LANGUAGE IDEOLOGIES

Practice and Theory

Edited by
BAMBIB. SCHIEFFELIN
KATHRYN A. WOOLARD
PAUL V.. KROSKRITY

New York Oxford

OXFORD UNIVER ITY PRESS
1998
Oxford University Press
Oxford New York
Athens Auckland Bangkok Bogota Bombay
Buenos Aires Calcum Cape Town Dar cs Salum
Ddhl Florence I long Kong uianbul lunchi
Kuala Lumpur M:i.dru Jl.hdrid Melbourne
MexicoCity Nairobi Paris Singapore
Taipei Tokyo Toronto Wanaw

and associated companitt in

Berlin Ibadan

Copyright© 1998 by Oxford University Press, lnc.

Published by Oxford Universu:y Press, Inc.
198 Madison Avenue, New York, New York 10016
Oxford is a registered trademark of Oxford University Press

All rights reserved. No pan of this publication may be reproduced,

stored in a retricwl sy.tem, or transmitted, in any form or by any means,
clcett0nic, mechanical, pho<ocopymg, recordmg, or othcrwue,
without the prior pcrmisSion of Oxford University Press.

Library ofCongrc.,Cmloging-in-Public,nion Data

Language ideologies: prmice and theory I edited by
Bambi B. Scluclfclin, l<Jthryn A. Woolard, ind Paul V. Kro.lcri ty.
p. cm. - (Oxford ,rudic, in anthropological linguisticr, lb)
lndudcs b,bliographical rcfcrcnccs.
ISBN O-t9-SIOS61-3; ISBN 0-19·Sl0S62·l (pbk.)
I Language and culture. 2. Ideology. I. Schicffelin, Bamba B.
U Woolard, Kathryn Ann. ITI. Krosknty, Paul V., 1949-
IV. ScriCL
P3S.L333 1998
400-dc.2 I 9723336

9 87654321

Printed in the United States of Amcrio:u

on acid-free pllper
LANGUAGE IDEOLOGIES
 1
Introduction
Language Ideology as a Field ofInquiry

KATHRYN A. WOOLARD

As Raymond William ob erved, "a definition oflanguage is always, implicitly or

explicitly, a definition of human beings in the world" (1977:21). The essays in tlus
volume examine definitions and conceptions of language in a wide range of set­
tings. They fo us on how such defining activity organizes individuals, institutions,
and their interrelations. Representations, whether C."<plicit or implicit, that construe
the intersection oflanguage and human beings in a social world are what we mean
by "language ideology. '
There is as much cultural variation in ideas about language and about how
communication works as a social process as there is in the very form of language
(Bauman 1983:16; Hymes 1974:13-14, 31). However, language ideology is of
anthropological irnportan e not simply because of its ethnographic variability but
because it is a mediating link between social forms and forms of talk (if I may be
forgiven a turn of phra e that emphasizes product over process). As all of the con­
tributions to this volume point out, ideologies oflanguage are not about language
alone. Rather, they envision and enact ties oflanguage to identity, to aesthetics, to
morality, and to epistemology. Through such linkages, th ey und erpin not only lin­
guistic form and u e but al o the very notion of the person and the social group, as
well as such fundamental social institutions as religious ritual, child socialization,
gender relations, the nation-state, schooling, and law.
Although efforts have been made recently to delimit language ideology, there
is no single core liter ature, and there are a number of different emphases. Linguis-

J
 INTRODUCTION
4

most �roadly as "!hared bodies o�

tic or language ideologies have been defined
e 10 the world (Rumsey 1990.
commonsense notions about the nature oflanguag
a�d on the �crivi:t ."a.cure of
346). With more emphasis on linguistic structure
defines lin�us�c 1�eolo­
ideology (to be discussed later in my essay), Silverstein
as a ranonalizat1on or
gies as "sets of beliefs about language articulated by users
:193). On the other hand,
justification ofperceived language soucture and use" (1979
ID' has been defined �s
with a greater emphasis on the social facet, language ideo!O
of language m
"self-evident ideas and objectives a group holds concenung roles
ofthe group"
the social experiences ofmembers as they contribute to the expre. �ion
cultural system of ideas about .ocial and linguistic
(Heath 1989:53) and as "the
relationships, together with their loading of moral and political interest s" (Irvine
1989:255).
I use the terms "linguistic ideology," "language ideology," and "ideologies oflan­
guage" interchangeably in this essay, although differences amo ng them can be de­
tected in separate traditions ofuse and are sometimes evident in the contributions to
this volume . At least three scholarly discussions explicitly invoke "linguistic ideol­
ogy" or "language ideology," sometimes in seeming mutual unawarenes:;. One �ig­
nificant, theoretically coherent body ofwork originates in linguistic anthropology and
concentrates on the relation of"linguistic ideology" to linguistic structures. This lit­
erature centers on Michael Silverstein's concept of rnerapragmatics, which encom­
passes implicit and explicit commentary on and signaling about language-in-use
(Silverstein 1976, 1979, 1981, 1985, 1993).1 A second focal area is contact between
languages or langua� varieties, and on this topic sociologists oflanguage and educa­
tionists , as well as linguists and anthropologists, have offered considerations of "lan­
guage ideology" (e.g., Heath 1989, 1991; Hornberger 1988a; Sonntag and Pool 1987),
�purist ideology" (Hill 1985; Hill and Hill 1980, 1986), and .. ideologies of
standard(ization)" (Milroy and Milroy 1985). Finally, the recently burgeoning his­
toriography ofpublic disc ourses on language has produced an explicit focus on "ide­
ologies of language," including the scientific ideol ogies of professional linguistics
Ooseph and Taylor 1990).
Beyond research that explicitly invokes the term "ideology" are countless studies
that addres� cultural conceptions oflanguage, in the guise ofmetalinguistics, atci­
tu�es'. presnge, standards, aesthetics, and � on. The field could profit from a re­
?"nking ofmuch of this material within an explicitly social-theoretical frame of
ideology analysis .
The point. of the comparative study of language ideology is to examine the
cultural and historical specificity ofconstruals oflanguage, not to distinguish ide­
ology oflanguag e from ideology in othe r d
omains ofhuman activity . Exclusion of
some cultural conceptualizations because language is not sufficie
ntly focal in them
wo�� h«; an ironic outcome to this attempt to denaruralize
our own intellectual
tradinon s companmentalizaa·on and re·fi . .
1 ca a· on of commurucative soo.u . _ 1 pracn•ces
•
(see Bnggs's and Kulick's chapters lOr funher comment: see also Lucy 1993a·
. c:
Mannheim 1986· Rumsey 1990) 0ur hope .1s not to resrnct '. vision
. . • but to focus
. ' · '
the a e ion °.fsch olars of language on the unavo
. � t idable significance of the ideo­
.1 g ic imens�on , as well as to provide firme
� r linguistic ground for students of
1 cology and dJscourse more generally.
Language Ideology as a Field ofInquiry 5

In this introductory essay, I first revi:w the general concept of ideology and
traditions ofits analysis. This is not a comprehensive overview of the immense lit­
erature on ideology, rather only a rehearsal of some of its recurrent themes, in
order to situate this newly coalescing field oflinguistic inquiry and to point out the
promise as well as the pitfalls.2 I then turn to literatures on language ideology in
particular, illustrating and reviewing a spectrum of approaches to cultural concep­
tions of language and of communicative behavior as an enactment of a collective
order (Silverstein 1987:1-2). Throughout the discussion I try to situate the chap­
ters of this volume, but I leave the extended discussion of each contribution to the
commentaries invited from Michael Silverstein, Susan Philips, and Susan Gal.

