Producing composite codeswitching: The role of the modularity of language production

Abstract

The basic characteristic of composite codeswitching is that the languages involved share responsibility for framing bilingual constituents. This paper points to evidence of this characteristic in the nature of morpheme distribution in mixed possessive constructions in Ewe–English codeswitching, spoken in Ghana. An Ewe semantic distinction between two types of possessive constructions is consistently neutralized when English possessum nominals are used instead of their Ewe counterparts, and the paper demonstrates that the neutralization of this distinction results from direct mapping of English-origin grammatical information about English nominals onto Ewe grammar. It explains that this mapping of information from one grammar onto another one is characteristic of composite codeswitching and that it is facilitated by the fact that language production is modular in the sense of Levelt ((1989). Speaking: From intention to articulation. Cambridge, MA: MIT Press) and Myers-Scotton ((1993). Duelling languages: Grammatical structure in codeswitching. Oxford, UK: Clarendon Press; (2002). Contact linguistics: Bilingual encounters and grammatical outcomes. Oxford, UK: Oxford University Press).

Language contact, language mixing and identity: The Akan spoken by Ghanaian immigrants in northern Italy

“…The examples quoted in Ansre (1971), Myers-Scotton (1993 and Amuzu (2005Amuzu ( , 2014 suggest that some other major African lingue franche in contact with English, such as Ewe, Swahili, Shona and Yoruba, may currently be undergoing a similar transition. 18 In order to illustrate the process which leads to language mixing, I feel that it is significant that Auer (1999) chooses a spontaneous interaction -reported in Blommaert (1992, p. 61) -between Swahili native speakers in Tanzania.…”

Situated Language Use in Africa

“…Being a hierarchical (i.e., "dominance" or "insertional") model, the MLF posits a single matrix language in which speakers insert various embedded language "islands." A discussion of how the MLF fares in comparison to non-hierarchical approaches (which describe CS grammar mostly in terms of syntactic equivalence between languages) falls outside the scope of the present discussion, but the reader may consult Kamwangamalu (1994), Finlayson et al (1998), Owens (2005), Amuzu (2014) or Baloji et al (2014) for relevant applications in African linguistics. Lamidi (e.g., 2008) discusses CS grammar within a generativist framework, while Bokamba has in many publications (e.g., 1989) been utterly critical of the universal constraints posited by equivalence-approaches.…”

Combining the Markedness Model and the Matrix Language Frame Model in Analysing Bilingual Speech

“…It is shown in those works that because of this, the slot in which an English content morpheme occurs may or may not correspond to the slot in which its Ewe equivalent occurs, i.e. (Amuzu, 2014a) b. Sukudzíkpɔ-lá (*ƒe) daa tsitsitɔ … head.teacher-DEF POSS sister older-one Amuzu (2014a) explains that the occurrence of both doctrines and elder sister in the [NP ʄe NP] structure has nothing to do with the distribution of their respective Ewe equivalent. The CS slot of an English content morpheme will only correspond to the slot in which its Ewe equivalent occurs if the Ewe equivalent has similar subcategorization features.…”