro mosa-

lodnelia, cdispirebi upiratesobas aniWeben aqtiur konstruqciebs, im Sem-

TxvevebSic ki, rodesac paciensi sulieria, anu, paciensis aqtualizaciisaTvis

da, Sesabamisad, pasiuri konstruqciebis formirebisaTvis yvelaze Sesaferisi

pirobebia.

4.1. qarTuli masala

qarTulSi mosalodneli pasivis nacvlad warmoebuli konstruqciebi specifi-

kuria: am konstruqciebSi subieti warmodgenilia an ganusazRvreli nacval-

saxelebiT viRac, raRac; an gvaqvs e.w. ganusazRvrelsubieqtiani winadadebebi

AD

34

subieqturi mesame piris mravlobiTobis maCvenebeli sufiqsiT zmnaSi.

magaliTad:

am kacs TavSi CaquCs urtyamen. boTls kraven fexs. mas viRac fexs urtyams. viRaca agdebs gogos kibidan. da sxv.

mxolod erT SemTxvevaSi dafiqsirda ‘situaciurad ucnobi agensis’

konkretuli arsebiTi saxeliT warmodgenis SemTxveva: tafas iRebs qali, rac, rogorc Cans, aixsneba zogadi warmodgeniT, rom ‘ZiriTadad qalebi amzadeben saWmels da saqme aqvT tafebTan’, ris safuZvelzec ‘ucnobi agensi’ cdispiris

mier Canacvlda konkretuli saxeliT qali:

sainteresod gveCveneba, kidev erTi konstruqcia _ viRacis nasroli yuTi agdebs mas (=gogos) magididan. es, usulo agensis Semcveli, konstruqcia

asaxavs angariSgasawev tendencias: savaraudo sulieri agensis Semoyvana ganu-

sazRvreli nacvalsaxeliT gadmocemuli atributuli konstruqciiT (viRacis nasroli) imgvarad, rom usulo participantis agensuri interpretacia ar

icvleba da is winadadebis subieqtad kvalificirdeba. am tendenciis kogni-

tiuri safuZvlebi aSkaraa _ usulo participantis interpretacia agensad

miuRebelia, Tumca misi subieqtad gaazreba (anu, agensTan funqcionaluri

gaTanabreba)_ dasaSvebi. es konstruqcia kidev erT zogad tendenciaze metyve-

lebs: usulo agensisa da sulieri paciensis SemTxvevaSic mosalodneli

pasivis nacvlad qarTulSi kvlav aqtiur konstruqcias vxvdebiT.

warmogidgenT “ucnobi agensis” Semcveli konstruqciebis statistikuri

maCveneblebis amsaxvel cxrils, romelic Sedgenilia Cven mier mopovebuli

masalis safuZvelze:

35

cxrili-1 aqtiuri

konstruqcia

ganusazRvreli

nacvalsaxelebiT

aqtiuri konstruqcia

ganusazRvrelsubieqtiani

mesame piris zmnuri formiT

pasiuri

konstruqcia

sxva tipis

konstruqcia

sul

10

10

0

1

21

miuxedavad imisa, rom masala ar aris didi raodenobis, konstruqciaTa gamo-

yenebis garkveul tendenciebze msjeloba mainc SesaZlebelia: ‘ucnobi agensis’

Semcveli situaciebis gadmosacemad: (1). saerTod ar gvxvdeba pasiuri

konstruqciebi; (2). ganusazRvreli subieqtis gamomxatveli zmnuri formebiT

(SIII piris mravlobiTi ricxvisa) agebuli aqtiuri konstruqciebi da ganu-

sazRvrelnacvalsaxeliani aqtiuri konstruqciebi Tanabrad misaRebia _ arCe-

vani, ZiriTadad, damokidebulia informantze: esa Tu is informanti, rogorc

wesi, irCevs erT-erT strategias da iSviaTad Tu cvlis mas. (3). sxva tipis

konstruqciebi ar dafiqsirda (Tu ar CavTvliT zemoT aRweril SemTxvevas _

tafas iRebs qali, sadac ucnobi agensi warmodgenilia konkretuli sazogado

saxeliT _ qali. am SemTxvevaSic konstruqcia aqtiuria).

is, rom qarTul sametyvelo teqstebSi ‘ucnobi agensis’ gamosaxatavad ar

dafiqsirda pasiuri konstruqciebi, Zlieri argumentia im interpretaciis

sasargeblod, romlis mixedviTac qarTulSi pasiur konstruqciebis mastimu-

lirebeli ar aris sainformacio struqturis konceptualur-funqcionaluri

aqtualizeba; e. i. is ar aris ubralod sintaqsuri mimarTebaTa cvlis Sedegi.

4.2. megruli masala

msgavsi daskvnebisken gvibiZgebs megrulisaTvis SemuSavebuli saeqsperimento

amocanebis gadawyvetis Sedegad mopovebuli masalis analizic _ arc erTma

informantma ucnobi agensis Semcveli situaciis aRwerisas ar gamoiyena

pasiuri konstruqcia da, qarTulis msgavsad, upiratesoba mianiWa aqtiur

konstruqciebs ganusazRvreli nacvalsaxeliT; Tumca, qarTulisagan

gansxvavebiT, megrul masalaSi SedarebiT iSviaTia konstruqciebi

ganusazRvrelsubieqtiani zmnis SIIIpl piris formiT; aseve, ar dafiqsirda sxva tipis konstruqciebi (ix. cxrili 2):

cxrili 2

aqtiuri

konstruqcia

ganusazRvreli

nacvalsaxelebiT

aqtiuri konstruqcia


zmnuri formiT

pasiuri

konstruqcia

sxva tipis

konstruqcia

sul

14

3

0

0

17

36

suraTebis aRwerisas, informantebi, qarTulisgan gansxvavebiT, megrulSi ufro

xSirad mimarTavdnen warsul dros (wyvetils), rac sainformacio struqturis

TvalsazrisiT metad saintereso faqtia: megrulis wyvetilSi gvaqvs mxolod

ergatiuli konstruqcia; ergatiuli konstruqcia ki, sazogadod, paciensis

(resp. pirdapiri obieqtis) konceptualur aqtualizaciaze migvaniSnebs da, am

TvalsazrisiT, funqciurad pasiur konstruqcias uTanabrdeba (konceptualuri

aqtualizaciisaTvis ix. [1, 2006; 2007]). magaliTebi:

midgaren-q tafe ge-Wof-u skami-Se somebody-ERG pan.NOM PR-take-AOR.S.3.SG chair-from viRacam tafa aiRo skamidan

koCi-s xe-s mi-o-g-an-a(n) man-DAT hand-DAT PR-OV-hit-TM-S.3.PL kacs xels urtyamen

is, rom megrulSi upiratesoba eniWeba konstruqcias ganusazRvreli

nacvalsaxeliT midgaren, SeiZleba aixsnas am nacvalsaxelis agebulebiT: midga-re-n (somebody-be-PRS.S.3.SG) “viRac aris (rom)”. anu, gvaqvs tipuri e.w.

gaglejili (Cleft) konstruqcia, romelic bevr enaSi (mag.: inglisurSi, germa-

nulSi, rusulSi da sxv.) fokusirebuli Semadgenlis gamoxatvis erT-erTi

gavrcelebuli modelia da, savaraudod, megrulSic ukeT gaxazavs ucnobi

agensis fokusur bunebas (e.i. miuxedavad imisa, rom fokusuri Semadgeneli

ucnobia, misi fokusuri buneba, mainc, gaxazulia), rac ganusazRvrelsubieqtian

winadadebaSi SeuZlebelia _ fokusirebuli axali (Tumca, ucnobi) informacia

(anu, ucnobi Ag) zedapirulad saerTod ar aris warmodgenili. midga-re-n formaSi yofna zmnis klitikur bunebaze (anu imaze, rom is bolomde erT

sityvad ar aris qceuli) miuTiTebs brunvis niSnis poziciuri, TiTqmis Tavi-

sufladmonacvle variantebis dadastureba masalaSi: midga-re-n-q//midga-q-re-n.

amgvarad, masalis analizma gviCvena, rom megrulic Tavs aridebs mosalodnel

pasiur konstruqciebs ucnobi agensis gamosaxatavad, magram usulo agensis

SemTxvevebSi, sadac qarTuli kvlav upiratesobas aqtiur konstruqcias

aniWebs, megruli ZiriTadad irCevs gansxvavebul strategias: rTul (zogjer

martivsac) winadadebebs, sadac warmodgenilia dinamikuri pasivebi: xvdeba, vardeba, ejajgureba, awveba da sxv; mag.: burTi moxvda gogos da is magididan vardeba (nacvlad qarTulSi dafiqsirebuli aqtiuri konstruqciisa: burTma gadmoagdo gogo magididan).

4.3. svanuri masala

qarTuli da megruli enebis kvaldakval svanuric msgavs strategiebs irCevs:

(1). ucnobi agensi gamoxatulia ganusazRvreli nacvalsaxeliT ervÀle, ‘re; (2). ganusazRvrelsubieqtiani konstruqcia zmnis SIIIpl piris maCvenebliT –x;

37

(3). erT SemTxvevaSi, qarTulis msgavsad, dafiqsirda aqtiuri konstruqcia,

sadac ucnobi agensi Canacvlebul iqna konkretuli sazogado saxeliT kaci:

mÀ Cuxakónk–a. kocÒl xanunRo ჴedni m‘rme mÀre i al mÀres TxumCu k–atxs xatyci. “kaci Cacucqulia. cota xnis Semdeb modis meore (sxva) kaci

da am kacs TavSi CaquCs urtyams.”

miuxedavad imisa, rom suraTze ar Cans, Tu vin urtyams kacs CaquCs, svanma

cdispirma ixmara kacis aRmniSvneli leqsema mÀre, rac, rogorc Cans, aixsneba zogadi warmodgeniT, rom ‘ZiriTadad kacebi arian agresiulebi da Cxuboben’. savaraudoa, rom swored aseTi kognitiuri interpretaciis safuZvelze

‘ucnobi agensi’ cdispiris mier Canacvlda konkretuli saxeliT kaci. svanuri masalis statistikuri analizi warmodgenilia mesame cxrilSi:

cxrili 3

amgvarad, samive qarTvelur enaSi ZiriTadad ergvarovani suraTia _ pasiuri

konstruqciebi ucnobi agensis Semcveli situaciebis gadmosacemad ar gvxvdeba.

es faqti Zlieri argumentia im interpretaciis sasargeblod, romlis

mixedviT qarTvelur enebSi pasiuri konstruqciebis mastimulirebeli ar aris

sainformacio struqturis konceptualur-funqcionaluri aqtualizeba; e. i.

is ar aris ubralod sintaqsuri mimarTebaTa cvlis Sedegi; da, amdenad, unda

veZioT pasiuri-aqtiuri morfosintaqsuri modelebis ganmsazRvreli sxva

(savaraudod, ara sintaqsuri, aramed semantikuri da/an kognitiuri) niSnebi.

qarTulSi aseTi savaraudo niSnebis analizs SemogTavazebT qvemoT.

5. vnebiTi gvaris semantikur-kognitiuri interpretacia

5.1. funqcionaluri midgoma

Teoriul enaTmecnierebaSi (kerZod, funqcionalur mimarTebaTa gramatikis

Teoriis farglebSi) pasiuri konstruqcia ganixileba Sesabamisi aqtiuri

konstruqciis konversiul formad, sadac paciensi dawinaurebulia sintaqsur

mimarTebaTa ierarqiulad organizebul rigSi (S>DO>IO) da qceulia

subieqtad, agensi ki transformirebulia windebulianFfrazad da, funqci-

onaluri TvalsazrisiT, aRar warmoadgens zmnis pirianobiT gansazRvrul

aqtiuri

konstruqcia

ganusazRvreli

nacvalsaxelebiT

aqtiuri konstruqcia


zmnuri formiT

pasiuri

konstruqcia

sxva tipis

konstruqcia

sul

11

6

0

1

21

38

birTvul arguments; amdenad, pasivi sintaqsur kategoriad moiazreba. mraval

enaSi aqtiuri konstruqciis konversiul (resp. pasiur) konstruqciaSi warmo-dgenili zmnuri forma morfologiuradac markirebulia da, Sesabamisad, am

enebSi gvaqvs vnebiTi gvaris morfosintaqsuri kategoria.

qarTulSi pasiuri konstruqcia yovelTvis ar warmoadgens Sesabamisi aqtiuri

konstruqciis konversiul formas; e.i. is ar aris mkacrad sintaqsuri kate-

goria, Tumca aqtiur-pasiur zmnur formaTa morfologiuri dapirispireba

saxezea. pasiur formaTa maxasiaTeblebia:

1. awmyoSi subieturi III piris niSani aqtiuri formebisTvis aris -s, pasiuri formebisTvis ki -a sufiqsi; 2. aseve, awmyos (da, sazogadod, I seriis) zmnuri formebisTvis

damaxasiaTebelia -eb- Temis niSani; 3. aqtiuri formebisgan gansxvavebiT, pasiurSi SeiZleba dadasturdes

(ara yvela pasivisTvis; ix. e. w. uniSno vnebiTebi) xmovani prefiqsebi -i-, -e- da sufiqsi -d-; es samive maxasiaTebeli erTad formalurad naTlad ganasxvavebs e.w.

aqtiur da pasiur zmnur formebs, Tumca cal-calke verc erTi maTgani ver

miiCneva mkacrad vnebiTi gvaris morfologiur markerad, ramdenadac:

1. -s da -a sufiqsis upirvelesi, mTavari funqciaa subieqturi III piris

markireba da isini aqtiuri-pasiur zmnuri formebis calsaxa markerebad ver

kvalificirdeba; mag., -s gvxvdeba pasiuri formebis kavSirebiTi kilos SIII piris formebSi; -a aqtiuri zmnebis warsul droebSi da sxv.

2. -eb- Temis niSnis zogadi funqcia unda iyos dinamikur formaTa

warmoeba da am funqciiT is aqtiur zmnebTanac gvxvdeba imave I seriis zmnur

formebSi;

3. xmovani prefiqsebic polifunqciuria. isini, zogadad, gamoxataven

zmnis valentobis cvlasTan dakavSirebul derivaciul cvlilebebs _

valentobiT gansazRvrul aqtantur saxelTa matebas an klebas. (mag., -i-prefiqsi gamoxatavs: saTaviso qcevas, sasxviso qcevas, refleqsur formebs,

xSiria deponensuri Sinaarsis zmnebTan, zogierT zmnebTan SeiZleba

potencialisic gamoxatos da ‘awarmoebs’ e.w. saSuali gvaris zmnaTa momaval

drosa da II seriis formebs. [7, 2001])

ase rom, saxezea formalurad gansxvavebuli ori modeli _ aqtiuri da

pasiuri, romelTa funqcionaluri kvalifikacia da semantikuri interpretacia

calsaxad naTeli ar aris da Semdgom analizs moiTxovs.

5.2. aqtiv-pasivis dapirispirebis kontinuumis saxiT warmodgenis modeli

mraval enaSi, qarTulis msavsad, aqtiv-pasivis dapirispireba ar warmoadgens

mkacrad sintaqsur kategorias da markirebis pasivisaTvis damaxasiaTebeli

modeli sxva kategoriebis formalizebisaTvisac gamoiyeneba. sazogadod,

cnobilia, rom msoflios mraval enaSi pasivis formaluri modeli iTavsebs

iseTi mimarTebebis markirebas, rogoricaa refleqsuri da reciprokaluri

39

mimarTebebi. zogierT enaSi, magaliTad, iaponurSi pasivis modeli iTavsebs,

aseve, spontanuri moqmedebebis, potencialisis, zrdilobiani formebisa da

mravlobiTi ricxvis markirebis funqciasac. bunebrivad gaCnda moTxovnileba

SemuSavebuliyo Teoriuli safuZvlebi amgvari faqtebis asaxsnelad. erT-erTi

aseTi Teoriulad axali interpretacia ekuTvnis iaponel enaTmecniers, m.

Sibatanis [13,1985]. mas miaCnia, rom aqtiv-pasivis dapirispireba ar warmoadgens

diskretul, binarul dapirispirebas da, ufro metad, kontinuumis saxe aqvs.

am kontinuumis ukiduresad polaruli erTeulebi Seesabameba prototipul

aqtivsa da pasivs; maT Soris moqceulia mravali ‘marginaluri’ Sinaarsi,

romelTac aqtivTan an pasivTan mxolod zogierTi niSani aerTianebT:

prototipuri aqtivi / medialuri formebi / prototipuri pasivi

<---------------------/------------------------------------/--------------------------->

am ‘marginaluri’ Sinaarsebis (vuwodoT maT medialuri) markirebisTvis enebi

irCeven gansxvavebul strategiebs: an qmnian axal, sruliad gansxvavebul

modelebs, an arsebul modelTagan am konkretuli SinaarsisaTvis yvelaze

axlo modelis mixedviT aformeben maT. aseT SemTxvevebSi, formaluri

modelebis funqcionaluri an semantikuri interpretacia mxolod sintaqsuri

mimarTebebis cvliT garTulebuli da, xSir SemTxvevaSi, SeuZlebelicaa.

5.3. qarTulSi aqtiv-pasivis kontinuumi

is, rom qarTulSi aqtiv-pasivis dapirispirebis gansxvavebul modelebs ver

gansazRvravs mkacrad sintaqsuri mimarTebebis cvla, cxadia (ix. zemo

msjeloba, Tavi 2). rac Seexeba semantikur niSnebs, aqac ufro faqizi

niuansebia misakvlevi, ramdenadac aseve cxadia, rom semantikuri dapirispireba

aqtiuri Sinaarsi : pasiuri Sinaarsi calsaxad ver faravs formalur dapiris-

pirebas aqtiuri modeli : pasiuri modeli; da pasiuri modeli zogjer aqtiur semantikasac gamoxatavs, magaliTad: eqaCeba, awveba, elaparakeba, uyveba, emaleba da sxv.

Tu qarTulSi aqtiv-pasivis dapirispirebas kontinuumis saxiT warmovadgenT,

sadac prototipuli aqtivi Seesabameba e.w. moqmedebiTi gvaris formebs, xolo

prototipuli pasivi – mis konversiul pasiur formebs, medialuri

Sinaarsebis markirebis procesi SeiZleba aixsnas zogadi kognitiuri

tendenciiT:

medialuri Sinaarsebis markirebisas ena irCevs an aqtiur an pasiur models da am strategias gansazRvravs enaTa mixedviT gansxvavebuli, specifikuri ‘gadawyvetileba’, Tu romeli semantikuri niSani miiCneva centralurad aqtiuri-pasiuri Sinaarsebis dapirispirebaSi. qarTulSi aseTi semantikuri niSnis (romelic gansazRvravs, Tu romeli

(aqtiuri Tu pasiuri modeli SeirCeva lingvisturi struqturirebis

40

kognitiur procesSi) gamosavlenad SeiZleba SemoTavazebul iqnes Semdegi

interpretacia: Tu medialuri (e.i. prototipulad arc aqtiuri da arc pasiuri) semantikis zmna predikaciisas ganasxvavebs mimarTulebas, zmna uTanabrdeba pasivs da irCevs markirebis pasiur models; xolo Tu medialuri semantikis zmna ver ganasxvavebs mimarTulebas, maSin is uTanabrdeba aqtivs da irCevs markirebis aqtiur models. formalurad es wesi advilad interpretirebadia:

Tu medialuri Sinaarsis zmna dairTavs zmniswin(eb)s, mas eqneba pasiuri forma, Tu ara – aqtiuri. (magaliTad.: dgeba/a-dgeba/Ca-dgeba/gada-dgeba; emaleba/da-emaleba; awveba/mi-awveba/da-awveba da a.S.; magram: cxovrobs, fiqrobs, muSaobs, dgas, Rirs da a.S.)

ramdenadac zmniswinebis funqcia qarTulSi mimarTulebisa da orientaciis

markirebis paralelurad moqmedebis sruli-usruli aspeqtis mixedviT

gansxvavebuli Sinaarsebis markirebacaa, SeiZleba davaskvnaT, rom aqtiur-

pasiur modelTa SerCevis ganmsazRvreli semantikuri niSania moqmedebis

‘dasrulebulobis’ mixedviT gansxvavebis SesaZlebloba:

Tu konkretuli zmnis semantika uSvebs moqmedebis dasrulebulobas (rac, zogadad, prototipuri aqtivis damaxasiaTebeli erT-erTi niSania), zmna formdeba prototipuli pasivisaTvis damaxasiaTebeli formaluri markerebiT _ SeirCeva pasiuri modeli. lingvisturi markirebis amgvari strategia, erTi SexedviT, SeiZleba

uCveulod mogveCenos. TiTqos, ufro mosalodneli unda yofiliyo aqtiuri

semantikisaTvis Rirebul niSans (moqmedebis dasrulebuloba-dausruleblobas)

aqtiuri modelis SerCeva ganepirobebina, magram Tu mxedvelobaSi miviRebT

zogadad formaluri markirebis ganmapirobebel kognitiur procesebs, advili

dasanaxi iqneba, rom qarTulSi lingvisturi struqturirebis ganmsazRvreli

pirobebi aixsneba markirebulobis mimarTebis ganmsazRvreli zogadi kogni-

tiuri safuZvlebiT:

ramdenadac pasiuri semantikisaTvis, sazogadod, naklebad mosalodnelia (e.i.

kognitiurad markirebulia) mimarTulebis gamoxatva (rac bunebrivia proto-

tipulad aqtiuri semantikis zmnebisTvis), is medialuri zmnebi, romlebic

uSveben amgvar Sinaarsebs, kognitiurad markirebulni arian da, Sesabamisad,

enobriv struqturebSic warmodgenilni arian markirebuli (resp. pasiuri)

modeliT; anu: kognitiuri markireba lingvisturi struqtuqturebis

markirebulobiT reprezentirdeba. ena amgvari modelis SerCeviT gvawvdis

mravali TvalsazrisiT datvirTul gramatikul informacias:

41

araprototipuli pasivi (da araprototipuli is imitomac aris, rom aqtiuris msgavsad ganasxvavebs moqmedebis dasrulebuloba-dausruleblobas), rogorc markirebuli, gamoixateba prototipul modelTagan formalurad yvelaze markirebuli (resp. pasiuri) modeliT. 5.4. gafarToebuli kontinuumi

SesaZlebelia amgvari interpretacia gavafarTovoT im mizniT, rom man moicvas

medialuri Sinaarsebis mTeli speqtri, e.w. statikuri vnebiTebic.

medialuri Sinaarsebis formaluri modelebis SerCevis procesis dros Tu

statikur vnebiTebsac gaviTvaliswinebT, lingvisturi struqturirebis

procesi orsafexurian procesad SeiZleba warmovadginoT, sadac gadamwyvet

rols ukve meore, aqtiur-pasiur SinaarsTa gansxvavebisaTvis aranakleb

Rirebuli, semantikuri dapirispireba ‘dinamika : statika’ TamaSobs

mniSvnelovan rols:

I safexuri: Tu medialuri Sinaarsi statikur mdgomareobebs gamoxatavs zmnas axasiaTebs awmyoSi meSvelzmniani uRvlileba; e.i. aseT SemTxvevaSi qarTuli ena qmnis aqtiv-pasivisgan gansxvavebul, axal

models. aseTi zmnebia, ak. SaniZis terminologiiT, e.w. statikuri vnebiTebi

(mag., dgia, afenia, kidia da sxv.) da saSual-vnebiTebi (mag., dgas, zis, wevs da sxv.). warmodgenili magaliTebidanac Cans, rom axali modelis mixedviT

gaformebuli statikuri zmnebi iyofa or qvejgufad: statikur zmnebi awmyoSi

SIII piris daboloebad pasiuri modelisaTvis damaxasiaTebel -a sufiqss

irCeven, saSual-vnebiTebi ki _ aqtiuri modelisaTvis damaxasiaTebel -s sufiqss. aq, rogorc Cans, modelTa SerCevisaTvis Rirebulia aqtiv-pasivis

prototipul SinaarsTa ganmasxvavebeli mesame dapirispireba:

Tu konkretuli medialuri Sinaarsis zmnis argumenti Tavad, Sinagani ZaliT aris garkveul mdgomareobaSi moqceuli (avtotiuria), zmna irCevs aqtiur models; winaaRmdeg SemTxvevaSi ki (e.i. Tu ar aris avtotiuri) pasiurs. (am niSnisa da aseTi interpretaciis gasaazreblad SeiZleba movixmoT aseTi

enobrivi testi: gamonaTqvamebi _ Tavad dgas, Tavad wevs, Tavad zis da a.S. _ misaRebia, maSin rodesac gamonaTqvamebi _ *Tavad gdia, *Tavad kidia, *Tavad afenia _ arabunebrivi da xelovnuria.)

II safexurze lingvisturi struqturireba mimdinareobs ukve dinamikuri-

statikuri niSniT diferencirebuli qvesistemebisTvis zemoT (ix. 6.3.)

Camoyalibebuli procesebis Sesabamisad.

42

5.5. aqtiv-pasivis dapirispireba _ semantikur-kognitiuri interpretacia

amgvarad, aqtiur-pasiur SinaarsTa markirebis ierarqiulad organizebuli

procesi Semdegnairad SeiZleba warmovadginoT:

I safexuri: prototipuli Sinaarsebi gamoxatulia ZiriTadi modelebiT,

romelTagan formalurad markirebulia sainformacio struqturis Tvalsaz-

risiT funqcionalurad aqtualizebuli paciensis Semcveli e.w. pasiuri kons-

truqcia (resp. aqtiuri konstruqciis Sesabamisi sintaqsurad konversiuli

pasivi);

II safexuri: medialuri (resp. prototipisgan gansxvavebuli) Sinaarsebis

gamoxatvisTvis gamoiyeneba ori gansxvavebuli strategia: 1. an iqmneba sruliad

axali modeli; an 2. SeirCeva prototipuli aqtivisTvis (2a); an proto-

tipuli pasivisTvis damaxasiaTebeli (3b) modelebi. modelTa SerCevis

strategias gansazRvravs zogadi kognitiuri procesebi da konkretuli

semantikuri niSnebi, sadac: pirvel safexurze modelebis SerCevas ganapirobebs

opozicia: “dinamika -statika”, meore safexurze ki: 1. dinamikur SinaarsTaA

qvesistemaSi mimarTulebisa da moqmedebis dasrulebulobis gansxvavebis

SesaZlebloba; xolo 2. statikur SinaarsTa qvesistemaSi – zmnis mTavari

argumentis mier mdgomareobis ganpirobebuloba. sqematurad es kognitiuri

procesi Semdegnairad SeiZleba warmovadginoT:

dinamika statika

(TemisniSniani modeli)M (meSvelzmniani modeli)

+mimarTuleba _mimarTuleba +‘avtotivi’ _‘avtotivi’

+dasrulebuloba _dasrulebuloba (SIII : -s) (SIII: -a)

(zmniswiniani) (uzmniswino)

dinamikuri vnebiTi/saSual-moqmedebiTi/saSual-vnebiTi/statikuri vnebiTi

6. daskvna

amgvarad, aqtiv-pasivis dapirispirebis kontinuumis saxiT warmodgena da

formaluri modelebis SerCevis ganmsazRvreli faqtorebis axsna lingvisturi

struqturirebis dinamikuri, kognitiuri procesebiT efeqturad asaxavs

aqtiur-pasiur formalur modelTa dapirispirebas. aseTi Teoriuli midgomiT

kidev erTxel dasturdeba is faqti, rom qarTulSi aqtiv-pasivis dapiris-

43

pirebis ganmsazRvleli aris ara informaciis struqturirebasTan dakavSire-

buli sintaqsuri faqtorebi, aramed kompleqsuri semantikuri niSnebi,

romlebic asaxaven am SinaarsTa lingvisturi struqturirebisas moqmed

specifikur kognitiur procesebsa da strategiebs. vfiqrobT, amgvari modgoma

naTels mohfens qarTulSi pasiuri modelebis gamoyenebis sxva Tavisebu-

rebebsac: e.w. gunebis vnebiTebi da deponensebi, potencialisisa da koleqtiuri

mniSvnelobiT naxmari vnebiTi gvaris formebi, uneburi moqmedebis gamomxatveli

vnebiTebi, TamaSobs-eTamaSeba tipis opoziciebi da sxv.

literatura 1. asaTiani 1990 _ r. asaTiani, gvaris morfologiuri kategoria qarTulsa da

qarTvelur enebSi, tipologiuri Ziebani, Tbilisi: “mecniereba”. 2001 _ R. Asatiani, Conceptual Structure of Reflexive and Middle, Proceedings of the

4th International Symposium on LLC, Amsterdam ILLC Scientific Publications, ed. Dick de Yongh.

2006 _ r. asaTiani, intonaciis roli winadadebis sainformacio struqturis

formirebaSi, saenaTmecniero krebuli, Tbilisi: Tsu gamomcemloba.

2007 _ r. asaTiani, informaciis struqturirebis sintaqsuri modelebi

qarTulSi. “semiotika-II“, i. WavWavaZis sax. universiteti. Tbilisi:

“universali”.

2007 _ R. Asatiani, The Main Devices of Foregrounding in the Information Structure of Georgian Sentences. Proceedings of Tbilisi Symposium on language, Logic and Computation 1005. Amsterdam: Spriger. 21-31.

2008 _ r. asaTiani, winadadebis sainformacio struqtura: qarTulSi vnebiTi

gvaris formaTa semantikur-kognitiuri interpretacia. g. kartoziasadmi miZRvnili saiubileo krebuli.

2. daviTiani 1973 _ ak. daviTiani, qarTuli enis sintaqsi. I. martivi winadadeba.

Tbilisi: “ganaTleba”.

3. diksoni 1979 _ R. M. W. Dixon, Ergativity. Cambridge: Cambridge Un. press. 4. enuqiZe 1981 _ l. enuqiZe, winadadebis aqtualuri danawevreba da misi mimarTeba

sintaqsuri da semantikuri analizis Tanamedrove meTodebTan Tanamedrove zogadi enaTmecnierebis sakiTxebi. VI. Tbilisi: enaTmecnierebis instituti.

5. kvaWaZe 1996 _ l. kvaWaZe,. Tanamedrove qarTuli enis sintaqsi. Tbilisi:

rubikoni.

6. kibriki 1997 _ A. Kibrik, Beyond Subject and Object: Toward a Comprehensive Relational Typology. Linguistic Typology.I.. Berlin-New York: Mouton de Gruyter. 279-346.

7. skopeteasi da sxv. 2006 _ Skopeteas, S., Fiedler I., Hellmuth R., Schwarz, A., Stoel, R., Fanselow, G., Féry, C., and Krifka M.,Questionnaire on Information Structure (QUIS). Interdisciplinary Studies on Information Structure 4. Working Papers of the SFB 632, Potsdam: Universitätsverlag Potsdam.

8. tuiti 1998 _ Tuite, K., 1998. Kartvelian morphosyntax. Munich: Lincom Europa. 9. SaniZe 1948 _ ak. SaniZe, qarTuli enis gramatika. II. sintaqsi. Tbilisi: Tsu

gamomcemloba.

1973 _ ak. SaniZe, qarTuli enis gramatikis safuZvlebi. Tbilisi: Tsu

gamomcemloba.

10. Ceifi 1971 _ W. L. Chafe, Meaning and the structure of language. Chicago and London: Chicago Un. Press.

44

11. Ciqobava 1968 _ arn. Ciqobava, martivi winadadebis problema qarTulSi. I. qvemdebare-damatebis sakiTxi Zvels qarTulSi. Tbilisi: ”mecniereba”.

12. harisi 1998 _ Al. C. Harris, Georgian Syntax: A Study in Relational Grammar. Cambridge etc.: Cambridge University Press.

Rusudan Asatiani

Information Structure of a Sentence: Peculiarities of Passive Constructions in the Kartvelian Languages

Abstract

The Kartvelian passive is quite different from the Indo-European one: In the Indo-European languages passive constructions are defined functionally: they are conversive of corresponding active constructions where patient is promoted to subject position, and agent is demoted and transferred into prepositional phrase. Kartvelian passive constructions do not always show conversion of an active ones. They are not simply defined by syntactic transformations; they are mostly governed by semantic peculiarities of a verb. Sometimes ‘passive constructions’ actually represent active semantics: dgeba ‘S/he is standing up’, ekačeba ’S/he pulls something’, ac’veba ‘S/he pushes something’ etc. It seems that in Kartvelian, as well as in some other languages, e. g. in Japanese [Shibatani, 1985], active-passive opposition forms a continuum where prototypical passive differs from so called middle forms. The peculiarities of Kartvelian passive define the restrictions of their usage in a process of information structuring. On the basis of an analysis of the semi-spontaneous data collected through the Questionnaire on Information Structure (QUIS is being developed within the Sonderforschungsbereich 632 “Information Structure” at the University of Potsdam and the Humboldt University Berlin funded by GSS [Skopeteas et all. 2006]), the passive constructions in Kartvelian are not defined by invisibility of agent and they do not always simply suppose changes of syntactic functions: When an invisible Agent is presented and the passive constructions are logically the most appropriate, Georgian, Megrelian and Svan informants prefer to produce active constructions with uncertain subject which is represented in verb forms by (1) either S.3.PL suffixes or (2) indefinite pronouns ‘somebody/something’ Thus, the analysis of the information structure of sentences in the Kartvelian languages gives one additional strong argument to interpret Kartvelian passive as a grammatical category mostly governed by semantic (and not by syntactic) features.

qeTevan grZeliZe

winadadebis sainformacio struqtura:

lokativebi qarTulSi

1. sakiTxis dasma

1.1. ena rogorc sakomunikacio saSualeba

enis umTavresi funqcia aris sakomunikacio. enis Seqmna sazogadoebis

CamoyalibebasTan mWidrod aris dakavSirebuli: rodesac gaCnda adamianTa

Soris informaciis gacvlis aucilebloba da survili, swored maSin Seiqmna

sociumi. sazogadoebaSi mimdinare cvlilebebi aisaxeba enaze. amitom vambobT,

rom ena sazogadoebrivi movlenaa.

ganviTarebis pirvel etapze enis funqcia, savaraudod, faqtebis martivi da

mSrali aRwera unda yofiliyo, magram droTa ganmavlobaSi sazogadoebis

moTxovnebi Seicvala. gaizarda pirovnebis roli, gaCndnen liderebi, Seiqmna

sxvadasxvagvari kultebi Tu religiebi, tomTa gaerTianebebi da qalaq-

saxelmwifoebi, Camoyalibda Taviseburi tradiciebi da maTs safuZvelze _

gansxvavebuli kulturebi.

sazogadod, adamiani eqceva im sazogadoebis rwmenisa da Cveulebebis

zegavlenis qveS, romelSic cxovrobs da mis mier realobis aRqma Sordeba

obieqturobas. garda amisa, misTvis ukve mniSvnelovania, ara marto gadasces an

miiRos informacia, aramed, saWiroebis SemTxvevaSi, Tavis sasargeblod

Secvalos igi. yovelive es aisaxeba enaze _ is ara mxolod sazogadoebis,

aramed individis samsaxurSicaa.

1.2. ena rogorc mamodelirebeli sistema

sazogadod, arsebobs erTi da imave informaciis sami saxe:

• pirveli arsebobs obieqtur realobaSi (A); • meore _ adamianis cnobierebaSi (es damokidebulia pirovnebaze, mis

ganaTlebaze, erovnebaze, rwmenaze da a. S.)(B); • da mesame formirdeba enis saSualebiT komunikaciisas (C). magaliTad, Tu gvaqvs raRac movlena A, is aRiqmeba rogorc B da

komunikaciisas SeiZleba gadaices rogorc C. B A C

45

46

realoba erTia, magram gansxvavebulia misi aRqma. zemoT moyvanil samkuTxedSi B zogadi aRmniSvnelia A-s usasrulo interpretaciebisa. aseve, C-c aris B-s enobrivi formulirebebis zogadi aRmniSvneli, magram enobrivi modelebis

raodenoba sasrulia. Tu erTsa da imave A-s SeiZleba Seesabamebodes

faqtobrivad usasrulo raodenobis B (B1, B2, B3, …Bn; sadac n usasruloa),

yovel calkeul B-s SeiZleba Seesabamebodes sasruli raodenobis C (C1, C2, C3, …Cm; sadacNm sasruli ricxvia). komunikaciis dros, SesaZloa, mosaubrem

informacia ise gadmosces, rogorc man aRiqva. am SemTxvevaSi SeiZleba viva-

raudoT, rom B=C. magram xSirad mosaubre uneblieT an mizanmimarTulad

ucvlis saxes informacias. am SemTxvevaSi SeiZleba arsebobdes C1, C2, C3, C4 da a.S.

informaciis saxecvlileba B-dan C-mde SeiZleba gamowveuli iyos ZiriTadad

ori mizeziT:

mosaubre uneblieT gadascems msmenels araobieqtur informacias,

anu ar awvdis iseTi saxiT, rogoriTac TviTon miiRo, radgan eqceva

emociuri an sxva faqtorebis gavlenis qveS;

mosaubres surs, rom msmenels miawodos araobieqturi an

subieqturi informacia, raTa gavlena moaxdinos masze.

calke SemTxvevaa msmenelis mier informaciis aRqma. am dros, isev subieqtur

faqtorTa gavleniT, msmenelma SeiZleba sxvadasxvagvarad aRiqvas Tundac

yvelaze obieqturi informacia, an saerTod ver gaigos, ras ambobs mosaubre.

es aris is SemTxveva, rodesac diskursis monawileTa Soris ar arsebobs

saerTo foni.

am SemTxvevebs qvemoT ganvixilavT. magram manamde mivubrundeT enis

mamodelirebel funqcias: ena aris erT-erTi faqtori, romelic ayalibebs Cvens

sinamdviles. adamianebi samyaros grZnobis organoebiT aRiqvamen, magram enob-

rivi mimarTebebi xSirad cvlis samyaros suraTs. Cven vcxovrobT enobrivi

informaciebis garemocvaSi. zogjer gvjera, rac gvesmis, zogjer _ ara, magram

mainc veqceviT enobrivi informaciis gavlenis qveS, radgan enaSi arsebobs

uamravi meqanizmi imisaTvis, rom Cven davijeroT. xSirad ena qmnis Cvens

samyaros.

1.3. winadadebis sainformacio struqtura

zemoT CamovTvaleT B-s cvlilebis mizezebi da aRvniSneT, rom enaSi

arsebobs meqanizmebi, romlebic emsaxureba mosaubris moTxovnebs diskursis

momentSi. magram ena uSualo dakvirvebaSi ar gveZleva. uSualo dakvirvebis

sagania metyveleba. yvela zemoT aRniSnuli enobrivi saSualeba realizdeba

47

swored metyvelebaSi. sametyvelo aqti ki Sedgeba konkretuli gamonaTqva-

mebisagan _ winadadebebisagan.

formalisti struqturalistebi amtkiceben, rom winadadebis analizisas

saWiroa Sinaarsis SeZlebisdagvarad ugulebelyofa da mTavaria gramatikuli

formebis sisworis kvleva. aseTi wminda formalisturi midgomis Sedegad

ikargeba winadadebis mravali semantikuri Tu pragmatikuli niuansi, radgan

winadadebis mTavari funqcia, iseve rogorc mTlianad enisa, aris informaciis

gadacema.

winadadeba ar aris mxolod sityvaTa meqanikuri jami. imisaTvis, rom man raRac

informacia gadmogvces, saWiroa mTeli rigi fonetikuri, morfologiuri,

sintaqsuri Tu leqsikuri saSualebebi, rac informaciis kodirebas, formalur

SefuTvas emsaxureba. informaciis struqturirebis dros informaciis

Semadgeneli garkveuli nawilebis gamosakveTad sxvadasxva enaSi sxvadasxva

lingvisturi modelebi SeiZleba Segvxvdes. yoveli ena irCevs misTvis misaReb

models da informaciis garkveuli nawilis gamosakveTad swored Tavisi

struqturis Sesabamis saSualebebs iyenebs.

magaliTi 1: Tu romelime enis sityvaTa rigis aramarkirebuli sintaqsuri

struqturaa SVO, maSin informaciis romelime nawilze logikuri maxvili

SeiZleba markirebuli OVS rigiT gamoixatos1: S V O O V S kaci wers werils. werils wers kaci.

am or winadadebas Soris gansxvaveba isaa, rom pirvel SemTxvevaSi informacia

naratiulia, msmenels faqtis ubralo aRwera miewodeba. winadadeba kaci wers werils neitraluri stilia, radgan is xazs ar usvams informaciis arc erT

nawils. magram winadadebaSi werils wers kaci markirebulia werili, rogorc informaciis yvelaze mniSvnelovani nawili konkretul momentSi. am SemTxvevaSi

mosaubres surs, rom msmenelma gansakuTrebuli yuradReba werils miaqcios. garda amisa, iqmneba STabeWdileba, rom msmenelisaTvis an msmenelTaTvis

cnobilia raRac damatebiTi informacia am faqtis Sesaxeb. am SemTxvevaSi,

sintaqsur poziciaTa cvlileba garkveulwilad artiklis rols asrulebs.

faqtobrivad, Cven vxedavT, rom erTi da igive faqti SeiZleba sxvadasxvanairad

gadmoices enobrivad da sxvadasxva zemoqmedeba iqonios msmenelze.

winadadebis sainformacio struqtura aerTianebs yvela im saSualebas,

romlebsac mimarTavs mosaubre garkveuli azris gadmosacemad. winadadebis

sainformacio struqturiT jer kidev antikuri xanis filosofosebi

interesdebodnen. magaliTad, aristoteles ,,ritorika~ aris swavleba mWevr-

1 OVS aris gramatikulad swori, magram SedarebiT iSviaTad gamoyenebuli konstruqcia Cveulebriv metyvelebaSi da, amdenad, is markirebulad miiCneva.

48

metyvelebis Sesaxeb, anu rogori Sinaarsi da struqtura unda hqondes

winadadebas, raTa man sasurveli zegavlena moaxdinos msmenelze.

winadadebis sainformacio struqturis Teoriuli analizi mxolod XX

saukunis 60-ian wlebSi daiwyo. C. Ceifi iyo erT-erTi pirveli, vinc am

sakiTxs miaqcia yuradReba. misi azriT, sainformacio struqtura informaciis

SefuTvas hgavs, romelic pasuxobs mosaubreTa sakomunikacio moTxovnebs

yovel konkretul SemTxvevaSi. C. Ceifi SefuTvaSi gulisxmobs iseT saSu-

alebebs, rogoric aris, magaliTad, sityvaTa rigis cvla (ix. zemoT

ganxiluli magaliTi 1.), Tumca m. krifkas azriT, termini ,,SefuTva~' srulad

ar asaxavs sainformacio struqturis mcvleli sxvadasxva formaluri

meqanizmis yvela funqcias, ramdenadac sxvadasxvagvari ,,SefuTvebi~ ara marto

formas, aramed Sinaarssac ucvlian informacias.

2. sainformacio struqturis saanalizod SemuSavebuli ZiriTadi cnebebi

imisaTvis, rom sainformacio struqtura SeviswavloT, saWiroa ganvmartoT

misi ganmsazRvreli ZiriTadi elementebi da TeoriaSi SemuSavebuli cnebebi da

terminebi.

2.1. upirveles yovlisa, ganvixiloT e. w. saerTo foni (Common Ground).

magaliTi 2: warmoidgineT, xalxiT gadavsebul avtobusSi xarT. am dros gesmiT: ,,skami!~ ras ifiqrebT? cxadia, mixvdebiT, rom viRacas dajdoma surs. am situaciaSi sityva ,,skami~ ekvivalentia winadadebisa ,,skamze dajdoma minda~. axla warmoidgineT, quCaSi rom igive sityva gaigonoT: Tqven CaTvliT, rom es sisulelea, an uadgilo da marTalic iqnebiT, radgan Cvens realobaSi quCaSi skamTan dakavSirebuli situacia naklebad warmosadgenia. albaT gagigoniaT gamoTqma: ,,mis sityvas fasi ara aqvs~, rac imas niSnavs, rom konkretuli mosaubris mier gadmocemuli informacia ar aris msmenelisaTvis sando, Rirebuli an situaciis Sesabamisi da mas iseve ar eqceva yuradReba, rogorc quCaSi warmoTqmul sityvas ,,skami!~.

magaliTi 3: j. lakofis da m. jonsonis wignSi ,,metaforebi, romlebiTac vcxovrobT~ aRwerilia kulturaTaSorisi gansxvavebebi, romlebic asaxaven saerTo fonebs Soris SesaZlebel sxvaobaTa ganmsazRvrel zogad warmodgenebs: magaliTad, SeiZleba romelime eris cnobierebaSi cekva ukavSirdebodes oms. Tu am eris warmomadgenlebTan moxvdeba adamiani, romlisTvisac cekva mxiarulebis simboloa, maSin maT Soris warmoiqmneba sakomunikacio velis rRveva da isini erTmaneTs ver gaugeben. am magaliTebidan Cans, rom komunikaciis dasamyareblad aucilebelia saerTo

konteqstis, anu saerTo fonis, arseboba. m. krifkas ganmartebiT, saerTo

foni aris is, rac aerTianebs mosaubresa da msmenels. es SeiZleba iyos

saerTo erovneba, kultura, religia, ganaTlebis garkveuli done da a. S.

49

Tu saerTo foni ar arsebobs, magaliTad, or ucxoels Soris, is unda

Seiqmnas enis daxmarebiT.

2.2. mocemuloba (Givenness) aris saerTo fonis is nawili, romelic

konkretuli mosaubrisaTvis ukve cnobilia da romelic, rogorc wesi,

SemosazRvravs saubris Temas da TandaTan, nabij-nabij farTovdeba saubrisas.

magaliTad, SeiZleba mosaubresa da msmenels bevri saerTo hqondeT, magram

konkretuli saubrisas mxolod erTi sakiTxi airCion da gaafarTovon Tavisi

,,codna~ am Temis Sesaxeb.

2.3. mocemulobis nawilia topiki. topiki aris saubris Tema _ is, vis an ris

Sesaxebac vsaubrobT da vagrovebT informacias, viZleviT komentars. m. krifkas

ganmartebiT:

topikuri Semadgeneli aris erTeuli an erTeulTa simravle, romlis Sesaxebac komentarSi gadmocemuli informaciaa mopovebuli. arsebobs sxvadasxva saxis topiki:

2.3.1. kontrastuli ewodeba topiks, romelsac diskursSi Semoaqvs ZvelTan

dapirispirebuli axali Semadgeneli.

magaliTi 4:

A: rogor ggonia, daTom gatexa Wiqa? B: ar vici. me (T) namdvilad ar gamitexavs.

2.3.2. nawilobrivi aris iseTi topiki, romelic gamoiyeneba im SemTxvevaSi,

roca mosaubres ar surs, an ar SeuZlia kiTxvaze pasuxis amomwuravad gacema

da gvawvdis moTxovnili informaciis mxolod raRac nawils.

magaliTi 5:

A: ras ityviT bavSvebis Sesaxeb? B: lika (T) TojiniT TamaSobs.

am SemTxvevaSi B-m ar icis (an ar unda, rom Tqvas) ras akeTebs meore bavSvi. kontrastuli da nawilobrivi topiki erTmaneTs Zalian hgavs. orive

gamoiyeneba im SemTxvevebSi, roca mosaubre (resp. mopasuxe, am SemTxvevaSi)

arasrulad pasuxobs. gansxvaveba imaSia, rom kontrastuli topiki gamoricxavs

erT konkretul SemTxvevas SesaZlo variantebidan. me namdvilad ar gamitexavs niSnavs, rom imaTSi, visac SeeZlo gaetexa Wiqa, me ar var. rac Seexeba

nawilobriv topiks, is mxolod sasurveli informaciis nawils gvawvdis.

2.3.4. implikaciuri topiki gamoiyeneba im SemTxvevaSi, rodesac mosaubres

surs, rom miniSnebis saSualebiT erTdroulad erTze meti informacia

mogvawodos.

50

magaliTi 6:

A: am kacs Svilebi hyavs? B: ai, mis Zmas ki nadvilad hyavs Svilebi. B-s pasuxi erTdroulad ramdenime informacias gvawvdis _ kacs Svilebi ar hyavs, mas hyavs Zma da am Zmas hyavs Svilebi. specifikuri konstruqcia (ai ... ki nadvilad) implikaciurad, arapirdapir gvawvdis kiTxvaSi formirebul

saZiebo informacias: kacs Svilebi ar hyavs. 2.4. saubris mizania topikze im informaciis Segroveba, romelic sainteresoa

konkretul SemTxvevaSi. yovel axal informacias, romelsac viRebT topikis

Sesaxeb, ewodeba fokusi. m. krifka fokuss ase ganmartavs:

fokusi miuTiTebs im alternativebis arsebobaze, romlebic SesaZlebelia enobriv gamonaTqvamTa interpretaciisas.

magaliTad, Tu vikiTxavT vin gatexa Wiqa? pasuxi SeiZleba iyos

sxvadasxvagvari: daTom, nikam, zuram, katam da a. S. yoveli maTgani aris

alternativa, romelic SeiZleba aRmoCndes pasuxSi da, Sesabamisad, is iqneba

fokusi.

fokusi, rogorc wesi, markirebulia, radgan axali informacia yvelaze

mniSvnelovania. magram saubrisas fokusi mudmivi arasodes ar aris. icvleba

ara marto topiki da fokusi, aramed mocemulobac da Sesabamisad, saerTo

fonic, radgan saubris mizani sworedac rom mocemulobis fokusebiT Sevseba

da saerTo fonis gafarToebaa.

magaliTi 7:

A: ra moxda? B: Wiqa gatyda(F). A: vin(F) [gatexa Wiqa] (T) (Giv)? B: daTom(F). A: vin(F) [aris daTo] (T) (Giv)? B: daTo(T) Cemi mezobelia(F).

arsebobs sxvadasxva saxis fokusi:

2.4.1. Verum fokusi (romelic SeiZleba iTargmnos rogorc damarwmunebeli

fokusi), gamoxatavs mkveTrad xazgasmul informacias. aseTi fokusi xSiria

inglisurSi da sailustaciod inglisur magaliTs movixmobT: I do like this music. It’s great. "do" zmna, Cveulebriv, neitralur gamonaTqvamebSi gramatikulad

Warbia, aq ki mas emfatikuri funqcia aqvs.

2.4.2 kompleqsuri (Complex) ewodeba iseT fokuss, romelic erTze met axal

Semadgenels aerTianebs. magaliTad:

51

magaliTi 8:

A: vin moiyvana daTom? B: [zura da lela](F). pasuxSi fokusi erT sityvaze ver daiyvaneba.

2.4.3. kontrastuli (Contrastive) fokusi gvaqvs iseT SemTxvevebSi, rodesac

kiTxvaSivea miTiTebuli is alternativebi, romelTaganac mopasuxem erT-erTi

unda airCios. magram SeiZleba isec moxdes, rom mopasuxem sulac ar airCios

arc erTi alternativa da sxva axali Semadgeneli Semoiyvanos. kontrastuli

fokusi uaxlovdeba topiks, radgan kiTxvaSi miTiTebuli alternativebi ukve

warmoadgens mocemulobis da, Sesabamisad, saerTo fonis nawils. isini

kiTxvaSia fokusi, magram pasuxSi ukve cnobili informaciis dakonkretebas

axdenen.

magaliTi 9:

A: Cais miirTmevT Tu yavas? B: yavas (an: Cais). magram pasuxi aseTic SeiZleba iyos: wyali mirCevnia, Tu SeiZleba. am

SemTxvevaSi wyali moulodneli fokusi iqneba, anu iseTi fokusi, romelic

kiTxvis formulirebiT sulac ar unda yofiliyo mosalodneli.

2.4.5. amomwuravi (Exhaustive) fokusi gviCvenebs, rom is aris erTaderTi swori alternativa da, rogorc wesi, gamoiyeneba amomwuravi mtkicebisas: mxolod daTom gatexa Wiqa. aseT SemTxvevaSi, fokusi Zlieri maxviliTaa gamoxatuli. iseT enebSi, sadac

maxvils an tonis simaRles fonologiuri Rirebuleba aqvs, fokusis

gamosaxatavad sxva, morfologiuri Tu sintaqsuri, saSualebebi iqneba saWiro. fokusi da topiki kiTxva-pasuxSi advili gasarCevia _ axali informacia (resp. fokusi), romelic avsebs saerTo fons, gansxvavdeba Zveli informaciisagan

(resp. topiki), romlis Sesaxebac vsaubrobT. magram fokusi arsebobs gabmul

teqstSic. magaliTad, zRaprebSi: iyo da ara iyo ra, iyo erTi mefe(F). mefes(T) hyavda sami asuli(F) da a.S. am SemTxvevaSi sainformacio struqturis analizisaTvis saWiroa aRvadginoT Sesabamisi savaraudo kiTxvebi. nebismieri

teqstis analizi SesaZlebelia sainformacio struqturis zemoT mocemul

Teoriul CarCoSi Camoyalibebuli terminebis daxmarebiT.

3. sainformacio struqturis mniSvneloba

rodesac zemoT aristoteles ,,ritorika~ vaxseneT, aRvniSneT, rom es aris

swavleba mWevrmetyvelebis Sesaxeb. aristotele gvaswavlis, rogori unda iyos

sityva, raTa gavlena iqonios msmenelze. igi gamoyofs sityvis ramdenime saxes

da TiToeul SemTxvevaSi _ saWiro enobriv saSualebebs. erTgan aristotele

ambobs: Tu romelime mxatvrul xerxs, magaliTad metaforas, Zalze xSirad

52

mivmarTavT, is dakargavs Tavis mxatvrul efeqts, anu msmenelze zemoqmedebis

unars. Tu yoveli calkeuli sityva iqneba markirebuli, maSin isini gane-

itraldebian da msmeneli maT gansakuTrebulad veRar aRiqvams. msmenelze

moqmedebs gamorCeuli da ara _ Cveulebrivi.

enis saSualebiT xdeba informaciis sasurveli SefuTva-struqturireba, rac

gamiznulia, raTa msmenels mosaubris mizandasaxulobis Sesabamisi informacia

miewodos.

rogor miiRweva es enaSi? filosofosma pol graisma gasuli saukunis 60-ian

wlebSi Camoayaliba efeqturi komunikaciisaTvis Rirebuli ZiriTadi maqsimebi

(e.w. graisis maqsimebi):

xarisxis maqsima: simarTle

ar Tqva is, risic ar gjera. ar Tqva is, raSic darwmunebuli ara xar.

raodenobis maqsima: informacia

Seni sityva unda moicavdes imden infor-

macias, ramdenic saWiroa konkretuli

saubrisaTvis.

ar Tqva imaze meti, vidre saWiroa.