What Is Ideology?

The word "ideology" is associated with a confusing tangle of commonsense and

semitechnical meanings (Friedrich 1989:300). As Silverstein discusses in his com­
mentary here, the term was first coined at the end of the eighteenth century by the
French philosopher Destutt de Tracy, a follower of Condillac who optimistically
hoped to develop a science of ideas and their basis in sensation. Destutt de Tracy
envisioned this positive science as a branch of zoology, one that not only would
allow the complete understanding of the human animal but that might ultimately
also serve the Enlightenment _E_roject of the rational regulation of society.
The term was soon given its negative connotation in Napoleon's effort to dis­
credit Destutt de Tracy and his colleagues, whose institutional position and work
were tied to republicanism. In Napoleon's use, ideology became "mere," and "ideo­
logue" a dismissive epithet for proponents of abstract theories not based in or ap­
propriate to human and political realities.
While Destutt de Tracy's intended meaning of a "science of ideas" has been
thoroughly abandoned, and Napoleonic negativism has proved quite enduring, there
is still significant variation even among social scientific meanings of the term. As
Eagleton says, the word "ideology" is itself a text, woven of a tissue of conceptual
strands (1991:1). In contemporary uses, several strands recur with particular sa­
lience. Although none of them is universal, and none is untroubled or untroubling,
I wis_h to single out four such recurring strands or themes that are picked up by
many if not all of the contributors to this volume.
1. The first common strand is an understanding of ideology as ideational or
conceptual, referring to mental phenomena; ideology has to do with conscious­
ness, subjective representa1ions, beliefs, ideas. Like Destutt ae Tracy's, some con­
temporary social scientific uses of the term focus almost entirely on this ideational
aspect, and when branding a phenomenon as "ideological," they do not intend the
social or critical dimensions that are discussed later in this essay.In the broadest
instance, ideology is taken to be the more intellectual constituent of culture, "the
basic notions that the members of a society hold abouta fairly definite ... area
such as honor, ... the division oflabor" (Friedrich 1989:301)-or, we could pro­
pose, language. But even more sharply focused and critical analyses ofideology often
share this emphasis on ideation and even explicit verbalization. In the sociologist
 INTRODUCTION
6

Alvin G ouldner's communication-based schema, for example, ideologies are dis­

cursive, rationalist "rep orts about the world" (197�:3,,1), or as J.B. Thompso n has
put it, "that part of consciousness which can� s�d (1984:85).
_
However, a primarily subjective, mentalist sinng 1S far from a uruversally accepted,
or even the most influential, view of ideology in the past few decades. In much re­
ce nt theory, ideology is not necessarily conscious, deliberate, or systema�ically
organized thought, or even thought at all; it is h:ha\f oml, practi�, Er�refl ecuve, or
_
structural. �cation-or, more simply, mearung--rather than 1deaoon m a men­
talist se nse is the core phe nomen on in these contemporary uses. And even the m o st
material asp ects of life are invested with meaning, rife with signification when they
are encompassed within the field of human action.
French structuralists and poststructuralists cast ideo logy not as a matter o f
consciousness or subjective representations but rather of lived relations, to use
Althusser's (1971) formulation. Eaglet on characterizes ideology 10 this �ense as "a
particular organization of signifying practices which goes to constirute human beings
as social subjects, and which produces the liv ed relatio ns by which such subjects
are connected to the dominant relations of pr od uction in s ociety" (1991:18). This
conception disperses ideology thr ougho ut the social o rder (McCarthy 1994:416)
and is recognizably related toBour dieu's use of Mdoxa" (as o pposed to heterod oxy
and orthodoxy), as well as to his noti on of habitus (1977). There arc also similari­
ties, discussed further in this volume by Susan Philips, to Raym o nd Williams's
(1977) interpretation of Grrunsci's idea of hegemony as the "saturation of conc;cious­
ne ss" an d "structures of feeling." The te nsion between different sitings of ideol­
ogy, between subjectively explicit an d co nstructive ly implicit (immane nt) versio ns,
is a recurre nt c once rn of co ntributors to this v olume.
A related dimension of variatio n is the degree to which ide ol ogy is held to be
a c oherent system of signification. In this volume, Hill foll o ws Eagleton (1991) in
expecting an ideology to form a relatively coherent system of meanings.But ide ol­
o� �n �e �cwed as piecemeal and internally contradicto ry. (Whether the contra­
d1ct1on lies m the conceptual model of the world or in the world which is accu­
rately mod eled is another point of some debate.) Volosinov, for example, does not
reserve tbe term Mideology" for organized systems of signification but writes of the
"lowest s�a� of behavioral ideology" as one that lacks logic or unity (1973:92).
�h3:3ctenzmg ideology as a social process, not a possession, Th crbom finds it m ore
like the cacophony of sounds and signs of a big city street than ... the text se­
renely communicating,vith the solitary reader or the teacher ... addressing a quiet,
domesticated audience" (1980:viii).
2. A second, and the most widely agreed -upon, strand is a conceptuali zation
.
of deology as derived from, roo ted in, reflectiv
'. e o f, or responsive to the experience
·
· on, even though .ideology so often (in some
or interests of a particular soCl·a1 positi
•
vie�, always� represents itself as universally true. 3 Unlike the
first view this em-
phasis on soaal and expen· ent.1'al ongms· · . . '.
necessarily denies explanatory indepen-
.
dence to ideology · It casts ·d
1 eo logy_as in some way dependent
. o n the material and
practical aspects of human lifie. The character
. . an d d egree o f that relation of depen-
dency vary across differen t theories, ranging from a view of the material and ideo-
Language Ideology as a Field ofInquiry 7

logical as mutually constituting and dialectical to views of ideology as secondary,

