Whereas humans have doubtlessly been switching codes for as long as there have been bilinguals, the phenomenon of code-switching has only received substantial attention in linguistics in the second half of the last century (Alvarez Cáccamo 1998 traces the first mention of the term back to Vogt 1954), and large-scale research on code-switching did not start before the 1970s. Of course it had not escaped previous research on language contact that speakers switch between languages (cf. Hermann Paul’s chapter on ‘language mixture’, 1898: 365–377); however, this was considered an externally induced mishap due to negative psychological or social factors (incompetence, laziness, lack of education, migration, conquest) which did not merit linguistic investigation and was, in any case, temporary (‘man wird von der Zweisprachigkeit wieder zur Einsprachigkeit gelangen’ [‘from bilingualism one will arrive at monolingualism in the end’], as Paul 1898: 366 put it). In light of this neglect, we can doubtlessly discover a monolingual bias which was characteristic of linguistics for a long time: The beginnings of the discipline (in the early nineteenth century) coincided – not by chance – with the triumph of the nation state and its ideology of ‘one nation one language’. In this ideology, multilingual language use had no place (Auer 2007; Heller 2007; Pujolar 2007), even though forms of language contact did (borrowing in particular, but also pidgin and creole formation). The enormous interest code-switching and related practices have found in linguistics over the last three decades is at least in part (i.e. in addition to internal developments in linguistics) due to the demise of the monolingual national language ideologies which have become less and less realistic in the age of globalization, transnationalism, and migration. Today, multilingual practices are an undisputed everyday phenomenon, even in the European nation states which hitherto considered themselves to be monolingual (see Stroud 2007 for a view beyond Europe).