Rirebulebis maqsima: Rirebuleba

Tqvi mxolod Rirebuli.

ragvarobis maqsima: simartive

Tqvi naTlad da martivad

eride orazrovnebas

Tqvi mokled

iyavi Tanmimdevruli

graisis maqsimebi mniSvnelovania im mxriv, rom isini miuTiTeben, Tu rogori

unda iyos saubris idealuri stili, magram saqme isaa, rom xSirad es maqsimebi

irRveva mizanmimarTulad. magaliTad, pirveli maqsimis darRveva yvelaze

xSiria, radgan arsebobs enobrivi saSualebebi, raTa simarTle daimalos ise,

rom msmenelTaTvis es SeumCneveli darCes. danarCeni maqsimebi xarisxis maqsimis

gavrcobas warmoadgens; magaliTad, zedmeti informacia da orazrovneba erT-

erTi xerxia imisa, rom Tavi avaridoT kiTxvas, romelzec pasuxi an ar

gvsurs, an ar viciT. sirTule da araTanmimdevrulobac aseT mizans emsaxureba.

bevri ucxo sityva da sintaqsuri konstruqciebis sirTule qmnis iseT STa-

53

beWdilebas, rom mosaubre ganaTlebulia da Wkviani, magram aseT stils umetes

SemTxvevaSi swored imitom mimarTaven, rom is niRbavs araRirebul infor-

macias da moCvenebiT Rirebulebas aniWebs mas. garda amisa, enas SeuZlia rea-

lobis iseTi interpretaciac, romelic msmenelSi garkveul emociebs aRZravs.

magaliTi 10:

mokvda _ neitraluri da odnav uxeSi;

gardaicvala _ pativiscemis gamomxatveli;

aRigava pirisagan miwisa _ amaRlebuli da arqauli;

gafrinda suli misi, rogorc mtredi _ mxatvruli;

fexebi gafSika, gaRma gavida da a. S. _ ironiuli da uxeSi.

magaliTi 11:

,,amartis ferad Secvala vardi cremlisa banaman~ (rusTaveli) _ cremlis

denam gaafiTra; saxe gauyviTlda.

,,dRem daixura pirbade, mTebma daxuWes Tvalebi~ (vaJa-fSavela) _ daRamda;

"The coldness of the moon had entered into them" (o. uaildi) _ (alublebi) civi

iyo. am magaliTebidan naTlad Cans, Tu rogor SeiZleba erTi da imave movlenis

sxvadasxvagvari enobrivi formulireba da damatebiT subieqturi niuansebis

miniWeba.

4. kvlevis miznebi da amocanebi

zemoT aRvniSneT, rom enas aqvs mamodelirebeli funqcia. samyaros pirvel

models qmnis Cveni xuTi ZiriTadi grZnoba. rodesac Cven vxedavT, magaliTad,

piramidas, misi ara obieqturi, aramed subieqturi xati aRibeWdeba Cvens

gonebaSi. magram, rodesac mis Sesaxeb gvesmis ise, rom arasodes gvinaxavs igi,

piramidis cneba ormagad subieqturia _ mas emateba rogorc Cveni, ise

mTxrobelis subieqturi STabeWdileba, romelic man Tavs mogvaxvia.

sainformacio struqturis codna gvexmareba, ganvsazRvroT adamianis

STabeWdilebebi ama Tu im saganze an movlenaze. Cveni kvlevis ZiriTadi

miznebia:

sainformacio struqturis terminebis daxmarebiT ganvsazRvroT,

Tu rogor xdeba realobis struqturireba da formalizeba;

ra enobriv _ fonetikursa Tu morfosintaqsur _ saSualebebs

mimarTavs mosaubre imisaTvis, rom aCvenos Tavisi damokidebuleba

realobisadmi;

54

rogor SeiZleba erTma da imave informaciamEsxvadasxva

mniSvneloba SeiZinos sainformacio struqturis sxvadasxvagvari

formirebis saSualebiT.

am zogad sakiTxebs ganvixilavT qarTulSi lokativebis Semcveli

struqturebis gaanalizebiT.

5. kvlevis meTodiB

Cveni kvleva efuZneba potsdam-berlinis proeqtis farglebSi SemuSavebuli

saeqsperimento amocanebis gadawyvetis Sedegad mopovebuli masalis analizs.

saeqsperimento amocana SemdegSi mdgomareobs: cdis pirebs miewodebaT

suraTebi, romlebzec asaxulia sxvadasxva sagani erTmaneTis mimarT sxvadasxva

poziciaSi. cdis pirebs evalebaT aRweron es cvlilebebi. isini ar unda iyvnen

winaswar informirebulni eqsperimentis miznebis Sesaxeb, radgan aseT

SemTxvevaSi maTi pasuxebi aucileblad motivirebuli iqneba da imoqmedebs e. w.

placebo efeqti2. es ar Sedis eqsperimentatoris interesebSi, ramdenadac

sainformacio struqturebis kvlevisaTvis gansakuTrebiT mniSvnelovania

spontanurad, maqsimalurad bunebriv pirobebSi warmoqmnili gamonaTqvamebi.

informantebis mier warmoTqmuli gamonaTqvamebis fonetikuri analizisaTvis

gamoviyeneT programa Praat, xolo masalis dasamuSaveblad programa EXMARaLDA.

6.Eeqsperimentis aRwera

cdis pirebs miewodebaT suraTebis seria. yoveli seria Sedgeba oTxi

suraTisgan, romlebzec gamosaxulia cxovelebi, sagnebi da adamianebi

erTmaneTTan sxvadasxva poziciaSi. seriebi isea Sedgenili, rom informaciulad

axal wevrs warmoadgens an lokativi an salokacio saxeli. cdis pirs evaleba,

aRweros es oTxi gansxvavebuli pozicia.

magaliTi 12:

2 fsiqologiuri termini: individma winaswar icis raime qmedebis Sedegis Sesaxeb da gaucnobiereblad moqmedebs ise, rom es Sedegi miRweul iqnes.

55

suraTebze erT SemTxvevaSi biWia xis win gamosaxuli, meore SemTxvevaSi ki _

gogo. cdis pirma unda aRweros cvlileba. mis mier warmoTqmuli winadadeba

iwereba programaSi Praat, romelic saSualebas gvaZlevs masala gavaanalizoT

sxvadasxvagvari fonetikuri parametrebiT:

biWi dgas xis win

gogo dgas xis win 7. masalis analizi da statistika

sul amgvari suraTebis 31 seria damuSavda. e. i. gasaanalizebelma masalam

Seadgina 124 winadadeba. winadadebebis zogadi struqtura aigeba sami ZiriTadi

elementis aucilebeli monawileobiT: salokacio saxeli (N), lokativi (L), da zmna (V); amdenad, Teoriulad SesaZlebeli modelebia: NVL, LVN, NLV, LNV, VNL, VLN. magram masalis analizma gviCvena, rom bolo ori modeli (zmniT

sawyis poziciaSi) saeqsperimento masalaSi ar dadasturda. es mosalodnelic

iyo, ramdenadac eqsperimentebis seria Tavidanve gamiznuli iyo mxolod N-sa

da V-s Soris arsebuli mimarTebebis gamosavlenad da ar moiazrebda zmnis

Time (s)0.0122387 2.64484

Pitch

(Hz)

75

5000.0122387 2.64484

GRG-RA-2-39-33_34_35_36

Time (s)2.63061 4.15324

Pitch

(Hz)

75

5002.63061 4.15324

GRG-RA-2-39-33_34_35_36

56

aqtualizaciis SemTxvevebs. darCenili modelebi sixSiris mixedviT (ix.

cxrili 1) gadanawilda Semdegi ierarqiis mixedviT: NVL>>LVN>>NLV>>LNV cxrili 1

suraTebis yovel seriaSi pirveli winadadebebi SeiZleba aramotivirebulad da

bunebrivad CavTvaloT, sadac neitralurad aris aRwerili mosaubris

damokidebuleba realobisadmi. magram misi strategia, e. w. naratiul

winadadebebSic, SeiZleba gansxvavebuli iyos; magaliTad, Tu mosaubre ambobs _ qvis win lomia _ misTvis mTavaria, Tu sad aris subieqti; anu, lokativia

fokusirebuli, salokacio saxeli ki topikalizebuli: is hyveba ambavs lomis Sesaxeb. imavdroulad is gvaZlevs axal informacias misi mdebareobis,

lokaciis Sesaxeb. meore modelSi ki _ lomi dgas qvis win _ piriqiT,

mosaubres ainteresebs vin/ra aris qvis win; anu: lokativia topikalizebuli,

salokacio saxeli ki fokusirebuli. unda aRiniSnos, rom, rogorc

statistikam gamoavlina, neitralur winadadebebSi topikalizebuli ZiriTadad

saxelia; anu: naratiul teqstebSi, rogorc wesi, informacias vagrovebT

situaciis ,,mTavari~, subieqturi wevris Sesaxeb. lokativi sawyis poziciaSi

aris mxolod im SemTxvevebSi, rodesac is situaciurad axalia (ix. cxrili

3). es kidev erTxel gvidasturebs da, amasTanave, gvixsnis, Tu ratom aris

qarTulSi dominirebuli, aramarkirebuli rigi _ SVO.

cxrili 2.

Semdeg suraTebSi cdis piri ukve aRwerda cvlilebebs, romlebic aucileblad

moiTxovdnen misgan garkveul markirebas _ fonetikurad (tonis aweviT),

morfologiurad (mag., ,,a~ klitikis an ,,c~ nawilakis darTviT), an

sintaqsurad (sityvaTa rigis cvliT).

magaliTi 13: biWi dgas xis win. gogo dgas xis win. xis gverdze dgas gogo. saxlis gverdze dgas gogo.

modeli

NVL LVN LNV NLV VNL VLN

sixSire 62 40 4 18 0 0

neitraluri winadadebebis

raodenoba

N sawyis

poziciaSi

L sawyis

poziciaSi

31 21 (68%) 10 (32%)

57

NVL; NVL; LVN; LVN _ am dadasturebuli modelebidan meoreSi axali

informacia (gogo) markirebulia mxolod fonetikurad _ intonaciuri

konturis talRovani cvliT (aRmavali-daRmavali), anu markirebuli toniT

gamoirCeva gogo). mesame winadadebaSi axali lokaciis markireba xorcieldeba

sintaqsurad _ NVL rigis cvliT markirebuli LVN rigis saxiT. am sintaqsur

cvlilebas axlavs intonaciuri konturis cvlac _ tonis aweva axasiaTebs

lokaciis maCvenebel mTel frazas: xis gverdze, ramdenadac es mTliani fraza

axali adgilmdebareobis maCvenebelia. meoTxe winadadebaSi informaciis cvla

kvlav mxolod tonis aweviT gamoixateba: tonis aweva xdeba sityvaze saxlis. axali informacia araneitalur winadadebebSi Semdegi ierarqiis Sesabamisad

gadanawilda (ix. cxrili 3): axali N >> axali L

cxrili 3.

araneitralur winadadebebSi (93 SemTxveva) erTi SexedviT Cans, TiTqos axali

lokativebi ufro xSirad ikaveben pirvel pozicias. magram es ase ar aris. Tu

am cxrils davukvirdebiT, davinaxavT, rom eqsperimentSi axal saxelTa

raodenobas sWarbobs axal lokativTa raodenoba. amitom, procentuli

analizis mixedviT, lokativi sawyis pozicias ikavebs 61-dan 55%-Si, xolo

saxeli 32-dan 67%-Si. garda amisa, lokativi ufro midrekilia bolo

poziciisaken, vidre saxeli _ 61-dan 38%-Si lokativi bolo adgilzea.

saxeli bolo poziciaSi mxolod SemTxvevaTa 15%-Si gvxvdeba. miuxedavad

imisa, rom magaliTebis raodenoba didi ar aris, mainc SeiZleba vimsjeloT

zogad tendenciebze da davaskvnaT: saxelebi bunebrivad midrekilia sawyisi

poziciisaken, lokativi sawyis poziciaSi gvxvdeba mxolod im SemTxvevebSi,

rodesac is axalia (resp. fokusirebuli). e.i. struqturirebisas moqmedebs ra ori ierarqia: (1) N >> L da (2) axali >> Zveli (ix. cxrili 2) _

,,gamarjvebuli~ gamodis (1). am ori ierarqiis SejerebiT vRebulobT axal

ierarqias: axali N >> axali L, rac realurad dadasturda masalis

procentuli analiziT (ix. cxrili 3). rac Seexeba zmnebs, isini sawyis

poziciaSi ar gvxvdeba, radgan, rogorc zemoT ukve SevniSneT, eqsperiments

ainteresebs mxolod LN da NL mimarTebebi; garda amisa, zmna sawyis poziciaSi metad mxatvrulia da literaturul stils ufro Seefereba vidre sasaubros.

pozicia

informacia

I II III sul

Aaxali

informacia

55 10 28 93

axali L

34(55%) 4(7%) 23(38%) 61(66%)

axali N

21(67%) 6(18%) 5(15%) 32(34%)

58

masalis mixedviT, zmna yvelaze xSiria meore poziciaSi, lokativsa da

salokacio saxels Soris (ix. cxrili 4):

cxrili 4.

rodesac zmna boloSia, SemTxvevaTa did umravlesobaSi is warmodgenilia

meSveli zmniT, ZiriTadad ,,a~ klitikis saxiT, romelic umetesad uerTdeba

salokacio saxels (ix. cxrili 5):

cxrili 5.

miuxedavad imisa, rom saanalizo masala Seadgenda mxolod 124 winadadebas, N-sa da L-s Soris arsebuli mimarTebebis ganmsazRvreli ZiriTadi modelebi da

garkveuli tendenciebi saxezea, rac saSualebas gvaZlevs daskvnebis saxiT

CamovayaliboT ZiriTadi debulebebi.

8. daskvnebi

eqsperimentebis Sedegad saanalizod mopovebulma masalam gviCvena, rom wina-

dadebis struqturireba ZiriTadad damokidebulia individsa da konkretuli

situaciis miseul aRqmaze: konkretuli situaciidan gamomdinare, ZiriTadad

mosaubre wyvets informaciis axal-Zvel informaciad kvalifikaciis stra-

tegias. masalis statistikuri analizis mixedviT, aRmoCnda, rom neitralur

SemTxvevaTa umravlesobaSi (31-dan 21-Si, anu 68%-Si), salokacio saxelia

pozicia

zmna

II III sul

damoukidebeli zmna 92 6 98

meSveli zmna 7 19 26

meSveli

zmnis

klitika

damoukidebeli

meSveli zmna

sul

NLV 0

1 1

LVN 0

4 4

LNV 16

2 18

NVL 1

2 3

59

pirvel adgilze, rac imas niSnavs, rom mosaubre Tvadapirvelad saubris Temis

(resp. topikis) kvalifikacias salokacio saxels (resp. subieqts) aniWebs. es ki, ramdenadac qarTulSi topikuri Semadgeneli miiswrafvis winadadebis

sawyisi poziciisaken, Tavis mxriv, ganapirobebs imas, rom dominanturi, ara-

markirebuli struqturuli rigi aris SVO. garda amisa, araneitraluri

winadadebebis analizma cxadyo, rom salokacio saxeli, rogorc fokusi, ufro

xSirad ikavebs sawyis pozicias, vidre fokusirebuli lokativi. is iSviaTad

gvxvdeba winadadebis boloSi, xolo winadadebis SuaSi misi gansakuTrebuli

markireba xorcieldeba ,,a~ klitikis darTviT. topikisa da fokusis lingvis-

turi warmodgenis ZiriTadi meqanizmia fonetikuri xasiaTis cvlilebebi,

romlebsac SeiZleba axldes sintaqsuri da morfologiuri konstruqciebis

cvla. sxvadasxvagvar zedapirul reprezentaciebs gansazRvravs lokativebis

Semcveli konstruqciebis agebisa da situaciis struqturirebis ganmapirobe-

beli Semdegi, kognitiuri xasiaTis, ierarqiuli mimarTebebi:

1. N >>L 2. axali informacia >> Zveli informacia

3. axali N >> axali L

rac, Tavis mxriv, ganapirobebs formalur modelebs Soris arsebul Semdeg

ierarqiul mimarTebebs:

1. NVL>>LVN>>NLV>>LNV

da, aseve, salokacio saxelis morfologiur markirebas meSveli zmnis

klitikis saSualebiT:

2. meSveli zmnis klitika >> damoukidebeli meSveli zmna

9. perspeqtivebi

warmoebulma kvlevam warmoaCina winadadebis sainformacio struqturis Ses-

wavlis Rirebuleba _ amgvari analizi mniSvnelovania rogorc zogadi enaTmec-

nierebis Teoriuli sakiTxebis gadasawyvetad, ise konkretul enaTa Seswavli-

saTvis. amgvari analizi saSualebas iZleva Teoriul safuZvelze SeviswavloT,

Tu rogor xdeba realobis struqturireba sxvadasxva enis specifikuri stra-

tegiebisda Sesabamisad da davadginoT gansxvavebebi da msgavsebebi enebs Soris.

radgan ena xalxis msoflmxedvelobis gamoxatulebaa, enobrivi gansxvavebu-

lobebis analizi garkveul informacias gvawvdis xalxTa kulturulogiur

Taviseburebebze. amasTanave, praqtikuli TvalsazrisiT, winadadebis sain-

formacio struqturis Seswavla daexmareba redaqtorebs, mTargmnelebs, masme-

diis muSakebsa da, sazogadod, teqstze momuSave nebismier pirs, konkretuli

mizandasaxulobidan gamomdinare, informaciis sasurveli saxiT aqtualizaciis

Sesabamisi modelis SerCevasa da adekvaturi teqstis agebaSi.

60

literatura aristotele, ritorika, `Tsu gamomcemloba~, Tbilisi, 1981.

r. asaTiani, martivi winadadebis tipologiuri analizi, Tanamedrove qarTuli

saliteraturo enis masalaze, , gamomcemloba `mecniereba~, Tbilisi,

1982.

r. asaTiani, intonaciis roli winadadebis sainformacio struqturis

formirebaSi, ènaTmecnierebis sakiTxebi~, gamomcemloba ùniversali~,

Tbilisi, 2006.

r. asaTiani, informaciis struqturirebis sintaqsuri modelebi qarTulSi,

`semiotika-II~, ilia WavWavaZis sax. universiteti, gamomcemloba

ùniversali~, 2007. Asatiani R., The Main Devices of Foregrounding in the Information Structure of

Georgian, The 6th International Symposium on LLC, Tbilisi-2005; Balden D.ten Cate&Henk W. Zeevat (Eds), Springer. 21-31. 2007.

Asatiani R., Word order and intonation in Georgian. (C. Fery and S. Scopeteas, co-authors),Lingua XXX, 1492, Elsevier, 1-26.2008.

S. afridoniZe, sityvaTganlageba axal qarTulSi, gamomcemloba `mecniereba~,

Tbilisi, 1986. Grice P., Logic and Conversation, in Cole, p., and Morgen, J.L. (eds.), Syntax and

Semantics, Volume 3, Speech Acts, pp. 41-58, New York, 1975 m. krifka, sainformacio struqturis ZiriTadi cnebebi (Targmna e.

soseliam). ena, logika, kompiuterizacia, II. Tbilisi: gamomcemloba

`degaprinti~ 7-57, 2008. Skopeteas S., Fiedler I., Hellmuth R., Schwarz, A., Stoel, R., Fanselow, G., Féry C.,

and Krifka M., Questionnaire on Information Structure (QUIS), Interdisciplinary Studies on Information Structure 4. Working Papers of the SFB 632, Potsdam: Universitätsverlag Potsdam, 2006.

Skopeteas, S., Fanselow, G., Effects of givenness and constraints on free word order. In Information Structure from different perspectives. Zimmermann, Malte and Caroline (eds.),Oxford Un. Press, 2007.

Chafe W. L., Meaning and the structure of language. Chicago and London: Chicago Un.Press. 1971.

n. wereTeli, r. asaTiani, winadadebis sainformacio struqtura:

implikaciuri topiki qarTulSi, ènaTmecnierebis sakiTxebi~, ~Tsu gamomcemloba~, Tbilisi, 2008.

61

Ketevan Grdzelidze

Sentence Information Structure: Locatives in Georgian

Abstract

Since ancient times there has been a great interest to what kind of information a sentence can convey. Many philosophers wrote that language can attract and influence, persuade or cheat the listener. But it was only in the second half of the XX century that the scientific studies of sentence information structure began. Analysing the sentence in linguistic terms has revealed and proved that language is individual and social at the same time. These studies help us to go deep to human perception of reality step by step and explain the secret of the magic power of language. The following article is about locatives in Georgian. Locative is a constituence of a sentence which indicates to the location of a subject. Studying locatives helps us to find out how the location of different objects is percepted. Analysing the data of 124 sentences we concluded that in most cases new information prevails the given one and that locatives, either new or old, tend to the final position of a sentence. We carried out our ‘field sessions’ according to Potsdam-Berlin experiment tasks. For analising we used the following programmes: Praat, EXMARaLDA.

62

naTia dundua sainformacio struqturis ganmsazRvreli formaluri modelebi

Tanamedrove sparsulSi

Sesavali: ZiriTadi cnebebi


ZiriTadi odenobebia. orive winadadebis aqtualizebul wevrs aRniSnavs:

fokusi – adre ucnob, axali informaciis Sesabamis aqtualizebul wevrs,

topiki – ukve cnobils, Zveli informaciisas.

formaluri TvalsazrisiT fokusi da topiki SeiZleba markirebul iqnen

sxvadasxva enobriv doneebze: (1) fonetika-fonologiis (intonacia, maxvili,

toni, specifikuri fonetikur-fonologiuri procesebi: gamJRereba-dayrueba,

asimilacia-disimilacia da sxv.); (2) morfologia-sintaqsisa (specialuri

brunva, Serwymuli nawilaki an klitika, sityvaTa rigis cvla, elifsisi,

specifikuri konstruqciebi, frazebis gagleja, tmesi da sxv.); da (3) leqsika-

semantikis (specialuri sityvebi, nawilakebi, artikli, kvantoruli sityvebi:

mxolod, marTlac, namdvilad, aseve, kidec da sxv.).

yvela zemoT CamoTvlili SesaZlebloba xSirad Tanaarsebobs informaciis

“SefuTvis” dros – warmoiqmneba rTuli, kopleqsuri SemTxvevebi da, amdenad,

rTuldeba formaluri analizic. yoveli ena Tavisi strategiis mixedviT

axorcielebs topik-fokusisa Tu sxva aqtualizebuli wevris formalizebas.

fokusisa da topikis formaluri modelebi

fokusisa da topikis Sesabamisi formaluri modelebis sakvlevad mimarTaven

ganmeorebiTi kiTxvebis dasmis meTods, risi meSveobiTac vlindeba sxvadasxva-

gvari nawilakebi, specifikuri sintaqsuri konstruqciebi da intonacia.

I. sxvadasxvagvari fokusis markirebis magaliTebi:

(1)- če qār mišavad? (ra xdeba?) - mard xāne rā misāzad. (kaci saxls aSenebs.)

(2)- ki misāzad xāne rā? (vin aSenebs saxls?)

- mard misāzad xāne rā. (kaci aSenebs saxls.)

(3)- mard či misāzad? (kaci ras aSenebs?) - xāne rā misāzad mard // mard xāne rā misāzad. (saxls aSenebs kaci//kaci saxls aSenebs.)

(4)- če kār mikonad mard? (ras akeTebs kaci?) - mard xāne rā misāzad // mard misāzad xāne rā // misāzad mard xāne rā. (kaci saxls aSenebs // kaci aSenebs saxls // aSenebs kaci saxls.)

63

(5)- ki či misāzad? (vin ras aSenebs?) - mard xāne rā misāzad // mard misāzad xāne rā // xāne rā mard misāzad // xāne rā misāzad mard. (kaci saxls aSenebs// kaci aSenebs saxls // saxls kaci aSenebs // saxls aSenebs kaci.) (6)- mard misāzad xāne rā? (kaci aSenebs saxls?) -bale, mard misāzad xāne rā // bale, in mardist ke xāne rā misāzad // vāq’e’an in mard ast ke xāne rā misāzad // rāsteš in mard ast ke xāne rā misāzad. (diax, kaci aSenebs saxls // diax, es kacia, vinc saxls aSenebs // namdvilad (rom) kaci aSenebs saxls // marTlac (rom) kaci aSenebs saxls.)

(7)- mard xāne rā misāzad? (- kaci saxls aSenebs?) - bale, xāne rā misāzad mard // bale, mard xāne rā misāzad // vāq’e’an in xāneist ke mard an rā misāzad... (diax, saxls aSenebs kaci // diax, kaci saxls aSenebs // marTlac rom saxlia is, rasac kaci aSenebs...) (8)- mard xāne rā misāzad? (kaci saxls aSenebs?) - bale, misāzad mard xāne rā // bale, mard xāne rā misāzad // vāq’e’an misāzad mard xāne rā. (diax, aSenebs kaci saxls // diax, kaci aSenebs saxls // marTlac (rom) aSenebs kaci saxls.)

da a.S.

zemoT ganxiluli masalis mixedviT fokusis markirebisTvis gamoiyofa Semdegi

modelebi:

a) markirebuli intonacia;

b) sityvaTa rigis Secvla (fokusis win gadasma) (+intonacia)

g) specifikuri sintaqsuri konstruqciebi (+intonacia)

d) nawilakebi (+sintaqsi +intonacia)

II. topikis markirebis nimuSebi:

(1)- če mitāvanid dar bāreie mard ra beguid? (ra SegiZliaT TqvaT kacis Sesaxeb?) - mard xāne rā sāxt. (- kacma aaSena saxli.)

(2)- dar bāreie xāne čizi midānid? (iciT raime saxlis Sesaxeb?) - xāne rā sāxt mard // xāne rā mard sāxt. (saxli aaSena kacma // saxli kacma aaSena.)

(3)- če mitāvanid dar bāreie in mard ra beguid? (ras ityviT am kacis Sesaxeb?) - in mard sāxt xāne rā // vāq’e’an in mard sāxt xāne rā // in mardist ke xāne rā sāxt // rāsteš in mard bud ke xāne rā sāxt. (am kacma aaSena saxli // namdvilad am kacma aaSena saxli // es is kacia, vinc saxli aaSena //marTlac (rom) es kaci iyo, romelmac saxli aaSena.)

64

rogorc vxedavT, topikis gamosaxatavad gamoiyo Semdegi modelebi:

a)markirebuli intonacia

b) sityvaTa rigis Secvla (topikis win gadasma da/an zmnis wina poziciaSi

dasma) (+intonacia)



ZiriTadi gansxvaveba fokusisa da topikis modelebs Soris intonaciaSia.

sityvaTa rigi

informaciis struqturirebisas markirebis erT-erTi saSualebaa sityvaTa

rigis cvla, rac damaxasiaTebelia e.w. “Tavisufali” sityvaTa rigis

enebisTvis, anu mdidari morfologiuri struqturis (resp. ganviTarebuli

afiqsaluri sistemis) mqone enebisTvis.

sparsulSi winadadebis sityvaTa rigi Tavisufalia. statistikurad yvelaze

gavrcelebulia SOV mimdevroba. yvela SesaZlo gadaadgileba amosavali ara-

markirebuli rigidan SesaZloa dakavSirebuli iyos winadadebis sainformacio

struqturasTan.

sityvaTa rigis TvalsazrisiT sainteresoa pirdapiri obieqtis pozicia

winadadebaSi. gansazRvruli obieqtis (resp. pirdapiri obieqtis) bazisuri

pozicia, romelic markirebulia gansazRvrulobis markeri Tandebuli rā-Ti, aris gansazRvruli zmnuri fraza (VP). ganusazRvreli obieqti (resp. iribi obieqti) gansxvavebulia sintaqsuri, semantikuri da morfologiuri maxasia-

TeblebiT da, amdenad, aqvs gansxvavebuli bazisuri, uSualod zmnis winamavali,

pozicia.

frazis struqturis wesebis mixedviT miiReba ori rigi(Erteschik-Shir, 2007): a. [VPDP[+specific] [v’ PPV]] b. [VP [V’PP [DP [-specific] V]]]

gansazRvruli pirdapiri obieqti SeiZleba gadaadgildes zmnizedis win, rac

imis maCvenebelia, rom is gadis zmnuri frazis (VP) gareT:

(1)Kimea porsid [CP ke [ketāb-rā]i diruz [VP Sepide [VP ti kojā gozāšt]]] Kimea-NOM ask-S3.SG. that book-rā yesterday Sepide-NOM where put-S3.SG Lit.: Kimea asked, as for the book, where did Sepide put (it) yesterday. “qimeam ikiTxa, Tu sad dado guSin sefidem wigni.” (2) Kimea porsid [CP ke [ketāb-rā]i diruz [VP Sepide [VP ti be ki dād]]] Kimea-NOM ask-S3.SG. that book-rā yesterday Sepide-NOM to who-DAT give-S3.SG.

65

Lit.: Kimea asked, as for the book, to whom did Sepide give (it) yesterday? “qimeam ikiTxa, Tu vis misca wigni guSin sefidem.”

rogorc sityva-sityviTi Targmanidan Cans, gadaadgilebuli obieqti interpre-

tirebulia topikad. gansazRvruli da ganusazRvreli obieqtebis gadaadgileba

SesaZlebelia, Tuki maTze modis maxvili da is kontrastulad aris gansaz-

Rvruli. es aris naCvenebi ganusazRvreli obieqtisTvis Semdeg magaliTSi:

(3) Kimea [(ye) ketāb]i barā -š ti xarid. Kimea-NOM (one) book-NOM for him buy-S3.SG. Lit.: Komea bought a BOOK for her. “qimeam erTi wigni misTvis iyida”.

aqedan Cans, rom Tumca ganusazRvreli aris axali informacia, is kvlavac

SeiZleba gadaadgildes, rodesac kontrastulia, radgan kontrasti gulis-

xmobs diskursul xazgasmas. kontrastuliwevrebi winadadebisa bunebrivad

eqceva topikebis klasSi. aqedan, is faqti, rom araspecifikuri kontrastuli

elementebi SeiZleba topikalizdes, ar aris gasakviri.

topikalizaciis sintaqsurad Warbi markerebi

topikalizaciis erT-erTi saSualebaa aseve sintaqsurad Warbi nacvalsaxelebis

dafiqsireba; magaliTad, [man]T raftam ([me]T wavedi), [man]T xaridam nān rā ([me]T viyide puri), [man]T bordam nān rā be madāram ([me]T wavuRe dedas puri)... aq man (me) Warbia sparsuli enis specifikidan gamomdinare, radganac pirveli

piri morfologiurad warmodgenilia zmnur formaSi pirveli subieqturi

piris markeriT (-am), misi dublireba sintaqsur konstruqciaSi man (me) nacvalsaxelis saxiT Warbia da migvaniSnebs pirveli piris topikalurobaze:

man raftam va na kasi digar (me wavedi da ara viRac sxva)...

sintaqsurad Warb markerad miCneulia winadadebis zogierT wevrTa gameorebac.

Cveulebriv, meordeba winadadebis aqtualizebuli wevri:

(1) to... to injāi... Sen... Sen aq xar... (sadeq’ čubak, “tangsir”) (2) sedaie... sedaie do’a miumad. xma, locvis xma modioda. (sadeq’ čubak, “tangsir”) (3) un do ta, un do ta ra migam, un do ta ra bebin. im orze, im orze geubnebi, im ors Sexede. (ebrahim golestān, “māhi o jofteš”) (4) āmār, ‘zize man, āmār. აmar, Zvirfaso, amar. (ebrahim golestān, “ba pesaram ruie rāh”) (5) kebrit, kebrit dāram, xāheš mikonam bexarid! asanTi, asanTi maqvs, gTxovT,

iyideT!

(6) āi, kebrit dāram, kebrit... ei, asanTi maqvs, asanTi... (“doxtarake kebritforuš”)

66

informaciis struqturireba da punqtuacia

Tanamedrove sparsul enaSi informaciis struqturirebisas gansakuTrebiT

sainteresod gveCveneba punqtuaciis roli. kerZod, xSir SemTxvevaSi werisas

aqtualizebuli wevris xazgasasmelad ixmareba mZime:

(1) dar xāne, kebrit ziād dāram. saxlSi, bevri asanTi maqvs.

(2) doxtarake kebritforuš, dar xiābānhāie sard o porbarf migozašt. asanTis

gamyidveli gogona, civ da Tovlian quCebSi dadis (“doxtarake kebritforuš”).

daskvna

• fokusi da topiki winadadebis sainformacio struqturis

ganmsazRvreli ZiriTadi odenobebia. orive winadadebis aqtualizebul

wevrs aRniSnavs: fokusi – adre ucnob, axali informaciis Sesabamis

aqtualizebul wevrs, topiki – ukve cnobils, Zveli informaciisas.

yoveli ena Tavisi strategiis mixedviT axorcielebs topik-fokusisa

Tu sxva aqtualizebuli wevris foრmalizebas.

• Tanamedrove sparsulSi fokusis markirebisTvis gamoiyofa Semdegi

modelebi:

a) markirebuli intonacia;

b) sityvaTa rigis cvla (fokusis win gadasma) (+intonacia)



• topikis gamosaxatavad gamoiyo Semdegi modelebi:

a) markirebuli intonacia

b) sityvaTa rigis cvla (topikis win gadasma da/an zmnis wina

poziciaSi dasma) (+intonacia)



• sparsulSi winadadebis sityvaTa rigi Tavisufalia. statistikurad

yvelaze gavrcelebulia SOV mimdevroba. yvela SesaZlo gadaadgileba

amosavali aramarkirebuli rigidan SesaZloa dakavSirebuli iyos

sainformacio struqturasTan.

• sityvaTa rigis TvalsazrisiT sainteresoa pirdapiri obieqtis pozicia

winadadebaSi. gansazRvruli obieqtis (resp. pirdapiri obieqtis)

bazisuri pozicia, romelic markirebulia gansazRvrulobis markeriT

(Tandebuli rā-Ti), SemosazRvrulia zmnuri fraziT (VP). • topikalizaciis erT-erTi saSualebaa aseve sintaqsurad Warbi

nacvalsaxelebis dafiqsireba.

• sintaqsurad Warb markerad miCneulia winadadebis zogierT wevrTa

gameorebac. Cveulebriv, meordeba winadadebis aqtualizebuli wevri.

67

• Tanamedrove sparsul enaSi informaciis struqturirebisas

gansakuTrebiT sainteresod gveCveneba punqtuaciis roli. kerZod, xSir

SemTxvevaSi werisas aqtualizebuli wevris xazgasasmelad ixmareba

mZime.

samomavlo perspeqtivebi

Catarebuli kvlevis safuZvelze aSkarad ikveTeba intonaciis gadamwyveti

roli aqtualizebuli wevris markirebaSi. momavalSi vfiqrobT eqsperimenti

CavataroT sparsuli enis matareblebze, rac neitraluri, fokusirebuli da

topikalizebuli nawilebis intonaciuri konturis fonetikuri analizis

saSualebas mogvcems.

literatura asaTiani, r., intonaciis roli winadadebis sainformacio struqturis

formirebaSi, enaTmecnierebis sakiTxebi, I-II, Tbilisi, 2006.

ivaniSvili, m., winadadebis sainformacio struqtura: sityvaTa rigi

qarTvelur enebSi, Tbilisi, 2009 (ix. amave krebulSi). Erteschik-Shir, N., Information Structure, Scrambling in Persian, Oxford University

Press, 2007. Karimi, S., On Object Positions, Specificity, and Scrambling in Persian, Blackwell

Publishing, Malden, 2003. Robert,s J.R., Scrambling in Persian: An RRG Approach, SIL International. http://linguistics.buffalo.edu/research/rrg.html, 2005. Natia Dundua

The Main Formal Models of Informational Structure in Modern Persian

Abstract

The word order in modern Persian is free, however, statistically the most widespread order is SOV. Each possible reordering from the basic unmarked word order can be the result of the various information structure of a sentence. Paper examines some models of topicalization: The basic position of specific objects, marked by the specificity marker -râ, is {SpecO,V}, which is changed in case of topicalization and SpecO does not buid the VP anymore. One way of the topicalization is the representation of syntactically redundant pronouns in a sentence.

68

Syntactically redundant entities as markers of topicalization can be found in sentences where the recurrence of some members takes place. Usually the actualized member of the sentence is repeated. In modern Persian while structuring the information the role of punctuation is very important. Particularly, topicalization in written texts is represented usually by comma.

69

marine ivaniSvili

winadadebis sainformacio struqtura:

sityvaTa rigi qarTvelur enebSi

1. Sesavali

bunebrivi enis universaluri maxasiaTebeli aris enobrivi sistemis bazisuri,

amosavali viTarebidan fakultaturuli gadaxris SesaZlebloba, rac

gansakuTrebiT TvalsaCinoa informaciis sasurveli saxiT struqturirebis

dros. SemadgenelTa wrfivi mimdevroba ganisazRvreba imis mixedviT, Tu

konteqsturad ra aris cnobili da ra _ ara. imis axsnas, Tu ra meqanizmi

udevs safuZvlad informaciis struqturirebas, wlebis manZilze cdilobdnen

fonetikur-fonologiuri, morfo-sintaqsuri, leqsikuri terminebisa Tu enob-

rivi sistemis semantikuri da pragmatikuli donis erTeulebis safuZvelze,

mimarTavdnen Sereuli midgomis xerxebsac. droTa viTarebaSi ixveweboda

terminologia, zustdeboda meTodologia. dRes am mimarTulebiT gamoqvey-

nebulia mravali Sroma. miRebuli Sedegebi metad mniSvnelovania kompiute-

ruli lingvistikis, xelovnuri inteleqtis, fsiqolingvistikis, nevrolin-

gvistikisa da enaTmecnierebis sxva Tanamedrove dargebisTvis.

informaciis aqtualizebisTvis sxvadasxva enaSi sxvadasxvagvari saSualebebi

arsebobs. yoveli ena irCevs misTvis misaReb strategias da informaciis

garkveuli nawilis gamosakveTad swored Tavisi struqturis Sesabamis

saSualebebs iyenebs.


ZiriTadi cnebebia. orive aRniSnavs winadadebis dawinaurebul wevrs: fokusi _

axali informaciis Semcvels, topiki _ ukve cnobilis. fokusisa da topikis

markirebis erT-erT models morfologia-sintaqsis doneze warmoadgens

sityvaTa rigis cvla.

2. sityvaTa rigi: mokle eqskursi

sityvaTa rigis mixedviT enebs yofen Tavisufali rigisa da sityvaTa

gansazRvruli rigis enebad. amgvari dayofa pirobiTia, radgan TiTqmis ar

arsebobs sityvaTa rigis mixedviT absoluturad Tavisufali an absoluturad

gansazRvruli enebi.

sityvaTa rigis SeswavlisaTvis arsebobs sul mcire ori tradicia:

`grinbergiseuli~ (Greenberg 1963), romelic dafuZnebulia im varaudze, rom

yovel enaSi arsebobs neitraluri, bazisuri, aramarkirebuli,

gramatikalizebuli sityvaTa rigi (Sdr. Hawkins 1983, Tomlin 1986, Dryer 1992), meore tradicia, `praRuli~ _ dakavSirebulia Cexi lingvistebis SromebTan

(Mathesius 1929/1983, Firbas 1964), romlebic ikvlevdnen aragramatikulad

gansazRvrul sityvaTa rigs slavur enebSi da xsnidnen mas `Tema/rema~ tipis

70

`pragmatikuli~ cnebebiT (alternatiuli terminebi: `topiki/komentari~ an

`fokusi/foni~).

es midgomebi ar ewinaaRmdegeba erTmaneTs, radganac principSi enis bazisuri

sityvaTa rigi SeiZleba iyos cvalebadic pragmatikuli faqtorebis

zegavleniT. yvelaze xSirad, marTlac, ase xdeba, vinaidan enebi, sadac

bazisuri sityvaTa rigi saerTod arasodes ar irRveva, Zalian iSviaTia

(inglisuri enaa yvelaze axlos am tipTan).

`grinbergiseuli~ tradicia bolo dros gamdidrda m. draeris SromaTa seriiT,

j. nikolsis gamokvlevebiT. gansakuTrebiT aRsaniSnavia meTiu draeris statia

`sityvaTa rigis grinbergiseuli korelaciebi~ (Dryer 1992). avtoris mizania Seamowmos mtkiceba imis Sesaxeb, rom elementTa rigebi sxvadasxva

Semadgenlobis CarCoebSi korelaciaSia erTmaneTTan _ magaliTad, enebs

bazisuri OV mimdevrobiT Cveulebriv, aseve, aqvT `zedsarTavi-arsebiTi~

mimdevroba da Tandebulebi _ windebulebis sanacvlod.

sityvaTa rigis Sesaxeb arsebuli literaturidan draeris Sromas gamoarCevs

is, rom aq gaTvaliswinebulia ara marto ZiriTadi elementebis rigi

klauzulaSi, aramed aseve rigebi sxva tipis SemadgenlebSic. draeris statias

aqvs sul mcire sami aspeqti, romelic aucilebelia ganvixiloT:

pirveli, draeri iyenebs enaTa amorCevis formalurad damuSavebul

proceduras, romelic mas saSualebas aZlevs gaaTavisuflos raodenobrivi

Sedegebi SemTxveviTobisa da damaxinjebisgan. msoflios yvela ena dayofilia

genetur jgufebad _ `genusebad~. draeris monacemTa baza Sedgeba 543 enisgan,

magram yovel konkretul daTvlaSi yoveli genusi (aseTi monacemTa bazaSi

225-ia) warmodgenilia erTi eniT. es saSualebas iZleva Tavidan iqnes acile-

buli daTvlis Sedegebis genetikuri damaxinjeba. garda amisa, enebi dayofilia

6 arealad (afrika, evrazia, samxreT-aRmosavleT azia da okeaneTi, avstralia

_ axali gvinea, Crdilo amerika da samxreT amerika). arealuri danawileba

saSualebas iZleva ganvsazRvroT, romeli movlenebia universaluri da rome-

lia damaxasiaTebeli gansazRvruli arealisTvis; amgvarad, Tavidan aris

acilebuli arealuri gadaxrebi. draeris SerCevis es procedura erT-erTi

yvelaze win waweuli meTodologiaa Tanamedrove lingvistikaSi (Sdr. Nichols 1992, Bybee, Perkins, Pagliuca 1994).

meore, draeri axerxebs daamtkicos erTi da ukuagdos sxva raodenobrivi

korelaciebi arsebul sityvaTa rigebs Soris. korelacia mocemul SemTxvevaSi

gagebulia mkacri statistikuri azriT. enis agebulebis maxasiaTeblad miRe-

bulia zmnisa da damatebis urTierTmimdevroba zmnur jgufSi, xolo yvela

sxva darCenili rigebis tipebi udardeba mas. amasTan, sityvaTaSeTanxmebis

nebismier tipSi erTi elementi warmoadgens zmnis analogs zmnur jgufSi,

meore ki damatebis analogs. zmnisa da damatebis rigTan korelaciaSia Semdegi

71

elementebis mimdevrobebi: zmnisa da qvemdebaris; windebulis/Tandebulisa da

saxelis; kavSirebis/damxmare zmnisa da ZiriTadi predikatis; zmnisa da senten-

ciuri aqtantis; kavSirisa da sentenciuri aqtantisa/sirkonstantis; kiTxviTi

elementisa da klauzulis, artiklisa da saxelis; saxelisa da genitiuri

gansazRvrebis; saxelisa da brunvis; zedsarTavisa da Sedarebis obieqtis;

zmnisa da windebuliani/Tandebuliani jgufebis; zmnisa da adgilis garemoebis.

zmnisa da damatebis mimdevrobasTan korelaciaSi ar arian Semdeg elementTa

rigebi: zedsarTavisa da saxelis; CvenebiTi nacvalsaxelisa da saxelis;

gaZlierebuli garemoebisa da zedsarTavis; uaryofiTi an drois nawilakebisa

da zmnis.

mesame, draeri gvTavazobs amgvari SemCneuli movlenebis axsnas. rigebis kore-

lacia xSirad ixsneba da ganzogaddeba avtorebis mier mwvervalisa (zmna zmnur

jgufSi da yvela misi analogi sxva sityvaTSeTanxmebaSi) da damokidebulis

(damateba zmnur jgufSi da misi analogebi) dapirispirebis safuZvelze.

draeri cdilobs gviCvenos, rom mwvervali/damokidebulis terminebSi axsnas ar

SeuZlia iwinaswarmetyvelos, winaswar gansazRvros yvela realuri korelacia

da, amasTan, uSvebs araswor korelacias (magaliTad, arsebiTisa da zedsar-

Tavis rigi). draeri gvTavazobs axsnis sxva princips, kerZod, SemadgenelTa

xeSi ganStoebis mimarTulebis (marjvniv/marcxniv) terminebSi. draeris

mixedviT, es principi didi sizustiT winaswarmetyvelebs dakvirvebis

elementTa korpuss, simravles. enaTa erT rigs aqvs tendencia marcxniv

ganStoebisa, sxvebs _ marjvniv da es zogadi tendenciebi ganapirobeben

konkretul sityvaTa rigebs sityvaTa SeTanxmebebis gansxvavebul tipebSi (Sdr. Hawkins 1990, 1994). draeris Sromebma ganapiroba `grinbergiseuli~ tradiciis

mniSvnelovani progresi, rac TanxmobaSia `pragmatikulad~ mocemul sityvaTa

rigis warmatebuli kvlevis SedegebTan (Payne 1992, Sdr. Kibrik-is recenzia 1995, Downing, Noonan 1995).

`praRelTa~ Tavdapirveli ideis Tanaxmad, ufro metad Tematuri elementebi

ganTavsdebian winadadebaSi adre, vidre rematulebi. am tendenciis savaraudo

universaloba eWvis qveS dadga rigi gamokvlevebis, gansakuTrebiT p. tomlinisa

da r. roudsis (Tomlin, Rhodes 1979/1992) Sromebis Semdeg, romlebmac ojiba

enaSi (algonkinuri ojaxi) sruliad sapirispiro tendencia aRmoaCines:

Tematuri informacia ganTavsdeba ufro gvian, vidre araTematuri. t. givonis

SromebSi msgavsi mtkiceba araerTxel formulirdeboda lamis universalurad

(mag., Givon 1984 : 206 ff). amJamad dagrovda sabuTebis didi raodenoba imis

Sesaxeb, rom principi `rematuli informacia TavSia~ (variaciebiT: axali

dasawyisSia, ganusazRvreli TavSia, mniSvnelovani TavSia, saswrafo TavSia)

msoflios enebSi Zalian gavrcelebulia, ix. Mithun 1987/1992, Payne 1992, Downing, Noonan 1995 : 13-17). p. dauningi cdilobs mogvces kognitiuri axsna

am rigis orive principisa, TumcaRa gaugebari rCeba, Tu ratom aris erTgan

upiratesi erTi, sxva SemTxvevaSi _ meore.

72

m. miTuni (Mithun 1995) aRniSnavs, rom principi `rema dasawyisSi~ TanxmobaSia prosodiul faqtorebTan, ramdenadac remac da dasawyisic ixreba prosodiuli

gamokveTisken. manve gamoTqva, aseve, varaudi (Mithun 1987/1992), rom bazisuri sityvaTa rigi _ ar aris universalurad misaRebi parametri, anu zogierTi

enisTvis am cnebas ubralod azri ara aqvs. miTunma gaaanaliza amerikisa da

avstraliis sami polisinTezuri enis teqstebi da mivida daskvnamde, rom am

enebSi gardamaval klauzulebSi orive aqtanti saxeluri jgufebiT imdenad

iSviaTad gamoixateba (1-3% SemTxvevaSi), rom mocemuli konstruqcia saerTod

periferiulia da arnairi samwevra mimdevroba ar SeiZleba CaiTvalos

bazisurad.

marTlac, grinbergidan moyolebuli ama Tu im rigis bazisurobis erT-erT

ZiriTad kriteriumad, Cveulebriv, iTvleba misi sixSiruli upiratesoba.

magaliTad, umetesoba slavur enebSi dasturdeba gansxvavebuli rigebi, magram

bazisuri SVO rigi gvxvdeba yvelaze xSirad. m. draeri (Dryer 1995) amtkicebs, rom ama Tu im rigis bazisurobis mTavar kriteriumad unda iTvlebodes misi

pragmatikuli aramarkirebuloba, romelic Teoriulad SeiZleba arc iyos

TanxmobaSi maqsimalur sixSiresTan diskursSi.

bevr SromaSi aRniSnulia, rom aragramatikul sityvaTa rigi iwodebodes

`Tavisufal~ rigad mTlad koreqtuli ar aris, radganac is sruliadac ar

aris nebismieri. ase, magaliTad, o’odxam enis (=papago; iuta – actekuri

ojaxi) ganxilvisas, sadac sityvaTa rigi misdevs models `ganusazRvreli _

zmna _ gansazRvruli~, doris peini gonebamaxvilurad SeniSnavs: `pirveli

lingvistebi rom o’odxam enis matareblebi yofiliyvnen da maT rom mieCniaT,

rom yvela SesaZlo ena moqmedebs funqciebsa da struqturebs Soris

Sesatyvisobis imave safuZvelze, razec maTi mSobliuri ena, maSin inglisuri

SeiZleba gangvexila rogorc ena `Tavisufali~ sityvaTa rigiT~ (Payne 1992b:162), ramdenadac inglisurad gansazRvruli da ganusazRvreli saxeluri

jgufi SeiZleba ganTavsdes klauzulis sxvadasxva nawilSi. aqedan gamomdinare,

gadaWarbebuli aqcenti mxolod sityvaTa rigis tipologiaze _ dafuZnebuli

qvemdebarisa da damatebis cnebebze _ mTlad gamarTlebuli ar aris. maT rig

naSromSi ama Tu im sityvaTa rigis upiratesoba dakavSirebulia diskursis

sxva maxasiaTeblebTan, kerZod, diskursis struqturasa da mis epizodebad

danawevrebasTan (Meyer 1992, Cooreman 1992), klauzulebis diskursul rolTan

(Myhill 1992, Luraghi 1995). sxva naSromebSi aRiniSneba diskursis tipis an

Janris roli am procesSi: erTsa da imave enaSi sityvaTa rigis wesebi

SeiZleba gansxvavdebodes sxvadasxva diskursSi (Quakenbush 1992, Dryer 1995, Kim 1995, Longacre 1995).

vfiqrobT, Semdgomi progresi sityvaTa rigis kvlevis sferoSi SesaZlebelia

mxolod maSin, Tu sxvanairad SevxedavT problemas, SevcvliT Tvalsazriss:

sityvaTa rigi unda Seiswavlebodes ara rogorc ganxilvis amosavali

wertili, aramed rogorc formalur xerxTagan erT-erTi, romelic warmo-

73

adgens gansazRvruli kognitiuri funqciebis realizacias sxva formalur

saSualebebTan erTad.

3. sityvaTa rigi: uaxlesi midgoma

oqsfordSi 2007 wels robert d. van valinis redaqtorobiT gamoica nomi

erteSek-Siris monografia `sainformacio struqtura~ (Erteschik-Shir, 2007). es Sroma seriidan Oxford Surveys in Syntax and Morphology warmoadgens

sainformacio struqturis inventarisa da arqiteqturis mowesrigebis cdas.

enaSi gansxvavebuli sityvaTa rigis SeswavlisTvis avtori ganixilavs beilinis

mier SemoTavazebul (Bailyn 2002a:82), enobriv sistemebSi gavrcelebul e.w.

`cocva~ (Scrambling) movlenas: `cocva~ aris zogadi, momcveli termini

wesebisTvis, romlebic warmoqmnian sityvaTa arakanonikuri rigis modelebs e.w.

`Tavisufali sityvaTa rigis enebSi~, iseTebSi rogoricaa, magaliTad, iaponuri,

rusuli, germanuli, hindi, qarTuli da bevri sxva. aseT enebSi Semadgenlebi

SeiZleba gamoCndnen sxvadasxvagvari zedapiruli rigebiT, winadadebis

birTvuli mniSvnelobis cvlilebis gareSe. termini modis rosisgan (Ross 1967), romelic mas iyenebda, rogorc stilistikur wess. mravali

gamokvlevisTvis is rCeba eqstra-gramatikulad misi zogadobis gamo da imis

gamo, rom ar aqvs aranairi semantikuri mniSvneloba. cocva gansxvavebulia

topikalizaciisgan imiT, rom is ar aris aucileblobiT SezRuduli bazisuri

winadadebiT, ase rom, yovelTvis advili ar aris gamoacalkevo gansakuTrebuli

SemTxvevebi, romelSic elementi `gacocebulia~ marcxena periferiisken. rosis

Tavdapirveli dakvirvebidan moyolebuli SemadgenelTa gadaadgileba

gansxvavdeboda rigi parametrebis mixedviT, maT Soris erT-erTia gansxvaveba A-bar da gadaadgilebas Soris. topikalizacia ganixileba an A-bar gadaadgilebis

terminebSi an rogorc damateba marcxena periferiaze. cocvam miiRo Sereuli

gageba SromaTa did nawilSi, romlebic eZRvneboda im sakiTxis gadawyvetas, Tu

rogor unda gaanalizebuliyo procesi _ rogorc A-bar Tu rogorc

gadaadgileba. kan geldereni (Gelderen 2003) cocvas gansazRvravs rogorc

sainformacio struqturiT warmoqmnil (sintaqsur) gadaadgilebas, xolo

topikalizacias _ rogorc cocvis tips. gansxvavebulia cocvis sul cota

sami tipi: 1) mokle cocva (VP-s farglebSi), 2) saSualo cocva (VP-s gare pozicia) da 3) daSorebuli distanciis cocva (CP sazRvris gadalaxviT)

(Takano 1998, Bailyn 2002b). miuxedavad im interesisa, romelsac es sakiTxebi

warmoSobs, Teoria, romelic lingvisturad gansazRvravda amaTgan erTis an

meoris gamoCenas, jer-jerobiT miuRwevelia. sintaqsur gamokvlevaTa umravle-

sobisTvis fakultatorobis sakiTxi aris yvelaze Znelad gadasaWreli, ramde-

nadac fakultaturi gadaadgileba sinamdvileSi iSviaTad aris mkacrad

fakultaturuli da xSirad motivirebulia sainformacio struqturiT (Bailyn 2000b).

zemoT warmodgenili msjeloba saSualebas gvaZlevs davaskvnaT, rom sainfor-

macio struqturisa da sityvaTa rigis kvleva mxolod sawyis safexurzea. am

mimarTulebiT muSaoba ara marto win waswevs sityvaTa rigze, rogorc aseTze,

74

Cvens warmodgenebs, aramed kidev ufro daazustebs sintaqsis rols

gramatikaSi.

