Abstract

Code-switching as a contact phenomenon has been studied for several decades from the point of view of sociolinguistics, psycholinguistics and structural linguistics, among others. Different theoretical models have been proposed, all building on modern, mostly oral data. However, code-switching is by no means confined to modern language communities. Intricate intrasentential mixes of two languages can already be found in manuscripts from the middle ages. This talk will focus on the morpho-syntactic characteristics of two datasets, Macaronic Sermons from 15th-century England and Martin Luther’s Table Talk from 16th-century Germany. We will assess how well Myers-Scotton’s Matrix Language Frame Model of code-switching holds for historical written texts, then consider possible interpretations of systematic discrepancies between the predictions of the model and the actual realization of grammatical morphemes, especially Latin case markers. The aims are twofold: to show how historical data can test a theoretical model based on modern code-switching, and to advocate using such a theoretical model to identify constant features of code-switching through the centuries.