Style shifting, codeswitching

7.2 Stylistic variation in English

Linguists investigating stylistic (or contextual) variation identify a set of
sociolinguistic variables and see how these are realized (what form they take) in
different contexts. There may be a stylistic continuum that is associated with
degree of formality: speakers use more prestige or high-status features in more
formal contexts and more vernacular features in more informal contexts. Speakers,
therefore, draw on different forms of English in different contexts, but there is no
single stylistic continuum: speaking style is better regarded as multidimensional.
STYLE AND AUDIENCE
The theory of audience design (by Allan Bell) suggests that the person or people
you are speaking to will have the greatest effect on the type of language you use.
Bell studied the varieties of English used by newsreaders on radio stations and
found that their pronunciation differed on different stations.
Radio stations might seem to be a special context because speakers are speaking
on behalf of a station rather than in their own voices. In face-to-face interactions
speakers use different varieties of English depending on the person they are
speaking to.
Speakers converge towards the speech of their interlocutors: they try to sound
similar to them. This notion derives from a theory concerned with motivations for
stylistic variation (accommodation theory). This theory suggests that speakers
will converge towards their interlocutors when they wish to reduce social distance,
or get on with one another. They will diverge (become linguistically less similar)
when they wish to emphasize their distinctiveness or increase social distance. It is
plausible to argue that speakers wish to get on with their interlocutors because they
need something from them too.
STYLE AND OTHER FACTORS
Edwards research was conducted as a set of interviews, with informants recorded
in small single-sex groups in a researchers flat. It is set out to construct contexts so
that the influence of different factors may be examined.
It is a quantitative research: linguistic features are identified that can be counted
up to allow numerical comparison between their use by different speakers and in
different contexts. Case studies are also presented, which give a fuller account of
individual speakers use of language.
In the research, whereas in black peer-group conversation one person shows an
overwhelming preference for Patois variants, in the white interview he chooses a
much higher proportion of English variants. While he still uses Patois features, the
number and range of these is very limited.
There is a danger, when presenting a generalized and quantitative account of
speaker style, of seeing speakers as responding somewhat mechanically to context.
One needs to take account of speakers feelings about language and about the
contexts in which they are speaking. This was borne out by Jenny Cheshire in one
study.

Her young informants used fewer vernacular features when they were recorded at
school than when they were recorded in a local adventure playground. But speakers
differed in the extent to which they adapted their speech. She suggests that
differences between speakers have to do with their familiarity with school and
knowledge of school conventions and also with how they feel about school (pupils
who identify with the school culture, or who get on with a teacher, are more likely to
adapt their speech and produce fewer non-standard forms). Speakers, then, are
making their own constructions of context: it is their perceptions of, and feelings
about, people and situations that affect the way they speak. Speakers continually
reassess the context and adjust their speaking style accordingly.
Michael Huspek took into account his informants feelings. From these, he was able
to identify features that varied according to the linguistic context and to the topic
the men were talking about. The speaker switches from an informal to a more
formal register to talk about a scientific topic. His switch was marked by the use of
some -ing forms. This form may also be used in relation to someone who is
respected by the speaker as well as to signal disrespect or resentment.
The -ing variants is recognized as a prestige form, hence it is used when workers
discuss the actions of high-prestige others. But the workers feelings about such
people are ambivalent and so the prestige form does not have entirely positive
connotations.
The quantitative data shows that males use more creole than females. This may be
a result of a higher level of education among the females and a greater acceptance
of standardized forms of English. The qualitative data provide evidence for the need
to take into account speakers feelings: the older males interviewed were more
critical about society than the females or younger males, sometimes showing anger
or frustration regarding white superiors at work and using more creole to define
their identity. Creole-English bilingualism plays a key role in defining identity of its
speakers.
WHAT ELSE NEEDS TO BE TAKEN INTO ACCOUNT?
Speakers vary the way they speak depending, at least, upon:

The
The
The
The

person or people they are speaking to

setting
format of the interaction
topic being discussed

Speakers are not merely responding to predetermined contextual features but to

their own interpretations, or constructions of context. In addition, speaking style
should not be seen simply as responsive.
Prestige and vernacular forms have a cluster of associations. Style is likely to
operate in more than one dimension, allowing speakers access to a more complex
range of social meanings. The same pronunciation feature could convey either
respect or disrespect: you need a certain amount of contextual knowledge to
interpret the use of different linguistic features.
In varying the way they speak, people will draw on a whole set of features,
including those that are less easy to measure. These will combine with, or be
counterbalanced by, other non-verbal features (facial expression).