4. sityvaTa rigi qarTvelur enebSi: gramatikuli tradicia

4.1. Tu ar CavTvliT ramdenime gamonakliss, qarTuli aris erTaderTi

qarTveluri enebidan, romlis sityvaTa rigi warmoadgens sistemuri Seswavlis

sagans. qarTvelur enebs aqvT ganviTarebuli morfologiuri struqtura da

mdidari afiqsaluri sistema, amitom ar aris aucilebloba sityvaTa rigis

mkacrad gansazRvrisa, rac Cveulebrivia uafiqso enebisaTvis. es aris mTavari

mizezi qarTvelur enebSi sityvaTa rigis Tavisuflebisa. amis Sesaxeb

aRniSnulia rogorc qarTul gramatikul tradiciaSi, aseve ucxoel

mkvlevarTa SromebSi (javaxiSvili 1929, SaniZe 1948; kiziria 1950; kaxaZe

1953; klimovi 1961; foCxua 1962; alxaziSvili 1966; kvaWaZe 1966; Ciqobava

1968; fogti 1971, 1988; sergia 1977; daviTiani 1973; komsi 1978; zardiaSvili

1978; boederi 1979; harisi 1981, 2000; enuqiZe 1981; asaTiani 1982;

sarjvelaZe 1984; afridoniZe 1986; kiziria 1987; hiuiti 1995; testeleci

1998).

literaturaSi aRniSnulia isic, rom, miuxedavad rigis Tavisuflebisa,

SeiniSneba garkveuli SezRudvebi da tendenciebi, kerZod: zedsarTavi,

Cveulebriv, saxelis win gvxvdeba; zmnizedebi da damatebebi, aseve, umetesad

warmodgenilia zmnis wina poziciaSi; kiTxviT winadadebebSi zmniswina pozicia

aseve kanonikuria kiTxviTi sityvisTvis sxvadasxva funqciiT da a.S. sintaqsuri

funqciebis statistikuri analizi gviCvenebs, rom kanonikuri sityvaTa rigi

aris SVO an SOV, anu statistikurad yvelaze xSiri swored es

mimdevrobebia, romlebic miiCneva amosaval, aramarkirebul sityvaTa rigad.

yoveli gadaxra am rigidan avtorTa umravlesobis mier Sefasebulia rogorc

`aqtualizacia~ maTesiusis Teoriis farglebSi. isini Semoifarglebian mxolod

zogadi cnobebiT sityvaTa rigisa da sainformacio struqturis Sesaxeb. bolo

dros, r. asaTianis informaciis struqturirebis modelebis Tanamedrove

meTodebiT kvlevam qarTul saenaTmecniero sivrceSi Seavso am kuTxiT

arsebuli xarvezi (asaTiani 2000, 2002, 2006, 2007), rac Tavis mxriv, gaxda

safuZveli saqarTvelos ganaTlebis saministros qarTvelologiis,

humanitaruli da socialuri mecnierebebis `rusTavelis fondis~ mier

dafinansebuli Tsu proeqtisa: `informaciis struqturirebis ZiriTadi

modelebi qarTvelur enebSi~ (2006-2009 ww.), daiwera samagistro Sromebic

(wereTeli 2006; grZeliZe 2007; janeliZe 2008).

4.2. qarTuli, megruli da svanuri TxrobiTi, martivi winadadebebi ZiriTad

SemadgenelTa rigis fleqsiurobis mixedviT erTmaneTisagan mniSvnelovnad ar

gansxvavdebian. qarvelur enebSi martivi sintaqsuri konstruqciebis struq-

tura efuZneba zmnis valentobasa da saxelis gansxvavebul brunvaTa formebs.

winadadebaSi ZiriTad SemadgenelTa _ S, V, O, Teoriulad SesaZlo eqvsive

75

mimdevroba dasaSvebia, rac vlindeba rogorc werilobiT teqstebSi, aseve

informantTaA zepirmetyvelebasa da dialogebSi.

(1) qarTuli

a. kac-i a – g – eb – s saxl – s man – NOM NV – build – THM – PRES.S3.SG house – DAT b. kaci saxls agebs

g. saxls agebs kaci

d. saxls kaci agebs

e. agebs kaci saxls

v. agebs saxls kaci

(2) megruli

a. koC – i o – g – an – s Fude – s man – NOM NV – build – THM – PRES.S3.SG house – DAT b. koCi Fudes ogans g. Fudes ogans koCi d. Fudes koCi ogans e. ogans koCi Fudes v. ogans Fudes koCi

(3) svanuri

a. mÀre- Ø a – g – em qor – s

man – NOM NV– build – THM.S3.SG house – DAT b. mÀre qors agem

g. qors agem mÀre

d. qors mÀre agem

e. agem mÀre qors

v. agem qors mÀre

winadadebis aqtualizebuli wevrebis (fokusis, topikis) nairsaxeobebis

sakvlevad svamen kiTxvebs, romlebzec pasuxi aCens markirebis gansxvavebul

modelebs: fonetikur-fonologiurs, morfosintaqsursa da sxv. magaliTad,

kiTxvebze:

(4) vin agebs saxls?

(5) ras agebs kaci? da a.S.

(4) kiTxvis pirdapir pasuxad mocemuli eqvsi variantidan yvelaze

xSiria a da b, xolo (5)-is pasuxad _ (g) da (d). statistikurad ufro

xSiria TFV da FVT.

76

5. sityvaTa rigi da Zveli/axali informacia

Tavisufali rigis enebSi Zveli/axali informaciis aqtualizaciisTvis sityvaTa

ganlageba asrulebs mniSvnelovan rols (Skopeteas and Fanselow 2007). am

TvalsazrisiT, winadadebis dasawyisi aris yvelaze Rirebuli pozicia

sametyvelo nakadis informaciulad winwamoweuli nawilis gamokveTisTvis.

statistika gviCvenebs, rom qarTvelur enebSi S-sTvis yvelaze xSiria

winadadebis Tavkiduri pozicia, SemadgenelTa momatebasTan erTad xdeba S-s gadaweva (resp. `gacoceba~) marcxniv. amasTan, rodesac zmna gardamavalia, gvaqvs SVO mimdevroba, gardauval zmnasTan e.w. `introduqciul~ winadadebebSi _ VS mimdevroba, magaliTad:

(6) qarT. iyo erTi xelmwife

megr. Fofe arTi xelmwife

svan. arda eSxu xelwif

winadadebaSi V-s adgili mkacrad gansazRvruli ar aris, Tumca SemadgenelTa

momatebasTan erTad aqac xdeba V-s gadmoweva marcxniv. pirdapiri obieqti zmnasTan mWidrod dakavSirebuli wevria, is umetesad

bolokidur an boloswina poziciaSi gvxvdeba. iribi obieqti meti

TavisuflebiT xasiaTdeba, SeiZleba gadaiwios marcxnivac da marjvnivac.

rodesac winadadebaSi orive obieqti saxezea, maSin rigi aseTia: Oir.VOpir.

sametyvelo nakadis (resp. informaciis) struqturirebisas informaciulad

gamokveTili wevrebi ikaveben winadadebis Tavkidur pozicias, amasTan, sawyisi

poziciis dakavebis mixedviT S>0, xolo Zveli informacia > axals. rodesac es

ori ierarqia S>0 da Zveli > axali ejaxeba erTmaneTs `gamarjvebulia~ Zveli

> axali ierarqia.

6. topikalizaciis sintaqsurad Warbi markerebi

topikalizaciis erT-erT saSualebad ganixileba sintaqsurad Warbi

erTeulebis xmareba, magaliTad: piris nacvalsaxelebis ([me]T wavedi), refleqsuri nacvalsaxelis (ninom [TavisTvis]T CainiSna) (asaTiani 2006).

topokalizaciis msgavs, sintaqsurad Warb, markerad migvaCnia winadadebis

zogierT wevrTa gameoreba. Cveulebriv meordeba winadadebis topikalizebuli

wevri. SeiZleba gamoiyos ori SemTxveva:

a. zmnis wina poziciaSi dadasturebuli sityvisa (saxeli, zmnizeda, nawilaki

da sxv.) da winadadebis Tavkidur poziciaSi dadasturebuli zmnis

topikalizacia SesaZlebelia gaZlierdes rekursiiT (magaliTebi qarTuli

literaturidan motanilia: afridoniZe 1986, mixedviT):

(7) qarT. wigni getyvis, [wigni]T megr. wingi giwiins, [wingi]T svan. lAYr jªqUnine [lAYr]T (8) qarT. marjvniv midis Cveni gza, [marjvniv]T megr. marjgvaniSe meurG˜ CqGni Sara, [marjgvaniSe]T

77

svan. mursgUenTe esRri gUSgUei SukU, [mursgUenTe]T (9) qarT. nu grcxvenia, [nu!]T megr. nu(m) giCqu onjRore, [nu(m)]T svan. nom(a) jiSgur, [nom(a)!] T (10) qarT. daviwviT, batono, [daviwviT!] T megr. dibWviT, patGn(i), [dibWviT!]T svan. CoTSixªnd, fusd, [CoTSixªnd!]T

zemoT moyvanil magaliTebSi rekursiis gziT aqtualizebuli winadadebis

wevrebi gacalkevebulia erTmaneTisgan sul mcire erTi sityviT.

b. winadadebis gameorebuli wevrebi (resp. topikebi)SeiZleba mosdevdnen

erTmaneTs:

(11) qarT. xelmwife, [xelmwife]T aqe! megr. xemwife [xemwife]T ocqve! svan. xelwif, [xelwif]T afaSdU…! (12) qarT. movSordeT, [movSordeT]T aqaurobas! megr. deveTxuaT, [deveTxuaT]T Taqianobas! svan. qªlRerd, [qªlRerd]T amxAnqa!

faqtiurad, ar gvxvdeba SemTxveva, rodesac winadadebaSi zmnis momdevno

Semadgeneli meordeba.

7. daskvna

• bunebrivi enis universaluri maxasiaTebeli aris enobrivi sistemis bazisuri,

amosavali viTarebidan fakultaturuli gadaxris SesaZlebloba, rac gansa-

kuTrebiT TvalsaCinoa informaciis sasurveli saxiT struqturirebis dros.

SemadgenelTa wrfivi mimdevroba ganisazRvreba imis mixedviT, Tu konteqs-

turad ra aris cnobili da ra _ ara. imis axsnas, Tu ra meqanizmi udevs

safuZvlad informaciis struqturirebas, wlebis manZilze cdilobdnen

fonetikur-fonologiuri, morfo-sintaqsuri Tu leqsikuri terminebisa da

enobrivi sistemis semantikuri, pragmatuli donis erTeulebis safuZvelze,

mimarTavdnen Sereuli midgomis xerxebsac. droTa viTarebaSi ixveweboda

terminologia, zustdeboda meTodologia. dRes am mimarTulebiT gamoqvey-

nebulia mravali Sroma, miRebuli Sedegebi metad mniSvnelovania enaTmecni-

erebis Tanamedrove dargebisTvis ganviTarebisTvis.

• informaciis aqtualizebisaTvis sxvadasxva enaSi sxvadasxvagvari saSualebebi

arsebobs. yoveli ena irCevs misTvis misaReb strategias da informaciis

garkveuli nawilis gamosakveTad swored Tavisi struqturis Sesabamis saSu-

alebebs iyenebs.

• fokusi da topiki winadadebis sainformacio struqturis ganmsazRvreli

ZiriTadi cnebebia. orive aRniSnavs winadadebis aqtualizebul wevrs: fokusi _

axali informaciis Semcvel aqtualizebul wevrs, topiki _ ukve cnobilis.

78

fokusisa da topikis markirebis erT-erT models morfologia-sintaqsis

doneze warmoadgens sityvaTa rigis cvla.

• qarTvelur enebSi sityvaTa rigi sazogadod Tavisufalia. enobriv masalaSi

SeiZleba dadasturebul iqnas nebismieri wyoba, SeiniSneba mxolod mcireodeni

SezRudvebi da upiratesobebi. sityvaTa gadalagebis logikurad SesaZlo yvela

kombinacia SeZleba ganvixiloT, rogorc realizcia ufro metad gansxvavebuli

informaciuli datvirTulobis maCvenebeli tendenciebisa, vidre wesebisa,

romlebic marTaven informaciis SefuTvis process.

• aramarkirebuli sityvaTa rigis mxolod gadalagebiT (intonaciuri

konturis cvlilebis gareSe) formdeba gansxvavebuli sainformacio struq-

turebi, Tumca sityvaTa yoveli gadalagebisTvis damaxasiaTebeli specifikuri

pragmatikuli Sinaarsis niuansebSi garkveva martivdeba, Tu gadalagebas Tan

axlavs intonaciis markirebac.

• aqtualizebuli informaciis (resp. winadadebis aqtualizebuli wevris)

winadadebis dasawyisSi moTavseba, win gadasma informaciis markirebis

tipologiurad gavrcelebuli strategiaa, romelsac qarTvelur enebSic

vxvdebiT.

• fokusirebuli wevri moTavsebulia winadadebis TavSi zmnis wina poziciaSi

(saTanado intonaciuri markirebiT); amdenad, saukeTeso mimdevrobaa [fokusi,

zmna, sxva wevrebi], Tumca iSviaTad fokusirebuli wevri SesaZlebelia

winadadebis bolokidur poziciaSic iyos warmodgenili.

• topikic aseve Tavsdeba winadadebis TavSi da/an zmnis wina poziciaSi

(saTanado intonaciuri markirebiT); Tumca iSviaTad aseve SesaZlebelia

topikis bolokidur poziciaSic ganTavseba.

• fokus-topikis winadadebebSi iqmneba erTgvari Sejibris situacia wina-

dadebis dasawyisSi ganTavebis TvalsazrisiT da orive fokusi an topiki

SeiZleba iyos `gamarjvebuli~: orive alternatiuli mimdevroba [T, F, V] an [F, V, T] SesaZlebelia.

• topikalizaciis markerad SeiZleba davasaxeloT winadadebis zogierTi

wevris gameoreba. Cveulebriv, meordeba winadadebis aqtualizebuli wevri,

gansakuTrebiT, naratiul teqstebsa da zepirmetyvelebaSi.

literatura asaTiani, r., 1982, martivi winadadebis tipologiuri analizi, `mecniereba~,

Tbilisi.

asaTiani, r., wereTeli n., 2009 winadadebis sainformacio struqtura:

implikaciuri topiki qarTulSi, enaTmecnierebis sakiTxebi, I, 2008, Tbilisi.

asaTiani, r., grZeliZe q., 2008, winadadebis sainformacio struqtura:

lokativebi qarTulSi (ibeWdeba).

asaTiani, r., janeliZe T., 2009, winadadebis sainformacio struqtura:

kontrastuli topiki da fokusi qarTulSi (ibeWdeba).

afridoniZe, S., 1986, sityvaTganlageba axal qarTulSi, `mecniereba~, Tbilisi.

79

daviTiani, ak., 1973, qarTuli enis sintaqsi, I, martivi winadadeba, `ganaTleba~,

Tbilisi.

enuqiZe, l., 1981, winadadebis aqtualuri danawevreba da misi mimarTeba

sintaqsuri da semantikuri analizis Tanamedrove meTodebTan, Tanamedrove

zogadi enaTmecnierebis sakiTxebi, t. VI, 96-110, Tbilisi.

kvaWaZe, l., 1966, Tanamedrove qarTuli enis sintaqsi, `rubikoni~, Tbilisi.

kiziria, a., 1950, Semasmenlis adgili winadadebaSi, qarTuli ena da

literatura skolaSi, 3, Tbilisi.

b 1987, saliteraturo qarTuli enis intonaciis sakiTxebi, `mecniereba~,

Tbilisi.

sarjvelaZe, z., 1984, qarTuli literaturuli enis istoriis Sesavali,

`ganaTleba~, Tbilisi.

sergia, v., 1977, sityvaTa rigi da winadadebis aqtualuri danawevreba qarTul

saliteraturo enaSi, Tsu Sromebi, enaTmecniereba, t. 187, 96-108, Tsu,

Tbilisi.

foCxua, b., 1962, sityvaTganlagebisTvis qarTulSi, ike, XIII, Tbilisi.

SaniZe, ak., 1948, qarTuli enis gramatika, II, sintaqsi, Tsu, Tbilisi.

Ciqobava, arn., 1968, martivi winadadebis problema qarTulSi, I `mecniereba~,

Tbilisi.

javaxiSvili, iv., 1929, Tanamedrove qarTuli saliteraturo enisaTvis,

mnaTobi, 1, 144-148, Tbilisi. Алхазишвили, А. А, 1959, Порядок слов и интонация в простом повествовательном предложении грузинского языка, Фонетический сборник ТГУ, посвященный акад. Г. С. Ахвледиани, Тбилиси. Кибрик, А. А., Плунгян, В. А., 1997, Функционализм, Фундаментальные Направления Современной Американской Лингвистики, Сборник Обзоров, изд. МГУ, Москва. Климов, Г. А., 1961, К вопросу о порядке членов атрибутивного комплекса в картвельских языках; Вокарев, А. А. (ed.). Вопросы Изучения Иберийско-Кавказских языков, М., изд. АН СССР, стр. 257-270. Bailyn, I. F., 2002a, Sirambling to Reduce Serambling – Part 1. Glot International 6(4): 83-90. 2002b. Scrambling to Reduce Scrambling – Part 2. Glot International 6(5):109-24. Boeder, W., 2004, The South Caucasian Language, Lingua 115 (2005) 5-89. Bybee, I. L., Perkins, R.D., Payliuca, W., 1994, The Evolution of Grammar: Tense, Aspect and Modality in the Languages of the World. Chicago: University of Chicago Press. Comrie, B., 1978, Ergativity, in: Lelmann, Winfried P. (ed.) Syntactic typology. Studies in the phenomenology of language. Austin & London: The University of Texas Press, pp. 329-394. Cooreman, A., 1992 – The Pragmatics of Word Order Variation in Chamorro Narrative. Text – Payne 1992c : 243-263. Downing, P., Noonan, M., (eds) 1995, Word Order in Discourse (Typological Studies in Language, vol. 30) Amsterdam : Benjamins. Dryer, M. S., 1922, 1995, 1996, Greenbergian Word-Order Correlations. Language 68 :81-138. Frequency and Pragmatically Unmarked Word Order. – Downing, Noonan, 1995 : 103-136, Focus, Pragmatic Prescepposition, and activated Propositions. Journal of Pragmatics 26(4): 475-523. Erterschik – Shir, N., 2007, Information Structure, The Syntax-Discourse Interface, Oxford University Press. Fery C., Fanselow, G., Krizka, M., 208, The Notions of Information Structure, Potstam – Berlin.

80

Firbas, J., 1964, On defining the Theme in Functional Sentence Analysis. Travaux linguistiques de Prague 1: 267-280. Gelderen, V. van, 2003, Scrambling Unscrambled. Ph. D. Dissertation, University of Leiden. Givón T., 1984-1990, Syntax: A Functional – Typological Introduction. vol. 1, 1984, vol. 2, Amsterdam : Benjamins. Greenberg, J. H., 1963 (1966), Some Universals of Grammar with Particular Reference to the Order of Meaningful Elements. Universals of Language ed. By J. H. Greenberg, 73-113. Cambridge, Mass: MIT Press. Harris, Al. C., 1981, 2000, Georgian Syntax. A Study in Relational Grammar. Cambridge: Cambridge University Press. Word Order Harmonies and Word Order Change in Georgian. In Sornicola, R., Poppe, E., Haley A. (eds), Stability Variation and Change of Word-Order Patterns over time. 133-163. Amsterdam/Philadelphia : Benjamins. Hawkins, J., 1983, 1990, 1994. Word Order Universals. N. Y.: Academic Press. Parsing Theory of Word Order Universals. Linguistic Inquiry 21:223-261; A Performance Theory of Word Order and Constituency. Cambridge: Cambridge University Press. Hewitt B. G., 1995, Georgian: A Structural Reference Grammar, Amsterdam, Philadelphia : Benjamins. Kibrik, A. A., 1995, Review of Payne (1992), Studies in Language 19:223-237. Kim, Kyu-hyun, 1995, WH-Clefts and left-Dislocation in English Conversation. – Downing and Nooan 1995 : 247-298. Longacre, R. E., 1995, Left Shifts in Strongly VSO Languages. Downing, Nooan 1995 : 331-354. Luraghi, S., 1995, The Pragmatics of Verbal Initial Sentences in Some Ancient Indo-European Languages. – Downing, Nooan, 1995 : 335-386. Mathesius, W., 1929/1983, Functional Linguistics. Vachek K. (ed.). Praguina. Amsterdam : Benjamins. Meyer, K.S., 1992. Word Order in Klamath. Payne 1992: 167-192. Mithun M., 1987/1992, Is Basic Word Order Universal? – Payne 1992 : 15-62. Myhill, J., 1992, Word Order and Temporal Sequencing. Payne 1992 : 265-278. Nichols, J., 1992, Linguistic Diversity in Space and Time. Chicago : University of Chicago Press. Payne, D., (ed.), 1992, Pragmatics of Word Order Flexibility. (Typological Studies in Language, vol. 22) Amsterdam : Benjamins. Nonidentifiable Information and Pragmatic Order Rules in ‘O’odham – Pane 1992: 137-166. Quakenbush, S., J., 1992, Word Order and Discourse Type: An Austronesian Example – Payne 1992 : 279-304. Ross, J. R., 1967, Constraints on Variables in Syntax. Ph. D. Dissertation, MIT, Cambridze, Mass. Skopeteas, S., Fanselow, G., 2007, Effects of givenness and constraints on free word order. In Information Structure from different perspectives. Zimmermann, Malte and Caroline (eds.), Oxford: Oxford Un. Press. Takano, Y., 1998, Object Shift and Scrambling. Natural Language and Linguistic Theory 16:817-89. Testelets, Y, G., Word Order in Kartvelian Languages, in: Siewierska Anna (ed.), Constituent order in the languages of Europe. Mouton de Gruyter: Berlin – New York. 1998, pp. 235-256. Tomlin, R. S., 1986, Basic Word Order : Functional Principies. London: Groom Helm. Tomlin, R., Rhodes, R., 1979/1992, Information Distribution in Ojibwa. – Payne 1995 : 117-136. Voght, H., 1971, Grammaire de la langue georgienne, Oslo. 1988, L’ordre des mots en georgien moderne. – In: Voght, Hans. Linguistique caucasienne et armenienne, Oslo : Norvegian University Press.

81

Marine Ivanishvili

Information Structure of a Sentence: Word Order in the Kartvelian Languages

Abstract

Kartvelian languages have a very developed morphology and they are rich in case affixes. Therefore, in Kartvelian and, generally speaking, in any languages with developed case patterns, there is no necessity to have a strict word order as it usually takes place in languages with poor morphology. Existence of various case patterns can be the main reason for word order flexibility: all kinds of reordering are permitted. Notwithstanding with a fact that the word order in Kartvelian languages in mostly free, yet, it is possible to distinguish statistically unmarked word order: SOV/SVO. Any deviation from this order is caused by the changes in information structure of a sentence: various focus and topic combinations can be expressed by the reordering. From the view of functional approach in particular of information packaging we got such results: • Kartvelian languages are the languages with a free word order with slight restrictions and preferences. All logically possible combinations of reordering can be realized as structures with various informational loadings, and tendencies rather than rules govern the processes of information packaging; • Only reordering of unmarked word order (without changing of pitch tracks) can form different information structures, yet the reordering along with the specific intonation seems to be the best strategy for various forms of foregrounding; • Fronting (putting the foregrounding part of information in the sentence-initial position) is the best strategy for marked information; • Focus tries to take preverbal position and usually occupies the sentence-initial position; thus, the best order is [Focus, Verb, Other Participants]; yet, in few cases, it can be also sentence-final; • Topic also tries to take the sentence-initial position; yet, sentence-final position is possible for it as well; • Focus-topic sentences make situations competitive due to the tendency of taking the sentence-initial position and: both alternates of order: [T, F, V] or [F, V, T] are possible; • Syntactically redundant entities as markers of topicalization can be found in the sentences where the recurrence of some members takes place; this device of topicalization is especially characteristic for narrative texts and colloquial language (speech).

82

ivane leJava

glotalizaciis xarisxi winadadebis fokusirebul da

topikalizebul wevrSi

bgeris glotalizaciis xarisxs gansazRvravs ramdenime statikuri da

dinamikuri parametri: grZlivoba Cqamis dasawyissa da momdevno bgeras

Soris, Cqamis grZlivoba da intensivoba, intensivobis mateba-klebis dina-

mika, ubgero intervalis arseboba da misi grZlivoba.

imis dasadgenad, Tu ra gavlenas axdens glotalizaciis xarisxze wina-

dadebis wevrTa topikalizacia da fokusireba qarTul metyvelebaSi,

Sedgenil iqna Sesabamisi kiTxvari, sadac Sedarebis TvalsaCinoebis mizniT

yuradReba ZiriTadad gamaxvilebulia erTsa da imave bgeraze (magaliTad:

biWi Wams Wads). es fraza neitraluria informaciis romelime Semadgeneli

elementis aqtualizebis TvalsazrisiT, rodesac pasuxia kiTxvaze _ ra xdeba?; martivi viwro fokusi gvaqvs kiTxvaze _ vin Wams Wads?, xolo

martivi topikia kiTxvaze _ ras ityviT biWis Sesaxeb?

ganxilulia Semdegi SemTxvevebi: martivi viwro fokusi, WeSmaritebis

fokusi, kompleqsuri fokusi, mravalricxovani fokusi, martivi topiki,

kontrastuli topiki da implikaciuri topiki. damatebiT, ufro detaluri

SedarebisaTvis gaanalizebulia neitraluri(N), martivi viwro fokusisa(F) da martivi topikis(T) magaliTebi: WaWa duRs, kata knavis. kiTxva-pasuxis audioCanawerebi gaanalizebulia metyvelebis analizis

cifruli programebis meSveobiT (WaveSurfer-1.8.5 da PPraat-4.6.36). warmoT-qmuli winadadebebis vizualizaciis Semdeg xdeboda sakvlevi bgerebis

segmentacia da bgeraTa fazebis droiTi da amplituduri maxasiaTeblebis

fiqsireba. kerZod, Cqamis grZlivoba, ubgero intervalis grZlivoba, Cqamis

intensivoba da misi mateba da kleba decibelebSi.

glotalizebuli bgerebis fizikuri parametrebi neitraluris(N), martivi

viwro fokusisa(F) da martivi topikis(T) mixedviT ase ganawildeba:

1. frazis Tavkiduri xSul-mskdomi k Tanxmovani Cqamis grZlivoba: F= T =N Cqamis intensivoba: F>T=N

Cqamis intensivobis mateba da kleba: F>T=N ubgero intervalis grZlivoba: F=T>N

83

2. frazis Tavkiduri afrikati W Tanxmovani Cqamis grZlivoba: F=T= N Cqamis intensivoba: F=T>N

Cqamis intensivobis mateba da kleba: F=T>N ubgero intervalis grZlivoba: F=T>N

3. frazis meore marcvlis xSul-mskdomi t Tanxmovani Cqamis grZlivoba: F=T=N Cqamis intensivoba: F=T=N Cqamis intensivobis mateba da kleba: F=T=N ubgero intervalis grZlivoba: F=T=N

4. frazis meore marcvlis afrikati W Tanxmovani

Cqamis grZlivoba: F≥T=N

Cqamis intensivoba: F=T=N

Cqamis intensivobis mateba da kleba: T>F=N ubgero intervalis grZlivoba: F=T>N

gaanalizebuli masalidan Cans, rom winadadebis wevrTa topikalizacia da

fokusireba iwvevs glotalizebuli bgeris fizikuri parametrebis garkveul

cvlilebebs.

glotalizebuli Tanxmovnebis im parametrTa cvlilebebis mixedviT, romlebic

gansazRvraven glotalizaciis xarisxs, irkveva, rom winadadebis fokusirebuli

da topikalizebuli wevris Tavkiduri Tanxmovnis glotalizaciis xarisxi _

rogorc mskdomis, ise afrikatisa _ metia, vidre neitralurisa.

odnav gansxvavebuli suraTia winadadebis meore marcvlis TanxmovnebSi imisda

mixedviT _ mskdomia Tu afrikati. mskdomi TanxmovnisaTvis yvela parametri

ucvlelia neitralurSi, topikalizebulsa da fokusirebulSi, xolo afri-

katisaTvis topikalizacia da fokusireba iwvevs glotalizaciis xarisxis

odnav gazrdas.

glotalizebuli Tanxmovnis Cqamis grZlivoba aris yvelaze stabiluri

parametri, romelic TiTqmis ar icvleba sityvis topikalizaciisa da

fokusirebisas.

84

[biWi Wams Wads] (N)

(pasuxi kiTxvaze: ra xdeba?)

[biWi] Wams Wads (F)

(pasuxi kiTxvaze: vin Wams Wads?)

85

[biWi], Wams Wads (T)

(pasuxi kiTxvaze: ras ityviT biWis Sesaxeb?)

[WaWa duRs] (N)


86

[WaWa] duRs (F) (pasuxi kiTxvaze: ra duRs?)

[WaWa] duRs (T)

(pasuxi kiTxvaze: ras ityviT WaWis Sesaxeb?)

87

[kata knavis] (N)


[kata] knavis (F) (pasuxi kiTxvaze: ra knavis?)

88

[kata] knavis (T) (pasuxi kiTxvaze: ras ityviT katis Sesaxeb?)

Ivane Lezhava

Glottalized Consonants of Focused and Topicalized Sentence Members in Georgian

Abstract

According to the changes of the parameters defining the degree of glottalization characteristic of glottalized consonants, it has been found out that the degree of glottalization of the initial consonant (being either plosive or affricate) of the focused and topicalized sentence members is higher than that of neutral ones. As for the consonants of the second syllable, the degree of glottalization characteristic of them depends on what the consonant is like: it is plosive or affricate. For the plosive consonant all parameters stay unchanged in neutral, topicalized and focused sentence members, though for affricate consonants topicalizing and focusing results in slight raising of the degree of glottalization. The duration of the noise of glottalized consonant is the most stable parameter, which hardly changes when the word is either topicalized or focused.

89

irine meliqiSvili

vnebiTi gvari da Tavisufal sityvaTa rigi

rogorc aqtantebis topikalizaciis alternativebi

winadadebis sainformacio struqturis cvlileba, anu winadadebis wevrTa

topikalizacia, sxvadasxva saSualebiT SeiZleba ganxorcieldes. am

saSualebaTagan intonaciasTan erTad ZiriTadia a)vnebiTi gvaris kons-

truqciaTa meSveobiT winadadebis wevrTa gadaweva subieqtis poziciaSi da

b)winadadebis wevrTa rigis SecvliT maTi dayeneba topikur poziciaSi.

qarTuli enisaTvis topikur pozicias warmoadgens winadadebis sawyisi,

ZiriTadad zmnis wina pozicia (asaTiani, wereTeli 2008). ena, misi struq-

turis Sesabamisad, SesaZlebelia upiratesobas aZlevdes erT-erTs am ori

saSualebidan, Tumca naklebi intensivobiT iyenebdes meoresac. Tu am Tval-

sazrisiT erTmaneTs SevudarebT, vTqvaT, inglisur da qarTul enebs, _

inglisuri aSkarad pasivizacias ufro intensiurad iyenebs topikalizaci-

isaTvis, qarTuli ki _ sityvaTa rigis cvlilebas. vfiqrobT, amis mizezi

SeiZleba davinaxoT imaSi, rom inglisuri fiqsirebuli sityvaTa rigis enaa,

xolo qarTulSi sityvaTa rigi advilad eqvemdebareba cvlilebas.

pasiuri transformaciis meSveobiT qvemdebaris poziciaSi sxvadasxva aqtan-

tebis gadawevis SesaZlebloba arsebobs. ramdenadac qvemdebare priorite-

tulad topiks warmoadgens, qvemdebaris poziciaSi pirdapiri, iribi,

ubralo damatebebis gadaweva winadadebis wevrTa topikalizaciis erT-erTi

saSualebaa. sainteresoa enaTa tipologia qvemdebaris poziciaSi aqtantebis

gadawevis TvalsazrisiT. am TvalsazrisiT perlmuterma da postalma Sem-

degi Semdegi universaluri ierarqia daadgines: S>DO>IO>OO (perlmuteri,

postali 1977), rac imas niSnavs, rom Tu enaSi xdeba ubralo damatebis

(OO – Obliqne Object) gadaweva qvemdebaris poziciaSi, maSin masSi gveqneba iribi (IO _ Indirect Object) da pirdapiri damatebebis (DO _ Direct Object) gadawevac da Tu enaSi xdeba iribi damatebis gadaweva, maSin am enaSi

gveqneba pirdapiri damatebis gadawevac, magram ara piriqiT.

Sesabamisad, qvemdebaris poziciaSi aqtantebis gadawevis TvalsazrisiT

gamoiyofa enaTa sami tipologiuri klasi: 1. enebi, romlebSic dasaSvebia

gadaweva DO > S (pirdapiri obieqti > subieqti), 2. enebi gadaweviT DO > S da IO > S (dasaSvebia pirdapiri da iribi obieqtis gadaweva subieqtis

poziciaSi) da 3. enebi gadaweviT DO > S, IO>S da OO > S (dasaSvebia rogorc pirdapiri da iribi, ise ubralo damatebis gadaweva subieqtis

poziciaSi. enaTa amgvari tipologia perlmuter-postalis ierarqiaze damya-

rebiT warmodgenili aqvs d. e. jonsons (jonsoni 1997; rusuli Targmani;

1982, 47-51).

90

enaTa pirveli tipis (DO>S) magaliTebad mas mohyavs franguli, germanuli,

albanuri enebi. Cveni mxridan SesaZlebelia davumatoT qarTuli da

rusuli. am enebSi qvemdebaris poziciaSi mxolod pirdapiri damatebis gada-

weva xdeba.

enaTa meore tipis magaliTebad d. jonsons mohyavs iaponuri da sanskriti;

rogorc viciT, am tips ganekuTvneba inglisuri enac: DO > S John threw the ball jonma isrola burTi The ball was thrown (by John) burTi nasroli iyo (jonis mier). IO > S John gave Mary a book jonma misca meris wigni Mary was given a book (by John) meris mieca wigni (jonis mier).

enaTa mesame tips ganekuTvneba malagasuri da sebuano. malagasurSi

DO>S da IO>S-Tan erTad SesaZlebelia OO>S, sadac OOObenefaqtivia:

Nividi ny vary ho an' ny ankizy ny vehivavy iyida es brinji Tvis am bavSvebis am qalma

„qalma bavSvebisaTvis iyida brinji~.

Nividianan' ny vihivavy ny vary ny ankizy nayidi iqna qalis mier (Som.) es brinji es bavSvebi (S) `bavSvebisTvis(S) nayidi iqna brinji qalis mier~.

am enaSi instrumentalisis gadawevac aris SesaZlebeli subieqtis pozi-

ciaSi:

Nividianan' ny vehivavy ny vary ny vola. nayidi iqna am qalis mier (Som. ) es brinji (DO) am fuliT (S). `am fuliT (S) brinji iyo nayidi qalis mier~. an `es fuli

gamoyenebuli iqna am qalis mier brinjis sayidlad~.

rogorc cnobilia, zmnasTan dakavSirebuli argumentebis topikalizacia

pasiuri transformaciis erT-erT ZiriTad funqcias warmoadgens. subieqti

prototipulad topikuri poziciaa. am poziciaSi pirdapiri, iribi, ubralo

damatebebis gadaweva cvlis winadadebis informaciul struqturas. buneb-

rivi topikis rolSi aRmoCndebian es aqtantebi, xolo amosavali struq-

turis subieqti Somerizacias (CamoqveiTebas) daeqvemdebareba. amgvarad,

fiqsirebuli sityvaTa rigis mqone enebisaTvis pasivizacia topikalizaciis

karg saSualebas warmoadgens.

iseT enebSi, sadac sityvaTa rigi Tavisufalia, topikalizaciis ZiriTad

saSualebas winadadebis wevrTa gadanacvleba warmoadgens. amitom pasiur

91

konstruqcias aseTi enebi am amocanis gansaxorcieleblad naklebad saWi-

roeben. pasivs, cxadia, sxva aspeqtebic aqvs, romlebic mis gamoyenebas

ganapirobeben am enebSi (pasivis gadakveTis Sesaxeb sxva kategoriebTan ix.

Sibatani 1985).

qarTul enaSi pasiur konstruqciaTa gamoyeneba sakmaod SezRudulia.

`vnebiTad~ miCneul formaTa umravlesobis miCneva pirdapiri obieqtis wevis

Sedegad SeuZlebelia. ucxo enaTa pasiuri konstruqciebis qarTuli orga-

nuli `vnebiTebiT~ Targmna xSirad xelovnur konstruqciebs gvaZlevs. Cven

vfiqrobT, rom qarTulisa da sxva qarTveluri enebisaTvis damaxasiaTebeli

winadadebis wevrTa Tavisufali rigi SeiZleba iyos am enebSi transforma-

ciuli (konversiuli) pasivis ganuviTareblobis erT-erTi mizezi.

qarTuli `vnebiTi~ warmoadgens zmnur formaTa rTul klass, romelic

morfologiurad aris markirebuli. `vnebiTebis~ klass aerTianebs awmyos

mwkrivis daboloeba -ebi(I da II pirSi)/-eba(III pirSi). am klasis oTxive

qvejgufis formebi: i-, e- prefiqsebiani, -d sufiqsiani da uniSno warmoeba

_ sakuTari arakonversiuli gramatikuli semantikis matarebelni arian.

i- prefiqsi refleqsuri semantikis mqonea, masSi S da DOp erT pirad aris

gaerTianebuli: imaleba = malavs sakuTar Tavs;

e- prefiqsi subieqtisa da iribi obieqtis urTierTobas asaxavs gardauval

zmnaSi _ amdenad uaxlovdeba reciprokul semantikas. is mas ebrZvis = isini ebrZvian erTmaneTs; is mas ekamaTeba = isini kamaToben erTmaneTTan.

-d sufiqsi gardaqceviTobas gamoxatavs: wiTldeba, brazdeba, mxiaruldeba; xdeba wiTeli, braziani, mxiaruli. uniSno vnebiTi Sinagan, organulad

mimdinare procesebs asaxavs: Tbeba, lRveba, Wkneba, Sreba, lpeba > Tavis-Tavad.

konkretuli semantikis formaTa am oTx qveklass unda hqondes zogadi,

maTi gamaerTianebeli semantikuri maxasiaTebeli, radgan isini qarTuli enis

gramatikulma logikam erT formalur klasad gaaerTiana.

qarTuli zmnebis oTxi ZiriTadi klasi, vfiqrobT, sami diferencialuri

niSnis safuZvelze SeiZleba daxasiaTdes. es niSnebia:

a)gardamavloba/gardauvloba,

b)dinamika/statika da g)miznobrioba/umiznoba (telikuroba/atelikuroba).

a. SaniZis mier gamoyofili oTxi zmnuri klasi gramatikuli semantikis

TvalsazrisiT ase SeiZleba davaxasiaToT:

92

Ggardamavloba dinamika telikuroba

moqmedebiTi

vnebiTi

saSuali

statikuri

+

_

_

_

+

+

+

_

+

+

_

_ SeniSvna: cxrilSi zmnuri klasebis daxasiaTeba CamoTvlili niSnebis dadasturebis

TvalsazrisiT da ar gulisxmobs interpretacias markirebis mimarTebis

TvalsazrisiT.

`vnebiTi~ gvaris zmnebi daxasiaTdeba rogorc gardauvali, dinamikuri da

telikuri. saSuali gvaris zmnebisagan maT ganasxvavebs telikuri (miznob-

rivi) semantika.

Sdr. duRdeba _ aduRebis procesSia, rac aduRebiT dasruldeba; duRs _ duRilis procesSia; es procesi ar gulisxmobs miznis miRwevas, igi

gagrZeldeba manam, sanam raime ar Sewyvets mas. `vnebiTs~ `moqmedebisagan~

ganasxvavebs gardamavlobis sema, magram aerTianebs telikuroba

(miznobrioba). rogorc moqmedebiT, ise vnebiT formas, aqvs mizani, romlis

Sesruleba Tu miRweva mis Sinagan logikaSi aris Cadebuli: awiTlebs da wiTldeba formebisaTvis igulisxmeba, rom es procesi dasruldeba. xolo

is, rac brwyinavs, curavs, naTobs _ araviTar mizans ar isaxavs da an

gagrZeldeba es procesi, an raime garegani faqtori daudebs mas zRvars.

es oTxi zmnuri klasi eTanadeba zmnaTa semantikur klasifikacias, romelic

devid doutim SemogvTavaza (douti 1979):

1. States _ mdgomareobebi qarTulisaTvis: statikuri.

2. Achievements _ miRwevebi ` vnebiTi

3. Accomplishments _ Sesrulebebi ` moqmedebiTi

4. activities _ aqtiurobani ` saSuali

zmnaTa es semantikuri klasebi qarTuli zmnis morfologiur klasebs

pirvelad d. holiskim SeuTanada (holiski 1981). Cven vfiqrobT, rom

qarTuli ena semantikuri markirebis enaa (meliqiSvili 2005). misi morfo-

logiuri wyoba ZiriTadad orientirebulia semantikur da ara sintaqsur

kategoriebze. es kargad Cans pirisa da qcevis kategoriebis analizisas _

piris binaruli opozicia orientirebulia aqtantebis semantikur maxasi-

aTeblebze, qcevis kategoria ki centriskenuli-centridanuli opoziciis

TvalsazrisiT sametyvelo aqtSi monawile I da II pirze. is, rom morfo-

logiuri kriteriumebis safuZvelze dadgenili qarTuli enis oTxwevra

klasifikacia zmnaTa universalur semantikur klasSi poulobs srul Sesa-

bamisobas, kidev erTi mowmobaa imisa, rom qarTuli ena semantikuri

markirebis enas warmoadgens.

93

rogorc wesi, qarTuli enis zmnebi sxvadasxva klasis formebiT SeiZleba

iyos warmodgenili. SesaZlebelia maTi gadayvana erTi klasidan meoreSi.

amasTan, sxvadasxva semantikis zmnebi sxvadasxva klasiT warmogvidgenen

amosaval, pirvelad formas.

1. moqmedebiTi > vnebiTi > statikuri

wers _ iwereba _ weria

Tesavs _ iTeseba _ Tesia

xatavs _ ixateba _ xatia

yris _ iyreba _ yria

2. saSuali > moqmedebiTi > vnebiTi

goravs _ agorebs _ gordeba

duRs _ adurebs _ duRdeba

mefobs _ amefebs _ mefdeba

wevs _ awvens _ wveba

3. vnebiTi > moqmedebiTi

Tbeba _ aTbobs

dneba _ adnobs

lpeba _ alpobs

statikur zmnebs yovelTvis eZebneba fardi moqmedebiTi klasis zmna, romel-

Tan mimarTebaSic Znelia Tqma _ romelia amosavali forma: statikuri Tu

moqmedebiTi.

yria _ yris

marxia _ marxavs

Tesia _ Tesavs

Tu statikur zmnebs mivakuTvnebT saSual-vnebiTi zmnebis klass, maSin

statikuri formebic iqneba amosavalTa Soris. amis uflebas iZleva am

zmnebis semantika da pirveli da meore piris formebi, romlebic statikur

zmnaTa msgavsad -var, -xar meSvelzmnuri elementis darTviT iwarmoeba: Sdr.

v-we-var, wev-xar, wevs

v-gdivar, gdixar, gdia

gansxvavebas qmnis mesame piris formebis daboloeba -s da -a, amas garda

momavalisa da aoristis warmoeba:

wevs _ i-wveba _ iwva

gdia _ e-gdeba _ egdo

wevs, dgas, zis tipis formebi gardamaval formebad SeiZleba miviCnioT

statikur da saSual zmnebs Soris. semantikurad isini statikurebia da

doutis klasifikaciis TvalsazrisiT statikurTa klasSi moTavsdebodnen.

saSual-vnebiTebis statikurebTan siaxloveze vrclad msjelobs ukve a.

SaniZe (SaniZe 1973, 320-321).

94

amgvarad, klasebis mixedviT zmnur gadanacvlebisaTvis SesaZlebelia amo-

savali forma iyos rogorc moqmedebiTi, ise saSuali da ise vnebiTi forma,

zmnis pirveladi semantikis Sesabamisad. zogi zmnis pirveladi semantika

iseTia, rom misTvis bunebrivi amosavali forma aris gardamavali, moqme-

debiTi, zogisa iseTi, rom misi amosavali forma bunebrivad aris vnebiTi

(gardauvali, telikuri), zogisa ki iseTi, rom amosavali forma bunebrivad

aris saSuali (gardauvali, atelikuri).

zemoT warmodgenili derivaciuli gadanacvlebebis safuZvelze SeiZleba

iTqvas: Tu pirveladi zmna moqmedebiTia, mas ewarmoeba `vnebiTisa~ da

statikuri formebi; Tu pirveladi zmna saSualia _ mas ewarmoeba vnebiTisa

da moqmedebiTis formebi; Tu amosavali zmna vnebiTia, mas ewarmoeba moqme-

debiTis formebi. e.i. moqmedebiTi da vnebiTi ver gadava saSualTa klasSi,

saSuali ki gadadis moqmedebiTisa da vnebiTis klasebSi. e.i. saSuali

gvevlineba mxolod pirveladi formebis saxiT da arasdros ar aris

derivati.

SesaZlebelia davaskvnaT, rom `vnebiTebis~ klasis ZiriTadi da pirveladi

funqcia ar unda yofiliyo transformaciul-konversiuli. es aris zmnaTa

garkveuli semantikur-formaluri klasi, romelic semantikurad ganisaz-

Rvreba rogorc gardauvali, dinamikuri, telikuri (doutis klasifikaciis

mixedviT `miRwevis~ klasi), xolo formalurad misi morfologiuri mar-

keria -ebi/-eba daboloeba. qarTuli ena, rogorc Tavisufali sityvaTa rigis

mqone ena, topikalizacias Tavisuflad awarmoebs winadadebaSi sityvebis

gadanacvlebiT, Sesabamisad, mas am mizniT vnebiTi gvaris gamoyenebis naklebi

saWiroeba aqvs. vfiqrobT, esec SesaZlebelia iyos imis mizezi, rom

qarTuli vnebiTi naklebad aris konversiuli xasiaTisa da is garkveuli

semantikis mqone (dinamikur, telikur) zmnur klasad warmogvidgeba.

literatura asaTiani, wereTeli 2008 _ r. asaTiani, n. wereTeli, winadadebis

sainformacio struqtura, implikaciuri topiki qarTulSi, enaTmecnierebis

sakiTxebi, II.

douti 1979 _ D. Dowty, Word Meaning and Montague Grammar, Dodrecht, Reidel. meliqiSvili 2005 _ i. meliqiSvili, pirdapiri semantikuri markireba,

rogorc qarTvelur enaTa morfologiis ganmsazRvreli principi da piris

niSanTa inversiis problema, tipologiuri Ziebani, V, Tbilisi.

perlmuteri, postali 1977 _ D. M. Perlmutter, P.M. Postal, Toward Universal Characterization of Passivization, Proceedings of the Third Annual Meeting of the Berkley Linguistic Society, Berkley. SaniZe 1973 _ a. SaniZe, qarTuli enis gramatikis safuZvlebi, Tbilisi.

Sibatani 1985 _ M. Shibatani, Passives and Related Constructions: A Prototype Analysis, Language, v, 61, # 4.

95

jonsoni, 1977 (1982) _ D. E. Johnson, On Relational Constraints on Grammars, I, Syntax and Semantics, v, 8, Academic Press, New York, 1977; rusuli Targmani

dabeWdilia krebulSi: Новое в зарубежной лингвистике. Москва, 1982. holiski 1981 _ D. A. Holisky, Aspect and Georgian Medial Verbs, Caravan Books, Delmar, New York.

Irine Melikishvili

Passivization and Flexible Word Order as Alternative Strategies for the Topcalization of Phrases

Abstract

Languages have a number of different strategies for signaling the topic of the sentence. Among these are the use of the passive voice and the placing a phrase in a topic position. In Georgian, a language with flexible word order, the change of the dominant word order is the main strategy for topicalization of phrases. This reduces the need of employment of the passive transformations for the same aim. Accordingly, the passive voice is weekly developed in Georgian. The verbs, considered as belonging to the class of passives, are not transforms of the active forms as a rule. They do not have always the active counterparts and possess their own specific semantics: i- verbs are reflexive (i-maleba “hides himself”), e- verbs have often the reciprocal meaning (e-brývis “fights with him”), the -d verbs have the semantics “become” (lamaz-d-eba “becomes pretty”), the unmarked passives denote internal processes, taking place within the subject (dneba “melts” êreba, xmeba “dries”). This verb class, marked morphologically with -ebi/-eba exponent, can be defined as the class of intransitive, dynamic, telic verbs. It corresponds to the verb class of Achievements in the verb classification of David Dowty. The week development of the transformational passive voice in Georgian must be due to the existence of other possibility for topicalization of phrases – the flexibility of the word order. The same is true for other Kartvelian languages – Megrelian-Laz and Svan.

96

eTer soselia

sxvaTa sityvis -o nawilaki

da misi funqcia teqstSi

-o nawilakis Sesaxeb msjeloba sasurvelia daviwyoT imis gansazRvriT, Tu

enobrivi analizis romeli donis erTeulad unda ganixilebodes igi.

rogorc cnobilia, -o nawilaki gvxvdeba sityvis absolutur boloSi

(auslautSi). TiTqos mosalodnelia, rom igi morfologiuri donis

erTeulad miviCnioT, ramdenadac morfologia swavlobs sityvis formas,

mis struqturas _ agebulebas mniSvnelobis mqone umciresi enobrivi

erTeulebis (morfemebis) urTierTmimarTebis TvalsazrisiT. rac Seexeba

TviT mniSvnelobas, igi SeiZleba iyos leqsikuri an gramatikuli. Tuki -o nawilaks morfologiuri donis erTeulad miviCnevT, calsaxad

SeiZleba iTqvas, rom igi araa leqsikuri mniSvnelobis matarebeli. amave

dros, misi mniSvneloba arc gramatikulia, ramdenadac gramatikuli

mniSvneloba gulisxmobs enobriv erTeulebs Soris garkveuli mimarTe-

bebis gamoxatvas.

Tumca SeiZleba iTqvas, rom -o nawilakis semantika ramdenadme aaxlosaa

gramatikul mniSvnelobasTan, radganac igi garkveul mimarTebas aRniSnavs.

magram es aris mimarTeba ara enobriv erTeulebs Soris, aramed

sametyvelo (lokuciur) aqtebs Soris. qvemoT dawvrilebiT ganvixilavT

am mimarTebas, aq ki gvinda aRvniSnoT, rom sxvaTa sityvis -o nawilaki

erTvis ara sityvas, aramed mTel gamonaTqvams da zogjer es gamonaTqvami

winadadebaze metia.

zemoTqmulis mixedviT cxadia, rom -o nawilaki ar warmoadgens morfo-

logiuri donis erTeuls. mas verc sintaqsur dones mivakuTvnebT, ramde-

nadac igi winadadebaSi arc raime funqcias asrulebs da arc winadadebis

wevrebs akavSirebs erTmaneTTan. rogorc Cans, -o nawilaki ufro maRali

ierarqiuli donis erTeulia. es aris teqstis, diskursis done, sadac

igi pragmatikul funqcias asrulebs.

imisaTvis, rom davadginoT -o nawilakis funqcia, saWiroa ganvsazRvroT

mis mier aRniSnuli mimarTeba, rac mxolod maSin gaxdeba SesaZlebeli,

Tu mas ganvixilavT erTobliobaSi sxva sityva-sityviT nawilakebTan

erTad. es nawilakebia: -meTqi, -Tqo (-Tqva). rogorc a. SaniZe aRniSnavs `qarTuli enis gramatikis safuZvlebSi~

(SaniZe 1980), `meTqi warmomdgaria ori sityvis SeerTebisagan: `me vTqvi~. amitomaa, rom igi moubari piris naTqvamis gameorebas an ganazraxi

sityvis Tqmas aRniSnavs.~ sailustraciod avtors moxmobili aqvs Semdegi

magaliTi:

97

`vuTxari: gagikeTeb-meTqi~.

zmnuri warmoSobisaa Tqva, Tqo nawilakebic, romelTa Sesaxeb a. SaniZe

(SaniZe 1980) aRniSnavs: ` Tqva, Tqo mesame piris formaa wyvetilisa [o

miRebulia ua jgufis SerwymiT]. aRniSnavs moubari piris danabarev

sityvas, romelic meore pirma mesames unda gadasces.~ sailustraciod aq

Semdegi magaliTia moxmobili:

àse moaxsene: me TviTon gavscem magas pasuxs-Tqva~ (`pasuxs-Tqo~).

rac Seexeba CvenTvis saintereso -o nawilaks, mis Sesaxeb avtori

aRniSnavs: ò gadmogvcems mesame piris naTqvams. igi sxvaTa sityvis

saxeliTaa cnobili.~ sailustraciod moxmobili magaliTi Semdegia:

`dedam SemogiTvala: cota fqvili gvasesxeo~.

iqve a. SaniZe dasZens, rom sxvaTa sityvis o xSirad axlavs andazebs,

radganac igulisxmeba: àmboben~, ùTqvamT~, `naTqvamia~.

warmodgenili gansazRvrebebidan cxadi xdeba, rom sityva-sityiTi nawi-

laki ukavSirdeba konkretul sametyvelo aqts, romelic garkveul

mimarTebaSia winamaval an momdevno sametyvelo aqtTan. amdenad, sityva-

sityviTi nawilaki Tavis TavSi gulisxmobs or sametyvelo aqts da

swored am aqtebs Soris arsebuli mimarTeba aris misi mniSvneloba.