entirely contingent, and/or superfluous.
3. The third major strand of ideology, often seen as following from the sec­
ond, is a direct !Ink to inhabitable positions of power-social, political, economic.
Ideology is seen as ideas, discourse, or signifying practices in the service 0£.die
struggle to acquire or maintain power. For some (e.g., V. I. Lenin), ideology may
be a tool of any protagonist in the contestation of power-that is, it can be "ours"
as well as "theirs," subaltern as well as dominant. But in the most restrictive for­
mulations of this connection, ideology is always the tool, property, or practice !f__
dominant social groups_; the cultural conceptions and practices of sufiordinate groups
are by definition nonideological. For J.B. Thompson, for example, ideology is sig­
nification that is "essentially linked to the process of sustaining asymmetrical rela­
tions of power-to maintaining domination . .. by disguising, legitimating, or
distorting those relations" (1984:4). 4
4. A fourth major strand in the text of the ideology concept, closely related
but not identical to the third, is precisely the last one pointed to by Thompson:
that of distortion, illusion, error, mystification, or rationalization. Such distortion
can derive from the defense of interest and power, but that is not the only source
recognized.When a theorist's emphasis is on the intellectual rather than the social
character of ideology, distortion can be seen to derive from limitations on human
perception and cognition, as well.s
The Marxist tradition has treated distortion as central to the ideology con­
cept, beginning with Marx and Engels's The German Ideology, which took up
Napoleon's pejoration of the term to criticize the "Young Hegelian" philosophers.
Perhaps the best-known phrasing of ideology as illusion is Engels's description of
it as "false consciousness" (see discussion in Eagleton 1991:89). A much quoted
metaphor of ideology as distortion is Marx and Engels's camera obscura, which
produces an upside-down image of the world (Marx and Engels 1989:47.) But it
is not only Marxists who hold to this conceptual focus. The sociologist Talcott
Parsons, for example, asserted that cognitive distortions are always present in ide­
ologies and that deviation from scientific objectivity (itself selective according to
community values), through selectivity and distortion, is an essential criterion of
ideology ([1959] 1970:294-295).
To many observers, this concept of error or illusion implies complementary
forms of knowledge as truth, as well as a privileged position (often reserved for
science) from which such truth is knowable. It is this implication in particular that
has led many social theorists, Foucault most influential among them, to eschew
the notion of ideology in favor of more encompassing concepts such as "discourse"
(1970, 1980).As is well known, Foucault argues that "truth" is constituted only
within discourses that sustain and are sustained by power.That is, all truth is con­
stituted by ideology, if ideology is understood to be power-linked discourse.

THE GREAT DIVIDE in studies of ideology lies between the second and third con­
ceptual foci, between neutral and negative values of the term. Uses that focus on
power and/or distortion share a fundamental critical stance toward ideology; such
 lNTROOUCTION
8

ve, parti ular, pragnu ti , or pcjorn­

uses have been variou l labeled critical, negati
he apply the term r JJ • to
tive.6 More globaliz.i.ng and intellccrualizing appro_ _
and are nonco IDJllltt al on the truth value of 1de 1-
all cul tur. al conceptual schemata .
or 1·a1 tent j_fitc
ogy. These have been di cu scd a neutral, des�ip�e •. n�tional,
cone ptions of ideology.7 Betraying my °'� lingmsttc ideology, I have _come t
mmcnt· fi r I a tl�a t
suspe tthat a phE!!_ological hibbole th can di �o where:, . � _
momentarily positioned on thi fundam ental m elketual d1,•1dc. pr nunc 1 non
. _
of"idca-olog( ([ay]deology) invokes the ideational and rep�· enran n· : , ·hilc �
pronunciation ..id-eology" ([1Jdeology) places power , nd mtcre (the 1<l lurktn
under thin over?) at the core of the phenomenon. But rguabl, • e\'Cn the mo t
doggedly neutral social- cientific uses are tinged with di ·approbation, the trul
neutral tance more often encoded by the choice o thcr label u ha ulturc,
worldvi w, belief, me11talitl, and so on.
Among anthropologist , there have been vo ti advocate fi r b th the nega­
tive and the neucral visions ofideology. Clifford Geem ([19 ] 1 73) argu ·d fore -
fully that social science mu t e chew negative u c and on ern i elf not with rhe
truth value ofldeologies but rather with the way that they mediate meaning fi r
social purposes. The Marxi t anthropologist Maurice Bloch ( 19 5), on the ocher
hand, re eive the term "ideology" for critical u e , indi ing m of r pr en­
tation that mask social processes and legitimate a social order. He adv catcs re­
taining a distinction between id ology and everyday o nition, derived from exp -
ricncc in interaction with a culturally constructed environment. John and Jean
Comaroff (1991) have also proposed a schema that di tingui he culture on rhe
on hand from more power-charged cultural form ofideoJogy and hegemon on
the other, in a wonomy that is perhap most clearly applicable to colonial and other
culture contact ituations. Susan Philips take chi di cu ion further in her om­
mcntary later in this volume.
Karl Mannheim, a founder of the sociology of knowledg , i one theori t who
attempted (albeit ambivalently) to neutrafu..e the negative annotations o the idc­
oJogy concept. In contrast to a "particular" conception of ideology (the negative
one that unma ks or debunks an ideology alwa · seen the intellectual property
of other: ) 1annheim ad\•ocated a �total conception" of ideolo , as tem of
thought that arc socially iruated and col lectively hared ([1936] 19 -). That· all
knowledge, including thar of the ocial analyst, i to c under tood
as ideological.
Mannheimian ideological anal i is fundamentally noncvaluathre,
ys indi. tingui h­
�blc from the sociology of knowledge, as i:t smdie the way knowledge ystcm are
miluen cd by the ocial and historical circum tance
in which th ey are iruatcd.
�- B. Thomp on (1990) riticizes 1annheim' total ideology concept fi r nc­
g lecttng power rela · on , but it might be argued that
the appr ach does not · much
neg le ct powcr ituate it as on aspect(surely an important and nC\
_ _� i rirahle one) of
c sooal poSt�oning o: cultural forms. If by
: idcolo we mean signirying pr -
ce �at n brute o t
. c? . � ubj ects, surely we hould also attend to, for example,
affih tio n, mnmacy, and ldentt_ ty, all
� of which are compJcxly imbci.cated with but
not d rcctl and s ply equatable to
! � � power.9 This larger intersection of ignifying
p ract e with oc1al relation is
'. I think, what both Heath and J rvi.ne are afi er in
defining cultural conceptualization
oflanguage a ideal gical when/be au e they
 Language Ideology as a Field ofInquiry 9

are also politically and morally loaded ideas about social experience, social relation­
ships, and group membership.
To be sure, almost any human act of sig__ni.fication in some r�ect�� .!Q..Q!"­
ganize social relations. 10 But this doe not neccs arily mean that enlarging the focus
of the ideology concept beyond signification in service of power ne essanly enlarges
the term to the paint of usele sne s. Although it doe not distinguish one form of
signification from another (almost all signification having an ideological aspect), the
concept can still very usefully hold onefaut of ignification to the light, what Mann­
heim called the "social and activi t roots� of thinking and ignification (1985:5).

