Andersen, Gisle (2001) Pragmatic Markers and Sociolinguistic Variation:
A Relevance-Theoretic Approach to the Language of Adolescents. John
Benjamins Publishing Company, hardback ISBN 1-158811-018-4 (US),
90-272-5103-7 (Eur), ix+354pp, $95.00, Pragmatics and Beyond NS 84.
Janet M. Fuller, Linguistics Department, Southern Illinois University
This book is aimed at an audience of researchers on
topics in pragmatics and sociolinguistics, and seeks to
fill a gap between research on pragmatic markers and work
on sociolinguistic variation. Andersen explicitly frames
her analysis in relevance theory and theories of
grammaticalization; implicit in her work is the general
variationist framework of correlating linguistic features
with social variables. Using a corpus of London adolescent
speech, and another of adult speech for comparison, she
examines two sets of features: first, "innit"/"is it" as
invariant tags and follow-ups, and second, the particle
"like" as a pragmatic marker with a variety of functions.
She presents evidence that both of these are found
primarily in adolescent speech, with some variation
according to gender, ethnicity, social class and location
of residence, and suggests that they are both undergoing
grammaticalization.
The format of the book is as follows. Chapter 1 is a
general introduction; chapter 2 is the theoretical
background applied in the analysis. Chapter 3 presents the
data and methods, and chapters 4 and 5 contain the
analyses. Chapter 6 is a brief summary.
The theoretical background chapter covers three topics:
relevance theory, grammaticalization theory, and a
discussion of what pragmatic markers are and how they fit
into these frameworks. The discussion of relevance theory
is an excellent resource in and of itself for those
interested in this theory, as Andersen presents the
essential points of relevance theory in a clear and concise
manner. The principle of relevance, which states that
speakers are assumed to produce just enough contextual
effects to be worth processing, is the backbone of this
theory; it is fleshed out with reliance on the "cognitive
environment", which allows speakers to underspecify their
utterances with regard to propositional meaning, and
hearers to interpret utterances according to a presumption
of optimal relevance.
The discussion of grammaticalization focuses on
a unidirectional cline from referential (propositional) to
non-referential meanings, meaning that lexical items come
to operate on textual and interpersonal levels. In
particular, pragmatic markers (also called "pragmatic
particles", "discourse markers" or "connectives")are
lexical items which have undergone, or are undergoing, such
a process. In general, pragmatic markers have a low degree
of lexical specificity and a high degree of context
sensitivity. As such, they are often claimed not to
contribute to propositional meaning. Andersen's discussion
of this feature of pragmatic markers is very valuable, as
she illustrates that not all pragmatic markers are outside
of propositional meaning of the utterances they modify. As
this is an often-claimed trait of pragmatic markers, this
discussion -- which shows that non-propositionality is a
frequent attribute of pragmatic markers, but not a
determining features -- is an important contribution to the
study of pragmatic markers in general. Andersen ties in
this argument to her discussion of grammaticalization by
suggesting that pragmatic markers with a lexical history
which have not been fully grammaticalized (e.g., "like",
"sort of", you know") may be problematic for the
propositional/non-propositional dichotomy, while others
which are fully grammaticalized fulfill this criterion.
As a final component of this chapter, Andersen ties in
relevance theory to the study of pragmatic markers.
Essentially, pragmatic markers function to indicate speaker
attitude and expectations of mutual manifestness of
propositions (also called "common ground" within other
approaches) and provide hearers with cues to correctly
interpret utterances. Andersen breaks down pragmatic
meaning into three basic aspects: subjective, which
describes the relation between the speaker and the
communicated proposition or assumption; interactional,
which can be either speaker or hearer oriented; and
textual, which contributes to and express coherence
relations.
The next chapter describes the data used for the
analysis. She analyzes adolescent speech from the Bergen
Corpus of London Teenage Language (COLT), which contains
roughly 100 hours of conversation of teens, recorded by 30
teenage "recruits" in natural settings. The recruits
responsible for recording the data provided information
which allowed categorizations of age, gender, social class,
and location of residence. Unfortunately, only the data
from the recruits themselves can be categorized according
to social class, and ethnic group membership was not
elicited from even the recruits. Because many of the
London boroughs in which the data were collected are areas
with great ethnic diversity, it is assumed that this
diversity is reflected in the conversations. Post hoc
information from the fieldworker and from the content of
the corpus reveal the ethnicity of some of the speakers,
and these speakers could be grouped into the general
categories of "White" and "Ethnicity minority". This less
than detailed demographic information about the speakers in
this corpus is the one drawback of this study. The lack of
a consistent classification criteria for the social
variable examined limits the findings on sociolinguistic
variation to a more speculative level than is desirable in
a study with this focus.