TiToeuli sityva-sityviTi nawilakiT aRniSnuli mimarTeba (sametyvelo

aqtebs Soris) sqematurad Semdegi saxiT SeiZleba warmovadginoT:

= I sametyv. aqti II sametyv. aqti (-meTqi) sqema 1. = I sametyv. Aaqti (-Tqo /-Tqva) II sametyv. aqti

sqema 2.

molap. msmen. molap. Mmsmen.

molap. msmen. molap. msmen.

98

=

/x I sametyv. aqti II sametyv. aqti (-o) sqema 3. -meTqi nawilaks (sqema 1) molaparake iyenebs II sametyvelo aqtSi da amgvarad

gadmogvcems im sityvebs, romlebic man winamaval sametyvelo aqtSi warmoTqva.

mimarTeba, romelsac aRniSnavs nawilaki -meTqi aris identurobis mimarTeba am ori sametyvelo aqtis molaparakeTa Soris.

-Tqo (-Tqva) nawilaks (sqema 2) molaparake iyenebs I sametyvelo aqtSi

da daurTavs im gamonaTqvams, romelic msmenelma momdevno sametyvelo

aqtSi (sadac is ukve molaparake iqneba) gadasces sxva msmenels. mimar-

Teba, romelsac aRniSnavs nawilaki -Tqo (-Tqva) aris identurobis

mimarTeba winamavali sametyvelo aqtis msmenelsa da momdevno sametyvelo

aqtis molaparakes Soris.

-o nawilaks (sqema 3) molaparake iyenebs II (pirobiTad) sametyvelo

aqtSi da daurTavs im gamonaTqvams, romelic warmoiTqva winamaval

sametyvelo aqtSi, sadac is arc molaparake yofila da arc msmeneli

(amitomac ewodeba –o-s sxvaTa sityvis nawilaki). mimarTeba, romelsac

aRniSnavs nawilaki -o aris identurobis mimarTeba winamavali sametyvelo

aqtis gare wevrsa da momdevno sametyvelo aqtis molaparakes Soris. me-3

sqemasTan dakavSirebiT SevniSnavT: X elementi SeiZleba identuri iyos

winamavali sametyvelo aqtis msmenelisa da am SemTxvevaSi TiTqos me-2

sqemas daemTxveva, Tumca gansxvaveba am or sqemas Soris mainc iqneba

(Tundac imis gamo, rom sityva-sityviTi nawilaki, me-2 sqemis mixedviT, I

sametyvelo aqtSi gamoiyeneba).

SesaZloa, -o nawilakiani sametyvelo aqti iyos uSualo gagZeleba -Tqo (-Tqva) nawilakiani sametyvelo aqtisa:

I sametyvelo aqti (deda eubneba Svils):

Sin rom mixval, mamas Cemgan uTxari, male moval da erTad visadiloT-Tqo.

II sametyvelo aqti (Svili eubneba mamas):

dedam Tqva, male moval da erTad visadiloTo.


99

aRniSnuli diskursuli situacia sqematurad Semdegnairad warmoidgineba:

= I sametyv. Aaqti (-Tqo /-Tqva) II sametyv. aqti (-o) sqema 4.

amgvarad, mas Semdeg, rac ganvixileT sityva-sityviTi nawilakebis

mniSvneloba, SeiZleba davaskvnaT, rom maTi funqcia diskursul-

pragmatikulia.

sityva-sityviT nawilakTagan amjerad yuradRebas SevaCerebT -o nawilakze.

Cveni mizania davadginoT, zemoT ganxiluli ZiriTadi funqciis garda,

damatebiT kidev ra funqciis Sesruleba SeuZlia sxvaTa sityvis nawi-

laks. kvlevis sawyis etapze saanalizo masalad SevarCieT qarTuli xal-

xuri zRaprebi, ramdenadac, erTi SexedviT, am tipis naratiul teqstebSi

sakmaod xSiria -o nawilakis gamoyeneba.

rogorc gamoirkva, -o nawilaki yvelaze xSirad gvxvdeba Tavisi ZiriTadi

funqciiT, e.i. roca molaparake konkretul sametyvelo aqtSi sityva-

sityviT gadmogvcems wina sametyvelo aqtSi realizebul gamonaTqvams.

gasagebia, rom am tipis gamonaTqvamebSi, rogorc wesi, sametyvelo

aqtivobis aRmniSvneli zmnebi gamoiyeneba, rogoricaa, magaliTad, uambo, uTxra, moaxsena, ÛSesZaxa, ÛSeekiTxa, SesCivla da sxva. sametyvelo aqtivobis aRmniSvneli zmna SeiZleba win uswrebdnen -o nawilakian gamonaTqvams:

mamam ikiTxa: _ movida Cemi Svilio? `xelmwifis Svili da bayayi~

zogjer amgvari zmna mosdevs -o nawilakian gamonaTqvams:

_ ... is Tu gviSvelis, sxva veravinao, _ uTxra Svils. `moxuci durgali da misi rZlebi~

zogjer igi gamonaTqvams nawilebad xleCs:

_ iq ierusalims ra xeli aqvs, _ uTxra mgelma, _ moemzade axlave unda SegWamoo.

`mglis berad Sedgoma~

aRniSnuli tipis zmnebi umetesad warsuli drois formiT gvxvdeba,

Tumca zogjer awmyos formiTac dasturdeba.


100

erTxel mama eubneba Svils: _ aba, Svilo, axla mometebuli jafa da Sroma gmarTebs, ojaxi unda fexze daayenoo.

`moxuci durgali da misi rZlebi~

am dros II sametyvelo aqti, roca avtori zRapars gviyveba, TiTqos

Tanadrouli xdeba im sametyvelo aqtisa, roca mama esaubreba Svils.

aRniSnuli tipis zmnebis garda sxvaTa sityvis nawilakTan SeiZleba

Segvxvdes zmnebi: gaifiqra, gadawyvita, guli gaikeTa, ... es ki imas

niSnavs, rom sxvaTa sityvis nawilakiT gadmocemuli gamonaTqvami

faqtobrivad arc yofila konkretul metyvelebaSi realizebuli:

mefis vaJi sul imas fiqrobs, saidan Semoevleba Cems saqmes, rom ocneba gavinaRdoo. `xvTisavari~

sxvaTa sityvis -o SeiZleba Segvxvdes sametyvelo aqtivobis (rogorc

realuri, ise virtualuri) gamomxatveli zmnebis gareSe, Tumca kon-

teqsti maT gulisxmobs (e.i sametyvelo aqtivobis gamomxatveli zmnebis

elifsis dros):

mTieli mivida moxuc kacTan. _ Zmobas, erTi Taiguli momecio. `lali, iadavardi da varskvlava~ Tavisi ZiriTadi funqciiT (e.i. roca molaparake konkretul sametyvelo

aqtSi sityva-sityviT gadmogvcems wina sametyvelo aqtSi realizebul

gamonaTqvams) -o nawilaki SeiZleba Segvxvdes sakmaod vrceli

gamonaTqvamis (romelic ramdenime winadadebisagan Sedgeba) bolos:

qalma uTxra: _ ager, cxra weliwadi sruldeba, TmiTa vkidivar. cxraTavianma devma momitaca, Cemi colad SerTva undoda da me ar vnebdebi. mcems, mtanjavs, mawvalebs, magrm raki erTxela vTqvi uari, ver gadamaTqmevina. exla es tye rom xmaurobs, dedamiwa rom inZreva, modis sadme. Ddaimale, Torem mogklavso. `lali, iadavardi da varskvlava~

-o nawilaki SeiZleba kiTxviT winadadebasac daerTos:

wamoiRes melia, gaatyaves da uTxres vaJsa: _ ra mogceT, rom eg melia mohkal da gadagvarCineo? `zRapari-gamocana~

zogjer wina sametyvelo aqtSi realizebuli gamonaTqvamis gadmocemisas -o nawilaki SeiZleba ramdenjerme Segvxvdes. am SemTxvevaSi sxvaTa

sityvis nawilaki iZens damatebiT funqcias, romelic gamonaTqvamis

101

informaciul struqturas ukavSirdeba. kerZod, igi mianiSnebs, rom -o nawilakis pirveli gamoCenis momdevno winadadeba fokusirebulia:

vaJma uTxra: _ qalbatono, eg ra aris, is unda naxoT, me rom gemze

maqvso. Tu gnebavT, wamobrZandiT da naxeTo! `monadiris Svili~

rogorc cnobilia, sainformacio struqtura mraval sxvadasxva TeoriaSi

ganixileba. erT-erTi maTgania semantikuri alternativebis Teoria,

romelic ZiriTadad fokusisa da fokusirebis problemas Seiswavlis (m.

krifka 2008). aRniSnuli Teoriis mixedviT, fokusi sazogadod miuTi-

Tebs, rom arsebobs alternativebi gamonaTqvamis (winadadebis) inter-

pretaciisa da rom fokusi aris konkretuli pasuxi SekiTxvaze,

romelsac ramdenime alternatiuli pasuxi aqvs. amdenad, fokusi gulis-

xmobs SekiTxvas. mocemul nawyvetSi Semdegi SekiTxva ivaraudeba:

rogor unda vnaxo is? amgvarad, zemoT moxmobil nawyvetSi fokusirebuli winadadeba Tavisi

fokusiT Semdegia:

Tu gnebavT, wamobrZandiTF da naxeTo!

an meore magaliTSi:

mere uTxra devma: _ ras ambob, dedakaco, Cems suls magaSi ra undao? Cemi suli niCbis tarSiao. `samni Zmani~

am SemTxvevaSi fokusi gulisxmobs SekiTxvas:

sad aris Cemi suli? amjerad fokusirebuli winadadeba Tavisi fokusiT Semdegia:

Cemi suli niCbis tarSiaF-o.

rodesac sxvaTa sityvis -o fokusze ar mianiSnebs, anu neitraluri

funqciiT gvxvdeba, igi, rogorc aRvniSneT, mxolod gamonaTqvamis

absolutur boloSia da, rogorc wesi, am gamonaTqvamis bolo wevri

zmna-Semasmenelia (zemoT moxmobili ramdenime magaliTi amis dasturia).

magram im SemTxvevaSi, Tu gamonaTqvamis bolo wevri, romelsac sxvaTa

sityvis nawilaki daerTvis, zmna-Semasmenlisgan gansxvavebuli romelime

sxva wevria, -o nawilaki kvlav fokusirebis niSania:

amoiRo devma erTi gasaRebi da vaJs misca. _ es im oTaxisaa, romelSiac me viyavi dabmuli. Sen is

oTaxi gaaRe da naxe, ra sanaxavebi da sakvirveli Tval-margalitia Sigao!

`samni Zmani~

102

am nawyvetSi fokusirebuli winadadeba Tavis fokusiT Semdegia:

naxe, ra sanaxavebi da sakvirveli Tval-margalitia SigaFo!

garda imisa, rom -o nawilaki informaciis struqturirebaSi monawi-

leobs, igi garkveuli emociis gamosaxatavadac gamoiyeneba, rac

gansakuTrebiT mokle SekiTxvebSi igrZnoba. magaliTad, zRaparSi `melia

da Citi~ mocemlia dialogi:

_ Cito, Cito, Ciorao! _ rao, batono melao?

Citis SekiTxvaSi kiTxviT sityvaze darTuli -o nawilaki gulisxmobs:

`ras mobrZanebT? ra gnebavT? ra saSinelebas mipireb?..~ da gamoxatavs

mRelvare, SiSiT savse molodins: Cits eSinia imisi, Tu ras etyvis mela.

sxvaTa sityvis nawilakis damtebiTi funqcia ufro mkafiod gamoixateba

im SemTxvevaSi, roca verbaluri aqtivobis gamomxatveli zmna araa

warmodgenili. magaliTad, roca zRaparSi `rwyili da WianWvela~ rwyili

midis RorTan da sTxovs jagars, Rori pasuxobs:

_ Sen rom CemTvis rko mogitaniao.

Semdeg rwyili midis muxasTan da sTxovs rkos. muxac msgavsadve

pasuxobs:

_ Sen rom CemTvis yvavi mogiSorebiao. am pasuxebSi -o nawilaki erTgvar sayvedurs gamoxatavs (rac faqtob-

rivad uars niSnavs) imis gamo, rom rwyils, erT SemTxvevaSi, RorisTvis

arasdros miutania rko, xolo meore SemTxvevaSi, muxisTvis arasdros

mouSorebia yvavi.

amjerad -o nawilakis zemoT ganxiluli funqciebiT SemovifarglebiT. es

is funqciebia, romlebic aSkarad gamoikveTa zRaprebSi. sxvaTa sityvis

nawilakis funqciaTa srulyofli analizisaTvis aucilebelia sxva tipis

teqstebis Seswavlac; aseve, Zalian saintereso Cans -o nawilakis funqci-

onireba zepir metyvelebaSi, rac kvlevis Sedgomi etapisTvis ivaraudeba.

literatura krifka 2008 _ m. krifka, sainformacio struqturis ZiriTadi cnebebi, krebulSi: `ena, logika, kompiuterizacia~, Tbilisi-vena.

qarTuli zRaprebi, (Seadgina e. virsalaZem), Tbilisi, 1984.

SaniZe 1980 _ a. SaniZe, qarTuli enis gramatikis safuZvlebi, Txzulebani

12 tomad, tomi III, Tbilisi.

103

Ether Soselia

Direct speech -o particle and its functions in text

Abstract

Direct speech -o particle together with -metki, -tko/-tkva other ones compiles a group of so-called word-for-word particles. The paper deals with -o particle and brings its functions out on the data of the Georgian Fairy-Tales. Direct speech -o particle is usually used to transmit the third person’s speech. This is its main function and generally verbs denoting speech activity appear with it. Usual position of -o particle is at the end of an utterance, even though the latter is quite extended. In addition, -o particle may have a certain function from the information structure point of view. It happens when -o particle is used twice (or more times) in the utterance. In this case the clause following each position of -o, except the one at the end, is focused. Besides, -o particle is used to convey some emotions, especially in short questions, when speech activity verbs are not presented.

104

nuca wereTeli, rusudan asaTiani

implikaciuri topiki qarTulSi

1. Sesavali

1.1. sakiTxis dasma

winadadebaSi informaciis struqturirebis sxvadasxvagvari saSualebebi cvlis

winadadebis pragmatikul Sinaarss da implikaciurad ufro farTo konteqstur

Sinaarsebs moicavs. am fenomenis sakvlevad Tanamedrove enaTmecnierebaSi Semu-

Savda garkveuli meTodologia da ganisazRvra ZiriTadi cnebebi, romelTa

saSualebiT, met-naklebi warmatebiT, SesaZlebelia winadadebis sainformacio

struqturis kvleva. fokusi da topiki winadadebis sainformacio struqturis

ganmsazRvreli swored aseTi ZiriTadi cnebebia.

1.2. fokusi

winadadebaSi fokusis garda warmodgenil informacias ewodeba foni (Common Ground). orive maTgani gamoxatavs sxvadasxva datvirTvis mqone informacias. fokusi ZiriTadad gamoxatavs axal informacias, romliTac sainformacio

nakadSi, ZiriTadad kiTxva-pasuxis diskursSi, ivseba informaciuli xarvezi,

xolo foni Zvel informacias, romelic mosaubreTa saerTo safuZvels qmnis.

mag.: kiTxvaze – vin daxata suraTi? – pasuxSi: biWma daxata suraTi –

fokusirebuli iqneba biWi, xolo fons (Common Groud) qmnis komunikaciis

aqtSi monawileTa saerTo ‘codna’ imis Sesaxeb, rom viRacam daxata suraTi. sakomunikacio aqtis dros, informaciis struqturirebisas, foni uwyvetad

icvleba da ivseba mopovebuli informaciisda Sesabamisad. u. l. Ceifi aris erT-erTi pirveli mecnierTagani, romelmac enaTmecnierebaSi

daamkvidra termini fokusi. igi saubrobda fokusze, rogorc informaciis

SefuTvis meqanizmze, romelic uSualod pasuxobs mosaubreTa sakomunikacio

moTxovnilebebs. u. Ceifi fokuss uwodebs axal informacias, romelic gamokve-

Tili intonaciis wyalobiT warmoaCens informaciis centralur nawils; xolo

foni gamoxatavs Zvel informacias, romelic safuZvels qmnis axali infor-

maciis mosapoveblad.

manfred krifkas azriT, yvelaze zusti gansazRvreba fokusisa unda asaxavdes

SesaZlebel pasuxTa alternatevebis rigs. mag.: zemoT moxmobil kiTxvaze – vin xatavs suraTs? – pasuxi SeiZleba iyos: biWi, gogo, kaci, qali, mxatvari da a.S. am SesaZlebel pasuxTa alternativebidan, situaciisda Sesabamisad, irCeva

erT-erTi, romelic swored am SerCevis wyalobiT aris fokusirebuli. aqedan

gamomdinare, m. krifka iZleva fokusis ufro zust definicias: fokusi miuTi-

Tebs im alternativebis arsebobaze, romlebic relevanturia lingvistur

gamoxatulebaTa interpretirebisTvis.

105

amdenad, fokusi, zogadad, gviCvenebs, rom arsebobs interpretaciis sxvadasxva

alternativebi. fokusis sxvadasxva qvetipi swored am zogadi ideis sxvada-

sxvagvari variaciaa. fokusirebuli SeiZleba iyos winadadebis sxvadasxva zomis

Semadgeneli: mTeli winadadeba, saxeluri Tu zmnuri frazebi an maTi nawilebi

(zedsarTavebi, CvenebiTi an kuTvnilebiTi nacvalsaxelebi, garemoebebi da sxv.).

zogjer gamoyofen farTo da viwro fokuss. qvemoT moxmobil magaliTSi

fokusi TandaTan viwrovdeba B1-dan B4-mde (C. Gessenhoven): A1: What else can you tell us about Helen? B1: She [used to drive a Renault CLIO]Foc A2: Has she driven any other cars besides Fords and Chevrolets? B2: She used to drive [a Renault CLIO]Foc A3: What kind of Renault did she drive? B3: She used to drive a Renault [CLIO]Foc A4: Does she drive a Renault CLIO? B4: She [USED TO]Foc drive a Renault Clio B1 farTo fokuss warmogvidgens, xolo B2, B3, B4 sxvadasxva saxis viwro fokusebs. yvela pasuxi erTi da imave winadadebiT aris warmodgenili, infor-

maciul sxvaobas qmnis mxolod intonaciuri konturi, rac sxvadasxva fokusis

maCvenebelia. am intonaciebis areva pasuxs araadekvaturad aqcevs. magaliTad, A1-ze B2, B3, an B4 pasuxTa Sesabamisi intonaciuri konturi araadekvaturi iqneba da sakomunikacio situaciis rRvevas gamoiwvevs.

arsebobs sxvadasxva tipis fokusebi, magaliTad:

1.Verum fokusi, romelic winadadebis WeSmarituli Rirebulebis

fokusia. is ZiriTadad gamoixateba damxmare zmnis aqcentirebiT: She DOES like broccoli 2. kompleqsuri (Complex) fokusi, romelic or an met Semadgenels aerTianebs da ver daiyvaneba erTamde: John only introduced BILL to SUE. 3. mravalricxovani (Multiple) fokusi, romelic erTsa da imave winadadebaSi warmogvidgens or (an met) damoukideblad fokusirebul Semadge-nels: John only introduced BILL only to SUE. 4. kontrastuli (Contrastive) fokusi gvaqvs iseT SemTxvevebSi, rodesac kiTxvaSive SemosazRvrulia alternativebis rigi. is gamoiyeneba mxolod nam-

dvilad kontrastuli azris gamosaxatavad: A: What do you want to drink tea or coffee? B: I want TEA. 5. amomwuravi (Exhaustive) fokusi gviCvenebs imas, rom is aris erTaderTi swori alternativa: It’s JOHN and BILL that stole a cookie. (anu: swored JOHN-ma da BILL-ma moipares namcxvari da sxvam aravin.) 6. Skaluri (Scalar) fokusis SemTxvevaSi (romelsac aseve emfatikur (Emphatic) fokuss uwodeben) alternativebi dalagebulia Skaluri principiT

da fokusi aris am alternativebis yvelaze dabali an yvelaze maRali (anu polaruli) elementi: WILD HORses wouldn’t drag me there.

106

1.3. topiki

fokusisagan gansxvavebiT, topiki ZiriTadad Zvel informacias warmogvidgens

da informaciis pragmatulad aqtualizebul nawils gamoxatavs: zogierT

SemTxvevaSi situaciis asaxvisas saWiro xdeba informaciis dazusteba, misi

pragmatuli Rirebulebebis gaxazva, movlenebs Soris arsebul (an ararsebul)

kavSirebze miniSneba, informaciis mosalodneloba-moulodnelobis asaxva,

informaciaSi implikaciuri an presupoziciuli mimarTebebis dafiqsireba da

sxva amgvari. topiki swored amgvar damatebiT informacias gadmogvcems da

ZiriTadad aris winadadebis is wevri, romlis Sesaxebac movipovebT infor-

macias. topikis dasadgenad yvelaze moxerxebulia frazebi: ras ityviT X-is Sesaxeb? Semdgomi informacia swored am X-is Sesaxebaa da, rogorc wesi,

gulisxmobs gamonaTqvams: rac Seexeba X-s, is... Tu fokusiT xdeba ucnobi

informaciis Sevseba, topiki, piriqiT, cnobilia da mis Sesaxeb vagrovebT dama-

tebiT informacias.

fon der gabelenci iyo pirveli mecnieri, romelmac topiki gamoiyena

mosaubris mier nagulisxmebi obieqtis aRsaniSnavad. im terminTa Soris,

romlebic ukavSirdeba komunikaciisas informaciis struqturirebas, topiki

aris iseTi cneba, romlis identificirebasac axdens mosaubre da romlis

Sesaxebac Semdgomi informacia, anu komentari, aris mocemuli. terminebi

‘topiki’ da ‘komentari’ xSirad gamoiyeneba e.w. ‘fsiqologiuri subieqtisa’ (resp. is, razec fiqrobs mosaubre) da ‘fsiqologiuri predikatis’ (resp. is, rasac am ukanasknelze fiqrobs mosaubre) mimarT. magaliTad, winadadeba Aristotle Onassis married Jacqueline Kennedy SeiZleba iyos sxvadasxvagvarad

interpretirebuli: Aa. [Aristotle Onassis]Topic [married Jacqueline Kennedy]Comment b. [Jacqueline Kennedy]Topic [married Aristotle Onassis]Comment a-SemTxvevaSi vsaubrobT, informacias movipovebT a. onasisis Sesaxeb da,

faqtiurad, vpasuxobT kiTxvaze: ras ityviT onasisis Sesaxeb?; xolo b-SemTxvevaSi – Jaklin kenedis Sesaxeb: ras ityviT kenedis Sesaxeb?

amgvarad, SeiZleba warmovadginoT topikis Semdegi definicia: topikuri Semad-

geneli aris erTeuli an erTeulTa simravle, romlis Sesaxebac komentarSi

gadmocemuli informaciaa mopovebuli.

arsebobs sxvadasxva tipis topikebi, magaliTad:

1. kontrastuli, romelic gviCvenebs kontrasts wina diskursul

TemasTan. is SeiZleba gamoiyenebul iqnes wina Temidan axal Temaze gada-

sasvlelad, ramdenadac axal informacias gvawvdis ara Zvel, aramed ZvelTan

dapirispirebul, topikalizebul, SedarebiT axal Semadgenelze: A: Do you think that Fritz would buy this suit? B: Well, [I]Top certainly wouldn’t. (am SemTxvevaSi mopasuxe B pirdapir ar pasuxobs A-s SekiTxvas, aramed iZleva gansxvavebul, Tumca kiTxvasTan dakavSirebul, kontrastul, pasuxs.)

107

2. nawilobrivi, romelic gamoiyeneba ZiriTadad im SemTxvevebSi,

rodesac mosaubre Tavs aridebs kiTxvaze

pirdapir, amomwurav pasuxs da Zveli informaciidan mxolod garkveuli nawi-

lis Sesaxeb gvawvdis axal informacias:

A: What did the pop stars wear? B: [The female pop stars]Top wore [caftans]Foc. (am SemTxvevaSic, mopasuxe B ar pasuxobs kiTxvaze pirdapir, yovel SemTxvevaSi, amomwuravad.)

3. implikaciuri, romelic gvawvdis informacias ara mxolod

moTxovnil wevrze, aramed nagulixmeb (ZiriTadad, sapirispiro situaciaSi

myof) wevrzec: A: What about the men, do they smoke cigarettes? B: [Only men]Top smoke cigarettes.(am pasuxSi igulisxmeba, rom mxolod

kacebi (da ara qalebi) ewevian sigarets)

1.4. topikisa da fokusis warmodgenis enobrivi saSualebebi

formaluri TvalsazrisiT, topiki da fokusi SeiZleba markirebul iqnes

sxvadasxva enobriv doneze:

1. fonetika-fonologiis (intonacia, maxvili, toni, specifikuri fone-

tikur-fonologiuri procesebi: gamJRereba-dayrueba, asimilacia-disimilacia

da sxv.);

2. morfologia-sintaqsis (morfologiuri markeri, specialuri brunva,

Serwymuli nawilaki an klitika; sityvaTa rigis cvla, elifsi, specifikuri

konstruqciebi, frazebis gagleja, tmesi da sxv.);

3. leqsika-semantikis (specialuri sityvebi, nawilakebi, artikli,

kvantoruli sityvebi: mxolod, marTlac, namdvilad, aseve, kidec da sxv.).

xSirad es formaluri meqanizmebi Tanaarsebobs da Sedegad vRebulobT rTul,

kompleqsur enobriv modelebs, romlebic emsaxureba informaciis struqtu-

rirebasa da SefuTvas winadadebaSi. formaluri meqanizmebi xSirad implika-

ciuri xasiaTisaa (mag., sityvaTa rigis cvla xSirad intonaciur gamokveTasac

iwvevs da garkveul SemTxvevebSi sintaqsuri mimarTebebis cvlasac; specialuri

nawilakebis gamoyeneba zogjer iwvevs winadadebis wevrTa Soris sintaqsur

mimarTebebis cvlas da specialur intonaciur gaformebasac moiTxovs; da

sxv.). topikisa da fokusis markirebis ZiriTadi universaluri meqanizmia spe-

cifikuri intonaciuri konturi, Tumca arsebobs enaTa mixedviT gansxvavebuli

sxvadasxvagvari formaluri modelebic. enebSi, romlebic orientirebulia

sainformacio nakadSi pragmatulad aqtualizebuli elementebis konceptualur

markirebaze, rogorc wesi, topikis aRmniSvneli specialuri morfologiuri

mawarmoebelic gamoiyofa (magaliTad, asea iaponurSi). miuxedavad imisa, rom es

procesebi enaTa mixedviT specifikuria (yoveli ena Tavisi strategiis

mixedviT axorcielebs topik-fokusisa Tu sxva aqtualizebuli wevris forma-

lizebas), universalur maxasiaTeblebsa da niSnebze saubari mainc SesaZle-

belia.

108

1.5. sakvlevi obieqtis dadgena; kvlevis miznebi da amocanebi

implikaciuri topiki, rogorc aRvniSneT, imiT aris gamorCeuli, rom is

damatebiT, damazustebel informacias gvawvdis saubris Temis Sesaxeb da,

imavdroulad, mas garkveulad upirispirebs, acalkevebs sxva nagulisxmebi

informaciisgan. specifikuri kontrastuli intonacia garkveuli Semadgenlisa

gulisxmobs kontrastul, sapirispiro informacias sxva obieqtze, rac nagu-

lisxmebia da gamomdinareobs an zogadi codnidan Cveni samyarosa da warmod-

genebis Sesaxeb (zogadi implikaciuri topiki) an, konkretuli situaciidan

gamomdinare, am konkretuli situaciuri codnidan (konteqsturad gansaz-

Rvruli implikaciuri topiki). magaliTad, winadadebaSi: [ai qalebi]Top ar ewevian, igulisxmeba Cveni codna da gamocdileba imis Sesaxeb, rom kacebi ewevian (zogadi implikaciuri topiki). xolo situaciidan: sanadirod wavidnen biWi da monadire. [ai monadirem]Top mokla iremi – gamomdinareobs, rom biWma ar mokla iremi (konteqsturi implikaciuri topiki). Cveni amocanaa, gavarkvioT:

1. gamoixateba Tu ara formalurad implikaciuri topiki qarTulSi;

2. Tu ki, konkretulad ra formaluri modelebi gamoiyeneba qarTulSi

implikaciuri topikis gamosaxatavad;

3. ra rols asrulebs am modelebSi intonacia;

4. ra saxis sintaqsuri konstruqciebia damaxasiaTebeli implikaciuri

topikisaTvis;

5. Tu gamoiyeneba (da romeli) nawilakebi implikaciuri topikisaTvis

qarTulSi. 1.6. meTodologia

ramdenadac topikisa da fokusis formaluri modelebis moZiebisaTvis gansa-

kuTrebiT mniSvnelovania bunebrivi sametyvelo situaciebi, saanalizo masalis

mosapoveblad SemuSavda swored aseT bunebriv situaciaTa mastimulirebeli

saeqsperimento amocanebi, romlebic eyrdnoba potsdam-berlinis proeqtis far-

glebSi SemuSavebul mravalferovan testebsa da kiTxvarebs. testebi Sed-

genilia situaciis ori ZiriTadi monawilis topikalizaciisTvis (agenti, mim-

Rebi): implikaciuri topikis realizeba mowmdeba mxolod ori monawilisTvis

(agenti, mimRebi). am amocanebis gadawyvetis pirobebSi cdispirTa mier sponta-

nurad warmoTqmul gamonaTqvamTa simravle qmnis saanalizo erTeulTa bazas.

2. masalis mopoveba

2.1. eqsperimentis aRwera

implikaciuri topikis ZiriTadi enobrivi modelebis gamosavlenad Catarda ori

tipis eqsperimenti.

pirveli amocana: cdispirebs eZleodaT ilustrirebuli baraTebi. maT unda

exelmZRvanelaT Semdegi instruqciiT: cdispirebs ise unda gaecaT pasuxi

dasmul SekiTxvaze, rom ar daesaxelebinaT is piri, romlis Sesaxebac gvsurda

109

mogvepovebina damatebiTi informacia. amdenad, topikalizebuli informacia

winadadebaSi fiqsirdeboda ara uSualod, aramed implikaciurad.

meore amocana: cdispirebs eZleoda erTi da igive, ornaxatiani, baraTebi. erT-

erTi cdispiris baraTze Semoxazuli iyo erT-erTi naxati. Semoxazulnaxatiani

baraTis mflobel informants ise unda gaeca pasuxi dasmul SekiTxvaze, rom

meore informanti, romlis baraTic ar Seicavda Semoxazul naxats, gamoecno,

romeli naxati iyo Semoxazuli.

eqsperimentSi monawileobas iRebda Teqvsmeti adamiani, ZiriTadad studentebi

– gogonebi da biWebi. muSaoba ZiriTadad wyvilebTan mimdinareobda. cda

tardeboda bunebrivi diskursis pirobebSi. cdispirebs ilustrirebuli fur-

clebi hqondaT darigebuli da maTi amocana iyo spontanurad, bunebrivad,

yovelgvari dafiqrebis gareSe gaecaT pasuxi dasmul SekiTxvaze an aReweraT

naxati. sul Catarda cameti sesia. 2.2. masalis Cawera

kiTxva-pasuxebis Cawera xorcieldeboda warmoTqmisTanave Praat programaSi. es programa saSualebas iZleva dafiqsirdes fonetikuri Canawerebi, romlis

speqtogramazec aisaxeba winadadebis mTliani intonaciuri konturi, tonuri

cvlilebebi, maxvili, intensivoba da xmovanTa formantebi.

2.3. masalis anotireba

mopovebuli masalis analizi-anotireba xdeboda programa Exmaralda-Si, sadac yovel winadadebas Seesabameboda sxvadasxva parametrebis Sesabamisad damuSa-

vebuli striqonebi:

1. saerTaSoriso, laTinuri anbanis safuZvelze miRebuli da

specialuri simboloebiT gamdidrebuli, transliteracia;

2. morfemebad daSlili mimdevrobebi;

3. morfemebis Sesabamisi glosebi;

4. inglisuri Targmani;

5. qarTuli anbanuri Canaweri;

6. fonetikuri Canaweri;

7. fonologiuri Canaweri;

8. metyvelebis nawilebis mixedviT anotirebuli Canaweri;

9. sintaqsuri funqciebis mixedviT anotirebuli Canaweri;

10. semantikuri rolebis mixedviT anotirebuli Canaweri;

11. Semadgenluri struqturis mixedviT anotirebuli Canaweri;

12. axali-Zveli informaciis mixedviT anotirebuli Canaweri;

13. topikis ganmsazRvreli Canaweri;

14. fokusis ganmsazRvreli Canaweri;

15. sulieroba-usulobis mixedviT anotirebuli Canaweri;

3.

pir

ins34P.1

ins

qal

34 P.3axu

ins

biW

mopovebuli

rveli amoc

struqtori-1 1: kaci ewev

struqcia: u

lze imgvara

kiTxva

pasuxi

-2

3: gogos dauravs qudi

struqcia: u

Wze imgvara

kiTxva

pasuxi

i masalis n

cana:

i aZlevs er

va

upasuxeT ki

ad, rom gas

a: vin eweva?

i: ai qali,

a ara biWs

upasuxeT ki

ad, rom gasa

a: vis axura

i: ai biWs, q

nimuSebi

rT baraTs o P.2:

iTxvas ise,

sagebi iyos,

?

ar eweva

iTxvas ise,

agebi iyos,

ravs qudi?

qudi ar axu

110

ori suraTi

qali svams

rom ar axs

, Tu ras ak

rom ar axs

Tu ras ake

xuravs

iT:

s

senoT kaci,

keTebs kaci.

senoT gogo

eTebs gogo

, ubralod

.

o, ubralod

.

miuTiTeT

d miuTiTeT

111

meore amocana: implikaciuri topiki (uaryofiTi implikacia) instruqtori

aZlevs erT baraTs ori suraTiT (romelzec ar aris Semoxazuli arc erTi

suraTi) erT-erT cdispirs, meores ki imave baraTs, sadac erT-erTi suraTi

Semoxazulia:

34-7

P.8a: qali (da ara kaci)

kiTxulobs wigns

P.8b: qali da Dkaci

kiTxuloben wigns

P.8a P.8b (Semoxazvis gareSe)

instruqcia pirvel informants:

Tqvens partniors aqvs igive suraTebi, rac Tqven. Tqvens wyvilze erTi suraTi

Semoxazulia. Tqvenma partniorma unda gamoicnos, romelia Semoxazuli da am

mizniT unda gkiTxoT, kiTxulobs Tu ara qali wigns Tqvens suraTze. Tqven upasuxebT rom kiTxulobs, rac cxadia orive suraTidan. Tqveni amocanaa, ise upasuxoT, rom Tqveni partniori mixvdes, kiTxulobs Tu ara meore piri wigns

da, Sesabamisad, is gamoicnobs, Tu romelia Semoxazuli suraTi.

instruqcia meore informants:

gTxovT hkiTxoT partniors, kiTxulobs Tu ara qali wigns Tqvens suraTze?

pasuxi: ki, mxolod qali kiTxulobs wigns.

112

meore amocana: implikaciuri (dadebiTi) topiki

34-8 P.9a: qali da kaci qudebis

gareSe P.9b: kaci qudis gareSe, qali

qudiT

= P.6a

= P.6b

=P.6a (Semoxazvis gareSe) =P.6b

instruqcia pirvel informants:

Tqvens partniors aqvs igive suraTebi, rac Tqven. Tqvens wyvilze erTi suraTi

Semoxazulia. Tqvenma partniorma unda gamoicnos, romelia Semoxazuli da am

mizniT unda gkiTxoT, axuravs Tu ara biWs qudi Tqvens suraTze. Tqven

upasuxebT, rom ar axuravs, rac cxadia orive suraTidan. Tqveni amocanaa, ise upasuxoT, rom Tqveni partniori mixvdes, axuravs Tu ara meore pirs qudi da,

Sesabamisad, is gamocnos, Tu romelia Semoxazuli suraTi.

instruqcia meore informants:

gTxovT hkiTxoT partniors, axuravs Tu ara biWs qudi Tqvens suraTze.

pasuxi: arc biWs ar axuravs qudi

4. masalis analizi

4.1. eqsperimentebis Sedegad dadasturebuli gamonaTqvamebi

eqsperimentis Sedegad dafiqsirda pasuxebi, sadac kargad Cans implikaciuri

topikis qarTuli enisaTvis damaxasiaTebeli formaluri modelebi (sul mopo-

vebuli masala moicavs 13 winadadebas). nimuSad warmovadgenT ramdenimes:

113

1. ai qali, ar eweva.

2. ai biWs, qudi ar axuravs

3. ara, biWi marto ar aris navSi

114

Catarebulma eqperimentma gviCvena, rom qarTul enaSi implikaciuri topiki,

upirveles yovlisa, gamoixateba specifikuri intonaciiT. aseve, gairkva, rom,

marTalia, ara gvaqvs specialuri morfologiuri markeri, magram nawilakebi,

romlebic mniSvnelovan rols TamaSoben implikaciuri xasiaTis informaciis

formirebaSi, garkveulwilad aseTi ‘markeris’ funqcias asruleben. mocemuli

cameti magaliTidan oTxjer dafiqsirda ai nawilaki, orjer dafiqsirda

mxolod nawilaki da TiTo-TiTojer dafiqsirda swored, arc da -c nawi-

lakebi. 4.2. nawilakebis analizi da maTi roli winadadebis sainformacio struqturis

formirebaSi

warmodgenili masalis mixedviT naTlad Cans nawilakebis gadamwyveti roli

implikaciuri topikis gamosaxatavad. masalaSi dadasturda ai, -c, mxolod, swored martivi da arc rTuli nawilakebi. am nawilakebis funqciebisa da

semantikis dasadgenad mogvyavs qarTul gramatikul tradiciaSi arsebuli

mosazrebebi:

ganmartebiT leqsikonSi ai nawilaki Semdegnairad aris daxasiaTebuli: 1.

miuTiTebs imaze, rac xdeba an rac unda moxdes:Aai ris Tqmac mindoda. ai erTi magaliTic. 2. igivea, rac hai:Aai, Se eSmako, Sena! ai gidi! – igivea, rac hai gidi.

115

sulxan-saba orbeliani ai-s aigivebs aa-sTan da ase gamartavs: “ixile aqa, Tu

viTar ganhyofs aobasa da myofsa. Mmyofi zogadi saxeli ars arsebisa da

SemTxveviTisa, rameTu arseba TaviT TvisiT myofi ars da ara sxvisa Soris

mqonebeli myofobisa da amisTvis qvemore damiweria myofsa Tana. xolo aÁ raRaTa ganeyofos arssa, ixile meWvretman: rameTu odes arsi anu myofi sTqva,

myis Tanave gvarsa warmouyeneb xedulisasa, raÁc xroa iyos gancdad xeduli

igi. xolo odes oden aÁ niSvnides, maT oden araobisagan ganhyofs da reca Tu a-Ta Soris warmoaCens, magram ugvarod da pirSeupyroblad, da merme uku-

anaÁsRa zeda daurTavs a-sa, viTarmed raÁao, da miugebT TanganvliT, viTarmed

cecxli, airi anu gonebaÁ da suli da eseviTari raÁve gvari myofTa.”

akaki SaniZe ki ai nawilaks Semdegnairad ganmartavs: ai misaTiTebeli

nawilakia: ai es kaci, rom geubnebodi, ai Seni wigni. SeiZleba SeuerTdes iq

da im sityvebs da miviRoT aiq, aim: aiq xidi aris, aim ficarzed ramdeni buzi

zis.

-c nawilaks Semdegi gansazRvreba aqvs: ca, c, c-c, c ki mimarTebiTi nawilakia;

gamaZlierebeli nawilakia.

-c nawilakis gameorebulad xmareba aucilebelia, Tu msjeloba ukuTqmiTia. am

SemTxvevaSi -c emateba ara winadadebis wevrs, aramed ukuTqmiT nawilaks, ris

gamoc miRebulia: arc-arc, aRarc-aRarc, verc-verc, nurc-nurc. swored ganimarteba Semdegnairad: 1. namdvilad, marTlac, zustad: swored ase

unda gakeTdes da ara sxvagvarad. biWi swored iq mivida, sadac elodnen. 2.

igivea, rac sworad: kiTxvaze sworad upasuxa.

mxolod nawilaki: 1. ara umetes imisa, rac vinc aris. swored is (isini),

swored im dros, im adgils da a.S., - marto: manqanaSi mxolod ori kaci ijda. mxolod axla dainaxa. jerjerobiT mxolod mercxlebi mofrindnen. 2. igivea, rac oRond: wadi, mxolod droze dabrundi! mxolod da mxolod =

martooden, mxolod. rogorc vxedavT, qarTul specialur literaturaSi TiTqmis araferia naTqvami

nawilakebis rolze winadadebis sainformacio struqturis formirebaSi. am

rolis gasaanalizeblad meti yuradReba unda mivaqcioT ara mxolod nawi-

lakebis semantikas, aramed maT funqciasac.

n. Sengelaia nawilakebs arasrul sityvebs uwodebs. ramdenadac arasruli sit-

yvebis forma aris martivi, imdenad rTulia maTi semantika da swored aseTi

arasruli sityvebis daxmarebiT SeiZleba gamoixatos sabolood molapa-

rakisTvis sasurveli informacia. arasrulAsityvaTa mniSvneloba araviTar

SemTxvevaSi ar warmoadgens raRac Semadgeneli mniSvnelobebis erTianobas.

116

isini warmoadgenen martiv formebs. maT calke ar gaaCniaT sakuTari mniS-

vneloba, isini mniSvnelobas iZenen winadadebaSi srulmniSvnelobian sityvebTan

erTad xmarebisas. arasruli sityvebi uzrunvelyofen logikurad azrobriv

mTlianobas, erTgvarovnebas da gabmulobas. arasruli sityvebis gareSe teqsti

ubralo mSrali winadadebebis mTlianobad iqceva.

am msjelobis dadasturebaa eqsperimentebiT mopovebuli masalac. metic, gair-

kva, rom nawilakebis roli teqstis mTlianobis formirebaSi ganisazRvreba

maTi, SeiZleba iTqvas, pirveladi funqciiT formalurad gamoxaton garkveuli

cvlilebebi winadadebis sainformacio struqturaSi.

5.Ddaskvnebi

5.1 nawilakebis roli informaciis struqturirebaSi

1. ai nawilaki cdispirebis mier gamoiyeneboda iseTi implikaciuri topikis

gamosaxatavad, romelic damatebiT, ZiriTadad sapirispiro, informacias

gvawvdida rogorc zogadi (34-1), aseve konkretuli situaciis meore wevrze:

34-2, 34-3, 43-4. implikaciuri topikis gamosaxatavad yvelaze xSirad swored

es nawilaki dafiqsirda. sazogadod, sainteresoa, ai nawilakis zogadi

funqcia: ‘miTiTeba+predikacia’, rac mas a. veJbickas mier navaraudebi erT-erTi

primitivis (demonstrativis, substantivisa da predikaciis elementebiT)

Sesabamis enobriv relizaciad warmogvidgens. swored am funqciebis gamo, is

bavSvTa metyvelebis adreuli formebis aucilebel Semadgenels warmoadgens,

rac, aseve, TanxmobaSia a. veJbickas mier winaswarmetyvelurad ‘aRdgenil’

(ramdenadac inglisurSi msgavsi, sami funqciis gamaerTianebeli, semantikuri

primitivis realizacia ar dasturdeba) primitivTan. 2. mxolod nawilaki sixSiriT meorea implikaciuri topikis markirebis

procesSi. is gulisxmobs, rom topikTa SesaZlebel variantTagan erTi da

mxolod erTia dasaSvebi situaciisda Sesabamisad da is mxolod aseT

konkretul SemTxvevebSia misaRebi: 34-6, 34-7.

3. EerTma cdispirma upiratesoba mianiWa marto nawilakis gamoyenebas mxolod nawilakis nacvlad. am nawilaks, rogorc Cans, igive funqciebi aqvs, rac

mxolod nawilaks: alternativebidan SearCios erTaderTi (34-9), Tumca

Tavisuflad monacvle leqsemebad maT mainc ver CavTvliT, ramdenadac arsebobs

konteqstebi, sadac isini erTmaneTs ver Caenacvlebian; mag.: marto kaci sikvdilSiac codoao. amis mizezi, rogorc Cans, maTi gansxvavebuli funqciebia:

mxolod kvantoruli zmnizedaa, marto nawilaks ki SeiZleba hqondes,

damatebiT, zedsarTavis funqciac, amitomac fraza marto kaci sintaqsuri

wyvilia (uSualo Semadgeneli winadadebisa; e.i. aq igi srulmniSvnelobiania),

maSin rodesac *mxolod kaci sintaqsur frazas ver qmnis (e.i. is nawilakad

rCeba). 4. aseve erT SemTxvevaSi dafiqsirda ar+c nawilaki: 34-8. am rTuli, nawar-

moebi nawilakis funqciaa SesaZlebel alternatiul topikTagan uaryofiTi

117

niSniT gamoarCios erT-erTi imgvarad, rom implikaciurad meoris igive

mdgomareoba daafiqsiros. es nawilaki informantma gamoiyena kiTxvaSi moce-

muli uaryofiTi mniSvnelobis dasadastureblad imgvarad, rom implikaciurad

meoris Sesaxebac msgavsi, uaryofiTi informacia mogvawoda.

5. swored nawilaki dafiqsirda im SemTxvevaSi, rodesac informanti impli-

kaciurad adasturebda kiTxvaSi mocemuli informaciis sapirispiro infor-

macias. saiteresoa, rom ‘uaryofiTi dadasturebis’ gasaZliereblad man

gamoiyena kompleqsuri fraza: swored rom (43-5).

6. -c nawilaki Segvxvda, aseve, informaciis (mxolod dadebiTis) dadastu-

rebisas (34-10). rogorc Cans, swored da -c nawilakebis funqcia aris

‘daTanxmeba, dadastureba’ situaciis meore wevris Sesaxeb damatebiT msgavsi

(dadebiTi an uaryofiTi) implikaciuri xasiaTis informaciis gadmocemis

mizniT.

7. aRsaniSnavia isic, rom nawilakebi, rogorc mosalodnelic iyo, ar

dadasturda martivi implikaciuri topikis SemTxvevebSi: 34-11, 34-12, 34-13.

5.2. intonaciis roli implikaciuri topikis formirebaSi

sazogadod, topikis ZiriTadi formaluri niSania specifikuri intonaciuri

konturi, romelic gansxvavdeba fokusirebuli an neitraluri frazis

intonaciisgan. misTvis damaxasiaTebelia aRmavali (magram ara imdenad, rogoric gvaqvs kiTxviT konstruqciebSi) intonacia frazis sasazRvro dabali intonaciiT (LH*Lp). aseTive intonaciuri konturi axasiaTebs implikaciur

topiks. nawilakebi, ZiriTadad, topikalizebul wevrTan erTad erT mTlian

fonologiur (intonaciur) frazas qmnian.

5.3. morfosintaqsuri modelebi

eqsperimentebma daadastura, rom qarTulSi implikaciuri topikisaTvis

specialuri morfologiuri markeri ar arsebobs, sintaqsuri Tavisebureba ki

aris is, rom topikalizebuli wevri miiswrafvis daikavos winadadebis sawyisi, ZiriTadad zmnis wina, pozicia. literatura asaTiani, r. 2006. intonaciis roli winadadebis sainformacio struqturis

formirebaSi; enaTmecnierebis sakiTxebi; Tbilisi: “universali”. 22-32.

orbeliani, s.s. 1991. leqsikoni qarTuli; Tbilisi: “merani”. 35-36.

SaniZe, a. 1973. qarTuli enis gramatikis safuZvlebi; Tbilisi: “Tbilisis

universitetis gamomcemloba”. 607-616.

Sengelaia, n. 2000. arasruli sityvebi da teqstis semantikuri mTlianoba; Tbilisi: “diogene”. 30-41

118

Ciqobava, a. WabaSvili, m. 1986. qarTuli enis ganmartebiTi leqsikoni; saqarTvelos ssr mecnierebaTa akademia, enaTmecnierebis instituti, qarTuli sabWoTa

enciklopedia. 19, 348, 451. Чейф, У. Л. 1975. Старая и новая информация. Значение и структура языка. Москва:

"Прогресс" . 241-267. Buring, D. 1995. Topic. Cologne University. Electronic Version. Gussenhoven, C. 2006. Notions and subnotions in information structure. Radboud Universiteit

Nejmegen and Queen Mary, University of London. Electronic Version. Krifka, M. 2006. Basic Notions of Information Srtucture. Humboldt Universitat zu Berlin and

Zentrum fur Allgemeine Sprachwissenshaft, Berlin. Electronic Version. Skopeteas, S. (in collaboration with Cornelia Endriss). Implicational Topic. Projekt D2:

Typologie der Informationsstruktur. Questionnaire on Information Structure. Universitat Potsdam/ Humboldt Universitat zu Berlin. Electronic Version.

Von der Gabelentz, G. 1869. Ideen zu einer vergleichenden Syntax; Zeitschrift fur Volkerpsychologie und Sprachwissenschaft. 6, 376-384.

Wierzbichka, A. 1996. Semantics (Primes and Universals). Oxford, New-York. Nutsa Tsereteli

The Information Structure of a Sentence:

Implicational Topic in Georgian

Abstract Structuring of information proceeds through the foregrounding of certain parts of the information. Any kind of ‘foregrounding’ (res. ‘Highlighting’, ‘Logical Emphasis’, ‘Promotion’, ‘Standing out as the first, important’ and etc) could be regarded as one, common phenomenon which represents the main strategy of structuring of linguistic structures. From this point of view Topic, Focus, Subject, Theme, Point of view and so on – are the same as far as they represent various forms of ‘foregrounding’. The goal of the work is to study implicational topic in Georgian. The data was experimentally accumulated on the basis of various experimental tasks that were carried out within natural discourse. The data is recorded in PRAAT Program and it is annotated in EXMARaLDA Program. On the basis of analyses of materials some conclusions were made and the various means of topicalization were suggested: Intonation patterns; Morpho-Syntactic devices and, especially, the role of Particles in the main models of implication topics in Georgian.

119

TinaTin janeliZe, rusudan asaTiani

mapirispirebeli kavSirebis Ffunqcionaluri analizi

qarTulSi


sazogadod, ‘obeqturi realobis’ lingvisturi struqturireba xorcieldeba

‘msgavsi da gansxvavebuli’ Sinaarsebis SepirispirebebiT, ris safuZvelzec

aigeba garkveuli formaluri opoziciebi. es opoziciebi warmogvidgenen, sul

cota, or erTeuls, romelTagan erT-erTi markirebulia, meore ki ara. aseTi

formaluri kontrasti, rogorc wesi, asaxavs informaciis struqturirebis

ganmsazRvrel ZiriTad strategias _ informaciis kontrastuli erTeulebis

gamovlenasa da maT kontrastul markirebas topikalizaciisa da/an fokusuri

Semadgenlis gamokveTis saSualebiT [1], [2], [5], [9, [10]. informaciulad gamokveTili kontrastuli erTeulebis SejerebisaTvis,

rogorc wesi, gamoiyeneba sxvadasxva saxis (ZiriTadad, mapirispirebeli) kav-

Sirebi. kavSirebis semantikuri da funqciuri gansxvavebulobebis dadgenisaTvis

aucilebelia gaviTvaliswinoT winadadebis sainformacio struqturis SefuT-

visa da informaciis struqturirebis gansxvavebuli strategiebi.

2. kavSiri rogorc metyvelebis nawili da misi specifika

kavSiri miuTiTebs im semantikur da logikur mimarTebebze, romelTa

saSualebiTac xorcieldeba teqstis Semadgeneli nawilebis erTmaneTTan

dakavSireba (resp. gadabma) [3], [8]. kavSiri arasodes icvlis formas. is

akavSirebs erTmaneTTan or an met winadadebas an winadadebis Semadgenel

wevrebs. kavSirebi uxvad arian warmodgenilni yvela saxis teqstSi, magram

TviTon ar arian damoukidebeli semantikuri Sinaarsis mqone denotaturi

erTeulebi. isini arafers aRniSnaven obieqtur realobaSi da mxolod im

siRrmiseul semantikur gadabmaze migvaniSneben, ris safuZvelzec gaerTiandnen

am teqstis nawilebi kontrastis, dapirispirebis an sxva specifikuri

Sinaarsebis gamosakveTad. 3. mapirispirebeli kavSirebi

dapirispirebuli Sinaarsis gamomxatveli informaciis Semadgeneli erTeulebis

gaerTianebisaTvis Rirebulia mapirispirebeli kavSirebi, romlebic ara marto

akavSireben winadadebebs an struqturulad erTgvarovan wevrebs erTmaneTTan,

aramed axal semantikasac aniWeben mTlian gamonaTqvams. maT SemoaqvT axali

mniSvneloba – kontrasti. qarTulSi am funqciiT gamoiyeneba Semdegi mapiris-

pirebeli kavSirebi: magram, Tumca, aramed, oRond, xolo, ki; da, aseve,

majgufebeli kavSirebi: Tu, da.

