|Type||Journal Article - Linguistik online|
|Title||Discovering the prehistory of multilingual situations in the lexicon. An empirical study on the Caucasian Urum vocabulary|
The transfer of linguistic entities in situations of language contact follows particular trends that may be generally subsumed under two dimensions. The first dimension refers to cross-linguistically established asymmetries with respect to the likelihood of borrowing for particular types of linguistic entity. For instance, core lexicon is less likely to be borrowed than non-core lexicon, the borrowing of nouns is more likely than the borrowing of verbs, word order borrowing is more likely for verb phrases than for adpositional phrases (see Matras 2007, for a summary of asymmetries in structural categories; see Swadesh 1955; Haspelmath/Tadmor 2009, for asymmetries in the lexicon; see also Thomason 2001: 70s.; Aikhenvald 2006: 5, for scales integrating lexical and grammatical phenomena). The second dimension refers to the culture-specific properties of individual contact situations. For instance, the use of words of Latin origin in scientific contexts, the use of English words for concepts relating to modern technology, or the borrowing of local place names by victorious invaders in several cases of language contact have straightforward socio-cultural determinants (see Thomason 2001: 66–69; Clyne 2003: 238–241; Appel/Muysken 2005: 165–170; Myers-Scotton 2006: 212–215; Haspelmath 2008: 51; Bartels 2009: 314–316).
The observation of such phenomena motivates inferences about the prehistory of language communities. For instance, the observation of common elements in the core lexicon implies a genetic relation. This is the basic assumption of the comparative method in historical linguistics (see Hock 1991: 384–345; Campbell 1999: 112; Rankin 2003: 187), as well as in the estimation of the time depth of genetic relationships in glottochronology (see Swadesh 1952, 1955; Lees 1953). Since socio-cultural contacts lead to the transfer of lexical items, the observation of borrowings in particular semantic domains is evidence for exchange in the corresponding domains of communication (e. g., see the description of loanwords in Archi, a language of the North Caucasus, in Chumakina 2009: 434–437). In this vein, Greenberg (1960: 206) interprets the presence of words of Kanuri origin in the Hausa vocabulary for 'writing' as evidence that the Kanem Empire exercised cultural influence on the Hausa states (see also the findings of a detailed recent investigation on Hausa in Awagana/Wolff 2009: 156, and on Kanuri in Löhr/Wolff 2009: 184).
The aim of this article is to draw inferences from the lexical inventory of Caucasian Urum, which is a variety of Anatolian Turkish spoken by ethnic Greek speakers on the Small Caucasus (Georgia). The majority of Caucasian Urum speakers are bilingual in Russian (93%), most of them are also competent in Georgian (83%), and they have intensive contact with Pontic Greek speakers in Georgia, who are considered to be homo-ethnic (see details in Section 2). Hence, we are dealing with a multilingual profile involving very different languages. The challenge of the present study is to draw inferences from the Urum vocabulary about the history of language contact, as summarized in (1).
(1) Likelihood of borrowings and historical inferences: research question
Knowing the likelihood for a concept to be borrowed across languages, which inferences can we draw from the origin of lexical items about the stratification of the involved languages in a contact situation?
Recent research on borrowings, in particular the World Loanword Database (= WOLD), opens new possibilities to the examination of linguistic relations manifested in the lexicon. The likelihood of borrowing was estimated for a large inventory of concepts based on the attested borrowings in a large cross-linguistic sample of 41 languages (see Tadmor 2009: 66). In order to answer the question in (1) we collected lexicological material in the field based on the WOLD-inventory (Haspelmath/Tadmor 2009; see also details of the data collection in Section 3). Based on this empirical data, we examine the following issues:
(a) Is the occurrence of borrowings informative for the stratification of the involved languages, i. e., for the distinction between substrate and superstrate languages (see Section 4)?
(b) What do we learn from the asymmetries in the frequency of borrowings in particular conceptual domains (see Section 5)?
(c) What does the likelihood of borrowings imply for the relation between Caucasian Urum and other related languages (see Section 6)?
|»||Georgia - General Population Census of 2002|