LINGUIST List 10.1822

Mon Nov 29 1999

Sum: re Query:10.1759:Code-Switching:Matrix Language

Editor for this issue: James Yuells <>


  1. Sophie ALBY, re Query:10.1759:Code-Switching:Matrix Language

Message 1: re Query:10.1759:Code-Switching:Matrix Language

Date: Mon, 29 Nov 1999 15:14:38 -0200
From: Sophie ALBY <>
Subject: re Query:10.1759:Code-Switching:Matrix Language

Thanks to Carol Myers-Scotton for her comments on my mail (19/11/99) :
"I am presently doing some doctoral work on french-kali'na (an amerindian
language spoken in French Guiana - Carib family) code-switching.

I am having trouble with the question of the 'base language' of the
sentences where I have occurences of code-mixing.
Ellen F. Prince and Susan Pintzuk (Bilingual Code-Switching and the
Open/Closed Class Distinction - University of Pennsylvania - January 1984)
states that :
"Following Joshi 1983 and others, we used the tensed verb of each tensed S
to determine the matrix language of that S".
Is there a general definition of matrix language that would fit for tensed S
and for other types of sentences ?"

Comments from Carol Myers-Scotton :
(1) First, under the Matrix Language Frame model (see the "afterword" in the
1997 paperback edition of Duelling Languages--Oxford University Press), I
definitely do not say that the Matrix Language (hereafter ML) can be
identified as the language of the finite verb. For evidence, see the
article by my colleague Jan Jake and me (Jake and Myers-Scotton) in the 1997
issue (the first issue) of the new journal, International Journal of
Bilingualism. We show that in an Arabic/English CS data set, the finite
verb of many, many CPs is in an Embedded Language Island. That is, Arabic
is the Matrix Language, but the IP is all in English (as an Embedded
Language Island).
For example: 'ana I don't think so.
transation: "emphatic I, I don't think so."
(2) I make clear that the ML can be identified on the basis of the two
principles I propose in Duelling Language as structuring a mixed
constituent. That is, the ML is the language of morpheme order (the
Morpheme Order Principle states that only one language provides the morpheme
order for a mixed constituent) and is the language of those system morphemes
that have relations external to their heads (The System Morpheme Principle).
That is, the ML is a theoretical construct used to "name" the source of
abstract grammatical structure in a bilingual CP (mixed constituent).
(3) From the above, it should be clear that the ML is not something to be
identified separately from these two principles. In the original version of
Duelling Languages (1993), I tried to identify the ML separately (with some
sociolinguistic criteria and also by counting morphemes. As I point out in
the 1997 edition of Duelling Languages, counting morphemes doesn't always
work ---although the ML is usually the source of more morphemes, it is true.

(4) The best way to think of the ML is NOT as a "language" itself, but
rather as the grammatical frame. In the case of classic CS, this
grammatical frame is synonymous with the frame of one of the participating
languages, it is true. But in other-than-classic CS, the frame may be a
composite of grammatical material from both (or more) languages involved.
That is, if either because the speaker does not have full psycholinguistic
access to the preferred ML or because the ML is "turning over" (for
socio-political motivations), then the ML is a composite. 
(5) When it is a composite, the ML typically still largely consists of
grammatical material from one of the languages, but the hypothesis is that
it gradually "turns over" to more of the grammatical material from the other
language. This is what happens in language shift. Or, at least it could be
one way that language shift is accomplished.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue