DOCUMENT RESUME

ED 198 745 FL 012 221

AUTHOR Byrnes, Dell
TITLE Language in Education: Ethnolinguistic Essays.
Language and Ethnography Series.
INSTITUTION Center fc': Applied Linguistics, Washington, D.C.
FEPORT NO ISBN-0-87281-134-4
PUB DATE Dec 80
NOTE 175p.
AVAILABLE FROM Center for Applied Linguistics, ,3520 Prospect Street,
N.W., Washington, DC 20007 ($10.50)
EDFS PRICE MF01/PC07 Plus Postage.
DESCRIPTORS *Anthrcpological Linguistics: Applied Linguistics:
*Epistemology: *Ethnography: *Ethnology: Narration:
Research Methodology: *Sociolinguistics: Story
Telling

ABSTRACT
Eight essays in ethnolinguistics were compiled for
this monograph. "Functions of Speech: An Evolutionary Approach"
represents an introduction to the application of linguistic knowledge
to the historical and sociological study of peoples. "Speech and
Language: On the Origins and Foundaticns of Inequality among
Speakers" expands on the theme of diversity, inequality, and
evolution, with discussions of writing and of the views of Bernstein
and Jurgen'Habermas. "Qualitative/Quantitative Research Methodologies
in Education: A Linguistic Perspective" addresses the development of
,The three middle chapters, "What Is Ethnography?"
"Ethnographic. Monitoring," and "Educational Ethnology," are concerned
in ccmplementary ways with what counts as legitimate knowledge and
who is counted as entitled to know. The status of narrative as a form
of knowledge is addressed in "Narrative Thinking and Story-Telling
Rights: A Fclkloristos Clue to a Critique of Education" (with
Courtney Cazden). The final chapter, "Language in Education: Forward
to Fundamentals" weaves together many of the themes of the book,
expressing a concern that an ethnographic or ethnolinguistic
Perspective nct be trivialized and vulgarized. (a)

***********************************************************************
Reproductions supplied by EDRS are the best that can be made
from the original document.
***********************************************************************
 1111111

ARV

Language in Education:
Ethnolinguistic Essays
by Dell Hymes
"PERMISSION TO REPRODUCE THIS U.S. DEPARTMENT OF HEALTH.
MATERIAL HAS BEEN GRANTED BY EDUCATION & WELFARE
NATIONAL INSTITUTE Of
Ce.vtbe r 4-t EDUCATION
THIS DOCUMENT HAS BEEN REPRO-
DUCED EXACTLY AS RECEIVED FROM
THE PERSON OR ORGANIZATION ORIGIN-
ATING IT POINTS OF VIEW OR OPINIONS
STATED DO NOT NECESSARILY REPRE-
SENT OFFICIAL NATIONAL INSTITUTE OF
TO THE EDUCATIONAL RESOURCES EDUCATION POSITION OR POLICY
INFORMATION CENTER (ERIC)."

Center for. Applied Linguistics

L
2
Library of Congress Cataloging in Publication Data
Hymes, Dell H.
Language in education.
(Language and ethnography series; 1)
Includes bibliographical references.
1. Anthropological linguistics Addresses, essays,
lectures. 2. Language and education--Addresses, essays,
lectures. I. Title. II. Series.
P35.H88 401'.9 80-27439
ISBN 0-87281-134-4
All essays in this volume are reprinted by permission. Sources
are as follows:
1. Anthropology and Education, edited by Frederick C. Gruber.
(The Martin G. Brumbaugh Lectures, Fifth Series.) Phila-
delphia: University of Pennsylvania Press, 1961, pp. 55-83.
(Also: Bobbs-Merrill Reprint Series in the Social sciences,
A-124 (1962).)
2. Daedalus (Summer, 1973), pp. 59-86. (Proceedings of the
American Academy of Arts and Sciences 102 (3).) (Issue:
Language as a Human Problem, organized by Einar Haugen
and Morton Bloomfield. Also published in book form.)
3. Anthropology and Education Quarterly 8:165-176, 1977.
4. Working Papers in Sociolinguistics 45. Austin: Southwest
Educational Development Laboratory, 1978.
5. Language Development in a Bilingual Setting, edited by
Eugene J. Briere, for the National Multilingual Multicultural
Materials Development Center, California State Polytechnic
University, Pomona. Los Angeles: National Dissemination
and Assessment Center, 1979, pp. 73-88.
6. Anthropology and Education Quarterly 11:3-8, 1980.
7. Keystone Folklore 22:21-36, 1978.
8. In Keynote Addresses from the Horace Mann Lecture Series
and the Paul Masoner International Lecture Series 1972-1978.
Pittsburgh: University of Pittsburgh, School of Education.
Also: Olga Garnio.a and Martha King, eds., Language, Chil-
dren and Socie. The Effect of Social Factors on Children
Learning to Communicate. (International Series in Psycho-
biology and Learning.) Oxford, New York: Pergamon Press,
1979, pp. 1-19.,

December 1980
Copyright © 1980
by the Center for Applied Linguistics
3520 Prospect Street, N.W.
Washington, D.C. 20007
Printed in the U.S.A.
 contents

introduction
functions of speech: an evolutionary approach 1

speech and language: on the origins and foundations

of inequality among speakers 19

qualitative/quantitative research methodologies in

education: a linguistic perspective 62

what is ethnography? 88

ethnographic monitoring 104

educational ethnology 119

with Courtney cazden: narrative thinking and story-
telling rights: a folklorist's due to a critique
of education 126

language in education: forward to fundamentals 139

iil
 introduction

These essays were written for different occasions but do have a

perspective in common. I shall try to point out some of the com-
mon features and implications, but first want to thank the Center
for Applied Linguistics for the opportunity of inaugurating this
series. It is particularly a pleasure because of a long personal
association with the Center. Let me thank Dora Johnson and
Marcia Taylor for bringing the book to print so rapidly and
effectively and Courtney Cazden for suggesting the collection to
them.
The term "ethnolinguistic" in the subtitle may be a bit sur-
prising, yet it expresses the common perspective of the essays
better than any other term known to me. One alternative would
be "sociolinguistic" and, indeed, I have written about this per-
spective under that heading (Gumperz and Hymes 1972, Hymes
1974, 1979). The history of such terms, indeed, would suggest
that "sociolinguistics" should be the general term for all involve-
ments of language in social life. The success and importance of
certain lines of work and certain issues, however, have led to a
widespread connotation for "sociolinguistics" as having specifically
to do with variation and chalige, studied along the lines developed
so inventively and successfully by William Labov. And part of
the domain of this book, especially the first two chapters, would
more easily fit what Joshua Fishman prefers to call the "sociology
of language." It seems best to admit that the inspiration for the
perspective of these essays is mainly anthropological tradition,
and to use a term, "ethnolinguistic," that openly suggests anthro-
pological concern for human culture throughout the world, for
open-ended inquiry of the sort best termed "ethnography," and
for cumulative comparison of cases of a sort best suggested by
"ethnology."
These essays can be considered a contribution to the nascent
fields of "educational linguistics" (cf. Spolsky 1978), but in that
context their perspective seems distinctive enough to make a
qualifying label useful. The focus is only partially on what
happens in schools and what is taught there. To a considerable
extent, the focus is on the societal context that shapes what can
happen in schools. The concern is with some of the ways in
which we think about language and knowledge in our disciplines
and in society at large.
The first essay, "Functions of Speech," is the first essay in
which the perspective known as "the ethnography of speaking"
was publicly advanced and was also the occasion of my first
vi Introduction
formal visit to the university of whose Graduate School of Edu-
cation I am now a member. In retrospect that visit, then an
isolated occasion, seems to have been premonitory. I hope that
readers will excuse the roughness, even crudeness, of the chap-
ter. The fundamental point is sound and essential. We need to
be able to think of language situations and educational situations
as part of the evolution of human societies. We need to be able
to think of languages and personal competencies as specific sets
of communicative means, shaped by particular histories and adap-
tive niches. We need to transcend the liberal assumption built
into so much of linguistic thought, that all sets of communicative
means are equal in the eyes of linguistic theory. They are not
equal in the eyes of history.
We inhabit a world whose languages are stratified in terms of
roles and lexical scope, broadly speaking, as world, national,
and local. (For example, English, Danish, Frisian.) Of course
a particular speaker of English may be linguistically inferior to a
particular speaker of Danish or Frisian, in command of a primary
language and in scope of repertoire. It remains that English, as
the name for a set of lexical and discourse resources, and a body
of materials employing them, has a different scale. It may be
that a person whose command of English was limited to just those
resources of technical terminology and routine in which English
was distinct would be a communicative monster, were that com-
mand complemented by no other. And the fact that a rapid-
loading gun can defeat a bow and arrow confers no normal
superiority on those with the gun (although they may think it
does), and it may be important to point that out. As analysts
of armament, nonetheless, we would be remiss if we ignored the
technical difference.
With regard to persons, what reason can there be to alter
circumstances if the circumstances have no substantive effect?
How can inequality be injustice if it does no injury? We must be
vigilant against false stereotypes and ascription of deficit, but
vigilance itself is insufficient. It postulates tacitly a natural
equivalence, as if social life could only interfere. The truth is
otherwise. Social life shapes communicative competence and does
so from infancy onward. Depending on gender, family, com-
munity, and religion, children are raised in terms of one con-
figuration of the use and meaning of language rather than an-
other. The particular configuration will affect the opportunities
and access they have for other uses and meanings of language.
Depending on social, economic, political factors, they will come
to be able to use and experience language in some ways and not
in others. Often enough, the result will be less than justice or
vision would require. It is my considered judgment that Indian
children raised in certain circumstances today suffer injustice
with regard to language. The past hundred years has seen the
demise of the Indian language with which they might identify,
which they might share, into whose cognitive and literary richness
 Language in Education vii
they cannot enter, yet they mostly do not have the opportunity
of mastering the corresponding dimensions of English. The cul-
tural autonomy of the Indian language, a century ago, allowed a
richness and satisfaction through language that subordination to
an educational system they do not control does not allow today.
Even the vitality of idiom and pronunciation in their Indian Eng-
lish is losing out in the younger generation, together with the
carrying over into English usage of narrative patterns and eti-
quette unconsciously cultivated and enjoyed. Domination does do
something to the life of language in a community, and we must
face the fact. The last embodiments of an ancient linguistic art
of narrative die, and we write, theoretically, as if nothing had
been lost. We escape into an Edenic vision of the equality of
languages and their uses that is contradicted by everything we
know about the history of the past hundred years, probably the
most vicious and destructive hundred years in the history of the
human race. We must despair a little if we are to do much good.
The theme of diversity, inequality, and evolution is elaborated
in the second essay where the original article is expanded by new
sections on writing and on the views of Bernstein and Jurgen
Habermas. The third essay addresses in part the development
of linguistics. I hope that it will help bring to the attention of
educators the importance of linguistics as embodiment of a method-
ology that is something of a third force, so far as the usual con-
trast between qualitative and quantitative methodologies is con-
cerned. Indeed, the essay was written because a conference de-
voted to that contrast ignored linguistics in that respect. So
much for the fame of French structuralism, one might say. But
the first fount of French structuralism, Claude Levi-Strauss,
acquired his sense of the qualitative rigor of linguistics in New
York City during World War II, and it was in American linguis-
tics and anthropology during the 1940s and 1950s that the exten-
sion of linguistic methodology to the rest of cultural life was first
explored. The work of men such as Hockett, Smith, Trager,
Pike, Goodenough, Lounsbury, and others in this regard is not
widely known today, perhaps, but it is part of a development to
be traced to the very roots of American linguistics in Sapir and
Bloomfield, and a development that informs the "ethnolinguistic"
perspective of this book. Models and formats apart, it is the
heritage of an understanding of what it means to establish units
and structure through commutation aid distribution that underlies
the belief in these essays and in the work of other educational
ethnographers, such as David Smith, that the openness of
ethnographic inquiry is not an invitation to subjectivity but a
path to intersubjective objectivity of a particular kind.
Chapter 3 also address what can be called the "arc" of
linguistics, from its separation out of philology, language depart-
ments, and anthropology earlier in this century to its inter-
dependence with psychology, computational sciences, philosophy,
and social science studies of discourse today.
viii introduction
The main direction of the development of linguistics as a
separate discipline has been from phonology (in the 1930s)
through morphology and syntax to discourse. This direction has
been founded on what may be loosely called the "referential"
function of language. The crucial aspect of phonology has been
the organization of speech sounds to distinguish and express ele-
ments of the lexicon and grammar. The crucial aspect of : grammar
has been its organization in relation to propositional content. The
most cultivated aspect of lexicon has been its categorization of
experience.
In recent years there has been growing attention to the
organization of discourse beyond the level of the sentence and to
the interdependence between the organization of discourse and
features of situations, including the identity and roles of par-
ticipants in situations. Perspectives for such attention are old,
but a considerable body of empirical work is recent. Interest in
the ways in which the use of language by and to men and women
differ; in the properties of discourse between professionals
(lawyers, psychiatrists, doctors, and teachers) and clients; in
politeness as a dimension of conversation; in literature; in what
is variously called "pragmatics," "semiotics ," "conversational
analysis," "ethnography of speaking," according to focus and
context, is rife. It appears to lack, however, a comprehensive
grounding in a view of the organization of linguistic means in the
service of such ends.
Much of what is done in the study of politeness, speech acts,
am'. the like is done on top of, or alongside of, standard con-
ceptions of the formal analysis of grammar. The formal models
are used, ignored, or rejected, but, apart from the work of
Michael Halliday, new, thorough-going models of the organization
of language are not discussed.
Such models would seem to be indispensable. When one
starts from the standpoint of talk by and to women, for example,
one must take into account more than is accounted for by stand-
ard formal models. The relevant features of speech may include
details of articulation, intonation, pitch, tempo, lexical choice,
syntactic choice that go together in a fashion or way of spealsing-y--,,
a style. Yet some of those features may not be part of a stand-
ard description of the language. The features are part of the
conventional speech of the persons in question, to be sure, but
a part that focuses on the organization of language in terms of the
"referential" function has set aside. It is the start from a
"higher" level, the organization of language in terms of inter-
action among persons, that brings the other features into view
and finds them essential parts of what language is and does.
In my own opinion, a comprehensive model of the organiza-
tion of language from the ground up, adequate to such inquiry,
would systematically recognize and investigate two complementary
elementary functions. One is that which can be called "referen-
tial" (or anything else), and the other is one which can be called

8
Language in Education ix
"stylistic." The same methodological foundation pertains to both.
Out of the stream of speech we identify certain features of sound
as structurally relevant to the "referential" function by showing
that they "contrast" in distinguishing the features distinguishing
words. "Pig",,and "big" are relevant in the structure of English
(but not necessarily of another language) because the substitu-
tion of the initial sound of one for that of the other results in a
different English word. Just such a /commutation test applies to
features whose substitution results in an utterance being conven-
tionally taken as more formal or self-conscious, more forceful in
intent, the speech of a person of one kind rather than of another;
"bi::: g" as different in degree from "big," but not a different
morpheme; "phig" (with heavily aspirated p) as more forceful
than "pig" with ordinary aspiration; whereas the difference be-
tween "p" and "b" in "pig" and "big.," is not expressive or stylis-
tic since it does make them different, morphemes. The two kinds
of commutation test and feature are interdependent. A feature
that serves the one function cannot simultaneously serve directly
the other: the difference in vowel length in "big," like the differ-
ence in aspiration in the two forms of "pig," can be expressive
just because it does not make the two forms different morphemes.
The difference between "pig" and "big" cannot be expressive
just because it does mark a difference of morpheme.. On the
other hand, a feature grounded structurally in the one function
can be employed indirectly in the pragmatic service of the other:
"Not that one, the bi:,;:g one" (expressive length used to dis-
criminate a referent) ; "Not beautiful, but certainly bootyful"
(said of a rich attorney's elderly ugly daughter, referential
difference enlisted in speech play). Together, the two kinds of
elementary diacritic function provide the true basis for analysis
of language, whether one's concern is with discourse or with
universals of structure.
With regard to the latter: it would commonly be held that
aspiration of consonants is not a universal of language structure.
It distinguishes words "referentially" in some languages (Hindi,
for example), but not others (English, for example). I would
suggest that aspiration is contrastively relevant in stylistic func-
tion in those languages in which it is not relevant in referential
function. If this sugge-slion proves valid, then aspiration is
functionally relevant in all languages. Its relevance is universal;
it is the function in which it is relevant that differs. The same
thing may hold true for vowel length and perhaps other features.
The very goal of formal analysis, a demonstration of what is uni-
versal, would seem to depend on returning to a more comprehen-
sive functional starting point in order to leap further.
With regard to analysis of discourse: if it is necessary to
attend to "stylistic" features in order to understand socially
interesting use of language, then we must encourage a broader
base of training in linguistics itself. The Chomskyan "revolution"
led to a widespread disparagement of phonetics and associated
 x Introduction
kinds of phonology. The ability to transcribe speech has not
been a standard part of linguistic t'aining for some years. And
training in each level of language has been understandably
focused on analysis of its place in models of language conceived
in terms of the "referential" function. Organization in terms of
the "stylistic" function has been set aside as something that
might be dealt with later on. We need people trained to identify
features serving "stylistic" function and to trace their distribu-
tion, interpret their configurations, in relation, not to grammar,
but to the organization of activities, events, and social relation-
ships.
The tradition of dialectology and sociolinguistic
contribute much. The frame of reference, however,variation can
would not
be language change but the varied ways in which varied means of
speech are organized and charged with meaning in community life.
A crosscultural and comparative perspective will be essential. It
is striking, for example, that the work of the Milroys in Belfast
discovers patterns of communication in some working-class neigh-
borhoods that are quite similar to patterns known from some
American Indian groups. The common element would appear to
be people whose several networks of relationship reinforce each
other. Those wh9,live in the same neighborhood also work to-
, gether to a great .extent and to a great eXtent are kin. The
finding suggests that recurrent social structure may be as much
or more an explanation of ways of speaking as geographical
Separation and cultural tradition.
These considerations can be brought to a head by contrasting
two formulations of the fundamental questions of linguistic theory.
Much of the work of Chomsky and others has been associated with
the questions: How is it that a child can 'acquire a grammar of
(technically) infinite capacity on the basis of finite experience?
What form 'of grammar can account for all the grammatical sen-
tences of a language while distinguishing them from the ungram-
matical? The perspective suggested here would address the ques-
tions: How do members of a group use language in order to con-
duct a certain activity? How do members of a group use language
in order to be taken as a certain kind of person, status, role, or
the like? The second set of questions entails a mode of organiza-
tion of linguistic means and a set of capabilities on the part of
speakers, just as does the first. Many linguists might accept
both sets of questions as pertinent, but think that the work
associated with the first set must be accomplished before there
can be answers to the second set. My argument is that one
cannot get from the one to the other. The work associated with
the first set leaves out of account features and relationships that
are essential to the second. In order to deal with the second set
of questions, one needs a broader starting point.
When one adopts the broader starting point, one finds, I
think, that the other starting point appears as a specialization
within its framework. The specialization establishes the nature of
Language in Education xi

many, but not all, the resources made use of in the organization
of speech in social life. One way to summarize the difference is
to contrast the organization of speech in terms of levels with
organization in terms of what may be called "lenses" (the term is
taken from my colleague, Claire Woods-Elliott). When we review
the development of linguistics as a discipline, focused around
issues first of phonology and then later of syntax, and still
later, now, of discourse, we see an evolution in terms of levels.
The connections between levels occupy much of our attention.
We do not much think of phonology, for example, as a sphere of
the organization of language which might be involved in social
and cultural life in interestingly different ways in different
places. We do know that some languages seem to revel in
onomatopoetic and sound symbolic possibilities and others to be
chary of them. We know that the differences between men and
women enter into the very phonological shape of words in a few
languages, although not in most. We are aware that it seems
possible to describe the phonology of some languages almost with-
out regard to the grammar, whereas in other languages the two
seem inextricable. Sometimes it seems necessary to rely on some-
thing phonological to delimit something grammatical, and sometimes
the other way round. Sometimes particular sectors of the lexicon
appear to have partly their own phonological character. When
languages are written, whether or not phonological shape is
represented, and if so, what portion, varies from culture to cul-
ture (cf. Chinese,, Arabic, English). Such considerations point_
to a view of phonology as a sphere of its own, a "lens" of its
own, as it were, through which to view the interaction of lan-
guage and social life, a "lens" that may have a greater part in
some societies, a lesser part in others. The fit of phonology
into a grammar, as one level among others, is itself something
that seems culturally variable.
A universal model of language is necessary to account for the
parameters of organization present in all languages. The fit
among spheres of language appears as more than a matter of re-
lations among levels. The spheres themselves seem to differ from
one case to the next in terms of the amount and kind of attention
given them over time by different communities and in terms of
the way and degree to which they have been integrated into the
rest of grammar, other modes of signalling, such as writing and
the symbolic and expressive activities of concern to the group.
Morphology, syntax, and lexicon appear as partly autonomous
lenses as well. The degree of independence and interdependence,
the degree of elaboration, the relation to writing and other forms
of signalling all vary in ways that say something about the place
of language in social life.
From this point of view, the description of the phonological
sphere or lens in a given case would comprise four aspects: the
inventory of phonological means, of course, including features
relevant both "referentially" and "stylistically"; attitudes and
xii Introduction
beliefs towards speech sound, including interest in sound symbol-
ism, speech play, correctness of pronunciation, and the like; the
organization and integration of phonological features in relation to
each other and in relation to the rest of the language; the organi-
zation and integration of phonological features in relation to kinds
of events, activities, and participants in them. The same four
aspects would apply to the sphere or lens of morphology, or syn-
tax, and of lexicon.
The middle chapters, 4 to 6, are concerned in complementary
ways with what counts as legitimate knowledge and who is counted
as entitled to know. The tempo of ethnography is in some re-
spects conservative; its results are the better for ripening in the
mind; insofar as it makes local practice intelligible, it may lessen
impetus or optimism for change. By legitimatizing the knowledge
of the participants in educational settings and by giving weight
to the universal human need for self-worth, ethnography is likely
to make it difficult- to argue solutions that take for granted the
fault or failure of teachers, of parents, or of some other category
of scapegoat. Ethnographic inquiry is likely to show people doing
the best they can with what they have to work with, given what
it is possible and reasonable for them to believe and do. The
ultimate result of ethnography, of course, may be radical. It
may suggest that some desired outcomes are impossible, given
what the society is willing to spend on schooling, and the conse-
quences in what children must find in schools, in terms of re-
sources, meaningful activities, minutes of meaningful guidance,
and communication. To empower teachers as having legitimate
knowledge may disrupt some practices. Just so, too, for princi-
pals in systems where principals are treated by those above them,
not as instructional leaders, but as hired hands. By making
particular situations palpable, credible, a living part of the imagi-
nation, ethnographic accounts may make it more difficult to impose
uniform general solutions that are arbitrary in local settings.
When some charge that ethnography does not permit generalize-
tion they may be shrewder than they know. Whose power is
hurt if the pretense of theoretically generalizable results is
stripped away? If educational practice is found to require, not
the application of general theory, but the discovery of new,
local knowledge?
In any case, the question of methodological preference is
secondary to the question of what we want to know. In anthro-
pology itself, ethnography comprises a wide variety of techniques
and methods and is combined with a variety of modes of compara-
tive analysis. If one truly wants to know about a culture, a
society, a way of life, one uses all there is to use. One does not
refuse to know something because it is known in a certain way.
Just so with schools and the educational configurations of neigh-
borhoods and communities. If we truly want to know them, we
will welcome and use every approach that can contribute.
Ethnography is indispensable and, to my way of thinking,
12
 Language in Education xiii
fundamental, but it is not the name of a methodological virgin
undefiled. It can embrace anything useful so long as it can make
the bed it lies in.
It remains that ethnography makes frequent use of narrative
accounts and itself is often cast in narrative form. The status of
narrative as a form of knowledge is addressed in chapter 7, and
I am indebted to the co-author, Courtney Cazden, for letting me
include it here. (I would also like to thank Lisa Delpit for her
thoughtful comments on it.)
Chapter 8 weaves together many of the themes of the book.
It and the preceding chapters should make clear that I share the
concern of Ray Rist (1980) and Elizabeth Brandt (1980) that an
ethnographic, or ethnolinguistic, perspective not be trivialized
and vulgarized. I am far from holding that educational ethnogra-
phy should be restricted to what ethnography means to one or
another anthropologist. Indeed, the history of ethnographic in-
quiry is checkered enough to provide precedent for a great
variety of practices, including sending students out with a check-
list or having informants brought to the veranda of one's hotel
for interview. The roots of ethnography in anthropology remain
important because they anchor educational ethnography in a
commitment to cumulative knowledge and comparative analysis of
particular eases. Such roots tie educational ethnography to a
rhythm of inquiry dictated by the nature of what it is one seeks
to know, as against the arbitrary timing of other forces. Valid
knowledge and successful change are both likely to require long-
term involvement and continuing relationships.
Chapter 8 ends by emphasizing the role of Schools of Educa-
tion, as those parts of institutions of learning and inquiry where
the problems of language in education can most readily and stead-
ily be addressed. Let me close this introduction with a few addi-
tional words on the long-term strategy that should inform such
efforts. (I use "strategy" in its standard sense of long-range
plans toward a goal, the sense in which it translates into English
concerns such as those of Mao Ze Dong when he assessed the
future of his country in the midst of its war with Japan and of
Antonio Gramsci when, gradually dying in one of Mussolini's jails,
he reflected on "wars of movement" and "wars of position" in
relation to the ultimate liberation of Italian society. This comment
is made necessary by the degenerate use of "strategy" in lin-
guistics sometimes to denote a single choice or device.)
I assume a shared concern to change schooling, and education
generally, for the better, and a willingness to consider an ethno-
graphic way of working an essential part of such change. Given
such assumptions, it is essential that our strategy or change
concentrate our energies on the true obstacles to bo overcome,
the limitations of our knowledge, and the limitationv ,af what can
be accomplished within the existing resources and organization of
education in this country. Our strategy must minimize diversion
of energies to conflict over labels, over disciplinary proprietorship
xiv Introduction
of a given method or perspec::Live, over the propriety of certain
techniques or fields or persons contributing to knowledge. It is
of course an inveterate characteristic of colleges and universities
to experience tension along such lines of fault in the scholarly
terrain. But if our energies are diverted to conflict between
liberal arts departments and professional school faculty, between
academic personnel and practitioners in schools, between "eth-
nography" and other modes of inquiry, we harm our major pur-
pose, and do the work of those who would prefer that what we
can discover not be known, and certainly that it not lead to
change.
Let us avoid, for example, christening our insights with
noble names so that the contribution to our' work of sociology,
for example, is taken as divided among "ethnomethodology" or
?In eopraxiology," "cognitive sociology," "wild sociology," "reflec-
tive sociology," "critical theory," "dramaturgical approaches in
sociology," "sociological behaviorism ," "phenomenological soci-
ology," "sociological introspection," "existential sociology" (see
Douglas and Johnson 1977:xiii for this catalogue and differenti-
ations among its elements). Such distinctions and tensions are
to be taken as secondary to the "principal contradiction" that
confronts us. That contradiction is between the dominant form
of legitimate knowledge and legitimate knowers, with respect to
education and schooling, and the legitimation of kinds of knowl-
edge and knowing that "ethnography," "ethnomethodology," and
many other trends in the philosophy and practice of social sci-
ence encourage. The fundamental point in common is an under-
standing of social life as something not given in advance and
a priori, but as having an ineradicable aspect of being consti-
tuted by its participants in an ongoing, evolving way. Those
who accept this point can agree on giving priority to discovery
of what is actually done in local settings and of what it means
to its participants. The concomitant of-that priority is an em-
powering of participants as sources of knowledge. What is ulti-
mately known about a situation may not match the view of any
one participant, but its validity and value will be compelling.
And the concomitant changes in practices of inquiry and in rela-
tions between knowers and known may themselves be quietly
radical, an essential step toward larger change. Work of this
kind is already going on in a number of places, and it is a rea-
son for hope, rather than despair, about the future of education.
In changing what goes on in our own work place and in its rela-
tionships to its immediate context, we clear and cultivate a soil,
from which much may grow.
References
Anderson, Perry. "The Antinomies of Antonio Gramsci." New
Left Review 100: 5-80, 1976-77