In addition to the COLT data, a subset of the British
National Corpus (BNC) containing adult speech is used for
comparison. Although there are problems with this
comparison -- the adult data are not given by location of
residence, and contain little ethnic variation -- it
provides some comparative basis from which to view the
findings on the COLT data.
Chapter 4 presents an analysis of "innit" and "is it"
as invariant tag questions and invariant follow-ups. Because
only paradigmatic use of these phrases are found in the
adult corpus, they appear to be undergoing grammaticalization
in adolescent speech. In the COLT corpus, all of these forms
co-occur with the use of canonical tags and follow-ups,
overall and in the speech of individuals.
"Innit" as a tag shows great flexibility in both form
and function -- much greater than in canonical tags -- and
comprises 26.8% of all tags used. There is little evidence
for invariable "is it" as a tag. The linguistic contexts
that favor the use of "innit" as a tag appear to be
contexts in which the canonical equivalent would be a
negative polarity tag in the third person, especially with
the present tense of "be"; use of this invariable tag is
also favored when the canonical tag realization would
involve a trisyllabic or syntactically awkward tag (e.g.,
"weren't they", "mightn't I").
As a follow-up, "innit" functions as a marker of
contextual alignment (what Andersen calls an "A-signal")
and appears in 100% of the contexts where this is
expressed; no canonical follow-ups are used in this
context. "Is it" as a follow-up functions to register
surprise and disbelief (i.e. divergence) following from the
previous speaker's utterance (a so-called "D-signal"), and
appears in 32.5% of the contexts, in variation with
canonical follow-ups.
Socially, the tag "innit" is an adolescent feature
which is found more among female speakers, low social class
members, and ethnic minorities, but the strongest
correlation is with the residential location, indicating
that it is as central phenomenon rather than a peripheral
one.
Finally, Andersen proposes the following diachronic
development for "innit": its use begins in third person
singular neuter contexts, it later becomes a tag throughout
the inflectional paradigm, and then comes to be used as a
follow-up.
Chapter 5 presents an analysis of the pragmatic
functions and sociolinguistic variation pragmatic marker
"like" in its. Andersen claims that "like" marks non-
literal resemblance between an utterance and its underlying
thought, and she links the development of these functions
of "like" to non-pragmatic marker uses, which have similar
semantic properties, through the process of reanalysis.
The marking of non-literal resemblance of the pragmatic
marker "like" includes such commonly reported functions as
marking approximation, introducing examples, and indicating
vague expressions as well as introducing quotations. While
"like" as a pragmatic marker is often outside of the truth
conditional meaning of an utterance, it is not always non-
truth conditional, as phrases such as "you wrote like four
sides" clearly differ truth conditionality from the
proposition "you wrote four sides". Yet even when it
contributes to the truth-conditionality of an utterance,
"like" also has a procedural function, i.e., it indicates
to the hearer the speaker's alignment with the element it
frames.
An analysis of the placement of "like" indicates that
it can modify anything from whole propositions to single
terms, and be syntactically bound (i.e., clause-internal)
or unbound. In these data, it is bound about two-thirds of
the time. It is less likely to occur within phrases with
high syntactic fixedness, and more likely to occur
immediately before the lexical material of a phrase, as
opposed to the grammatical words.
Socially, "like" is favored by adolescents, although it
is found in the speech of those in their 30s and 40s. It
has been primarily adopted by adolescent girls in their
late teens in the COLT data, and although it is used by
speakers of all social classes, is used at a significantly
higher rate in the high class group than the middle and low
groups. It is also primarily a feature of white adolescent
speech, with no clear pattern in terms of location of
residence, although it is less favored in areas where there
are many ethnic minorities.
The final chapter of this book is quite brief, and is
divided between the subject of age-grading and suggestions
for further research. Although Andersen has, throughout
the book, argued for a language change analysis for the
linguistic features she analyzes, she concedes in this
chapter that age-grading must not be ruled out. This leads
smoothly into the section on suggestions for further
research, a major aspect of which is a plea for more
research on adolescent-specific language use.
Despite its one drawback -- the limitations for
conclusions about sociolinguistic variation due to the
inadequate demographic information about speakers in the
main corpus -- this book is a worthwhile read. I applaud
her use of theoretical frameworks for her analysis, and the
two features she has examined are well-deserving of
attention. Her contributions on the feature of invariable
"innit" are largely unique, as this feature has not been
previously studied in great detail. The discussion of
"like" -- a more traveled topic in pragmatics -- contributes
to the study of this pragmatic marker by analyzing its
function within a pragmatic theory and linking linguistic
and social development of this particle. This chapter on
"like" is, in my opinion, the highlight of the book, as it
most successfully links the theoretical background and
empirical findings in a clear and straightforward analysis.
Janet M. Fuller is an Assistant Professor of Linguistics at Southern
Illinois University at Carbondale. Her research interests lie in
the areas of sociolinguistics, bilingualism and language contact,
discourse markers, and language and gender.
|