120

winadadebaSi kontrastuli informaciis warmodgenisaTvis SesaZlebelia saer-

Tod ar iyos gamoyenebuli mapirispirebeli kavSiri da Sinaarsobrivi

dapirispireba formalurad mxolod Sesabamisi intonaciiT gadmoices. mapiris-

pirebeli kavSirebi formalurad aZliereben kontrastul mimarTebebs winada-

debaSi da axdenen winadadebis aqtualizebuli wevris xazgasmas, gamoyofas.

kavSirebiT gaerTianebisas Sinaarsobrivi dapirispireba formaluradac aris

specialuri sityviT (resp. kavSiriT) gamoxatuli; magaliTad:

(1) erTi xmaurobda, meore wynarad ijda. (2) erTi xmaurobda, meore ki wynarad ijda.

pirvel winadadebaSi, ukavSiro SeerTebisas, Sinaarsobrivi dapirispireba

xmaurobda: wynarad ijda, mxolod kontrastuli intonaciis saSualebiT

miiRweva, meoreSi ki am kontrasts formalurad aZlierebs kavSiri ki, romelic intonaciasTan erTad migvaniSnebs semantikurad Sepirispirebul

movlenebze. garda Sinaarsobrivad dapirispirebuli gamonaTqvamebisa, SesaZlebelia, aseve,

iseTi gamonaTqvamebis kontrastuli gaerTianeba, romlebic Sinaarsobrivad ar

gulisxmoben ucilobel kontrasts; magaliTad:

(3) kaci werda, qali kiTxulobda. (4) kaci werda, qali ki kiTxulobda. am SemTxvevebSi, (3) da (4) winadadebebis Sinaarsobrivad aradapirispirebuli

Semadgenlebi _ werda:kiTxulobda _ kontrastul opozicias qmnian mxolod

da mxolod Sesabamisi intonaciis (3) an intonaciasTan erTad formalurad

warmodgenili ki mapirispirebeli kavSiris (4) saSualebiT. radganac yvela mapirispirebel kavSirs erTi saerTo, zogadi funqcia aqvs _

winadadebaSi gamoxatos kontrasti/dapirispireba, xSir SemTxvevaSi SesaZle-

belia maTi erTmaneTiT Canacvleba:

(5) megobarTan mivedi, magram saxlSi ar damxvda. (6) megobarTan mivedi, oRond saxlSi ar damxvda. (7) megobarTan mivedi, Tumca saxlSi ar damxvda. (8) megobarTan mivedi, xolo is saxlSi ar damxvda. (9) megobarTan mivedi, saxlSi ki ar damxvda. (10) megobarTan mivedi da is saxlSi ar damxvda. zemoT warmodgenil magaliTebSi dapirispireba gamoxatulia sxvadasxva kav-

SiriT. rogorc vxedavT maT ZiriTadi Sinaarsi erTnairi aqvT, gansxvavebulia

mxolod ‘kontrastis’ saxe. winadadebebi dalagebulia kontrastis/dapirispi-

rebis xasiaTisa da xarisxis cvlis mixedviT _ (10)-dan (5)-mde kontrasti

sul ufro da ufro Zlieria, ufro metad aris xazgasmuli winadadebebs

Soris kontrastuli dapirispireba; SeiZleba iTqvas, rom es kavSirebi

kontrasts sxvadasxva xarisxiT gamoxataven.

121

4. mapirispirebeli kavSirebis funqciur-semantikuri sxvaobis amsaxveli sqema

marTalia, zogierT konteqstSi mapirispirebeli kavSirebis erTmaneTiT Cana-

cvleba SesaZlebelia, magram es kavSirebi sinonimurni ar arian _ maT TavianTi

funqcia da datvirTva aqvT da gamoirCevian mxolod maTTvis damaxasiaTebeli

niSan-TvisebebiTa da TaviseburebebiT. am gansxvavebulobebis gamosavlenad

gavaanalizeT sxvadasxvagvari teqstebi da zepirmetyvelebis nimuSebi, gaviTva-

liswineT qarTul gramatikul tradiciaSi dafiqsirebuli Tvalsazrisebi [3],

[4], [6], [8] da Cveni dakvirvebis Sedegebi SevajameT garkveuli sqemis saxiT,

romelic asaxavs mapirispirebeli kavSirebis funqciur da semantikuri

gansxvavebulobebs:

kontrasti

+dapirispireba _dapirispireba

+sawinaaRmdego -sawinaaRmdego +erTi mainc _erTi mainc

WeSmariti (=yvela)

WeSmariti

+koreqcia _koreqcia +Zlieri -Zlieri

+molodini -molodini

+survili -survili

aramed/oRond/Tumca/magram/ki//kidev / xolo / Tu / da

122

5. mapirispirebeli kavSirebis funqciaTa ganmasxvavebeli diferencialuri

niSnebi

rogorc sqemidan Cans, mapirispirebel kavSirebis dasaxasiaTeblad gamoiyo 7

binaruli diferencialuri niSani, romlebic Rirebulia am kavSirebs Soris

gansxvavebulobebis aRsawerad. ganvixiloT es niSnebi:

I. +/_ [dapirispireba]

yvela mapirispirebeli kavSiris saerTo funqcia aris is, rom isini gamoxataven

kontrasts, magram kontrastis gamoxatva aucileblobiT ar gulisxmobs

garkveul erTeulebs Soris dapirispirebis gaxazvas. enobrivad kontrastuli

SeiZleba iyos nebismieri ori (an meti) erTgvarovani Semadgeneli, romelTa

Soris Sinaarsobrivi dapirispirebis arseboba aucilebeli arc aris (SesaZ-

lebelia gamoricxulic ki iyos, magaliTad, sinonimebis SemTxvevaSi). kavSirebi

da da Tu SesaZlebelia aerTianebdnen kontrastuli xasiaTis Semadgenlebs,

Tumca isini am Semadgenlebis dapirispirebas ar gamoxataven; magaliTad:

(11) is xan ucnaurad tokavda da xan saocrad SeifarTxalebda xolme (ak. wer.). (12) saqme iqamde mivida, rom dRes Tu xval skolidan unda gamoericxaT (i. gogeb.). (13) dainaxes Tu ara es mercxlebma, swrafad daanebes Tavi da afrindnen maRla (i. gogeb.).

maSasadame, da da Tu majgufebeli kavSirebi SeiZleba davaxasiaToT +/–

[dapirispireba] binaruli diferencialuri niSnis uaryofiTi SinaarsiT

(_[dapirispireba]), xolo mapirispirebeli kavSirebi: magram, Tumca, aramed, oRond, xolo, ki _ amave diferencialuri niSnis dadebiTi SinaarsiT

(+[dapirispireba])

II. +/_[erTi mainc]

miuxedavad imisa, rom da da Tu kavSirebi, orive, aradapirispirebuli

kontrastuli gaerTianebis gamosaxatavad SeiZleba iyos gamoyenebuli, maT

Soris funqciuri sxvaoba mainc didia:

da majgufebeli kavSiria da misi umTavresi funqcia dakavSirebaa _ mas

SeuZlia nebismieri saxis erTgvari wevrebisa Tu winadadebebis gaerTianeba.

Sedegad miRebuli gamonaTqvami WeSmariti iqneba mxolod da mxolod im

SemTxvevaSi, rodesac gaerTianebul SemadgenelTagan yoveli iqneba WeSmariti.

Tu kavSiric aerTianebs erTgvar wevrebsa Tu winadadebebs, magram iseT

SemTxvevaSi, rodesac gaerTianebul SemadgenelTagan erT-erTi mainc aris

WeSmariti. ase rom, SeiZleba iTqvas, rom da kavSiri Seesabameba formaluri

logikis koniunqcias, xolo Tu _ diziunqcias. Tu xSiria SekiTxvebSi, sadac fokusTa SesaZlebeli variantebis raodenoba

SemosazRvrulia; magaliTad:

123

(14) A: nika kiTxulobs Tu wers? P SesaZlebel amomwurav pasuxTagan yoveli varianti iqneba SemosazRvruli,

dakonkretebuli:

(14) BB-1: nika kiTxulobs! BB-2: nika wers! A BB-3: nika kidevac kiTxulobs da kidevac wers! M

magram Tu davsvamT kiTxvas:

(15) AA: ras akeTebs nika?

SesaZlebel pasuxTa variantebi faqtiurad SemousazRvreli iqneba: (15) BB: nika kiTxulobs (wers, xatavs, TamaSobs...).…

am funqciiT Tu kavSiri seleqciur, kontrastul fokuszea orientirebuli

maSin, rodesac da kavSiriT gaerTianebul Semadgenlebian kiTxvaze pasuxi

moiTxovs mTlianad azris dadasturebas an uaryofas:

(16) AA: nika wers da kiTxulobs? P BB-1: ki.

pasuxi ki gulisxmobs, rom nika orive moqmedebas asrulebs: kidec wers da

kidec kiTxulobs; xolo pasuxi ara gvatyobinebs, rom nika an arc erT

moqmedebas ar asrulebs (e.i arc wers da arc kiTxulobs), an erT-erTs mainc

asrulebs, amotom amomwuravi pasuxi am kiTxvaze unda iyos dazustebeli:

(16) B-2: ara, mxolod kiTxulobs. B-3: ara, mxolod wers. B-4: ara, arc kiTxulobs da arc wers.

rodesac Tu kavSiriT gaerTianebulia ori (an meti) winadadeba, maSin saqme

gvaqvs daqvemdebarebul konstruqciasTan, sadac azris dasrulebisaTvis

aucileblad unda iyos warmodgenili kidev erTi (mTavari) winadadeba, romlis

daqvemdebarebulic iqneba Tu kavSiriT gaerTianebuli, TavisTavad rTuli,

winadadeba:

(17) nino cekvavda Tu nika mReroda, maris es ukve aRar ainteresebda.

aseTi rTuli qvewyobili konstruqcia ar aris aucilebeli da kavSiriT

gaerTianebuli winadadebebis SemTxvevaSi _ azri dasrulebulia ‘mesame’

winadadebis gareSec:

(18) nino cekvavda da nika mReroda.A aqve SevecdebiT Tu kavSiris funqciur gamijvna-gansxvavebis warmoCenas an - an macalkevebeli kavSirisagan: Cveulebriv, Tu-s dros gaerTianebul

SemadgenelTagan orive an erT-erTi mainc WeSmaritia; magaliTad:

(19) didi Tu patara TavgamodebiT muSaobda (vaJa.). A

124

aq igulisxmeba, rom yoveli _ didi da patara _ damoukideblad muSaobda

TavgamodebiT (zmna am SemTxvevaSi mxolobiTSia). gansxvavebiT da kavSiriT

gaerTianebisas igulisxmeba, rom yvela erTad _ didi da patara _

TavgamodebiT muSaobda (zmna am SemTxvevaSi mravlobiTSia). gansxvavebuli

viTarebaa (an) _ an kavSiriT gaerTianebul SemadgenlebTan:

(20) an dRes mogvardeba yvelaferi, an xval. aq igulisxmeba, rom mxolod erT SemTxvevaSi (an dRes, an xval) mogvardeba

yvelaferi. (Sdr: dRes Tu xval yvelaferi mogvardeba _ niSnavs, rom nebismier dRes (an dRes, an xval) SeiZleba mogvardes yvelaferi.)

amdenad, SeiZleba iTqvas, rom maSin, rodesac Tu kavSiri Seesabameba logikur

diziunqcias (resp. +[erTi mainc]), kavSiri an _ an fardia logikis e. w.

gamomricxavi diziunqciisa (resp. +[mxolod erTi]); xolo orive upirispirdeba

da kavSirs, romelic logikis koniuqciis Sesabamisia (resp. _[erTi mainc]; anu +[yvela]).

III. +/_[sawinaaRmdego]

kontrastul dapirispirebas SesaZlebelia qmnidnen rogorc sawinaaRmdego

Sinaarsis (magaliTad, antonimebi; leqsemebi, romlebic uSveben semantikurad

sapirispiro Sinaarss; formalurad uaryofiTi da dadebiTi konstruqciebis

SepirispirebiT miRebuli gamonaTqvamebi; da sxv.), aseve, arasawinaaRmdego

Sinaarsis gamomxatveli gansxvavebuli Semadgenlebi. Tu dapirispireba

sawinaaRmdego Semadgenlebs aerTianebs, maSin mizanSewonilia oRond, Tumca, aramed, magram kavSirebis gamoyeneba; xolo Tu Sepirispirebulia gansxvavebuli,

magram arasawinaaRmdego Sinaarsis mqone Semadgenlebi, ZiriTadad gamoiyeneba

ki//kidev, xolo kavSirebi. e.i. oRond, Tumca, aramed, magram kavSirebi SeiZleba

daxasiaTdes, rogorc kavSirebi, romlebic aerTianeben +[sawinaaRmdego]

Sinaarsebs, xolo ki//kidev, xolo kavSirebi, rogorc kavSirebi, romlebic

aerTianeben –[sawinaaRmdego] Sinaarsebs. aRsaniSnavia, rom, ramdenadac am kavSi-

rebis funqciaa ise gaaerTianos gamonaTqvamebi, rom gaxazos (an ar gaxazos)

maT Soris sawinaaRmdego (an arasawinaarmdego) Sinaarsebi, isini SeiZleba

gamoyenebul iqnen am funqciiT im SemTxvevaSic, rodesac dapirispirebul erTe-

ulTa semantika aucileblobiT ar aris SesabamisobaSi gaerTianebul Semadge-

nelTa sawinaaRmdego an arasawinaaRmdego SinaarsebTan; anu +[sawinaaRmdego]

kavSirebiT SeiZleba gaerTiandes, aseve, semantikurad arasawinaaRmdego Sinaar-

sis Semadgenlebi da, piriqiT, –[sawinaaRmdego] kavSirebiT _ arasawinaaRmdego

Sinaarsebi, Tuki konkretuli situacia gulisxmobs aseT dapirispirebas.

magaliTebi:

M(21) mas brZola undoda, Tumca am yvelafriT Zalian daRlili iyo (i. gogeb.). (22) jer ar menaxa misi lekuri, Tumca gagonebiT ki bevri gamegona (ak. wer.). (23) Cven, irmebi, eniT ar vlaparakobT, aramed TvalebiT (vaJa.).

125

(24) Zmani xerxeuliZeni aramc Tu ar Sedrknen, aramed lomebad gadaiqcnen (i. gog.). (25) niniam daaxala daTvs Tofi, magram aacdina (vaJa.). (26) wvimam gadaiRo, magram caze mze mainc ar Canda (vaJa.). (27) ivarjiSe, oRond Zalian nu gadaiRlebi. (28) imRereT, oRond nu abRavldebiT (m. jav). es (21) _ (28) magaliTebi upirispirdeba (29) _ (32) magaliTebs, romlebic

arasawinaaRmdego SinaarsTa kontrastul dapirispirebas gamoxataven:

A(29) Zalian moiwyina, tirils ki ar apirebda (ak. wer.). (30) zamTari iyo, bavSvs ki qudi ar exura (ak. wer.). (31) erTi sacekvaod wavida, xolo meorem dasveneba gadawyvita. A (ak.

wer.)

(32) nadiris xorciT ZaRlebs asuqebdnen, xolo tyavs amarilebdnen da axmobdnen (m. jav).

amdenad, diferencialuri niSani +/_[sawinaaRmdego] asaxavs kavSirebis

funqciur sxvaobas _ gaaerTianon da daupirispiron Semadgenlebi imgvarad,

rom warmoaCinon am dapirispirebis sawinaaRmdego an arasawinaaRmdego xasiaTi. IV. +/_[dapirispirebul SemadgenelTa Zlieri gamokveTa]

arasawinaaRmdego, magram dapirispirebuli Semadgenlebis gamaerTianebeli

kavSirebi ki/kidev, xolo funqciurad msgavsi kavSirebia. isini xSirad

enacvlebian erTmaneTs da gamoxataven SedarebiT ‘rbil’ dapirispirebas, Tumca

es monacvleobac Tavisufali ar aris, ramdenadac maTi poziciebi Semosaz-

Rvrulia: xolo iwyebs dapirispirebis amsaxvel meore Semadgenlur konst-

ruqcias, ki ki mosdevs meore winadadebaSi warmodgenil dapirispirebul

Semadgenels. xolo kavSirisgan gansxvavebiT, ki SeiZleba erTmaneTs upirispi-

rebdes ara mxolod winadadebebs, aramed martiv Semadgenlebsac. aqedan gamom-

dinare, ki-s ufro metad SeuZlia gamokveTos dapirispirebis konkretuli

obieqti _ topiki an fokusi.

(33) zamTari iyo. yvelas qudi exura, sabralo bavSvs (T) ki ara. (34) AA: vin wers? B: qali(F) wers, kaci(F) ki ara. AA: ras wers? BB: werils(F) wers, wigns(F) ki ara. amdenad, xolo/ki//kidev kavSirebis funqciebis gansasxvaveblad Rirebulad

SeiZleba miviCnioT Semdegi diferencialuri niSani: +/_[dapirispirebul

SemadgenelTa Zlieri gamokveTa]. es niSani asaxavs im faqts, rom ki//kidev kavSiri ufro Zlierad gamoxatavs SemadgenelTa topikur/fokusur bunebas,

xolo kavSiri ki mTlianad winadadebebis dapirispirebas gulisxmobs.

rac Seexeba sawinaaRmdego dapirispirebis gamomxatvel kavSirebs:

magram, Tumca, oRond, aramed _ maTi urTierTCanacvlebadobis xarisxic

maRalia, Tumca urTierTgamomricxavi gamoyenebis SemTxvevebis analizi

126

saSualebas gvaZlevs maTi funqciebis ganmasxvavebeli ZiriTadi diferencialuri

niSnebsac iqnes mikvleuli.

V. +/_[koreqcia]

komunikaciis dros xSirad xdeba garkveul gamonaTqvamTa koreqcia, Sesworeba:

adre gamoTqmuli mosazrebis uaryofa da axlis mtkiceba. aseTi ori

Semadgenlis dakavSirebisas, bunebrivia, saxezea sawinaaRmdego mtkicebis dapi-

rispirebul SemadgenelTa gaerTianeba. amgvar SemTxvevaSi yvelaze mizanSewo-

nilia aramed kavSiris gamoyeneba: aramed amyarebs kavSirs movlenebs Soris

imgvarad, rom uaryofs mocemul dadebiT mtkicebas da mas sawinaaRmdego

mtkicebas upirispirebs; magaliTad:

(35) AA: rogorc Cans, magrad wvims. BB: ki ar wvims, aramed mxolod quxs. aramed kavSiris am funqciidan gamomdinare, formalurad mkacrad

gansazRvrulia amgvari koreqtivis amsaxveli gamonaTqvamis formaluri

struqturac: rogorc wesi, aramed kavSiriT gaerTianebul Semadgenlebidan

pirvelia uaryofiTi formiT warmodgenili, radganac is asaxavs adre

mocemuli msjelobis sapirispiro mtkicebas, meore ki ukve koreqtirebul

mtkicebas gamoxatavs. amis gamo, pirvel Semadgenel winadadebasTan, rogorc

wesi, vxvdebiT uaryofiT nawilakebs: ar, ki ar, nu, nu ki, ara Tu da, aseve, kiTxviT nawilaks gana, romelic gamoxatavs gakvirvebas, eWvis Setanas da,

amdenad, semantikurad uaryofis Sinaarsi aqvs (es nawilaki, rogorc wesi,

ixmareba kiTxviT winadadebebSi, rodesac kiTxvis avtori uaryofiT pasuxs

moelis). magaliTebi:

K(36) ki ar gavaleb, aramed gTxov. (38) ki nu gabrazdebi, aramed dafiqrdi. (39) mefe erekle am dros iyo ara uwindeli Zal-RoniT savse arwivi, aramed oTxmoci wlis moxuci (i. gogeb.). G(40) gana marto codna niSnavs bevr rames, aramed saqmec unda SegveZlos (ak. wer.).

maSasadame, kavSiri aramed sawinaaRmdego dapirispirebis gamomxatvel danarCen

kavSirebs: oRond, Tumca, magram _ upirispirdeba +/_[koreqcia]

diferencialuri niSnis dadebiTi maCvenebliT. VI. +/_[molodini]

sazogadod, movlenaTa Sepirispirebis dros SesaZlebelia dafiqsirdes

mosaubris molodinis sawinaaRmdego movlenebi. am SemTxvevaSi gamoiyeneba

kavSirebi: oRond, Tumca. iseTi gaerTianeba, sadac gaxazulia sawinaaRmdegoTa

dapirispireba, Tumca mosaubris molodinze yuradReba gamaxvilebuli ar aris,

rogorc wesi, magram kavSiriT xorcieldeba. magram kavSiri yvelaze Zlierad

gamoxatavs sawinaaRmdego movlenebis dapirispirebul gaerTianebas, Tumca aq

127

movlenebs Soris urTierTkavSiris da, Sesabamisad, erTis gamo meoris

moulodnelobis gamokveTa ar xdeba:

(41) wvims, magram ar quxs. sapirispirod, oRond da Tumca kavSirebi aseTi molodinis SesaZleblobas

gaxazaven: (42) wvims, Tumca ar quxs. (43) wvims, oRond ar quxs.

am SemTxvevaSi, ramdenadac wvimis dros xSiria quxilic, mosaubre imgvarad

aRwers situacias, rom afiqsirebs am zogadi molodinis sawinaaRmdego

viTarebas. SesaZlebelia mosaubrem airCios gansxvavebuli strategia da

yuradReba ar gaamaxvilos aRniSnul molodinze, maSin gveqneboda

gansxvavebuli winadadeba:

(44) wvims, magram ar quxs.

vnaxoT am saxis sxva magaliTebic:

(45) meama, magram cud gunebaze avdeqi mainc (ak. wer.). (46) xoxbebi yurs ugdebdnen, magram ar elaparekebodnen (vaJa.).

Tu ar CavTvliT am individualur maxasiaTeblebs, kavSirebi magram, Tumca, oRond SesaZlebelia Tavisuflad Caenacvlon erTmaneTs. siZlieris mixedviT

magram kavSiri xolo/ki kavSirebTan SedarebiT yvelaze metad gamoxatavs

dapirispirebas da zogjer mas kidev ufro aZlierebs nawilaki mainc: (47) wvimam gadaiRo, magram caze mze mainc ar Canda (vaJa.). VII. +/_[survili]

miuxedavad imisa, rom Tumca da oRond funqciurad Zalian hgvanan erTmaneTs,

mainc moxda am kavSirebSi funqciuri sxvaobis gamovlena. isini TiTqmis

yovelTvis enacvlebian erTmaneTs, magram aris konstruqciebi, sadac oRond kavSiri molodinTan erTad gamoxatavs mosaubris Zlier survilsa an SiSs

imisa, rom misi molodini SeiZleba ar gamarTldes.

M (48) modi, oRond ar daigviano! M (49) modi, Tumca ar daigviano! oRond kavSiris Tumca-Ti Canacvlebam (49) winadadebaSi mogvca gansxvavebuli

Sinaarsi: mosaubre afrTxilebs msmenels, ar daigvianos, maSin rodesac (48)

winadadebaSi gamoxatulia mosaubris Zlier survili da/an SiSi imisa, rom

saubris adresati SeiZleba dagvianebiT movides. vnaxoT am saxis sxva

magaliTebic:

(50) Sen oRond imecadine, am saqmes me mivxedav. (51) ivarjiSe, oRond Zalian nu gadaiRlebi. (52) SegiZlia gaemgzavro, Tumca ara mgonia, gaqceviT saqmes uSvelo.

128

(53) sirbili Zalian damRlelia, Tumca janmrTelobisTvis sasargebloa. oRond kavSirs dapirispirebasTan erTad aqvs, aseve, daTmobis funqcia da

SedarebiT ‘rbil’ molodinsa Tu gafrTxilebas gamoxatavs:

(54) iTamaSeT, oRond Zalian nu ixmaurebT! (55) wadi saTevzaod, oRond Rrma wyalSi ar Sexvide!

sazogadod, brZanebiTi konstruqciebi survilis semantikiT ufro

oRondkavSiriani formebisTvis aris bunebrivi (oRond imRere!), Tumca kavSiriani formebisTvis ki damaxasiaTebeli ar aris (*Tumca, imRere!). Tumca kavSiriani brZanebiTebi ufro gafrTxilebas gamoxataven, vidre survils:

Tumca, imRere ra, me ra menarvleba, Sen Sercxvebi!.

aseTi specifikuri konstruqciebi, sadac mosaubris sasurveli molodinia asa-

xuli, gvaZlevs biZgs am kavSirebis gansasxvaveblad gamovyoT diferencialuri

niSani +/_[survili], romlis mixedviTac oRond daxasiaTdeba, rogorc

+[survili]-s gamomxatveli, Tumca ki _ gafrTxilebis, -[survilis]

gamomxatveli.

6. daskvnebi

amgvarad, warmodgenili 7 binaruli diferencialuri niSani sakmarisi aRmoCnda

mapirispirebel kavSirTa semantikuri da/an funqciuri gansxvavebulobebis

aRsawerad. diferencialuri niSnebis mixedviT kavSirTa funqciebis

ganmasxvavebeli xe-struqtura (ix. gv. 5); SeiZleba Semdegi cxrilis saxiTac

warmovadginoT:

(*cxrilSi ‘0’ miuTiTebs, rom aRniSnuli niSani mocemuli kavSirisaTvis Warbia.)

aRsaniSnavia isic, rom mapirispirebeli kavSirebis funqciuri sxvaobebis

amsaxveli xe-struqtura imgvarad aris warmodgenili, rom asaxavs am

aramed oRond Tumca magram ki xolo Tu da

dapirispireba + + + + + + _ _

erTi mainc 0 0 0 0 0 0 + _

sawinaaRmdego + + + + _ _ 0 0

Zlieri

gamokveTa

0 0 0 0 + _ 0 0

koreqcia + _ _ _ 0 0 0 0

molodini 0 + _ _ 0 0 0 0

survili 0 + _ 0 0 0 0 0

129

kavSirebis mier dapirispirebulobis gadmocemis mixedviT mzard (marjvnidan

marcxniv) ierarqiul mimarTebebs:

aramed > oRond > Tumca > magram > ki//kidev > xolo > Tu > da

sazogadod, warmodgenil semantikur-funqciur diferencialur niSnebs ufro

sarekomendacio xasiaTi aqvT, vidre mkacrad gramatikuli mimarTebebis aRweris

pretenzia. aucilebelia SevniSnoT, rom zemoT mocemul sqemaSi minus niSnis

qveS warmodgenilma kavSirma, rogorc aramarkirebulma, SeiZleba zogierT

SemTxvevaSi sapirispiro, plius niSniT nagulisxmebi Sinaarsic gamoxatos.

magaliTad, da kavSiriT SeiZleba gavaerTianoT sapirispiro movlenebi, magram kavSirma SeiZleba molodinic da zogjer survilic gamoxatos; magram piriqiT,

plius niSnis qveS moTavsebulma kavSirma rom gamoxatos minusianis semantika,

Zneli dasaSvebia. am dros kavSiris arasworad gamoyenebasTan gveqneba saqme.

sqemis sarekomendacio xasiaTi Semdegnairad SeiZleba CamovayaliboT: Tu

movlenaTa gaerTianebisas gvsurs pliusiT gamoxatuli semantikis dafiqsireba,

umjobesia, Sesabamisi, mis qvemoT moTavsebuli, kavSiri gamoviyenoT; xolo, Tu

gvsurs mxolod dapirispireba/kontrasti davafiqsiroT, maSin mis qvemoT

moTavsebuli yvela kavSiris gamoyeneba Tavisuflad dasaSvebia. sazogadod,

informaciis struqturireba da movlenaTa amsaxveli gamonaTqvamebis gaerTi-

anebis strategiis SerCeva, situaciis garda, mniSvnelovanwilad mosaubris

mizansa da survilzea damokidebuli. kavSirebis polifunqciuroba amis saSu-

alebas iZleva. kavSirebis funqciebis garkveva da maTi sworad gamoyeneba

saTqmelis ufro mokled da naTlad gamoxatvis SesaZleblobas iZleva.

literatura 1. asaTiani, r., 2007. intonaciis roli winadadebis sainformacio

struqturis formorebaSi, enaTmecnierebis sakiTxebi, Tbilisi: “Tsu

gamocemloba”. 22-33.

2. asaTiani, r., 2007. informaciis struqturirebis sintaqsuri modelebi

qarTulSi. “semiotika-II“, WavWavaZis sax. universiteti. Tbilisi:

“universali”. 7-29.

3. gamyreliZe, T., kiknaZe, z., Saduri, i., Sengelaia, n., 2003. Teoriuli

enaTmecnierebis kursi, Tbilisi; “Tsu gamomcemloba”.

4. kvaWaZe, l., 1966. qarTuli enis sintaqsi, Tbilisi: “ganaTleba”.

5. krifka, m., 2008. sainformacio struqturis ZiriTadi cnebebi; ena, logika, kompiuterizacia, II t. Tbilisi: “degaprinti”. 7-40

6. SaniZe, ak., 1980. qarTuli enis gramatikis safuZvlebi, III t. Tbilisi:

“Tsu gamomcemloba”.

7. Sengelaia, n., 2000. arasruli sityvebi da teqstis semantikuri

mTlianoba, Tbilisi: “diogene”.

8. ZiZiguri, S., 1973. kavSirebi qarTul enaSi, Tbilisi: “Tsu

gamomcemloba”. 9. Чейф, У. Л. 1975. Старая и новая информация. Значение и структура

языка. Москва: "Прогресс" . 241-267.

130

10. Asatiani, R., The Main Devices of Foregrounding in the Information Structure of Georgian, The 6th International Symposium on LLC, Tbilisi-2005. Balden D.ten Cate&Henk W. Zeevat (Eds). Springer. 21-31. 2007.

Tinatin Janelidze, Rusudan Asatiani

Functional Analysis of Conjunctions Showing an Informational Contrast in Georgian

Abstract

Linguistic structuring of reality based on the notions ‘same-different’ proceeds through ‘oppositions’. An opposition means that there are at least two items one of which is ‘marked’ and another is ‘unmarked’. Structuring of information, its packaging, also proceeds through the foregrounding of such contrastive constituents: one part of information stands out against a background of the other part of information. As a result different formal models of various types of topics and focuses arise. The paper examines the role of conjunctions in the process of structuring of contrastive topics and/or focuses in Georgian. On the basis of functional-semantic analysis of conjunctions showing contrast seven binary differential semantic features are suggested: +/-[Contrast], +/-[Opposition], +/-[Correction], +/- [Expectation], +/-[Wish], +/-[At least one], +/-[Strong underlining]. The various combinations of the features express main functions of the conjunctions, which are relevant for the formal representations of contrastive topics and/or focuses.

131

marika jiqia erTi tipis qvewyobili winadadebis Sesaxeb megrulSi

qarTul, megrul-Wanursa da svanur cocxal metyvelebaSi dadasturebuli

qvewyoba saintereso sakvlevia (a) saliteraturo enisa da cocxali metyve-

lebis urTierTmimarTebis, (b) qvewyobiT sistemebTan Sedarebisa da maT Soris

msgavseba-gansxvavebaTa gamovlenis, (g) kavSirisa Tu korelatis gamoyenebis, (d)

mTavar winadadebaSi mocemuli asaxsneli wevrisa Tu (e) damokidebul winada-

debaTa klasificirebis TvalsazrisiT.

megrulSi dadasturebuli yvela qvewyobili konstruqcia amomwuravad aris

gaanalizebuli maia lomias monografiaSi `hipotaqsis sakiTxebi megrulSi~

(lomia 2006). radgan Cveni dakvirvebis obieqtic megruli qvewyobili winada-

debis erTi saxeobaa, saWirod CavTvaleT mokled gadmogveca am naSromis

ZiriTadi debulebebi.

megrulis hipotaqsur konstruqciebSi monawileoben korelatebi, mimarTebiTi

nacvalsaxelebi, mimarTebiTi zmnizedebi da maqvemdebarebeli kavSirebi.

mimarTebiTi nacvalsaxelebi: namuT ‘romelic’, miT ‘vinc’, muT ‘rac’, muzmaT

‘ramdenadac’ muferi(T) ‘ranairic’, mudaneri ‘ramdennairic’ _ qvewyobil winada-

debebSi gvevlineba wevr-kavSirebad.

asevea mimarTebiTi zmnizedebic: soTi ‘sadac’, sodeT ‘sadac’, sovre ‘saiTkenac’,

soiSax ‘sanam’, muJansGT ‘roca’, muWoT ‘rogorc’, mu dros ‘rodesac’,

namuganiSe ‘romeli mxridanac’, muS mamalas ‘rogor Cqara’ (es ukanaskneli

mxolod megrulSi dasturdeba).

mimarTebiTnacvalsaxelian rTul qvewyobil winadadebebSi dadgenilia mTavar

da damokidebul winadadebaTa sami saerTo da erTi Taviseburi ganlageba:

mTavari + damokidebuli + korelati, rac qarTuli zRaprebis enasa da

dialeqtebSic dasturdeba.

naSromSi gamovlenilia mimarTebiTzmnizediani rTuli qvewyobili winadadebis

komponentTa oTxi SesaZlo ganlageba.

maqvemdebarebelkavSirian qvewyobil winadadebSi gvxvdeba sakuTriv megruli

kavSirebi _ -ni, namuda, -da _ da qarTulidan nasesxebi ramdenime maqvemde-

barebeli kavSiri. amaTgan -ni poziciurad gansxvavdeba sxva qarTvelur enebSi

dadasturebuli fardi kavSirebisagan: igi enklitikis kvalobaze mierTvis

damokidebuli winadadebis Semasmenels. namudas-s SemTxvevaSi damokidebuli

132

mosdevs mTavars. -da kavSiriT pirobiT damokidebuli winadadeba ukavSirdeba

mTavars.

megrulis damaxasiaTebeli qvewyobili konstruqciebia: kiTxviTi hipotaqsuri

konstruqcia, romlis mTavars damokidebuli -ni ‘rom’ kavSiriT uerTdeba: mu ore, sak-iT me-gi-Ru-ni? ‘ra aris, tomriT rom migaqvs?’ am konstruqciaSi

mTavari winadadebis qvemdebares marTavs damokidebuli winadadebis Semasmeneli.

megrulSi erTmaneTis gverdigverd arsebobs ganusazRvrelnacvalsaxeliani

martivi winadadeba da hipotaqsuri konstruqcia, sadac ganusazRvreli

nacvalsaxeli damokidebuli winadadebis saxiT aris warmodgenili; zogjer es

nacvalsaxeli mTavari winadadebis Semasmenlis mier imarTvis brunvaSi, an

brunvis niSani masve erTvis: mi(n)-dga-q re-ni ‘viRacam rom aris’.

hipotaqsur konstruqciebSi dadasturebulia leqsikuri aqtualizaciis

SemTxvevebic. esaa: a) vareno/varduo ‘ar aris?/ar iyo?’, rac faqtobrivad

xSiradaa ritorikul frazebSi; b) sityvaTa gameoreba da g) inicialuri saxe-

lobiTis konstruqcia.

gvxvdeba Tqma-brZanebisa da guneba-ganwyobilebis gamomxatvel zmnaTa Semcveli

specifikuri konstruqciebi.

sagrante proeqtis farglebSi am sakiTxebis sakvlevad SemuSavebuli meTodo-

logiiT ganisazRvra sainformacio struqturebis ZiriTadi cnebebi: fokusi da

topiki. cdispirebTan muSaobisas da savele pirobebSi Caweril megrul

teqstebSi davakvirdiT -ni ‘rom’ kavSirian da fer ‘iseTi, imnairi’ korelatis

Semcvel qvewyobil struqturebs, romlebic m. lomias zemoxsenebul naSromSi

specialuri kvlevis sagnad ar aris qceuli, Tumca naxsenebia -ni kavSiriani

damokidebuli winadadebis komponentTa Tavisebur ganlagebasTan kavSirSi

(lomia 2006, 95). Cven SevecadeT, megruli hipotaqsis es erTi kerZo saxeoba

sainformacio struqturebis parametrTa TvalsazrisiT Segvemowmebina Semdegi

winadadebebis magaliTebze:

osuri Fvilun(s) kars mikoxen(i) fer borbolias.

osur-i Fvilun-(s) kar-s mi-kox-e-n(i) fer borbolia-s. woman-NOM kill-TM-PRS.S.3.SG door-DAT PRV-sit-PRS.S.3.SG-CONJ(=that) alike spider-DAT

‘qali klavs karze rom zis im obobas’.

hipotaqsuri konstruqciis mTavari komponenti _ osuri Fvilun(s).... fer borbolias _ gaxleCilia damokidebuli komponentiT _ kars mikoxen(i); magram es ar aris martivi gaxleCa (split), radgan mimdevrobas kars mikoxen(i) fer borbolias myari Sesityvebis Tvisebebi aRmoaCnda da erT frazemad iqna aRqmuli. am aSkarad hipotaqsuri konstruqciis Tavisebureba is aris, rom

133

damokidebuliseuli qvemdebaris elifsisi SeumCnevelia mTavari winadadebis

pirdapir damatebasTan Tanxvdomis gamo. gansxvavebuli brunva mTavari winada-

debis wevrTa Tanamimdevrobis farglebSi problemas ar qmnis.

es winadadeba aris cdispirisaTvis SeTavazebuli suraTis aRweris Sedegi.

Zalian Wirs qarTul TargmanSi sasveni niSnis dasma, megrulSi ki es

SeuZlebelia.

sxva cdispirma igive suraTi ase gadmosca:

kars mikoxen fer borboli-s Fviluns osuri.

kar-s mi-koxe-n fer borbolia-s Fvilu-n-s osur-i. door-DAT PRV-sit-PRS.S.3.SG-CONJ(=that) alike spider-DAT kill-TM-PRS.S.3.SG woman-NOM

‘karze rom zis im obobas klavs qali’.

bondRvis mikoxen fer borbolias yviluns osuri.

bondRvi-s mi-kox-e-n fer borbolia-s Fvilu-n-s osur-i. (spider’s web)-DAT PRV-sit-PRS.S.3.SG-CONJ(=that) alik spider-DAT kill-TM-PRS.S.3.SG woman-NOM

‘ablabudaSi rom zis im obobas klavs qali’.

megrulSi sasveni niSnis dasma am variantSic Wirs imitom, rom segmentSi _

fer borbolias Fviluns osuri _ mTavari winadadebis wevrebi mijriT

aRmoCndnen.

mesame cdispirma suraTi ase aRwera:

ZRab kars mikoxen fer borbolias miotyobans.

ZRab kar-s mi-kox-e-n fer borbolia-s mi-o-tyob-an-s. girl-NOM door-DAT PRV-sit-PRS.S.3.SG-CONJ(=that) alike spider-DAT PRV-VV-creep-TM-PRS.S.3.SG

‘gogo karze rom zis im obobas epareba’.

ZRab qsers gexen fer borbolias miotyobans.

Zrab qser-s ge-xe-n fer borbolia-s mi-o-tyob-an-s. girl-NOM web-DAT PRV-sit-PRS.S.3.SG-CONJ(=that) alike spider-DAT PRV-VV-creep-TM-PRS.S.3.SG

‘gogo qselSi rom zis iseT obobas epareba’.

mTavari winadadeba _ ZRab fer borbolias miotyobans _ gaxleCilia

damokidebuliT _ qsers gexen.

informaciis struqturirebis TvalsazrisiT fokusi miuTiTebs im

alternativebis arsebobaze, romlebic relevanturia lingvistur

gamosaxulebaTa interpretirebisaTvis. Cvens SemTxvevaSi fokusirebulia

qvewyobili winadadebis damokidebuli segmenti – karze an qselSi rom oboba

134

zis, xolo sintaqsuri konstruqciis mTavar nawils kargad Seewyoba topikis

dasadgeni yvelaze moxerxebuli kiTxva:

ras ityviT gogonas (qalis) Sesaxeb? qali (gogo) epareba obobas, qali klavs obobas.

topiki da fokusi sxvadasxva doneebzea markirebuli. Cvens SemTxvevaSi mor-

fologiur markers SesaZloa gauTanabrdes korelati fer(i) ‘iseTi, imnairi’. sxva donis markeria sityvaTa rigis cvla, kerZod ki gaxleCili mTavari

winadadebis wevrebisa da korelatis ganlageba. ganxiluli winadadebebis

magaliTze topikisa da fokusis warmodgenis formalur enobriv saSu-

alebaTagan figurirebs ramdenime: morfosintaqsuri markeri, sityvaTa rigis

cvla da winadadebis gaxleCa. fonetika-fonologiis donis movlenebidan

aRsaniSnavia intonacia. megruli winadadeba erTi amosunTqviT warmoiTqmis,

pauzis gareSe.

radgan korelati mTavari winadadebis Semadgeneli nawilia, gamodis, rom

CvenTvis saintereso qvewyobili winadadebis wevrTa TanamimdevrobaSi _

mTavari+damokidebuli+korelati _ mTavari ixliCeba -ni kavSiriani damo-

kidebuliT, mas ki yovelTvis mosdevs korelati feri. korelatis boloSi dasmiT mTavarsa da damokidebul winada-debebs Soris mijna erTgvarad iSleba da rTuli winadadebis orive komponenti erT mTlian intonaciur-artikulaciur monakveTad iqceva (danelia-canava 1991, Sesavali werili, 9).

korelati hipotaqsuri konstruqciis bolos damaxasiaTebelia qarTulis dia-

leqtebisTvisac:

aswavla, ravarc uTxra, ise (qarT. dialeqt., 1961, 435,8).

iuqme, ravarc Cven uqmofT, isTeio (qarT. dialeqt., 1961, 428,32).

-n(i) kavSiriani qvewyobili winadadebisaTvis specifikuria korelatis

moTavseba winadadebaTa sazRvarze. am dros damokidebuli uswrebs mTavars

(lomia 2006, 95). -ni kavSiriani wevri damokidebuli winadadebis boloSi

dgas da korelati yovelTvis mosdevs mas.

kibes moursi(n) fer boSis moxvadu burTi ukaxale.

kibe-s mo-ur-si(n) fer boSi-s mo-xvad-u burT-i ukaxa-le. Staircase-DAT PRV-walk-PRS.S.3.SG-0-CONJ(=that) boy-DAT PRV-hit-PST.S.3.SGball-NOM back(ADV)

‘kibeze rom Camodis, im biWs moxvda burTi uknidan’.

aq raRa xdeba?

damokidebulia kibe-s mo-ur-s-i-(n);

korelatTan gaTanabrebulia fer;

mTavaria boS-is mo-xvad-u burT-i ukaxale.

am SemTxvevaSi hipotaqsuri konstruqciis komponentebs Soris mijna

mkrTaldeba.

135

sizmari miRud mudgarduni vergudun feri do mumexvariT (xubua 1937, 15,25). sizmar-i mi-Rud mu-dga-rdG-ni v-e-rgu-du-n feri do mu-m-exvar-iT. dream-NOM O.1-VV-have-PST-S.3.SG somebody-was-CONJ(=that) PTC(=not)-to be useful- PST-S.3.SG- CONJ(=that) alike and PRV-O.1-VV-help-0-S.2.PL

‘sizmari mqonda, raRac rom iyo, rom ar vargoda iseTi da momexmareT’.

ma gi-w-u-e ‘‘he’’-ma-qG, Ti-n-wkuma raS-is-G marTax-i qi-gi-aSqvi, u-nwara-Si ve e-Fid-a-sG ni, efer-i (yifSiZe 88, 5).

me (Tu) giTxari ‘‘he’’-meTqi, maSin raSs maTraxi dahkari, ufro mware rom ar

arsebobdes, iseTi’.

zogjer mTavari winadadeba mxolod korelatiTaa manifestirebuli:

qu-da-a-gur-u, dRa-s ko-s va ma-(a)-gi-n-e-n feri (xubua 1937,195,32-33). ‘Seaswavla, verasodes kaciSvili rom ver mougebs iseTi [xerxi]’.

am SemTxvevaSi asaxsneli wevri (misaTiTebeli sityva) Cavardnilia da wina

konteqstiT ivaraudeba.

zogjer -ni kavSiriani damokidebuli gaSifrulia mis win mdgomi gansazRvrebis

an garemoebis SinaarsiT:

Sara-s Tol-e-laRal-ir-i irem-i qo-Zir-G, Tol-i va u-Ru-du-n fer-i (xubua 1937, 242,22).

‘gzaze TvalamoRebuli iremi naxa, Tvali rom ar hqonda, iseTi’.

aseT winadadebebSi informacia iseTnairad gadmoicema, rom mTxrobelma gansxva-

vebuli intonaciiT warmoaCinos sayuradRebo movlena, xolo msmenelma advi-

lad daimaxsovros igi.

zogjer ki winadadebis bolos korelati da misaTiTebeli sityvaa:

Te-s Ji qG-mn-a-Wkad-es, sum-i uRu kambeS-is va ma-Rer-dG-ni, fer-i rkina

(yifSiZe 106,10).

‘mas zemoT daaWedes sami uReli kameCi rom ver moitanda, iseTi rkina.’

qo-Zir-es, koC-iSi Tol-i u-jg-uS-is va Zir-un-du ni, efer-i cira (yifSiZe 88,20).

‘naxes, kacis Tvali ukeTess rom ver naxavda, iseTi gogo.’

ori damokidebulis Semcvel rTul sintaqsur konstruqciaSi korelati fer da misaTiTebeli sityva gaTiSulia:

zaJigalka do sigaret-i u-k-eb-u-n fer koC-i ge-re pepelnica gi-a-Zu-n fer

stol-iS woxoe.

‘sanTebela da sigareti rom ukavia, is kaci dgas, saferfle rom devs, im

magidis win’.

136

zaJigalka do sigaret-i u-k-eb-u-n fer koC-i ge-re uSquri gi-a-Zu-n fer sto-

ÓS woxoe.

‘sanTebela da sigareti rom ukavia, is kaci dgas, vaSli rom devs, im

magidis win’.

me-ul-a Sara-s- G -ni, arT-i koC-q qe-Se-xvad-es, lakv-i me-Funs-u-ni fer-q-i

(danelia, canava 1991,92,30-31).

‘Saraze rom midiodnen, erTi kaci SexvdaT, lekvi rom mihyavda, iseTi.

iSviaTad amgvari ganlagebis ‘rom’ kavSiriani qvewyobili winadadeba qarTul

dialeqtebSic dasturdeba xan mxolod korelatiT:

dedinacvali aRar uSvebs, am gers oqros Tmebi ro aq, imitom (qarT. dialeqt.

1961,450,36-37).

ai qaliSvilebi mieci, me rom giTxra, imaso (qarT. dialeqt. 1961, 427,31-32).

xan _ korelatiTa da misaTiTebeli sityviT:

mova, baRSi rom devi mokla, imisi Zma (qarT. dialeqt. 1961, 452,25-26).

SeiZleba davaskvnaT, rom qvewyobil winadadebebSi mTavar da damokidebul

komponentTa ganlagebis SesaZlo SemTxvevebidan wyoba _ mTavari +

damokidebuli + korelati (misaTiTebeli sityva) _ megrulSi upiratesia

(qarTulTan SedarebiT).

literatura yifSiZe 1914 _ И. Кипшидзе, Грамматика мингрельского (иверского) языка, с хрестоматией и словарем, СПб, 1914. lomia 2006 _ m. lomia, hipotaqsis sakiTxebi megrulSi, gamomcemloba

,,universali’’, Tbilisi.

asaTiani, wereTeli 2008 _ r. asaTiani, n. wereTeli, winadadebis sainformacio

struqtura: implikaciuri topiki qarTulSi _ enaTmecnierebis sakiTxebi,

Tbilisis universitetis gamomcemloba.

xubua 1937 _ m. xubua, megruli teqstebi, tfilisi;

qarT. dialeqt., 1961 _ iv. gigineiSvili, v. Tofuria, iv. qavTaraZe, qarTuli

dialeqtologia, I, Tbilisi.

danelia, canava 1991 _ qarTuli xalxuri sityviereba, megruli teqstebi, II,

gamosces k. daneliam da a. canavam.

137

Marika Jikia One Type of Compound Sentences in Megrelian

Abstract

The compound sentences in Megrelian are well studied (M.Lomia. Issues of Hypotaxes in Megrelian. Tbilisi 2006). One type of compound sentences with the conjunction -n(i) ‘what, which’ and correlate per(i)’like’ is analyzed in the article from the point of view of focus and topic. In the sentences with -ni peri the main clauses are split through the subordinate clauses and the boundaries between the parts of sentences disappear. The both (or more) components of compound sentences form a whole intonational-articulational unity. The construction is peculiar for Megrelian. Such sentences are not attested either in Georgian or in Svan.

138

RUSUDAN ASATIANI

THE MAIN DEVICES OF FOREGROUNDING IN THE INFORMATION STRUCTURE OF GEORGIAN SENTENCES*

Abstract Structuring of information proceeds through the foregrounding of certain parts of the information. In general, foregrounding can be realized on various linguistic levels and it is possible to distinguish: Conceptual, Functional, Discourse and Pragmatic devices, which can be represented by various formal means: Phonetic-Phonological, Morphological-Syntactic and Lexical-Pragmatic. All the devices can co-occur during the information packaging. Some of them are obligatory and are on the high level of the hierarchically organized processes of foregrounding (e.g. conceptual or functional foregrounding); some of them are optional and they are defined by the specific discourse and/or pragmatic values of a sentence (e.g. focus or topic); some forms of foregrounding are implicational (e.g. sometimes reordering implies emphasis of intonation) and so on. The relations between the different kinds of foregrounding are language specific, but it seems possible to speak about universal models of formalization of the information structures. In Georgian there is no morphological topic marker, but all other devices of foregrounding are possible. The paper examines the main models of such devices. 1. Introduction Linguistic structuring of reality based on the notions ‘same-different’ proceeds through ‘oppositions’. An opposition means that there are at least two items one of which is ‘marked’ and another is ‘unmarked’. Structuring of the information, its packaging, also proceeds through oppositions where one part of the information stands out against a background of the other part of the information. From the communicational, pragmatic point of view, this information is highlighted, important and represents the foregrounding of a certain part of information. Any kind of ‘foregrounding’ (res. ‘Highlighting’, ‘Logical Emphasis’, ‘Promotion’, ‘Standing out as the first, important’ and etc) could be regarded as one, common phenomenon which represents the main strategy of structuring of linguistic structures. From this point of view Topic, Focus, Subject, Theme, Point of view and so on – are the same as far as they represent various forms of ‘foregrounding’. It is supposed that such a wide, generalized interpretation of ‘topicalization’ make more clear, what happens when we have mixed forms of ‘foregrounding’. Foregrounding, according to such a wide interpretation, can be realized on various linguistic levels: * This work was fulfilled within the Potsdam Project D2: “Information Structure of a sentence”.