14
Language in Education xv
Brandt, Elizabeth. "Popularity and Peril: Ethnography and Edu-
cation." Paper presented at 40th annual meeting of the
Society for Applied Anthropology, held March 19-22, 1980,
in Denver, Colorado
Douglas, Jack D. and John M. Johnson, eds. Existential
Sociology. Cambridge, London, New York: Cambridge Uni-
versity Press, 1977
Gumperz, John J. and Dell Hymes, eds. Directions in Socio-
linguistics: The Ethnography of Communication. New York:
Holt, Rinehart and Winston, 1972
Hymes, Dell. "Introduction." In Courtney Cazden, Vera Johns,
and Dell Hymes, eds. Functions of Language in the Class-
room. New York: Teachers College Press, 1972
. Foundations in Sociolinguistics: An Ethnographic Approach.
Philadelphia: University of Pennsylvania Press, 1974
. Soziolinguistik. Zur Ethnographie der Kornrnunikation.
Eingeleitet and herausgegeben von Florian Coulmas. Frank-
furt am Main: Suhrkamp, 1979
. "Blitzkreig and Protracted Warfare; or Wars of Movement
and of Position; or Rist-slapping." Keynote address delivered
at the Ethnography in Education Research Forum, March 14,
1980, Philadelphia, Pennsylvania
Mao Tse-Tung (Mao Ze Dong). On Protracted War. Peking:
Languages Press, 1967. (First edition, 1954; from a series
of lectures, May 26-June 3, 1938.)
Milroy, Lesley. Language and Social Networks. (Language in
Society, 2.) Baltimore: University Park Press, 1980
Rist, Ray C. "Blitzkrieg Ethnography: On the Transformation
of a Method into a Movement." Educational Researcher 9: 8-
10, 1980
Spolsky, Bernard. Educational Linguistics: An Introduction.
Rowley, Mass.: Newbury House, 1978
Woods-Elliott, Claire and Dell Hymes. "Issues in Literacy:
Different Lenses." Commissioned paper for Stephen Reder,
Northwest Regional Educational Laboratory, for report to
National institute of Education, 1980

15
 functions of speech: an
evolutionary approach

Let me begin by stating the thesis that lies behind my title. I

want to controvert two widely accepted views, first, that all lan-
guages are functionally equivalent, and second, that all languages
are evolutionarily on a par. I want to maintain that the role of
speech is not the same in every society, and that the differences
can best be understood from an evolutionary point of view; that
we must understand speech habits as functionally varying in their
adaptation to particular social and natural environments, and recog-
nize that there are ways in which some languages are evolutionarily
more advanced than others. Letting "speech habits" stand for the
gamut of linguistic phenomena and "functions" for the varied roles
these play, I am arguing for an evolutionary, comparative approach
to functions of speech. Such an approach does not now exist in
anthropology. I shall indicate reasons for the present neglect,
and try to show that by overcoming it, anthropology will contrib-
ute to both its own theory and the foundations of education.
There can be, particularly, a contribution to some problems of
education in the rapidly changing modern world, especially in
underdeveloped and linguistically complex areas. We are all aware
that, given the great surge throughout the world toward social
and economic progress, the only feasible goal is for all to share as
equitably and peacefully as possible in the fruits of industrialized
civilization. To attain this goal in many areas requires the intro-
duction of new educational forms and content, and we must help
in this introduction while maintaining and enhancing the quality of
education as part of a democratic way of life in our own country.
And while success depends much upon problems which are political
and economic, it also involves problems which have to do with the
functions of speech.
As an instance of a problem encountered widely, we can cite
the Mezquital Otomi of Mexico. The need here is through educa-
tion to enable a group to overcome its poverty and isolation from
the national society. Dr. Manuel Gamio, father of applied anthro-
pology in Mexico, once commented that Otomi cultural character
had changed very little during the twenty-five years of his active
work among them. Part of the problem is an arid environment,
but a missionary linguist writes that "the comparatively high de-
gree of monolingualism in the tribe, forming an immediate barrier
to fusion with the official system of education executed in Spanish,
a language foreign to most of the members of the tribe, is an
obstacle to progress tantamount to the imposing economic one."1
And the obstacle of monolingualism in such a case can best be
overcome with the help of an adequate analysis of the functions of
1
 2 Functions of Speech
speech from a general point of view. Let us do this by consider-
ing what it means for a child (or an adult) to master the speech
habits of a group, to function as a linguistically normal member of
it.
FUNCTIONS OF SPEECH
When we think of learning a language, we may think first of
rules of pronunciation, grammar, and vocabulary, but there is
clearly more than this to the acquisition of a form of speech. A
person could master these rules but still be unable to use them.
He could produce any possible utterance but not know which
possible utterance to produce in a given situation, or whether to
produce any. If he spoke, he might say something phonologically,
grammatically, and semantically correct, but wrong, because in-
appropriate. He might find hearers (or correspondents) "taking
him the wrong way" or responding in ways that indicated that,
although understood, he was not a normal member of the speech
community.
In a society speech as an activity is not a simple function of
the structure and meanings of the language or languages involved.
Nor is the speech activity random. Like the languages, it is pat-
terned, governed by rules, and this patterning also must be
learned by linguistically normal participants in the society. More-
over, the patterning of speech activity is not the same from
society to society, or from group to group within societies such
as our own.
The nature of such patterning, as well as its cross-cultural
variation, can be brought out by considering four aspects of it:
(1) in terms of the materials of speech, there is the patterning of
utterances in discourse; (2) in terms of the individual participants,
there is the patterning of expression and interpretation of person-
ality; (3) in terms of the social system, there is the patterning of
speech situations; and (4) in terms of cultural values and outlook,
there is the patterning of attitudes and conceptions about speech.
Let us briefly take up each in turn.
1. Beyond the syntactic structure of sentences (with which
grammars usually deal), utterances have an organization into what
we may call 'routines.' By 'linguistic routine,' I refer to sequen-
tial organization, what follows what, either on the part of a single
individual or in interchange between more than one. 2 Routines
range from reciting the alphabet, counting, and greeting, to the
sonnet form, the marriage ceremony, and the direction of a buffalo
hunt. Obviously, societies and groups differ both in the content
of equivalent routines, such as those for greeting, and in the
kinds and numbers of their routines. The more complex the
society, the greater the number and variety of routines, and the
greater the variation in control of routines by individuals. 3
2. Persons who participate in speech activity learn the pat-
terning of its use as medium for personality and role-playing.
Language in Education 3

Cues expressed and perceived in speech may enable individuals to

place, and to adjust quite subtly to, each Other. This complex
process ranges from tempo and general handling of voice dynamics
to choice of expressions and over-all style. The individual learns
both signaling patterns outside language proper and the integra-
tion of these in speaking (and, correspondingly, in writing).
Such signals differ from group to group, of course, and can be
misinterpreted, either in themselves or as part of other behavior.
Thus, as James Sledd observes:
British speakers have far more final rising pitches in
statements than do Americans, whose favorite intonation
pattern /231#/ sounds brusque to British ears. British-
ers are also likely to use a greater range of pitches than
Americans, more frequent and extreme pitch changes,
and more numerous expressive devices. ...In some
parts of the United States, an adult male who talked so
would be suspect. 84
In a school for IVIesquaki Fox Indian children near Tama, Iowa,
many white teachers who probably regard their classroom behavior
as normal have had loudness of voice, together with verbal direct-
ness, interpreted by Indian pupils as "mean"-ness and a tendency
to "get mad."
The relative importance of speech to personality, vis-à-vis
other modes of activity and communication, varies from person to
person and group to group, and so do the range of expression
and interpretation of personality possible in speech, the extent to
which speech is a form of gratification (oral or other), and the
importance of speech for role performance and attaining rewards,
especially those depending on personal interaction. Among the
Ngoni of Africa, rules of speaking etiquette are strict, and skill
in speech is greatly encouraged, for such skill is considered part
of what it means to be a true Ngoni.5 Contrast this to concep-
tions of the "strong, silent type," the "man of few words," etc.
in sectors of American society.
3. Social systems are often regarded as patterned relation-
ships among roles and among groups such as families, lineages,
and corporations, and there are speech patterns diagnostic or
characteristic of particular roles and groups, just as there are
speech patterns diagnostic or characteristic of particular person-
alities. These of course differ cross-culturally in content and
relative importance. If we also look at a social system in terms of
the behavioral activity involved, we can see it as a network of
interaction in situations or behavior settings, and can discover
related patterns of speech. For example, societies differ in the
settings in which speech is prescribed, proscribed, or simply
optional. We so commonly think a social situation requires some-
thing to be said that writers have described this as a universal
need.6 Certainly, were someone to come to your or my house,
4 Functions of Speech
sit silently for half an hour, and leave still silent, we should not
consider it normal. In some America_n Indian groups this would
constitute an acceptable social visit. For them, physical presence
is enough; the situation is defined as one in which speech is not
necessary when one has nothing to say.
If we look at a social network in terms of speech settings,
we can discriminate a set of factors whose interrelations may serve
to describe its patterns of speech activity, and so provide a basis
for comparing the functions of speech in different social systems.
These factors can be termed: a sender (or source); a receiver
(or destination); a message (viewed in terms of its form or shape);
a channel; a code; a topic; a context (setting, situation, scene).
All are compresent in speech activity. Societies differ in what can
function as an instance of each factor, and in the relations of
appropriateness which obtain among the factors in given cases.
There is a system of speech activity in a society, then, because
not all possible combinations of particular senders, receivers,
message forms, channels, codes, topics, and contexts can occur.
A teacher in a school for Navaho children may discover that
one boy cannot speak to a girl classmate because she stands in
a certain kinship relation to him. A society may traditionally per-
mit only certain individuals to use the channel of writing, and
among the otherwise non-literate Hanunoo of the Philippines, writ-
ing is used only among young people in courtship and love affairs.
Education for birth control may encounter the barrier that such a
topic cannot be discussed among or in the presence of both sexes,
including husbands and wives. A teacher may misinterpret an
ornate and allusive style in an examination as an attempt to con-
ceal ignorance of the answers, not realizing that Puerto Rican stu-
dents may deem it the only style appropriate to such an occasion.
One teacher in a project of fundamental education may find it hard
to teach children to define the classroom situation as one in which
they do not talk to each other, and in which they speak to her
only when asked or acknowledged. In another society a teacher
may find it equally difficult to bring children to define the class-
room situation as one in which they can speak at all, they having
learned to regard instruction as a situation in which they function
as receivers only.
These scattered examples must suffice here, except to note,
under the code factor, the importance of levels, styles, and func-
tional varieties of a language, and in some societies, of entirely
different languages, in relation to particular settings, channels,
senders, and receivers. Here rules of appropriateness may make
a great difference, especially if they differ for teacher and stu-
dents because of differences in class or cultural background.
4. Cultural attitudes and conceptions regarding speech differ
notably from society to society and also from class to class. The
pattern of such attitudes and conceptions permeates the role of
speech in personality and social structure. Reciprocally, differ-
ences with regard to the interest in and valuation of speech (or of

19
 Language in Education 5

a particular linguistic code) may have correlates in differences

with regard to how speech enters into the socialization and early
education of children. It is clear from ethnographic sources that
societies differ as to their conceptions of
language, and of the process of languagechildren as users of
learning; as to the stage
in children's speech development at which major socialization pres-
sure is exerted; as to the extent to which interest in speech and
speech play is encouraged or discouraged; as to the extent to
which speech is a mode of reward and punishment for children;
and as to the portions of culture which are linguistically communi-
cated. Some societies are permissive about eating and toilet train-
ing until the child can understand verbal explanations, whereas
others conceive a newborn child as capable of understanding
speech, and lecture it from the cradle. Adult skills are trans-
mitted verbally for the most part in many societies, but among
societies such as the Kaska of northwestern Canada, children
learn them almost wholly through observation and imitation.
Differences as to the functions of speech in adult life probably
are related to such differences as these in the functions of speech
in childhood, but there has not been the systematic comparative
study which would permit us to be sure.
In any event, although we tend to think first of cases in
which language has been integral to a group's sense of identity
and unity, and in which it is thus a focus of pride, it is clear
that here also the function of speech may vary.
functions and three correlated attitudes have beenFour distinct
differentiated
by students of the development of standard languages, and these
can be applied generally to all languages. ' The first two functions
are separatist and unifying, jointly associated with an attitude of
language loyalty. There is a prestige function associated with
language pride, and a frame of reference function associated with
awareness of a norm.
Two South American peoples contrast sharply
the separatist and unifying functions. The Fulniowith regard to
of Brazil have
abandoned their homes several times in the past three centuries to
avoid assimilation by Brazilian national society. The preservation
of their language and an annual religious ceremony have been the
basis as well as the symbol of their distinct identity. The Gua-
queries, a Venezuelan group, seem to have abandoned their lan-
guage and native religion perhaps as early as the eighteenth cen-
tury, but the society thrives as a distinct identity within the
Venezuelan nation, through maintenance of a special socio-economic
base. 8
Two North American groups contrast sharply with regard to
the prestige function. Three centuries ago a Tewa-speaking
fled from the Spanish to find refuge on one of the Hopi mesasgroup in
Arizona. There, as the Hopi-Tewa, they have maintained a posi-
tion as a specially regarded and privileged minority--a situation in
which attitudes toward language have been a major factor.
to their Tewa dialect has had a separatist and unifying role,Loyalty
as

20
6 Functions of Speech
has their persuasion of the Hopi, through the reiteration of a
myth and constant ridicule, that no Hopi can learn their lan-
guage. They in turn have a reputation as polyglots for their
own knowledge of Hopi, and often of Navaho and English; they
haiPe maintained pride in their language, and have won linguistic
prestige for it and themselves. 9 In contrast, the Eastern Chero-
kees of North Carolina, a remnant group, retain their language-in
large part, but without pride. It is a source neither of prestige
nor of unity, persisting only in a separatist function with a nega-
tive, anti-White language loyalty.10
In Mexico the Zapotecs of the Isthmus of Tehuantepec resem-
ble the Hopi-Tewa in the fact that, as a group. bilingual in Span-
ish and Zapotec, they retain pride in their first language and
national identity, and these are accorded prestige by those
around them. In contrast, language has been salient in the cul-
tural persistence of the Otomi against Spanish pressure, but the
Otomi have accepted an outside valuation of their language as in-
ferior to Spanish, and feel no prestige in its use. Language
loyalty to Otomi makes imposition of education in Spanish alone
impossible, but acceptance of prestige for Spanish alone makes
education in Otomi alone unacceptable; it suggests an attempt to
keep them in an inferior status. Bilingual education, using diglot
texts, wins acceptance, by reassurance that knowledge of Spanish
is the end in view. (Such bilingual education may, of course,
come to enhance the prestige of Otomi.)
How groups differ in the degree to which a language serves
as a frame of reference in the sense of awareness of a norm is
noticeable in attitudes among themselves towards incorrectness or
slovenliness of speech. Among some American Indian groups such
as the Washo and Paiute, a child might receive as a nickname a
word it frequently mispronounced. Attitudes towards correctness
among foreigners may depend upon the identity of the speaker.
Many Frenchmen find a Spanish or Italian accent charming but a
Germanic accent unbearable. The choice of teaching personnel
and procedures obviously would pose a different problem among
linguistic sticklers, such as the Ngoni of Africa, from that posed
among linguistically more laissez-faire peoples.
The functions of writing systems in these respects are often
significant for attempts to introduce literacy and new education.
Often, as among the Otomi, a successful orthography for the
native language, if it is to be easily accepted, must depart from
scientific accuracy to resemble a prestigeful other written lan-
guage. Native conceptions (folk-linguistics) enter too, as when
the writing of tones with accent marks was found impractical
among the Soytaltepec Mazatec of Mexico. They conceive of tones,
not as high and low, but as thick and thin, and it makes no
sense to try to teach them the rule that the mark for the "high"
tone slants up and that for the "low" tone slants down. Printing
expense and legibility make the use of their own metaphor of thick
 Language in Education 7

and thin impractical also, but superscript numbers for tones have
proved successful."
Even in this cursory survey, we see that the role of language
and linguistic activity can vary greatly from group to group, and
we can begin to see more clearly how this variation matters for
practical problems of raising the educational level of the world. In
programs of fundamental education such as UNESCO has sponsored,
for example, literacy must often be introduced. To this end, one
among several dialects or languages often must be chosen as the
medium of education, and
or constructed. Whether often an orthography must be selected
literacy is already present or not, new
speech habits and verbal training must be introduced, necessarily
by particular sources to particular receivers, using a particular
code with messages of particular forms via particular channels,
about particular topics and in particular settings--and all this from
and to people for whom there already exist definite patternings of
linguistic routines, of personality expression via speech, of uses
of speech in social situations, of attitudes and conceptions toward
speech. It seems reasonable that success in such an educational
venture will be enhanced by an understanding of this existing
structure, because the innovators' efforts will be perceived and
judged in terms of it, and innovations
greater success than those which crosswhich mesh with it will have
its grain.
There is direct analogy with the
sounds of another language in terms fact
that one perceives the
one's own. 12 This phenomenon perception of the structure of sounds in
terms of one's own--has been studied by linguistsof another system in
as interference
between two systems, most notably by Uriel Weinreich of Columbia
University in his book Languages in Contact. 13 When both systems
in question are known, it is possible to predict quite accurately
where and what kind of interference will occur, and what kinds of
substitutions and interpretations will be made, as speakers of one
language learn the other. In consequence, it is possible to design
materials for the teaching of one language specifically for speakers
of another, and to anticipate the particular advantages and disad-
vantages their own system will confer in the task.
This suggests the nature of the contribution that anthropology
can make to such problems in education. It would be a matter of
applied anthropology, defined as "the formal utilization of social
science knowledge...to understand regularities in cultural pro-
cesses and to achieve directed culture change." 14 An adequate
comparative study of the functions of speech would imply a de-
scriptive science of the functions of speech, just as there is a
descriptive science which deals with language structures. Stich a
science would provide a basis for detailed analysis of the differ-
ing systems of speech activity which meet in an educational situ-
ation, and such analysis would make it possible to predict or at
least to anticipate more effectively the interference which a pro-
gram of literacy, bilingual education, and so forth would en-
counter. Even the broad conceptual analysis outlined above can
8 Functions of Speech
help by calling attention to aspects of the problem, such as the
Mesquaki children's perception of teachers as "mean," or the need
for a successful written form of Otomi to resemble that of Spanish.
But there must be detailed empirical studies, from which can
emerge a more refined theory and the descriptive science I have
advocated.
It is remarkable that no such comparative study of speech
functions exists. Anthropology is noted for just this sort of
cross-cultural perspective when it is a matter of religion, of kin-
ship, of sexual behavior, of adolescent crisis. Why not when it
is a matter of the functioning of speech in society?
The answer lies in the theoretical perspective on the function-
ing of speech now usual in anthropology, a perspective which is
non-evolutionary and minimizes cross-cultural variation. So I must
now sketch an evolutionary perspective, as framework for a short
critique of current anthropological views and a basis for a broader
concluding interpretation of the educational aspects of functions of
speech.
AN EVOLUTIONARY PERSPECTIVE
There is no single or monolithic body of evolutionary theory in
anthropology and biology, but there is a body of recent literature
from which we can single out some essential features for applica-
tion to speech.15
First, there is an essential distinction between two kinds of
evolutionary study, namely of specific evolution and of general
evolution. Specific evolution is concerned with individual lines of
evolution, the development and adaptation of particular groups in
particular environments. General evolution is concerned with the
course of evolution as a whole. It abstracts from and often cuts
across individual lines of evolution to consider types, as these
have emerged and as these represent broad levels of evolutionary
advance. General evolution might consider the relation between the
mammal and marsupial types, and the advance which led to the
dominance of the former. Specific evolution would examine such
questions as the adaptation of the whale to marine existence, the
bat to flight, and the radiation of the kangaroo line of marsupials
into various ecological niches in Australia. In terms of specific
evolution, advance means improved adaptation to the particular en-
vironment relevant to a group and in relation to those with whom
it is in .direct competition in that environment; outside this context
it is relativistic. In terms of general evolution, advance means
progress which emerges in the course of specific evolution and has
consequences for it, but considered in a broader spatial, temporal,
and environmental context. Thus a familiar case is the successful
adaptation to a specific environment which proves fatal in the long
run. Criteria for general evolutionary advance include "change
in the direction of increase in range and variety of adjustments to
its environment" and succession of dominant types.16
 Language in Education

When we study evolution specifically, we find that its focus

is upon a population, and a set of traits associated therewith;
that it analyzes the variation in traits within a population, and
the differential retention of traits within the population over a
period of time; that it interprets
tation of the population within its this process through the adap-
environment (and of the traits
to one another), in connection with the pressures
tively affect the retention of traits and hence this which selec-
adaptation. It
sees a population and its characteristics as participating in a con-
tinuous process of change, and it interprets the change
characteristics of the population in broadly contextual andand the
func-
tional terms.
It is easy enough to see linguistic change in these terms:
speech community has a certain set of speech habits, whose inci-a
dence varies within the population and which are differentially re-
tained, as a result of selective pressures (such as the social and
natural environment, prestige of speakers, customs such as tabu
and word-play, and internal requirements for maintenance of the
linguistic code), the whole being adaptive both to the environment
of the speech habits and to the maintenance of the code.
If we carry through such a view, however, we can find our-
selves in a stance quite different from that typical of the attitude
toward language today. Our broad category of speech habits in
relation to a population , as the unit of primary focus, does
the first instance isolate the formal structure of the linguisticnot in
code, the usual object of linguistic
of the uses of speech. Both wouldattention, from the patterning
equally be analytical abstrac-
tions from the same phenomenal reality, the speech activity of the
population. We begin by considering the totality of the speech
habits of the population, and so subsume at first the presence not
only of different types, varieties, and dialects, but even of differ-
ent languages as parts of the whole.
Since one readily takes a functional view of traits involved
with selective pressures, it becomes quite natural to analyze indi-
vidual speech habits or sets of habits, including those of separate
languages, in terms of competition within the environment of a
population, and to see this competition as turning partly on the
merits of the habits themselves. Such a view requires one to con-
sider what the relevant environment for the adaptation of a popu-
lation's speech habits really is at present, and is necessarily con-
cerned with the real locus of change of such habits in the speech
activity of definite individuals living in a definite society. Such
an evolutionary view will direct attention toward the variation
within and between the speech habits of populations, and will
give due impcirtance to the differences in the functions of speech
associated with a set of speech habits of a population, to the
consequences of this variation, and to the evolutionary survival,
development, or disappearance of traits or sets of traits.17
In short, the evolving units are sets of speech habits as
characteristics of populations- -units which sometimes will, and

2.1
10 Functions of Speech
sometimes will not, coincide with the historical units known to us
as languages. We begin by examining natural totalities of speech
habits firmly embedded in environmental context, cultural and
physical, as adaptive to that context as an integral part of the
whole socio-cultural adaptation of the population.
If we examine our recurrent example, the Mezquital
from this perspective, we find that an Otomi child grows into, and
acquires its education as a member of, a population whose speech
habits comprise a majority of Otomi provenience, a minority of
Spanish origin. The child becomes well aware of the competition,
selection, and specialization, among these two sets of habits. The
Otomi habits are dominant in the sphere of subsistence and in
most of social life, occupying a. privileged position in the early
socialization of young and the loyalities of all, and generally in the
tribal environment within which its adaptation has almost wholly
taken place. But Spanish habits are dominant in certain situations
such as the market and the classroom, and as the relevant environ-
ment of Otomi life and speech habits shifts and enlarges, the posi-
tion of Spanish speech habits is enhanced, and more situations are
encountered in which Spanish has selective advantage. The pres-
tige which all accord to Spanish speech habits, and the experi-
enced relatively greater utility of these situations in the expand-
ing sphere of the environment, underlie the general expectation
that the relative function of Spanish will continually expand - -an
expectation which in turn helps bring about the expected state
of affairs. As for development towards filling the enlarged en-
vironment on the part of Otomi speech habits, this occurs only in-
directly as a by-product of bilingual education and the inculcation
of Spanish.
Although the published analysis of the Otomi case is one of
the few such, it is brief, and its focus is not upon the kinds of
questions and of data which an evolutionary perspective requires.18
The discussion is cast in terms of the proposition that all lan-
guages are functionally equivalent and equal, neutral in their own
cultural settings, and. Otomi is seen as having been forced into a
status of ascribed inferiority, because of inferiority ascribed to
its associated, culture. This is considered to be entirely a social-
psychological matter of prestige, having no basis in anything con-
nected with Otomi as a language. Now this assumption is remark-
able, since the discussion does mention concrete linguistic differ-
ences in the functional value of Otomi and Spanish in certain situ-
ations, and despite the declared equality of Otomi, the possibility
that Otomi can be developed to meet the modern educational needs
of its 'speakers is not considered.
While the superiority of Spanish in the situation is partly a
matter of attitude and prestige, it is at best ingenuous not to see
it as also partly a matter of the actual linguistic superiority of
Spanish. The failure to see this is in part due to heavy reliance
on the view that there is no evolutionary superiority among lan-
guages, while in fact the superiority of Spanish in the situation