Some research has adopted a more complex model of style, seeing this as
multidimensional and as representing different, perhaps competing, aspects of
social identity.
MULTISTYLE
R.B. Le Page and Andre Tabouret-Kellers work was carried out in multilingual
communities but their ideas have been influential among researchers with an
interest
Slanging in Singapore
in
Americanised pronunciations occur in several contexts to transmit a Westernised
identity. This adoption of an Americanised accents, as opposed to a Singaporean
accent, is called slanging in Singapore.
monolingual stylistic variation. They suggest that the desire to identify with, or
distinguish from, particular social groups is a major factor in influencing speakers
choice of language variety. But they also allow for fluctuating patterns of usage and
for the fact that speakers may have various motivations to speak in certain ways.

7.3 Switching in and out of English

In bilingual and multilingual communities, style may be expressed by the selection
of one language in preference to another. By opting for English or another language,
speakers may tap a whole set of social meanings with which the language has
become associated. But bilingual speakers need not to keep their languages
separate. They can codeswitch to switch back and forth between languages, thus
capitalizing on the associations of each language.
Codeswitching is generally seen in a positive light by todays researchers as use of
language that is systematic, skilled and socially meaningful. In the past,
codeswitching was indicative of a speakers incomplete control of the language.
Many communities still view codeswitching as a defective use of language.
WHY SWITCH?
Bilingual codeswitching is meaningful: it fulfils certain functions in an interaction.
Myers-Scottons markedness model suggests that particular codes are associated
with, and therefore expected in, particular contexts. Codeswitching itself may be
the unmarked (expected) choice in certain contexts.
A speakers choice of language has to do with maintaining or negotiating a certain
type of social identity. The use of a particular language also gives access to rights
and obligations associated with that identity. It allows speakers (simultaneous)
access to rights and obligations associated with different social identities.
Switching may sometimes operate to initiate a change to relationships, or to make
salient different aspects of the context (a switch to English may communicate
authority).
Codeswitching is useful in cases of uncertainty about relationships: it allows
speakers to feel their way and negotiate identities in relation to others.
Researchers interested in the meaning or functions of codeswitching have tried to
establish social meanings at a very general level: we codes (associated with

home and family) and they codes (associated with more public contexts). In
many bilingual contexts, English functions as the they code because it is often
associated with education, formality and public rather than private arenas. This
suggests a view of meaning as something rather fixed and static. In Kenya, English
can encode both social distance and solidarity, depending on the context. One of
the values of codeswitch is that it permits a certain amount of ambiguity in contexts
and relations between people.
Codeswitching is related to different aspects of a speakers meaning to particular
switches. During mainly creole conversation, a switch to English may be used for an
aside. In contrast, a switch from English to creole marks out a sequence as salient:
it is the part of the utterance that other parties in the interaction respond to.
While speakers use both creole and English at home and among peers to discuss a
range of topics, creole may feel closer to the heart and mind and thus may impart
greater salience to an utterance.
Codeswitching approaches have tended to be used to look at the variable language
use of bilingual or bidialectal speakers where switches are relatively easy to identify.
But in practice the distinction between (bilingual or bidialectal) codeswitching and
(monolingual) style shifting becomes rather blurred.
Quantitative analyses of style and qualitative analyses of codeswitching can be
regarded as different methods, underpinned by different views of what is important
about language, as much as responses to different sorts of data.
SWITCHING AND GRAMMAR
The examples of codeswitching show that, as well as fulfilling a number of social
functions, switching can take a variety of different forms. Speakers may switch from
one language t another at a clause boundary or a long sequence in one language
may be followed by a switch to another. But often switches occur within a clause
and involve a more intimate mix of two or more languages.
Switches are not random: they follow certain patterns and so are subject to
grammatical constraints.
One model called the Matrix Langague Frame (MLF) model argues that within
any stretch of codeswitching one language can be seen as the main, matrix
language. This provides a frame into which items from the other language(s) may
be embedded. It is the grammar of the matrix language that affects the form of
codeswitching. When single words from another language are embedded, the
matrix language word order applies, and the matrix language also supplies the
syntactically relevant morphemes. Content morphemes are distinguished
from system morphemes, which signal grammatical relationships rather than
carrying semantic content. The MFL model predicts that any system morphemes
that signal relations between items in a sentence will come from the matrix
language and be in the surface order demanded by the matrix language (the
morpheme order principle).