139

1.1. Conceptual Foregrounding During the linguistic structuring of the extra-linguistic situations some languages conventionally conceptualize as the central part of the information either Agent or Patient. In result, either Nominative (which shows agent’s foregrounding) or Ergative (which shows patient’s foregrounding) constructions arise. The first construction formally emphasizes who is acting, while the second emphasizes what is done. From the grammatical point of view, conceptual foregrounding is represented by the unmarked, Nominative case: In the nominative languages it is the Agent, who always stands in nominative, while in the ergative languages it is the Patient (and not the Agent) who appears in nominative. There are some languages which ignore semantic roles. The informational dimension plays a crucial role in the grammatical structures of such languages. This dimension helps the speaker and the hearer to package and retrieve the information: The highlighted part of the information (res. foregrounding of it) is formally marked by a special marker and it is possible to distinguish topicalized, foregrounding part of the information by morphological affixes. 1.2. Functional Foregrounding Patient’s foregrounding in the nominative languages, where agent is conceptually highlighted part, can further (on the second stage of foregrounding) be achieved by the changes of functional roles and as a result passive constructions rise. In the passive construction Patient is functionally promoted and it is defined as the Subject. The term Subject actually denotes foregrounding of a central part of information to whom or what the information concerns. Active construction shows Agent’s foregrounding (that means: Agent is the Subject), while Passive construction shows Patient’s foregrounding (that means: Patient is the Subject). 1.3. Focus During the communicative act, in the discourse, it is usual to stop the gap, which can occur in the information flow. In the dynamic linguistic structures, e.g. in dialogues such supplement of information is fulfilled by question-answer pairs: The demanded information in questions is given in the answers as the highlighted one: That is, foregrounding of the demanded information takes place. Such foregrounding can be reinterpreted as focusing and focal part of information is called Focus. In most cases, the Focus has a specific, marked intonation. It is represented in various languages by the different formal devices. 1.4. Topic From the pragmatic point of view, sometimes it is necessary to make the information more exact, more precise and accurate or hypernymic in order to stress the contrast between the events, to clarify their implicational relations or bridging, to emphasize new or old information, to underline parallel events and so on. All these are reached by

140

foregrounding of the contrasted parts of the information. This process is called topicalization and the foregrounding part of information is called Topic. Conceptual and Functional foregrounding are obligatory. They are always represented in any linguistic structures. Focus is characteristic for the dialogue systems. As for the Topics, they are optional and defined only by the specific situations. 2. Grammatical Models of Foregrounding From the formal point of view, foregrounding can be marked on various linguistic levels: Phonetic-Phonological (The almost universal device of foregrounding is the highlighting of a certain part of information by the marked intonation, which is different from the neutral one. Stress and other supra-segmental means are also possible.); Morphological-Syntactic (Some languages have special morphological markers (cases, particles, clitics) or specific syntactic constructions (reordering of unmarked word order, cleft constructions, different kinds of split, elliptic (short) answers and etc.); Lexical-Pragmatic (It is also possible to use special words, quantifiers or particles for the foregrounding (indeed, certainly, also, just, only, etc.) Besides the emphasis of the definite part of the information, such items add to the whole sentence specific semantics). These devices denote the further foregrounding of any part of the information that is already structuralized and constructed on the conceptual or/and the functional linguistic levels. All these devices can co-occur during the information packaging. Some of them are obligatory and are on the high level of the hierarchically organized processes of foregrounding (e.g. conceptual or functional foregrounding); some of them are optional and they are defined by the specific discourse and/or pragmatic values of a sentence (e.g. focus or topic); some of them are implicational (e.g. sometimes the reordering implies emphasis of intonation) and so on. Different devices and strategies are characteristic for the various languages. The relations between the different kinds of foregrounding are language specific, but it seems possible to speak about the universal signs of this linguistic processes. 3. Georgian Data In the Georgian language there is no morphological topic marker but all other devices of foregrounding are possible. 3.1. Conceptual Foregrounding The Georgian Language shows split ergativity: The Present Tense forms build the Nominative Constructions where conceptual foregrounding means to put the Agent in the central position: monadire-0 k’l-av-s irem-s hunter-Nom kill-Prs.-S.3 deer-Dat (‘The hunter kills the deer’)

141

The Aorist and Perfect Tense forms build the Ergative construction where conceptual foregrounding puts the Patient in the central position: Aorist: monadire-m mo-k’l-a irem-i hunter-Erg Prev- kill-Aor.S.3.Sg deer-Nom Perfect: monadire-s mo-u-k’l-av-s irem-i hunter-Dat Prev-Perf.Vers.-kill-Th.-S.3g deer-Nom 3.2. Functional Foregrounding The Passivization is a regular way for the Patient’s foregrounding for the Present Tense forms: Active: monadire-0 k’l-av-s irem-s hunter-Nom kill-Prs.-S.3 deer-Dat (‘The hunter kills the deer’) Passive: irem-i i-k’vl-eb-a monadir-is mier deer-Nom Prev.-Pass.-kill-Prs.-S.3 hunter-Gen by (‘The deer is killed by the hunter’) The Passive construction is not always formally clearly distinguishable by the verb forms in the Aorist: Active: monadire-m mo-i-k’l-a irem-i tav-is-tvis hunter-Erg Prev.-S.Vers.-kill-Aor.S.3 deer-Nom self-Gen-For (‘The hunter killed the deer for himself’) Passive: irem-i mo-i-k’l-a monadir-is mier deer-Nom Prev.-Pass.-kill-Prs.S.3 hunter-Gen by (‘The deer is killed by the hunter’) Finally, the Passive constructions are almost excluded in the Perfect Tense Forms. Conceptually this fact is understandable: In ergative constructions (such constructions are characteristic for Aorist and Perfect) Patient is already defined as a conceptually foregrounding one and from the informational point of view its further functional foregrounding seems to be redundant. 3.3. Focus The Focus in Georgian is represented by the special rising or wave-like-raising intonation, which differs from the non-focal, neutral intonation. Reordering and, especially, Fronting of the focal part of information is also characteristic. Because the Georgian language has free word order, all logically possible combinations of reordering can be realized as structures with different informational loading. It is difficult to describe and to explain all semantic or pragmatic nuances of these combinations. Intonation+Fronting seems to be the best and the clearest formal device of Focusing. Passivization is not an effective device for focusing because an answer usually has the same functional interpretation as the question has; That is, if a question is formulated by the active construction, an answer will be formulated as the active construction as well

142

and vice versa: the passive question implies passive answer. Focus usually stands before the verb; so, the best order is: Focus - Verb. This regularity must be a result of one the strongest syntactic restriction of word order in Georgian: Question words always are in preverbal position and consequently focus which replaces Wh-words in answers usually appears in the same position. The rising intonation of focus also should be a result of the regularity of intonation phrasing in Georgian: The verb has a tendency to be integrated into the p-phrase of a preceding or a following argument and as one unite it has the boundary tone of a prosodic phrase (p-phrase) which is canonically rising. Here are some typical examples: ra xdeba? (What is happening?) monadire(hunter.Nom) k’lav-s(kills-Prs) irem-s(deer-Dat) vin k’lavs irems? (Who kills the deer?) monadire k’lavs irems ras k’lavs monadire? (What does the hunter kill?) irems k’lavs monadire ras ak’etebs monadire? (What does the hunter do?) k’lavs monadire irems vin ras k’lavs? (Who kills what?) monadire irems k’lavs monadire k’lavs irems irems monadire k’lavs irems k’lavs monadire In answers (especially to repeated questions, which demand to give more accurate information) special particles and definite syntactic constructions arise: monadire k’lavs irems? (Does the hunter kill the deer? Is it the hunter who kills the deer?) diax (yes.Pol), monadire k’lavs irems diaxac (yes.Pol-also=yes.mimicking), monadire k’lavs irems diax, es (this) monadire-a (is), vinc (who) k’lavs irems namdvilad (really) monadire k’lavs irems sc’ored (just,exactly) (rom(that),) monadire k’lavs irems martlac (indeed, right-also) (rom (that),) monadire k’lavs irems martlacda (indeed-and), monadire k’lavs irems irems k’lavs monadire? (Does the hunter kill the deer?/Is it the deer which is killed by the hunter?) diax (yes.Pol), irems k’lavs monadire diaxac (yes.Pol-also=yes.mimicking), irems k’lavs monadire diax, es (this) iremi-a (is), visac (whom) monadire k’lavs namdvilad (really) irems k’lavs monadire sc’ored (just,exactly) (rom(that),) irems k’lavs monadire martlac (indeed, right-also) (rom (that),) irems k’lavs monadire martlacda (indeed-and), irems k’lavs monadire monadire k’lavs irems? (Does the hunter kill the deer?/Does the hunter kill the deer or does he not?) diax (yes.Pol), k’lavs monadire irems

143

diaxac (yes.Pol-also=yes.mimicking), k’lavs monadire irems k’lavs monadire irems, aba (well!) ara (no)? (Of course, the

hunter do kill) namdvilad (really) k’lavs monadire irems sc’ored (just,exactly) (rom(that),) k’lavs monadire irems martlac (indeed, right-also) (rom (that),) k’lavs monadire irems martlacda (indeed-and), k’lavs monadire irems And, so on. Summarizing all the data, we can distinguish the following models for the Focusing: 1. Marked Intonation; 2. Reordering (Fronting) (+Intonation) 3. Syntactic Constructions (+Intonation) 4. Particles (+Syntax+Intonation) 3.4. Topic The intonation is the main device for the Topicalization. The Topic intonation differs from the Focus and the Neutral intonations: it is rising-falling (L*H*L). All devices which are characteristic for the Focus are also possible for the Topic. There can be found also specific particles and constructions. Here are some typical examples: ra-s it’q’vit monadir-is shesaxeb? what-Dat say.Fut-S.2.Pl hunter-Gen about (What about the hunter?) monadire-m mo-k’l-a irem-i hunter-Erg Prev-kill-Aor.S.3 deer-Nom ([The hunter]T killed the deer) ici-t rame irm-is shesaxeb? know-Prs-S.2.Pl something.Nom deer-Gen about (Do you know something about the deer?) irem-i mo-k’l-a monadire-m deer-Nom Prev-kill-Aor.S.3 humter-Erg (The hunter killed [the deer]T) irem-i mo-i-k’l-a monadir-is mier deer-Nom Prev-Pass-kill-Aor.S.3 hunter-Gen by ([The deer]T is killed by the hunter) ra-s it’q’vi-t am monadir-is shesaxeb? what-Dat say.Fut-S.2.Pl this.Gen hunter-Gen about (What about this hunter?)

144

am monadire-m mo-k’l-a irem-i this.Erg hunter-Erg Prev-kill-Aor.S.3 deer-Nom ([This hunter]T killed a deer) ai am monadire-m mo-k’l-a irem-i ([Precisely this hunter]T killed a deer.) here.is this.Erg hunter-Erg Prev-kill-Aor.S.3 deer-Nom namdvilad (really) am monadirem mo-k’l-a iremi ([Really this hunter]T killed a deer./ This is really so that this hunter (and not other) killed a deer./ It is really this hunter who killed a deer. ) sc’ored (just, exactly) (rom(that),) am monadirem mo-k’l-a iremi ([Exactly this hunter]T killed a deer./ This is exactly true, that this hunter killed a deer. It is just this hunter who killed a deer.) martlac (indeed, right-also) (rom (that),) am monadirem mo-k’l-a iremi ([Indeed this hunter]T killed a deer./ It is indeed this hunter who killed a deer.) martlacda (indeed-and), am monadirem mo-k’l-a iremi (Indeed, this is true that it is this hunter who killed a deer.) es is monadire-a, vinc iremi mo-k’l-a this.Nom that.Nom hunter.Nom=be.Prs.S.3 who deer Prev-kill-Aor.S.3 (This is that hunter who killed a deer) ici-t rame am irm-is shesaxeb? know-Prs-S.2.Pl something.Nom this.Gen deer-Gen about es irem-i mo-k’l-a monadire-m this.Nom deer-Nom Prev-kill-Aor.S.3 hunter-Erg (The hunter killed [this deer]T) ai es irem-i mo-k’l-a monadire-m here.is this.Nom deer-Nom Prev-kill-Aor.S.3 hunter-Erg (The hunter killed [precisely this deer]) sc’ored rom es irem-i mo-k’l-a monadire-m just that this.Nom deer-Nom Prev-kill-Aor.S.3 hunter-Erg (The hunter killed [just this deer]) martlac rom es irem-i mo-k’l-a monadire-m indeed that this.Nom deer-Nom Prev-kill-Aor.S.3 hunter-Erg And so on. Same constructions are usual also for the corresponding passive constructions:

145

ai es irem-i mo-i-k’l-a monadir-is mier here.is deer-Nom Prev-Pass-kill-Aor.S.3 hunter-Gen by ([This deer]T is killed by the hunter) And so on. In Georgian specific constructions more often classified as ways to introduce topics, like “As for”, “As far as … is concerned”, “Concerning”, “As regards …”, represent mostly syntactic devices of toplicalization: ra-c she-e-x-eb-a monadire-s, what.Nom-Part Prev-Pass-concern-Th.Suf-Pass.Pres.3.Sg hunter-Dat sts’ored rom is klav-s irem-s exactly that he.Nom kill-Pres.3.Sg deer-Dat

(Concerning the hunter, [just it is he]T who kills a deer.) Summarizing all the data, we can distinguish the following models for the Topicalization: 1. Marked Intonation (different from focus and neutral intonations); 2. Reordering (Fronting) (+Intonation) 3. Syntactic Constructions (+Intonation) 4. Particles (+Syntax+Intonation) 4. Mixed Forms of Foregrounding Different kinds of foregrounding can co-occur and we can speak about the different degrees of ‘Foregrounding’: It is supposed that increasing of formal devices represents rising of the degree of foregrounding and ‘stages’ can conventionally represent this complicated process. As an example, let us consider the sentence: kal-ma gat’exa magida (The woman broke the table) woman-Erg broke Table.Nom 1st stage (Conceptual Foregrounding): Ergative construction represents the Patient ( magida ‘table.Nom’) foregrounding; 2nd stage (Functional Foregrounding): Active construction denotes the Agent (kal-ma ‘woman-Erg’) foregrounding 3rd stage: Intonation emphasis shows different kinds of different foregrounding. It depends on the wider context and on the type of intonation are these highlighting parts Focus or Topic ones? (Underlining in the below examples mark specific changes of an intonation): kalma gat’exa magida kalma gat’exa magida kalma gat’exa magida

146

4th stage: The reordering also works as the marker of foregrounding: kalma gat’exa magida (neutral word order) kalma magida gat’exa (neutral word order) gat’exa kalma magida gat’exa magida kalma magida gat’exa kalma magida kalma gat’exa But the reordering with a certain pitch accent shows clearly a higher degree of foregrounding. The most usual position for Topic is the beginning of the sentence and for the Focus the position before the verb. Fronting together with intonation emphasis gives the highest degree of foregrounding: 5th stage: kalma gat’exa magida gat’exa kalma magida magida gat’exa kalma It is also possible to use specific particles: 6th stage: ai kalma gat’exa magida (The woman (not the other one) broke the table) ai magida gat’exa kalma (The table (not the other thing) was broken by the woman)

ai gat’exa magida kalma (The woman was broken (neither bought, nor made or etc.) the table’)

(The particle ai (‘here is’) implies also fronting and specific intonation.) Specific syntactic constructions (cleft, split…) along with the certain particles show the highlighted part of information as well. 7th stage: es magidaa, kalma rom gat’exa (‘It is the table that the woman broke’) magida, gat’exa kalma, xis (‘The table, the woman broke, wooden’) sts’ored rom magida gat’exa kalma (‘(It is) precisely the table that the woman broke’) If we change the active construction into the passive one, the sentence magida gat’q’da (kalis mier) would show the different foregrounding on the 2nd stage where magida has turned into the Subject. All the possibilities which are characteristic for the topicalization or focusing in active constructions can be used in the passive construction as well: 3rd stage: magida gat’q’da (kalis mier) magida gat’q’da (kalis mier) magida gat’q’da kalis mier 4th stage: gat’q’da magida (kalis mier)

147

kalis mier gat’q’da magida 5th stage: ai magida gat’q’da (kalis mier) ai gat’q’da magida (kalis mier) ai kalis mier gat’q’da magida 6th stage: es magidaa, (kalis mier) rom gat’q’da magida gat’q’da (kalis mier), xis sts’ored rom magida gat’q’da (kalis mier) So, the following hierarchy occurs: 1st stage – Aorist and Perfect show Patient’s foregrounding, while the Present shows Agent’s foregrounding; 2nd stage – Passive construction shows Patient’s further foregrounding in Present; 3rd stage – Marked intonation shows different kinds of Topics or Focuses; 4th stage – Reordering+Intonation (Fronting+Intonation); 5th stage – Syntactic constructions (+Intonation); 6th stage – Particles (+Syntax+Intonation). The first and second stages are obligatory, other stages are optional. We assume that the hierarchy of stages (1<2)<3<4<5<6 presents the rising of the degree of foregrounding. References Asatiani, R. 2000. subiekt’isa da obiekt’is akt’ualizaciis dziritadi mekanizmebi kartulshi (Main devices of the subject and object topicalization in the Arabic language). Typological Researches, IV. Tbilisi: mecniereba. Asatiani, R. 2002. aktualizaciis dziritadi tipologiurad gansxvavebuli modelebi (Main typologically different models of topicalization). Tbilisi: TSU publ. Büring, D. 1995. Topic. Cologne University: Electronic publication. Chikobava, A. 1968. mart’ivi c’inadadebis problema kartulshi (Problem of the simple sentence in Georgian). Tbilisi: mecniereba. Shanidze, A. 1973. kartuli enis gramat’ik’is sapudzvlebi (The foundations of the Georgian language grammar). Tbilisi: mecniereba. Vallduvi, E. 1995. Discourse Configurational Languages. edited by Katalin E. Kiss. New York-Oxford: Oxford Un.Press.

148

STAVROS SKOPETEAS, CAROLINE FÉRY, RUSUDAN ASATIANI

WORD ORDER AND INTONATION IN GEORGIAN

Abstract Georgian is famous for its word order flexibility: all permutations of constituent order are possible and the choice among them is primarily determined by information structure. In this paper, we show that word order is not the only means to encode information structure in this language, but it is used in combination with sentence prosody. After a preliminary description of the use of prosodic phrasing and intonation for this purpose, we address the question of the interrelation between these two strategies. Based on experimental evidence, we investigate the interaction of focus with word order and prosody, and we conclude that some aspects of word order variation are pragmatically vacuous and can be accommodated in any context if they are realized with an appropriate prosodic structure, while other word order phenomena are quite restrictive and cannot be overridden through prosodic manipulations. Keywords: Georgian, word order, intonation, focus, acceptability judgments. 1. Introduction Georgian is well known for its extreme word order freedom, but less so for a comparable richness in tonal structures. The present paper investigates the relation between word order and intonation, and presents the results of a perception experiment in which both components were varied.

In the next section, the syntactic and intonational issues are introduced. The syntactic issues are based on the abundant literature on Georgian morphosyntax (see section 2.1). The detailed discussions of syntactic issues contrast sharply with the paucity of studies on prosody. Indeed, we could rely on few existing studies, as very little has been published about this subject. The survey of the intonational properties of Georgian found in this paper is based entirely on our own research, and has to be considered as preliminary (see section 2.2). Section 3 of this paper presents an experimental study which addresses the issue of the interrelation of prosody and word order for the encoding of information structure. Section 4 presents and discusses the results, and section 5 concludes.

2. Background 2.1. Word order Georgian is characterized as a ‘free word order’ language: all permutations between major clausal constituents are grammatical (see Aronson, 1982: 47; Asatiani, 1982; Boeder, 1989: 160; Davitiani, 1973: 151; Harris, 1981: 22; Hewitt, 1995: 528). The fact that the alternative word orders occur with considerable frequency in texts, as already shown in corpus studies (see Vogt, 1971: 222; Apridonidze, 1986: 136ff.), does not imply that the choice of word order in discourse is random. Aim of this section is to summarize the empirical facts about word order alternation in a theory-neutral way. Verb finality. The following examples illustrate the most frequent orders of clauses involving a transitive verb and two lexically realized arguments, SOV in (1a) and SVO in (1b). Both orders may occur in all-new contexts.

149

(1) (a) ǯarisk’ac-i monadire-s da-č’r-i-s. soldier‐NOM hunter‐DAT PR(FUT)‐cut‐THM‐S.3.SG (b) ǯarisk’ac-i da-č’r-i-s monadire-s. soldier‐NOM PR(FUT)‐cut‐THM‐S.3.SG hunter‐DAT ‘The soldier will wound the hunter.’ Most authors adopt the view that the basic order is V-final (see Pochkhua, 1962: 122, Aronson, 1982: 47; McGinnis, 1997: 8; Harris, 2000: 141-146; Boeder, 2005: 64). The OV order occurs more frequently in the available corpus studies on literary texts (Apridonidze, 1986: 136-143; Kvachadze, 1996: 258-261; Pochkhua, 1962: 122; Vogt, 1971: 222) and equally frequent to the VO orders in a corpus study on fairy tales (Vogt, 1971: 222), which suggests that stylistic factors might have an effect on the choice of verb placement as well (Asatiani, 2007b). The syntactic evidence for the assumption of V-finality is admittedly weak, but all available criteria (e.g., the order in sequences of finite and non-finite verbs in Harris, 2000: 145, object placement with coordinated verbs in Skopeteas and Fanselow, 2008b, etc.) suggest a V-final word order. Verb fronting. Assuming that the canonical order is V-final, we should accommodate the empirical fact that the SVO order in (1b) may also occur in pragmatically neutral contexts, as is already observed in the literature (see Anderson, 1984: 186; Harris, 1981: 22; Hewitt, 1995: 528). We assume that these sentences involve an optional head-fronting operation. The assumption of ‘optional’ V-fronting implies that VO orders are not necessarily the result of a movement operation that targets a position that is associated with a discrete information structural function. V-fronting is a semantically vacuous operation that may be optionally selected in discourse in order to meet preferences on the linearization of the involved constituents (for further discussion, see Skopeteas and Fanselow, 2008b). The exact conditions that determine the choice between VO and OV are not yet studied in detail, but the freedom in the alternation between these orders is acknowledged in all studies on Georgian word order. It is important to note that postverbal arguments in clauses involving V-fronting may be narrowly focused (in which case a particular prosodic structure is used, as we show in this paper, see section 3.3). The freedom in verb placement does not hold for the sentence initial position. Hence, the VSO order illustrated in (2) is a grammatical order, but it occurs in very restricted contexts, e.g. at the beginning of narratives (see Tuite, 1998: 41f.). The different status of V-initial orders is reflected in the corpus study of Apridonidze (1986): in sentences with three to six major constituents, the V is encountered at the sentence initial position only a 3.3% of the corpus occurrences, while it occurs 53.9% in the final position and 42.8% in medial positions (total: 23 253 sentences). Hence, we assume that V-initial orders result from a different syntactic operation (of V-movement to a higher clausal position) which is licensed by a restricted subset of contexts. This implies that the optional fronting of the verb takes place within a lower layer of the clause, that does not include the initial position.1

1 This observation led Anderson (1984: 186) to claim that word order of V-projections is not specified in Georgian, assuming that a V-projection contains the V and its internal arguments.

150

(2) da-č’r-i-s ǯarisk’ac-i monadire-s. PR(FUT)‐cut‐THM‐S.3.SG soldier‐NOM hunter‐DAT ‘The soldier will wound the hunter.’ XPFOCV adjacency. Narrow focus has a very robust effect on Georgian word order that is already known in the literature: narrowly focused constituents (as well as interrogative pronouns) appear left adjacent to the finite verb (see Alkhazishvili, 1959; Harris, 1981: 14, 1993: 1385; Kvachadze, 1996: 250; McGinnis, 1997: 8 citing Nash, 1995; Asatiani, 2007a; Skopeteas and Fanselow, 2008b). Hence, a subset of the non-V-final clauses in Georgian results from the rule for XPFOCV adjacency, which is a distinct phenomenon from V-fronting. While V-fronting is an optional operation that depends on linearization preferences, narrowly focused constituents induce an obligatory attraction of the finite verb, which renders XPFOCYP V structures non-acceptable. However, XPFOCV applies only for narrow focus on preverbal arguments. Postverbal arguments may be narrowly focused too, which – if we adopt a hierarchical application of the rules at stake – suggests that optional V-fronting takes place before the operation that establishes XPFOCV adjacency (see further discussion in Skopeteas and Fanselow, 2008b). Argument scrambling. Argument order follows the order SpO2pO1 (O1=direct object; O2=indirect object) (see Amiridze, 2006: 52), which reflects the embeddedness of the arguments in verb projections. Deviations from the default order are possible and occur very frequently in discourse. Example (3) illustrates the scrambled version of example (1a). (3) monadire-s ǯarisk’ac-i da-č’r-i-s. hunter‐DAT soldier‐NOM PR(FUT)‐cut‐THM‐S.3.SG ‘The soldier will wound the hunter.’ The binding possibilities of the non-canonical orders (McGinnis 1999, 2004, Skopeteas and Fanselow 2008a) as well as long distance dependencies (Skopeteas and Fanselow 2008a) show that the reordering of arguments in Georgian has the properties of scrambling. The choice of scrambled orders in discourse may be influenced by general pragmatic preferences such as ‘givenpnew’, but is not the result of movement to designated information structural positions (see evidence from language production in Asatiani, 2007b, Skopeteas and Fanselow, 2008a). In sum, word order in Georgian exhibits considerable freedom. The V may be optionally fronted in the layer of the clause below the initial constituent. Clause initial verb placement is possible, but it is the result of another syntactic operation that is licensed by restricted contextual conditions. There is a syntactic operation that requires the adjacency of preverbal narrow focused XPs with the verb. Arguments follow the SpO2pO1 order and deviations from this order are instances of scrambling. However, it should be noted that V-fronting is optional independently of the constituent that occupies the initial position. Hence, in sentences with a fronted O (e.g., due to object-topicalization), OSV alternates with OVS in the same way SOV alternates with SVO in neutral contexts.

151

2.2. Intonational issues Some general properties of Georgian intonation are already known through previous studies (Alkhazishvili, 1959; Kiziria, 1987; Tevdoradze, 1978; Zhghenti, 1963). A fundamental issue reported in the literature concerns lexical stress. Authors do not agree on the existence and location of lexical stress, though a majority of researchers assume initial stress; we follow Alkhazishvili (1959), Tevdoradze (1978), and Zhghenti (1963), who claim that pitch accent assignment applies at the post-lexical level. A special prosodic pattern of focused constituents is already mentioned in Alkhazishvili (1959) and Kiziria (1987: 56). The first autosegmental account of the association between intonational tones and phonological phrases is Bush (1999); see also Müller (2007), who finds exactly the same pattern for Georgian questions as Bush. Bush provides an insightful analysis of the boundary tones in question sentences, which he analyzes as a complex LPHP, where L stands for low tone and H for high tone, and a subscript P signals the boundary of a prosodic phrase. This complex boundary tone can be followed by an additional LI or HI, thus an additional boundary tone for an intonation phrase. One of his examples is reproduced in (4). (4) HI LPHP LI [[ rusul-ad lap’arak’-ob ]P]I Russian‐ADV (S.2)speak‐THM.PRS(S.2.SG) ‘Do you speak Russian? As will become clear below, his analysis of the complex contour found at the end of questions as boundary tones is in line with our own results, as we assume that most of the tonal excursions are to be analyzed as boundary tones. Jun, Vicenik & Lofstedt (2007) give an overview of Georgian intonation based on the speech of one speaker who realized declaratives, wh-questions, yes-no questions, focus sentences and other types of sentences. Their analysis of declarative sentences agrees with ours, but the analysis of sentences containing a focus does not, since they find that a narrow focus is always accompanied by a pitch accent (H* or L+H*), while we find that a focus can have a flat and low intonation, depending on the place of the sentence it appears. Given the limits of this paper, we can only outline the main intonational properties of Georgian here, but we report the interested reader to Féry & Skopeteas (2008) for a more detailed account of the phrasing and tonal pattern of declarative sentences containing different focus structures. Our data are described in section 3.3. Prosodic phrasing. First, prosodic phrasing is pervasive and based on syntactic structure. Every constituent forms its own prosodic phrase (p-phrase), with the exception of the verb, which may be integrated into the phrase of an adjacent argument. Tonal structure. Non-final p-phrases have either a rising, or a falling pattern. If they have a rising pattern, a low pitch accent, transcribed as L*, appears very early in the p-phrase. A high tone, which we analyze as a p-phrase boundary tone HP, appears late in the p-phrase. The rise starts relatively late in the last syllable of every p-phrase. If the contour of a p-phrase is falling, it starts at the level of the preceding HP, and falls immediately. It

152

is transcribed as H*L. In longer words, like in some verbs, a second drop can be located on the second accent of these words. As a rule, the first p-phrase has the clearest intonational contour, and the following p-phrases use only reduced F0 range. The last p-phrase of a sentence is always falling, but in many cases is already so low that no contour is realized at all. We analyze it as L* LI. Intonation phrase boundary tones. The final p-phrase of a declarative sentence ends low, and a low boundary tone ends every declarative sentence (see also Alkhazishvili, 1959, Tevdoradze, 1978, Zhghenti, 1963, who observe that the melodic structure of an affirmative utterance is falling). We postulate a LI, which is the boundary tone of the intonation phrase (i-phrase). Downstep. An all-new sentence may be uttered in a single i-phrase, and every p-phrase is downstepped relative to the preceding one, which means that the high part of a p-phrase is lower than the high part of the preceding p-phrase. As already mentioned, only the first p-phrase is realized with a full contour and a large range. Prosodic prominence. Prosodic prominence in Georgian is different from what we are used from other languages. Only sentence initially is a narrow focus optionally accompanied by an increase in F0 height. In the sentence-medial and final position, a narrow focus is accompanied by a lower and flatter contour, and by optional deletion of prosodic boundaries. Prosodic prominence is thus replaced by prosodic leveling. An initial high pitch accent is rather exceptional. In this case, the p-phrase ends with a low boundary tone, LP. The next pair of examples illustrates an interesting difference between the realization of a word as part of an all-new sentence or as a narrow focus. The all-new realization is shown in (5). There is a low tone on the first syllable (L*), and a high tone on the second syllable of bavšví ‘child’ (HP). The subject constituent is phrased separately from the verb. The falling pattern on the verb is interpreted as a H*L, downstepped relative to the high boundary tone of bavšv-i . The last tone is the low boundary tone of the whole i-phrase. (5) {What is happening?} L* HP H*L LI [[bavšv-i]P [i-cin-i-s]P]I child‐NOM PV‐laugh‐PRS‐S.3.SG ‘The child is laughing.’

153

Figure 1 All-new realization

When the subject is narrowly focused, as in (6), there is a rise which already starts on

the first syllable; we analyze this as a pitch accent LH*, and this pitch accent is followed by a low boundary tone LP. The verb has again the pattern H* L. In this case, as well, there are two p-phrases, but the tonal structure is different. (6) {Who is laughing?} LH*L LP H* L LI [[bavšv-i]P [i-cin-i-s]P]I child‐NOM PV‐laugh‐PRS‐S.3.SG ‘The child is laughing.’

L HP H L LI

bavš v-i i-cin-i-s

20

250

50

100

150

200

Time (s)0 1.05626

154

Figure 2 Narrow focus realization

The contrastive accent has a striking effect: in (6) the word bavšvi has a peak located on the first syllable, whereas in the rising contour in (5), the peak is late in the second syllable. The phonological structure is different. A peak on a word has thus two sources: it may originate from a high pitch accent accompanying a narrow focus or a topic in the initial p-phrase; but most often the peaks in the tonal structure are created by boundary tones. 3. Experimental study The aim of the experimental study described in this section is to gain insights into the interaction between intonation and word order in the encoding of information structure in a language with great freedom in both respects. The method involves manipulation of these two factors and the elicitation of acceptability judgments on context-target sentence pairs by 60 naïve listeners (native speakers of Georgian). 3.1. Hypotheses The aim of the experiment is to shed light on the interaction between intonation and word order in the expression of information structure. Hence, our expectations concerning the influence of these factors depend on the assumptions about the relation between prosodic constraints and syntactic representations. Most accounts on the syntax-prosody interface are based on the idea that prosodic constraints apply on the output of syntactic rules (see Selkirk, 1986; Truckenbrodt, 1999; Büring, 2000 among others). This implies that syntactic well-formedness is independent of prosody, i.e. that prosody applies on the subset of structures that are syntactically well-formed (Golston 2005). Though this view may be challenged by a range of phenomena that suggest that particular prosodic constraints may outrank syntactic constraints (see Zec and Inkelas 1990, Harford and

L H* L LP H* L LI

bavš v- i i-cin-i-s

20

25

50

10

15

20

Time (s)0 0.90204

155

Demuth 1999), it generally holds for the kind of phenomena that are dealt with in our paper that word orders that are syntactically non-well-formed cannot be rendered grammatical through prosodic manipulations. Turning now to phenomena that relate to the contextual felicity (rather than to the grammaticality) of particular expressions, a different relation between prosody and syntax is suggested by the linguistic data. Empirical studies on intonational languages show that the role of prosody outranks the role of syntax on the felicity conditions of a particular linguistic expression. Keller & Alexopoulou (2001) present evidence from Modern Greek that violations of prosodic constraints have a stronger negative effect on contextual felicity than violations of word order constraints and they conclude that prosodic structures are ‘grammaticalized’ expressions of information structure, while word orders are not. The hypothesis put forth in this paper is that two types of word order manipulation have to be distinguished. We assume that some word order possibilities are not uniquely associated with particular information structure; their occurrence in particular contexts is a matter of preference. Word order markedness of this kind may be contextually accommodated by a marked prosodic structure that distributes phonological prominence in the way that fits to the context. Another subset of word order possibilities has strong information structural requirements. We assume that the latter word orders are licit only if the information structural requirements are met; if not, their contextual felicity cannot be ‘repaired’ by a felicitous prosody. We refer to the former type of word order markedness as resulting from the violation of ‘weak’ word order constraints and to the latter as resulting from the violation of ‘strong’ word order constraints. If we apply this distinction to the observations about Georgian syntax in section 2.1, we may formulate our expectations about the interaction of prosody and word order in this language. We argued with reference to the discussion in the previous literature that argument reordering and V-fronting are optional operations in Georgian. Their occurrence may be influenced by contextual factors, such as the preference for a givenpnew order for scrambling or particular stylistic or prosodic preferences for V-fronting. We hypothesize that sentences involving either operation are marked, but a felicitous prosody may repair the effects of their markedness in any context. On the other hand, we argued that V-initial sentences result from a syntactic operation that differs from V-fronting and that this operation is licensed by a very restricted type of contexts. Hence, we assume that, when the contextual requirements are not met, the infelicity of the V-initial structure will not be overridden by a felicitous prosody. Furthermore, we argued that XPFOCV adjacency is obligatory for narrow focus in Georgian. In the same vein, a felicitous prosody is not expected to accommodate violations of this rule. In sum, we assume that scrambling or V-fronting correspond to a weak constraint violation, while V-initiality or a violation of XPFOCV adjacency correspond to a strong constraint violation. Expressions involving violations of the former but not of the latter type can be rendered felicitous when presented with a marked prosodic structure that fits the contextual requirements.

156

In the context of question-answer pairs, that we examine in our experimental study, we assume that when a marked prosodic realization of the answer meets the requirements of the question, scrambling and V-fronting will not have a substantial effect on the intuition of contextual felicity. V-initiality and violations of the XPFOCV adjacency are expected to have a strong negative effect. When the prosodic realization of the answer does not fit the question, a strong effect on contextual felicity should result. In this case, the prosodically suboptimal sentences should not be equally rejected, but rather exhibit gradience reflecting their syntactic properties, as it has been already observed in similar experiments of this type (see Keller & Alexopoulou 2001). Finally, similar studies on English and German (Gussenhoven, 1983; Alter et al., 2001; Féry and Stoel, 2006) have firmly established that, in these languages, a prenuclear accent is readily accepted, whereas a postnuclear one provokes strong negative reactions. A related question for Georgian is to investigate whether postnuclearity also correlates with deaccenting. We expect that in prosodically non-congruent sentences there will be an influence from the position of the accent: postnuclear accents are expected to have a stronger impact on the acceptability judgments. (7) Hypotheses

a. SYNTACTIC VIOLATIONS: Strong violations have a stronger impact on acceptability than weak violations.

b. PROSODIC VIOLATIONS: A non-felicitous prosody has an additive negative effect to the felicity of syntactic representations.

c. PROSODY AND SYNTACTIC VIOLATIONS: A marked licit prosody only overrides the negative effect of weak word order violations, but cannot have the same effect when strong violations occur.

d. ACCENT PLACEMENT: Postnuclear non-licit accents have a stronger impact on the acceptability judgments than prenuclear non-licit accents.

3.2. Method Items and conditions In order to test the hypotheses in 3.1, we designed an experiment based on question/answer pairs. The experiment included three factors: a contextual factor determined by the question (5 levels), the word order of the answer (4 levels), and the prosodic realization of the answer (2 levels). Crossing the levels of all factors gave 5×4×2=40 experimental conditions. These conditions were implemented in four different sentences, performed in two different tenses each, in order to compensate possible effects of case marking, which is determined by the inflectional properties of the verb in Georgian. In sum, our material contained a total of 40×8=320 question/answer pairs. Four different verbs were used in all examined conditions. In order to have four verbs with exactly the same behavior (without differences in use or frequency which could cause variance in the results), we chose four causative verbs, which are shown in (10a-d). In Georgian, causative verbs have the same syntax as ditransitive verbs, licensing a subject (causer), an indirect object (causee), and a direct object (theme). Causee

157

arguments are dealt with as indirect objects in Georgian grammar since they have identical case marking, word order, and agreement properties with recipient constituents. (8) (a) item 1 dato nino-s c’ign-s c’a-ak’itx-eb-s.

Dato(NOM) Nino-DAT book-DAT PR(FUT)-(IO.3)CAUS.read-THM-S.3.SG ‘Dato will cause Nino to read a book.’

(b) item 2 deda gogo-s c’eril-s da-ac’erin-eb-s.

mother(NOM) girl-DAT letter-DAT PR(FUT)-(IO.3)CAUS.write-THM-S.3.SG ‘The mother will cause the daughter to write a letter.’

(c) item 3 kal-i k’ac-eb-s muxa-s mo-ač’revin-eb-s.

woman-NOM man-PL-DAT oak-DAT PR(FUT)-(IO.3)CAUS.cut-THM-S.3.SG ‘The woman will cause the men to cut the oak.’ (d) item 4 bavšv-i k’at’a-s tagv-s da-ač’erin-eb-s.

child-NOM cat-DAT mouse-DAT PR(FUT)-(IO.3)CAUS.catch-THM-S.3.SG ‘The child will cause the cat to catch the mouse.’

Case marking in Georgian depends on the inflectional properties of the verb. In the verb class that occurs in the stimuli of our experiment, agent constituents bear nominative case in the present and future indicative (direct case marking pattern), while they bear dative case in the perfect tenses (inverse case marking pattern). Theme constituents bear dative case in the direct case marking pattern and nominative case in the inverse pattern. However, case marking does not influence word order or binding asymmetries between agents and patients (see Harris 1981, Skopeteas, Fanselow, and Asatiani 2008). In the inverse case marking pattern, agents are cross-referenced by the same person affixes as goal constituents in the direct pattern, while themes in inverse pattern are cross-referenced by the same affixes as agents in the direct pattern. In order to outbalance a possible impact of case marking in our results, all sentences were used in two different tenses (future indicative and perfect) which license two different case markings on the arguments: causer=nominative, causee=dative, and theme=dative in the future, and causer=dative, causee=postpositional, and theme=nominative in the perfect (compare 9 with the illustrative example for item 1 in example 8). (9) item 1, perfect tense & inverse case marking dato-s nino-s-tvis c’ign-i c’a-uk’itxvinebi-a.

Dato-DAT Nino-GEN-for book-NOM PR-(INVS.3)CAUS.read:PRF-INVO.3 ‘Dato has (apparently) caused Nino to read a book.’ All four items were produced in four different word orders. The aim of the word order

manipulation was to test hypotheses concerning word order markedness. For this purpose, we chose (a) the canonical word order SO2O1V (Word Order 1, WO1, see (12a)), (b) a word order involving argument reordering, namely O2O1SV (WO2, see (12b)), (c) a word order involving argument reordering (henceforth ‘A-reordering’) and V-fronting, namely O1SVO2 (WO3, see (12c)), and (d) a word order involving V-initiality and also reordering of the arguments, namely VSO1O2 (WO4, see 12d). In the sense of the distinction between weak and strong word order violations, that we presented in section 3.1, WO1 does not display any violation, WO2 displays a weak violation (A-reordering), WO3

158

display two weak violations (A-reordering and V-fronting), and WO4 displays a weak and a strong word order violation (A-reordering and V-initiality). Additionally, WO1-WO3 may display a violation of XPFOCV adjacency depending on context (i.e., whenever the context licenses focus on a preverbal argument that is not adjacent to the verb). (10) (a) WO1 (SO2O1V) dato nino-s c’ign-s c’a-ak’itx-eb-s.

Dato(NOM) Nino-DAT book-DAT PR(FUT)-(IO.3)CAUS.read-THM-S.3.SG ‘Dato will cause Nino to read a book.’ (b) WO2 (O2O1SV) nino-s c’ign-s dato c’a-ak’itx-eb-s. (c) WO3 (O1SVO2) c’ign-s dato c’a-ak’itx-eb-s nino-s. (d) WO4 (VSO1O2) c’a-ak’itx-eb-s dato c’ign-s nino-s. Finally, all four orders were combined with five different context questions. The answer types that are investigated appear in (11), along with an example of a question. (11a) asks for an all-new sentence. (11b-d) ask for a narrow focus on an argument: subject, direct object or indirect object. (11e) asks for a multiple focus.2 These contexts were chosen in order to allow an examination of the interaction of argument focus with the four word orders. Further contexts that would be of interest are V-focus and VP-focus, which were not included in our experimental design. (11) (a) All-new pattern (wide focus) answer to ‘What is happening?’ (b) Subject focus answer to ‘Who will cause Nino to read a book?’ (c) Direct object focus answer to ‘What will Dato cause Nino to read?’ (d) Indirect object focus answer to ‘Whom will Dato cause to read a book?’ (e) Multiple focus (subject and direct object) answer to ‘Who will cause Nino to read what?’

We call ‘congruent’ a prosodic realization that is felicitous in a particular context, and ‘non-congruent’ a non-felicitous one. For each context question, each answer was presented in two different intonation patterns: one realization was congruent to the context, i.e., the prosodically prominent part was the part under question, and the other realization was non-congruent. This is shown in (12) with an example for each case (the

2 We do not expect an influence of the word order of the question on the acceptability scorings. The word orders of the questions used in the experiment do not coincide with any of the word orders of the answers, as shown in following: (a) All-New Question: SV? (b) Subject Focus Question: SVO2O1? (c) Direct Object Focus Question: O1VSO2? (d) Indirect Object Focus Question: O2VSO1? (e) Multiple Constituent Question: SVO2O1?

159

exact prosodic realizations of these sentences are discussed in section 3.3). Capitals indicate prosodic prominence. (12) (a) Question ‘Who will cause Nino to read a book?’ (b) WO1 (SO2O1V), congruent prosody DATO nino-s c’ign-s c’a-ak’itx-eb-s.

Dato(NOM) Nino-DAT book-DAT PR(FUT)-(IO.3)CAUS.read-THM-S.3.SG ‘DATO will cause Nino to read a book.’ (c) WO1 (SO2O1V), non-congruent prosody dato NINO-S c’ign-s c’a-ak’itx-eb-s.

Dato(NOM) Nino-DAT book-DAT PR(FUT)-(IO.3)CAUS.read-THM-S.3.SG ‘Dato will cause NINO to read a book.’ In order to keep the material to be judged within comfortable limits, the material used in the perception experiment included for every context and every word order, the prosody hypothesized to be congruent and only one of the prosodies hypothesized to be non-congruent. Which non-congruent answer was used in each case is shown in Table 1. For instance, in the all-new context, and in WO1, the non-congruent answer was the one with a prosodically prominent indirect object in the leftmost top cell of Table 1 (the prosodically prominent argument is underlined). Table 1 Prosodically prominent constituents in non-congruent answers WO1 WO2 WO3 WO4 all-new question SO2O1V O2O1SV O1SVO2 VSO1O2 subject question SO2O1V O2O1SV O1SVO2 VSO1O2 direct object question SO2O1V O2O1SV O1SVO2 VSO1O2 indirect object question SO2O1V O2O1SV O1SVO2 VSO1O2 multiple constituent question SO2O1V O2O1SV O1SVO2 VSO1O2 Material The sentences used in the production experiment were spoken by two native speakers of Georgian, both women in their twenties, with a background in linguistics. The recordings took place in separate sessions and were conducted by the second author in Tbilisi, in September 2005. The first speaker performed the questions and the second speaker performed the answers. The speakers were aware of the goal of the experiment, were instructed to speak as naturally as possible, and were allowed to correct themselves as often as they wanted. The second speaker read the targeted answers in the context of written all-new or constituent questions (depending on experimental condition) and was instructed to put emphasis to the information under question. The performances were recorded on a DAT recorder (SONY 100). Later on, some informal naturalness checks of the recordings of both speakers were made with other Georgian speakers, one of these being the third author. All listeners confirmed that the speakers were very natural. Method Question/answer pairs were presented acoustically to listeners who gave judgments on a scale from 1 to 5 about their appropriateness: 1 was the best, 5 the worst, and 2-4 were intermediate levels. The participants were instructed to listen carefully to the question/answer pairs and to give a higher score if the answer sounded natural and made

160

sense as an answer to the preceding question. The written instructions made clear that the ratings should not reflect their intuitions about “correct” Georgian, but rather the speaker had to consider whether they thought that the question/answer pair could occur in a natural conversation. The use of scalar judgments is based on the assumption that native speakers perceive different degrees of appropriateness of several structures in context (see Cowart, 1997, for methodological aspects, as well as Keller and Alexopoulou, 2001, for a similar experiment on Greek, and Féry and Stoel, 2006, for some results on the perception of similar pairs in German). Sixty (male and female) native speakers of Georgian, all students at the University of Tbilisi, participated to the experiment, which was conducted by the first and third authors and two Georgian student assistants. The experiment took place at the University of Tbilisi and was performed in two days. The informants were paid for their participation. The 320 pairs were divided into 4 subsets, and each informant was asked to rate only one set. As a result, every person had to give a rating for 80 question/answer pairs, which contained each experimental condition twice. The subsets of sentences were rated by 15 persons each. The question/answer pairs had been implemented in a DMDX presentation and were automatically randomized for each subject. The duration of the entire experimental session was approximately 20 minutes. Subjects first went through a training session containing five question/answer pairs, and when they felt confident with the task, they went on to the actual experiment. 3.3. Intonation pattern of the stimuli The first sentence illustrates the tonal contours of the material used in the experiment. The speaker was very consistent in her productions of the four sentences, so that the illustrations are representative for the whole material. To keep the discussion short, one example for each focus pattern is given here, that shows the main characteristics of the tonal patterns.3 A special feature of Georgian intonation that we have already mentioned in section 2.2 is the clear phrasing in each realization. A constituent is generally phrased individually and realized with the tonal excursions typical for a p-phrase, with the exception of the verb. In our data, the verb is nearly always included in the same p-phrase as an adjacent argument. In general, it is the preceding argument, like the direct object in WO1 (SO2O1V). When the verb is sentence initial, there is no preceding argument and the verb is included in the phrase of the following argument, which is the subject in WO4 (VSO1O2). As for the tonal contour, recall that non-final phrases have two possible contours: a rising contour, analyzed as L* HP, or a falling one, analyzed as (L)H*L LP (see section 2.2). We also mentioned the fact that the final p-phrase is often so low that no contour is realized. In this case, we assume that only the final boundary tone (LP or LI) is assigned to the p-

3 According to the intuition of the third author, a native speaker of Georgian, there is always a lexical stress on the first syllable of each word in our examples, and the verbs can also have a secondary stress on the third syllable. As in our sentences the verb had only little prominence (because the sentences were elicited either in ‘all-new’ contexts or in contexts inducing argument focus), there is no tonal reflex of secondary stress in the pitch tracks.

161

phrase. Another variation concerning the falling pattern is that the high pitch accent H* may be preceded by a rise, that we analyze as a trailing tone L coming before the high pitch accent. In this case the contour of the p-phrase is LH*L LP/I. In the all-new sentences, an illustration of which is given in (15a) and Figure 3, the non-final p-phrases were realized by our speaker with the default contour L* HP. We found a single counterexample: in WO3 (O1SVO2), the direct object carried a falling contour, which may reflect a different spontaneous interpretation of the information structure of the sentence, due to the marked word order. The p-phrase containing the verb, the last one in (15a), often started with a high tone followed by an immediate fall. The final boundary tone is low as it is the last p-phrase in the sentence. An additional feature of Georgian intonation is visible from the pitch track in Figure 3: downstep of the H tones of each p-phrase in an i-phrase, expressing cohesion between the p-phrases. Each H tone is lower than the preceding one. (13) (a) All-new congruent pattern L* HP L* HP H*L LI [dato]P [nino-s]P [c’ign-s c’a-ak’itx-eb-s]P Dato(NOM) Nino-DAT book-DAT PR(FUT)-(IO.3)CAUS.read-THM-S.3.SG Figure 3 Item 1 in WO1 (SO2O1V): all-new context

L* HP L* HP H* L LI

dato nino-s c'ign-s c'a-ak'itx-eb-s

80

400

200

300

Time (s)0 2.98347

162

(15b) is a sentence with narrow focus on the subject. The accompanying pitch track in Figure 4 shows that the focused p-phrase may be realized with a falling contour (though this correlation is not obligatory). In this sentence also, we find downstep of the H tones. The last p-phrase already starts at a low level and remains low and flat throughout. This is an instance of the low p-phrase that we analyze as L* LI.

(b) Subject focus congruent pattern

LH*L LP L* HP L* LI [DATO]P [nino-s]P [c’ign-s c’a-ak’itx-eb-s]P Figure 4 Item 1 in WO1 (SO2O1V): subject focus

L H* L LP L* HP L* LI

DATO nino-s c'ign-s c'a-ak'itx-eb-s

80

400

200

300

Time (s)0 2.52254

163

The pitch track in Figure 5 illustrates a rather special contour. Both unfocused arguments, the subject and the indirect object, as well as the focused argument, the direct object, have a pitch accent, but the one on the direct object is much higher than the others.

(c) Direct object congruent pattern LH*L LP LH*L LP LH*L LI [dato]P [nino-s]P [C’IGN-S c’a-ak’itx-eb-s]P Figure 5 Item 1 in WO1 (SO2O1V): direct object focus

L H*LLPLH* L LP L H* L LI

dato nino-s C'IGN-S c'a-ak'itx-eb-s

80

400

200

300

Time (s)0 3.5078

164

In Figure 6, with focus on the indirect object, the prefocal subject does not exhibit the default contour L* HP, but has a low contour L* LP. The focused argument, by contrast, has an unmarked L* HP contour. Such sentences occurred only rarely in our data.

(d) Indirect object congruent pattern

L* LP L* HP H*L LI [dato]P [NINO-S]P [c’ign-s c’a-ak’itx-eb-s]P Figure 6 Item 1 in WO1 (SO2O1V): indirect object focus

L* LP L* HP H* L LI

dato NINO-S c'ign-s c'a-ak'itx-eb-s

80

400

200

300

Time (s)0 2.85689

165

Figure 7 shows a multiple focus on the subject and on the direct object. Both the subject and the object have a pitch accent (H*L). The object forms a final p-phrase together with the verb. The accented effect comes from the unusual height of the p-phrase-initial high tone on c’ign-s ‘the book’, not visible in the pitch track because of the voiceless affricate starting this word.

(e) Multiple focus congruent pattern L H*L LP L* HP H*L LI [DATO]P [nino-s]P [C’IGN-S c’a-ak’itx-eb-s]P Figure 7 Item 1 in WO1 (SO2O1V): multiple focus

L H* L LP L* HP H* L LI

DATO nino-s C'IGN-S c'a-ak'itx-eb-s

80

400

200

300

Time (s)0 3.48905

166

A final feature of our data remains to be remarked upon. The word orders WO3 (O1SVO2) and WO4 (VSO1O2) had instances of sentence-final narrow focus. All eight instances (distributed over the four sentences) displayed the special low and flat contour L* LI. Additionally, some consonants show more tenseness and aspiration. An example of the final focus (L*LI) is shown in (14), with item 2. In (14), a falling pitch accent H*L is realized on the first p-phrase and a rising contour appears in the second p-phrase. (14) H*L LP L* HP L* LI

[[c’eril-s]P [deda da-ac’erin-eb-s]P [GOGO-S]P]I letter-DAT mother(NOM) PR(FUT)-(IO.3)CAUS.write-THM-S.3.SG girl-DAT Figure 8 Item 2 in WO3 (O1SVO2): indirect object focus

To sum up, the production experiment revealed the following features of Georgian intonation. A. Comparison between the tonal contours of an all-new sentence and of a narrow focus Some regular differences between the realization of words in an all-new sentence and in narrow focus could be identified. First, an all-new sentence mostly contains non-prominent phrases, realized tonally as a sequence L* HP, especially when the word order is unmarked. When the narrowly focused word is initial, it is often realized with a rise-fall LH*L LP. However, the rise-fall contour is not bi-uniquely associated with focus, since it also occurs with non-focal constituents (see Figure 5). In the case of non-final focus, the remainder of the sentence has a compressed range. However, at least when the subject or the indirect object is initial (in WO1 (SO2O1V) and WO2 (O2O1SV)), the remainder of the

H* L LP L* HP L* LI

c'eril-s deda da-ac'erin-eb-s GOGO-S

80

400

200

300

Time (s)0 3.04694

167

sentence is fully intonated. In WO3 (O1SVO2), it is the direct object which is initial; in this case our speaker realized a narrowly focused word with a very high pitch accent in the rise-fall, and the remainder of the sentence was more compressed than in the other cases. A medial narrow focus (neither sentence-initial, nor sentence-final) is realized just with a rise when it is integrated with the following verb (direct object in WO1 (SO2O1V), subject focus in WO2 (O2O1SV) and WO3 (O1SVO2)), but when it is phrased individually, it is again often realized with a rise-fall (subject in WO3 (O1SVO2) and direct object in WO4 (VSO1O2)). The indirect object in WO1 (SO2O1V) is an exception. It is phrased individually, but it is realized just with a steep rise. The reason could be that the subject in this pattern is realized with a rise-fall, maybe because the speaker realized it as a topic, and two full rise-falls in a row are avoided. A sentence-final focus is realized differently from all the others: it is realized separately, in a p-phrase preceded by a short break, at a low level and with a flat intonation. It was analyzed as L* LI. In all cases, a crucial property of the narrow focus is what we call emphatic realization: the consonants are realized with intensity, and the words are longer than in the sentences with wide focus. But this property is exaggerated on a final constituent. In sum, we assume that the correlates of narrow focus are a high pitch accent when non-final and a low pitch accent when final, a clear phrasing except when preverbal, and tenseness as well as intensity in the segments of the focused word. B. Prenuclear part Before the high tone of a narrow focus, the full tonal structure is usually present. In verb-final WO1 (SO2O1V) for instance, even if the focus is on one of the objects, the subject and the prenuclear object have the same tonal contour as the one they have in the all-new pattern. The same can be said for the WO2 (O2O1SV) and WO3 (O1SVO2), if the focused word is not final. In WO4 (VSO1O2), there was some compression and restructuring of the prenuclear material, due to the high markedness of this word order. As will be seen in the next section, this pattern was perceived as extremely marginal by the informants, and it would be unwise to draw generalizations on the Georgian intonation from this pattern. In a sentence with final focus, the prefocal stretch of discourse is clearly compressed. It is also realized more rapidly and without any correlate of the emphasis we find in narrowly focused words. C. Postnuclear part In many languages, the material after a narrow focus (postnuclear) is the place where deaccenting is observed. But in our elicited material, it is remarkable how little material is deaccented. Some examples of a sentence-initial narrow focus are illustrated above. But still, the remaining material – except for the verb – is realized with a full tonal pattern. Again, it is difficult to express generalizations on deaccenting from our experimental material, and more studies on Georgian intonation are necessary.

168

4. Results Sixty subjects participated in the perception experiment and gave 80 judgments each. Of the total of 4 800 trials, 455 trials were non-valid (subjects failed to select any value from 1 to 5 within the time window). Some additional judgments had to be eliminated due to errors in the combination of question/answer pairs in the DMDX presentation, such that the final data set contained 3 797 (79.1%) valid judgments. The means of judgments reported in the subsequent sections relate to the valid raw data. Valid judgments have been normalized through transformation into standard scores in order to eliminate individual differences in the way subjects perceived the rating scale (see Cowart, 1997: 114). In the final subject/treatment table, missing values (which resulted from the elimination of the non-valid data) have been imputed through a regression analysis carried out on the valid values as predictors for the estimation of the missing ones; standardized residuals have been added to the predicted values in order to correct the reduction of the standard errors which results from the imputation procedure (see Rietveld and Van Hout, 2005: 202ff.). All statistical analyses have been conducted on the normalized data after the imputation procedure. In the subsequent sections, we discuss the interaction between different word orders and prosodic realizations on the means of judgments. The experimental design contains three factors, i.e., context, word order, and prosodic structure (see 3.2). The effects and the interaction of these factors will be presented in a nested design in the following (section 4.1 to 4.5)Our hypotheses in 3.1 relate to the interaction of prosody and word order for the encoding of particular information structures. The identification of the levels of these factors depends on contextual condition (congruent vs. non-congruent prosody, violations of XPFOCV adjacency). Hence, it suggests itself to nest the data set on the basis of the factor ‘context’. In section 4.6 we summarize our major findings and test our hypotheses in the entire data set. 4.1. All-new In an all-new sentence, every constituent is focused, and, as a result, the constraint of XPFOCV adjacency that relates to narrow focus does not apply. The judgments are expected to reflect the word order markedness that results from A-reordering, V-fronting, and V-initiality. These expectations are presented in Table 2: based on the assumption that multiple violations will have an additive effect on the acceptability of the presented orders, and the assumption of different strength of the involved violations, the violation profile in Table 2 allows for prediction about the acceptability of the examined orders. Table 2 Word order violations in all-new questions all-new STRONG WEAK predicted rank WO1 (SO2O1V) 1 WO2 (O2O1SV) A-reordering 2 WO3 (O1SVO2) A-reordering, V-fronting 3 WO4 (VSO1O2) V-initial A-reordering 4 Two prosodic realizations have been examined: The congruent prosodic structure corresponds to a “neutral” intonation, in which no particular prosodic prominence was

169

rendered to any constituent (see the orders lacking underlining in Appendix). Neutral intonation of this kind is not expected to have an accommodating effect on marked word order. The incongruent prosodic structure was intended for a narrowly focused argument. In other words, only one constituent is prosodically focused, the other ones are prosodically realized as given. The predictions of Table 2 were borne out, as shown in the obtained results in

Figure 9 (error bars show the standard error of the means at a 0.95 level). A repeated measures analysis of variance revealed a significant main effect for prosodic congruence (F1,59=50.32, p <.001) and for word order (F1,59=21.36, p <.001), as well as a significant interaction between the two factors (F1,59=6.98, p <.01). Figure 9 Judgments in all-new questions

Figure 9 and subsequent figures are based the normalized data (see normalization procedure in the previous section). The non-normalized means of the obtained judgments as well as the corresponding normalized values (standard scores) are given in the Appendix. The highest (non-normalized) mean in all parts of the experiment is 4.6 (see Appendix), i.e. no question/answer pair was given the maximal value 5 by all speakers. We can only speculate about the sources of this variation: first, certain speakers did not use the maximal value at all. Second, particular items may have had negative effects on the intuition of acceptability though they instantiate a fully grammatical syntactic configuration, etc. However, the experimental data does not allow for generalizations concerning the absolute numeric values. The crucial point is the difference between the observed means that provides evidence for a differential impact of the compared experimental conditions. Following Keller and Alexopoulou (2001: 3.5), we assume that

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

SO2O1V O2O1SV O1SVO2 VSO1O2

w ord orders

z-sc

ores

of a

ccep

tabi

lity

congruent prosody

non-congruent prosody

170

two means of acceptability judgments significantly differ whenever their numerical difference is larger than one standard error. In particular, we assume that x>y holds when both x is higher than the higher bound of y (y+SEy) and y is lower than the lower bound of x (x–SEx). Prosodically congruent sentences completely confirm the predictions of Table 2. Prosodically non-congruent tokens obtained lower scores, and do not provide evidence for all predicted levels of word order markedness, since only WO1 (SO2O1V) reached a distinct level of acceptability compared to the other orders. With respect to the strength of the negative effect of the involved factors, we observe that a strong word order violation, such as the one exemplified through WO4 (VSO1O2), has a comparable effect on acceptability to the non-congruent prosody on the canonical word order WO1 (SO2O1V). 4.2. Subject focus In this condition, the context question induces a narrow focus on the subject. The availability of a narrowly focused XP has consequences for the XPFOCV adjacency constraint. Orders that provide a preverbal focused argument that is not left adjacent to the verb exhibit a strong violation, which holds for WO1 (SO2O1V). The further constraint violations are identical to the all-new contexts, discussed in section 4.1 and are summarized in Table 3. The answers to constituent questions are realized with a marked prosodic structure, i.e. a prosody that renders prominent the questioned constituent. As we argued in 3.1, weak word order violations are expected to be accommodated when a marked prosodic structure places prosodic prominence of the argument that is expected to be focused in the given context. Table 3 Word order violations in subject questions S focus STRONG WEAK predicted

rank WO1 (SO2O1V)

XPFOCV violation

3

WO2 (O2O1SV)

A-reordering 1

WO3 (O1SVO2)

A-reordering, V-fronting

2

WO4 (VSO1O2)

V-initial A-reordering 4

Figure 10 presents the obtained results in this contextual condition. A repeated measures analysis of variance showed two significant main effects for prosodic congruence (F1,59=78.84, p <.001) and for word order (F1,59=24.54, p <.001). There is no significant interaction between the two factors, which confirms the observation that may be gathered from Error! Not a valid bookmark self-reference. that the effects of these factors are cumulatively combined in the result.

171

Figure 10 Judgments in subject questions

The placement of the focused subject in the position which is left adjacent to the verb increases acceptability, as may be seen in the acceptability means for WO2 (O2O1SV) and WO3 (O1SVO2) (compare also with Figure 9). The difference between WO2 (O2O1SV) and WO3 (O1SVO2) is slightly larger than a standard error (see Appendix) in the prosodically congruent realizations. Note, furthermore, that WO1 (SO2O1V) obtained very high scores that only slightly differ from WO3 (O1SVO2). This is due to the fact that SO2O1V is the canonical word order in the language and it retains a high acceptability level across contexts. The result in Figure 10 is very informative for our hypothesis concerning strong and weak word order violations. Weak violations, such as the reordering of the arguments in WO2 (O2O1SV) and WO3 (O1SVO2) are largely ignored when the prosody is appropriate in the context. This does not hold for WO4 (VSO1O2) that incurs a strong violation. Note that the negative effect of a strong word order violation has the same strength with the effect of a non-congruent prosodic structure on the contextually licit orders WO2 (O2O1SV) and WO3 (O1SVO2). 4.3. Direct object focus The next context question licenses a focus feature on the direct object. XPFOCV is satisfied by the canonical order WO1 (SO2O1V), since in this order the focused constituent is placed immediately before the verb. The orders WO2 (O2O1SV) and WO3 (O1SVO2) incur a strong XPFOCV violation in this contextual condition next to the weak violations relating to A-reordering and V-fronting. This constraint does not apply when

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1


w ord orders

z-sc

ores

of a

ccep

tabi

lity

congruent prosody


172

the verb is fronted as in WO4 (VSO1O2), since postverbal focus is always possible in Georgian (see discussion in section 2.1). Table 4 Word order violations in direct object questions O1 focus STRONG WEAK predicted

rank WO1 (SO2O1V)

1

WO2 (O2O1SV)

XPFOCV violation

A-reordering 2

WO3 (O1SVO2)

XPFOCV violation


3

WO4 (VSO1O2)


The results are presented in

Figure 11. A repeated measures analysis of variance revealed a significant main effect of prosody (F1,59=47.5, p <.001) and a significant effect of word order (F1,59=54.03, p <.001), as well as a significant interaction between word order and prosody (F1,59=6.99, p <.01). Figure 11 Judgments in direct object questions

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1


w ord orders

z-sc

ores

of a

ccep

tabi

lity

congruent prosody


173

As expected, WO1 (SO2O1V) was judged the most appropriate answer to a direct object question. The further orders incur a strong word order violation and obtained a scoring that is at the same level with the scoring of the optimal word order with incongruent prosody as well as the order exhibiting a strong word order violation. If we take into account the means differences that are larger than a standard error, the ranking between the three suboptimal orders with congruent prosody is (WO2|WO4)>WO3, which is exactly in line with our predictions in Table 4. In the prosodically incongruent orders, we observe lower acceptability levels for all orders; the marked word order WO4 (VSO1O2) reached a very low acceptability level which is not predicted by Table 4. 4.4. Indirect object focus None of the examined orders realizes an indirect object focus structure by placing this constituent in the preverbal position. Since the postverbal argument position can also bear focused information, WO3 (O1SVO2) is predicted to be the best among the candidates though it exhibits two weak word order violations: A-reordering and V-fronting. WO1 (SO2O1V) and WO2 (O2O1SV) incur a strong violation of XPFOCV adjacency. Between the two orders we expect an advantage for WO1 (SO2O1V), reflecting the fact that this is the canonical order. WO4 (VSO1O2) displays focus on a postverbal argument, which is possible in Georgian (recall the discussion around Fig.8 in section 3.3), but V-initiality is not licit in the context of an indirect object question. The violation profile is presented in Table 5. Table 5 Word order violations in indirect object questions O2 focus STRONG WEAK predicted

rank WO1 (SO2O1V)

XPFOCV violation

2

WO2 (O2O1SV)

XPFOCV violation

A-reordering 3

WO3 (O1SVO2)


1

WO4 (VSO1O2)


Similarly with the previous context questions, the prosodic realization is decisive for the judgments. A repeated measures analysis of variance revealed a significant main effect of prosody (F1,59=49.76, p <.001). The word order factor did not reach significance, nor did the interaction between prosody and word order.

174

Figure 12 Judgments in indirect object questions

A comparison of these results with the previous ones reveals that no order reached a high scoring. WO3 (O1SVO2) reached the highest scoring, probably due to the fact that all other orders exhibit a strong violation. However, it does not significantly differ from WO1 (SO2O1V) which has a general advantage across contexts as the canonical order. The mean of judgments for WO4 (VSO1O2) in the congruent realization is the highest scoring that this order obtained in all experimental conditions; we may speculate that right peripheral placement of the focused constituent has a positive influence on the degree of acceptability, a hypothesis that has to be tested in future research. In the non-congruent realizations, a low and flat prosody is expected on the narrowly focused indirect object in WO3 (O1SVO2), but instead a prosody with an early accent is provided. This may be the reason why acceptability decreased more than in the case of WO2 (O2O1SV) with a late accent. 4.5. Multiple constituent focus In the experiment, multiple constituent questions are combined with single pair answers. Since the question requests information about the subject and the direct object, we expect that, when either of these arguments is in the preverbal position, no violation of XPFOCV occurs. The expected ranking appears in Table 6. Table 6 Word order violations in multiple constituent questions S&O1 focus STRONG WEAK predicted rank WO1 (SO2O1V) 1 WO2 (O2O1SV) A-reordering 2 WO3 (O1SVO2) A-reordering, V-fronting 3 WO4 (VSO1O2) V-initial A-reordering 4

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1


w ord orders

z-sc

ores

of a

ccep

tabi

lity

congruent prosody


175

The obtained data is presented in Figure 13. A repeated measures analysis of variance revealed a significant main effect for prosody (F1,59=14.3, p <.001) and a significant main effect for word order (F1,59=16.87, p <.001), but no significant interaction between the two factors. Figure 12 Judgments in multiple constituent questions

In general, this context induced low judgments in comparison to the previous question types, which may be due to the fact that multiple constituent questions are less common than single constituent questions. Our hypotheses predict the acceptability rank WO1>WO2>WO3>WO4, and the obtained result partly confirms this expectation both in congruent as well as in non-congruent prosodic realizations: (WO1|WO2)>(WO3|WO4). The difference between congruent and non-congruent realizations is smaller in this context, which probably results from the general tendency of the speakers to reject these question/answer pairs. 4.6. Major findings A repeated-measures analysis of variance in the overall results revealed a significant main

effect of prosodic structure (F1,59= 220.11, p <.01), a significant main effect of word order (F1,59= 64.31, p <.001), and a significant main effect of context (F1,59= 6.95, p <.01). The following sections outline the major generalizations that are supported by our data set.

A. May a congruent prosody accommodate word order markedness? Our hypothesis concerning the interaction between prosody and syntax is that a marked prosody that is contextually licensed may override the negative effect of weak word order violations such as A-reordering and V-fronting, but not the negative effect of strong word order violations, which are exemplified through V-initiality and violations of XFOCV

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1


w ord orders

z-sc

ores

of a

ccep

tabi

lity

congruent prosody


176

adjacency in our data set. In the discussion of the previous sections, we observed that this hypothesis explains several aspects of the obtained contrasts. We take the expressions that do not display any violation at all to provide a baseline that shows the average speakers’ reaction, when the question/answer pair is completely felicitous. This baseline is given by the WO1 (SO2O1V) in the ‘all-new’ and the ‘direct object focus’ contexts; the means of these conditions are given in the left column of Table 7. WO2 (O2O1SV) and WO3 (O1SVO2) incur weak violations of word order constraints (A-reordering the former, A-reordering and V-fronting the latter), which have a negative effect on the judgments, as shown in the ‘all-new’ contexts (line 1 in the middle column of Table 7 presents the mean of both orders). However, when these orders are presented with a marked prosody that is contextually licensed, the effects of word order markedness disappear. This case is exemplified by WO2 (O2O1SV) in subject questions and by WO3 (O1SVO2) in subject and indirect object questions, in which these orders do not incur any strong violation (line 2 in the middle column of Table 7 presents the mean of these three experimental conditions). In contrast to weak word order violations, strong word order violations cannot be accommodated by prosody. This contrast is illustrated by the speakers’ reactions to WO4 (VSO1O2) (the right column of Table 7 presents the mean of judgments for WO4: line 1 for the ‘all-new’ contexts with a neutral prosody, and line 2 for the single constituent questions in which this order is realized with a marked congruent prosody).4 A repeated measures analysis of variance on the two factors presented in Table 7 revealed a significant main effect for word order violations (F1,59=43.91, p <.001). The factor prosodic markedness does not have a significant main effect, which is expected since both neutral and marked prosodic realizations are congruent to the context and are not expected to differ in acceptability. The interaction between the two factors is not significant. Taken that the acceptability of the canonical word order in all-new contexts is a baseline for the estimation of felicitous question/answer pairs, we observe that the mean of the canonical word order with a contextually licensed marked prosody does not significantly differs. The critical evidence for our hypothesis relates to the expressions with a weak word order violation: the acceptability of these orders differs significantly from the baseline (t59 = 4.81, p < 0.001), when these orders are presented in a neutral context with neutral prosody, but it does not differ significantly when they are presented with a marked prosody that is licensed by the context (and provided that they do not incur a XPFOCV violation). In contrast, the speakers’ reaction to strong word order violations is significantly lower than the baseline, both in all-new contexts with unmarked prosody (t59 = 5.15, p < 0.001), as well as in the context of constituent questions with congruent prosody (t59 = 5.77, p < 0.001). Finally, the difference between the means of strong word order violations with either neutral or marked prosody did not reach significance.

4 The data from multiple constituent questions are ignored in these measurements, since they involve an overall negative effect that depends on this type of context.

177

Table 5 Prosodic markedness and word order constraints no violation weak violation strong violation means SE means SE means SE neutral prosody 0.69 ±0.09 0.24 ±0.09 -0.06 ±0.15 marked prosody 0.63 ±0.07 0.66 ±0.05 0.12 ±0.07 This data confirms our hypothesis that a marked prosodic structure that is congruent with the context may accommodate the negative effect of weak word order violations on contextual felicity but not the negative effect of strong word order violations.

B. Do prosodic and syntactic infelicities have a cumulative effect? The effects of prosodic structure are very consistent across conditions. Speakers judged incongruent intonation on average 0.88 points lower in the 1-5 scale (non-normalized values). Highly significant main effects for prosodic congruence were obtained in all contexts. We observed that the interaction between prosody and word order did not reach significance in subject questions (see 4.2), indirect object questions (see 4.4), and multiple constituent questions (see 4.5). The absence of interaction reflects the fact that non-congruent prosody had an additive negative effect of more or less the same size in most cases. The interaction of word order and prosody in the entire data set is obtained from the comparison of the three levels of word order violations that we assumed in the previous section: (a) no violation: WO1 (SO2O1V) in the ‘all-new’ and the ‘direct object focus’, (b) weak word order violation, containing WO2 (O2O1SV) and WO3 (O1SVO2) in the contexts in which these orders only incur weak word order violations, and (c) strong word order violation, containing WO4 (VSO1O2) as well as the further orders in contexts in which they incur an XPFOCV violation. These means are summarized in Table 8. A repeated measures analysis of variance reveals a significant main effect for prosody (F1,59=77.03, p <.001) and a significant main effect for word order constraints (F1,59=74.14, p <.001). The interaction between prosody and word order is not significant, which is in line with the observation that a non-congruent prosody has a consistent additive effect (of 0.6 points in the normalized data) to the felicity of the syntactic properties of the expression at issue. Table 8 Prosodic congruence and word order constraints no violation weak violation strong violation means SE means SE means SE congruent prosody 0.66 ±0.06 0.65 ±0.05 0.21 ±0.04 incongruent prosody -0.01 ±0.07 -0.02 ±0.06 -0.48 ±0.05

C. Is there an asymmetry in the acceptability of prenuclear and postnuclear accents? The negative effect of a non-congruent prosodic realization may differ depending on accent placement. Based on previous studies (see 3.1), we hypothesized that a late non-congruent pitch accent may have a stronger negative effect on the acceptability judgments that an early non-congruent accent. Two non-congruent answers in each question display

178

an earlier prosodically prominent constituent than expected, and two answers display a prosodically prominent constituent later than expected. Table summarizes the differences between congruent and non-congruent prosodic realizations in single constituent questions. In line with our hypothesis, the size of the negative effect of realizations with a late accent is greater than the negative effect of an early accent in the context of subject questions. However, in the other contexts, our hypothesis is not borne out, since the smallest negative effects are obtained in realizations with late accents (O1SVO2 in direct object questions and O2O1SV in indirect object questions). Hence, accent placement does not seem to play a noticeable role in our data set, probably due to the fact that prosodic prominence was realized with particular clarity in the stimulus material. We assume that speakers easily recognized the prosodically prominent part of the utterance and their judgments reflect their negative reaction to the mismatch without being particularly influenced by the properties of the accent placement. Table 9 Difference between congruent and non-congruent prosodic realizations subject questions dir. object

questions ind. object questions

order difference

order difference

order difference

early accent O2O1SV 0.69 SO2O1

V 0.45 O1SVO2 0.75

O1SVO2 0.61 O2O1SV

0.61 VSO1O2 0.83

late accent SO2O1V 0.87 O1SVO2

0.4 SO2O1V 0.84

VSO1O2 0.79 VSO1O2

0.9 O2O1SV 0.55

D. Does case inversion interact with information structure? The results from the manipulation of case marking show no interaction between case inversion and information structure. As mentioned above, all our items were implemented in two different aspects in order to counterbalance effects of the different case marking patterns on information structure (see example 10 in section 3.2). The background of this decision in the experimental design is that it is pointed out that case marking in Georgian corresponds to differences in the discourse properties of agents and patients (see Asatiani 2007a; Harris 1985: 295-300). In particular, the nominative marking of patients in the perfect tense reflects the fact that this constituent is “discourse prominent” in this aspect/mood (note that Georgian perfects have an evidential function). We assume that the discourse prominent constituent is the constituent the answer is about. In the question/answer pairs that we examined, this assumption implies that the answer is more felicitous when the given argument, i.e., the argument introduced in the question, bears nominative case. In subject focus questions, the nominative marked direct object (perfect tense) should be more felicitous than the dative marked direct object (future tense). This prediction is not borne out: a repeated measures analysis of variance revealed that there is no significant main effect of case marking in this context (p <.92; word order

179

has a significant main effect, see section 4.3). The opposite prediction may be tested in direct object questions. In this context, the subject is expected to obtain higher scorings when it bears nominative case, i.e., in the future tense. A repeated measures analysis of variance revealed again that there is no significant main effect of case marking (p <.17; word order has a significant main effect, see section 4.4). Hence, our data shows that both nominative and dative arguments may serve as equally felicitous hosts for either a focused or a backgrounded constituent. This result implies that case marking is not sensitive to information structure in question/answer pairs. However, this finding should not be over-generalized. Case marking depends on the tense properties of the examined Georgian verbs. Since the answer is expected to have the same tense properties with the question, it is not expected that the preference for a particular case marking of discourse prominent arguments will affect the choice of verb inflection in a question/answer pair. This conclusion does not exclude other types of possible interaction between tense and discourse state of the arguments that are not tested in this context: e.g., inverse marking occurs in the perfect tenses, which have evidential function in Georgian, and it may be that epistemic uses of verbs occur less frequently with an agent constituent (see Harris 1981; Blevins 2005). Future research will hopefully examine such correlations that may provide empirical evidence for the relation between case inversion and information structure. 5. Conclusion This paper has presented the results of a perception experiment investigating the interaction of word order and intonation in Georgian, a language with a rich tonal structure and free word order. Most of the hypotheses, formulated on the basis of what is known about this language, could be confirmed. We examined four major phenomena of Georgian word order: V-fronting, argument reordering, left adjacency of focused XPs to the verb, and V-movement to the sentence initial position. Based on the observations in previous works (Apridonidze, 1986; Asatiani, 2007a; Harris, 1981, 1993; McGinnis, 1997; Skopeteas and Fanselow, 2008b; Tuite, 1998), we assumed that the two former operations are ‘optional’ in the sense that they are chosen in order to satisfy discourse related preferences that affect the linearization, but do not display the properties of movement that targets particular positions that are associated with discrete semantic or pragmatic functions. The latter two operations are restrictive: a violation of the preference for left adjacency of focused XPs to the verb results in loss of acceptability, and placement of the verb in the sentence initial position is contextually restricted to a particular type of contexts. Our experimental evidence has shown that violations of the former type are completely ignored when these expressions are presented with a marked prosodic structure licensed by the context. Violations of the latter type result in loss of acceptability that cannot be overridden by prosody. Furthermore, the experimental results have shown that a non-congruent prosody has a very robust negative effect on acceptability. Prosodic infelicities had an additive effect to

180

word order infelicities, showing that the felicity of an expression in the context is the product of its syntactic and prosodic properties. To conclude, previous studies have shown that prosodic prominence on a constituent is a strong indicator for the focus properties of the utterance in intonational languages. Whenever the prosodic properties do not correspond the contextual expectations, a clear negative reaction of the speakers is induced. In languages with free word order, word order is sensitive to information structure too, but word order infelicities may be overridden by an appropriate prosodic structure. This observation may lead to the conclusion that prosodic constraints outrank syntactic constraints in the encoding of information structure. Glosses 2 2nd person 3 3rd person ADV adverbializer AOR aorist CAUS causative DAT dative ERG ergative FUT future GEN genitive INVO inverted object INVS inverted subject IO indirect object (person affix) NOM nominative PL plural PRF perfect PRS present PR preverb PV preradical vowel S subject (person affix) SG singular THM thematic suffix Acknowledgments This paper is a product of the project “Typology of Information Structure,” which is part of the SFB 632 “Information Structure” at the University of Potsdam and Humboldt University Berlin (sponsored by the DFG). An important part of the observations about word order come from common work with Gisbert Fanselow on Georgian syntax. During the preparation of the final version of the paper we profited from the substantial comments of three anonymous reviewers. Special thanks are due to Shorena Bartaia, Anna Tsutkerashvili, and Nutsa Tsereteli, who contributed to the development and execution of our experiments and Kirsten Bock for her comments on a pre-final draft. Many thanks also to Ani Asatiani, Nana Chidrashvili, and Natja Dundua for their assistance in the execution of the experiments in Tbilisi.

181

References Alkhazishvili, A., 1959. porjadok slov i intonacija v prostom povestvovateljnom

predlojenii gruzinskogo jazyka [Word order and intonation in simple extended sentences in Georgian]. Phonetics (Moscow) I, 367-414.

Alter, K., Mleinek, I., Umbach, C., Rohe, T., 2001. Kontrastprosodie in Sprachproduktion und -perzeption. In: Steube, A., Umbach, C. (Eds.), Kontrast: lexikalisch, semantisch, phonologisch. Linguistische Arbeitsberichte 77. Institut für Linguistik, Universität Leipzig, pp. 59-81.

Anderson, S. R., 1984. On representations in morphology: Case, agreement, and inversion in Georgian. Natural Language and Linguistic Theory 2, 157-218.

Apridonidze, Sh., 1986. sit’q’vatganlageba axal kartulši [Word order in modern Georgian]. Mecniereba, Tbilisi.

Aronson, H. I., 1982. Georgian: A Reading Grammar. Slavica, Chicago. Asatiani, R., 1982. mart’ivi c’inadadebis t’ip’ologiuri analizi (tanamedrove kartuli

salit’erat’uro enis masalaze [The typological analysis of a simple sentence (On the data of modern literary Georgian)]. Mecniereba, Tbilisi.

Asatiani, R., 2007a. The main devices of foregrounding in the information structure of Georgian sentences. In: Ten Cate, B., Zeevat, H. (Eds.), Proceedings of the Tbilisi Symposium on Language, Logic and Computation 2005. Springer, Amsterdam, pp. 21-30.

Asatiani, R., 2007b. Georgian. In: Skopeteas, S., Hellmuth, S., Fanselow, G., Féry, C., (Eds.), The expression of information structure: The interaction of syntax and phonology in cross-linguistic perspective. Mouton De Gruyter, Berlin.

Boeder, W., 1989, Verbal person marking, noun phrase and word order in Georgian. In: Marácz, L., Muysken, P. (Eds.), Configurationality: The Typology of Asymmetries. Foris, Dordrecht, pp. 159-184.

Boeder, W., 2005, The South Caucasian languages. Lingua 115, 5-89. Bush, R., 1999. Georgian yes–no question intonation. Phonology at Santa Cruz, Vol. 6.

UC Santa Cruz, Santa Cruz, CA, pp. 1–11. Cowart, W., 1997. Experimental Syntax: Applying Objective Methods to Sentence

Judgments. Sage Publications, Thousand Oaks, CA. Davitiani, A., 1973. kartuli enis sint’aksi [Syntax of the Georgian language]. Ganatleba,

Tbilisi. Féry, C., Samek-Lodovici, V., 2006. Focus projection and prosodic prominence in nested

foci. Language 82.1, 131-150. Féry, C., Skopeteas, S., 2008, Prosodic structure of Georgian: Evidence from language

production. Féry, C., Stoel, R., 2006. Gradient perception of intonation. In: Fanselow, G., Féry, C.,

Schlesewsky, M., Vogel, R. (Eds.), Gradience in Grammar: Generative Perspectives. Oxford University Press, Oxford, pp. 145-166.

Golston, C., 1995, Syntax outranks phonology: Evidence from Ancient Greek, Phonology 12.3, 343-368.

Grimshaw, J., 1997. Projections, heads, and optimality. Linguistic Inquiry 28, 373-422. Gussenhoven, C., 1983. Testing the reality of focus domains. Language and Speech 26,

61-80.

182

Harford, C., Demuth, K., 1999. Prosody outranks syntax: An Optimality approach to subject inversion in Bantu relatives. Linguistic Analysis 29.1-2, 47-68.

Harris, A. C., 1981. Georgian Syntax: A Study in Relational Grammar. Cambridge University Press, Cambridge.

Harris, A. C., 1985. Diachronic Syntax: The Kartvelian Case. Academic Press, New York.

Harris, A. C., 1993. Georgian. In: Jacobs, J., von Stechow, A., Sternefeld, W., Vennemann, Th. (Eds.), Syntax: An International Handbook of Contemporary Research. De Gruyter, Berlin/New York, pp. 1377-1397.

Harris, A. C., 2000. Word order harmonies and word order change in Georgian. In: Sornicola, R., Poppe, E., Haley, A. (Eds.), Stability, Variation, and Change of Word-Order Patterns over Time. Benjamins, Amsterdam/Philadelphia, pp. 133-163.

Hewitt, B. G., 1995. Georgian: A Structural Reference Grammar. Benjamins, Amsterdam/Philadelphia.

Jun, S.-A., Vicenik, Ch., Lofstedt, I. 2007. Intonational Phonology of Georgian. PowerPoint Presentation given at the Workshop on Intonational phonology: Understudied or Fieldwork languages, Saarbrücken, Germany.

Keller, F., Alexopoulou, Th., 2001. Phonology competes with syntax: Experimental evidence for the interaction of word order and accent placement in the realization of information structure. Cognition 79.3, 301-371.

Kiziria, N., 1987. salit’erat’uro kartulis int’onaciis sak’itxebi [Issues of intonation of literary Georgian]. Mecniereba, Tbilisi.

Krifka, M., 2001. For a structured meaning account of questions and answers. In: Féry, C., Sternefeld, W. (Eds.), Audiatur Vox Sapientiae: A Festschrift for Arnim von Stechow, Studia grammatica 52. Akademie Verlag, Berlin, pp. 287-319.

Kvachadze, L., 1996. tanamedrove kartuli enis sint’aksi [ Syntax of the contemporary Georgian language]. (the fourth revised edition). Rubikoni, Tbilisi.

McGinnis, M., 1997. Case and locality in L-Syntax: Evidence from Georgian. In: Harley, H. (Ed.), The UPenn/MIT Roundtable on Argument Structure and Aspect. MITWPL 32.

McGinnis, Martha 1999, A-scrambling exists! In: Michelle Minnick & Na-Rae Han (eds.), Proceedings of the 23th Annual Penn Linguistics Colloquium. Penn Working Papers in Linguistics. Department of Linguistics, University of Pennsylvania,

McGinnis, Martha 2004, Lethal ambiguity. Linguistic Inquiry 35.1, 47-95. Müller, G. 2007. Frageintonation im Georgischen. Master’s thesis. University of

Cologne. Nash, L., 1995. Portée argumentale at marquage casuel dans les langues SOV et dans les

langues ergatives: l’exemple du géorgien. Ph.D., Université de Paris VIII. Pochkhua, B., 1962. sit’q’vata rigi kartulši [Word order in Georgian]. Ibero-Caucasian

Linguistics (Tbilisi) 13, 109-125. Rietveld, T., Van Hout, R., 2005. Statistics in Language Research: Analysis of Variance.

Mouton De Gruyter, Berlin/New York. Skopeteas, S., Fanselow, G., 2008a. Effects of givenness and constraints on free word

order. To appear in: Zimmermann, M., Féry, C. (Eds.), Information Structure from Different Perspectives. Oxford University Press, Oxford.

Skopeteas, S., Fanselow, G., 2008b. Focus in Georgian and the expression of contrast. To appear in: Cook, Philippa & Repp, Sophie (Eds.), Special issue on Contrast, Lingua.

183

Skopeteas S., Fanselow, G., Asatiani, R., 2008. Case Inversion in Georgian. Ms., Univ. of Potsdam.

Skopeteas S., Féry, C., 2007. Contrastive topics in pairing answers: A cross-linguistic production study. In: Featherston, S., Sternefeld, W. (Eds.), Linguistic Evidence, Vol. 2. Mouton de Gruyter, Berlin.

Skopeteas, S., Fiedler, I., Hellmuth, S., Schwarz, A., Stoel, R., Fanselow, G., Féry, C., Krifka, M., 2006. Questionnaire on Information Structure (QUIS). Interdisciplinary Studies on Information Structure 4. Working Papers of the SFB 632, Universitätsverlag Potsdam, Potsdam.

Tevdoradze, I., 1978. kartuli enis p’rosodiis sak’itxebi [Issues of prosody of the Georgian language]. Tbilisi State University Press, Tbilisi.

Tuite, K., 1998. Kartvelian Morphosyntax: Number, Agreement, and Morphosyntactic Orientation in the South Caucasian Languages. Lincom Europa, Munich.

Vogt, H., 1971. Grammaire de la langue géorgienne. Universitaetsvorlaget, Oslo. Zec, D., Inkelas, Sh. 1990, Prosodically Constrained Syntax. In: Inkelas, S., Zec, D.

(Eds.), The Phonology-Syntax Connection, Center for the Study of Language and Information, The University of Chicago Press, Chicago, pp. 365-378.

Zhghenti, S., 1963. kartuli enis rit’mik’ul-melodik’uri st’ruk’t’ura [Rhythmic and melodic structure of the Georgian language]. Codna, Tbilisi.

Appendix: Means of judgments and standard scores

focus prosody order judgments standard scores

mean mean SE5 SD all congruent SO2O1V 4.5 0.69 0.09 0.75 O2O1SV 4 0.31 0.12 0.93 O1SVO2 3.7 0.16 0.12 0.93 VSO1O2 3.4 -0.06 0.15 1.18

non-congruent SO2O1V 3.1 -0.18 0.11 0.89

O2O1SV 2.7 -0.46 0.11 0.81 O1SVO2 2.7 -0.48 0.11 0.91 VSO1O2 2.8 -0.46 0.11 0.85 subject congruent SO2O1V 4.2 0.55 0.08 0.62 O2O1SV 4.6 0.81 0.06 0.48 O1SVO2 4.4 0.66 0.06 0.49 VSO1O2 3.4 0.01 0.09 0.74

non-congruent SO2O1V 3 -0.32 0.09 0.71

O2O1SV 3.6 0.12 0.08 0.66 O1SVO2 3.4 0.05 0.09 0.71 VSO1O2 2.2 -0.78 0.08 0.65

5 SE = standard error; SD = standard deviation.

184

direct object congruent SO2O1V 4.4 0.63 0.07 0.53 O2O1SV 3.8 0.23 0.07 0.58 O1SVO2 3.4 0.03 0.09 0.72 VSO1O2 3.7 0.14 0.09 0.75

non-congruent SO2O1V 3.6 0.18 0.09 0.75

O2O1SV 2.9 -0.38 0.09 0.71 O1SVO2 2.8 -0.37 0.08 0.69 VSO1O2 2.3 -0.76 0.08 0.68 indirect object congruent SO2O1V 4.1 0.43 0.07 0.59 O2O1SV 3.9 0.32 0.08 0.65 O1SVO2 4.2 0.49 0.09 0.76 VSO1O2 3.8 0.21 0.09 0.74


O2O1SV 3 -0.23 0.11 0.82 O1SVO2 3 -0.26 0.11 0.87 VSO1O2 2.5 -0.62 0.09 0.72 multiple congruent SO2O1V 3.7 0.19 0.09 0.73 O2O1SV 3.8 0.26 0.08 0.67 O1SVO2 3.4 -0.01 0.08 0.68 VSO1O2 3.3 -0.09 0.11 0.81


O2O1SV 3.5 0.01 0.09 0.76 O1SVO2 2.8 -0.35 0.11 0.77 VSO1O2 2.9 -0.33 0.09 0.73

185

STAVROS SKOPETEAS, GISBERT FANSELO, RUSUDAN ASATIANI

CASE INVERSION IN GEORGIAN:

SYNTACTIC PROPERTIES AND SENTENCE PROCESSING*

1. Preliminaries That the human sentence processing mechanism makes use of all available cues for gram-matical function assignment and structure building in incremental sentence processing is a very natural assumption that has been confirmed in a number of experimental studies. Bates, McWhinney, and Kliegl (1974) were among the first showing this by demonstrating that overt case marking, morphological agreement information and syntactic position are used to different degrees in different languages in a sentence interpretation experiment. Early online-studies concerned with the effect of explicit morphological marking used morphological information that is (relatively) unambiguous in nature. Thus, Krems (1984) found increased reading times for German sentences beginning with an NP unambiguously marked for accusative case that almost always marks direct objects only, a result later confirmed by Hemforth (1993) and Fanselow, Kliegl, and Schlesewsky (1999), see also Felser, Clahsen, and Münthe (2003), Fiebach, Schlesewsky, and Friederici (2001) and Matzke et al. (2002) for related ERP-based results. An unambiguous case marking need not, however, indicate the grammatical function and/or structural position of the NP bearing it in an unambiguous way. While subjects are marked with nominative case, and indirect (second) objects with dative case in a language like German, the correlation is not a biunique one, since nominative and dative NPs occur in the position of direct (first) objects, too – the former do so in passive and unaccusative contexts, the latter do so when they instantiate idiosyncratic case frames of a number of verbs such as helfen ‘help’. The question arises whether (and if so, how) explicit morphological information that is syntactically ambiguous is used in online sentence processing. Recent results of Bornkessel, Schlesewsky, and Friederici (2002, 2003) suggest that such information is put to use, and in a way much reminiscent of ambiguous structural information: morphological cases are linked to interpretation preferences that can be

* Thanks go to Nutsa Tsereteli and Shorena Bartaia, who collaborated in the development of the experimental stimuli and the performance of the experiments as well as to Caroline Féry for her advice during the development of the experiments, and to Ina Bornkessel-Schlesewsky, Robin Hörnig, Matthias Schlesewsky, Barbara Stiebels, and Shravan Vasishth for stimulating discussions. We are grateful to Shorena Bartaia, Tamar Khisanishvili, and Tamar Kvaskhvadze that contributed their intuitions about the grammaticality of Georgian sentences. Many thanks also to Ani Asatiani, Nana Chidrashvili, and Natja Dundua for their assistance to the performance of the experiments in Tbilisi. This paper evolved within the project D2 ‘Typology of Information Structure’ which is part of the SFB 632 ‘Information Structure’ at the University of Potsdam/Humboldt University Berlin (funded by the DFG).

186

expressed in various ways, either as inviting hypothesis concerning the relative thematic position of the NP arguments (such that nominative NPs are preferentially interpreted as the highest argument of a predicate) as in the proposal of Bornkessel et al., or as preferences for the location of the NPs in the structural representation of the clause (such that datives trigger a default placement into the indirect object position (the specifier of VP in certain recent syntactic models, see e.g. Chomsky 2003), while nominatives are preferentially located in the subject position (the specifier of vP in such models). As suggested by the findings of Bates et al. (1982), languages seem to make use of morphologically explicit but syntactically ambiguous case information to different degrees. Röhm, Schlesewsky, and Bornkessel-Schlesewsky (2007) present electrophysiological evidence for the claim that morphological case plays a very minor role in the online processing of Icelandic, a language with rigid constituent order and a rich morphological case system showing very little correspondence between case and grammatical function. This is on contrast with the apparently strong impact of morphological case in German, in which there is at least a system of unidirectional correspondences between case and grammatical function (if subject then nominative, if indirect object then dative, if accusative then direct object). In addition, German differs from Icelandic in being a free constituent order language (with an SOV rather than SVO default order). In the present paper, we will consider the influence of explicit case marking on sentence processing in Georgian. Georgian may help us disentangle the reasons for the different treatment of case in German and Icelandic. Just like German, Georgian is a free word order language, and, arguably, it even shares the SOV base order with German. Thus, if the availability of scrambling is the factor that differentiates German from Icelandic with respect to the role of case in online processing, Georgian and German should behave alike in the latter domain as well. On the other hand, there is no overwhelming correspondence between explicit case and grammatical function in Georgian. In the verbal domain, we find nominative and ergative case, while dative and accusative are morphologically identical (but they may trigger different 3rd person agreement on the verb). Structural subjects may appear as ergative, nominative, or dative-accusative phrases, direct objects may bear dative-accusative or nominative, while indirect objects seem to be confined dative-accusative case. In turn, the ergative is the only case linked to grammatical function (if ergative, then subject). Given that the ergative is a peripheral case (it shows up with the aorist only), and that indirect objects are a rare category, one can argue that Georgian is closer to Icelandic that to German with respect to the grammatical interpretation of case, which may trigger parallels in case processing if that aspect of the morphology-syntax correspondence is relevant. In order to assess the role of explicit case information in the processing of Georgian, we carried out two reaction time experiments (see section 3). In particular, we examine the role of case inversion which is exemplified in (1). Depending on the inflectional properties of the verb (conjugation class and tense), arguments in Georgian show two different case marking patterns (among others): in the direct pattern, which is illustrated in (1a), the nominative marked constituent bears the θ-role of an actor and the theme

187

constituent bears an oblique case, namely the dative; in the inverse pattern, which is illustrated in (1b), the actor constituent bears the dative case, while the nominative argument is the theme. (1) (a) ǯarisk’ac-i monadire-s da- č’r-i-s.

soldier-(NOM) hunter-DAT PR(FUT)-(IO.3)cut-THM-S.3.SG ‘The soldier will wound the hunter.’ (b) ǯarisk’ac-s monadire da-u- č’r-i-a. soldier-DAT hunter(NOM) PR(PRF)-(INV.S.3)PV-cut-PF-INV.O.3(INV.S.SG) ‘The soldier has wounded the hunter.’ Before we can describe the experimental findings, a few remarks on the case system of Georgian are in order (see section 2). 2. Case system of Georgian 2.1. Basic facts In interaction with aspectual/modal categories, Georgian verb (conjugation) classes license three different case marking patterns which are presented in Table 1 (see Harris 1981:1, Blevins 2005). The roles indicated in this table should be understood as abstractions out of the θ-roles (‘proto-roles’ in the sense of Dowty 1991, or ‘macroroles’ in the sense of van Valin and LaPolla 1997). An ‘actor’ stands for the highest participant involved in the event, which may be an agent or an experiencer depending on the event at issue; the ‘goal’ encompasses recipients and beneficiaries; the ‘theme’ is the participant which is affected or effected through the event or the (non-agentive) stimulus of an experiential event. The case affix labeled ‘dative’ is the product of case syncretism between accusative and dative (this affix is labeled ‘dative’ following the Georgian grammatical tradition, see Tschenkéli 1958, Harris 1981; Hewitt 1995 among others). Synchronically, this affix shows mixed properties as it will be exemplified below: when it is licensed by thematic properties, it is used for the marking of goals and non-volitional actors, which is a function reminiscent of the dative in familiar languages; when it is structurally licensed, it marks the lower argument, hence having the properties of an accusative. The case affix labeled ‘ergative’ (also called ‘narrative’ in Georgian grammar; see Harris 1990, 1993; Hewitt 1987) marks actors.

Table 1. Proto-roles and case marking patterns (to be revised) case marking pattern actor undergoer goal

A ergative nominative dative B nominative dative dative C dative nominative (postpositional)

The case marked arguments are cross-referenced through person affixes of the verb. In the patterns A and B, each of the three arguments is cross-referenced by a distinct set of affixes. In pattern C, the nominative argument corresponds to the affix set that marks actors in the other patterns, and the dative argument to the affix set that marks goal

188

constituents in the other patterns, whereby the postpositional goal is not cross-referenced at all, which shows that it has adjunct status.1 Case-marking patterns are determined by the verb (conjugation) class in interaction with its tense/aspect/mood properties. Following the Georgian grammatical tradition, inflectional categories are classified in three series (or ‘screeves’): Series I: present (indicative/subjunctive), future (indicative/subjunctive), imperfect, conditional; Series II: aorist indicative, optative; and Series III: present perfect, pluperfect. Georgian verbs are divided in four conjugation classes which – in interaction with the series – determine different case marking patterns for the clausal arguments as it is presented in Table 2. Class membership is not fully predictable by the semantic properties of the verb (see Aronson 1989), but some rough correlations with syntactic/semantic properties are identifiable (see Harris 1981:228ff.): class 1 mainly contains active transitive verbs and a large class of derived causatives; class 2 contains unaccusatives, analytic passives which are formed with the class 2 verb ikneba, as well as synthetic passives with change of state semantics; class 3 contains unergative verbs and class 4 mainly contains experiencer verbs.

Table 2. Case marking in classes and series (see Harris 1981:2, 118) series I series II series III

classes 1&3 B A C class 2 B class 4 C

The following examples illustrate the facts about case marking. Example (2) illustrates the case patterns of a class 1 verb. The present form (series I) in (2a) licenses case marking pattern B, the aorist form (series II) in (2b) licenses pattern A and the perfect form (series III) in (2c) licenses pattern C. (2) (a) Class 1, pattern B dato nino-s c’ign-s a-čuk-eb-s.

Dato(NOM) Nino-DAT book-DAT (IO.3)PV-donate-THM-S.3.SG ‘Dato donates/will donate a book to Nino.’ (b) Class 1, pattern A dato-m nino-s c’ign-i a-čuk-a.

Dato-ERG Nino-DAT book-NOM (IO.3)PV-donate-AOR.S.3.SG ‘Dato donated a book to Nino.’ (c) Class 1, pattern C dato-s nino-s-tvis c’ign-i u-čuk-eb-i-a.

Dato-DAT Nino-GEN-for book-NOM (INV.S.3)PV-donate-THM-PF- INV.O.3(INV.S.SG)

‘(Apparently) Dato has given a book to Nino.’

1 There three sets of affixes display many cases of overlap, resulting in cases of conflict whenever more than one argument markers compete for a single available slot. Without proceeding into details, the resolution of these conflicts is determined by the interaction of morpho-tactic constraints and constraints derived by the person and argument hierarchies (see detailed account in Anderson 1984, Carmack 1997).

189

Two classes of verbs deviate from the basic pattern illustrated in (2), as shown in (3). These verbs license an invariable case marking pattern which is not influenced by the tense properties of the verb. Examples (3a-b) illustrate the syntactic behaviour of a class 2 verb which licenses pattern B and a class 4 verb which licenses pattern C. (3) (a) Class 2, pattern B dato nino-s elodeba/daeloda/dalodebia

Dato-(NOM) Nino-DAT (IO.3)wait(S.3.SG) (present/aorist/perfect) ‘Dato waits/waited/has waited for Nino.’ (b) Class 4, pattern C dato-s nino sʒuls/sʒulda/sʒulebia.

Dato-DAT Nino(NOM) (IO.3)hate(S.3.SG) (present/past/perfect) ‘Dato hates/hated/has hated Nino.’ This data shows that the only Georgian case that is uni-directionally associated with a particular role is the ergative. Nominative and dative affixes are mapped to different arguments which suggests that they are only probabilistic predictors for argument structure. 2.2. Inherent vs. structural cases In line with recent accounts on Georgian case marking (Joppen-Hellwig 2001, Blevins 2005), we assume two distinct layers of case licensing, a non-structural and a structural one. The non-structural layer encompasses cases that are either determined by the role properties of the argument (inherent case), or are idiosyncratically stipulated by particular verbal heads through lexical specification (lexical case) (see Chomsky 1981, Fanselow 2000, Woolford 2006, Butt and King 2005 among others). The structural layer applies to the set of eligible arguments that did not received case through a lexical or thematic rule. At this layer, case is licensed in a hierarchical manner, such that the highest eligible argument receives the highest available case (following the hierarchy nominative > accusative > dative) (see Wunderlich 1997, Stiebels 2000). Depending on the theoretical framework, argument hierarchy is either determined by the embeddedness in the syntactic configuration or by the hierarchy of theta roles (see a comparison of the alternative accounts in Fanselow 2000). The case patterns in Georgian (see Table 1) may be accounted for, if we assume the layers of case licensing as given in Table 3 (compare previous accounts in Joppen-Hellwig 2001, Blevins 2005). Non-structural cases are specified either through lexical specification of the verbal head (this holds for dativeλ in the B2 pattern) or through the thematic properties of the arguments (see dativeθ in all other cases). The case marking of the remaining arguments is subject to structural licensing (marked as s in Table 3). Structural licensing takes place in a hierarchical manner, such that the highest case (nominative > dative) marks the highest eligible argument (actor > undergoer). Furthermore, Table 3 is a refinement of Table 1, since it contains also some additional information about the classes of verbs that do not occur with all three arguments.

190

Table 3. Layers of case licensing* case marking pattern actor undergoer goal

A1 (class 1, series II) ergativeθ sH dativeθ A2 (class 3, series II) ergativeθ - dativeθ B1 (class 1, series I) sH sL dativeθ B2 (class 3, series I) sH - dativeθ B3 (class 2) sH dativeλ (postpositional)θ C1 (class 1/3, series III) dativeθ sH (postpositional)θ C2 (class 4) dativeθ sH -

* xθ = inherent case; xλ = lexical case; sH = structural case, higher in the argument hierarchy; sL = structural case, lower in the argument hierarchy Evidence for the assumptions in Table 3 comes from three types of phenomena that are discussed in the following sections: (a) thematic relatedness of the non-structural cases (see 2.2.1), (b) eligibility for A-movement (see 2.2.2), and (c) stratal uniqueness effects (see 2.2.3). 2.2.1. Thematic relatedness Cases that are licensed by a thematic rule are restricted to a particular range of thematic relations to the verbal head. This is quite straightforward for the marking of goal (recipient and beneficiary) arguments in patterns A and B1/2. That the licensing condition goal → dativeθ fails to apply for pattern C1 is an effect of stratal uniqueness that is discussed in section 2.2.3. The case affix labeled ‘ergative’ is not a structural case in Georgian. It is thematically related to the actor role, a fact that is extensively discussed by Harris (1990). As already mentioned in section 2.1, conjugation class 3 contains unergatives verbs, while conjugation class 2 contains unaccusatives verbs. Consequently, licensing the ergative affix in series II depends on the thematic properties of the argument: it marks actors of either transitive (class 1) or intransitive (class 3) verbs, but not undergoers of either transitive (class 1) or intransitive (class 2) verbs. The dativeθ in pattern C occurs with two types of arguments. First, it marks actors of experiencer verbs in class 4. Actors of these verbs do not exercise volitional control on the event. Evidence for the non-volitionality of experiencers is provided through distributional diagnostics: the occurrence of an adverb that requires the volitional involvement of the actor is not acceptable with class 4 verbs, see (4). (4) #dato-s nino ganzr-ax s-ʒul-s.

Dato-DAT Nino(NOM) purpose-on INV.S.3-hate-INV.O.3(INV.S.SG) ‘Dato hates Nino on purpose.’ Furthermore, the dativeθ in pattern C marks actors of transitive verbs in series III (perfect tense). The perfect in Georgian involves stative aspect and has epistemic properties: the speaker asserts that he has not direct evidence that the conveyed proposition took place. These aspectual/modal properties do not exclude volitional actors, as shown through the grammaticality of example (5) that involves a Class 1 verb in the perfect tense and an adverb asserting the volitional involvement of the actor constituent. However, native speaker intuitions suggest that utterances in the perfect tense involve an assertion about the truth value of a particular proposition letting the volitional or non-volitional

191

involvement of the speaker unspecified (see Harris 1981 and Asatiani and Ivanishvili 2007 for further discussion). Some distributional evidence comes from the interaction with negation. Negative facts in the past are encoded in the perfect, while positive facts are encoded in the aorist (see Joppen-Hellwig 2001: 142). (5) dato-s nino-s-tvis c’ign-i ganzr-ax u-čuk-eb-i-a.

Dato-DAT Nino-GEN-for book-NOM purpose-on (INV.S.3)PV-donate- THM-PF-INV.O.3(INV.S.SG)

‘(Apparently) Dato has given Nino a book on purpose.’ It should be noted that the thematic range of dative actors of perfect tense verbs is identical to the range of nominative actors of present/future verbs in class 1. The thematic relatedness of perfect actors does not imply a restriction to the external arguments that occur in this position, but it refers to the interpretative properties of utterances in the perfect and to their implications for the involvement of the actor constituent. The dativeλ in pattern B3 is a lexical case, i.e. it is lexically determined by the verbal head. Thematically, these arguments are undergoers which is in line with the view that case licensing by the verbal head can only apply within the VP proper (see Fanselow 2000, Woolford 2006). That the undergoer-dative in pattern B3 (class 2 verbs) is licensed differently than the undergoer-dative in pattern B1 (class 1 and 3 verbs) is supported by evidence from eligibility for A-movement and stratal uniqueness effects that are discussed in the next sections. 2.2.2. Eligibility for A-movement Structural case is licensed by the syntactic configuration, i.e. either through the verb in case of internal arguments or through some functional head such as tense or agreement in case of external arguments (see Haider 1999). By consequence, structural case marking is expected to not be preserved under those types of A-movement that affect the syntactic function of the arguments, e.g. passive or raising. In the following, we discuss data from the passivization of class B1 and class B2.2 It is worth noting that only undergoer-datives and not goal-datives are eligible for A-movement of the passive type which is further evidence for the view that the former but not the latter arguments are structurally case marked (see McGinnis 1998a, 1998b). Passivization of Class 1 verbs (case marking pattern B1) is illustrated in (6). The undergoer-dative in (6a) is not preserved under A-movement (passivization) in (6b) which supports the view that case marking is structurally licensed in this class: the highest argument in the configuration receives nominative case (see also Tuite 1987).

2 The case preservation criterion does not apply for pattern A and pattern C themes. This constituent is the highest non-inherently marked argument in either voice, hence it always receives nominative case due to its structural prominence and not due to case preservation (see examples in Tuite 1987: 297).

192

(6) Class 1, series I (a) Nino Vano-s č’ri-s.

Nino(NOM) Vano-DAT wound-S.3.SG ‘Nino wounds Vano.’ (b) Vano i-č’r-eb-a (Nino-s mier).

Vano(NOM) PASS-wound-THM-S.3.SG Nino-GEN by ‘Vano is wounded (by Nino).’ In contrast to class 1 verbs, class 2 verbs (case marking pattern B2) do not allow for A-movement of the passive type, as illustrated in (7) and (8). (7) (a) Nino Vano-s st’umr-ob-s.

Nino(NOM) Vano-DAT (IO.3)visit-THM-S.3.SG ‘Nino is visiting Vano.’ (b) *Vano Nino-s mier i-st’umr-eb-a.

Vano(NOM) Nino-GEN by PV-visit-THM-PASS-S.3.SG ‘Vano is visited by Nino’. (8) (a) Nino Vano-s e-čxub-eb-a.

Nino(NOM) Vano-DAT PV-(IO.3)fight-THM-S.3.SG ‘Nino is fighting with Vano.’ (b) *Vano Ninos mier i-čxub-eb-a.

Vano-DAT Nino-GEN by PASS-fight-THM-S.3.SG ‘Vano is fighted by Nino.’ (colloquial) This data supports the view that the alleged case marking pattern B contains two markably different subtypes: subtype B1 involves undergoers that are structurally marked for dative case; subtype B3 involves undergoers that are marked for dative case through the lexical specification of the verb. 2.2.3. Stratal uniqueness effects Stratal uniqueness requires that the case-to-argument correspondences at each layer of case licensing are biunique. Restrictions to the use of the same case affix for more than one arguments are evidence that the case licensing of these arguments takes place in the same licensing layer (structural vs. non-structural). The facts in Table 1 already suggest that stratal uniqueness determines the Georgian case patterns. While two dative arguments are possible in pattern B1, they are excluded in pattern C1.This difference is straightforwardly explained by the assumption of two different layers of case licensing as presented in Table 3. Double occurrence of the dative affix is not banned, if one of them is licensed thematically and the other structurally (pattern B1), but it is not possible, when both arguments bear non-structural case (pattern C1). Previous accounts that subsume class 1 and class 2 verbs under the same case marking pattern (pattern B) do not mention the fact, that the goal argument of class 2 verbs is not dative-marked. Indeed, a benefactive constituent with these verbs is encoded through a PP and not through dative case, as exemplified in (9).