25
 Language in Education 11

is in part a consequence of its being one of a type of evolution-

arily more advanced language. Let us now turn to the study of
general evolution, which deals with this question.
We mentioned. "increase in range and variety of adjustments to
environment" and succession of dominant types as two
evolutionary advance as between general types. Whencriteria of
such cri-
teria are applied to culture, it is generally agreed that some cul-
tures are technologically more advanced than others. Vocabulary
is the linguistic analogue of technology...." Clearly the lexical con-
tent of standard languages shows increase in range and variety of
adjustments to environment in comparison to dialects.
minority languages, 'where these are not supported byor aregional
standard
or
language outside the situation. World languages such as English,
French, Spanish, Russian, and Chinese show such increase, and
have spread as representatives of a dominant type, quite apart
from military conquests. The existence of linguistic science itself,
and of the self-consciousness and awareness and control which go
into the construction of logics and systems of mathematical manipu-
lation, argue for the advanced status, as a type, of languages
which participate in, and indeed make possible, such activities.
The same holds for linguistic routines in philosophy, literature,
religion , and science. It has been argued that mathematics, logic,
et al., are not language itself, but "post-language. "3. 9 Even, if
this view is taken, it remains that ordinary language is the medium
in which "post-language" 'systems must ultimately be interpreted,
and not all natural languages car perform this function. Indeed,
differentiation and specialization of function is an important aspect
of evolutionary change in languages, as has long been recognized
by students of the development of standardized languages. Of
course, increase in number and diversity of functions of a lan-
guage is a response to change in other aspects of a culture. This
is true also of increase in the content and complexity of the vocab-
ulary of a language. Some scholars may point to this
reason for disregarding such changes as not properly afact as a
linguistic
problem, or as not part of language. The argument does not hold.
Many, if not all, linguistic changes have sociocultural roots. Lexi-
cal borrowing, a standard
As part of general evolutiontopic in linguistics, is a case in point.
ity in the lexical content and functions of then,
in language,
a
increased complex-
language cannot be
disregarded. The response is linguistic, even if the stimulus is
not.
We may bypass the question of evolutionary advance in gram-
matical features, a question raised especially in this century by
leading French linguists, and also the question of increased
efficiency and economy in language evolution. 2 ° Both possibilities
may be regarded, not
of general evolutionaryasadvance
disproved, but as unproved. The reality
in the sphere of language seems
clear.
It may be pointed out that any set of speech habits is capable
of expanding in content and functions sufficiently to serve a
12 Functions of Speech
complex civilization and its associated systems of thought. Yes,
of course, potentially it can so serve, but we must distinguish
between potential and actual development, recognizing that some
languages are actually of the more advanced type while others
are not.
If this distinction is valid, why is it not part of the common
perspective of linguistics and anthropology today or at least a
subject of discussion? Particularly now that an evolutionary per-
spective toward culture is being renewed in anthropology, how
can a part of culture, language, be omitted? The answer lies in
a dominant outlook, whose focus is upon the single language and
its most highly formal, structured aspect, its grammar and pho-
nology considered in abstraction in the first instance apart from
cultural and natural context. Vocabulary, the aspect most closely
tied to this context, is likely to be treated as residual. The
thrust has been to exclude from central concern those realms of
phenomena and bits of data which do not seem to fit into formal
structures. Such structures are abstracted from variation, the
occurrence of which, though not denied, has been submerged
under the dominant presumption of regularity and homogeneity
throughout a speech community. As a background against which
to set off the structure of the formal code, speech activity com-
monly has been considered random, perhaps a matter of individual
and unpredictable choice. As for the functions of speech, these
have been seen as universally equivalent, 21 while competition be-
tween sets of speech habits, languages, or parts thereof, has fre-
quently been taken to be a purely social matter, not a matter in-
volving the adaptive merits of the habits involved, and often inter-
preted under the blanket term of an unanalyzed differential "pres-
tige." Thus important questions about linguistic change have gone
unanswered because the focus is not upon the actual locus speech
change, and the relation of language to culture becomes a theo-
retical problem.
Now there are exceptions to each part of the picture I have
sketched. It is not a question here of a monolithic ideology. It
is a question of emphases, a dominant outlook and direction, and
the ter-ins in which matters tend to be couched. Moreover, one
must understand that these arose in answer to definite needs,
and that with them great advances in knowledge of language have
been made.
I see the dominance of the view just outlined as arising out of
a battle that had to be won early in this century for the autonomy
and legitimacy of formal linguistic structure as an object of study
in its own right, as distinct from historical and psychological prob-
lems and explanations. And this carried with it, especially in
anthropology, the impliciition,- of equality of all languages for such
study. This autonomy of structural linguistics is a theme in two
classics, the Cours de linguistique generale of Ferdinand de
Saussure22 and the Language of Edward Sapir.23 The need for
this focus carried with it the de-emphasis upon cultural

2
 Language in Education 13

entanglements that we have noted, and, especially in anthropology,

an emphatic non-evolutionary view. There still linger misconcep-
tions about the existence of so-called "primitive" languages, whose
meager vocabularies must be eked out by. gesture, which lack
grammars, definite systems of sounds, and abstract terms, and
which are more variable and change more rapidly because of being
unwritten. All this was demonstrably untrue, and stood in the
way of a general science of linguistics, whose material must be
the rich diversity actually at hand in the world's languages.
"Equality, diversity, relativity" became a linguistic theme.
The rejected notions about "prirative" languages (along with
equally mistaken notions about the superiority of an Indo-European
type, or of one of its exemplars, Latin, and the use of such as
ideal models for description) were of course evolutionary in one
sense. But now the notion of evolution was rejected in toto.
Whatever differences might obtain between simpler and more ad-
vanced cultures, no correlative difference was found to hold be-
tween the structures of their languages. "When it comes to lin-
guistic form, Plato walks with the Macedonian swineherd, Con-
fucius with the head-hunting savage of Assam." 2 No No measures
which would satisfactorily rate language structures as more or less
advanced appeared that were free from cultural bias. One evolu-
tionary typology would have put Chinese at the bottom of the
scale--a patent absurdity. Also, efforts by distinguished scholars,
such as Jespersen, to show trends toward progress in efficiency,
suffered from limitation in data to two language families, Indo-
European and Semitic, and from inadequacies of conceptual analy-
sis. In view of these inadequacies and errors, it is not surprising
that the evolutionary notions of the day dropped out. And,
despite present-day attacks on that generation of cultural anthro-
pologists for being anti-evolutionary, we must remember that evolu-
tionary theory in biology then was rather disunited and uncertain,
not at all the vital force and stimulus, it is today, and that to be
against certain evolutionary stereotypes was to adopt a democratic
and progressive stance.
Now that the battle against the mistaken evolutionary ideas has
been won, and the study of formal structure well established, it is
time to take up the evolutionary question again. And we can see
that the fight against notions of "primitive" languages, whose
echoes still reverberate today, confused three levels of evolution-
ary advance, and so jumped too quickly to the conclusion that all
languages are evolutionarily. equal. There is the level of "primi-
tive" languages, proto- forms below the status of full languages,
then the status of full languages, and finally, the advanced status
we have indicated as occupied by world languages and some others.
The fight against misconceptions about "primitive" languages did
not distinguish the two latter stages, so that to deny the equality
of all languages was taken to imply that the less advanced were
"primitive." No known languages are. All known languages have
achieved the middle status. All languages have achieved the level
14 Functions of Speech
of basic or primary efficiency, such that they can fully adapt, in
time, to the needs of any population. In this sense all languages
are potentially equal, as we observed above, and hence capable of
adaptation to the needs of a complex industrial civilization. This
is just what has happened historically in the case of English,
which in its Old English period would certainly not have been ade-
quate to modern technology and science. But not all languages
are equally efficient compared with one another, either in terms
of specific evolution in meeting particular needs, or in terms of
general evolution in meeting the needs of modern complex civiliza-
tion. Already for many local languages, and ultimately for all.
the direction of change in the world is one which is making modern
complex civilization part of their relevant environment, within
whose context they must compete. The ideal image of a single
"neutral" language in a single, homogeneous cultural context hardly
holds any longer. For populations such as the Otomi the relevant
environment is one in which sets of habits of differing origins com-
pete as means of developing the new forms and content of speech
activity--among which education may be included--that successful
adaptation requires.
In all this, equality in primary or basic efficiency is not
enough, nor is the difference purely a matter of social, non-
linguistic factors. In the particular time and place of competition,
one set of speech habits is as such functionally superior. Partly
the superiority is mutual and relative, specific to particular
niches, e.g., that of Spanish routines for the market place, and
of Otomi for the usual subsistence activities of tribal life. But
partly it is a matter of general superiority, and here the perspec-
tive of general evolution provides a sober and realistic attitude.
Otomi, like Anglo-Saxon and many languages around the world to-
day, could become a medium of technology, science, and philoso-
phy. Just so, any normal infant, wherever born, could partici-
pate in any culture, however complex. But the human infant need
only be raised in the cultural environment for the potentiality to
be realized in one lifetime, while the realization of the potential
of languages often takes much longer. Even granted the will, the
cost in money, personnel, and time may be prohibitive for a poor
nation or a newly struggling one. This is a poignant fact, for
the decisions that must be made in -view of it are often hard.
There is a widespread respect for cultural autonomy and
integrity. And, as the UNESCO report, The Use of Vernacular
Languages in Education, 25 shows a child learns to read most
efficiently if taugl-in its native language first, even when it is
then to learn to read in a second language. Yet what if, once
literate, the child finds-that-there is nothing to read in this first
language, because nothing has been written and the country can-
not afford to duplicate the needed educational materials many times
over in different languages? (The absence of the needed written
materials is a problem even in such places as Egypt and Puerto
Rico, where certain aspects of advanced education have had to be
 Language in Education 15

conducted in English.) And there are many people of great talent

whose efforts to develop a literature in a local language, as for
example in Ghana, must, for similar reasons, come to naught.
The languages could so have developed, but they have not done
so in their existing adaptation, and now it is too late.
So the selective pressure among the languages of the world
continues in environments rapidly changing with technological and
social revolutions. 26 From a scientific and humanistic point of
view, it is a hard loss to see much of this diversity disappear or
become constricted in local uses, not because of inadequacy in its
own terms, but because the terms have changed, and the chance
for development through creative nationalism is lost to all but a
handful. The scientific value of a language is independent of its
political importance, just as the scientific import of a plant or ani-
mal does not depend upon its utility as food, and accordingly,
some linguists are devoting their energies to recording and analyz-
ing the languages about to disappear, so that future theories
language can have the broadest possible base, and so that we about may
come as close as possible to enjoying the full light that language
can shed on the range of human nature and creativity. With
plants and animals, discovery of a new process or type or rare
form is worth more to science than millions of additional pigs and
potatoes, and so it is with languages.
The reduction of linguistic diversity is a loss for humanistic
educational values too, and perhaps a matter of concern for our
own future adaptation. Any one form of language of necessity
selects a small portion of the total range of ways of categorizing
and analyzing experience that language can embody. To a large
extent, the growth of science transcends the framework of any one
language, but insofar as the particulars of our first language
shape- our later thought and use of language, the existence of di-
verse languages is of value as a means of transcending the per
spective of any one, valuable perhaps
voir of potential change. And for sucheven to mankind as a reser-
transcendence, records of
past languages are never so generally effective as living examples.
Let us hope, then, that the attrition of the world's languages will
leave us not entirely impoverished, but still with some store of
diversity.
CONCLUSION
To sum up: I believe that an evolutionary approach to speech
can be unifying and vivifying. In linguistics itself it can, by its
generality and functional perspective, integrate many separate
concerns--genetic classification, linguistic areas, dialectology, bi-
lingualism, standard language studies, linguistic acculturation,
and the like -that deal with language change. In anthropology it
can remove the embarrassing contradiction between an evolutionary
view of culture and a non-evolutionary view of culture's part, lan-
guage, and point toward integration of linguistic and other
16 Functions of Speech
anthropological studies. In education it can, for instance, pro-
vide perspective on questions of correctness in speech. But
chiefly, for the problems of education in large parts of the world,
it can contribute to linguistics and anthropology by focusing on
speech habits in relation to populations; by emphasizing a process
of change through variation, adaptation, and selection; and by
providing a framework and incentive for a descriptive science of
the functions of speech. From this, I hope, will come the com-
parative perspective which anthropology should provide on the
ways in which speech activity enters into the process of educa-
tion.
NOTES
Ethel Emilia Wallis, "Sociolinguistics in Relation to Mez-
1.
quital [Otomi] Transition Education," Estudios Antropologicos
publicado en homenaje al doctor Manuel Gamio (Mexico, D.F.:
Direccion General de Publicaciones, 1956), p. 524.
2. "Repertoire" can refer to the array of resources, words,
phrases, constructions, etc. which are generally available for use
in routines and from which are formed the alternatives available
at particular points in routines.
3. Frequently, different routines are peculiar to different
languages, as in the use of an esoteric foreign language for cere-
monial activity, and of a foreign language, jargon, kione, or the
like for diplomacy, trade, or other communication with members of
other groups. The Otomi use a "market Spanish" in one situation,
but their own language in basic subsistence activities. An essen-
tial task of bilingual education among the Otomi is the introduction
of additional linguistic routines in Otomi and Spanish.
4. James Sledd, "Review of R. Kingdon, The Groundwork of
English Intonation," in Language, Vol. 36, No. 1 (1960), p. 178.
5. Margaret Read, Children of Their Fathers: Growing Up
Among the Ngoni of Nyasaland (New Haven: Yale University
Press, 1960).
6. Edward Sapir, "Language," Encyclopedia of the Social
Sciences (New York: Macmillan, 1933), Vol. 9, pp. 155-69; re-
printed in David G. Mandelbaum (ed.), Selected Writings of Ed-
ward Sapir (Berkeley and Los Angeles: University of California
Press, 1949), p. 16.
7. See Paul Garvin and Madeleine Mathiot, "The Urbanization
of the Guarani Language," in A. F. C. Wallace (ed.), Men and
Cultures: Selected Papers of the Fifth International Congress of
Anthropological and Ethnological Sciences (Philadelphia: Uni-
versity of Pennsylvania Press, 1960), pp. 783-90.
8. W. D. Hohenthal and Thomas McCorkle, "The Problem of
Aboriginal Persistence," Southwestern Journal of Anthropology,
Vol. 11, No. 3 (1956), pp. 288 -300..
9. Edward P. Dozier, "Resistance to Acculturation and
Assimilation in an Indian Pueblo," American Anthropologist,
 Language in Education 17

Vol. 53, No. 1 (1951), pp. 56-66.

10. John Gulick, "Language and Passive Resistance Among
the Eastern Cherokees," Ethnohistory, Vol. 5, No. 1 (1958), pp.
60-81.
11. For the full account, see Sarah C. Gudschinsky, "To-
neme Representation in Mazatec Orthography," Word, Vol. 15,
No. 3 (1959), pp. 446-52.
12. For excellent analysis and examples, see Hans Wolff,
"Phonemic Structure and the Teaching of Pronunciation," Lan-
guage Learning, Vol. 6, No. 3-4 (1956), pp. 17-23.
13. Uriel Weinreich, Languages in Contact (New York: Lin-
guistic Circle of New York, 1953).
14. George M. Foster, "Applied Anthropology and Modern
Life," in Estudios Antropologicos publicados en homenaje al doctor
Manuel Gamio (Mexico, D.F.: Direction General de Publicaciones,
1956), pp. 332-33.
15. See especially Marshall Sahlins.and Elman Service (eds.),
Evolution and Culture (Ann Arbor: University of Michigan Press,
1960); Evolution and Anthropology: A Centennial Appraisal (Wash-
ington, D.C.: Anthropological Society of Washington, 1959);
Julian S. Huxley, "Evolution, Cultural and Biological," in William L.
Thomas, Jr. (ed.), Current Anthropology (Chicago: Universify of
Chicago Press, 1956), pp. 3-25.
16. See Huxley, op. cit., and George G. Simpson, The Mean-
ing of Evolution (New Haven: Yale University Press, 1952),
chap. XV.
17. Several lines of fresh interest in speech variation are now
emerging, and we
the next decade. can expect a number of significant studies in
The fresh concern manifests itself in various
guises and under various names, but it is especially associated
with the revitalization of dialectology. An example is the work of
the Linguistic Survey of Scotland; a theoretical approach akin to
that advocated here has been sketched by Trevor Hill, a member
of the Survey, in his article, "Institutional Linguistics," Orbis,
Vol. 8, No. 2 (1958), pp. 441-55.
18. It is because Miss Wallis' article is so valuable and stimu-
lating, being a discussion which presents both concrete detail
and theoretical assumptions, that it is the focus here for critical
analysis of some of these assumptions in terms of a different theo-
retical perspect've. Miss Wallis relies extensively upon the au-
thority of Edwa-d Sapir for her perspective, and, as the discus-
sion immediately following indicates, Sapir fer her perspective,
and, as the discussion immediately following indicates, Sapir was
a principal exponent of the non-evolutionary view whIch it is now
necessary to transcend.
19. Joseph Greenberg, "Language and Evolutionary Theory,"
Essays in Linguistics (Chicago: University of Chicago Press,
1956), pp. 56-65.
20. Evolutionary advance in grammatical features is suggested
by A. Mei llet, Introduction a l'etude comparative des langues
18 Functions of Speech
indo-europeenes (Paris: Libraire Hachette, 8th ed., 1937), pp.
424-5; Marcel Cohen, Le Langage, Structure et Evolution (Paris:
Editions Sociales, 1950), p. 112; and Henri Frei, "Systemes de
deictiques," Acta Linguistica, Vol. 4, No. 3 (1944), pp. 111-129.
Meillet and Cohen cite particularly the association in various Indo-
European languages between loss of the category of dual and de-
velopment of more complex civilization. Frei writes that "Linguists
agree in attributing the disappearance of the dual to the march of
civilization," citing A. Cuny, Le nombre duel en Grec (These de
Paris, 1906) and J. Wackernagel, Vorlesungen Ether Syntax, Vol. 1,
2nd ed. (Bale, 1926), pp. 74-5, and going on to interpret as a
parallel the tendency of deictic systems to evolve toward a binary
type (such as "here":"there").
The best recent attempt to analyse the development of economy
and efficiency in languages is probably that of W. Koenraads,
Studien fiber sprachokon.omische Entwicklungen im Deutschen
(Amsterdam: Muelenhoff, 1953). There are valuable comments in
the reviews of the book by H. Hoenigswald (Language, Vol. 30,
pp. 591-3 (1954), and Uriel Weinreich (Word, Vol. 11, pp. 237-40
(1955)).
21. For example, "A necessary condition for socialization in
man is the learning and use of a language. But different lan-
guages are functionally equivalent in this respect, and one lan-
guage is comparable with another because human speech has cer-
tain common denominators" (A. 1. Hallowell, "Culture, Personality,
and Society," in A. L. Kroeber and others, Anthropology Today
[Chicago: University of Chicago Press, 1953], p. 612). Such
statements point out the important common denominator of the
functioning of language in education and socialization, but ignore
the important differences.
22. F. de Saussure, Cours de linguistique generale, pub-
lished by C. Bally and E. Sechehaye (Paris: Payot, 1916), and
now in English translation, Course in General Linguistics (New
York: Philosophical Library, 1959).
23. Edward Sapir, Language (New York: Harcourt, Brace,
1921).
24. Edward Sapir, op. cit., p. 234.
25. The Use of Vernacular Languages in Education (Paris:
UNESCO, Monographs on Fundamental Education: VIII, 1953).
See the extended review by William E. Bull in the International
Journal of American Linguistics, Vol. 21, No. 3 (1955), pp. 288-
94.
26. Some scholars are directing attention to the linguistic
aspects of this subject, notably (in the United States) Charles A.
Ferguson, Paul Garvin, John J. Gumperz, and Uriel Weinreich.
See Charles A. Ferguson and John J. Gumperz (eds.), Linguistic
Diversity in South Asia: Studies in Regional, Social, and Func-
tional Variation (Bloomington, Indiana: Indiana University Re-
search Center in Anthropology, Folklore and Linguistics, Publi-
cation #13; Part III, International Journal of American Linguistics,
Vol. 26, No. 3, July 1960).
 speech and language: on the
origins and foundations of
inequality among speakers
I conceive of two sorts of inequality in the human species;
one, which I call natural or physical, because it is estab-
lished by nature and consists in the difference of ages,
health, bodily strengths, and qualities of mind or soul;
the other, which may be called moral or political inequal-
ity, because it depends upon a sort of convention and is
established, or at least authorized, by the consent of men.
The tatter consists in the different privileges that some
men enjoy to the prejudice of others, such as to be richer,
more honored, more powerful than they," or even to make
themselves obeyed by them.
Rousseau (1775) 3-
I use the second paragraph f_+' 'Rousseau's second Discourse as an
epigraph, and adapt its title, because I want to call attention to a
link between his concerns and ours. Like him, we think knowledge
of human nature essential and pursue it; like him, we think the
present condition of mankind unjust, and seek to transform it.
These two concerns, for example, provide the frame for Noam
Chomsky's recent Russell lectures. 2 Unlike Chomsky, but like
Rousseau, moreover, some linguists are beginning to attend to a
conception of linguistic structure as interdependent with social
circumstances, and as subject to human needs and evolutionary
adaptation. And like Rousseau, our image of the linguistic world,
the standard by which we judge the present situation, harks back
to an, earlier stage of human society. Here Rousseau has the ad-
vantage of us. He knew he did this, and specified the limitations
of it (see the end of note h to the Discourse). We do it implicitly,
fidling back on a "Herderian" conception of the world as composed
of individual language-and-culture units, for lack of another way
of seeing the resources of language as an aspect of human groups,
because we have not thought through new ways of seeing how lin-
guistic resources do, in fact, come organized in the world. Thus
we have no accepted way of joining our understanding of inequal-
ity with our understanding of the nature of language.
Chomsky's Russell lectures are a case in point. The first lec-
ture, "On Interpreting the World," presents implications of a cer-
tain conception of the nature of language and of the goals of lin-
guistic research, leading to a humanistic, libertarian conception of
man. The second lecture, "On Changing the World," is about in-
justice, its roots in inequality of power, and the failure of scholars
and governments to deal with the true issues in these respects.
19
20 Speech and Language
There is little or no linguistics in the second lecture, just as
there is little or nothing of social reality in the first. Such
principled schizophrenia besets linguistics today; the scientific
and social goals of its practitioners are commonly compartmental-
ized. Such an alienation from experience and social reality of one
of "the many kinds of segmental scientists of man," against which
Edward Sapir warned years agt),3 does not mirror either the true
nature of language or its relation to social life; rather, it reflects
a certain ideological conception of that nature and that relation,
one which diverts and divorces linguistics from the contribution,
desperately needed, that it might make to the understanding of
language as a human problem.
The heart of the matter is this. A dominant conception of the
goals of "linguistic theory"" encourages one to think of language
exclusively in terms of the vast potentiality of formal grammar,
and to think of that potentiality exclusively in terms of its uni-
versality. But a perspective which treats language only as an
attribute of man leaves language as an attribute of men unintelligi-
ble. In actuality language is in large part what users have made
of it. Navajo is what it is in part because it is a human language,
and in part because it is the language of the Navajo. The gen-
eric potentiality of the human faculty for language is realized dif-
ferently, as to direction and as to degree, in different human com-
munities, and is useless except insofar as it is so realized. The
thrust of Chomskyan linguistics has been to depreciate the actual-
ity of language under the guise of rejecting an outmoded philoso-
phy of science, but we must be able to see beyond its ideological
use and recognize that one cannot change a world if one's theory
permits no purchase on it. Thus, one of the problems to be over-
come with regard to language is the linguist's usual conception of
it. A broader, differently based notion of the form in which we
encounter and use language in the world, a notion which I shall
call ways of speaking, is needed.
Let me subsume further consideration of how it is that linguis-
tics is part of the problem, under the following consideration of
some of the other dimensions of language and of some general
sources of inequality with regard to it. In both sections I shall
try to indicate the need for a conception of ways of speaking.
Some Dimensions of Language as a Human Problem
It is striking that we have no general perspective on language as
a human problem, not even an integrated body of works in search
of one. Salient problems, such as translation, multilingualism,
literacy, and language development, have long attracted attention,
but mostly as practical matters constituting "applications" of lin-
guistics, rather than as proper, theoretically pertinent parts of it.
There are notable exceptions, as in the work of Einar Haugen, but
for about a generation most linguistic thought in the United States
has seen in the role of language in human life only something to
 Language in Education 21

praise, not something to question and study. Perhaps this situa-

tion reflects a phase in the alternation of "high" and "low" evalu-
ations of language to which the philosopher Urban called atten-
tion.5 The skeptical period after the First World War did see
leading American theorists of language devote themselves to lan-
guage problems, such as those involving new vehicles for inter-
national communication (Jespersen, Sapir), the teaching of reading
(Bloomfield), literacy (Swadesh), language as an instrument and
hence a shaper of thought (Sapir, Whorf), and linguistic aspects
of psychiatric and other interpersonal communication (Trager,
Hockett, in the early 1950's) . Perhaps this issue of Dcedalus is
a sign that the climate of opinion is shifting once again toward a
balanced recognition of language as "at one and the same time
helping and retarding us," as Sapir put it in one context. 6
In any case, it is unusual today to think of language as some-
thing to overcome, yet four broad dimensions of language can use-
fully be considered in just that way: diversity of language,
medium of language (spoken, written), structure of language,
and functioning of language. Of each we can ask,
(1) when, where, and how it came to be seen as a problem;
(2) from what vantage point it is seen as a problem (in rela-
tion to other vantage points from which it may not be so seen);
(3) in what ways the problem has been approached or over-
come as a practical task and also as an intellectual, conceptual
task;
(4) what its consequences for the study of language itself
have been;
(5) what kinds of study, to 'which linguists might contribute,
are now needed.
I cannot do more than raise such questions here; limitations of
knowledge would prevent my doing more, if limitations of space
did not. To raise such questions may, I hope, help to stimulate
the development of a general perspective.
Overcoming Diversity of Language. This problem may be the
most familiar, and the historical solutions to it form an important
part of the subject matter of linguistics itself: lingua francas,
koines, pidgins and creoles, standardized languages, diffusion
and areal convergence, multilingual repertoires, and constructed
auxiliary languages. The myths and lexicons of many cultures
show a widespread and presumably ancient recognition of the di-
versity of language, although not uniformly in the mold of the
Tower of Babel. The Busama of New Guinea and the Quileute of
the present state of Washington believed that originally each per-
son had a separate language, and that community of language was
a subsequent development created by a culture hero or trans-
former. Thus it is an interesting question whether it is unity or
diversity, within or between speech communities, that has seemed
the thing requiring an intellectual explanation.
In Western civilization the dominant intellectual response to the
existence of diversity has been to seek an original unity, either of
22 Speech and Language

