Aspects of a formal theory of bilingualism
Thursday 7 November, 10.30-11.00, E0104
Most formal theories of the syntactic competence of the native speaker are designed to deal with what Chomsky (1965) has called “the ideal speaker/listener”, where it is assumed that the speaker is monolingual. Since then, most formal work has retained this simplification, mainly because it has made it easier to construct theories of complex empirical phenomena. Because of this, phenomena of a more complex character, like various kinds of language mixing phenomena seen in bilingual situations, have never received much attention (though see MacSwan 1999; see also Muysken 2000). Given that the latter phenomena are the default in today’s world, it is important that formal approaches to grammar are designed such that they can analyze bilingual (and multilingual) situations. The goal of the present talk is to take a few steps towards such a theory, which will increase our understanding of the interplay between language diversity and universal design: In order to understand cognitive aspects of language mixing, it is also necessary to understand the grammar of bilingual speakers.
This talk will discuss bilingual situations of mixing/code-switching where one language is the native language, but where the other language is acquired later (by the same individual) as a second language. This is typical of immigration into a foreign speech community. Norwegian immigrants in the US since 1850 until today provide an example of such immigration and language mixing. In his seminal work, Haugen (1953) discusses examples such as (1).
(1) Så play-de dom game-r
then play-PAST they game-PL
‘Then, they played games’
In this example, some of the content words have been replaced by English words, but these English words have Norwegian inflection. We will use additional material collected by several researchers since the 1930ies of the Norwegian immigrant language in America that existed from about 1850 and (for a few speakers) up to our time (available online through The Text Laboratory, University of Oslo).
We will show what the main patterns are in language mixing between Norwegian and American English, and investigate whether they fit into a specific formal theory of the grammar of bilingual speakers. This formal theory will be an exoskeletal theory (Borer 2005), where we distinguish between the content words and the functional words (cf. Åfarli 2007, Lohndal 2012 for exoskeletal theories). We will show that the Norwegian data fits into the following pattern, which is based on other data from Muysken (2000) and Myers Scotton (2002).
(2) a. LSEC + INFLMAIN
b. LMAIN + INFLMAIN
c. *LSEC + INFLSEC (except in bigger code-switched chunks)
d. *LMAIN + INFLSEC
Here L = lexical content morpheme and INFL = inflectional morpheme, and the subscripts MAIN and SEC show whether the morpheme comes from the main or secondary language. This pattern shows that lexical content words cannot come from the main language when the inflectional morphemes come from the secondary language. We will show how an exoskeletal theory predicts this pattern and demonstrate that the contact situation between Norwegian and American English fits into the same picture.