MacSwan, Jeff (1998) "A Minimalist Approach to
Intrasentential Code Switching", Garland Publishing;
Outstanding Dissertations in Linguistics
Reviewed by Cecilia Montes-Alcala, UCSB.
INTRODUCTION
This dissertation addresses grammatical aspects of
intrasentential code switching and their relevance to
education. It is organized in six chapters, the first one
serving as a general introduction to the field, as well as
setting the unifying thesis of the work. Chapter 2 reviews
the relevant literature in bilingualism, code switching,
syntactic theory, and studies on Nahuatl and Spanish.
Chapter 3 addresses the research design, and chapter 4
presents the findings. Chapter 6 deals with educational
policy and teaching in bilingual education.
For the purposes of this review, I focus on chapter 5,
which constitutes the core findings of the dissertation. The
basic claim is that in the spirit of minimalism nothing
constrains code switching, apart from the requirements of
the mixed grammars.
Code Switching on Minimalist Assumptions
Adopting the Minimalist approach to syntax(Chomsky, 1995),
MacSwan states that "nothing constrains code switching apart
from the requirements of the mixed grammars." This would
entail that we ignore the differences between particular
languages for the purposes of linguistic theory, and the
language-specific requirements would be represented in
morphology (parametric variation.) In minimalist terms, a
conflict in language-specific requirements is just a
conflict of lexical features. The computational system
selects items from one or the other lexicon, or it can
select items from both lexicons, and then we would have a
code switching sample. Although much work has been devoted
to proposing grammatical clashes for code switching, MacSwan
pursues an explanation in terms of conflicts in the lexical
requirements, rather than code switching-specific
mechanisms.
The Spanish-Nahuatl Corpus.
MacSwan applies a number of previous approaches to his
Spanish-Nahuatl corpus and shows them all to be lacking.
Among these, Poplack's (1980, 1981) Free Morpheme
Constraint, and Equivalence Constraint; Joshi's (1985)
constrain on closed-class items; Di Sciullo, et al.'s (1986)
anti-government condition; Mahootian's (1993) approach; and
Belazi, et al.'s (1994) Functional Head Constrain are all
refuted based on numerous counterexamples found in the
corpus. MacSwan concludes by stating that all these
proposals are empirically incorrect, and the analysis
certainly flawed.
Language specific differences in functional categories
explain some properties of code switching: in particular
constructions with pronouns and agreement morphemes. A
switch between a Spanish pronoun and a Nahuatl verb may
occur for third person (which has a null subject agreement
morpheme in Nahuatl), but not for first or second persons
(which are not null.) MacSwan arrives at the conclusion that
Nahuatl NP's must be (usually) arguments. Nahuatl pronouns
and DP's do not overtly mark case or gender distinction,
while Spanish pronouns and DP's have morphological marking
for both. There is thus a mismatch of Nahuatl verbs, which
have weak features, with Spanish verbs, which have strong
features. Similarly, there is a gender mismatch: Nahuatl has
no overt gender marking, while Spanish gender is two-valued.
As it appears, then, code switches between DP's and
predicates in languages with like gender systems should be
allowed, otherwise disallowed. Code switching between a
Spanish verb and a Nahuatl direct object is disallowed
unless a Spanish clitic doubles the object. Under the
analysis presented here, no Spanish subject or object may
occur in a construction with a Nahuatl verb bearing a
subject agreement morpheme, and no Nahuatl DP's are allowed
with Spanish verbs.
With respect to embedded clauses, Spanish verbs of
speaking may take a Nahuatl CP complement and vice versa,
but switched IP complements are always ill-formed. Much
attention has been given to V-V sequences. The conclusion
seems to be that languages cannot be switched in V-V
compounds. MacSwan proposes that this is due to his PF
Disjunction Theorem, which bars code switching within a PF
component.
For durative constructions, a switch between Spanish
auxiliary (estar) and a Nahuatl durative is allowed only
when the latter does not have inflectional material. Note
that Nahuatl does not employ auxiliaries before present
participles like Spanish (MacSwan assumes a null copula.)
For negatives, the data shows that a switch between a
Spanish negation and a Nahuatl verb is unacceptable, but a
switch between a Nahuatl negation and a Spanish verb is
allowed. MacSwan explains this by assuming that the Spanish
no is a clitic (like French ne). Under this assumption, the
property of Neg would attract V, and therefore the PF
Disjunction Theorem would bar a switch here, while
constructions with the Nahuatl negation are allowed because
it does not attract V.
Gender features in DP's are also examined. A Nahuatl
determiner before a Spanish noun is well-formed, but not
vice versa. Baker (1996) argues that Nahuatl has no "true
determiners", but these are rather adjuncts to NP, and that
explains their flexibility regarding word order. Movement of
N to D is seen in Spanish, but not in Nahuatl. Thus, no
problem arises if a Spanish N does not check its features
with a Nahuatl D, but if a Spanish D attracts a Nahuatl N
the construction will be ill-formed because of a gender
feature conflict, and a violation of the PF Disjunction
Theorem. An interesting fact also is that constructions with
Spanish feminine D's are worse than those with masculine
D's, because Spanish has masculine gender as the default
form, and it is more acceptable with the Nahuatl null gender
system. There are some counterexamples involving the verb
"have" for which MacSwan gives no explanation.
MacSwan concludes that all the samples could be
analyzed in terms of mechanisms independently motivated for
the analysis of monolingual data and, therefore, code
switching phenomena can be explained without appealing to ad
hoc constraints specific to code switching. The underlying
assumption is that those do not exist, and once again,
nothing constrains code switching apart from the
requirements of the mixed grammars.
MacSwan also considers some conflicting findings in
other code switching corpora and develops similar
conclusions.
Conclusions
The dissertation concludes that there are no code
switching-specific constraints. The Nahuatl-Spanish data
presented here has been analyzed in terms of principles
motivated to explain monolingual data without specific
reference to the bilingual phenomena. This leads to the
conclusion that nothing constrains code switching apart from
the requirements of the mixed grammars. Furthermore, code
switchers have the same grammatical competence as
monolinguals for the languages they use, although MacSwan does
not explain what a "native bilingual code switcher" really is,
but this has important potential implications for
educational policies and teaching.
CRITIQUE
As we have seen, much of the literature on code
switching regards this phenomenon as something isolated
from, or in opposition to, monolingualism. The advance
proposed in this dissertation is its claim that other theories
of codeswitching do not account for all the empirical data,
and moreover, that code switching data can be explained in
the same terms as monolingual data. Without necessarily
agreeing with this claim in absolute terms, I must admit
this statement constitutes a big step towards a better
understanding of this natural phenomenon, and a good point
of departure from the social stigma that code switching has
always carried. The last chapter on educational policy
offers a new insight in using code switching as a tool for
learning, rather than as an obstacle for bilingual
education, which is certainly commendable. I have no objections
regarding the basic thesis of the dissertation, nor with its
implications.
However, I do have certain minor problems in relation to the
methodology used in the data. More specifically, regarding
data collection, we can pose the following questions: is any
data collected valid to refute the proposed theories? How
can one distinguish what MacSwan calls a "native bilingual
code switcher" from a non-native one?
However, all in all, this book constitutes a rather
original piece of work which contributes to a better
understanding of code switching within the framework of the
Minimalist Program.
The Reviewer:
Cecilia Montes-Alcala is a Ph.D. candidate in the
Linguistics Program of the Spanish & Portuguese Department
at the University of California, Santa Barbara. Her main
research interests are Sociolinguistics (bilingualism,
code-switching) and Applied Linguistics (second language
acquisition.)
|