historical or of psychological origin (sometimes of both). The

dominant practical response has been to impose a novel unity in
the form of the hegemony of one language or standard. The
presence of the Tower of Babel story in the civilization's sacred
book legitimated, and perhaps stimulated, efforts to relate lan-
guages in terms of an original unity and played a great part in
the cumulative development of linguistic research. Indeed, some
rather sophisticated work and criticism on this subject can be
found from the Renaissance onward, and the dating of the origin
of linguistic science with the comparative-historical work of the
early nineteenth century reflects its institutionalization as much
as or more than its intellectual originality. 7 The force of Chris-
tian and humanitarian concern to establish the monogenesis of
man through the monogenesis of language was felt strongly well
through the nineteenth century, from the dominance of the "ethno-
logical question" in the first part through the controversies involv-
ing Max Muller, Darwin, Broca, and others. 8 The special interest
of Europeans in Indo-European origins became increasingly im-
portant in the latter part of the century, the idea of a common
linguistic origin stimulating and legitimating studies of common
cultural origins and developments. Humanitarian motives played a
part as wellMatthew Arnold appealing to Indo-European brother-
hood as a reason for the English to respect Celtic (Irish) culture
and perhaps the Irish, and Sir Henry Maine making a similar ap-
peal on behalf of the peoples of India. Sheer intellectual curiosity
and satisfaction must always be assigned a large part in motivat-
ing work in comparative-historical linguistics, and humanistic con-
cern has probably played a part in the major contemporary effort
to establish empirically a common historical origin for languages,
that of the late Morris Swadesh. 9
The most salient effort to establish a conceptual unity of
human languages today is, of course, linked with the views of
Noam Chomsky. Concern for such a unity is itself old and con-
tinuous--the appearance of disinterest among part of a genera-
tion of U.S. linguists before and after the Second World War was
a local aberration whose importance is primarily due to Chomsky's
reaction against it. He has reached back to the seventeenth and
eighteenth centuries for an ancestral tradition, 1° when he had
only to take up the tradition in this country of Boas and Sapir,
or the European tradition, partially transplanted to this country,
of Trubetzkoy and Jakobson. In both of these traditions some
significant things were being said about the universals of lan-
guage in the 1930's and early 1940's. It is true, however, that
the history of the tradition of general linguistics stretching back
through the nineteenth century (and, Jakobson would argue, con-
tinuing straight back through the Enlightenment to origins in
medieval speculative grammar), had been lost from sight in Ameri-
can linguistics, and a sense of it is only now being recovered.
It is true, too, that since Herder and von Humboldt, the tradi-
tion does not much appeal to Chomsky, since its universalism is
Language in Education 23

combined with an intense interest in typology, that is, in the

characterization of specific languages as well as, and as an in-
strument of, the characterization of language.
Here we touch on the inescapable limitation of either kind of
effort to conceive the unity of human language. Although one
used to speak of the discovery of a genetic relationship as "re-
ducing" the number of linguistic groups, both the language and
the thought were badly misleading. Languages may disappear
through the destruction of their speakers, but not through the
publication of linguistic papers and maps. The newly related lan-
guages remain to be accounted for in their differences and de-
velopments as well as in terms of the portion (often quite small)
of their makeup that shows their common origin. Likewise, the
discovery of putative universals in linguistic structure does not
erase the differences. Indeed, the more one emphasizes uni-
versals, in association with a self-developing, powerful faculty of
language within persons themselves, the more mysterious actual
languages become. Why are there more than one, or two, or
three? If the internal faculty of language is so constraining, must
not social, historical, adaptive forces have been even more con-
straining, to produce the specific plenitude of languages actually
found? For Chinookan is not Sahaptin is not Klamath is not
Takelma is not Coos is not Siuslaw is not Tsimshian is not Wintu
is not Maidu is not Miwok is not Yokuts is not Costanoan...(is
not Tonkawa, is not Zuni, is not Mixe, is not Zoque, is not any
of the numerous Mayan languages, or affiliates of Mayan, if one
extends the horizon). The many differences do not disappear,
and the likenesses, indeed, are far from all Chomskyan univer-
sals; some likenesses exist because of a genetic common origin
(Penutian), some because of areal adaptations (Northwest Coast
for some, California for others), some because of diffusion, some
because of limited possibilities and implications at la Greenberg).
Franz Boas once argued against exclusive concentration on genetic
classification, calling the full historical development of languages
the true problem." A similar point can be made today as against
concentration on putative universals. Most of language begins
where abstract universals leave off. In the tradition from Herder
and von Humboldt through Boas and Sapir, languages are "con-
crete universals," and most of language as a human problem is
bound up with the adjective of that term.
Both of these modes of overcoming diversity of language
intellectually, genetic classification and the search for putative
universals, locate their solutions in time. There is a past refer-
ence, a historical origin of languages or an evolutionary origin of
the faculty of language, and there is a present and future refer-
ence, one which draws the moral of the unity that is found.
Neither speaks to the present and future in terms of the processes
actually shaping the place of language in human life, for the
faculty of language presumably remains constant and genetic
diversification of languages is literally a .thing of the past. The
24 Speech and Language
major process of the present and foreseeable future is the adapta-
tion of languages and varieties to one another and their integra-
tion into special roles and complex speech communities. The .nder-
standing of this process is the true problem that diversity o lan-
guage poses, both to mankind and to those who study mankind's
languages,
The essence of the problem appears as communication, intelli-
gibility. Some are concerned with the problem at the level of the
world as a whole, and efforts to choose or shape a common lan-
guage for the world continue.12 Some project this contemporary
concern onto the past, speaking of a "stubborn mystery" in the
"profoundly startling, 'anti-economic' multiplicity of languages
spoken on this crowded planet."13 Such a view is anachronistic,
however, for the diversity was not "anti-economic" when it came
into being; it was just as much a "naturally selected, maximalized
efficiency of adjustment to local need and ecology" as the great
variety of fauna and flora to which Steiner refers in the phrase
just quoted. Universal processes of change inherent in language,
its transmission and use, together with separation and separate
adaptation of communities over the course of many centuries, suf-
fice to explain the diversity. Simply the accumulation of unshared
changes would in time make the languages of separate groups mutu-
ally unintelligible. There is .of course more to it than physical and
temporal distance (as Steiner insightfully suggests); there is social
distance as well. Boundaries are deliberately created and main-
tained, as well as given by default. Some aspects of the struc-
tures of languages are likely due to this. If the surface form of
a means of communication is simplified greatly when there is need
to overcome barriers, as it is in the formation of pidgin languages,
then the surface form of means of communication may be compli-
cated when there is a desire to raise or maintain barriers. 14 This
latter process may have something to do with the fact that the sur-
face structf,ife-s of languages spoken in small, cheek-by-jowl com-
munities so often are markedly complex, and the surface structures
Of languages spoken over wide ranges less so. (The observation
would seem to apply at least to North American Indian languages
and Oceania) .
In any case, the problem is one of more than languages; it is
one of speech communities. Here the inadequacy of dominant con-
cepts and methods in linguistics is most painfully apparent. The
great triumph of linguistic science in the nineteenth century, the
comparative-historical method, deals with speech communities as
the source and result of genetic diversification. The great triumph
of linguistic science in the twentieth century, structural method,
deals with speech communities as equivalent to language.15 Genetic
diversification can hardly be said to occur any longer, and a speech
community comprising a single language hardly exists. The study
of complex speech communities must benefit mightily from the tools
and results both of historical linguistics, for the unraveling and
Language in Education 25

interpretation of change, and of structural linguistics, for the ex-

plicit analysis of linguistic form. But it cannot simply apply them,
it must extend them and develop new tools.
The needs can be expressed in terms of what is between
speech communities and what is within them. Despite .their well-
known differences as to psycholog3r, both Bloomfield and Chomsky
reduce the concept of speech community to that of a language.16
This will not do. The boundaries between speech communities are
thought of first of all as boundaries of communication, but com-
munication, or mutual intelligibility as it is often phrased, is not
solely a function of a certain objective degree of difference be-
tween two languages or some series of related languages. One
and the same degree of "objective" linguistic differentiation may
be taken to demarcate boundaries in one case, and may be de-
preciated in another, depending on the social and political cir-
cumstances.17 And intelligibility itself is not only a complex
function of features of linguistic form (phonological, lexical, syn-
tactic), but also of norms of interaction and conduct in conver-
sation, and of attitudes towards differences in all these respects.
In Nigeria one linguist found that as soon as members of a cer-
tain community recognized a related hinterland dialect, they re-
fused to understand it; 18 other communities are noted for the
effort they make to understand despite great difference. Such
considerations cut across language boundaries. One may be at a
loss to understand fellow speakers of his own language if his
assumptions as to appropriate topics, what follows what, and the
functions of speech are different (as happens often enough in
classrooms between teachers of one background and students of
another), and many of us have had the experience of following a
discussion in a language of which we have little grasp, when the
topics, technical terminology, and norms of conduct are profes-
sionally shared.
To repeat, communication cannot be equated with a "common"
language. A term such as "the English language" comprises all
linguistic varieties that owe their basic resources to the historical
tradition known as English. That "language" is no longer an ex-
clusive possession of the English, or even of the English and the
Americans--there are perhaps more users of English in the Third
World, and they have their own rights to its resources and future.
Many varieties of "English" are not mutually intelligible within
Great Britain and the United States as well as elsewhere. In
fact, it is an important clarification if we can agree to restrict
the term "language" (and the term "dialect") to just this sort of
meaning: identification of a historically derived set of resources
whose social functioning-- organization into used varieties, mutual
intelligibility, etc. -is not given by the fact of historical deriva-
tion itself, but is problematic, needing to be determined, and
calling for other concepts and terms.
We are in poorly explored territory here. Even with consider-
ation restricted to groups which can communicate, there is a gamut

4
26 Speech and Language
from "I can make myself understood" at one end to "he talks the
same language" at the other. Probably it is best to employ terms
such as "field" and "network" for the larger spheres within which
a person operates communicatively, and to reserve the term "com-
munity" for more integral units. Clearly the boundary (and the
internal organization) of a speech community is not a question
solely of degree of interaction among persons (as Bloomfield said,
and others have continued to say), but a question equally of
membership, of identity and identification. If interaction were
enough, school children would speak the TV and teacher English
they constantly hear. Some indeed can so speak, but do not
necessarily choose to do so. A few years ago I was asked by
teachers at Columbia Point why the children in the school did not
show the influence of TV, or, more pointedly, of daily exposure '
to the talk of the teachers. A mother present made a telling ob-
servation: she had indeed heard children talk that way, but on
the-playground, playing school; when playing school stopped,
that way of talking stopped too.
Community, in this sense, is a dynamic, complex, and some-
times subtle thing. There are latent or obsolescent speech com-
munities on some Indian reservations in this country, brought
into being now principally by the visit of a linguist or anthro-
pologist who also can, use the language and shows respect for the
uses to which it can be put. There are emergent communities,
such as New York City would appear to be, in the sense that
they share norms for the evaluation of certain variables (such as
post-vocalic r), that have developed in this century. There are
other communities whos-e stigmata are variable and signs of severe
insecurity, like those of New York, or the community of portenos
in Buenos Aires, comprised principally of immigrants concerned to
maintain their distance and prestige vis-t-vis speakers from the
provinces (who, ironically enough, have lived in the country much
longer). There can be multiple membership, and there is much
scope rot. false perception; authorities, both governmental and
.rational, are often ignorant of the existence of varieties of
language and communication under their noses. An unsuspected
variety of creolized English was discovered recently on an island
off Australia by the chance of a tape recorder being left on in a
room where two children were playing. When the linguist heard
the tape and could not understand it, he came to realize what it
was. That such a language was known by the children was en-
tirely unknown to the school. Indians who have been beaten as
children for using their Indian tongue or blacks who have been
shamed for using "deep" Creole will not necessarily trot the lan-
guage out for an idle inquirer. In general, when we recognize
that this diversity of speech communities involves social as well
as linguistic realities, we must face the fact that there are differ-
ent vantage points from which diversity may be viewed. One
person's obstacle may be someone else's source of identity. In
the United States and Canada today one can find Indians seeking
Language in Education 27

to learn the Indian language they did not acquire as children.

Leveling of language seems neither inevitable nor desirable in the
world today. It is common to mock efforts at preservation and
revitalization of languages as outmoded romanticism, but the
mockery may express a view of human nature and human needs
whose shallowness bodes ill for us.
What is within a speech community in linguistic terms has
begun to be understood better through recent work in socio-
linguistics. Empirical and theoretical work has begun to provide
a way of seeing the subject "steadily as a whole." It suggests
that one think of a community (or any group, or person) in
terms, not of a single language, but of a repertoire. A reper-
toire comprises a set of ways of speaking. Ways of speaking,
in turn, comprise speech styles, on the one hand, and contexts
of discourse, on the other, together with the relations of appro-
priateness obtaining between styles and contexts. Membership in
a speech community consists in sharing one or more of its ways
of speaking--that is, not in knowledge of a speech style (or any
other purely linguistic entity, such as a language) alone, but in
terms of knowledge of appropriate use as well. There are rules
of use without which rules of syntax are useless. Moreover, the
linguistic." features that enter into speech styles are not only the
"referentially-based" features usually dealt with in linguistics to-
day, but also the "stylistic" features that are complementary to
them, and inseparable from them in communication. Just as social
meaning is an integral part of the definition and demarcation of
speech communities, so it is an integral part of the organization
of linguistic features within them. (Cf. Bernstein's concept of
"restricted" and "elaborated" code, classical diglossia, liturgy.)
The sphere adequate to the description of speech communities, of
linguistic diversity as a human problem, can be said to be: means
of speech, and their meanings to those who use them."
No one has ever denied the facts of multilingualism and hetero-
geneity of speech community in the world, but little has been done
to enable us to comprehend and deal with them. Until-now a
"Herderian" conception of a world of independent one-language-
one-culture units, a conception appropriate enough, perhaps, to
a world pristinely peopled by hunters and gatherers and small-
scale horticulturalists, has been tacitly fallen back upon. There
now begins to be work to characterize complex linguistic communi-
ties and to describe speech communities adequately. Such de-
scription must extend to the place of speech itself in the life of
a community: whether it is a resource to bz., hoarded or something
freely expended, whether it is essential or not to public roles,
whether it is conceived as intrinsically good or dangerous, what
its proper role in socialization and demonstration of competence is
conceived to be , and so forth. 2° Through such work one can
hope to provide adequate foundations for assessing diversity of
language as both a human problem and a human resource.
"Diversity" could stand as the heading for all of the problems
connected with speech and language,-once our focus is enlarged
28 Speech and Language
from languages as such to speech communitiesexisting diversity
as an obstacle, and sometimes diversity that it is desired to main-
tain or achieve. Nevertheless, it is worthwhile to comment sepa-
rately on three topics that have been singled out for attention in
their own right. These are problems connected with the media,
the structures, and the functions of language.
Overcoming the Medium of Language. Not long ago one might
have said that most of the world was attempting to overcome the
spokeness of language through programs of literacy, while some
of the advanced sectors of civiliza' .on--the advertising and com-
munications industries, and the university--were hailing the im-
minent transcendence of language in graphic form. McLuhan is
less prominent now, but these twin poles of spoken and written
language-remain very much with us. A good deal has been said
about speaking and writing, about oral and literate cultures, 21
and I have no new generalization to add, but I do have a bit of
skepticism to advance. We really know very little as to the role
of the medium of language. Technological determinism is not
generally popular, for good reason, so it is puzzling to find it
avidly welcomed in the sphere of communication. There is no
more reason to regard it as gospel there than elsewhere. Cer-
tainly, it is impossible to generalize validly about "oral" vs.
"literate" cultures as uniform types. Popular social science does
seem to thrive on three-stage evolutionary sequences -David
Riesman, Margaret Mead, Charles Reich have all, like McLuhan,
employed them--but if dogmatic Marxism is not to be allowed such
schemes, again for good reason, it really seem_ s a little unfair to
tolerate it in dogmatic McLuhanism.
In such theses, nevertheless, lies a major threat and fasci-
nation of media. Is use of one medium of communication rather
than another simply transfer of an underlying competence that
remains constant? Or is there more to it than that? Is the com-
municative medium itself partly constitutive of meaning, even of
reality, even perhaps of language itself?22
Undoubtedly the adaptation of communication to an oral-audi-
tory_channelto mouth, throat, air, and ear--has helped shape
human language [e. g. to the range in number of phonological
units in languages (a-medium with different properties might have
facilitated more or permitted less), and as to relation among units
(sequences, kinds of change) conditioned by the physical charac-
teristics of the sounds]. So much has become a matter of con-
siderable interest to linguists. Not so the adaptation of verbal
communication to a manual-visual channel, to hand, things scrip-
turable, and eye. The origin and history of writing systems has
sometimes attracted interest, but as a separate specialty; linguists
do not think of the written channel as shaping, hence partly ex-
plaining, their object of study. 23 Writing is seen as a record of
something already existing. Interest in the history of writing has
to do with the nature of different modes of representation of

40
Language in Education 29

language, their evolution , diffusion, and effect on. what one can
know about languages represented by them. Debates about writ-
ing have to do with adequacy of different representations (past
or present), and, more generally, with the adequacy of any
written norm as basis for linguistic analysis.
Many modern linguists, reacting against the inadequacies of
conventional writing systems, and the role of conventional writing
systems as symbols of cultural domination, have insisted that
written forms are entirely derivative of speech, entirely second-
ary, arbitrary, not, as so often thought in traditional cultures,
intrinsic to what is expressed in them. (Hall (1975), for example,
maintains such a view vigorously.) Many linguists-- associated with
Chomsky's approach have looked more kindly upon conventional
English orthography, reacting against a preceding generation's
reaction against tradition, perhaps, and certainly on general
principle. Whereas the preceding generation emphasized study of
the spoken form of language, Chomsky and others depreciate the
spoken form as a highly imperfect, even "degenerate," manifesta-
tion of structure. The net effect is the same. The issue is the
accuracy of a system of writing in representing something else,
the something else being primary. As a secondary realization of
the structure of language (whether in speech or in the mind),
writing has little or no theoretical interest of its own. Inde-
pendent linguists, such as Dwight Bolinger, H.J. Uldall (1944),
and Josef Vachek (1944-49), have defended the partial autonomy
of writing, as something requiring investigation, but their lead
has been little followed.
The views of American linguists have perhaps been unduly
influenced- by the situation of English. The patent discrepancy
between conventional spelling and actual pronunciation helped one
approach dismiss writing; a reaction has led others to profess to
find a close fit between conventional spelling and a phonological
structure imputed to the language (Chomsky and Halle 1968).
(There is of course an interesting fit between conventional spell-
ing and earlier stages of actual pronunciation (e.g., the "gh" in
"right," "night ," "light" was once pronounced as in German
Recht, Nacht, Licht).) I am sure this view is mistaken, but no
general principle about writing is at stake, only analysis of Eng-
lish. There are languages whose written form is not a spelling
at all (Chinese), and there are languages whose conventional
written form matches spoken phonology quite well (Spanish).
Wherever English belongs between these two poles, its status is
misleading as a basis for thinking about the relation between
speech and writing.
The point is this: the general issue is not the degree to which
one mode is an accurate expression of the other, or of underlying
structure. Such a formulation limits attention to the relation be-
tween speech and writing as one of representation. The general
issue is the relation between speech and writing as modes of
30 Speech and Language
action. It is in their status as modes of action that speech and
writing fundamentally are related to each other in a society. Di-
versity and inequality are not manifested in matters of represen-
tation alone; they are manifest in what it means to speak, to
write (or hear and read), at all, and of course in what it means
to do so in one or another way. In sum, the fundamental relation
between speech and writing is not that of successive, or correla-
tive, levels of linguistic structure. The fundamental relation is
that of choice of means within communicative repertoires. (Clearly
my use of "speech," "speaker," etc. in much of this chapter must
be understood as surrogate for all communicative modes, wherever
speech is not specifically contrasted with others.)
This perspective, that of choice of means embedded in acts,
helps keep in view two considerations essential to study of writ-
ing: graphic means are not neutral, but have social meaning;
graphic means have scope and organization of their own.
As to the first: linguists involved in practical work, such as
literacy, standard language planning, and education have long had
reason to know that to choose what form of language is to be
written, and to choose how it is to be written, are never purely
technical matters. Cultural values and social hierarchies are in-
volved. A notable consequence for the situation of language in
the world is that many languages, and varieties of language, have
not been thought worth writing, or even capable of being written;
their written forms, and what exists in them in writing, has come
largely from outsiders with a religious or scientific mission. The
efforts of outsiders have not always been welcome, and in any
case, the sheer fact of the existence of a written form has not
been sufficient reason for it to be used. A social interest must
be mobilized, as many missionaries have found. (Such facts show
that it is silly to explain writing by appeal to its obvious advan-
tages, as if the advantages were self-realizing; more is said on
the cultural role of writing below.) Or a social interest may have
to be overcome. A technically advantageous form of writing may
be rejected because of the prestige of some alternative or to pro-
tect some interest. (Thus, the Korean hangul had to wait several
centuries and a change of social order to be generally adopted;
the government of Somali had to resolve an orthographic impasse
by fiat (the solution was indeed in the interest of the country as
a whole); Chinese plans for Romanization seem to ha- been
shelved.]
Such questions occur within the United States with regard to
the place of varieties of language other than standard English in
classrooms, and modes of writing English dialects and Indian lan-
guages. A linguist's concern for the efficiency and universality
of a phonetic orthography may encounter a Native American's
preference for something emblematically different from English
symbols, while a linguist taking the standard orthography for
granted in his work may unwittingly reinforce social prejudice.
Language in Education 31

Black parents may react strongly against the suggestion that

their children be taught with materials that represent the speech
of their community, as something distinct from standard English,
while black college students may protest against being penalized
for departures from standard orthographic practice.
Social meaning is not limited to ethnic, regional, or dialect
differences. Joseph Jaquith has pointed out a contrast between
conventional and vernacular spellings, particularly in signs and
advertisements, associated with the durability, cost, prestige of
a product or service. The vernacular spellings employ phonetic
and quasi-phonetic substitutions ("rite," "kwik"), syllabary-like
uses of alphabet letters ("E-Z"), etc. (The vernacular spellings
of words such as "rite," "nite," "12te," incidentally, is evidence
against imputed psychological reality and phonological fit for the
"gh" in the conventional spelling.) Quite within the scope of the
standard language, then, graphic competence in American English
embraces more than one variety of spelling. The relation between
speech and writing has to be discussed in terms of styles of writ-
ing, as well as of speech.
The above examples have dwelt on the representational rela-
tion of writing to speech, but of course scriptoria' competence is
not limited to knowledge of how to represent speech, or struc-
tures strictly common to writing and speech alike. In keeping
with a tendency that might be called "communicative plenitude"
meaningfulness expands to fill available means--the significance of
graphic signs is not restricted to representation of phonic ones or
of an element of structure indicated by both. Nor, of course, is
the significance of a phonic sign restricted to manifestation of a
graphic one, or of an element indicated-by-both. Users notice
not only the respect in which such signs convey the "referentially-
based" relationships of grammar, but also the respects in which
such signs, and some of their "referentially" indifferent details,
are associated with persons, places, purposes, and styles, are
susceptible to play and aesthetic patterning, etc., and elaborate
these possibilities. Such elaboration gives rise to devices and
relationships that are specific to each medium, having no exact
counterpart in the other, but being part of what one can do with
language only when language is being used in the medium in ques-
tion. Within the field of language, of competence in language,
styles of speech and styles of writing become partly autonomous
families of symbolic form. Such growth in the range of means is
one respect in which the resources of languages change in scope
in the course of history. Part of the competence in language of
many people is shaped by, must be partly explained by, the
availability and characteristics of graphic channels.
We are often reminded of how much is missed when linguistic
analysis is based on examples that omit essential features of
speech, such as intonation and voice quality. This is indeed a
crucial obstacle to be overcome by linguistics if it is to deal
32 Speech and Language
adequately with language. One seldom thinks of the converse, of
how much is missed when one neglects features specific to writing
and print. It is as if the field of competence in language had the
shape of a butterfly, one wing specific to speaking, one to writ-
ing, the body common to both. Linguistic analysis has focused on
the body, as it were, and while that is vital, so are the wings.
Put another way: sometimes to speak is to read aloud, some-
times to write is to transcribe. Such cases of strict equivalence
are special cases, interesting just because of that. Unfortunately,
linguistic analysis has proceeded as if such cases were general.
Having emphasized that speech and writing are not isomorphic,
and have autonomy, I do not want to seem to imply that they are
wholly disjunct. It would be a mistake to postulate a universal,
absolute contrast between styles of speaking and styles of writing.
We need instead to broach the more general topic of communicative
styles. The organization of communicative means may follow lines
dictated by modalities, but need not. A style may integrate fea-
tures from different components of structure (so that a style of
graphic English might select and group together features of
orthography, morphology, syntax, diction, discourse); with re-
gard to any one component, may select some and not other fea-
tures (e.g. , one spelling, alternant, construction rather than an-
other); and may have features uniquely its own (as in a special-
ized typography); and may integrate features from more than one
medium. The integration of spoken and gestural signs within a
communicative paradigm should be well known (for an excellent
analysis of a case, see Sherzer 1973; cf. Hymes 1974a:102). Inte-
gration of spoken and graphic signs should not be surprising.
The metalinguistic use of finger-indication of written characters in
the midst of conversation is well enough known from Japan. Ad
hoc hand depiction of letters does occur in interactions in the
United States (quite independently of sign language). The rela-
tion between spoken presentation and styles associated with print
(what might be called "scriptive styles") has undergone great
change in the last generation or so, both in lectures and public
talk generally. There has also been a marked rise of engaged
performance, as distinct from prudent reading, by poets. Identi-
fication of the social meaning of styles, and analysis of their
appropriateness and effect, must deal with such shifting and
mingling. 24
In sum, the point of view from which to grasp the relation
between speech and writing, as media or modalities of language,
is function. How are the features of modalities organized for the
purposes of those who employ them?
Having emphasized that speech and writing are to be seen as
modes of action, I do not want to seem to suggest that they are
everywhere the same modes of action. Their degree of autonomy
from each other, their relative hierarchy, their integration into
communicative styles, all these are problematic, and to be
 Language in Education 33
determined ethnographically. As a general principle, one can
assume that a difference of means may condition differences in
what is accomplished, and that choice of a style specific to
speech, or specific to writing, or mingling both, affects meaning
and outcome. There is little hard knowledge, however, as to
repertoires of choice and strategies for choosing.
As a general principle, one may assume that difference of
means will condition differences in what is accomplished; that
would seem to hold for the comparative study of. symbolic forms
as a whole, including those of speech and writing. That speech
and writing are not simply interchangeable, and have been de-
veloped historically in ways at least partly autonomous, is obvious.
There is little hard knowledge,
tonomy and the consequences ofhowever,
it.
as to the degree of au-
One thing we do know is that a given society may define the
role of any one medium quite differently from another society, as
to scope and as to purpose. I have elaborated this theme with
regard to speaking elsewhere. Here, let -me illustrate it briefly
with regard to writing.25 For one thing, new writing systems
continue to be independently invented--one was devised in 1904
by Silas John Edwards, a Western Apache shaman and leader of a
nativistic religious movement. The 'sole purpose of the writing
system is to record the sixty-two prayers Silas John received in
his vision and to provide for their ritual performance. Compe-
tence in the system has been restricted to a small number of
specialists. Discovery and study of this system by Keith Basso
has shown that existing schemes for the analysis of writing sys-
tems fail to characterize it adequately, ,and probably fail as well
for many other systems, having been devised with evolutionary,
a priori aims, rather than with the aim of understanding individual
systems in their own terms. The development of an ethnography
of writing, such as Basso is undertaking, is long overdue.26
Here belongs also study of the many surrogate codes found round
the world--drum-language, whistle-talk, horn-language, and the
like -for their relation to speech is analytically the same as that
of writing,27 and they go together with the various modalities of
graphic communication (handwriting, handprinting, typing, typo-
graphic printing, etc.) and the various modalities of oral com-
munication (chanting, singing, declamation, whispering, etc.)
a general account of the relations between linguistic means andin
ends.
As to ends, the Hanunoo of the Philippines are literate- -they
have a system of writing derivative of the Indian Devanagari--but
they use it exclusively for love-letters, just as the Buan of New
Guinea use their writing. In central Oregon the town of Madras
has many signs, but the nearby Indian reservation, Warm Springs,
has almost none, and those only where strangers impinge- -the
residents of Warm Springs do not need the information signs
give. 2 8 Recently Vista workers tried to help prepare Warm

