Albert Ortmann & Tinatin Kiguradze (University of Düsseldorf) The factors of variation in definiteness marking in the history of Georgian According to some scholars of Old Georgian, the language’s system of encoding definiteness must have been fairly straightforward: „Nouns without the article were indefinite, nouns with the article definite“ (Fähnrich 1991: 150; cf. also Shanidze 1930). As soon as one looks closer, however, one encounters numerous occurrences of undetermined nouns in definite contexts. Throughout a period of about one-thousand years, there is considerable variation, both inter- and intra-textually, as to the use or non-use of articles. The goal of our paper is to achieve a more precise understanding of this variation by investigating the following conditions: – the semantic and pragmatic factors which govern the use of articles – the variation in terms of text sorts – the role of contact, in particular the influence of Greek The historical development of definiteness marking in Georgian appears to be strikingly unusual in that in a certain stage of later Old Georgian (roughly, 8th to 11th century) definiteness is marked more systematically, and in wider semantic environments, than in both earlier and later stages of the language. The forms of the Old-Georgian postnominal distal demonstrative pronoun igi/man/mas/mis were also used as definite articles, and they furthermore had the function of 3rd person pronouns. This picture is familiar from the history of many other languages; as a matter of fact, it is considered to be a common path of grammaticalization (cf. Lyons 1999: 134). Crucially, however, whereas in other systems such as, e.g., those of Germanic and Romance, this development is typically followed by phonological reduction, semantic weakening/loss of deictic features, and an increase of obligatoriness in contexts of definiteness, the development of Georgian clearly differs in this respect. The demonstrative function of the forms at issue has been maintained throughout. Even in Modern Georgian, the very markers, though in a modified phonological shape (is in the nominative, and im in the other structural cases), fulfil the sole function of (prenominal) demonstrative pronouns, while definiteness is no longer expressed. The course of this development, however, was not a steady one. In earlier Old Georgian text, the demonstrative pronoun in its use as definite article occurs fairly systematically in contexts of deitic/situational definiteness (especially anaphoricity), but hardly occurs in contexts of nondeictic/non-textual definiteness. This may be illustrated from the following examples, which are from the very earliest Old Georgian text, an autochthonous (that is, non-translated) hagiographical narration, and from a non-clerical text, respectively: (1) čika-j igi p’ir-sa šealec’a da ġvino-j igi daitxia. glass.NOM DET.NOM mouth-DAT throw and wine-NOM DET-NOM spill ‘She threw her the glass into the face and the wine was spilled.’ [Sus VI] (2) p’ilo-j igi saxe ars elephant-NOM DET.NOM face.NOM COP.3SG ‘The elephant is the face of death.’ sik’udil-isa death-GEN [Bal] The nouns for glass, wine, and elephant were all previously introduced in the text; accordingly, they are marked for definiteness. By contrast, the nouns for mouth, face, and death are not anaphoric, hence not accompanied by the article; they are, however, semantically definite (in the sense of unique reference; cf. Löbner 1985) since they each refer unambiguously to an entity that is identifiable to the reader. In later texts, the domain of the use of articles is extended so as cover precisely these contexts of non-deictic/semantic definiteness as well, as in the following examples from the 9th century: 2 (3) a. gza-sa mas č’ešmarit’-sa ertġmteeb-isa-sa way-DAT DET.DAT true-DAT monotheism-GEN-DAT ‘the true path of monotheism’ b. gandevna sibnelej igi sul-ta ban-AOR.3SG darkness-NOM DET.NOM soul-GEN.PL ‘he banned the darkness of the believers’ souls’ [Luc 9] morc’mune-taj. believer-GEN.PL [Luc 11-12] Note, crucially, that this extension of the use of the article is typically found in religious translations. In autochthonous historical texts, by contrast, definiteness still tends to be signalled only where it is textually established. In other words, the article is usually absent in cases of nonanaphoric definiteness (ex. from the 11th century): (4) odes babilon-s godol-i aġašenes when Babylon-DAT tower-NOM build.AOR.3PL ‘after they had built the tower of Babylon’ [Leon I, 9-10] The asymmetry between autochthonous texts and translations is also reflected in the subsequent gradual decrease of the use of articles. In autochthonous non-clerical texts, articles more and more fail to occur even in contexts of anaphoricity as early as in the 12th century: (5) sxva ʒe ar esva mepe-sa other.NOM son.NOM NEG sit_to_him king-DAT ‘the king had no (other) son’ [Vepx 33] In translations of religious texts, by contrast, the exponence of anaphoric definiteness is, at least optionally, maintained significantly longer, as the following passage about Sarah (taken from the Biblia Mcxetica; 16th century) illustrates: (6) ixiles megvip’t’-el-ta dedak’ac-i igi, rametu šuenier iq’o priad see.AOR.3PL Egyptian-PL-DAT woman-NOM DET since beautiful be-AOR.3SG very ‘The Egyptians saw the woman because she was very beautiful.’ [Genesis 12, 14] We argue that the extension of the article use one encounters in translations was influenced by the systematic use of articles in virtually all contexts of definiteness in Greek, from which by far most hagiographical texts were translated. This also accounts for the longer survival of articles and to the more complex history of decrease in translations compared to autochthonous texts (first in contexts of non-deictic definiteness, and later in deictic/textual/anaphoric contexts). On the other hand, the picture that emerges with autochthonous non-clerical texts can be regarded as the natural development of the loss of a functional category, that is, not influenced by external factors such as language contact, and interacting with a global word order change in Kartvelian towards prenominal modification and reduction of case concord. References Fähnrich, Heinz (1991) Old Georgian. In: Alice C. Harris (ed.): The indigenous languages of the Caucasus, Vol. 1: The South Caucasian Languages, 129-217. Delmar, N.Y.: Caravan. Löbner, Sebastian (1985) Definites. Journal of Semantics 4. 279-326. Lyons, Christopher (1999) Definiteness. Cambridge: Cambridge University Press. Shanidze, Akaki G. (1930/1980) kartuli gramatikis sapudzvlebi (“Foundations of Georgian grammar”, vol. III). Tbilisi: sakartvelos ssr mecnierebata akademia. Sources [Bal] Balavariani. Quoted from: Sardjveladze, Zurab (1997) dzveli kartuli ena. („The Old Georgian Language“). Tbilisi: tbilisis saxelmtsipo pedagogiuri universiteti. [Sus] Iacob Tzurtaveli: Martyrium Susanicae. (http://titus.uni-frankfurt.de/texte/etcs/cauc/ageo/gh/gh1/gh1.htm) [Luc] Bishop Pomphilos: Martyrium Luciani. (http://titus.uni-frankfurt.de/texte/etca/cauc/ageo/keimena/keimena2/keime.htm) [Leon] Leontius Ruensis, Vitae regum Iberorum (http://titus.uni-frankfurt.de/texte/etcs/cauc/ageo/kcx1/kcx1.htm) [Serap] Vita Serapionis Zarzmensis http://titus.uni-frankfurt.de/texte/etcs/cauc/ageo/gh/gh1/gh1.htm [Vepx] Šota Rustaveli: Vepxist’q’aosani. http://titus.uni-frankfurt.de/texte/etcs/cauc/mgeo/vepx/vepx.htm