193

(9) (a) Nino Vano-s-tvis Dato-s e-lap’arak-eb-a. Nino(NOM) Vano-GEN-for Dato-DAT PR-(IO.3)speak-THM-S.3.SG

‘Vano is speaking with Dato for (the sake of) Vano.’ (b) Nino Vano-s-tvis Dato-s st’umr-ob-s.

Nino(NOM) Vano-GEN-for Dato-DAT (IO.3)visit-THM-S.3.SG ‘Vano is visiting Dato for (the sake of) Vano.’ (c) *Nino Vano-s Dato-s e-lap’arak-eb-a.

Nino(NOM) Vano-DAT Dato-DAT PV-(IO.3)speak-THM-S.3.SG (d) *Nino Vano-s Dato-s st’umrob-s.

Nino(NOM) Vano-DAT Dato-DAT PV-(IO.3)visit-THM-S.3.SG Further evidence for stratal uniqueness is provided by causative verbs (see Harris 1981). Causatives are class 1 verbs independently of the conjugation class of the base verb (see Harris 1981:132), hence the syntactic properties of causatives may not be related to the syntactic properties of class 2 and class 4 verbs. The facts from class 1 verbs in the three case marking patterns are presented in (10). (10) (a) Series I, pattern B soso dato-s nino-s-tvis c’ign-s

Soso(NOM) Dato-DAT Nino-GEN-for book-DAT a-čuk-eb-in-eb-s.

(IO.3)PV-donate-THM-CAUS-S.3.SG ‘Soso lets Dato to give Nino a book.’ (b) Series II, pattern A soso-m dato-s nino-s-tvis c’ign-i

Soso-ERG Dato-DAT Nino-GEN-for book-NOM a-čuk-eb-in-a.

PV(IO.3)-donate-CAUS-AOR.S.3.SG ‘Soso let Dato to give Nino a book.’ (c) Series III, pattern C soso-s dato-s-tvis nino-ze c’ign-i

Soso-DAT Dato-GEN-for Nino(DAT)-to book-NOM u-čuk-eb-in-eb-i-a.

PF-(INV.S.3)-donate-THM-CAUS-THM-INV.O.3(INV.S.SG) ‘(Apparently) Soso has let Dato to give Nino a book.’ Table 4 summarizes the case marking patterns presented in(10a-c) in comparison to the basic patterns of Table 3. The causee argument is marked by an inherent dative in patterns A1 and B1, as it may be shown from the fact that this case excludes the use of the dative for the goal constituent and it does not conflict with the structural dative marking of undergoers in pattern B1. Furthermore, in pattern C1 the inherent dative of the causer does not allow for a second inherent dative for the causee which is rendered through a PP instead.

194

Table 4. Case licensing of causative verbs causer actor/causee undergoer goal

A1 basic ergativeθ sH dativeθ causative ergativeθ dativeθ sH (postpositional)θ B1 basic sH sL dativeθ causative sH dativeθ sL (postpositional)θ C1 basic dativeθ sH (postpositional)θ causative dativeθ (postpositional)θ sH (postpositional)θ 2.2.4.Summary This section has elaborated the conclusion of the last section concerning the non-unidirectional mapping of cases to arguments in Georgian. It has shown that there are two distinct and ordered layers of licensing case: the first layer is non-structural and determines the case according to case marking information in the subcategorization frame of the verb. The second layer is structural and distributes case marking according to the position in the syntactic configuration to those arguments that do not already bear non-structural case. 2.3. Case marking and syntactic asymmetries The issue of the syntactic status of case in Georgian is highly controversial, since some authors argue that the nominative marked argument is the subject, while others argue that the actor argument is the subject across case marking patterns. The controversy obviously depends on what each author assumes to be a subject and will not be reported here in detail. The important point for our purposes is that there is no syntactic evidence that case marking affects argument asymmetries. In the following, we discuss evidence concerning the opposition between the direct base marking pattern (B) and the inverse case marking pattern (C), since this is the locus of the existing controversy and also the issue of our empirical study in section 3. 2.3.1. Word order Georgian is a language with flexible word order: all six possible permutations of a clause with a V, an S, and an O are possible and determined by information structural properties (see Skopeteas and Fanselow 2008a about scrambling given constituents, and 2008b about movement operations related to focus). Though there is some controversy concerning the directionality of verb projections, all accounts agree that subjects precede objects in the canonical word order. This generalization holds independently of case marking, as it is shown in the following. We illustrate this generalization by means of NPs that are incompatible with topicalization, namely negative words. The (a)-examples below show that the order {NPNOM NPDAT} is only acceptable in (11) with class 1 (series I) and class 2 verbs, i.e., when it in line with the order {actor undergoer}. The (b)-examples show that the order {NPDAT NPNOM} is acceptable in (12), i.e., when it in line with the order {actor undergoer}, which holds for class 1 (series III) and class 4 verbs.3

3 Note that the suboptimal {undergoer actor} orders are not categorically excluded. This is due to the fact that there is a restricted type of contexts that may indeed license this order. For instance, (10b) may be a felicitous answer to the questions ‘Who will wound nobody?’/‘Who does wait for nobody?’, that induce V-attraction to the head of the functional projection that hosts the focused negative word at its specifier (see details on this operation in Skopeteas and Fanselow 2008b).

195

(10) Pattern B: Class 1 (series I) and Class 2 (a) aravi-n aravi-s da-č’r-i-s/

nobody-NOM nobody-DAT PR(FUT)-(IO.3)cut-THM-S.3.SG / e-lod-eb-a. PV-(IO.3)wait-THM-S.3.SG

‘Nobody will wound/waits for nobody.’ (b) ??aravi-s aravi-n dač’ris/elodeba. (11) Pattern C: Class 1 (series III) and Class 4 (a) ??aravi-n aravi-s da-u-č’r-i-a/ . nobody-NOM nobody-DAT PR(PRF)-(INV.S.3)PV-cut-PF-INV.O.3(INV.S.SG)

s-ýul-s INV.S.3-hate-INV.O.3(INV.S.SG)

‘Nobody has wounded nobody.’ (b) aravi-s aravi-n da-u-č’r-i-a/s-ýul-s. The asymmetry presented in (10)-(11) should not be interpreted as a subject/object asymmetry. The next examples show that the same asymmetry holds between the theme and goal constituents of a ditransitive verb. In all examples, word order reflects embeddedness: the most embedded constituent surfaces closer to the verbal head. This holds independently of case properties as (12) and (13) illustrate for dative vs. postpositional arguments. (12) Ditransitive verb, pattern B (a) aravi-s araper-s a-čuk-eb.

nobody-DAT nothing-DAT PV-(IO.3)give-THM(S.2.SG) ‘You give nothing to nobody.’ (b) ??araper-s aravi-s a-čuk-eb. (13) Ditransitive verb, pattern C (a) aravi-s-tvis araper-i g-i-čuk-eb-i-a.

nobody-GEN-for nothing-DAT SINV.2-PV-give-THM-OINV.3.SG ‘You have given nothing to nobody.’ (b) ??araper-i aravi-s-tvisi-čuk-eb-i-a. The data pattern shown in (10)-(13) by means of negative words equally holds for a number of parallel constructions: indefinite quantifiers (vi-ghaca vi-ghaca-s V ‘somebody-NOM somebody-DAT’) and multiple constituent questions (vi-n vi-s ‘who-NOM who-DAT’) that provide evidence for superiority effects on wh- movement (superiority in the direct case marking pattern are dealt with in Amiridze 2006: 64). In all cases, that are not reported here in the interested of space, the data pattern is exactly the same, providing hence robust evidence that case marking does not interact with a change in the syntactic relations. 2.3.2. Control predicates Infinitives may be embedded in matrix predicates in control constructions, which provide a further diagnostic for argument asymmetry. The question in this construction is which argument of the embedded verb may be referentially controlled by the NP in the matrix

Since this type of context is restricted, native speakers perceive the order in (10b) as non-acceptable when it is presented out of the blue.

196

predicate. As it is exemplified in (14)-(16), it is only the actor argument which allows for control and this holds independently of case marking.4 (14) Class 1 (a) monadire i-c’q’-eb-s ǯarisk’ac-is cem-a-s hunter(NOM) PV-start-THM-S.3.SG soldier-GEN hit-INF-DAT ‘The hunter starts hitting the soldier.’ (b) ǯarisk’ac-i i-c’q’-eb-s monadir-is mier cem-a-s soldier-NOM PV-start-THM-S.3.SG hunter-GEN by hit-INF- DAT (15) Class 2 (a) monadire i-c’q’eb-s ǯarisk’ac-is da-lod-eb-a-s hunter(NOM) PV-start-THM-S.3.SG soldier-GEN PR-wait-THM-INF-DAT ‘The hunter starts waiting for the soldier.’ (b) ǯarisk’ac-i i-c’q’-eb-s monadir-is mier da-lod-eb-a-s soldier-NOM PV-start-THM-S.3.SG soldier-GEN by PR-wait-THM-INF-DAT (16) Class 4 (a) monadire5 i-c’q’-eb-s ǯarisk’ac-is še-ʒul-eb-a-s hunter(NOM) PV-start-THM-S.3.SG soldier-GEN PR-hate-THM-INF-DAT ‘The hunter starts hating the soldier.’ (b) ǯarisk’aci i-c’q-eb-s monadiri-s mier še-ʒuleb-a-s. soldier-NOM PV-start-THM-S.3.SG hunter -GEN by PR-hate-THM-INF-DAT 2.3.3. Binding properties Binding reflexive NPs has been a central argument of Harris (1981) for the subject status of dative arguments in inversion. The generalizations in Harris (1981) have been challenged in Asatiani (1982) and Amiridze (2005, 2006) who show that Georgian also allows for subject reflexive NPs (mostly for causative verbs), attested however with particular verb types and having constrained semantic properties. These observations do not lead to a complete refutation of the original claim, but they put an essential limitation: binding reflexive subjects is not categorically non-grammatical, but its occurrence is restricted, while binding reflexive objects is unrestricted. Crucially for our understanding of case inversion, all accounts agree that there is no impact of case marking on the reported binding asymmetries. It has been already observed that the same binding asymmetries that apply between subjects and objects also apply between subjects and indirect objects as well as between indirect objects and direct objects (see McGinnis 1998b, Amiridze 2006). Hence, without replicating examples that are already available in

4 This test is only applicable for the opposition of different verb classes, but not for the opposition between different inflectional categories (verb series), since the embedded verb is an infinitive. 5 Notice that the dative experiencer of class 4 verbs is nominative marked in the matrix predicate which may be thought to be counterevidence to the assumption that this argument is inherently case marked. Crucially, the embedded predicate is a nominalized verb form, i.e. an inflectional form that cannot project dative marked actor arguments (an actor constituent with the infinitive is expressed as an adjunct, see (16b)). Hence, the above examples provide only evidence for the eligibility of arguments of an embedded predicate for control phenomena, but are irrelevant for the observation of case preservation effects.

197

previous literature, we consider that the available evidence from binding reinforces our view that argument asymmetries are not influenced by case marking in Georgian. Georgian belongs to the languages that provide reconstruction effects in NPs with a reflexive determiner. Objects may scramble to a position from which they c-command subjects and in this position they may bind reflexive determiners (see McGinnis 2004, Skopeteas & Fanselow 2007). The examples (17)- (18) show that reconstruction effects are obtained independently of case marking. The (a) examples show that binding a reflexive determiner embedded within a subject NP is not grammatical in the canonical order. Scrambling the object past the subject as in the (b) example renders the configuration grammatical, a phenomenon that is accounted as movement to an argument position (see Mahajan 1990, Fanselow 2001, Haider & Rosengren 2003, among others). (17) Class 1/future and Class 2/future: pattern B (a) *tavis-ii/j biʒa q’vela msaxiobi-s RFL.3.SG-NOM uncle(NOM) every actor-DAT da-č’r-i-s/ da-e-lod-eb-a. PR(FUT)-cut-THM-S.3.SG PR(FUT)-PV-(IO.3)wait-THM-S.3.SG ‘Hisi uncle will wound every actori.’/‘Hisi uncle waits for every actori.’ (b) q’vela msaxiobi-s tavis-ii/*j biʒa every[DAT] actor-DAT RFL.3.SG-NOM uncle(NOM) da-č’r-i-s/ da-e-lod-eb-a. PR(FUT)-cut-THM-S.3.SG PR(FUT)-PV-(IO.3)wait-THM-S.3.SG ‘Hisi uncle will wound every actori.’/‘Hisi uncle waits for every actori.’ Experiencer verbs and class 1 (series III) verbs show exactly the same asymmetry, which gives further support to the view that the hierarchical relation between the arguments is the same in the direct and the inverse case marking pattern. (18) Class 1/perfect and Class 4: pattern C (a) *tavisi/j biʒa-s q’vela msaxiobi-i

RFL.3.SG uncle-DAT every actor-NOM da-u-č’r-i-a/ s-ʒul-s PR(PRF)-(INV.S.3)PV-cut-PF- INV.O.3(INV.S.SG) INV.S.3-hate-INV.O.3(INV.S.SG) ‘Hisi uncle wounded every actori.’/‘Hisi uncle hates every actori.’(b) q’vela msaxiobi-i tavisi/*j biʒa-s every actor-NOM RFL.3.SG uncle-DAT da-u-č’r-i-a/ s-ʒul-s PR(PRF)-(INV.S.3)PV-cut-PF-INV.O.3(INV.S.SG) INV.S.3-hate-INV.O.3(INV.S.SG) ‘Hisi uncle wounded every actori.’/‘Hisi uncle hates every actori.’ 2.3.4. Summary This section presented data from word order, control constructions, and binding. In all these phenomena, whenever evidence for a syntactic asymmetry is available, it holds that {NPNOM > NPDAT} in the direct case marking pattern and {NPDAT > NPNOM} in the direct case marking pattern, which amounts to saying that {actor > undergoer} across case

198

marking patterns, or simply that case marking does not interact with argument asymmetries. It should be added that case marking does not interact with information structure either, which is compatible with the data presented so far. Since information structure is encoded through word order (see Asatiani 2000, 2007, Skopeteas and Fanselow 2008a, Skopeteas and Fanselow 2008b, Skopeteas, Féry and Asatiani 2008) and since case does not interact with word order, it is expected that case marking does not interact with information structure too. This expectation is also experimentally confirmed in Skopeteas, Féry and Asatiani (2008), a study on the contextual felicity of several constellations of word order, case (direct vs. inverse patterns), and prosody. The results of this experimental study show that case marking has no influence on the intuition of contextual felicity. 3. Syntactic processing of case 3.1. Preliminaries We may now return to the examples of case inversion in (1) that are repeated below. After the outline of the Georgian case system in section 2, we are now in a position to hypothesize, how these utterances are syntactically processed by native speakers. (19) (a) ǯarisk’ac-i monadire-s da-č’r-i-s.

soldier-NOM hunter-DAT PR(FUT)-cut-THM-S.3.SG ‘The soldier will wound the hunter.’ (b) ǯarisk’ac-s monadire da-u-č’r-i-a.

soldier-DAT hunter(NOM) PR(PRF)-(INV.S.3)PV-cut-PF- INV.O.3(INV.S.SG)

‘The soldier has wounded the hunter.’ Following our account in section 2.2, there is no uni-directional case-to-argument association in Georgian. When these cases are determined by the syntactic configuration, nominative marks the higher constituent and dative the most embedded one (this holds for typical transitive verbs of the class 1 in series I). However, particular verbal heads may license a dative marked actor, goal or undergoer, in which case nominative marks the highest non-inherently case marked argument. In view of syntactic processing, the absence of uni-directional association implies that morphological case is a probabilistic cue about argument structure in Georgian before processing the verbal head. However, note that the two case licensing layers (structural and non-structural) are crucially asymmetric. If the selectional properties of the verbal head license particular case markings, these specifications are decisive. We subsume the licensing conditions for inherent and lexical cases under a common constraint that prevents deviations from the specifications of the verbal head concerning case licensing. This constraint applies on pairs of the form {role→case} that correspond to the non-structural cases licensed by the verbal head (either lexically or through its inflectional properties). The relevant pairs for the marking of actor and undergoer are two: {actor→dativeθ (class 4 verbs, perfect series of class 1 verbs), undergoer→dativeλ (class 2 verbs)}. The constraint on the selectional properties is formulated in (20).

199

(20) Constraint I: SELECTION Do not violate specifications of the predicate concerning non-structural case licensing. Structural case marking is strictly determined by locality, i.e., the highest eligible (non-structurally case marked) argument receives the highest case. This case licensing condition is formulated in terms of a violable constraint that ranges over two sets: the set arguments {external, internal} and the set of morphological cases {nominative, dative} and bans disharmonic alignments between these two hierarchies. (21) Constraint II: LOCALITY Do not mark the highest argument with the non-highest eligible structural case. Word order is also a probabilistic cue in Georgian. As shown in section 2.3.1, actors precede undergoers in the canonical word order, but deviations from the canonical order are possible and may be licensed through information structure (see in particular Harris 1981, Apridonidze 1986, Asatiani 2007, Skopeteas, Féry and Asatiani 2008, Skopeteas and Fanselow 2008a and 2008b). Notably, undergoer constituents may scramble over actor constituents, when the latter but not the former are discourse-anaphoric. In view of these facts, we assume that the default interpretation of an {argument, argument} sequence will be that the preceding argument is an actor. These facts are straightforwardly captured by the assumption of an {actor, undergoer} basic order and a constraint that renders non-canonical word orders suboptimal (in the sense of Grimsaw 1997). (22) Constraint III: STAY Do not assume movement from the base configuration. The constraints above apply in strict order. Structural case licensing has the status of an elsewhere condition, i.e., it applies for the subset of arguments that do not bear a non-structural case. In terms of constraint interaction, this generalization is captured through the constraint ranking: the constraint I (SELECTION) dominates constraint II (LOCALITY). Word order is always dominated by the case licensing constraints, since scrambling lexical object constituents over lexical subject constituents is always possible (and indeed frequently attested in production, see Apridonidze 1986, Skopeteas & Fanselow 2007; converging data from experimental judgments are presented in Skopeteas & Féry & Asatiani 2007). The ranking of the constraints at issue is given in (23). (23) SELECTION >> LOCALITY >> STAY 3.2. Incremental optimization We assume that the hearer develops assumptions about clause structure which are derived by the constraint interaction in (23) and these assumptions are subject to revision during the incremental parsing of the utterance (see model of incremental optimization in Lamers and De Hoop 2006). At the time point <t1> of the utterance in (24), the speaker has processed two lexical NPs with particular case and word order properties (i.e., NPNOM NPDAT). At this point, (s)he can develop assumptions about clause structure through LOCALITY and STAY. Both constraints suggest that the nominative and first NP is the actor constituent, as illustrated in Fig. 1. At the time point <t2>, the speaker has processed the verb. The verb belongs to class 1/series I, i.e. it does not license any inherent case and SELECTION does not apply. Hence, the optimal interpretation in <t1> is identical to the optimal interpretation in <t2>.

200

(24) monadire ǯarisk’ac-s da-č’r-i-s. hunter-(NOM ) soldier-DAT <t1> PR(FUT)-cut-THM-S.3.SG <t2>

‘The hunter will wound the soldier.’ Fig. 1. Incremental optimization: actor=nom&first, V: class 1, series I

NPNOM NPDAT... <t1>

...V <t2>

SELECTION LOCALITY STAY

ACT→nom&first ACT→nom&first ACT→dat&non-first ACT→dat&non-first * *

Though superficially identical, the situation with class 2 verbs is slightly different (see (25)). Verbs of this class are lexically specified for a dative undergoer, hence the input at <t2> involves the information that the undergoer bears dativeλ. By consequence, the SELECTION constraint applies at <t2>, reinforcing the assumption already made by LOCALITY and STAY at <t1>. (25) monadire ǯarisk’ac-s e-lod-eb-a.

hunter-(NOM) soldier-DAT <t1> (IO.3)PV-wait-THM-S.3.SG <t2> ‘The hunter waits for the soldier.’

Fig. 2. Incremental optimization: actor=nom&first, V: class 2 NPNOM NPDAT... <t1>

...V {UND→dat} <t2>


ACT→nom&first ACT→nom&first ACT→dat&non-first ACT→dat&non-first * * *

In sentence (26), the NPDAT NPNOM order creates a conflicting situation at <t1>. LOCALITY predicts that the actor is the nominative argument, while STAY makes the opposite prediction. The conflict at <t1> is resolved by the constraint ranking and is confirmed at <t2>, since the class 1 series I verbal head does not license any inherent argument and SELECTION does not apply. (26) ǯarisk’ac -s monadire- da-č’r-i-s.

soldier-DAT hunter-(NOM) <t1> PR(FUT)-cut-THM-S.3.SG <t2> ‘The hunter will wound the soldier.’

Fig. 3. Incremental optimization: actor=nom&non-first, V: class 1, series I NPDAT NPNOM... <t1>

...V <t2>


ACT→nom&non-first ACT→nom&non-first * ACT→dat&first ACT→dat&first *

SELECTION applies with a class 2 verb (see (27)). Incremental optimization proceeds in the same way as for class 1 series I verbs (see Fig. 3), but the verb which is processed at <t2> is lexically specified for a dativeλ undergoer. The outcome of the application of SELECTION is in line with the interpretation that the hearer already had computed on the basis of LOCALITY and STAY at <t1>. (27) ǯarisk’ac -s monadire e-lod-eb-a.

soldier-DAT hunter-(NOM) <t1> (IO.3)PV-wait-THM-S.3.SG <t2> ‘The hunter waits for the soldier.’

201

Fig. 4. Incremental optimization: actor=nom&non-first, V: class 2 NPDAT NPNOM... <t1>

...V{UND→dat} <t2>


ACT→nom&non-first ACT→nom&non-first * ACT→dat&first ACT→dat&first * *

Sentences (28)-(29) present sentences in which the verb licenses a dativeθ actor. Up to <t1>, the hearer has processed the same case and order information as in sentences (26)-(27) and is confronted with the same conflict between STAY and LOCALITY, which suggests a nominative actor. Processing the selectional properties of the verbal head in <t2> involves a revision of the assumed argument structure, since the class 1, series III verb in (28) and the class 4 verb in (29) license a dativeθ actor. This revision is presented in Fig. 5: the optimal candidate at <t1> differs from the optimal candidate at <t2>. (28) ǯarisk’ac -s monadire da-u-č’r-i-a.

soldier-DAT hunter-(NOM) <t1> PR(PRF)-(INV.S.3)PV-cut-PF- INV.O.3(INV.S.SG) <t2>

‘The soldier has wound the hunter.’ (29) ǯarisk’ac-s monadire sʒul-s.

soldier-DAT hunter-NOM <t1> INV.S.3-hate-INV.O.3(INV.S.SG) <t2> ‘The soldier hates the hunter.’

Fig. 5 Incremental optimization: actor=dat&first, V: class 1 (series III) or class 4 NPDAT NPNOM... <t1>

...V{ACT→dat} <t2>


ACT→nom&non-first ACT→nom&non-first * * ACT→dat&first ACT→dat&first *

Finally, sentences (30)-(31) illustrate the revision of a stronger initial assumption about the thematic properties of the processed NPs. Both LOCALITY and STAY suggest that NPNOM is an actor. However, the verb processed at <t2> licenses a dativeθ actor. (30) monadire ǯarisk’ac-s da-u-č’r-i-a.

hunter -NOM soldier -DAT <t1> PR(PRF)-(INV.S.3)PV-cut-PF- INV.O.3(INV.S.SG) <t2>

‘The soldier has wound the hunter.’ (31) monadire ǯarisk’ac-s s-ʒul-s.

hunter(NOM) soldier-DAT <t1> INV.S.3-hate-INV.O.3(INV.S.SG) <t2> ‘The soldier hates the hunter.’

Fig. 6 Incremental optimization: actor=dat&non-first, V: class 1 (III) or class 4 NPNOM NPDAT... <t1>

...V{ACT→dat} <t2>


ACT→nom&first ACT→nom&first * ACT→dat&non-first ACT→dat&non-first * *

The constraint interactions presented so far differ in two respects: (a) some sentences involve a revision of the assumed argument structure at <t2> (see Fig. 5 and Fig. 6), (b) the assumptions at <t1> form a scale of two different degrees of “strength” depending on the constraint interactions that motivate them. A strong assumption at <t1> is motivated

202

by converging evidence from both LOCALITY and STAY (see Fig. 1, Fig. 2, and Fig. 6). A weak assumption at <t1> is motivated by LOCALITY but involves conflicting evidence from the lower ranked constraint STAY (see Fig. 3, Fig. 4, and Fig. 5). Combining these differences results in a scale that reflects the extent to which the optimal candidate in <t2> is predictable at the point <t1> (see Table 5).

Table 5. Predictions predictions thematic

revision at <t2>

strength of assumption at <t1>

NPNOM NPDAT... V({UND→dat}) < NPDAT NPNOM... V({UND→dat}) < NPDAT NPNOM... V{act→dat} < NPNOM NPDAT... V{act→dat}

- - yes yes

strong weak weak strong

Aim of the empirical study in the following is to examine if the scalar assumption in Table 5 have a discernible effect on the reaction times in sentence processing. 3.3. Method Our study consists of two independent experiments. The first experiment examines the contrast between series I and series III in class 1 verbs, and the second experiment examines the contrast between class 2 and class 4 verbs. The experimental conditions are summarized in 32)- (33). Each experiment contains two factors, ‘case’ {level I: actor→nominative, level II: actor→dative} and ‘order’ {level I: actor→first, level II: actor→non-first}, which are crossed, hence rendering 4 experimental conditions. 32) Experiment I: series’ contrast (a) actor → nominative and first (class 1, series I); see (24) (b) actor → nominative and non-first (class 1, series I); see (26) (c) actor → dative and first (class 1, series III); see (28) (d) actor → dative and non-first (class 1, series III); see (30) (33) Experiment II: classes’ contrast (a) actor → nominative and first (class 2); see (25) (b) actor → nominative and non-first (class 2); see (27) (c) actor → dative and first (class 4); see (29) (d) actor → dative and non-first (class 4); see (31) These conditions were implemented in a DMDX presentation. Native speakers were first shown a nominative and a dative NP on the screen (either in the nominative-first or in the dative-first order) and a slot at the place of the missing verb, as it is illustrated in (34). (34) jariskaci monadires _______.

soldier-NOM hunter-DAT After 5 sec., the verb appears in the clause final slot: (35) ǯarisk’ac-i monadire-s dauč’ria.

soldier-NOM hunter-DAT PR(PRF)-(INV.S.3)PV-cut-PF-INV.O.3(INV.S.SG) The speakers were instructed to judge if this sentence is “right” or “false” in Georgian by typing the corresponding keys (left SHIFT=right, right SHIFT=false) as soon as they can.

203

The clock turned on when the verb appeared on the screen and the reaction time was stored within a time window of 10 sec. For each experiment, 16 items were developed in a factorial design rendering four data points per participant for each experimental condition. Each item contained two NPs with animate referents which were used in all four conditions. For the first experiment, each experimental item contained a class 1 verb which was used in two different inflectional forms, namely future (series I) and perfect (series III) (see list of the verbs in the appendix). For the second experiment, we used verbs that lexically license a case marking: 8 items contained class 2 verbs (direct case marking) and 8 items contained class 4 verbs (inverse case marking). The full list of verbs is given in the Appendix. The 16 items of both experiments were distributed in four DMDX presentations, so that every participant has seen each item once. Each presentation contained four trials per experimental condition for each experiment (total: 4 trials × 4 conditions × 2 experiments = 32 trials). Additionally, 96 fillers were used which were identical in the four presentations. 32 fillers were presenting well formed sentences with very different clause structures (intransitives, passives, verbs with adjuncts, etc.). The remaining 64 fillers displayed several violations of the subcategorization rules for verbs in order to draw the attention of the participants to the case properties of the arguments. The proportion of “right” vs. “false” sentences was 1:1 in each experimental session. The 32 target sentences and the 96 fillers were randomized through the software for each performance. Each DMDX presentation was performed by 8 participants. In sum, 32 participants, all students in the Humanities at the University of Tbilisi (age range 19 to 25) took part in the experiment and were paid for their contribution. The experiment was performed at the University of Tbilisi in September 2005. The choice of verbs was based on the appendix of verb classes in Harris (1981). Since the examined conditions are necessarily based on different inflectional forms of the same verb (experiment I) or on different verbs (experiment II), differences in the processing time of individual verbs or inflectional forms (due to their length, frequency, etc.) may have an influence to our result. In order to estimate the influence of item specific properties to the result, we have performed a lexical decision experiment. The verb forms used in experiments I and II were inserted in two DMDX presentations. Each presentation contained the 16 class 1 verbs (8 in series I and 8 in series III), 4 verbs of each class examined in experiment II, and 72 fillers (total: 16 + 2 × 4 + 72 = 96 elements). 10 participants performed the lexical decision experiment (students at the University of Tbilisi age range 20 to 25), that were different from the participants of the main experiments I and II. The resulting data set contains 5 measurements of reaction time for each verb form used in either experiment. 3.4. Results Experiments I and II resulted in a data set of 32 (target sentences) × 32 (participants) = 1 024 reaction time measurements. Some measurements were classified as non-valid either because they contained negative judgments (58 measurements), or because the participant did not give any judgment within the time window (7 measurements). After

204

excluding this data, we computed the z-transformation of the gathered RTs per participant and excluded 12 measurements that strongly deviated from participant’s means (z > ±2.5). The measurements presented below are calculated on the remaining (1 024 – 58 – 7 – 12 =) 947 valid RTs. 3.4.1. Experiment I: Influence of inflectional contrast (series I vs. series III) Experiment I examines the contrast between future (series I) and perfect (series III) of class 1 verbs. According to our account in section 2.2, actors and undergoers of future tense class 1 verbs are structurally case-marked, while actors of perfect tense verbs bear a non-structural dativeθ. The obtained means of the RTs as well as the means of the logarithmized measurements are given in Table 6.

Table 6. Means of raw and logarithmized measurements in Experiment I First→ACT Non-first→ACT msec log msec log Nom→ACT (series I) 2 171 7.60 2 319 7.66 Dat→ACT (series III) 2 563 7.75 2 632 7.80

As it may be observed in Fig. 7, the preference for nominative actors has a strong effect on the reaction times. The preference for initial actors has an additional weak effect. A repeated measures ANOVA was carried out on the logarithmized data, and an α level of .05 was adopted. This analysis revealed a significant main effect of case marking (F1,31 = 18.27, p < .001); the effect of the different word orders as well as the interaction between the two factors did not reach significance.

Fig. 7 Reaction times in Experiment I

On the basis of the results from the lexical decision test, we calculated the means of reaction times for each verb form. The natural logarithm of the means was inserted to a linear regression analysis as predictor for the logarithmized reaction times. The regression analysis was carried out for the whole data set of experiment I and gave a highly significant regression coefficient (t = 3.74, p < .001) which implies that the differences in the processing of the individual verb forms have a significant impact on the obtained reaction times. In order to eliminate this impact, we carried out an evaluation of the residuals per experimental condition, assuming that these values give an estimate for the actual effect of the investigated factors independently of the processing cost of individual verb forms (see Table 7).

1700

1900

2100

2300

2500

2700

Nom→ACT Dat→ACT

mea

ns o

f RTs

(in

mse

c.)

First→ACT

Last→ACT

205

Table 7. Means of residuals in Experiment I First→ACT Non-first→ACT Nom→ACT (series I) -.055 .007 Dat→ACT (series III) .011 .064

A repeated measures ANOVA at an α level of .05 revealed a significant main effect of case marking (F1,31 = 4.52, p < .05), but no significant effect of order nor of the interaction between the two factors. Hence, we conclude that the significant main effect of case marking is independent of the differences in the processing of the individual verb forms. 3.4.2. Experiment II: Influence of lexical contrast (class 2 vs. class 4) Experiment II examines the contrast between class 2 and class 4 verbs. Class 2 verbs license a dativeλ undergoer, while class 4 verbs license a dativeθ actor (see section 2.2). The means of the obtained RTs and the corresponding logarithmized values are given in Table 8. RTs in experiment II were overall lower that the RTs of experiment I. This difference is not relevant for our hypotheses; we may speculate that it relates to the fact that the verbs of experiment I were complex inflectional forms (future or perfect), while the verbs of experiment II were presented in present tense.

Table 8. Means and logarithized means in Experiment II First→ACT Non-first→ACT msec log(rt) msec log(rt) Nom→ACT (class 2) 1 822 7.43 1 857 7.45 Dat→ACT (class 4) 1 870 7.46 2 264 7.65

Fig. 8 shows that experiment II yields a different interaction effect than experiment I. A repeated measures ANOVA, carried out on the logarithmized data at an α level of .05, revealed a significant main effect of case (F1,31 = 9.73, p < .01), a significant main effect for order (F1,31 = 8.67, p < .01), and a significant interaction between the two factors (F1,31 = 9.44, p < .01).

Fig. 8 Reaction times gathered in Experiment II

Paired sample (2-tailed) t-tests revealed significant differences for all comparisons with the ‘Dat & Non-first →ACT’ condition (and only those).

1700

1900

2100

2300

2500

2700

Nom→ACT Dat→ACT

mea

ns o

f RTs

(in

mse

c.)

First→ACT

Last→ACT

206

(36) (a) Nom & First <rt Dat & Non-first (t31 = 3.81, p < .001) (b) Nom & Non-first <rt Dat & Non-first (t31 = 4.19, p < .001) (c) Dat & First <rt Dat & Non-first (t31 = 3.50, p < .001) The verb forms used in experiment II were independently tested in a lexical decision test. For the calculation of the effect of the individual verb forms on the obtained RTs we used the same procedure as in experiment I (see section 3.4.1). A linear regression analysis on the RTs of the individual verbs (in the lexical decision text) and the corresponding RTs in experiment II revealed a significant regression coefficient (t = 2.92, p < .01). On the basis of a regression analysis carried out separately for every participant, we computed the means of the residuals that are given in Table 9.

Table 9. Means of residuals in Experiment II First→ACT Non-first→ACT Nom→ACT (class 2) -.071 -.062 Dat→ACT (class 4) -.003 .171

A repeated measures ANOVA at an α level of ,05 revealed a significant main effect for case (F1,31 = 30.03, p < .001), a significant main effect of order (F1,31 = 9.47, p < .01), and a significant interaction effect (F1,31 = 9.72, p < .01). Paired t-tests (2-tailed) computed on the means of residuals gave the same significant differences with the logarithmized data (see (36)), as shown in (37). I.e., the only condition that involves a significantly higher processing cost is the condition of a dative actor that is not first in the linear order. (37) (a) Nom & First <rt Dat & Non-first (t31 = 5.59, p < .001) (b) Nom & Non-first <rt Dat & Non-first (t31 = 5.88, p < .001) (c) Dat & First <rt Dat & Non-first (t31 = 3.57, p < .001) 3.5. Discussion The experimental results are summarized in Table 10. Based on the significant differences we can establish two levels of processing cost for each experiment. Experiment I yielded a significant main effect of case that does not interact with order: this result implies two levels of effect size, α and α + β. In experiment II, only the order {NPNOM NPDAT} with a dative actor differed significantly from all other experimental conditions; this result also corresponds to two levels of effect size, α´ and α´ + β´. The resulting data patterns of both experiments are fully in line with our predictions in Table 5, however they are not identical to each other.

Table 10. Summary of predictions and results predictions results Exp. I

NPNOM NPDAT... V < NPDAT NPNOM... V < NPDAT NPNOM... V{act→dat} < NPNOM NPDAT... V{act→dat}

α α α + β α + β

Exp. II

NPNOM NPDAT... V{und→dat} < NPDAT NPNOM... V{UND→dat} < NPDAT NPNOM... V{act→dat} < NPNOM NPDAT... V{act→dat}

α´ α´ α´ α´ + β´

Both experiments show no significant differences between {NPNOM NPDAT} and {NPDAT NPNOM} orders when the nominative is interpreted as the actor argument. This is in line

207

for the view that these conditions do not involve a revision of the assumed thematic properties of the processed NPs (see Table 5) which follows from the assumption that structural case information (LOCALITY) outranks word order (STAY). The experimental findings differ with respect to the processing of sentences involving non-nominative actors. The results from experiment I show that reaction times increase whenever the verb licenses a dative actor. Hence, revising the assumptions based on LOCALITY is related to high processing cost. The data pattern of experiment II differs in that the {NPDAT NPNOM} order with a dative actor is processed within a short time. Hence, this order is sanctioned when the hearer parses an experiencer verb which lexically licenses a dative actor. This difference is not predicted by our account on Georgian case and cannot be explained through the data presented so far. In our future research on Georgian case marking, we are planning to examine if this finding may be replicated in acceptability judgments and if it will yield a robust effect in a self-paced study on sentence comprehension. In terms of a preliminary account of the observed difference, we hypothesize that it relates to the difference between lexically and inflectionally licensed dative actors, already mentioned in 2.2.1. It has been observed in German that a reanalysis towards a dative-first word order involves less processing effort than a reanalysis towards an accusative-first word order, since only the latter involves a reanalysis of the constituent structure (see revision of behavioral studies and qualitative neuro-physiological correlates in Bornkessel et al. 2004, Bornkessel & Schlesewsky 2007). In this view, it is not surprising that the thematic reanalysis of initial dative arguments as experiencers with class 4 verbs yields low processing cost. However, this explanation seems to be contradicted by the finding in the perfect tense of class 1 verbs. According to the observation in German, we may hypothesize that the processing of initial datives involves a reanalysis of the constituent structure with perfect tense verbs but not with experiencer verbs. The obtained difference suggests that the crucial factor is the distinction between lexical vs. inflectional licensing of non-structural cases. Dative actors of class 4 verbs are thematically determined by their relation to the lexical semantics of the verb. With class 1 verbs, dative actors in the perfect and nominative actors in the present denote the same range of thematic relations with the verb and mark both the external argument. Their case properties reflect the impact of the modal properties of the perfect to the relevance of actor’s involvement to the event (following assumptions in the literature summarized in section 2.2.1). We may thus speculate that the revision of the thematic properties of the morphological case with this verb group is associated with a reanalysis of the assumed constituent structure. Further research is required in order to test the implications of this claim and its compatibility with non-structural case licensing. 4. Conclusions The aim of this article is to account on the morphological and syntactic properties of case inversion in Georgian and to examine their implications for syntactic processing. We have shown that the complex case marking patterns in Georgian arise through two layers

208

of case licensing, non-structural and structural. Evidence for this distinction comes from the thematic relatedness of case affixes, from stratal uniqueness effects and from case preservation effects. The crucial point is the morphological dative which shows mixed properties arising from the syncretism of a structural case (accusative) and an inherent case (dative). As a structural case, the dative affix marks the lower structurally case marked argument. As a non-structural case, the dative affix marks experiencer arguments of class 4 verbs, goal arguments of class 1 verbs, actor arguments of class 1 verbs in the perfect tense, and is lexically licensed as a marker of the theme argument of class 2 verbs. Furthermore, we presented the findings of two experiments on incremental processing of sentences with two case arguments that are thematically ambiguous. The results of these experiments provided evidence that case marking is a more reliable cue than word order in processing thematically ambiguous expressions in Georgian. Scrambling the undergoer constituent over the actor constituent of verbs with a nominative actor does not result in additional processing effort. Furthermore, the experimental findings suggest a difference between dative experiencers and dative actors, since only the revision of the thematic properties of the latter is associated with high processing effort. At this stage of our understanding of Georgian case system, this difference may be only speculatively accounted for. Further research is planned in order to test the robustness of this experimental finding. As stated in the preliminary section, Georgian sentence processing presents an interesting puzzle, since this language displays free word order (such as German) and no uni-directional case-to-argument associations (such as Icelandic). The experimental findings suggest that case information is indeed a stronger cue for assumptions about argument structure in this language. In the absence of a verbal head speakers build their assumptions according to the default rules of structural case marking and less according to word order. Abbreviations 1 1st person NOM nominative 2 2nd person NV neutral version 3 3rd person OV object version AOR aorist PART participle CAUS causative PASS passive voice DAT dative PF perfect ERG ergative PL plural FUT future PR preverb GEN genitive PV preradical vowel THM thematic suffisex RFL reflexive INV.S inversive subject S subject (person affix) INV.O inversive object SG singular IO indirect object (person affix)

209

References Amiridze, Nino 2005. Georgian reflexives in subject function in special contexts. In:

Müller, Stefan (ed.), Proceedings of the HPSG05 Conference. Stanford: CSLI Publications, 455-472.

Amiridze, Nino 2006. Reflexivization strategies in Georgian. Utrecht: LOT Dissertation Series.

Anderson, Stephen R. 1984, On representations in morphology: case, agreement, and inversion in Georgian. Natural Language and Linguistic Theory 2, 157-218.

Apridonidze, Sh. 1986, sit’q’vatganlageba axal kartulši [word order in Modern Georgian]. Tbilisi: Mecniereba.

Aronson, Howard I. 1989, Inflection vs. derivation in Georgian conjugation. In: Aronson, Howard I. (ed.), The non-Slavic languages of the USSR: Linguistic studies. Chicago: Chicago Linguistic Society, 1-19.

Asatiani, Rusudan & Marina Ivanishvili 2007, Semantics and typology of dative subjects (on the Georgian data). In: P. Dekker and R. van Rooy (eds.), Proceedings of the Fourteenth Amsterdam Colloquium-2003, The University of Amsterdam, 2003, p.69-75.

Asatiani, Rusudan 1982, mart’ivi c’inadadebis t’ip’ologiuri analizi [Typological Analysis of Simple Sentence]. Mecniereba, Tbilisi.

Asatiani, Rududan 2007, The main devices of foregrounding in the information structure. In: Balder D.ten Cate and Henk W. Zeevat (eds.), Logic, language and Computation, 6th International Tbilisi Symposium on LLC, Tbilisi 2005, Revised selected papers. Springer-verlag Berlin Heidelberg, 2007, 21-31.

Asatiani, Rusudan 2007, Information Structure in Georgian. To appear in: Stavros Skopeteas, Sam Hellmuth, Gisbert Fanselow, Caroline Féry (eds.), The expression of information structure: The interaction of syntax and phonology in cross-linguistic perspective. Berlin: De Gruyter.

Bates, Elisabeth, Sandra McNew, Brian MacWhinney, Antonella Devescovi, Stan Smith (1982), Functional constraints on sentence processing: A cross-linguistic study. Cognition 11, 245-299.

Bates, McWhinney, and Kliegl 1974, Blevins, James P. 2005, Thematic inversion in Georgian. Ms., University of

Cambridge. Bornkessel, Ina and Matthias Schlesewsky 2007, The extended argument dependency

model: a neurocognitive approach to sentence comprehension across languages. To appear in Psychological Review.

Bornkessel, Ina, McElree, Brian, Schlesewsky, Matthias, and Friederici, Angela 2004, Multi-dimensional contributions to garden path strength: Dissociating phrase structure from case marking. Journal of Memory and Language 51, 495-522.

Bornkessel, Ina, Schlesewsky, Matthias, and Friederici, Angela 2002, Grammar overrides frequency: Evidence from the online processing of flexible word order. Cognition 85, B21-B30.

Bornkessel, Ina, Schlesewsky, Matthias, and Friederici, Angela 2003, Eliciting thematic reanalysis effects: The role of structure-independent information during parsing. Language and Cognitive Processes 18, 268-298.

210

Butt, Miriam and King, Tracy Holloway 2005, The status of case. In Dayal, V. and Mahajan, A. (eds.), Clause structure in South Asian languages. Kluwer.

Chomsky, Noam 1981, Lectures on Government and Binding. Dordrecht: Foris. Dowty, David, 1991, Thematic proto-roles and argument selection. Language 67,

547-619. Fanselow, Gisbert 2000, Optimal exceptions. In: Barbara Stiebels and Dieter

Wunderlich (eds.), The lexicon in focus. Berlin: Akademie Verlag, 173–209. Fanselow, Gisbert 2001, Features, θ-roles, and free constituent order. Linguistic

Inquiry 32, 405-437. Fanselow, Gisbert 2002, Quirky Subjects and Other Specifiers. In Barbara Kaufmann

& Barbara Stiebels (eds.), More Than Words. Berlin: Akademie Verlag, 227-250. Fanselow, Gisbert, Kliegl, R and Schlesewsky, M. 1999, Processing difficulty and

Principles of Grammar. In S. Kemper and R. Kliegl (eds.), Constraints on Language: Aging, Grammar and Memory. Dordrecht: Kluwer.

Felser, Claudia, Clahsen, Harald and Münthe, T. 2003, Storage and Intergration in the Processing of Filler-Gap Dependencies: An ERP Study of Topicalization and wh- movement in German. Brain and Language 87, 345-354.

Fiebach, Christian, Matthias Schlesewsky, and Angela Friederici 2002, Separating syntactic memory costs and syntactic integration costs during parsing: The processing of German WH questions. Journal of Memory and Language.47: 250-272.

Grimshaw, Jane 1997, Projections, Heads and Optimality. Linguistic Inquiry 28, 373-422.

Haider, Hubert & Rosengren, Inger 2003, Scrambling: Nontriggered chain formation in OV languages. Journal of Germanic Linguistics 15.3, 203-267.

Haider, Hubert 1999, The license to check: structural case plus economy yields Burzio’s generalization. In: Reuland, E. (ed.), Argument and case: Explaining Burzio’s Generalization. Amsterdam: Benjamins, 31-54.

Harris, Alice C. 1981, Georgian Syntax: A Study in Relational Grammar. Cambridge etc.: Cambridge University Press.

Harris, A.C., 1990. Georgian: a language with active case marking. Lingua 80, 47–65. Harris, A.C., 1993, Georgian. In: Jacobs, Joachim, Arnim von Stechow, Wolfgang

Sternefeld & Theo Vennemann (eds.), Syntax: An International Handbook of Contemporary Research. Berlin, New York: De Gruyter, 1377–1397.

Hemforth, B. 1993, Kognitives Parsing: Repräsentation und Verarbeitung sprachlichen Wissens. Sankt Augustin: Infix.

Hewitt, B.G. 1995, Georgian: A Structural Reference Grammar. Amsterdam, Philadelphia: Benjamins.

Joppen-Hellwig, Sandra 2001, Verbklassen und Argumentlinking: Nicht-kanonische Argumente, Expletiva und vierstellige Kausativa in Ergativ- versus Akkusativsprachen. Tübingen: Niemeyer.

Krems, J. 1984, Erwartungsgeleitete Sprachverarbeitung. Frankfurt/Main: Lang. Lamers, Monique & de Hoop, Helen 2006, Incremental distinguishability of subject

and object. In Kulikov, L., Malchukov, A., de Swart, P. (ed.), Case, Valency and Transitivity. Amsterdam: Benjamins, 269-287.

211

Mahajan, Anoop K. 1990, The A/A-bar distinction and Movement Theory, Ph.D. dissertation, Massachusetts Institute of Technology, Cambridge, Massachusetts.

Matzke et al. (2002) McGinnis, Martha 1997a, Case and locality in L-syntax: Evidence from Georgian. In

Harley, Heidi (ed.), MITWPL 32: The UPenn/MIT Roundtable on Argument Structure and Aspect.

McGinnis, Martha 1997b, Reflexive external arguments and lethal ambiguity. In Emily Curtis, James Lyle, & Gabriel Webster (eds.), Proceedings of WCCFL 16. Stanford: CSLI Publications, 303-317

McGinnis, Martha 2004, Lethal ambiguity. Linguistic Inquiry 35.1, 47-95. Röhm, D., Bornkessel-Schlesewsky, I. & Schlesewsky, M. 2007, The internal

structure of the N400: Frequency characteristics of language related ERP component. Chaos and Complexity Letters 2.2.

Skopeteas, Stavros & Féry, Caroline & Asatiani, Rusudan 2007, Word order and intonation in Georgian. Ms., to appear in Lingua.

Skopeteas, Stavros and Gisbert Fanselow 2007a, Effects of givenness and constraints on free word order. In Information Structure from different perspectives. Submitted to: Malte Zimmerman and Caroline Féry (eds.). Oxford: Oxford University Press.

Skopeteas, Stavros and Gisbert Fanselow 2007b, Focus in Georgian and the expression of contrast. Submitted to: Philippa Cook & Sophie Repp (eds.), Lingua, Special Issue ‘Contrast in Grammar’.

Skopeteas, Stavros, Caroline Féry, and Rusudan Asatiani 2007, Word order and intonation in Georgian. Lingua.

Skopeteas, Stavros, Ines Fiedler, Sam Hellmuth, Anne Schwarz, Ruben Stoel, Gisbert Fanselow, Caroline Féry, and Manfred Krifka 2006. Questionnaire on Information Structure (QUIS). Interdisciplinary Studies on Information Structure (ISIS), Vol. 4. Universitätsverlag Potsdam, Germany.

Stiebels, Barbara 2000, Typologie des Argumentlinkings: Ökonomie und Expressivität. Habilitationsschrift, Heinrich-Heine University of Düsseldorf.

Tschenkéli, Kita 1958, Einführung in die georgische Sprache, Bd. I-II. Zürich: Amirani Verlag.

Tuite, Kevin 1987, Indirect transitives in Georgian. Proceedings of the Berkeley Linguistics Society 13, 296-309.

Tuite, Kevin 1998, Kartvelian morphosyntax. Munich: Lincom Europa. Woolford, Ellen 2006, Lexical case, inherent case, and argument structure. Linguistic

Inquiry 37.1, 111-130. Wunderlich, Dieter 1997, Cause and the structure of verbs. Linguistic Inquiry 28, 27-

68.

212

Appendix: Verbs used in the comprehension experiment Table 11. Part I

class transcription translation 1 dač’ris ‘he/she wounds him/her’ 1 dabans ‘he/she washes him/her’ 1 dak’argavs ‘he/she looses him/her’ 1 irčevs ‘he/she chooses him/her’ 1 aɣmertebs ‘he/she admires him/her’ 1 k’lavs ‘he/she kills him/her’ 1 xedavs ‘he/she sees him/her’ 1 at’arebs ‘he/she carries him/her’ 1 malavs ‘he/she hides him/her’ 1 daɣlis ‘he/she makes him/her tired’ 1 acek’vebs ‘he/she causes him/her to dance’ 1 avarǯišebs ‘he/she causes him/her to exercise’ 1 amɣerebs ‘he/she causes him/her to sing’ 1 atamašebs ‘he/she causes him/her to play’ 1 amušavebs ‘he/she causes him/her to work’ 1 amepebs ‘he/she causes him/her to reign’

Table 12. Part II

class transcription translation 2 elodeba ‘he/she waits him/her’ 2 šeepereba ‘he/she suits him/her’ 2 šeesabameba ‘he/she fits him/her’ 2 šeesat’q’viseba ‘he/she corresponds him/her’ 2 etamašeba ‘he/she plays with him/her’ 2 elap’arak’eba ‘he/she speaks with him/her’ 2 ečxubeba ‘he/she fights him/her’ 2 st’umrobs ‘he/she visits him/her’ 4 uq’vars ‘he/she loves him/her’ 4 sʒuls ‘he/she hates him/her’ 4 mosc’ons ‘he/she likes him/her’ 4 avic’q’deba ‘he/she forgets him/her’ 4 axsovs ‘he/she remembers him/her’ 4 enat’reba ‘he/she misses him/her’ 46 sč’irdeba ‘he/she needs him/her’ 4 aint’eresebs ‘he/she interests him/her’

6 The verbs sč’irdeba and aint’eresebs are classified as class 4 verbs in the appendix of Harris (1981), on which the selection of the above verbs is based. These verbs share some morphological properties with class 2 verbs, however they both have inversive argument structure, hence they have to be considered together with the class 4 verbs with respect to the experimental hypotheses.

213

Sinaarsi

rusudan asaTiani .............................................................................................................. 3 intonaciis roli winadadebis sainformacio

struqturis formirebaSi

rusudan asaTiani ........................................................................................................... 14 informaciis struqturirebis sintaqsuri

modelebi qarTulSi

rusudan asaTiani ........................................................................................................... 30 vnebiTi gvaris konstruqciaTa pragmatikuli

Tavisebureba qarTvelur enebSi

qeTevan grZeliZe ............................................................................................................ 45 winadadebis sainformacio struqtura:

lokativebi qarTulSi

naTia dundua ............................................................................................................ 62 sainformacio struqturis ganmsazRvreli

formaluri modelebi Tanamedrove sparsulSi

marine ivaniSvili ............................................................................................................. 69 winadadebis sainformacio struqtura:

sityvaTa rigi qarTvelur enebSi

ivane leJava ............................................................................................................ 82 glotalizaciis xarisxi winadadebis

fokusirebul da topikalizebul wevrTan

irine meliqiSvili .......................................................................................................... 89 vnebiTi gvari da Tavisufal sityvaTa rigi

rogorc aqtantebis topikalizaciis alternativebi

eTer soselia ........................................................................................................... 96 sxvaTa sityvis -o nawilaki da misi funqcia teqstSi

nuca wereTeli, rusudan asaTiani ........................................................................ 104………

implikaciuri topiki qarTulSi

TinaTin janeliZe, rusudan asaTiani ........................................................... 119 mapirispirebeli kavSirebis Ffunqcionaluri

analizi qarTulSi

marika jiqia ........................................................................................................... 131 erTi tipis qvewyobili winadadebis Sesaxeb megrulSi

Rusudan Asatiani ........................................................................................................... 138 THE MAIN DEVICES OF FOREGROUNDING IN THE INFORMATION STRUCTURE OF GEORGIAN SENTENCES

Stavros Skopeteas, Caroline Féry, Rusudan Asatiani ............................................ 148 WORD ORDER AND INTONATION IN GEORGIAN

Stavros Skopeteas, Gisbert Fanselo, Rusudan Asatiani ........................................ 185 CASE INVERSION IN GEORGIAN: SYNTACTIC PROPERTIES AND SENTENCE PROCESSING

informaciis struqturirebis ZiriTadi modelebi qarTvelur enebSi · (d) nawilakebi...

Documents

Transcript of informaciis struqturirebis ZiriTadi modelebi qarTvelur enebSi · (d) nawilakebi...