48
34. Speech and Language
Springs children for school by asking Indian parents to read to
them in preschool years. U.S. schools tend to presuppose that
sort of preparation, and middle-class families provide it, showing
attention and affection by reading bed-time stories and the like,
but Warm Springs parents show attention and affection in quite
other ways, had no need of reading to do so, and the effort got
nowhere. The general question of the consequences of literacy
has been forcefully raised for contemporary European society by
Richard Hoggart in a seminal book.29
In general, many generalizations about the consequences of
writing and the properties of speaking make necessities out of
possibilities. Writing, for example, can preserve information,
but need not be used to do so (recall IBM's shredder, Auden's
"Better Burn This"), and we ought to beware of a possible ethno-
centrism in this regard. Classical Indian civilization committed
vital texts to memory, through careful training in sutras, for fear
of the perishability of material things. Classical Chinese cal-
ligraphy, the cuneiform of Assyrian merchants, and the style of
hand taught to generations of Reed students by Lloyd J. Rey-
nolds, are rather different kinds of things. Television may have
great impact, but one cannot tell from what is on the screen
alone. In any given household, does the set run on unattended?
Is the picture even on? Is silence enforced when a favorite pro-
gram or the news comes on? Or is a program treated as a re-
source for family interaction?
We have had a great deal more study of means than of mean-
ings. There appear to be many more books on the alphabet than
on the role of writing as actually observed in a community; many
more pronouncements on speech than ethnographies of speaking;
many more debates about television and content analyses of pro-
grams than first-hand accounts of what happens in the rooms in
which sets are turned on. The perspective broached above with
regard to speech communities applies here, since media are a
constituent of the organization of ways of speaking (i.e. , ways
of communication). We need particularly to know the meanings of
media relative to one another within the context of given roles,
setting, and purposes, for the etiquette of these things enters
into whatever constitutive role a medium may have, including the
opportunity or lack of it that persons and groups may have to
use the medium. In England a typed letter is not acceptable in
some contexts in which it would be taken for granted in the United
States; the family Christmas letter in the United States is a genre
that can be socially located; subgroups in the United States differ
dramatically in their assumptions as to what should be photo-
graphed and by whom. 3° At Warm Springs reservation last Aug-
ust, at the burial of a young boy killed in a car accident, his
team-mates from the Madras High School spoke haltingly in their
turn beside the grave and presented the parents with a photo-
graph of the boy in athletic uniform, "as we would like to

43
Language in Education 35

remember him"--a shocking thing, which the parents stoically let

pass--for the last sight of the dead person, which bears the
greatest emotional distress, had already been endured in the
church before coming to the cemetery. When the rites were com-
plete, Baptist and Longhouse, when all the men, then all the
women, had filed past the gravesite, taking each in turn a hand-
ful of dirt from a shovel held out by the uncle of the boy, and
dropping it on the half-visible coffin within' the site, when the
burial mound had been raised over the coffin, the old women's
singing ended, and the many flowers and the toy deer fixed
round the mound, then, as people began to leave, the bereaved
parents were stood at one end of the mound, facing the other,
where their friends gathered to photograph them across it. That
picture, of the manifestation of solidarity and concern on the part
of so many, evident in the flowers, might be welcome.
The several media, of course, may occur together in _several
mixes and hierarchies, in relation to each other and in relation to
modalities such as touch. Communities seem to differ as to
whether tactile or vocal acts, or both together, are the indis-
pensable or ultimate components of rituals of curing, for example.
In some parts of Africa, languages are evaluated partly in terms
of their greeting systems, and the Haya of northern Tanzania,
who are acquiring Swahili, find it less satisfying than their own
language, for in a Haya greeting one touches as well as talks.31
Finally, the use of media and modalities needs to be related
to the norms by which a community takes responsibility for per-
formance and interpretation of kinds of communication. My stress
here obviously is on the qualitative basis of assessing media as a
human problem. Statistics on radios and newspapers and the like
barely scratch the surface. I think it entirely possible that a
medium may have a constitutive effect in one community and not
in another, due to its qualitative role, its social meaning and func-
tion, even though frequencies of occurrence may be the same in
both. We have to do here with the question of identities and
identifications, mentioned earlier with regard to varieties of lan-
guage in schools. We need, in short, a great deal of eth-
nography.
Overcoming the Structure of Language. Concern to overcome
the structure of language seems to have centered around the func-
tion of naming, either to achieve a uniform relation between lan-
guage and meaning as a semantic ideal, or to avoid it as a spirit-
ual desert or death. Early in the development of Indo-European
studies, when modern languages were thought degenerate in form,
the great pioneer of reconstruction, Franz Bopp, sought to infer
an original Indo-European structure in which meanings and mor-
phemes went hand in hand, reflecting perhaps an original, neces-
sary relationship. Others have sought to realize a semantic ideal
in the present, by constructing an artificial language, or by
36 Speech and Language

reconstructing an existing ont: to convey the universal meanings

required by science and philosophy. One thinks especially of
the late seventeenth century (Dalgarno, Bishop Wilkes, Leibniz)
and the early twentieth century (Russell, the early Wittgenstein,
Carnap, Bergmann, and others). Still others have thought that
the ideal relationship between meaning and form might be glimpsed
in the future, once linguists had worked thrOugh the diverse
structures of existing languages to the higher level of structure
beyond them. Such was Whorf's vision. 32
At an opposite extreme would be a philosopher like Brice
Parain, who despairs of the adequacy of language, and of course
adherents of the Zen tradition that regards language's inveterate
distinguishing of things as a trap to be transcended. Inter-
mediate would be the conscious defense of other modes of mean-
ing than that envisioned in the "semantic ideal," in particular,
the defenses of poetry and of religious language.33 And here
would belong conceptions of literary and religious use of language
as necessarily in defiance of other, conventional modes of use.
Much of philosophy and some of linguistics seem to have found
their way back to an open-ended conception of the modes of
meaning in language, and are experiencing great surges of inter-
est in poetics and rhetoric.
Such work is of the greatest importance, but it does leave
the general question of the adequacy of language, or of a particu-
lar language, in abeyance. It would seem that the structures of
languages have never been wholly satisfactory to their users, for
they have never let them rest. Shifts in the obligatory gram-
matical categories of languages over time, like the shift from
aspect to tense in Indo-European, bespeak shifts in what was
deemed essential to convey. Conscious reports of such concerns
may have appeared first in classical Greece, when Plato complained
that the processual character of Greek verbs favored his philo-
sophical opponents, although, at the time, devices such as the
suffix -itos for forming abstract nouns were growing in produc-
tivity. When in the fourth century A.D. Marius Victorinus tried
to translate Plotinus from Greek into Latin, there was no adequate
abstract terminology in his contemporary Latin, and his clumsy
efforts to coin one met with little acceptance, thus inhibiting the
spread of the Neo-Platonic philosophy in that period. Some cen-
turies later "theologisms" had evolved in Latin which quite
matched the terms of the Greek fathers in precision and maneuver-
ability.34 In the early modern period, English writers lamented
the inadequacies of English and set out to remedy them.35 At
Warm Springs, some fifteen years ago, a speaker of Wasco (a
Chinookan language), acknowledging Wasco's lack of a term for a
contemporary object, said that when he was a boy, if one of the
old men had come out of his house and seen such a thing, he
would have coined a word for it, "just like that" (with a sharp
gesture). There are no such old men anymore to coin words or
shape experience into the discourse of myth. 6 Such fates are
Language in Education 37

common, though not much attended to by linguits. The official

preference is to stress the potentiality of a language and to ig-
nore the circumstances and consequences of its limitations. Yet
every language is an instrument shaped by its history and pat-
terns of use, such that for a given speaker and setting it can
do some things well, some clumsily, and others not intelligibly at
all. The cost, as between expressing things easily and con-
cisely, and expressing them with difficulty and at great length,
is a real cost, commonly operative, and a constraint on the theo-
retical potentiality of language in daily life. Here is the irreduci-
ble element of truth in what is known as the "Whorf hypothesis":
means condition what can be _ddne with them, and in the case of
languages, the meanings that can be created and conveyed. The
Chomskyan image of human creativity in language is a partial
truth whose partiality can be dangerous if it leads us to think
of any constraints on linguistic communication either as nugatory
or as wholly negative. As to the force of such constraints, the
testimony of writers and the comparative history of literary lan-
guages should, perhaps, suffice here.37 As to their positive
side, we seem to need to repeat the development of thought dis-
cerned by Cassirer in Goethe, Herder, and W. von Humboldt:
To them, the Spinozistic thesis, that definition is limita-
tion, is valid only where it applies to external limitation,
such as the form given to an object by a force not its
own. But within the free sphere of one's personality
such checking heightens personality; it truly acquires
form only by forming itself.... Every universal in the
sphere of culture, whether discovered in language, art,
religion, or philosophy, is as individual as it is uni-
versal. For in this sphere we perceive the universal
only within the actuality of the particular; only in it
can the cultural universal find its actualization, its
realization as a cultural universal. 38
We need, of course, ethnography to discover the specific forms
which the realization of universality takes in particular communi-
ties, and, where the question is one of speech, we need eth-
nographies of speaking.
Whorf himself led in describing the organization of linguistic
features pertinent to cultural values and world views as cutting
across the usual sectors of linguistic description, and as involv-
ing "concatenations that run across...departmental lines" (that
is, the lines of the usual rubrics of linguistic, ethnographic, or
sociological description that divide the study of a culture and
language as a whole). 39 Whorf referred to the required organi-
zation of features as a fashion of speaking, and one can see in
his notion an anticipation, though not developed by him, of the
sociolinguistic concept of ways of speaking. The crucial differ-
ence is that to the notion of speech styles, the sociolinguistic
38 Speech and Language
approach adds the notion of contexts of situation and patterns
relating style and context to each other.
Here, as before, the great interest is not merely in diversity
or uniformity, but in the possibility that such differences shape
or constitute worlds. Do semantic-syntactic structures do so?
Sapir and Whorf thought that for the naive speaker they did,
although contrastive study of language structures was a way to
overcome the effect. What Chomsky describes as the seemingly
untrammeled "creative aspect" of language use was treated by
Sapir as true, but not true in the same way for speakers of
different languages. Each language has a formal completeness
(i.e. , it shares fully in the generic potentiality of human lan-
guage), but does so in terms of an orientation, a "form-feeling"
of its own, so as to constitute quite a unique frame of reference
toward being in the world. A monolingual's sense of unlimited
adequacy is founded on universality, not of form or meaning, but
of function, and that very sense, being unreflecting, may con-
fine him all the more. The particular strengths f a given lan-
guage are inseparable from its limitations. This is what Sapir
(preceding and giving the lead to Whorf) called
a kind of relativity that is generally hidden from us by
our naive acceptance of fixed habits of speech as guides
to an objective understanding of the nature of experi-
ence. This is the relativity of concepts, or, as it might
be called, the relativity of the form of thought.... It is
the appreciation of the relativity of the form of thought
which results from linguistic study that is perhaps the
most liberalizing thing about it. What fetters the mind
and benumbs the spirit is ever the dogged acceptance of
absolutes. 4
I think this is as fair a statement of the evidence and parameters
of the situation today as it was a half-century ago when Sapir
wrote it. I cite Sapir here partly because I think that linguis-
tics in the United States, having worked its way through a
decade or so of superficial positivism, shows signs of having
worked its way through another decade or so of superficial
rationalism, and a readiness to pick up the thread of the com-
plexly adequate approach that began to emerge in the years just
before the Second World War in the work of men like Sapir,
Firth, Trubetzkoy, and Jakobson.
To return to relativity: the type associated with Sapir and
Whorf in any case is underlain by a more fundamental kind. The
consequences of the relativity of the structure of language depend
upon the relativity of the function of language. Take, for exam-
ple, the common case of multilingualism. Inference as to the
shaping effect of some one language on thought and the world
must be qualified immediately in terms of the place of the speak-
er's languages in his biography and mode of life. Moreover,
communities differ in the roles they assign to language itself in
Language in Education 39

socialization, acquisition of cultural knowledge, and performance.

Community differences extend to the role of languages in naming
the worlds they help to shape or constitute. In central Oregon,
for example, English speakers typically go up a level in taxonomy
when asked to name a plant for which they lack a term: "some
kind of bush"; Sahaptin speakers analogize: "sort of an A," or
"between an A and a B" (A and B being specific plants); Wasco
speakers demur: "No, no name for that," in keeping with a cul-
tural preference for precision and certainty of reference.41
This second type of linguistic relativity, concerned with the
functions of languages, has more than a critical, cautionary im-
port. As a sociolinguistic approach, it calls attention to the
organization of linguistic features in social interaction, and cur-
rent work has begun to show that description of fashions of
speaking can reveal basic cultural values and orientations. The
worlds so revealed are not the ontological and epistemological
worlds of physical relationships, of concern to Whorf, but the
worlds of social relationships. What are disclosed are not orien-
tations toward space, time, vibratory phenomena, and the like,
but orientations towards persons,. roles, statuses, rights and
duties, deference and demeanor.42 Such an approach obviously
requires an ethnographic base. 4 3
Overcoming the Function of Language. Diversity is a rubric
under which the phenomena of language as a human problem can
be grasped; the questions which underlie our concern with di-
versity can be summed up in the term, function. What differ-
ences do language diversities make through their role in human
lives? Some of these differences have been touched upon, and I
want to take space for only general consideration here. Linguists
have mostly taken the functions of language for granted, but it
is necessary to investigate them. Such investigation is indeed
going on, but mostly not in linguistics. It is a striking fact
that problems of overcoming some of the ordinary functioning of
language in modern life attract increased attention from philoso-
phers, writers, and sociological analysts of the condition of com-
munication in society, while many linguists proceed as if mankind
became more unified each time they used the word "universal";
freer and more capable of solving its problems each time they in-
voked linguistic competence and creativity. (This is what I mean
by superficial rationalism.)
Serious analysis of the functioning of language is to be found
in England and the continent much more than in the United States.
Let me merely mention here Merleau-Ponty on the "prose of the
world," Heidegger on speaking as "showing," Brice Parain (al-
ready cited) on the inadequacy of language, Barthes on l'ecriture,
LeFebvre on discours, Sartre on precoded interpretations of
events such as the Hungarian uprising, and Ricoeur on hermeneu-
tics, and state briefly the significance of two approaches, those
of Bernstein and of Habermas.
40 Speech and Language
Bernstein's work has a significance apart from how one
assesses his particular studies, which have been considerably
shaped by the exigencies of support for practical concerns. His
theoretical views, which precede these studies, are rooted in a
belief that the role of language in constituting social reality is
crucial to any general sociological theory, and that that role has
not yet been understood because it has been approached in terms
of an unexamined concept of language. For Bernstein, linguistic
features affect the transmission and transformation of social
realities through their organization into what he calls codes;
that is, through selective organization of linguistic features, not
through the agency of a "language" (e.g., English) as such. He
is noted for his twin notions of restricted and elaborated codes.
This dichotomy has not always done the texture of his thought
good service, for the two notions have had to subsume a series
of dimensions that ought analytically to be separated, and that
combine differently in different communities. (See an analysis in
Hymes 1974, ch. 4, and discussion below.) 44 Nevertheless, one
dimension essential to his work is essential to understanding lan-
guage as a human problem in the contemporary world. It is the
dimension of a contrast between more implicit and more explicit
styles.
Let me interpret Bernstein's view of a major importance of the
contrast. It is not that one of the styles is "good ," the other
"bad:" Each has its place. The more implicit style, in which
many understandings can be taken for granted, is essential to
efficient communication in some circumstances, and to ways of life
in others. But, and this is an aspect of Ber-nstein's view that
has often been overlooked, the more explicit style is associated
with predominantly universalistic or context-free meanings, while
the more implicit style is associated with particularistic or context-
specific meanings. And, argues Bernstein, the universalistic
meanings possible to the more explicit style are essential, if one is
to be able to analyze means of communication themselves, the ways
in which meanings come organized in a community in the service
of particular interests and cultural hegemony, and so gain the
knowledge and leverage necessary for the transformation of social
relationships.
Bernstein has surely put his finger on a crucial issue. There
is inequality in command of verbal resources, and in access to
them, and it is not the case that inequality would be overcome
simply by ending prejudice and discrimination against diverse
forms of speech. Some discrimination among verbal abilities and
products is not prejudice, but accurate judgment. The transfor-
mation of society to a juster, more equal way of life requires
transformation of genuine inequalities in verbal resource. But-
here is the crux--we know very little accurately about the distri-
bution of verbal resource and ability in our society. We know too
little to be able to specify the nature of the intrinsic inequalities
 Language in Education 41

and to judge appropriate remedies. We must be thankful to Bern-

stein for the courage to insist on an essential truth--within one
and the same "language-," there are socially shaped contrasts in
way of speaking and verbal resource--but we must go beyond his
analytic scheme.
The implication of Bernstein's argument is that command of the
more explicit style (his "elaborated code") should made common
to all. To apply such a remedy, one would have be to enable others
to identify reliably the more explicit style, on the one hand, and
the desired kind of cognitive power, on the other, and one would
have assumed a necessary link between them. Let me suggest
some of the difficulties. It would hardly suffice to equate the
style with the proprieties of the standard language (although some
would be tempted to do so); nor can one equate cognitive power
with profusion of words. Certain kinds of analysis of social life
no doubt require certain kinds of verbal resource, but we are far
from knowing how much of the verbal style in which we couch
such analysis now is necessary, how much merely customary.
There are verbal repertoires without something of what is neces-
sary--in this I agree with Bernstein. But is the remedy only a
matter of lack of certain concepts and terms? Or of certain modes
of analytic statement (together with verbal means that facilitate
them)? Or of an entire orientation toward meanings, as Bern-
stein suggests?
There is further difficulty in linking the one style to uni-
versalistic, context-free meanings. No use of language, of
course, is ever wholly context-free. The indexical function, as
Pierce called it, is ever-present and ever essential to interpreta-
tion (as Harold Garfinkel has especially stressed in developing
the perspective known as "ethnomethodology") . Certainly there
are differences in degree of dependence and independence, but
their relationship to forms of social life and cognitive power is not
self-evident. The distribution of these things within our society
is little known. We may think of science and scholarship as deal-
ing in universalistic, context-free meanings, but their work has
become highly particularistic and context-dependent, if one thinks
in terms of ability and opportunity to share in it. There are
large elements of faith and authority, both for those outside these
fields and for those within them (as studies in the sociology of
science and knowledge show). If public communicability of analy-
tic knowledge is considered, then adaptation to particular contexts
of understanding may have an essential role. Some forms of
knowledge, indeed, may require "literary" rather than "scientific"
methods for their effective transmission, and it is not clear where
such verbal methods fit within the contrast in question. The
understanding of the perspective of others that is necessary to
desired forms of change requires uses of language with narrative
and expressive qualities; these qualities often partake of particu-
laristic, context-dependent meanings. It may be that some who
42 Speech and Language
would be said to have an "elaborated code" need greater command ti
of such qualities, and the devices that convey them, to make
their efforts to change ideas and practices effective. It may be
that some who would be said to have a "restricted code" have
sufficient analytic power, but need command of certain of such
qualities and devices in order to be heard by some they seek to
reach. Finally, we tend to think of explicitness as frankness,
as egalitarian and democratic (at least in public communication),
yet in some societies (cf. Rosaldo 1973) explicitness is experi-
enced as authoritarian, whereas implicitness, allusion, and in-
directness is essential to traditional, reciprocal, consensual modes
of resolving issues.
It seems that Bernstein's analytic scheme has inherited a long-
standing tendency to dichotomize kinds of meaning and communi-
cation, and to consider kinds primarily in terms of a cognitive
ideal, whereas the actual fabric of relationships among kinds of
meaning, communicative style, and social consequences is intri-
cate.
This is not to depreciate the importance of Bernstein's work.
More than anyone else in sociolinguistics, he has called attention
forcefully to essential dimensions of the organization of ways of
speaking and styles of speech. A contrast or polarity of explicit-
ness and implicitness is probably a universal dimension of means
of speech. The same is true of simplification and complication of
message-form, another dimension associated prominently with the
notions of "restricted" and "elaborated" codes. One or tt.,a other
is frequently the salient feature of an important, institutionalized
use of language. There is need for analytic clarification of these
dimensions, as elements of general linguistic theory, and for a
wide range of descriptive and comparative studies. Contemporary
linguistics has given attention to simplification and complication
as aspects of pidginization and creolization, but their universal
relevance has been neglected (cf. Introduction, Part III, in
Hymes 1971; a critical analysis of the notions has been broached
by Mary Hope Lee).
Bernstein himself does not claim validity for his analysis be-
yond the English situation. The fact that his work attracts
international attention indicates that it corresponds to something
real in other situations; the proper use of the stimulus of his
work is not to impose its categories, or conjure with them, but
to discover the way in which the dimensions to which he calls
attention do come organized in the given case.
In doing so, it is necessary to differentiate dimensions from
one another (e.g. , explicitness/implicitness, and complication/
simplificationexplicitness may go together with either a complex
or simplified message-form). And it is necessary to disentangle
three factors of communicative events. A broad dichotomy, such
as "restricted" vs. "elaborated" (or some particular dimension),
may easily suggest a contrast that subsumes message-form,
 Language in Education 43
content, and context restriction (or elaboration), say, in form,
content, and context, all three. The three factors are in fact
analytically, and often empirically, distinct. One must discover
the relationships among them.
In applying the global contrast, or a contrast of dimension,
then, one must not begin with a simple two-fold choice (as many
have done and continue to do). There may be restriction or
elaboration in respect to each factor. Such a contrast (symbol-
ized here R /E) generates eight possible types of relationship, as
the following table shows.

Message-form Content Context

(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
The eight-fold framework provides a more adequate, because
more'differentiated, starting point, but chiefly it simply illus-
trates the need of differentiation and needs to be superseded in
the light of empirical work.
Bernstein himself has elaborated his initial dichotomy in
different respects, and it is instructive to consider each criti-
cally. One kind of elaboration develops additional distinctions of
content and application (again, cf. reference 44). These dis-
tinctions have their own interest; the critical observation to be
made is that the basis of elaboration remains binary contrast in
range of alternatives. At each point, one category has greater,
one lesser, range. Here is the source of a limitation of Bern-
stein's development of his initial insight, and of applications of
such ideas. First, binary contrasts may be inadequate to the
actual organization of ranges of alternatives. The locally rele-
vant, -validcategorization may not be binary, but quantitatively
variable along a scale, may be ternary, etc. If one discovers
local norms, there may be a contrast between two styles, one
more and one less explicit, or complex. There may also be an
unmarked norm, from which a second and third style are dis-
tinguished, as markedly more and less explicit, or simple and
complex, respectively. Binary categories, however suggestive,
prejudge.
Second, there is a persistent tendency to interpret the wider
range of alternatives (the "elaborated" category) as more valu-
able, even though Bernstein himself sometimes cautions against
 Speech and Language
this. It is hard to avoid such interpretation, especially if one
thinks in terms of a cognitive ideal. "More" suggests more in-
formation, more precision, etc. Yet in actual life, forms of
message with the widest range of contexts open to them may be
the least valued, others with a narrower range valued the most.
A message more elaborated in form may be considered more trivial
in content. Evaluation cannot be built into the descriptive frame-
work. Local orientations toward meaning, values, must be dis-
covered. Consider an illustration of rows (5, 6) above. Within
institutions and circles of high prestige, most of the elaboration
of form of a genre of message may be treated as incidental re-
affirmation, even if not predictable in detail, and the key to
interpretation, found in manner or nuance of expression; such
as orientation is very similar to the orientation described for "re-
stricted code" use in Bernstein's initial formulations, and identi-
fied there with lower social status. Again, consider an illus-
tration of rows (3, 4). A message may be restricted in form,
highly predictable, context-determined, yet considered rich and
open in content. Men and women of high status, commanding
what their community considers valued elaborated forms, may
give much of themselves to repeated experience of a message, a
piece of music, ritual sequence, literary or religious text, find-
ing, not increase of information so much as increase of connec-
tion, resonance, depth. Spareness, predictability, context-
dependence or form may go together with either shallowness or
with depth, with poverty or with richness, of meaning, and so
may prolixity, unpredictability, context-independent of form.
The value of the meaning is analytically independent of the code-
characteristics. There may be a tight connection in particular
cases, but one cannot prejudge what will be connected to what.
It is indeed an important step toward a pragmatically ade-
quate basis of research just to transect the initial dichotomy with
this one of valued meaning, so as to be able to speak of "deep
elaborated /shallow elaborated" and "deep restricted /shallow re-
stricted" codes, variants, or styles. (Nothing of course depends
on the particular adjectives "deep" and "shallow"; another pair,
such as "thick" and "thin," would serve.)
This four-fold distinction resembles, but seems different
from, an elaboration that Bernstein himself has made. He has
come to distinguish "codes" from "variants" (Bernstein 1973).
Earlier, some persons had been said to have both elaborated and
restricted codes, and others only restricted codes. Now each
"code" is considered to have both "restricted" and "elaborated"
variants. Obviously this is not to reduce the two codes to
equivalence. "Code" continues to designate an underlying, selec-
tive orientation toward kinds and possibilities of meaning. And
it seems that a person is still considered to come to have essen-
tially one code-orientation or the other.
Now, recognition of parallel variants in both codes does mean
 Language in Education 45
r
that one cannot readily determiAe the presence of a "code" from
the form of messages. Since much of Bernstein's earlier work
sought to identify codes from-features of- message-form, it is
called into question by the change. If the implications of the
change are pursued, then future research must concentrate on
communicative strategies in natural settings, and employ partici-
pant observation fully. Operative orientations toward meanings
cannot be assessed adequately from text apart from context.
(An apparently "restricted" utterance may be merely practical in
a context in which something is known to one's hearer; an appar-
ently "elaborated" utterance may be elaborated from pedantic
habit, not cognitive force.) To adapt and revise an assumption
once formulated by Martin Joos, 1.:)37 and large "text does not sig-
nal its own strategy." Much of what is needed for assessment of
orientations is accessible only in persons, not in transcripts.
And it becomes essential to speak in the first instance, not of
"codes," but of "styles," as I have done earlier in interpreting
Bernstein's view of the implicit /explicit dimension. To speak of
a "style" leaves open the meaning of the style
it. "Code," in Bernstein's work, does not. to those who use
As noted, Bernstein seems to consider that a person comes
to have essentially either one code-orientation or the other. The
restricted variant of the elaborated code seems intended to ac-
count mainly for predictable aspects of social interaction, greet-
ings, casual conversation, and the like (as did the earlier attribu-
tion of a restricted code to elaborated code users), rather than
for the experience, say, of a middle-class Christian Scientist
hearing the Bible and Mary Baker Eddy in fixed text, Sunday
after Sunday, and finding new meaning. Bernstein sometimes
attends to such situations, but the thrust of his analysis con-
tinues to be that the distribution of code-orientations in the
society is tantamount to a distribution of people. (Else why dis-
tinguish distinct codes as underlying parallel variants?)
I want to suggest that there is something answering to the
two types of code-orientation, but not, on that basis alone, two
types of people. I want to suggest that persons in fact have
alternative code-orientations, that such indeed is the common
state of affairs in modern society, and that the central problem
is not that some people have one, that others do not (as most
users of Bernstein's ideas have assumed). The central problem
is the management of the relation between the two. If people
differ as types in terms of code-orientations, it is in terms of
types of management of the relation. There may be many types
of management; there is no reason to assume in advance that
there are just two. (For an account of an analogous situation
at a national level, cf. Neustupny 1974.)
In sum, each ideal type of code-orientation identified by
Befnstein has a necessary part in the life of a person, whatever-
the person's social origin and experience. Each person must to
46 Speech and Language
some extent project an analysis of the social life and change in
which he or she is caught up, and each must to some extent
"traditionalize" some sphere of experience and relationships.
Both orientations are to some degree inevitable for all. To
understand people in this regard, then, one must think of them
as having repertoires of code-orientation, and as having to adapt
to a communicative ecology that favors now one, now another,
element of the repertoire, there being often enough serious ten-
sion between person and niche. Many people can be thought of
as having to spend much of their waking life in "verbal passing',"
employing a style constrained by job or group, and unable to
satisfy felt needs for use of language in other ways. Such eco-
logical deprivation may involve lack of others with whom to pur-
sue certain kinds of cognitive elaboration and play, or lack of
others with whom to have certain meanings taken for granted.
Many life choices are made for the sake of "someone to talk to"
in these senses. The problem, then, is not absence of the
orientation in the person, nor absolute absence of contexts for
an orientation, but a specific network of relations between orien-
tations, contents, and contexts.
This analysis brings us to the way in which problems of
modern society have been interpreted as problems of contrasting
code-orientation by Jurgen Habermas.
Habermas develops a contrast analogous to Bernstein's. His
starting point is not observation of class and family differences
in communication, but analysis of theories of knowledge and com-
munication in science and everyday life. Starting from the neo-
Marxian tradition of the Frankfurt School, with its attention to
the Hegelian roots and cultural problems of Marxist thought,
Habermas has turned to the positive contributions of a psycho-
analytic perspective, and the possibilities of groundi.ng in the
nature of language and communication a theory that is critical
of society and emancipatory in aim. He may be said to give a
reinterpretation of Marxian categories of analysis in communi-
cative and linguistic terms.
Like Bernstein, Habermas contrasts two orientations toward
communication. One is a technical cognitive interest, and has
to do with activity guided by technical rules based on empirical
knowledge; such activity comprises "instrumental action systems"
or, generally, "purposive rational action systems." Scientific,
technological, and to some extent bureaucratic modes of ration-
ality and communication are based on this interest. There is,
however, another equally fundamental and valid orientation, a
practical cognitive interest, which has to do with activity
guided by the symbolic processes of everyday life. It is typi-
cally dialogic and narrative in its forms of verification and expla-
nation., and involves interpretive understanding and indeed re-
flexive self - understanding. It cannot be reduced to the models
and formalizations of instrumental action.

61
 Language in Education 47

It is Habermas' view that whereas the "free market" concept

was the dominant ideological rationalization of the capitalist order
in the nineteenth century, the notion of "technological progress"
serves that role today, and that a great threat to human life in
modern society is the invasion of spheres of practical symbolic
interaction by the technological orientation. Value preferences
and special interests masquerade in the idiom of instrumental
necessity; personal and expressive dimensions of meaning become
inadmissible over a greater and greater range of activity. Offi-
cial social science in its positivistic interpretation of its task
actually aids in the maintenance and establishment of technological
control, in contrast to those trends in social science concerned
with understanding socio-cultural life-worlds and with extending
intersubjective understanding (what may loosely be called a
family of "interpretive" approaches), and those trends concerned
with analyzing received modes of authority in the interest of
emancipating people from them.
Whereas the criterion of critical evaluation advanced by Marx
stressed material inequality, and the contradiction between pro-
duction for use and production for profit (use-values vs.
exchange-values), Habermas stresses communicative inequality
and the conflict between an ideal speech situation and communi-
cation distorted and repressed by actual patterns of socialization
and interaction. To quote him (from his article, "Summation and
Response," Continuum (Spring-Summer 1970), p. 131, as cited in
Schroyer 1975:161):
We name a speaking-situation ideal where the communication
is not only not hindered by external, contingent influences,
but also not hindered by forces which result from the
structure of communication itself. Only then does the
peculiarly unforced compulsion of a better argument domi-
nate.
This conception has left Marxism and much of social science
behind. Habermas' ideal adds an invaluable dimension, necessary
to critical analysis of social appearances, but there is no ade-
quate link to ongoing social processes and projected states of
affairs. Real situations can be criticized in terms of the ideal.
No means of progressing toward the ideal, other than criticism,
is given. This is why Schroyer (1975) in a sympathetic account
considers Habermas only to complement, not to replace, a Marx-
ian analysis of inequality and change. One can go further and
suggest that Habermas' analysis contributes only the generic
notion of the ideal communicative situation, as a notion that must
be integrated into the foundations of a linguistic theory adequate
to social life. He does not contribute a satisfactory formulation
of the notion.
Notice that the ideal, as formulated by Habermas, is in the
end analogous to the ideal implicit in Bernstein's treatment of
48 Speech and Language
the "elaborated code." The need for a complementary orientation
is sympathetically recognized by both, and Habermas gives it
foundation in a thorough-going critique of narrow conceptions of
knowledge. But in the end, the role of the mode of symbolic
interaction, for Habermas, is to permit complete explicitness.
The explicitness is rooted, not in a "code" as such, but in the
dialogic relations of the participants in a communicative situation,
and that is a decisive advance. But the possibility of a positive
role for "restrictions" within symbolic interaction is forgotten.
In the light of the ideal (quoted above), all restrictions fall
short.
This, I submit, is "utopian," not in the good sense of an
imagined ideal, but in the bad sense of an ideal whose unrealiza-
bility distorts evaluation of situations and efforts toward change.
The ideal of unrestricted speech is said to be inherent in human
communication, and it seems to be assumed that the logic of his-
torical development moves toward its realization. An ethnographi-
cally informed analysis of ideals of communication suggests other-
wise.
The ideal of unrestricted speech is not the sole ideal inherent
in attested ways of speaking. In some societies, indeed, unre-
stricted speech is viewed as dangerous, and the view is perva-
sively institutionalized (e.g., traditional Ashanti society (Hogan
1971), Burundi (Albert 1964, 1974), Malagasy (Keenan 1973)).
Speech as a source of mischief and evil is a recurrent theme in
cultural wisdom. Such cases might be said to represent a stage
of human history that is to be transcended, and the particular
practices may indeed yield to change (they tend to be associated
with a dichtomization of sex roles, for example), but the particu-
lars are expressions of functions that appear to be perennial
requisites of social life.
Habermas' unique ideal of unrestricted communication is
specifically a cognitive ideal of colloquy. It is an ideal of the
right and contribution of every member of a group in the resolu-
tion of problems. As such, it is an advance over a purely scien-
tific ideal, for it comprises political decision as well (it resembles
the ideal of persuasion long advocated by Kenneth Burke, and
taken by many to be regulative for science and democracy). But
the ideal does not speak to the perennial requisite of structure.
What may appear as restriction is from another point of view
simply the existence of structure. And it is not possible to en-
visage viable social life without structure in the sense at least of
shared understandings of rights and duties, norms of interactions
grounds of authority, and the like. Even the most free conver-
sational situation, if there is taking of turns, begins inherently
to show elements of restrictive structure. Habermas presumably
is concerned simply that no structure prevent a member of a
group from having a right to participate in decision. But if one
considers the possibility as well of an obligation to contribute
 Language in Education 49

what one knows and wants, the lack of right to remain silent or
refuse commitment to a consensus -real enough issues--one has
raised again the matter of constraint. In general, the universal-
ity of appropriateness as a meaning and ideal of speech is equiva-
lent to the inherent presence of a principle of structure in human
speech situations.
Moreover, not all social life is problem-solving and decision-
making in the pertinent sense. There
tions in uses of language that embody are universally satisfac-
play, employ unequally
shared performance abilities, accept ritual-life repetitions of
words accepted as authoritative. One can not envisage a viable
life in which every point is open to dialogic determination at
every point--in which every one can say everything to everyone
in every way at every moment in every place. It can be objected
that this last is an unfair reductio ad absurdum, but Habermas
does not show how to avoid such a reduction. Such a reduction
can be avoided only by a theory communicative competence
that is based on more than purelyofrational reflection, that
built up through patient study and comparison of ideals of iscom-
munication developed in actual communities.
In sum, every known community embodies alternatives to the
unique cognitive ideal, and any community (such as a revolu-
tionary group) that could bring closer approximation to the ideal
would have to embody alternatives. 5 But the theory can only
criticize communicative structure in the light of its absence; it
cannot address real structures and choices among them. Yet
there lies the true problem for any community and person, revo-
lutionary or not. The problem can be phrased in terms of
Haberrnast ideal: What costs in communicative inequality should
be accepted in order to gain the benefit of greater equality than
now obtains? It would be more adequate to say: What kinds of
communicative inequality are acceptable, what unacceptable, in
the light of the historical situation and aspirations of a given
community? (I say "given community," for if a community wished
to maintain certain forms of communicative structure, Habermas'
ideal would not condone its being "forced to be free.")
The psychiatrically informed ideal of the ending of repressed
communication seems faulty on similar grounds. The cognitive
ideal, presumably, is that repression that prevents solution of
life-problems should be overcome. From this it does not follow
that no repression is permissible, that a life should be lived in
the light of an ideal of access to all unconsciously held knowledge.
One could not play tennis that way--but to be serious, such a
prescription would resemble the Christian ideal of a life without
hidden sin, and might entail neurasthenia if rigorously observed.
A healthy view of the relation between conscious and unconscious
knowledge seems to me to be found in the conjunction of a state-
ment by Sapir and a gloss by Zellig Harris (Sapir 1949(1927):
559, Harris 1951: 330) :
50 Speech and Language
Complete analysis and the conscious control that comes
with a complete analysis are at best but the medicine of
society, not its food.
Which means: Don't take it as food; but also: Do take
it as medicine.
Bernstein and Habermas are important, influential, and repre-
sentative in their pioneering efforts to analyze the problems of
linguistic and communicative inequality. Each falls back in a
crucial respect upon a cognitive ideal to which the absence of
restriction, hence "more is better," is intrinsic. Such an ideal
is essential to certain aspects of social problems, but not suf-
ficient to all. Both scholars are able to criticize cultural situ-
ations, but not to articulate alternative situations that answer
the cultural nature of human life, that give a legitimate place to
the practicalities of ordinary life and the full range of needs of
human nature. Both have a sense of the range of needs, to be
sure, and other contemporary theorists of language have hardly
addressed the issues at all. Bernstein and Habermas are repre-
sentative of this situation particularly in the fact that the analy-
sis of each revolves around a dichotomy. The influence of each
reflects the fact that the dichotomy touches something real in
our experience, and also the fact that the realities involved have
only begun to be analyzed. The dichotomies are symptoms of
initial recognition of an issue, first approximations in addressing
it. Knowledge and successful change require us to be able to
link the insights expressed in such dichotomies to actual situa-
tions. Such linkage depends on ethnography, and ethnography
will lead to reconstruction of the initial theories in more articu-
lated, diversified form. Communicative theory, as a foundation
of social theory and practice, will be informed by typologies of
cases, and initial dichotomies will give way to sets of dimensions,
diversely hierarchized and apportioned, in justice to the experi-
ence and aspirations of specific communities.
In sum, the problem of overcoming the function of language
is first of all a problem of discovering tie functions language
does have. It is important to imagine ideal states of affairs, but
the process of imagining and, equally, of implementing ideal com-
municative situations should be, as Habermas implies and Bern-
stein undoubtedly would agree, an open, dialogic one. If diverse
communities and cultural traditions are counted among the voices,
the outcome of such a process is likely to be a plurality of con-
ceptions of what communicative structure should be. For many
communities, the goals of transformation will be not only to over-
come obstacles to openness, but also to overcome threats to pat-
terns interwoven with the meaning of a way of life. If we seek
to evaluate such things critically, a comparative ethnographic
perspective is essential, in order to overcome the obstacle of un-
witting ethnocentrism in our efforts to think about principles and

65
Language in Education 51

premises of verbal interaction. We are likely to extrapolate and

project ideal notions of our own tradition, unwittingly misrepre-
senting the realities of our own conduct and the ideals of others
both. For example, many of us would be likely to link Habermas'
speaker-situation ideal of unforced compulsion with the explicit-
ness of Bernstein's ideal-type "elaborated code." We tend to
think of explicitness (in public communication at least) as frank-
ness, directness, and as egalitarian and democratic in its impli-
cations. In many societies, however, explicitness and directness
are experienced as authoritarian, something associated with im-
posed decisions. Implicitness, allusion, and indirectness are
associated with traditional modes of resolving issues in reciprocal
colloquy (cf. Rosa Id° 1973, both on the Ilongot of the Philippines
and on the general issue).
Thinking About Linguistic Inequality
Occasionally linguists have been so carried away by ideological
certitude as to state that all languages are equally complex. This
is of course not so. It is known that languages differ in sheer
number of lexical elements by an order of magnitude of about two
to one as between world languages and local languages. They
differ in number and in proportion or abstract, superordinate
terms. They differ in elaboration of expressive and stylistic de-
vices-- lexical, grammatical, and phonological. Languages differ
in number of phoneme-like units, in complexity of morphopho-
nemics, in complexity of word-structure (both phonological and
morphological), in degree of utilization of morphophonemically
permitted morpheme-shapes, etc.
The usual view is that such things are distinctions without a
difference, that all languages are equally adapted to the needs of
those who use them. Leaving aside that such equality might be
an equality of imperfect adaptations, speech communities round
the world simply do not find this to be the case. They are found
to prefer one language for a purpose as against another, to ac-
quire some languages and give up others because of their suita-
bility for certain purposes. No Third World government can
afford to assume the equality of the languages within its domain.
The usual answer to this objection is that all languages are
potentially equal. In fact this is so in one vital respect; all
languages are indeed capable of adaptive growth, and it is a
victory of anthropologically oriented linguistic work, particularly,
to have established this point. The difficulty with the usual
answer is twofold. First, given that each language constitutes
an already formed starting point, it is not at all clear that ex-
pansion of resources, however far, would result in languages
beihg interchangeable, let alone identical. Limiting consideration
to world languages, we find that many who command more than
one prefer one to another for one or more purposes, and that
52 Speech and Language

this is often enough a function of the resources of the languages

themselves. The other difficulty is that the realization of po-
tentiality entails costs. The Chomskyan image of the child ideally
acquiring mastery of language by an immanent unfolding misleads
us here. It has an element of truth to which the world should
hearken, but it omits the costs, and the constitutive role of
social factors. Most of the languages of the world will not be
developed, as was Anglo-Saxon, into world languages over the
course of centuries. (It is speculated that Japanese may be the
last language to join that particular club.)
I regret to differ from admired colleagues on this general
issue, but it seems necessary, if linguistic work is to make its
contribution to solution of human problems, not to blink realities.
How could languages be other than different, if languages have
any role at all in human life? To a great extent, langu.ages, as
I have said, are what has been made of them. Thereis an ele-
ment of truth in the thesis of potentiality and an element of truth
in the thesis of equivalent adaptation across communities, but
both theses fall short of contemporary reality, where languages
are not in fact found unmolested, as it were, one to a community,
each working out its own destiny in an autonomous community.
Not to take the step to that reality is to fall back on the "Herd-
erian" image, a falling back that is all too common. If that image
were a reality, then the analysis of linguistic inequality would
perhaps be only an academic exercise for scholars who take
pleasure in languages the way one may take pleasure in kinds
of music. Given our world, however, analysis of linguistic in-
equality is of great practical import.
What, then, are the sources and consequences of linguistic
inequality? The kinds of diversity already discussed contribute,
of course, but the plain fact is that having hardly raised the
question we have no clear notion. A Parsonian set of categories
can serve as an initial guide.
First, languages differ in their makeup as adaptive resources;
the linguistic resources of speech communities differ in what can
be done with them, as has been indicated. A generation ago
some kinds of difference were regarded with a spirit of relativis-
tic tolerance, as the special virtues of the languages that had
them, and so one got at least some account of their lexical and
grammatical strengths. The present temper, however, treats
mention of differences as grounds for suspicion of prejudice, if
not raasm, so that poor Whorf, who believed fervently in the
universal grounding of language, and extolled the superiority of
Hopi, has become, like Machiavelli, a perjorative symbol for un-
pleasant facts to which he called attention. Until this temper
changes, we are not likely to learn much about this fundamental
aspect of language.
Second, linguistic resources differ as an aspect of persons
and personalities. In addition to the variability inevitable on
Language in Education 53

genetic grounds, there is the variability due to social pattern-

ing. Conceptions of male and female roles, or of specialized
roles, including that of leadership, may differ markedly among
speech communities so that eloquence or other verbal skills may
be necessary for normal adult roles in one society (commonly for
men, not women), and essential to no important role at all in
another. The requirements of a speaking role may be simple,
or subtle and difficult as they are in the special bind of a tra-
ditional Quaker minister who had to speak out of spiritual silence
and, desirably, after periods of doubting his calling.64
Dif-
ferences in verbal skills desired, of course, feed back upon the
ways in which the linguistic resources of a community are
elaborated.
Third, linguistic resources differ according to the institu-
tions of a community. So far as I know , comparative analysis of
institutions has not much considered the ways in which they do
and do not require or foster particular developments of verbal
skill and resource, or at least has not phrased its findings as
contributions to the understanding of language. There are in-
deed some analyses of the development of the verbal style and
resources of particular sciences, of science as a social movement,
and of religious and political movements. My impression is, how-
ever, that one finds case studies, but not coordinated efforts
toward a comparative analysis and a theory.
Fourth, linguistic resources differ according to the values and
beliefs of a community. Infants' vocalizations, for example, may
be postulated as a special language, one with serious conse-
quences, such that special interpreters are required, so that a
child's wishes can be known and its soul kept from returning to
whence it came. The shaping of linguistic resources by religious
concerns appears to be attracting a surge of interest-47 A com-
munity's values and beliefs may implicitly identify spontaneous
speech as a danger to the cultural order, as among the tradi-
tional Ashanti, or they may treat speaking and especially elabor-
ate speaking, as a badge of inferiority, both between persons
and among the orders of a social hierarchy, as is the case with
the Wolof of Senegal. The normal condition of a community may
be constant chatter on the one hand, or pervasive quiet on the
other, according to how speech is valued.
Such a guide to differences does not in itself go beyond a
"Herderian" perspective of discrete speech communities, each
part of the cultural plenitude of the world. Such description
bears on inequality, however, when speech communities are
viewed in a larger context. Differences by themselves would
constitute inequality only in the sense of lack of equivalence,
not in the sense of inadequacy. But just as the resources of a
speech community must be described as speech styles in relation
to contexts of situation, so must they also be assessed in relation
to their contexts when the perspective is that of human problems.
54 Speech and Language
The essential thing seems to me to be to assess the situation of a
speech community in terms of the relation between its abilities
and its opportunities. Every speech community is to some degree
caught up in a changing relationship with a larger context, in
which opportunities for the meaningful use of traditionally
fostered abilities may be declining, and novel opportunities (or
requirements) for which members have not been traditionally pre-
pared may be impinging. The term competence should be em-
ployed within just such a perspective. It should not be used as
a synonym for ideal grammatical knowledge as by Chomsky, or
extended to a speech community collectively as by De Camp, or
extended to ideal communicative knowledge as by Habermas, or
done away with as Labov would seem to prefer; rather, compe-
tence should retain its normal sense of the actual ability of a
person. Just such a term is needed to assess the processes at
work in actual speech communities, and their consequences for
persons. Competence as a term for ideal knowledge may over-
come inequality conceptually for linguists, but only as a term for
the abilities of persons, assessed in relation to contexts of use,
can it help to overcome inequality practically for the members of
speech communities.
Conclusion
To sum up: from one standpoint the history of human society can
be seen as a history of diversity of language, of diversity as a
problem--both diversity of languages as such, and diversity as
to their media, structures, and functions. From another stand-
point, that same diversity has been a resource and an oppor-
tunity--for scholars to understand the potentialities of human
language, and for speakers to develop the potentialities of their
forms of life and of their identities.
From antiquity it has been the mark of a true science of man,
of greatness in a science of man, to attempt to comprehend the
known diversity of cultures and history. Herodotus did so in a
narrative of his age's great conflict between East and West, in-
corporating the ethnology of his world. The Enlightenment,
while recognizing a debt to antiquity, was conscious also of the
superiority and the challenge of a new horizon provided by its
knowledge of manners and customs from the New World, and from
remoter Africa and Asia; the Victorian evolutionists, while recog-
nizing an Enlightenment precedent, were conscious of a superior-
ity and the challenge of a new horizon provided by the recent
recognition of the great prehistoric antiquity of man. In this
century there has been no new horizon of data in space or time
that has vivified the whole (unless one counts primate studies
and finds of fossil man as such), but a principle of methodologi-
cal relativism has been gradually established that is of equal im-
portance. Now we are a juncture where only the future of
Language in Education 55

man offers the challenge of a new horizon to a science of man;

