Applications of the MLF to Spanish-English data
A lthough there have been many studies adopting the approaches of other
frameworks (e.g. Herring et al., 2010; Pfaff, 1979; Poplack, 1980; Timm, 1975;
among many others), the study of Spanish-English CS within the MLF
framework had not previously received much attention. To our knowledge
the only other research in this area has been conducted by Smith (2006),
who reports that in his Spanish-English corpus collected from Latin
American immigrants in Georgia, the most common type of bilingual clause
in the corpus is a Spanish clause with a single English lexeme insertion, so
that the ML of such a clause is identified as Spanish. English clauses with a
single Spanish morpheme insertion are much less frequent. Smith argues
that as the bilingual Spanish/English sentences reach a certain threshold of
English content, Spanish morphosyntactic patterns begin to converge toward
those of English. Smith analysed data from naturally occurring conversa-
tions by 56 children and adults of both genders. Smith (2006: 209) makes use
of the concept of ‘community ML’, which he appears to identify unquestion-
ably with the community’s first language (Spanish) as opposed to the lan-
guage acquired by immigrants in their host country (English).
His analysis revealed that percentages of monolingual and code-switched
utterances pattern in relation to the percentages of utterances showing con-
vergence. This indicates that the informants’ Spanish does not begin to con-
verge toward English until fewer than 70% of their utterances are
monolingual Spanish. When fewer than 40% of their utterances are mono-
lingual Spanish, English is more frequently used than Spanish and English
sentences show more convergence (towards Spanish) than vice versa.
Those who use more monolingual English and who code-switch more
and converge more are more likely to be in one or more of the following
social categories: younger age, female, higher socioeconomic status, more
time spent in the US and in US schools, more reported preference for and
use of English. Those who use more monolingual Spanish and who code-
switch and converge less are more likely to be in one or more of the follow-
ing social categories: older age, male, lower socioeconomic status, less time
spent in the US and in US schools, less reported preference for and use of
English. While Mexicans as a group code-switch more than non-Mexicans,
and non-Mexicans converge more than Mexicans, still it is the younger
Mexicans who do most of the CS and younger non-Mexicans who do most
of the converging.
The infl uence of speaker-based variables
Apart from the work of Smith (2006), outlined above, no studies to our
knowledge have examined the effects of sociolinguistic variables specifi-
cally on ML distribution. Although Finlayson, Calteaux & Myers-Scotton
(1998) used the MLF model as part of their study on CS in South Africa,
118 Part 3: Bilingual Language Use