One of the prevailing views in the study of language change today is that—more often than not—diachronic innovations that occur in settings of language contact owe their development to a combination of language-internal and ‑external (i.e., contact-related) factors. In this talk, I offer support for this view by examining a set of cyclical changes in the expression of spatial relations by means of adpositional relators in the Modern Greek dialects of Asia Minor (Cappadocian, Pharasiot, Silliot, Pontic), which developed for a significant amount of time in the context of intense language contact with Turkish.
The Asia Minor Greek dialects display two sets of relators, which they inherited from Late Medieval Greek: (a) one set of highly grammaticalised and semantically bleached prepositions, and (b) one set of adverbs of a more lexical nature that combined with simple prepositions to encode spatial meanings in a more salient fashion. Simple prepositions could only occur on the left of their complements ([prep + np]). The position of adverbs, however, was variable: they could appear either before simple prepositions or after the prepositional complement ([adv + prep + np] ~ [prep + np + adv]). In inner Asia Minor Greek, the similarity of the latter option with corresponding Turkish postpositional structures led to the promotion of the postnominal order as the only available grammatical option, thus resulting in the formation of circumpositions ([prep + np + adv]), a crosslinguistically rare adpositional type. At a later stage and in a few varieties, the more grammaticalised and therefore less meaningful simple preposition se was lost from circumpositions, leaving postpositional adverbs as the only overt relators ([np + adv], i.e., [np + postp]). In that, Asia Minor Greek and Turkish converged fully with respect to the encoding of spatial relations. Convergence, however, was only unidirectional as Turkish did not undergo any changes that would make its adpositional system more like that of Greek, a typical symptom of displacive contact (in the sense of Aikhenvald 2007).
Apart from highlighting the role of multiple causation as the most fruitful approach to the study of contact-induced language change, the case-in-point further confirms Heine’s (2008) proposal that language contact rarely results in the creation of previously unavailable word orders. Rather, it most commonly favours the conventionalisation of originally pragmatically marked alternatives that happen to correspond to structural patterns in the model language. It also corroborates Stilo’s (2009) thesis that when languages that differ with respect to head ordering are spoken in the same geographical area, contact may often result in the merger of the two opposites into one hybridised pattern framing the head; for example, the contact between a prepositional and a postpositional language may result in the development of circumpositions as exactly is the case with Asia Minor Greek.