the choices for its future appear to be irrelevance, the service
of domination, or the service of liberation through universaliza-
tion. That is, the sciences of man have developed in the matrix
of a certain relationship between one part of the world and the
rest; a relationship defined in terms, not of aspirations, but of
activities. Anthropology, for instance, is fairly described as the
study of colored people by whites." That matrix has changed
irreversibly. A science of man limited to certain societies or
interests was always implicitly a contradiction in terms; increas-
ingly, it has become an impossibility or a monstrosity. Knowl-
edge about people is a resource, like control of oil and of armies.
Nations cannot accept permanent inferiority in this regard. For
the social scientist, the problem is complicated by the relations
not only between his own country and others, but by the rela-
tions between the governments of other countries and their own
peoples; for usually any knowledge that he can gain that is worth
the having entails entering into a relationship of mutuality and
trust with the people he is studying. Thus universalization of
the science of man must mean extension not only to all countries
of participation, but to all communities. The proper role of the
scientist, and the goal of his efforts, should not be "extractive,"
but mediative. It should be to help communities be ethnographers
of their own situations, to relate their knowledge usefully to
general knowledge, not merely to test and document. a
role could be the safeguard of both the intellectual andand the ethi-
cal purposes of the science itself.
The study of language has had a checkered career in the
history just sketched. It first became a self-conscious activity,
and to a great extent has developed since, as an instrument of
exclusion and domination. The analysis of Sanskrit in ancient
India, of classical songs and writings in ancient China, of Greek
and then Latin in the ancient Mediterranean, of nascent national
languages in the Renaissance (e.g., Nebrijars grammar of Cas-
tilian) , v. ere all in the interest of cultural hegemony.
in our own century, through the decisive work of Boas,It Sapir,
is only
and other anthropologically oriented linguists (as components of
the general triumph of "methodological relativism" in the human
sciences), that every form of human speech has gained the
"right," as it were, to contribute on equal footing to the general
theory of human language.
The present situation of linguistics in the United States is
quite mixed, where it is not obscure. Chomskyan theory holds
out the liberation of mankind as an aspiration, but its practice
can contribute only conceptually at best, if it does not in fact
stand as an obstacle to the kind of work that is actually needed.
This paper has argued for the study of speech communities as
actual communities of sp-mkers. In this way we can go beyond
a liberal humanism which merely recognizes the abstract
56 Speech and Language
potentiality of all languages, to a humanism which can deal with
concrete situations, with the inequalities that actually obtain, and
help to transform them through knowledge of the ways in which
language is actually organized as a human problem and resource.
References
1. Jean--Jacques Rousseau, Discourse on the Origin and
Foundations of Inequality Among Men (1756), trans. Roger D.
and Judith R. Masters, The First and Second Discourses, ed.
Roger D. Masters (New York: St. Martin's Press, 1964).
2. Noam Chomsky, Problems of Knowledge and Freedom
(New York: Pantheon Books, 1971).
3. Edward Sapir, "Psychiatric and Cultural Pitfalls in the
Business of Getting a Living" (1939), Selected Writings of Ed-
ward Sapir, ed. D. G. Mandelbaum (Berkeley and Los Angeles:
University of California Press, 1949), p. 578.
4. "Linguistic Theory" ought to refer to a general theory
of language, or at least a general theory of the aspects of lan-
guage dealt with by linguists, but it has been appropriated re-
cently for just those aspects of language dealt with in trans-
formational generative grammar--another instance of Chomsky's
skill as a-polemicist. Hence the quotation marks.
5. W. M. Urban, Language and Reality: The Philosophy of
Language and the Principles of Symbolism (London: George Allen,
1939), p. 23.
6. Edward Sapir, "Language," Encyclopedia of Social
Sciences, IX (New York: Macmillan, 1933), pp. 155-169, cited
from Selected Writings of Edward Sapir, ed. Mandelbaum, p. 11.
7. G. 3. Metcalf, "The Development of Comparative Lin-
guists in the Seventeenth and Eighteenth Centuries: Precursors
to Sir William Jones," and P. Diderichsen, "The Foundation of
Comparative Linguistics: Revolution or Continuation?" Studies
in the History of Linguistics, ed. Hymes (Bloomington: Indiana
University Press, forthcoming).
8. Discussed in Hymes, "Lexicostatistics and Glottochronology
in the Nineteenth Century" (with notes toward a general history),
Proceedings of the Conference on Genetic Lexicostatistics (tenta-
tive title), ed. I. Dyen (The Hague: Mouton, forthcoming).
9. M. Swadesh, Origin and Diversification of Languages
(Chicago: Aldine, 1971).
10. N. Chomsky, Cartesian Linguistics (New York: Harper
and Row, 1966).
11. F. Boas, "The Classification of American Languages,"
American Anthropologist, XXII (1920), 367-376.
12. "The Problem of Linguistic Communication in the Modern
World," La Monda Lingvo-Problemo, III, No. 9 (1971), 129-176.
13. G. Steiner, Extraterritorial (New York: Atheneum, 1971),
p. 70.
 Language in Education 57
14. Hymes, "latroduction to Part III," Pidginization and
Creolization of Languages, ed. Hymes (London and New York:
Cambridge University Press, 1971), p. 73.
15. There are noble exceptions--Schuchardt in the seven-
teenth century, for one, and the Prague School and J. R. Frith
in the twentieth century, but the main thrust of successive de-
velopments has been as described. Transformational grammar is
included under structural method here because it shares the same
assumptions when contrasted to a functional approach; of the
contrast drawn in Hymes, "Why Linguistics Needs the Sociologist ,"
Social Research, XXXIV, No. 4 (1967), 632-647.
16. L. Bloomfield, Language (New York: Holt, Rinehart &
Winston, 1933), Ch. 3, and N. Chomsky, Aspects of the Theory
of Syntax (Cambridge: M.I.T. Press, 1965), p. 3.
17. Assumptions as to the bases of mutual intelligibility, and
as to relations among linguistic boundaries, ethnic boundaries,
and communication are analyzed in Hymes, "Linguistic Problems
in Defining the Concept of 'Tribe,'" Essays on the Problem of
Tribe, ed. J. Helm (Seattle: University of Washington Press for
the American Ethnological Society, 1968).
18. H. Wolff, "Intelligibility and Inter-Ethnic Attitudes,"
Anthropological Linguistics, I, No. 3 (1959), 34-41.
19. This conception is dealt with in more detail in my "Intro-
duction" to Language in Society, I, No. 1 (1972), 1-14, and "The
Scope of Sociolinguistics," Report of the 23rd Annual Round Table
Meeting on Linguistics and Language Study; Sociolinguistics, ed.
R. W. Shuy (Washington, D.C.: Georgetown University Press).
20. On complex linguistic communities, see C.A. Ferguson,
"National Sociolinguistic Profiles," Sociolinguistics, ed. W. Bright
(The Hague: Mouton, 1966).
of speaking, see Hymes, "TwoOn comparative study of the role
Types of Linguistic Relativity,"
ibid. , and "Models of the Interaction of Language and Social
Life," Directions in Sociolinguistics: The Ethnography of Com-
munication, eds. J. Gumperz and D. Hymes (New York: Holt,
Rinehart & Winston, 1972), pp. 35-71. The work of John Gum-
perz and William Labov has been of special importance to the
understanding of the problems dealt with in this and the preced-
ing note.
21. D. Wade, "The Limits of the Electronic Media ," T L S
Essays and Reviews, V (May 1972), 515-516.
22. The following 14 paragraphs are added to the original
article. The additional references, in order of occurrence, are:
Dwight L. Bolinger, "Visual Morphemes," Language, 22 (1946),
333-340; R.A. Hall, Jr., "Review of 3. Vachek, Written Language:
General Problems and Problems of English," Language, 51 (1975),
461-5; H. J. Uldall, "Speech and Writing," Acta Linguistica, 4
(1944), 11-16; J. Vachek, "Some Remarks on Writing and Phonetic
Transcription," Acta Linguistica, 5 (1944-49), 86-93; N. Chomsky
and M. Halle, Sound Pattern of English (New York: Harper and
58 Speech and Language
Row, 1968); Joseph Jaquith, unpublished MS.; Joel F. Sherzer,
"Verbal and Nonverbal Deixis in San Bias Cuna," Language in
Society, 2(1) (1973), 117-132; D. Hymes, Foundations in Socio-
linguistics (Philadelphia: University of Pennsylvania Press,
1974), p. 102.
23. The possibility is considered by Joseph Greenberg
(Anthropological Linguistics (New York: Random House, 1968),
p. 133) in the course of a lucid account of approaches to lan-
guage classification. He suggests the semantic features of rich-
ness in quasi-synonymity as a possible example of a character-
istic that would permit one to treat languages that exist in written
form as a class from the standpoint of linguistics proper. Green-
berg's requirement is that the linguistically external fact of a
functional role go together with internal facts of structure. In
what follows I argue that the fact of written form does not itself
uniquely determine functional role; there is need for a typology
within the category, "language with a written form." Greenberg's
requirement would still hold for several types within the category,
provided the notion of "internal facts of structure: is interpreted
in the broad sense of the organization of means of speech (and
writing), not in the narrow sense of grammar proper.
24. Both plenitude and integration are perhaps illustrated by
traditional and vernacular styles pointed out by Jaquith. Each
may entail something in speech. A sign in one spelling may per-
haps be read aloud in one way, a sign in the other in another.
If so , if "right" et al. go with one spoken style, "rite" et al.
with another, the differences cannot be in the sounds the letters
are considered to spell. The spelled sounds are the same. The
differences would be in sub-phonemic detail, tempo, voice quality,
intonation, and perhaps other aspects of manner.
25. See works cited in reference 20 and, on writing, "Toward
Ethnographies of Communication," The Ethnography of Communi-
cation, eds. J. Gumperz and D. Hymes (Washington, D.C.:
American Anthropological Association, 1965), pp. 24-25.
26. K. Basso and N. Anderson, "The Painted Symbols of
Silas John: A Western Apache Writing System," Science, CLXXIX
(forthcoming in 1973).
27. T. Stern, "Drum and Whistle Languages: An Analysis of
Speech Surrogates ," American Anthropologist, LIX (1957), 487
506.
28. These observations are from the work of Susan Philips,
to be presented in an article in Foundations of Language De-
velopment, eds. E. and E. Lenneberg, sponsored by UNESCO.
29. R. Hoggart, The Uses of Literacy (London: Chatto and
Windus, 1957); note the introduction to the French. edition,
Working Papers in Cultural Studies, trans. J. C. Passeron
(Birmingham: Centre for Contemporary Cultural Studies, Uni-
versity of Birmingham, 1971), pp. 120-131.

73
Language in Education 59

30. The ethnography of taking pictures in U,S. society is

being studied by Richard Chalfen of Temple University; see Sol
Worth, Through Navaho Eyes (Bloomington: Indiana University
Press, 1972).
31. I owe this information to Sheila Seitel.
32. B. L. Whorf, "Language, Mind and Reality," 1942; cited
from Language, Thought, and Reality: Selected Writings of
Benjamin Lee Whorf, ed. J. B. Carroll (Cambridge: The Tech-
nology Press, 1956), pp. 246-270.
33. B. Parain, Petite Metaphysique de la Parole (Paris:
Gallimard, 1969), translated as A Metaphysics of Language
(Garden City, N.Y.: Doubleday, Anchor Books, 1971); K. Burke,
"Semantic and Poetic Meaning," The Philosophy of Literary Form
(Baton Rouge: Louisiana State University Press, 1941); E. L.
Mascall, Words and Images (London: Darton, Longman and Todd,
1968); I. T. Ramsey, Religious Language (New York: Macmillan,
1957).
34. From a comment by G. E. von Grunebaum, in Language
in Culture, ed. H. Hoijer (Chicago: University of Chicago Press,
1954), pp. 228-229.
35. R. F. Jones, The Triumph of the English Language
(Stanford: Stanford University Press, 1953); cf. F. Brunot, "La
Propagation du frangais en France jusqu'a la fin de l'Ancien
Regime," Histoire de la Langue Frangaise des Origines a 1900,
VII, 2nd ed. (Paris: Colin, 1947), and E. A. Blackall, The
Emergence of German as a Literary Language (Cambridge: Cam-
bridge University Press, 1959).
36. E. Sapir, Wishram Texts (Leiden: E. J. Brill, 1909),
p. 48, lines 1-2.
37. E.g., Eliot's "one has only learnt to get the better of
words /For the thing one no longer has to say, or the way in
which/One is no longer disposed to say it." Four Quartets
(New York: Harcourt Brace, 1943), p. 16. The general question
of the "Herderian" standpoint and of the mixed standing of lin-
guistic resources as determinants is reviewed in Hymes, "Lin-
guistic Aspects of Comparative Political Research," The Method-
ology of Comparative Research, eds. R. T. Holt and J. E.
Turner (New York: The Free Press, 1970), pp. 295-341.
38. E. Cassirer, The Logic of the Humanities (New Haven:
Yale University Press, 1961), pp. 24-25.
39. B. L. Whorf, The Relation.of Language to Habitual
Thought and Behavior (1941), cited from Carroll, Selected Writ-
ings of B. L. Whorf, pp. 158-159.
40. E. Sapir, The Grammarian and His Language (1924),
cited from Selected Writings of E. Sapir, 'ed. Mandelbaum,
pp. 153, 157.
41. From work of David French. On the general issue, see
my papers cited in reference 20.
60 Speech and Language
42. H. M. Hogan, "An Ethnography of Communication among
the Ashanti," Penn-Texas Working Papers in Sociolinguistics , I
(Austin: University of Texas, Department of Anthropology,
1971); R. Darnell, "Prolegomena to Typologies of Speech Use,"
Texas Working Papers in Sociolinguistics (Austin: University of
Texas, Department of Anthropology, 1972), and papers by J. T.
Irvine, E. 0. Keenan, and J. F. Sherzer in The Ethnography of
Speaking, eds. R. Bauman and J. F. Sherzer (London and New
York: Cambridge University Press, forthcoming).
43. M. Cole, J. Gay, J. A. Glick, D. W. Sharp, The Cul-
tural Context of Learning and Thinking (New York: Basic Books,
1971) is an excellent demonstration of the necessity of ethnogra-
phy for assessment of linguistic and cognitive abilities, even
though, unfortunately, the authors do not disclose the linguistic
characteristics of the material on which their work rests.
44. The following 34 paragraphs are added to the original
article. The references in this section, in order of occurrence,
are: D. Hymes, Foundations in Sociolinguistics (Philadelphia:
University of Pennsylvania Press, 1974); B. Bernstein, "A
Critique of the Concept 'Compensatory Education," Functions of
Language in the Classroom, eds. C. Cazden, V. John-Steiner,
and D. Hymes (New York: Teachers College Press, 1972); M.
Rosaldo, "I have Nothing to Hide: The Language of Ilongot
Oratory," Language in Society, 2 (1973), 193-223; Hymes, "In-
troduction, Part III," Pidginization and Creolization of Languages
(London and New York: Cambridge University Press, 1971); B.
Bernstein, Class, Codes, and Control, II (London: Routledge,
Kegan Paul, 1973); Martin Joos, "Linguistic Prospects in the
United States," Trends in European and American Linguistics
1930-1960, eds. C. Mohrmann, A. Sommerfelt, and J. Whatmough
(Utrecht and Antwerp: Spectrum, 1961); J. Habermas, Knowl-
edge and Human Interests (Boston: Beacon Press, 1971); T.
Schroyer, "A Reconceptualization of Critical Theory," Radical
Sociology, eds. J. D. Colfax and J. L. Roach (New York: Basi;.;
Books, 1971); Schroyer, "Toward a Critical Theory for Advanced
Industrial Society," Recent Sociology No. 2, ed. H. P. Dreitzel
(New York: Macmillan, 1970); Habermas, "Toward a Theory of
Communicative Competence ," in Dreitzel, op. cit.; J. B. Neu-
stupny, "The Modernization of the Japanese System of Communi-
cation," Language in Society, 3 (1974), 33-48; T. Schroyer,. "The
Critique of Domination," The Origins and Development of Critical
Theory (Boston: Beacon Press, 1975); Helen M. Hogan, "An
Ethnography of Communication Among the Ashanti," Penn-Texas
Working Papers in Sociolinguistics, I (Austin: University of Texas,
Department of Anthropology, 1971); Ethel Albert, "'Rhetoric,'
'Logic,' and 'Poetics' in Burundi: Cultural Patterning of Speech
Behavior," Directions in Sociolinguistics, eds. J. Gumperz and
D. Hymes (New York: Holt, Rinehart and Winston, 1972); Elinor
Ochs Keenan, "A Sliding Sense of Obligatoriness: The Poly-
Structure of Malagasy Oratory," Language in Society, 2 (1973),
Language in Education 61

225-243; E. Sapir, "The Unconscious Patterning of Behavior in

Society," The Unconscious: A Symposium, ed. E. S. Dummer
(New York: Knopf, 1927), reprinted in Selected Writings of
Edward Sapir, ed. D. G. Mandelbaum (Berkeley and Los Angeles:
University of California Press, 1949), 544 559; Zellig S. Harris,
"Review of Selected Writings of Edward Sapir," Language, 27
(1951), 288-333.
45. Cf. these lines from a poem from a conscientious-objector
camp in Oregon in World War II:
The pacifist speaks, / Face to face with.his own kind, /
And seeks to fashion a common course / That all may
mark. / But whatever he offers, / Finds already framed
in another's thought / A divergent approach. / The
binding belief that each allows / Is cruxed on rejection: /
Thou shalt not kill. [stanza] But for all the rest, / What
Voice shall speak from the burning bush, / In the work-
site 1100119. , / When the loaf is broken, / And brief and
rebuttal countercross, / And no one wins? (William Ever-
son).
46. R. Bauman, "Speaking in the Light: The Role of the
Quaker Minister," The Eihnography of Speaking, eds. Bauman
and Sherzer.
47. Papers on language and religion from a working group
at the 1972 Georgetown Round Table Conference; it is expected
that these papers will be published under the editorship of
W. Samarin.
48. W. S. Willis, Jr., "Skeletons in the Anthropological
Closet," Reinventing Anthropology, ed. D. Hymes (New York:
Pantheon, 1973). This discussion draws on my introduction to
the book.
qualitative/quantitative research
methodologies in education:
a linguistic perspective'

The study of language has a special role to play when one

seeks to come to terms with the relation between quantitative and
qualitative methods. The rise of linguistics in this century as an
autonomous discipline is based on the discovery of a qualitative
methodology. The success of linguists in discovering relationships
that are capable of rigorous formulation, of patent reliability and
validity, without recourse to numbers, has stood as an object
lesson. It is an object lesson that has been heeded most of all
in anthropology, where it is familiar in writings of Sapir, Kluck-
hohn, Levi-Strauss, Goodenough, Lounsbury, Frake, and others,
and has spawned a series of special approaches and debates.
(One can mention componential analysis, ethnoscience, structural
analysis of myth, paralinguistics and kinesics, and various forms
of semiotics.) For whatever reason, this import of linguistics has
not been particularly discussed in educational anthropology. I
should like to sketch its history and present standing, so as to
indicate both the value and the limitations of the perspective it
brings. Linguistics is increasingly being extended today through
attention to social context and use. Such attention entails ethnog-
raphy, and I will end by trying to say how the linguistics and
the ethnography fit.

A few dates and historical reference points are needed. Most

people may not realize that there were no departments of linguis-
tics in this country before the Second World War. The profes-
sional association of linguists, the Linguistic Society of America,
is only 51 years old, roughly half the age of the major social
science associations founded in the latter part of the preceding
century. Fifty years ago what we consider the study of language
was mostly the study of individual languages and language fami-
lies, Indo-European having pride of place. Study of general lin-
guistics, and study of Indo-Euorpean languages, as its founda-
tion, were often considered equivalent. In the 1930s the intro-
ductory courses at the first Linguistic Institutes sponsored by the
new Linguistic Society focused on Indo-European languages.
There were of course students of language in general. But
if one sought a career in the study of language, one pretty much
62
7 A;
Language in Education 63
had to choose between becoming a specialist in the languages and
literatures of some major language group of European fount, or
becoming an anthropologist who could write down languages mostly
unwritten. All this changed and changed dramatically. It
changed in connection with the exploitation of a little-noticed gap
in the existing academic citadel: the sounds of language. The
study of speech sounds was hithergo either taken up within indi-
vidual languages, language families (Romance, Germanic, Slavic),
or taken up as an aspect of psychophysics, of phonetics as a
distinct physical science. For the former purpose, the analysis
was specific to the languages in question. For the latter pur-
pose, one sought exactness of measurement. For many students
of language, the two activities were wholly separate categories.
To study languages was a human (or "moral" or "mental") science:
a Geisteswissenschaft. To study speech sounds, those physical
phenomena, was something altogether different: a Naturwissen-
schaft.
What happened in the 1920s and 1930s was that the men we
now revere as founders of the discipline of linguistics- -men such
as Edward Sapir and Leonard Bloomfield in the United States-
created a methodology, a qualitative methodology, which on the
one hand generalized the insights into particular patterns of
speech sound from the study of particular languages, and on the
other hand transcended the phonetics of pure physical measure-
ment. What they accomplished is loosely called the discovery of
the concept of the phoneme and the creation of phonology as an
entirely general science of the systematic properties of the dimen-
sion of language having to do with sounds. What they did, in
effect, was to integrate the study of sounds in general into a
Geisteswissenschaft of language study by replacing one conception
of rigor with another. Rigor of measurement was replaced by
rigor of functional contrast.
The classical locus of this discovery in American linguistics is
the 1925 paper by Sapir, "Sound patterns of language." From
that paper has flowed not only much of the development of method-
ology in linguistics, but also much of the effect that linguistics
has had on conceptions of methodology for the study of behavior
more generally (see Hymes 1970).
Sapir's essential point was the distinction between a physical
event and an element in a system of signs. The distinction was
dramatized by consideration of cases in which one and the same
physical feature could have entirely different significance, depend-
ing first of all on whether or not it was an element in a system of
signs, and, if it was, then on the relations into which it entered
in the particular system of signs to which it belonged. As to the
first, Sapir considered the difference between a breath through
pursed lips to extinguish a candle, and such a breath as the be-
ginning of an English word such as "when" (when pronounced in
the standard form with the aspiration (hwen)). 2 The initial
breath of the English word can distinguish forms within the
64 Qualitative/Quantitative Research Methodologies
language (hwen : wen (of the skin)), can be an object of atten-
tion as a difference (sometimes stigmatized) between styles of
speech with regard to the adverb "when" itself, and makes possi-
ble exact or approximate puns (as in commenting on someone's
reiterated "When?" with "This must be W(h)ensday."). More
strikingly for the subsequent development of phonology, linguis-
tics, and anthropology, Sapir compared hypothetical inventories
of sounds for whole languages. Two languages might have identi-
cal inventories of sounds, according to observations of physical
properties; yet when the functional relations among the sounds
within the system of the language were considered, the two lan-
guages might be found to have quite different patterns, con-
figurations, or structures of elements. In each, for example,
one might hear both "p" and "b ," "t" and "d," "k" and "g."
In the one, such a minimal difference might' be functional, serv-
ing to distinguish words by itself. A word beginning with (p)
would be a different word from one otherwise the same but be-
ginning with (b), and so on. In the second language, the differ-
ence between the two types of sound might not be functionally
relevant. It might be a predictable alternation ((b) perhaps
occurring always between vowels, and (p) never). In the second
language there would be, from the standpoint of functional rele-
vance, just one series of stops that could best be written /p t k/
(since it is the "voiced" sounds, b, d, g, that are predictable
from their environment). In the first language, there would be
two series, /p t k/ and /b d g/. To repeat, the difference be-
tween the languages would lie, not in the presence or absence of
observed sounds, but in the status of the observed sounds within
the system of the language. And the principle that determines
the status is qualitative, an all-or-nothing principle that leads to
invariant, fixed reference points. From this perspective, there
is not such a thing as more or less of such a unit. There is
rigor in the work, and a branch of formal scientific inquiry to
which to appeal, but it is qualitative and discrete mathematics,
not statistics or experimental measurement.
Sapir went on to complete the picture by considering two in-
ventories that were different as observed sets of sounds but
identical, once analyzed in terms of functional relevance, as ele-
ments within a system of mutually contrasting points in a pattern.
A briM illustration: one language might have (p t k) as stops,
all functionally relevant. A second language might have (p t k)
and (b d g) as well, tut, like the second language in the pre-
vious example, no relevant distinction between the two series.
In sum, there would be three observed stops in the one language,
six in the other, but just three systemically relevant units in
each.
Some linguists resisted the development of phonology, feeling
that it began to leave behind the concrete realities of the sounds
of language. For to the principle of contrastive relevance (often
called the principle of communication), was added concern with
 Language in Education 65

symmetry and simplicity of the systems disclosed, and concern

with elegant solutions to the sometimes complex consequences of
tension between the phonological and the other sectors of a lan-
guage. Sound patterns of languages are subject to the strains
of historical change and communicative specialization, to the some-
times contrary pushes and pulls of external adequacy + internal
economy, with grammatical and lexical considerations
taking priority. Logical models invite a Conception ofsometimes
a language
as a monolithic system, with the meaning at one end and sounds
or letters at the other, but history and comparative perspective
quickly show that a more adequate conception is one of languages
as composed of interconnected major sectors, somewhat like inter-
connected continental shelves whose occasional displacements can
create untidy interfaces. The interrelation between phonology
and the rest of a language is often one such untidy interface. In
some languages the interrelation can be specified with few detours
and only occasional mounds and valleys, whereas in others it is
rather as if a mountain range had been thrown up.
II