Why Ideologyf

We have, then, no easy con ensu on the meaning and use of the term in question,
"ideology." Given this welter, its invocation in the research program announced in
this volume and essay should not be construed as a fierce commitment to termi­
nology. Indeed, the initial choice was as much circumstantial as considered, and
there may well be better ones.
Again, there is little point to attempting to legislate a ingle interpretation of
ideology from the range ofuscful meanings (Eagleton 1991:1). Particularly in a vol­
rune whose genesis has constituted a first foray into identifying a field of inquiry ,
recognition of multiple existing tradition has been essentia l Although most con­
tributors' formulations are critical, they explore various di men ions ofideology.
A source of particular interest to mo t contributors, and of unresolved tension
for some (see especially the exchange between Kroskrity and Briggs), is the problem
of alternate sitings ofideology. Ideology is variously discovered in linguistic practice
itself; in explicit talk about language, ·that is j metalinguistic or metapragmatic
discourse; and in the regimentation of language use through more implicit meta­
pragmatics. {"Implicit metapragmatics'' means linguistic signaling that is part of the
stream of language use in process and that imultaneously indicates how to interpret
that language-in-use; see Lucy 1993a and Silverstein 1993 for full discus ions, and
see also Gumperz's (1982) notion of"contextualization cues.") Irvine insists and care­
fully demonstrates that linguistic ideology must be treated as distinct from and not
just implicit in and discoverable from linguistic structure and practice. In the com­
parative tudy of honorification, she argues, only by looking at the patterning of
linguistic structure, usage, and ideology can cro s-linguisric similarities and thus
the core of the phenomenon be revealed. Kroskrity argues that a focus on overt ideo­
logical comesration hould not lead us to lose sight of ideology as doxa, naturalized
dominant ideologies that rarely ri c ro dis ursive con ciousness. Blommaert and
Verschueren sirnila.cly argue for a methodology that emphasizes the unsaid, the un­
expressed as umptions th t implicitly frame a text and enable its coherence. Briggs ,
in contrast, suggest that such an emphasis not only privileges the analyst's perspec­
tive but may contribute to the analyst's unintended collusion in reifying the perspec­
tive of only a sector of a community.
Some contributors simultaneously examine ideology on more than one leveJ:
Spituln.ik compares language ideolo ies that are embodied by the practice of
 INTRODUCTION
10
age _ideologies d�scoverable in dis­
radio broadcasting in Zambia with explicit langu
at ideology �s 1� breaks th ro�g�
course about these practices. Mertz, too, looks
anc mdic�t�rs to exp_liot
multiple layers oflanguage use, from implicit metapragm
of e �plic.im:ss of 1de­
metapragmatic regimentation. Philips contrasts the degre�
u c B?�· raises ques­
ology in the different chapters she discusses and, n�t �
d in d1scourse data.
tions about the privilege of reading out ideology_ as 1�plicate _
larly produc­
Silverstein identifies institutionalized and interaction ntuals as parttcu
and thus for the
tive (because privileged and value-setting) sites for the enactment,
researcher's discovery, oflanguage ideology.
In spite of the many possible points of divergence, the essays gathered �ere
share a fundamental emphasis on the social origins of thought and representation,
on their roots in or responsiveness to th e experience ofa particular social position.
Seeking to bring out the social dimensions of cultural conceptualizations of lan­
guage, most contributors go further, insisting on the tie ofcultural conceptions to
social power as the crucial feature ofthe phenomena under study, alt h oug h power
is variously conceived. This recogrution ofrh e social derivation ofrepresentational
practices does not simply debunk them, as long as we recognize that there is no
privileged knowledge, including the scientific, that escapes grounding in social life.
There is a danger that in taking the subject matter of, for example, the et h ­
nography of speaking and reconsidering it as ideology, we will simply repeat w hat
anthropology has "always talked about anyway" as culture (Asad 1979:26). A natu­
ralizing move that drains the conceptual ofits contingent historical content, mak­
ing it seem universal or timeless, is often cited as key to ideological process. Ironi­
cally, ant hropology too often h as participated in a kind of naturalization of th e
cultural, casting culture as a sh ared and timeless prime motivator. The emphasis
ofideological analysis on the social and experiential origins ofsystems ofsignifica­
tion helps counter such naturalization. We intend by th e term a reminder that th e
cultural conceptions we study are partial, interest-laden, contestable, and contested
(Hill and Mannheim 1992:382). Cultural frames have social histories, and th is
demands t hat we ask how seemingly essential and natural meanings of and about
langua�e are socially produced as effective and powerful. As Silverstein and P h ilips
botfi u�derline in th eir essays, th is implies a met hodological stance, a commitment
to consider the relevance ofsocial relations, and particularly of power relations, to
the nature ofcultural forms.
Such a coo:imitment does not entail an outmoded base-superstructure model,
. _
10 wh ch i:natenal �fe and relations are seen as primary
'. and real and the ideological
as denvanve, �red1cta�le or illusory (a theme Philips develops further in her com­
m_entary). S�1al relat1?ns and materiality are not presymbolic but rat her arc con­
stituted, notJ�st sust�med, th roug h symbolic activity (Thompson 1990:58); ide­
ology and soCJal relations are understood in this volume as
mutuall y constiruti,•e.
In t�eir early _ work in The German Ideology, M arx and
Engels attacked as ideologi­
cal (i.e , false in socially significant ways) the notion
: t h at ideas are autonomous and
efficacious. Contributors to this .
' vo1 umc do not cast ideolog
. . y as autonomous, but
they do view 1t as having effiects (though not " .
. efficac1ous� .m th e sense of h aving
desired effects) · Ideology-not as I·deas
so much as construed practice-is conse-
Language Ideology as a Field oj'Jnquiry 11

quential, for both social and linguistic process, although not always consequential
in the way its practitioners might envision.
This understanding ofideology as active and effective distinguishes our enter­
prise from many versions of Mannheim-influenced sociology of knowledge, in
which the problem addressed by ideological analysis is the more unidirectional social
determination of thought. This isn't the sole concern of the research reviewed in
this essay and represented in this volume, which can be as interested in the ideo­
logical determination of (linguistic and social) structures as in the structural deter­
mination of that ideology (see especially the discussion of the work of Michael
Silverstein in the following section). The point is not just to analyze and critique
the social roots of linguistic ideologies but to analyze their efficacy, the way they
transform the material reality they comment on. The emphasis is on what E agleton,
harking back to Austin's speech act theory, calls the performative aspect ofideol­
ogy under its constative guise: ideology creates and acts in a social world while it
masquerades as a description of that world (1991: 19).
In the following sections I turn from the general concept of ideology to the
review of studies oflanguage ideology in particular. The work discussed includes a
full range of scholars' notions of ideology, as well as much research that has gone
on under other rubrics, such as culture, worldview, and metalinguistics. Although
I try to touch on the widest possible literature in this sometimes breathless survey,
I cannot pretend to be exhaustive (only exhausting). My purpose is to contextualize
some of the principal and currently most productive approaches to inquiry and to
situate the work in this volume in relation to those trends.

Approaches to Language Ideology

Ideology at the Intersection ofLanguage Use and Structure

A dominant view in American anthropology and linguistics has long cast ideology
as a somewhat unfortunate, though perhaps socioculturally interesting, distraction
from prima.ry and thus "real" linguistic data. Franz Boas (1911) proposed that lan­
guage is a cultural system whose prim ary structure is little influenced by secondary
rationalizations and so is an exempl ary target ofanalysis. The American structural
linguist Leonard Bloomfield ([1927] 1970, 1944) actually gave considerable atten­
tion to speakers' evaluations of speech forms. Although he characterized these as
part of the linguist's data, his acerbic remarks make it clear that he saw them as a
"detour," oflittle relevance to the explanation ofthe structure of"normal" language
(1933:22). Modern linguistics in the Bloomfieldian tradition has generally assumed
that linguistic ideology and prescriptive norms have little significant-or, paradoxi­
cally, only pernicious-effect on speech forms (although they may be recognized
as having some less negligible effect on writing). 11
In contrast to this received wisdom, Michael Silverstein has argued that a grasp
oflanguage ideology is essential to understanding the evolution oflinguistic struc­
ture: "The total linguistic fact, the datum for a science of language, is irreducibly
 lNTRODUCTIO
12
ual interaction of meanjngful ign orm
dialectic in nature. It is an unstable mut
n u · medi ted b the fa t of cultural
contextualiud to ituations ofintere ted huma
to the cx��t that langu ge u _e i
ideology" (1985:220). According to Silverstein,
concep� langu . as ooally
teleological-that i , to the extent that speakers
ideas about the �eanmg, fun n, and value
purposive action-we must look at their
lly bared
of language in order to understand the degree of
empirically occurring linguisti c form . . . .
rtth Lingw -
For Silver tein ideology can quite actively and ncretel 'd1 t
2 ronoun alter­
tic structure it repre ents.1 In analyse of gender in English, T/V p
nation, andJavan e e peech level, Silverste in (e. g ., 1979, 1985) h hown that
ideology, understoo d as rationaliza tion not onl exp lain, but all · affec , lin­
"rationalizi n " it, often b y makin g it more r · g ul a r. ]deolo gy hu
guistic strucrure, g
constitutes an essential moment of the p henomenon of anal gical tin i ti chan ge.
Ideological tenets are derived from ome as p e t of cx p crien e and hen g eneralized
beyond that core and econdarily imp ed on a br der · tegory phenomena;
thi broader cate gory then undergoe restru twin . tru ture ondition ideology,
which then reinforces and expand the original tru tur , di tonin g langua ge in
the name of making it more like itsclf(cf. Bourdieu 1991). In, rno c th r joins the
conceptual to the active side of ideology (or the con tative to the perfurmati e, to
echo Eagleton's invocation of Au tin), this appro:1 h h , that r .. under rand"
one's own linguistic usage is potentially to change ic ( ilverstein 1 7 : 233). Rumsey
has nicely restated this view:

Language structure and linguistic ideology are n t entirely independent o each other,
nor i either determined entirely by the other. Instead rhe structure provide formal
categories ofa kind that are panicufarl conducive to "mi recognition. And partly
a result ofthar mi recognition, might not the linguistic rem gradual!· change so
to approximate that for which it was misrecognized? (1990:35n

Important linguistic changes can be �t off b uch ideological interpretation

ofl�nguage structure-in-u e. But becau e th derive only from larger ocial dia­
ey
leca� su<:11 changes are likely to take an unintended direction, as exem lified in
p
the histoncal case of second-person pronoun alternation in En lish that ilve .rstein
g
(1 985) analyzes and Kulick reviews in his cha ter in thl volume (see Ehrlich aod
_ p
King 1994). In the seventeenth century, Qyakers insi ted on u c of"thou" form
for all seco nd-person singular addres , rationalizin th.i usage orcilng to the
. _ g
emergmg lingui tic ideology of the time as more truthful
be au e faithful to nu­
merical realities of the objective world. This racti e
p on ec ndarily ideologizcd
by the larger so iety as an index of ti matized
g Qyakcr identity, ct off a back.la h
m�vem;nt
�ay from any productive u e of "thou" by that larger community. A
shiftto yo u was ompletedb 1700(
y ilverstein1985:246).lntbj v lume,Kulick
p re sents �e lo s of the local langua e of Ga un in
g p Papua ew Guin ea as a parallel
case of uruntended hi.ft • tn·
ggerc d by a complex of indexical as ociations o the
vern u�ar and the langua e of wide
; g r ommunication.
' rnngton ( � 988) ob erves thar althou h
g it is tandard in ociolingui ti analysi
to I ook f,or relations between s
truc turaJ ch ange and communicati ve function, .1t .1
Language Ideology as a Field ofInquiry 13

more controversial to invoke a notion of native speaker awareness as an explanatory

link. As Itvine (1989) has pointed out, Labovian correlational sociolinguistics sug­
gests a direct relation between lingiu stic variation and social differentiation.This
correlation is better viewed as mediated by an ideological interpretation of the meaning
of language use, as Irvine demonstrates in her contribution to this volume.
Labov himself, construing ideology as overt political discourse, has explicitly
discounted the power of ideology to affect speech forms (1979:329). However,
Labovian research attends to measures of "subjective reactions" and of"linguistic
insecurity"; in rus commentary in this volume, Silverstein characterizes the latter
as ideological allegiance to the standard register. Labov differentiates mechanisms
of change from below and above the level of speakers' awareness and argu es that
only changes from below are extensive and systematic, while conscious self­
correction-wruch he labels ideology-leads to only sporadic and haphazard ef­
fects on linguistic forms (1979:329).
Errington (1988) argues that Labov's generalization is most applicable to pho­
nological variation, wruch may be mediated less by speakers' understandings of their
conscious communicative projects.In his work on Indonesian speech levels, Errington
has developed the notion of pragmatic salieoce-"native speakers' awareness of the
social significance of different leveled linguistic altemants" (1985:294-95).More
"pragmatically salient" classes of variables are more susccptibl: to rationalization and
strategic use, being (mis)recognized by speakers as more crucial linguistic mediators
of social relations (sec also Philips 1991 and discussion in Agha 1994). Because such
awareness and use drive linguistic change, says Errington, these variables require a
fundamentally different, participant-oriented analysis (see also Han.ks 1993, Meeuwis
and Brisard 1993, Mertz 1993). lrvine takes up this theme of pragmatically salient
phenomena in her essay, examining the relation between ideologies and patterns of
honorification not only in Javanese but in several African languages, and briefly con­
sidering possible consequences for language change.
Examples from European language communities, and especially American
English, reveal a tendency to see reference or propositionality as the essence of lan­
guage, to confuse or at least to merge the indexical functions of language with the
referential function, and to assume that the divisions and structures of langu age
should-and in the best circumstances do-transparently fit the structures of the
"real world" (see especially Silverstein 1979, 1981, 1985, 1987). Such views of
language are held by expert as well as lay observers, as Irvine comments: "many
writers ... in linguistics and the social sciences ... have assumed that referential
communication is the only function of language" (1989:250; see also Briggs 1986,
Reddy 1979).
Silverstein has been understood by some to suggest that this Western objectify­
ing drive for reference and a focus on the surface-segmentable aspects ofreferentially
evaluated language is a nearly uni,,crsal phenomenon. More accurately, Silverstein's
claim is that the referential structure is universally a structuring condition of the con­
sciousness of pragmatic functions. The rationalizing accounts people devise to ex­
plain language beyond that vary widely, from reference in our own tradition to fully
pragmatic theories for Javanese, which derive the power of langage u from theories of
interaction.ll
 I TROOUCTION
14

e- e�cnrable � . and consc�

Although not universal, the focus on surfac
oUe non ofwo W1 cl · attested
uent conception oflanguage as a grammarless
1991) . In hi haptc r thi volu.m
Blommaert 1994a, Glinert
&.g. .
!n � .
. · ly qu�non
a tngger to rcfl
siders such a view among the Tolowa, USU1g 1t as . �
tru � . (Irvine similar�y m� a
professional linguistics' privileging oflanguage .
comparison ofWolof andJavanc c ideolo
g ies ofhon on 1c::at:10n to rccon 1� ra��n
on uucr f lin tic
of professional linguists' privileging of refere nee th rough the
(1990) h argued that the ocus on
altemants as referentially "the rune.") Rum ey
not dichoto-
lexicon is not characteristic of Au tra1.i n aborigjnal cultures, whl do
(1 2) has lhat Ilo
mize talk and action or word and things. R sa do l

th.inkoflanguage in terms of:iction rather than referen e. Amon akers,

the language's disappearance is ideologized differently by� o encrati ns, sh wing
how cultural conception flanguage structure are ind <l root d in ocial po ition.
Younger speakers see Language as a collection of wor , ommodifi d obj to be
brought out and displayed, while for older speakers ch language i · matter o myths,
not words (Moore 1988, 1993). HiU takes up the qu ti n of re ercntial bi in her
chapter, identifying an "en_a 've" Mexicano langu ge i J · that empb izcs n r
the relation hip between reference and reality but rath r Lhe proper ac -ompli hmcnt
of human relationships through dialogue.

