Linguistic Confusion or Multilingualism? : Fragmentary Finnish in Old Swedish Charters c. 1350–1526.

This thesis examines fragmentary Finnish in Late Old Swedish charters (c. 1350–1526) issued in the Finnish part of the Swedish realm, the diocese of Åbo. Consisting mostly of proper names, albeit occasionally displaying Finnish inflectional and derivational morphology, these fragments have previously not generally been regarded as representing actual written Finnish, but rather as onomastic loans or transcriptions of oral language by more or less monolingual Swedish scribes. This thesis attempts a description and analysis of the Swedish–Finnish language mixture, to see to what extent the embedding of Finnish segments in these Swedish-language charters can be said to reflect scribal proficiency in Finnish or a lack thereof. The thesis relies on theoretical and empirical findings in the fields of code-switching and historical sociolinguistics. To provide a socio-historical context for the linguistic analysis, sociolinguistic conditions in medieval Finland and the textual genre of medieval charters are outlined. The bilingual segments in the data are then described and compared with models of code-switching from modern studies, to see whether their form corresponds to patterns that could be expected of more or less balanced bilinguals. The choice between Swedish and Finnish linguistic variants is also considered in the light of textual and sociolinguistic factors, and a study is made of Finnish grammatical transfer in the scribal Swedish of medieval Finland. Although the scarcity of the medieval data does not allow definite conclusions, the tentative results reveal a language mixture that is mainly well formed, though limited in scope and with some instances of scribal errors that could be due to a lack of proficiency in Finnish. On the other hand, the insertion of Finnish segments shows a stylistic patterning that suggests a linguistic awareness on the part of the scribes, and the choice of Swedish prepositions in certain constructions differs quantitatively from the norm in non-Finnish parts of medieval Sweden, in a way that can partly be attributed to the influence of Finnish locative case semantics. While it is apparent that proficiency levels in Finnish must have varied somewhat among medieval scribes in Finland, the results point to a more or less bilingual proficiency, or at least extensive passive knowledge of Finnish