LINGUIST List 6.1320

Wed Sep 27 1995

Sum: How to restrict code-switching inside a constituent

Editor for this issue: Ann Dizdar <>


  1. Stephan Pirsch, Re: How to restrict code-switching inside a constituent

Message 1: Re: How to restrict code-switching inside a constituent

Date: Tue, 26 Sep 1995 23:57:59 Re: How to restrict code-switching inside a constituent
From: Stephan Pirsch <>
Subject: Re: How to restrict code-switching inside a constituent

My question was:

 I'am trying to accord a proposal made by Zanuttini (91) for a phrasal
 category NegP with the 'governmemt constraint' on Code-Switching built up by
 di Sciullo et al. in 1986.
 In Zanuttini (91), NegP selects TP as its complement, i.e. is higher than Infl
 Zanuttini suggests that the verb may raise at S-structure up to the head of
 NegP where the negation marker is located. Inside one constituent (the head
 of NegP) a switch between the negation marker and the verb might not be
 possible. I'didn't find in the literature on Code-Switching any proposal
 which would allow a switch inside one constituent.
 My problem is that I don't know how to explain it, i.e. that a switch
 between negation marker and verb in this configuration is not possible.
 Is it maybe possible to consider the negation marker in this situation as a
 bound morpheme to the verb, so that I could argue with the 'bound morpheme
 constraint'proposed by Poplack (1980)?
 It would be very nice if someone could help me on this problem


1 From: Shahrzad Mahootian <>

Hello! I think i've go tthe solution to your (non)problem: I wrote my
dissertation on codeswitching in 1993 ("A null theory of codeswitching",
Nortwestern University). After exploring all existing
models i developed my own for exactly the same reason that you've come
across: within constituent switching. There are many such examples
throughout codeswitching literature. i propose that codeswitching does
not require special constraints, rather it follows the same rules and
principles that apply to produce monolingual sequences. the basic insight
is this: Heads determine the syntactic properties of their complements in
codeswitching and monolingual contexts alike.

Given the above principle the "rule" for codeswitching becomes:
heads dictate what goes in head-complement structures, and the
complements may be in either language. Therefore in the case of a
NEGP analysis we would predict That Neg would be in language A
leaving its complements to be ineither language A or B. i use
alexicalized tree adjoing grammar (TAG) formalism to generate the
structures which combine to derive codeswitched (and monolingual)
sequences. The model works with any lexicalized formalism of your choice.
the only relevant factor is that the lexical items shoulder the
responsibility of projecting their syntactic requirements and thereby
determine the phrase structure position, syntactic category and feature
content of their complements.Take a simple switch between a determiner
and an NP such as :

in KITCHEN xeyli kaesif-e (Farsi/Eng. Mahootian 1993)
this _______ very dirty-is
`this kitchen is very dirty'

the lexical projections of the structures involved in the derivation are
given below (structures the speaker would have access to).

 a. DP (farsi)	b.	NP(farsi)	c. DP (eng)	d. NP(eng)
	 |			|		 |		 |
	 D'			N'		 D'		 N'
	/ \			|		 / \ 	 |
	D NP			N		D	NP	 N
	|			|	 this		kitchen
	in	 	 ashpaezxune
	(this)			(kitchen)

The codeswitched sequence in the DP in the example above is derived by
combining the structure for the farsi determiner `in' (meaning `this')
shown in (a) with the structure for the English NP `kitchen' shown in (d).
This derivational procedure is called substitution in TAG formalism and is
one of two basic procedures used in the course of deriving monolingual
phrases. The English NP `kitchen' is substituted into the empty NP node in
the farsi DP projected by the determiner `in' (shown in (a)).

Note that the monolingual phrases `this kitchen' and `in ashpaezxune' can
be derived from the same set of structures (a-d).

This was the quick and very general version of how my model accounts for
all codeswitches between bound and free morphemes, and between languages
that are typologically different (for example popstpositional and
prepositional languages such as japanese and english...). If you are
interested in more details i'd be happy to send you a copy of my diss as
well as copies of articles which will be appearing in Lingua and
Linguistics Inquiry.

I would appreciate greatly if you could share with me a some of your
neg-verb switches. i have examples of switches between other verbal
inflections and the verb and have predicted switched between neg-and verb
but have no such examples in my corpus.

shahrzad mahootain


2 From: Edward J Rubin <>

	In your post to linguist, you asked for help explaining the
impossibility of a switch between the functional head NEG and a verb.
You should take a look at the proposal that I made with my colleagues Hedi
Belazi and Jacqueline Toribio in a paper entitled "Code Switching and
X-Bar Theory: The Functional Head Constraint", which appears in Linguistic
Inquiry vol. 25, no. 2. We discuss precisely this situation, as a
sub-case of a more general phenomenon. You might also be interested in
another paper by Jacqueline Toribio and me that appears in Current Issues
in Linguist Theory 123: Contempory Research in Romance Linguistics, ed.
by Jon Amastae, Grant Goodall, Mario Montalbetti and Marianne Phinney,
John Benjamins pub. comp., 1995, which is titled "Feature-Checking and the
syntax of language contact".
	I hope these papers help,
	Ed Rubin
- --------------------------------------------------------------------------

3 From: rakesh bhatt

 In many languages the verb moves via a head-to-head movement (cf.
 Travis 1984, Baker 1988, and more recently Hornstein and Lightfoot 1994).
 In all these cases a lower head is incorporated into a higher head
 (forming adjunction structures). But each head-movement must respect ECP
 ( proper antecedent govt, or one could invoke Baker's Head mvt
 constraint). The point is that some aspect of govt is respected (if
 something intervenes between moving site and moved site--the structure is
 bad). I am sure you know this story, but I thought just in case ...
 Now you can blame the impossibility of switching between verb and Neg
 on the properties of head mvt (and presumably Free morpheme constraint,
 although that has been challenged in Bokamba 1988, among several others).
 Sine the head moves to another head that "governs" it, it may not be
 therefore possible to find a switch sine govt blocks switching (according
 to Disciullio (sp??).

 Rakesh Bhatt
- ----------------------------------------------------------------------------

Thank you very much for the efforts you put into these replies. Your
suggestions helped me a lot with my work.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue