11 Interactional
Sociolinguistics:
A Personal Perspective
JOHN J. GUMPERZ
Introduction: Background
Interactional sociolinguistics
(IS) is an approach to discourse analysis that has its
origin in the search for
replicable methods of qualitative analysis that account for our
ability to interpret what
participants intend to convey in everyday communicative
practice. It is well known that
conversationalists always rely on knowledge that goes
beyond grammar and lexicon to
make themselves heard. But how such knowledge
affects understanding is still
not sufficiently understood.
My perspective on verbal
communication is grounded in earlier work on ethnography
of communication (Hymes 1961);
Hymes’s key insight was that instead of
seeking to explain talk as
directly reflecting the beliefs and values of communities,
structuralist abstractions that
are notoriously difficult to operationalize, it should be
more fruitful to concentrate on
situations of speaking or, to use Roman Jakobson’s
term, speech events. Events are
arguably more concretely available for ethnographic
investigation (Gumperz and Hymes
1964, 1972). They constitute units of interaction
subject to direct analysis by
established empirical means. At the same time, what
happens in such events frequently
enters into public discussion, so that replicable
information on relevant beliefs
and values can readily be obtained through focused
ethnographic inquiry.
transpires in longer sequences
and yields hypotheses on how native speakers think in
everyday interaction. IS is one
of several traditions concerned with these issues.1
To look at talk as it occurs in
speech events is to look at communicative practices.
Along with others I claim that
such practices constitute an intermediate and in many
ways analytically distinct level
of organization. A sociological predecessor here is
Erving Goffman, who proposed the
concept of “Interaction Order” as a distinct level
of discursive
organization bridging the linguistic and the social .
Garfinkel sees interaction as
constituted by goal-oriented moves,
and his main concern is with the
interpretive processes through which interactional
outcomes are achieved. Based on a
variety of illustrative examples taken from what
he refers to as naturally
organized situations, he argues that everyday talk can never
be precise and detailed enough to
convey what is really intended, so that interactants
inevitably and necessarily rely
on what he calls “practical reasoning” and unstated,
taken-for-granted background
knowledge to fill in for what is left unsaid. He goes on
to point out that in so doing
they display a built-in, deeply internalized, and for the
most part unverbalized sense of
social order. Yet apart from advocating that analysts
resort to historical methods to
trace how specific understandings come about so as to
recover what types of knowledge
are at work, Garfinkel gives no further specifics of
how interpretive processes work
in everyday talk.
It is the philosopher Paul Grice
(1989) who lays the foundations for a truly social
perspective on speaking, with his
emphasis on conversational cooperation as a precondition
for understanding. Arguing that
communicating is by its very nature an
intentional process, Grice goes
on to develop a theory of meaning that brackets the
traditional semanticists’ concern
with word-to-world relationships or denotation, to
focus not on utterance
interpretation as such, but on implicature – roughly, what a
speaker intends to
convey by means of a message.
background knowledge enough? We
assume that information about contextual frames
is communicated as part of the
process of interacting, and therefore it becomes necessary
to be clearer about the specifics
of what happens in the interaction as such, to
assess what is intended.
Conversational analysts also set out to do this, and their
work has brilliantly shown what
can be learned through turn-by-turn sequential
analyses. But I suggest that
sequential analysis cannot by itself account for situated
interpretation. It describes just
one of the many indexical processes that affect
inferencing. I argue that
assessments of communicative intent at any one point in an
exchange take the form of
hypotheses that are either confirmed or rejected in the
course of the exchange. That is,
I adopt the conversational analysts’ focus on members’
procedures but apply it to
inferencing. The analytical problem then becomes not
just to determine what is meant,
but to discover how interpretive assessments relate
to the linguistic signaling
processes through which they are negotiated.
1 Diversity as a
Central IS Theme
A main IS theme is the inherent
linguistic and cultural diversity of today’s communicative
environments. Research on the
communicative import of diversity has been
and continues to be plagued by
deep theoretical divisions. On the one hand there are
those who regard communicative
practices as shaped by habitus: embodied dispositions
to act and to perceive the world
that directly reflect the macrosocietal conditions,
political and economic forces,
and relationships of power in which they were
acquired (Bourdieu 1977, 1994).
They argue that it is to such conditioning factors that
we must look for
insights into the nature of diversity
IS seeks to bridge the gap
between these two approaches by focusing on communicative
practice as the everyday-world
site where societal and interactive forces merge.
Hanks (1996) defines
communicative practice as largely resting on the discursive
practices of actors acting in
pursuit of their goals and aspirations. Therefore speaking,
when seen in a practice
perspective, is not just a matter of individuals’ encoding and
decoding messages.
As in-depth, discourse-level
analyses of situated performances became available,
it soon became evident that
speech event categorizations cannot be treated as
extralinguistically defined
givens. More often than not, participants’ definition of
what the relevant event is and
what it means in an encounter emerges in and through
the performance itself (Bauman
1986; Bauman and Briggs 1990; Hymes 1981). As
Hanks puts it in an article on
genre and related questions of language use: “The idea
of objectivist rules is replaced
by schemes and strategies, leading one to view genre as
a set of focal and prototypical
elements which actors use variously and which never
become fixed in a unitary
structure” (1987: 681, quoted in Bauman and Briggs 1990).
The analytical issue thus shifts
from the search for grammar-like rules of language
use as traditionally conceived,
to questions such as (1) how and by what signaling
devices language functions to
evoke the contextual presuppositions that affect interpretation,
and (2) what presuppositions are
at work in particular talk exchanges.
Thus the IS approach to diversity
is essentially a semiotic one, which allows for a
shifting balance between multiple
inputs. Such an approach accounts for the fact that
what count as different systems
at the level of denotational structures can come to
convey information at the level
of communicative structure.
My interpretation is of course
not the only possible one. I relied on background
knowledge acquired through past
communicative experience to infer what was intended.
To the extent that background
knowledge is not shared, interpretations may
differ. What the presuppositions
are that enter into conversational inference and how
they are reflected in talk vary,
among other things, with speakers’ and listeners’
communicative background. Sharing
of inferential procedures cannot be taken for
granted; it must be
demonstrated through ethnographically informed, in-depth analysis of what
transpires in an encounter
They are taken
from a set of selection
interviews recorded in the mid-1970s in the British Midlands.
The applicants are applying for
paid traineeships at a publicly funded institution,
offering instruction in skills
that are in short supply:
(1) Electrician:
a. Interviewer: have you
visited the skills center?
b. Applicant: yes, I did.
c. Interviewer: so you’ve
had a look at the workshops?
d. Applicant: yes.
e. Interviewer: you know
what the training allowance is? do you?
f. Applicant: yeah.
g. Interviewer: Do you
know how much you’ve got to live on for the period
of time.
(2) Bricklayer:
a. Interviewer: have you
visited the skills center?
b. Applicant: yep. I’ve
been there. yeah.
c. Interviewer: so you’ve
had a chance to look around?
and did you look in at the brick
shop?
d. Applicant: ah yeah. we
had a look around the brickshop.
and uhm, it look o.k. I mean
it’s- . . .
e. Interviewer: all right.
f. Applicant: pretty good
yeah.
Conversational inference relies
on two types of verbal signs:
symbolic signs that convey
information via the well-known lexical and grammatical
rules and indexical signs that
signal by direct association between sign and context.
Terms like “here” and “there” or
“this” and “that” are typical examples of indexicality,
in that what is intended in any
one instance can only be understood with reference
to some physical or discursive
environment. But context also can be and often is
communicatively evoked through
talk, and it is that evocation process that is at work
in code-switching.
Finally, and perhaps most importantly,
indirect (not overtly lexicalized) signaling
mechanisms are for the most part
culturally or subculturally specific. In fact prosody
and “accent” (in the sense of
phonetically marked features of pronunciation), for example,
are among the principal means by
which we identify where people are from and
“who” they are, and assess their
social identity, as happened in the above examples.
. Apart from focusing on
interpretations as such, IS analysis attempts
to illustrate how these tasks are
accomplished. It is for this reason that the analysis
places so much stress on
contextualization processes.
2 IS Method
In empirical studies, IS analysts
have worked out a set of procedures along the
following lines. First there is
an initial period of ethnographic research designed to
(1) provide insight into the
local communicative ecology; (2) discover recurrent
encounter types most likely to
yield communicative data relevant to the research problem
at hand; and (3) find out through
observation, interviewing key participants, and
checking one’s own interpretations
with them how local actors handle the problems
they encounter and what their
expectations and presuppositions are. In the second
stage, the ethnographic findings
provide the basis for selecting events reflecting representative
sets of interactions for
recording. (4) The next phase of the analysis begins
with scanning the recorded
materials at two levels of organization: (a) content and (b)
pronunciation and
prosodic organization.