Ethnography ofSpeaking
The ethnography of speaking was chartered to tudy "ways of speaking" from the
point of view of events, ac , and tyle . Hym es (1974: 1) insi t d early on that a
community' own theory of peech mu t be considered part of any serious eth­
nography, and from its inception the ethnography of peaking h gi en rem­
ati attention to ideologies oflanguage, primarily in the neutral ense o cultural
conceptions, particularly those embodied in explicit metalinguistics (e.g., Bauman
and herzer 1974, Gumperz and H es 1972). Ethnographers of speaking have
also pursued the grounding oflanguage beliefi in other cultural and ocial pro e se
ym

�e.g:, Feld �d chie.ffelin 1981, Katricl 1986, Rodman 1991). Language ocial-
1zanon tudies, o
� : �pie, h�vc demonstrated connection among folk theoric
oflangu:ige acqumnon, lingu.isb .
practices, and key cul rural id about pet50nhood
(Och and chie.ffelin 1984).14
Over ti.me, the field has moved toward more concerted attentio ro the rela­
. n
tion �tween uch local lingui ti theories and practice
� . Langu:ige ideology h been
made incre 1ngly explicit as a fore that shapes verbal practices and genre
� from
o atory to disputing. Genres them elves have of
� come to be recognized not
di ou.rs.c features but rather as gorienti.ng framcw
or: ' interpretive pro cdure and
sets of expectations" (Hanks 1987:670). IS

(l 96�pe
� �
t act theory, a developed in the work of
an)e John S arlc (1969), was initially
�
o ra p hy of s e aki.n . Later
he philosophers
welcomed as compatible with chc eth­
however it stimulated critical reflections on lingui -
J. .ustin

� �
tic ideology. Silverstein (1979:210) :inn,
..d that A us ti' n ,s 1.deas about 1anguagc .. acts"
an d "fior��-s » were projections of cove
-c,-
rt categories typical in the metapragm ti di -
course O anguage s such as English. On the basis
of her fieldwork with the llongor
Language Ideology as a Field ofInquiry 15

of the Philippines, Rosaldo (1982) concurred that speech act theory was based in
a specifically Western linguistic ideology, what Verschueren (1985:22) character­
ized as a privatized view oflanguage that emphasizes the psychological state ofthe
speaker while downplaying the social consequences ofspeech (see also Pratt 1981).
Ethnographers, particularly ofPacific societies, have since argued that the central­
ity of intentionality within speech act theory is rooted in Western conceptions of
the selfand that it is inappropriate to other societies, where it obscures local methods
of producing meaning.16
As is true ofcultural anthropologists generally, ethnographers ofspeaking have
increasingly incorporated considerations of power in their analyses, again leading
to a more explicit focus on linguistic ideology. Bauman's (1983) historical ethnog­
raphy oflanguage and silence in Qyaker ideology was an important development,
since it addressed not a neutral variety of ideology but a more formal, conscious,
and politically strategic form. Research on language and gender that has responded
critically to essentialist readings of gendered behavior and values has helped iden­
tify the mediating role oflanguage ideology in the organization ofpower. 17 Point­
ing to the "paradoxical power of silence," Gal (1991) in particular reminds us that
the social meaning ofcommunicative forms can never be taken as natural and trans­
parent but must always be examined as cultural construction. In this volume, Kulick
analyzes the indirect gendering and consequent loss of a bilingual repertoire and
the ties of these processes to new forms ofsocial power in Gapun (see also Kulick
1992, Woolard 1995).
The ethnography oflanguage and schooling and oflanguage and the law simi­
larly made early moves to incorporate dimensl9ns of _power and ideology into the
analysis ofcommunicative practices. Numerous studies in both areas examined how
these institutions arrogate truth and value to some linguistic strategies and forms while
ruling others out ofbounds. Mertz's chapter in this volume brings together work on
both institutions, examining the regimentation of linguistic practice in American law
schools as a key constituent of the epistemology of the legal profession.18
Finally, in considering ideological dimensions ofcommunication, the ethnog­
raphy of speaking has shifted toward the recognition of variability and contradic­
tions. As Briggs discusses in this volume, early ethnographic critiques of univer­
salist speech act theory, such as Rosaldo's, rested on their own assumptions of
cultural uniformity. New directions in the ethnography ofspeaking move away from
the imposition of such homogeneous cultural templates. Claims about "the lan­
guage ideology ofthex" are increasingly viewed as problematic. Verschueren (1985)
has noted, for example, that English speakers and other Westerners can be seen to
hold ideologies rather similar to those ofthe Ilongots that Rosaldo discusses, de­
pending on the kind of dnta we look at (see also Rumsey 1990).
Current research recognizes struggles among multiple conceptualizations of
talk within a community and even contradictions within individuals {e.g., Briggs
1996a; Gal 1993; Urciuoli 1991, 1996). ln her contribution here, Hill locates a
Mexicano discourse ofnostalgia and respect that is common only in the sense it is
known to, but not produced by, all. Rather, women and men possessed of little
local capital participate in a counterdiscourse about language. Hill argues, how­
ever, that both discourses build on a more fundamentally shared Mexicano linguistic
 1 TROOUCTIO
16