Single words that fit in with the morphology of the matrix language have
sometimes been relegated to a separate category of nonce, or one-off
borrowings.
Codeswitching and borrowing
Embedded
items may
Codeswitched items are regarded as belonging to another language, so that
show greater
someone who codeswitches has to have access to two linguistic systems.
or less
Borrowed items are felt to have become part of the matrix language. All
integration
languages have borrowed items. Cultural borrowings enter the language
into the
abruptly as the need for them arises, whereas core borrowings enter
structure of
gradually, via codeswitching: they are subject to the same social
the matrix
motivations and grammatical constraints. As they become used more
language in
frequently, they are on their way to becoming borrowings, sometimes
terms of
displacing original terms. There is a continuum operating between
phonology, syntax and morphology (and that researchers have used such formal
criteria to distinguish between different types of embedded items).

7.4 Designer English?

Qualitative researches of codeswitching highlighted speakers strategic use of
different language varieties and that this was masked in quantitative studies
totaling the occurrence of linguistic variants in different contexts. Speakers are able
to design their speech to take on particular identities.
This suggests that speaker style has its origin in variation between groups of
speakers. Research in English-speaking communities has found that most linguistic
features that show variation do vary among social groups as well as stylistically.
Stylistic variation is always less extreme than social variation. Style variation
derives from and mirrors the social variation. The reflection is less distinct than the
original: style differentiation is less sharp than the social.
MLF model seems to allow a more dynamic relationship between individual
speaking style and established social meanings. She argues that speakers are
aware of patterns of language use that are unmarked or expected in particular
contexts from their experience. They usually choose a speaking style that fits in this
context. But they are being creative in the limited sense of choosing one option, the
unmarked pattern. In so doing, they are helping to re-establish this as normal or
expected. Sometimes they may make a marked choice in an attempt to redefine a
relationship.
The work on bilingual codeswitching discussed here, as well as some studies of
monolingual style shifting have focused on the role of style in managing or
(re)negotiating speakers social identities. Other interpretations have seen style as
primarily a response to an audience: audience design and accommodation theory.
Accommodation theory suggests that speakers will converge towards the speech of
their interlocutor in order to emphasize solidarity and diverge from their
interlocutors speech in order to increase social distance. This is based on the
assumption that convergence will be positively evaluated and divergence negatively
evaluated. But some codeswitching can be interpreted as a way to communicate
authority and divergence.

Bells theory of audience design distinguishes between different types of audience.

It is not only by the person addressed who will affect someones speech but also
others who are involved in the interaction. Bell argues that style is not always
responsive: it may have an initiative function when a speaker switches style to
redefine a relationship. Bell suggests that on such occasions speakers are
addressing their audience as if the audience were someone else. Often speakers are
switching towards a referee, someone not involved in the interaction but who is
nevertheless salient. Speech divergence can be redefined as initiative shifting since
the speaker is not simply diverging away from the addressee but towards another
reference group. Finally, Bell argues that audience design provides a comprehensive
and integrative model of speaker style: other contextual factors that influence
peoples speech may be re-interpreted in terms of audience.
Speakers associate classes of topics or settings with classes of persons. They
therefore shift style when talking on those topics or in those settings as if they were
talking to addressees whom they associate with the topic or setting. The basis of all
style shift according to non-personal factors lies then in audience-designed shift.
Accommodation theory has two developments particularly relevant. It now
recognizes that speakers do not always accommodate to how their addressee
actually speaks. There are obvious limitations, like accents. But, in addition,
speakers sometimes converge towards the variety they expect their addressee to
speak, or that is associated with their addressee, rather than to the variety the
addressee actually speaks.
Difference in status between the speakers is likely to be a factor. But there may be
several reasons why it is more, or less, in a speakers interests to converge. In some
situations, it may be important to maintain aspects of a distinctive identity, but
without necessarily implying hostility. This is called complementarity. It is not
always the case that convergent speakers intend to decrease social distance, nor
that convergence will be positively evaluated.
Finally, what is interpreted as accommodation may be an artefact. This may be
because of a wish to appear in a certain way rather than simply due to a desire to
converge.
Identity-based theories such as the markedness model have different origins, and
different emphases. Speakers do not adopt certain language varieties in order to lay
claim to a certain identity, but this is always in relation to other participants. At a
general level, speakers are taking account both of their own identities and those of
their interlocutors in designing the way they speak. One common feature of the
three theories is that design is seen most frequently as responsive, as when
speakers fall into expected patterns of convergence or complementarity. But
speakers may also make marked or divergent choices in a bid to redefine a
relationship.
To conclude, there is an important question to some of the basic assumptions
made. One of these addresses the nature of explanation: the researchers concerned
have sought to explain why switching occurs where it does; however, does a full
explanation not also need to account for why codeswitching does not occur in
seemingly similar circumstances? Research should, therefore, also recognize that
patterns of switching will differ across communities.