Once isolated, events are
transcribed and interactional texts (that is, transcripts that
account for all the
communicatively significant, verbal and nonverbal signs perceived)
(Silverstein 1992) are prepared
by setting down on paper all those perceptual cues:
verbal and nonverbal, segmental
and nonsegmental, prosodic, paralinguistic, and
others that, as past and
ongoing research shows, speakers and listeners demonstrably
. In the first extract, the
questioning is
designed to test the applicant’s
knowledge of the course:
(3) a. Interviewer: and
you’ve put here, that you want to apply for that course
because there are more jobs in .
. . the trade.
b. Applicant: yeah (low).
c. Interviewer: so perhaps
you could explain to Mr. C. apart from that reason,
why else you want to
apply for electrical work.
d. Applicant: I think I
like . . . this job in my- , as a profession.
e. Instructor: and why do
you think you’ll like it?
f. Applicant: why?
g. Instructor: could you
explain to me why?
h. Applicant: why do I
like it? I think it is more job prospect.
By using stress to foreground the
word “trade” the interviewer is drawing the
applicant’s attention to the term
the applicant used in the written questionnaire he
filled out before the interview,
relying on him to infer what she intended to convey
by this strategy.
The consequences of the
miscommunication that results become clear in the following
segment, when the instructor
turns to the topic of the applicant’s previous experience
with electrical work:
(4) i. Instructor: what
sorts of work have you done before in this particular field?
j. Applicant: what do you
mean please.
k. Instructor: well,
electrical installation and maintenance. some of it involves
jobs done in your home. in your
own home have you done work
in your own home?
l. Applicant: yes sir.
m. Instructor: yeah, and
what sorts of jobs have you done?
n. Applicant: well I-, I
wired up my own house.
o. Instructor: you’ve
wired your own house?
p. Applicant: yeah.
q. Instructor: yeah?
r. Applicant: it is
passed, by the authority, electricity board.
s. Instructor: yeah?
t. Applicant: first time.
u. Instructor: so having
wired your own house, could you tell me what the
“consumer box” is?
v. Applicant: yeah, where
the fuses is.
w. Instructor: where the
fuses are. all right fine. have you done anything other
than wiring your own house?
The three principals in this
example have lived in the region for over a decade and,
apart from the Asian’s accent and
minor grammatical oddities, they all speak English
well. Moreover, they agree on
what a selection interview is about and understand
what is being said at
the level of literal or denotational meaning.
By revealing the underlying
interpretive process at work in an encounter, which is
otherwise bound to remain hidden,
IS analysis of key situations in institutional life
can provide insights into the
interpretive and ideological bases of communicative
assessments, while at the same
time enabling participants to learn from some of the
difficulties arising in their
contacts with others.
3 Conclusion
The intercultural encounters I
have discussed constitute an extreme case where participants
represent historically and
linguistically quite distinct traditions. All the participants
had lived and worked in western
industrial settings for much of their adult
life, but they brought into that
different linguistic and cultural background experiences
which continue to
resonate in these encounters.
As some of the shorter examples
cited above indicate, IS analysis is
applicable to communicative
situations of all kinds, monolingual or multilingual, as a
means of monitoring the
communication processes that are so important in institutional
life.
NOTES
1 For other related approaches see, for
example, Bauman (1986); Briggs (1996);
Fairclough (1995); Guenthner (1993);
Hill and Irvine (1993); Kallmeyer
(1994); Sarangi and Roberts (1999);
Sherzer (1983); Silverstein and Urban
(1996); Tannen (1984, 1989); Young
(1994).
REFERENCES
Auer, P. ed. 1998. Code-Switching in
Conversation. London: Routledge.
Bauman, R. 1986. Story Performance and
Event: Contextual Studies of Oral
Narrative. Cambridge: Cambridge
University Press.
Bauman, R. and Briggs, C. 1990. Poetics
and performance as critical
perspective on social life. Annual
Review of Anthropology.
Blom, J. P. and Gumperz, J. J. 1972. Social
knowledge in linguistic structures:
code-switching in Norway. In
J. J. Gumperz and D. Hymes, eds,
Directions in Sociolinguistics: The
Ethnography of Communication. New
York: Holt, Rhinehart and Winston,
pp. 407–34.
Briggs, C. ed. 1996. Disorderly Discourse,
Narrative Conflict and Inequality.
Oxford: Oxford University Press.
Bourdieu, P. 1977. Outline of a Theory of
Practice. Cambridge: Cambridge
University Press.
Bourdieu, P. 1994. Language and Symbolic
Power. Cambridge: Polity.
Brown, P. and Levinson, S. C. 1979. Social
structure, groups and interaction. In
H. Giles and K. Scherer, eds, Social
Markers in Speech. Cambridge:
Cambridge University Press.
Cook-Gumperz, J. and Gumperz, J. 1994