Language Contact and Conflict

In multilingual communirie where ther r:rugglc o,· r
an-
language, re earchers hav long trea ed langua e ideal · (in ne gui e or
other) as socially, politically, and even lingui tically. ignift JC. Th dirion al topi
,- of sociolinguisti inqui ry in these settings have be •n language main enance and
shift, contact-indu ed ungui tic change, the linkage oflangu gt' t ethni iry and
nationalism language attitude , and langua e planni.n and d vdopmc:nr. All of
these draw u into Ifarl Mannheim' question of rhc " o iaJ nd ,1 ti i t roor "of
conceptions oflnnguage(s). Dimen ion oflangua e ideology that ha · been treated
in such work include: ideas of what ounts a langu g · and, under! ing the e, the
very notion that there arr di tin tly identifiable language , bj ct rha n be
"had"-isolated, named, counted, and fetishized; value o c
iat d wirh particular
languagevarictie ycommunity members; as umpti n that identi and allegiance
are indexed by language use.
The extensive body of research on langua e attitude in multilingu al commu­
nities grew up in a ociaJ psychological framework ( cc Baker 1992, Giles ct al.
1987). However we can recast the intrapcrsonal attitude a ociall , d rMd, in­
tellcctualir.ed or behavioral ideology akin to Bourdieu' .. habiru " ( tti.na i 19 3;
Bourdicu 1991; Woolard 1985, 1989a)_l9 On the other hand, rudie oflaEK.Uagc
maintenance and shift initially implicated macrosocial event dire t � . tcr
re earcb ha insi tcd that it i only through the interpreti ve fi1 er f belie about
language, cognition, and ocial relations that political and c nomi events have
an e�ecton l guagc mainrenance orshift(Menz 19 9:109). In · v lumc, Kuli ·
�
examine the 1de I gical dyn amic in one such hift.20
The identification of a language with a people and 11 con qucn diagno ·i of
pcople hood by the criterion oflanguagc. have been the fundamen :tl tene oflan­
gu agc 1_ d °logy t which this tradition of research has attended (
. 7 e I lyme · 1984).
It 1� a 1 � that the equation oflanguage and nation i not a
� natural fact but rather
a.luston al tdcological construct. This construction is convent
i nally dated to latt.�
� ig� tecn th-century German Romanticism and Johann
Her er' amous character·
1zat1on oflanguage ' th geruu
. · o f a peop1· e, an d thu ·1t ·1s often re erred to as th e
Romantic or Herderian concept oflanguage (see K ke 1990; see also Humbol t
oep
Language Ideology as a Field ofInquiry 17

1988). But in fact Herder's formulation can be traced to the French Enlighten­
ment and the French philosopher Condillac (Aarsleff 1982, Olender 1992).
Exported through colonialism, this Herderian or nationalist ideology oflan­
gu age is globally hegemonic today. In this volume, Blommaert and Verschueren
trace these assumptions in Western European news reporting, and Spitulnik finds
them running throughout Zambian radio decisions. Modern linguistic theory itself
has been seen as framed and constrained by the one language / one people assump­
tion (Le Page 1988, Romaine 1989). State policies as well as challenges to the state
around the world are structured by this nationalist ideology oflanguage and iden­
tity.21 As Blommaert and Verschueren show in this volume, it underpins ethnic
struggles to such an extent that lack of a distinct language can cast doubt on the
legitimacy of a group's claim to nationhood.
The belief that distinctly identifiable languages can and should be isolated,
named, and counted enters nor only into minority and majority nationalisms but
into various strategics ofsocial domination. For example, ideas about what is or is
not a "real" language have contributed to profound decisions about the civility and
even the humanity of others, particularly subjects of colonial domination in the
Americas and elsewhere.22 As Spitulnik's discussion of Zambian mass media,
Errington's oflndonesian language development, and Collins's ofAmerican In­
dian language education programs demonstrate, rankings oflanguages continue to
be invoked to regulate the access ofspeech varieties to prestigious institutional uses
and of their speakers to domains ofpower and privilege.23
Written form, lexical elaboration, rules for word formation, and historical
derivation all may be seized on in diagnosing "real language" and ranking the can­
didates (see, e.g., Ferguson and Gumperz 1960, Haugen 1972, Olender 1992).
Evaluations of oral language are often implicitly based in literate standards, as
Bloomfield noted long ago, although speakers ofsome minoritized languages hold
them in high esteem precisely because they "cannot be written" (King 1994, Tay­
lor 1989). The question of whether a variety has a grammar plays an important
part in such debates and diagnostics (Eckert 1983). Academic linguists' extension
ofthe concept ofgrammar from the explicitly artifacrual product ofliterate schol­
arly inteivcntion to an underlying natural system only exacerbates the polemics (see
controversies reviewed in Morgan 1994).
An equation of change not just with grammarlessness but with decay also
pervades judgments about the status oflanguages. Language mixing, codeswitching,
and creolization thus make speech varieties particularly vulnerable to folk and
prescriptivist evaluation as grammarless and/or decadent and therefore as less than
fully formed (Jourdan 1991, Ludwig 1989, Romaine 1994).
Movement:; to save minority languages ironically are often structured, willy­
nilly, around the same received notions oflanguage that have led to their oppres­
sion and/or suppression. Although in some minority language movements the stan­
dard terms of evaluation are subverted (Posner 1993, Thiers 1993, Urla 1995),
minority language activists often find themsdves imposing standards, elevating \
literate forms and uses, and negatively sanctioning variability in order to demon- }
Strate the reality, validity, and integrity of their languages. 24 Or again, culturally
cohesive indigenous groups that enter into struggles for state recognition in the
 INTRODUCTION
18

th ir intcm.iJ lingu' ti di ercnccs

nationalist ideological climate may recon true
.
as defming cthni di tinction Oackson 1995) . .
e uation of one lan g u a g e/one people ha came m • tcm:e
Along with the q " .
tonue . the one lust and
g
on the authenticity and morn.I igni.ficance of"mother _
elf (I laugen l 91,
therefore real langu ge of a speaker cransparenr to he true
kutnabb-Kan and Phillip on 1989). Anothe r tenet o en lu temi, 1th the
Herderian ideology in both folk and cientific view dem· d
� . lin i ti puri m
.
essential to the urvival of minority lan a
gu g es, a kind o po hc1 of th bound.an
that have been drawn to create di tinct language onns. Puri an and anal t have
criticized all of the e umptions as inappropriate in crtin �·here rnul ilingual­
ism is more typical and where the lingu istic re p rtoirc · Outd or mpl . �ut
whether waged in Corsi a, India, or the America n ourhw t, the ru � ,le gain t
the nationalist ideological complex has been diffi ult t win ( nz; ld ·. 1987, Jaffe
1993, Khubchanda.ni 1983, Pattanayak 1988).
Although the validity of the nationali t ideal o langu· h o en been
debated or debunked, le anention has traditionally been ivcn to unde tanding
how the view oflanguagc not only as discrete, di tinctive cntiti but embJcm­
aticof elf and community comes ro take hold in so man •d' er nt ttin (Fi. hman
1989). ln his critical es ay on so ial cientific notion o idcolo cncrall , ecm.
([1964] 197 ) long ago called for systematic attention to the iaJ and, ·h cl w uld
prefer to call emi tic proces c through which ideol ogi · come to ignify. The amc
must be id about ideological con eptions oflangua ·
In analyzin politicized contest over the "true" national b.nguagc and tan­
dard fonn , we might k whi h lingui tic features are ci1..ed on, and throu h what
semiotic proc e are they interpret d as representing the ollc tiV1ty? Errington'
idea of pragmatic alicnce, di cussed earlier, points out one dire tion in whi h the
analy is might proceed ( ee al o Thiers 1993). Although lingui ti V'lriation may
a�pe ar to community member and in correla.tional s iolingui tac imply a
diag ra m of ocial differentiation, anal ts have begun to examin · rhe ideological
ys
p r�uction and signifying tructure of that diagr-am in both folk and expert model
(Irvine 19�9:253 ). All of the contributors to this volume reco
_ gniu: chat impl u. ing
langu age tn pa.mcular wa i not what form ocial
gr p idcntitie, or relation
ou ,
�nor d�s th� group rel�tion automatically give rise co ling i ti di tin ion ; rather
u
1deologtcal interpretation of uch use oflan
gu a e aJ } mcdi te the c ctfe
The onccpru al hcma of the semiotician C.
. Peirce (1 74) ha be nu d
to �al;� t�c �race of emiotic mediation b · which chunk o lingui ic m -
�cnal gam 1gn 1 ficance a representations of particular
pulation , a Gill pur it
in herco mrncn�.25Wor�g th ePeirician n tion
. � � o of ind · lity, tlversrein
(1996a, 1n p
�cul3:i sec his d1 cuss1on in thi volume) h d eveloped a ncrJ.I
theory of the 1 dc:x1cal order� that i
� particul arly productive on thi qu · tion. In
the rransformati n of first-order indcxi
cality into second-order indcxi licy, in-
tan e of pcech that are tatisticall
. y assoc iable with an aggregate of individual
are typified �y community members
_ or ex perts as particular wa of P :iking that
are c.hc at1 d a c tego nc lly
� � � � � ass �ciate d with types of people.26
Whc� � lm m t c form-in-use 1
. � ! thu ideologized a di tinctiv . n I a impli­
cating a di nnctive kind of people,
it is often further mi recognized, in Bourdieu'
Language Ideology as a Field oJI11quily 19

term, or revalorized, as transparently emblematic of social, political, intellectual,

or moral character (see Silverstein's discussion in this volume; see also Bourdieu
1991).27 So, for example, the speaker of the British prestige speech form known as
Received Pronunciation is heard not just as a member of a socially privileged sec­
tor of English society bur also as a person ofgreater intellectual and personal worth.
G. B. Shaw's Pygmalion is a poignant if comedic exposition of such reinterpreta­
tion. In this volume, Hill, Kulick, Spitulnik, Errington, and Collins all explore the
complexes of values taken on by Language varieties in contact situations and their
interplay with social phenomena such as class, gender and affect.28
lconicity, the interpretation oflinguistic form not just as a dependable index of
a social group bur as a transparent depiction ofthe distinctive qualities ofthe group,
is one ofthree semiotic processes that Irvine and Gal (1995) have suggested are wide­
spread in linguistic ideologizations. For example, in a particular community ideol­
ogy, the 'simple folk' might be characterized iconically by 'plain speech,' in contrast
to the 'ornate' speech of another social stratum. Such iconic readings give rise to the
other processes Irvine and Gal have identified: recursivity, in which an opposition
salient at one level is projected onto other levels ofa linguistic and social relation­
ship, and erasure, the rendering invisible of some sociolinguistic activities
or actors in a �y that bolsters the iconic reading oflanguage differences. So, as'sim­
plicity' proliferates in the speech of the 'simple folk,' that which might be character­
ized as complex or ornate is ignored, compartmentalized, or even sometimes Stripped
from their repertoire through the process oferasure.
The attribution of social, moral, and political meanings to specific language
varieties and the erasure ofcontradictions and variation affect patterns of language
acquisition, style-switching, and shift, as Hill, Kulick, and Schieffelin and Doucet
show, as well as institutional use, as Spitulnik and Errington discuss in this volume.29
Moreover, in liberal democratic societies, the misrecognition or revalorization ofthe
indexical character of language may make discrimination on linguistic grounds
publicly acceptable where the corresponding ethnic or racial discrimination is not
(Gumperz 1992, Lopez 1991, Milroy and Milroy 1985, Sonntag and Pool 1987).
For example, although penalizing a student for being African American may be ille­
gal, penalizing a student for speaking African American Veracular n English is not.
Where discrimination against Asian Americans in job promotion is illegal, passing
over or dismissing an Asian American because of an "accent" that others claim is
difficult to comprehend is not (Lippi-Green 1994, Matsuda 1991).
A crude version ofWhorfian thinking that treats English as a sine qua non of
democratic thought runs through the tradition of American language policy and
at times has also enabled attacks on the rights ofminority-langu age citizens (Baron
1990, Leibowitz 1976, Mertz 1982). The identification ofdiscriminatory dynam­
ics in the recent English-only backlash against bilingual policies in the United States
has led a number oflinguists and social scientists to engage in ideological critique. 30
However, simply asserting that struggles over language are "really about" racism
does not constitute analysis. Tearing aside the curtain of mystification in a �iz­
ard of Oz theory of ideology," as Asad (1979:622) has phrased it, begs the ques­
tions ofhow and why language comes not only to stand for social groups but to do
so in a manner that is both dependably readable and nonetheless socially accept-
 INTRODUCTION
20
able . This de mands attention to both the se miotic and the social process of lan­
guage ide ology, along the lines that Silverste in and Irvine and Gal have dcv�loJ><: d.
Communities not only evaluate but may appropriate some part ofthe linguis­
tic resources of groups with whom they are in contact and in tension, refiguring
and incorporating linguistic structures in ways that reveal linguistic and social ide­
ologies (Hill 1985; see also Gumpcrz. and Wilson 1971).31 Lingui�tic borrowi�g
might appear to indicate speakers' high regard for the donor langua�e. But �ill
(1993a, 1995) argues that socially grounded linguistic analysis ofAnglo-American
borrowings and of humorous misrcnderings of Spanish reveals them as racist dis­
tancing strategies that reduce complex Latino experience to a ubordinatcd, com­
modity identity. In a similar examp le, the appropriation of creole speech, music,
and dress by white adolescents in South London, who sec only ·sty le" (again,
commodified), is in tension with black adolescent view� of these codes as part of
the ir distinctive identity (Hewitt 1986). A classic example ofthe negative loading
of langu age appropriation is Basso's (1979) dcscnpuon of a Western Apache
metalinguistic joking genre that uses English to parody "\Vhitcman" conver.;ational
pragmatics, in a scathing representation of and comment on cthnolinguisric dif­
ferences and ine quality. Rampton (1995), however, argues that "language cross­
ing," a limite d form of borrowing and mixing used by Asian, Anglo, and Carib­
be an youths in England, can be a means ofironizing and transgre�sing recognized
ethnic boundaries and ofsituationally forging an alte rnative shared identity.
Ifthe linguistic ideology under lying borrowing or mime sis demands scrutiny,
so does the ideology underlying differentiation. Irvine (1989) points out that the
formal characteristics of the counte rcultural argots that the linguist M. A. K.
H�d�y dub�d an�language s are not arbitrary and that the y again suggest the
me�1atton ofideological conce ptualizations oflinguistic structures. Linguistic in­
versio�s su� as metathesis, so often characteristic of antilanguages, are an image
o� the inversion ofcultural values. Similar ly, subordinate languages in contact situ­
.
att�ns can acqUlfe not only the functional but also some of the fo rm al properties of
an�angu�e \ Speakers ofmoribund varieties ofXinca in Guate mala, for example,
go hog-wild (Campbe ll and Muntzel 1989:189) with glottalizcd consonants,
.
which :ue exo�. c fro the int of view of the dominant Spanish language. This
� �
res�lts 1� the kind ofdmortion oflinguistic structures that Silverstein has described,
� dJSto1:1on
.
that �akes a language more like itse lf. In this case, importantly, it is a
lingwsnc selfthat 15 most disnnct1ve
· · from 1ts · sooally
· dominant counterpart, Sparush
.
(see Th.iers 1993).