The point of this extended analogy is that language is not as

neat as linguists sometimes make it out to be. Herein lies the
limitation of the lesson from linguistics. Any consideration of
qualitative methodology in the study of human life must take into
account the success of linguistics in establishing a domain of
study, central to human life, that has a methodology that is at
once qualitative and rigorous. But our consideration cannot
leave matters there. The student armed with qualitative method-
ology can be just as a priori in assumption, just as prone to
overlook disquieting empirical facts, just as heavy-handed in the
service of his methodological god as can the quantitative researcher
of fabled evil. In short, the success of linguistics is often ap-
pealed to, and rightly so, as evidence that quantitative method-
ologies are not sufficient, not the only model of rigorous science,
in the human sphere. That lesson is a crucial lesson. There is
a tendency perhaps for the sophisticated statistician or sensitive
experimentalist to believe that all methodologies ultimately reduce
to his. Qualitative insight and observation may be given great
scope, mindless counting depored, but still the belief is that the
final test comes with the quantitative or experimental design.
This belief is unfounded, and linguistics shows it to be so. At
the same time it is essential not to fall into the trap of believing
that the foundations of linguistics as presently practiced are ade-
quate and secure, such that quantitative measurement and experi-
mental design can only complement and come after findings ob-
tained by other means. This belief is the inverse of the other,
and it is prevalent in American linguistics. It is equally un-
founded.
66 Qualitative/Quantitative Research Methodologies
The history of psycholinguistic research since the early 1940s
shows the truth of the matter. At first in the study of the ac-
quisition of language it was necessary to learn what in fact was
acquired. Gradually psychologists interested in child development
and language acquisition became -knowledgeable about- phonological--
and morphological units. The work of Roger Brown and numerous
associates is noteworthy here. Once the rule-governed nature of
language was utilized in planning research, one could investigate
the presence or absence, and the stages of acquisition, of the
specific rules. One could go beyond gross measures of length of
utterance and the like to specific properties of the system con-
cerned. It is rather like being able to go beyond comparisons of
motors in terms of external properties such as size or color, to
analyses of engines in terms of internal properties such as com-
bustion pressure or piston rate.
When the course of modern linguistics reached syntax (having
started out, as we have seen, with a focus on phonology), and
when controversy over models of syntax was resolved effectively
in favor of transformational-generative grammar, begot by Chomsky
out of Harris, it seemed to some psycholinguists that almost a
millenium was at hand. George Miller, who had been prominently
associated with the development of information theory, became a
convert--a nice example of a conversion from a quantitative to a
qualitative "paradigm" (although not with loss of experimental de-
sign). Experiments based on the Chomskian model gave initially
exciting results. It seemed that the grammatical model and psycho-
logical reality were twins, and the job of psychology was to devise
ingenious experiments on the basis of the linguistic model. A few
years later the bloom was off that particular rose. The relation
between psychological reality (the mechanisms of the mind) and
grammatical theory (the mechanisms of a model of grammar) came
to seem increasingly remote. Indeed, a number of psychologists
have come to the conclusion that experimental analysis of relations
between linguistic elements is itself a primary source of knowledge.
A certain command of linguistics is required in order to deal with
the units of language, but where relationships among the units is
in question in terms of alternative models, experimental study need
not wait upon the linguist. To be sure, Chomsky has consistently
maintained that other kinds of study concerned with language
must wait upon the outcome of his. But that contention is in-
creasingly ignored. We see some of the productive outcome of
such independence in studies discussed by Cazden (1977) and
Shuy (1977), and the use of such terms as "ecology of language,"
"functional linguistics," and "communicative competence."
In this history, there is a second methodological lesson from
linguistics. It has to do with validity as much as does the first.
The- -first lesson has to do with validity in the sense of structure.
the second has to do with validity in the sense of function. The
two are indeed interrelated. Wh Sapir showed with regard to
phonology was that recognition of structure depended upon
Language in Education 67

recognition of functional relevance. The acutest ear, the most

careful design, could not take the fundamental first step in the
analysis of sound patterns in language, so long as the presence
of pattern was not understood to depend upon the lin "st's ver-
sion of experimental control, the test of commutation.
The fundamental elements of a system were determined in
terms, not of the relationship of sound to sound alone, but in
terms of the relationship of sound to sound in the service of dis-
tinguishing units of another level (words, sentences). And in-
ternal analysis of the relationships among such elements might
result in patterns that were rather different than observable
patterns.
Linguists both learned and neglected this lesson in subse-
quent stages of their discipline. They learned it for phonology,
as against phonetics, and for morphology, but many resisted it
for a time when it became an issue with regard to syntax. The
structure of sentences was studied in terms of similarities in the
distribution of elements within sentences of the same type. Sen-
tences such as "John is easy to please" and "John is eager to
please" would be seen as sentences of the same type, and "easy"
and "eager" as words of the same type. They contrasted as
words, of course (another instance of form/meaning covariation),
but not in terms of grammatical function. Chomsky's view,
crudely put, was that fundamental syntactic structure depended
upon recognition of functional relevance at a further level. This
level was discernible when sentences of a different type and dif-
ferent in overt pattern were seen to be related, sharing invariant
sets of grammatical functions and derivable from one another, or
from a common base, by regular rules, and seemingly similar sen-
tences and words to be different by the same token. "John is
easy to please," "It is easy to please John," "Pleasing John is
easy ," show a common core of meaning and functional relationship
among the elements "John," "please," and "easy." And "John is
easy to please" no longer appears the simple analogue of "John is
eager to please ," when the same commutation test across that set
of sentence types yields unacceptable sequences, *"It is eager to
please John ," *"Pleasing John is eager." (The asterisk marks the
unacceptable sequences.)
It is fair to see here a parallel to the lesson Sapir taught in
"Sound Patterns of Language." A major characteristic of the syn-
tactic work inspired by Chomsky was that seeming diversities
among sentences were found to have an underlying unity, and
seeming likenesses an underlying difference.
Having established this lesson in syntax, Chomsky was to be
confronted by students who insisted on applying it again in se-
mantics. Syntactic relationships that were clear and distinct
according to his model came to seem not so to them, when viewed
from the standpoint of semantic relationships. The dispute be-
tween those insisting on the primacy of syntactic relationships and
those insisting on the primacy of semantic relations continues.
68 Qualitative/Quantitative Research Methodologies
And Chomsky, having established syntax to his own satisfaction
as the core of language, insisted that studies of use, of styles
and such, was as dependent on the results of syntax, as any
other study of language. But this is a partial truth. To be
sure, as Cazden points out (quoting Crystal), one must attend
to the specific units of language or one will not see any relation-
ships at all (just as ignorance of the speech sounds of a foreign
language will yield a sense of noise, not of phonology). But the
relationships that are there will not all come into view if one stays
at a given level. Each functional sector or level of language
organizes units in a way not given by the units themselves. To
use an old example of mine, the functional category of greetings
may range. from single morphemes to complex sentences, from "Hi"
to "Well, be a son of a gun, if it isn't Sid Mintz" (Byrnes
1964). Nothing in syntactic analysis itself would bring these two
together. One has to start with the category of greeting itself,
and discover what elements and relationships among elements may
serve it. Shuy's studies of functional language illustrate this
principle in their examples of alternative ways to accomplish re-
quests, directions, instructions, and the like.
We are almost to ethnography now, but not quite. In their
recent papers on assessing language development both Cazden
(1977) and Shuy (1977) point out the need for ethnography, im-
plicitly at least. Cazden asks, how does one decide what com-
munication functions are of the most worth, and where does the
list of communicative competencies end? If an answer is not to
be imposed a priori, ethnographic inquiry into the communicative
repertoire of a community is essential. Again, Shuy, using ex-
tended observation and videotaping in a school setting, can recog-
nize functions and probe them experimentally because of conso-
nance with his own cultural knowledge. I shall try to indicate
the character of a fully ethnographic approach below, but first
let me finish the path begun with linguistics.
The path so far described for the course of linguistic method-
ology is step-wise. A level of functional relevance is recognized,
the step of structure dependent upon it analyzed; then something
of a kick and a leap must occur to move the field as a whole to
the next step, so easily does the student of language become im-
mersed in familiar form. The leap now before the field, though
continuous with the rest, produces in many a sense of falling out-
side linguistics itself. It is the leap to the study of the relation-
ships among linguistic elements in the service of speech styles.''
From the standpoint of a speech style, one has to do not with
an additional level of language, possessing an additional set of
units, parallel to phonemes, morphemes, syntactic constructions,
semantic features. True, the term "styleme" has been used, but
it can really only refer to units (or co-occurring units) already
identified and seen, from the standpoint of style, to be character-
istic or expressive. With styles, one has to do with a novel
organization of units at perhaps all the standard levels. What
 Language in Education 69
distinguishes a formal style, say, from an informal style, may
have to do with pronunciation, choice of words, choice of syn-
tactic construction, and preferred and inadmissible meanings. A
style is more a configuration than a level. And the elements of
a style may differ in scope from those of levels such as pho-
nology and syntax. It is possible to give a sense of an "archaic"
or "archaicizing" style (seriously or humorously) by occasional
use of a few salient features--say, a "thee" and a "thou," a
"natheless" and a "howbeit ," a syntactic inversion or two. The
rest of what occurs may be indifferent.
In sum, the difference of a style, as a configuration, from a
structural level of language is this. In phonology one has to do
with elements and relationships that are exhaustive of sentences
in one of their aspects. All of a sentence (or a discourse) can
be represented as a sequence of phonological units, mapped in
terms of phonological units. The same is true for morphological
units, syntactic units, and semantic units. Indeed, when lin-
guists speak of their subject matter as having to do with the re-
lation between sound and meaning, with the mapping of the inter-
vening structures, it is the exhaustive kind of level that is
thought of as intervening. There is "total accountability," so to
speak, for the linguistically relevant features at each level.
Styles need not be like that. To be sure, they can be. In the
Yana language spoken by Ishi, of whom Theodora Kroeber has
written so well, men's speech and women's speech were distin-
guished in the phonological ending of every word. But the
differences between men's and women's speech styles generally
in American society are not evident in every word. Such gender-
linked styles are indeed superb evidence of the need for a func-
tional starting point. They entail differences that appear only
when one sets out to discover them, starting from men and
women, rather than from grammar.
We have, then, to do with language in which traits may
constitute the relevant difference. And while some differences
among styles may depend upon presence or absence, be all or
nothing contrasts, others depend upon proportions and frequen-
cies. (Shuy (1977) discusses some of these cases.) We recog-
nize such phenomena when we speak loosely of someone having a
"touch" of an accent or of someone having a "thick" accent, or of
a high proportion of features at one end of the scale as "deep."
We have also to do with language in a respect in which it is
inescapably sensitive to situation. Progress in linguistics has
mainly been independent of social context, becuase it could be
assumed that the features being analyzed were common to all
users and uses of a language. That assumption is never wholly
correct, and the relation between what is analyzed by a linguist
as "English" and what you or I can say and understand may be
very problematic. The fundamental point here is that when we
reach consideration of style, we inevitably reach consideration of
styles. Even when a speaker of a language can be thought of as
70 Qualitative/Quantitative Research Methodologies
having a single grammar, he or she cannot be thought of as hav-
ing a single style. When we reach consideration of styles, we
must consider speakers as having, not a grammar, but a verbal
repertoire. In some cases that repertoire may comprise more than
one language. In every case the consistent continuation of the
principle of functional relevance leads to the questions--What are
the differences by which the styles in a speaker's repertoire can
be described as contrasting? What are the dimensions underlying
those differences? (What are the relations between the styles and
their occasions of use?)
We have reached, in effect, a study of language that is in-
separable from a study of social life, and in which quantitative
differences are inseparable from qualitative effects.
Ill

Many linguists may say that such a study of language is not

linguistics, but some other field, perhaps anthropology, psy-
chology, sociology. Whatever its label, it is beginning to emerge
into prominence, and it is the sort of study of language that is
fundamental to education. From one standpoint, such a study of
language may be "applied" linguistics, especially if it is concerned
with language use in s....thools. But "applied" taken alone is a mis-
nomer. Linguists do not now know enough about these phe-
nomena for others to come to them to ask simply for application
of knowledge already in hand. Research into these questions is
not applied, but is foundational and at the frontiers of linguistics.
Its practical relevance is obvious, but it is no less concerned with
issues of theory for that. The plain fact is that practical needs
and theoretical challenges coincide here, as they do in so many
other places. And there are not enough who are taking them up.
In this regard we pay a price for the isolation of linguistic
and educational research from each other, for pc irization between
qualitative and quantitative methodologies, for the lack of a suf-
ficient cadre of linguistic ethnographers. Perhaps we must always
-pay this price. Perhaps the values that are iintitutionalized in
our academic disciplines, our institutions and government, and our
culture, are such as to prevent the growth of the work that is
needed. It is so easy for modes of work to become frozen in
doctrinaire niches. Language is a subject beset by prejudice
and preformed opinion. In attempting to change the way in which
it is studied and understood, one may unwittingly be challenging
deep-set assumptions of the society. Perhaps language develop-
ment is assessed as it is today, for the most part, because to do
so supports the present order of things. Perhaps the vested
interest of an elite in the notion that change and the masses cor-
rupt language, and the vested interest of a highly stratified,
bureaucratized social order with a democratic frosting, such that
individuals must be considered to have ended up where they do

su
Language in Education 71

as a result of their own doing, converge. Just as we would not

know what to do if schools failed to keep millions of young
people out of the job market, so we would not know what to do
if schools succeeded in producing millions of young people with
the language competence they take as an ideal. Or rather, per-
haps we would indeed know what to do. If accents and dialects
and vernaculars were to disappear and no longer be available as
ways of discriminating, if everyone spoke standard English, we
might simply substitute a finer lens in the microscope of correct-
ness. No more "he do" for "he does?" But plenty of "transpire"
for "occur," confusion of "1.:ifer" and "imply," jarring plain "im-
pact" as a verb instead of "have an impact on." After all, a
great many of the distinctions upon which we now insist came into
existence only as a result of the rise of the middle classes in the
eighteenth and nineteenth centuries. If new ones are needed, or
more weight need be given to old ones in order to maintain the
desired level of linguistic insecurity in the populace, the neces-
sary sense that most people do not deserve better because of
linguistic inadequaay, it could surely be done (cf. pp. 104-118
in this volume).
All this is speculative, of course. If it is too dramatic, my
excuse must be that there seems to me an issue here that con-
stantly eludes us, and must somehow be forced into attention.
Before we can make satisfactory contributions to the assessment
of language development, we must know far more than we do
about the role of the assessment of language development in the
history of American schooling and American society. Fortunately,
a few scholars are beginning to pioneer in research in the his-
tory of our language attitudes and policies (e. g. , Shirley Brice
Heath, Glendon Drake). The histories I have seen give little
attention to it. A broad picture is clear enough: wipe out the;
Indian languages, erase linguistic differences due to immigrant
origin, disvalue or stereotype dialect, insist on a, single standard
as a badge of intellectual and personal virtue. Little seems to be
known about the formation of these views in schools of education,
their implementation in schools and school districts, the tcnsions,
interactions, and adjustments in specific regions, where specific
configurations of linguistic difference and verbal repertoire pre-
vailed. One senses a pervasive difference in attitude today be-
tween groups differently situated in the class structure, a pattern
of difference perhaps between the Eastern seaboard and the West,
but without adequate documentation.
Perhaps we need to step back, imaginatively, in a way analo-
gous to "zero-budgeting." What if there were no assessment of
language development at all? Would anything be lost? Most of
mankind in time and space, after all, has not had explicit assess-
ment of language development of the sort with which we are con-
cerned. Why do we? Why do we have to? How did it start?
What functions, latent as well as manifest, does it serve? Is it
possible that language development of children in our schools
72 Qualitative/Quantitative Research Methodologies
would improve if it were not assessed? To what extent are the
functions and effects of assessment different for different chil-
dren, according to region, class, sex, program?
Some essential light can be shed by knowledge of assessment
of language development as it occurs outside the classroom. In-
sofar as parents and communities have not come to accept the
figures of formal tests as the only criterion of achievement, there
must remain at least residues of informal assessment according to
the norms of local cultures. One would want to know what kinds
of use of language are valued, which users of language are
valued, how these values are exhibited, experienced, and ac-
quired. One would want to know how the relationship between
language use in school and language use outside of school is
viewed, where there is continuity, where conflict, where com-
partmentalization.
Research outside schools could discover evidence of abilities
in community situations that might put the display of abilities in
schools and test situations into perspective. It is well known
that display of abilities may be tied to situations. Sometimes it
is a matter of appropriate content, sometimes a matter of appro-
priate norms of interaction. In their study of Kpelle children,
Cole and Gay found that children who had difficulty in school
with lessons in mathematics that dealt with certain principles
could be seen to employ these same principles in certain work
situations. Susan Philips has documented the cultural pattern
that underlies the "shyness" of Indian children from Warm
Springs reservation, when directly questioned by teachers. The
teachers perceive the children often simply as "shy," not talking.
Observation of the children in play and home situations shows
that they can be very talkative indeed.
Here may be an opportunity to unite research with effective
change. One might start with what the teachers and the school
personnel perceived as the problems associated with language
development of a group of children. Given this definition of the
problem, as perceived in the school situation, one could under-
take a study of the language activity of the children in the full
round of their lives,. putting the phenomena of the classroom into
perspective as part of that round. One could involve the teach-
ers as participants in the ethnographic study to a certain extent,
or at least as participants in an on-going seminar in which the
ethnographic inquiry was regularly reviewed and discussed. The
results of the study, of course, might confirm the teachers'
initial perception. If the results indicated a different interpre-
tation of the children's language activity in school, the teachers
would not he confronted with it cold. They would have par-
ticipated in its development, and understand the process by
which it was reached. Such involvement in the process might
make acceptance of the product more likely.
 Language in Education 73

IV
Let me now take up a conception of ethnography, in order to
make clear what the term implies for me, as method and discipline
that can be vital for educational research in language.
One hears the word "ethnography" more often these days in
educational and linguistic circles, and one also begins to hear the
question, "What is ethnography?" There is no single answer.
Almost anything that involves direct contact with people as a
source of information may find itself included under the label,
especially if the contact is made by an anthropologist.
ception I sketch here is shared by some but not by all. The con-
It is
intimately connected with the sketch of the development of lin-
guistic methodology given above.
It is important to distinguish between ":ethnography" and "field
work." There are two related reasons for this. First, "field
work" is a suitable general term for any contact with people as
sources of information; second, and most important, not all "field
work" in this sense is "ethnography" in the sense I 'intend.
There are again two reasons.
First, contact, having been there, is not enough. Sometimes
the claims of anthropologists to a distinct and even superior
methodology embody an element of "I was there." This ought to
give us pause when we reflect on how many people we met there
whose views of the people we studied we would not trust. Those
of us who work with Native American communities often enough
meet the local resident who "knows all about the Indians" from
having lived there. We often discover that what is knoirn is
limited to what Indians were willing to disclose in certain, partial
contexts, or is colored and constrained by an economic c r social
relationship that closes off certain kinds of. knowledge as uncom-
fortable.
If the anthropological methodology in field work is effective,
wit is based on more than being there, however romantic some of
us may make the field work experience sound. It may, to be
sure, be based on insights and intuitions, but these are nour-
ished and controlled by a certain kind of training. It is this
training that is c;ften missed by those who have not had it, giv-
ing rise to an equation of ethnography with sheer presence in
the field. Indeed, a focus on field work may easily miss the
training, because the training commonly occurs apart from the
field. It has to do with the systematic, comparative knowledge
of phenomena and systems like those under study which the
ethnographer brings to the description and interpretation of the
particular case. It has to do with the knowledge that enables him
or her to recognize in a funny use of words for "aunt" and
"uncle" a kinship system of an Omaha, or Crow, or other type;
in a problem of attendance at schools or jobs the persistence of a
seasonal round; in a consistent failure to say "Thank you," not
 Qualitative/Quantitative Research Methodologies
ingratitude, but a pattern of reciprocity that avoids closure in
single situations.
Such inferences presuppose the skills to obtain the infor-
mation from which they are inferred, and these too entail more
than presence and observation. These skills involve the ques-
tions asked in the mind, if not in speech, that guide presence
and observation. This brings us to the second way in which I
should like to restrict the term "ethnography." I should like to
give "ethnography" the connotation of inquiry that is open to
questions and answers not foreseen, for which possible observa-
tions need not be precoded, and for which the test of validity
need not fit within a prestructured model. When anthropologists
limit their inquiry to observations and questions for which the
set of alternative answers is already fixed, I should like to say
that that may be field work, but not ethn, graphy.
These two poles of validation through field work, then--"I
was there" contact, precoded content--represent a Scylla and
Charybdis between which true ethnography steers. The steering
is not reducible to a routine; that is what makes it hard, and
seemingly ineffable at times. Sensitive awareness, empathy, and
intuition are not ruled out, far from it, but merely not enough.
Pre-existing models and frameworks are inseparable from the
requisite training, but one must be able to get beyond them. It
all comes down, unfortunately, to being attentive and smart. Be-
ing there won't allow one to sop it up; methodology won't allow
one to grind it out. The steering is indeed cybernetic, a matter
of feedback, of dialectical interplay, if you will. It has to be so,
because of thekinds of situations in which one works and the
kinds of knowledge one seeks. One works in situations which re-
quire the trust of others, accommodation to their activities, parti-
cipation in ways that often preclude writing or recording at the
time. One sacrifices certain kinds of reliability for the validity
that one hopes and often finds to come with depth. In a sense,
one is half in the position of a child or newcomer learning the
ways of the community. One has not the time and full immersion
of the child or newcomer, but a measure of orientation through
training that compensates. Just so in acquiring a local language:
the limited opportunity to hear and use it is compensated for by a
training in what to listen for and what to do with what one hears.
In a word, ethnography is inquiry that begins with recogni-
tion that one is at work in situations that are, indeed, massively
prestructured, but prestructureU by the history and ways of
those among whom one inquires. At the heart of it is a process
of which linguistic inquiry is indeed a model, if we set aside any
particular model of grammar, and think of linguistic inquiry in
the generic sense as the interpretation of codes.
For the ethnographer in this sense, the world of inquiry is
neither merely a source of raw data for general schemes, nor a
gallery of essences that can only be intuited and expressively
 Language in Education 75

expressed. It is a world of many codes, of many structures.

Not a single natural world, indeed, but a plurality of worlds
(Lebenswelt); worlds that are constituted in the lives and experi-
ence of participants in a group or activity, in important part
through selecting and grouping and reinterpreting received tra-
ditions, traditions which from the point of view of other tradi7
tions may seem unintelligible or irrational. From the standpoint
of a merely universalizing or generalizing science, such traditions
and worlds may seem arbitrary and parochial. Yet even a science
that wishes to rise from human worlds to something
if it wishes to effect change, must reach into these called "Man,"
worlds, be
mediated by them, if change is to be consonant with intention.
From the standpoint of a science dedicated to generalization and
universals, the specifics of each world may seem simply boundary
conditions, specific constants and ranges to which the parameters
of general theory must be adjusted. From the standpoint of a
science imbued strongly with a historical sensitivity, the specifics
may contain qualities of emergence. To the one -view, qualities
that are rare or unique may seem something that can be set aside
because of their infrequency. To the other view they may seem
opportunities for insight, configurations that disclose hitherto un-
realized and unsuspected potentialities. To take an example from
language: a general theory of language can regard a specific lan-
guage as an exemplification, and perhaps test, of features of
the design of language in general. Some awould hold that only
universal considerations are important. Others would regard a
specific language from a typological point of view remaining con-
cerned with universal language design, but concerned as well
with the recurrence of major
ing to reveal potentialities of types of structure, themselves seem-
language structure that recur inde-
pendently of history, and that are not easily reducible to a single
model, if the model is at all rich in content. Still others, myself
included, accepting and valuing the preceding interests, would
want to keep in view a third concern. Navajo is what it is be-
cause it is an instance of human language; it is an instance of
certain types of language structure that have great interest; it
is also what it is because it is the language of the Navajo. To a
great extent its structures are what they are because of possi-
bilities and impossibilities inherent in
by the mind. Its flesh and blood, as language structures mediated
it were, the meanings it has
for those who use it, the texture that it takes and gives in their
speech and reflection, are what they are because of the specific
experience of those who have spoken and continue to speak it.
What role the language can play in the modern world, in schools,
is to be understood in terms of that history, valuation, and out-
look. Linguists long ago, in what must seem the ancient time of
Boas and Sapir, established that there is nothing intrinsic to the
structure of any language that precludes its adaptation and
elaboration to serve new needs of whatever kind. If there are
limitations and disabilities, when a language is confronted with new

5
76 Qualitative/Quantitative Research Methodologies
circumstances, these reflect nothing about its potentiality; they
reflect the fact that its vocabulary and idiom, its conventional
speech acts, routines, and genres, the assumptions as to eti-
quette of speech, have evolved and been embedded in a certain
way of life. (As is true, of course, of any language, even those
which become world languages through their adaptation to the
needs of commerce and science, and their association with world
powers. Their near universal currency is a demographic, politi-
cal, and cultural fact, not one due to any unique structural
property.)
It follows that there can be field work with a language, even
field work devoted to applied goals, educational goals, that falls
short of ethnography. One can devise an orthography to permit
the use of a language in primers and bilingual education, but
knowledge of the role of writing and reading, of language in
visual form, is needed, if -the written material is to be used. In-
deed, knowledge of the meanings associated with the alphabets
and letters known to the people is necessary, if the alphabet it-
self is to be successful. One cannot make general assumptions
about the role of literacy. One has to find out through ethnog-
raphy what it means in the case in hand. In general, it is not
enough to decipher the code of the language itself; one has to
decipher the codes associated with the use of the language, as
an element in the verbal repertoire of the community.
A brief citation from Levi-Strauss on this role of ethnography.
He once observed that if an object of art came to Paris, and the
code was known, it would go to the Louvre, but if the code was
not known, it would go to the Musee de l'Homme (the ethnological
museum).
A further part of this view of ethnography as inquiry into
worlds is a view of these worlds as inherently adapting and chang-
ing, recreated and reinterpreted by individuals in their own lives
and in relation to the experience of the group as a whole. From
this standpoint, one taken initially in American anthropology, so
far as I know, by Sapir in his writings of the 1930s, a cultural
world has not been accounted for if treated in terms of its condi-
tions alone. 5 One such condition, a major concern in the develop-
ment of anthropology, is of course historical provenience and
transmission, cultures as interesting and distinctive wholes that
persons acquire, manifest, transmit, but as their locus rather than
their source. From this point of view, what is cultural tends to
be equated with what is in fact common or shared. From Sapirts
.point of view,. the fundamental nature of the cultural is that it is
capable of being shared, that is, of being communicated. A shara-
ble symbolic trait, something capable of becoming more generally
part of a group's repertoire of codes, is already within the sphere
of the cultural.' Such a point of view is necessary to cope anthro-
pologically with the modern world, where the overt signs and dia-
critics of cultural traditions float, jo-itle, and merge as if each
city were an eddy, left behind by a flood that swept all detachable
Language in Education 77

bits of culture about the world. Sometimes anthropologists :lave

been able to see an object of study only in cultural worlds like
those sketched in their textbooks, saliently distinct. Indians
driving new cars, buying cases of pop, watching color television
--where's the cultural world there? There is one, and one not so
wholly like the non-Indian as appears, but it may go unwitnessed,
if we shake our heads and mourn that the god of cultures has
long ago finished his task of creation. When it comes to cultural
worlds, the seventh day will never come. (This theme is de-
veloped in Hymes 1973; see esp. p. 34.)
Such a view, restricting anthropology to the "other" cultures
that historically fell to its lot in the great handing out of subject
matters a century ago, seems to rest on the assumption that iso-
lation, or at least strong barrier, is necessary for the flow of
culture to acquire distinctive form, and that this requisite is in-
creasingly absent. There is truth in the assumption, but so to
interpret it involves an inadequate conception of the nature of
boundaries. On the one hand, the salient boundaries, marked by
a language, a geographical barrier, a political line, have proven
very permeable. On the other hand, the sense of a distinct cul-
tural world depends ultimately upon taking something as a bound-
ary. It is a function of self-definition, identification, of meaning
given to whatever differences may obtain. The differences may
be few in number, may be less in physical and observable traits
than in configuration, or simply in shared experience and what
Raymond Williams has called "structure of feeling." Such a basis
for boundary may have been more important than usual zealizP-i
in the cultures traditionally studied by anthropologists.' Such L.
basis may be prevalent within a conglomerate social structure such
as our own. Perhaps each of us moves too much in a round of
activities and people that matches our own conception of a world
to appreciate the diversity about us. We can share a city such
as San Francisco or Philadelphia with hundreds of thousands, and
meet only professional colleagues. But if the paths of each were
traced, and the meanings glinting on either side gleaned and
understood, a multitude of distinct worlds might become evident.
Beyond our own rounds, and the spheres defined as public prob-
lems by media, perhaps lie a great many worlds unmentioned and
out of sight.
If this is so, then there is plenty of work for ethnography,
and work that only ethnographers do (though the assistance of
novelists is to be welcomed). The existence and character of
these worlds, their bearing on schools and education, can become
known only through participation.
Let me give an example. The Warm Springs Reservation,
where my wife and I work with speakers of two Indian languages,
has been a distinct political entity 1::3r more than a century. Al-
though three different peoples were brought together on it, and
a lively awareness of the original tribal affiliations persists, re-
inforced by financial considerations in treatysettlements,.much
78 Qualitative/Quantitative Research Methodologies
intermarriage and interaction has resulted in a common sense of
membership in the "Confederated Tribes." A certain amount ht.s
been learned and recorded about the aboriginal culture of thu
people, and about its persistent elements, some of which--one of
the languages, certain rituals; certain patterns of activity, etc.-
remain today. A certain amount of attention has been focused on
the Reservation as an example of Indian people seeking economic
self-determination and self-sufficiency. The policies adopted in
this regard, the various activities, the consequences, can be
known. Various other aspects of life, such as housing and health,
attract attention because of the involvement of governmental
agencies. So far as I know, no one has addressed the question,
what is it like to grow up and live at Warm Springs Reservation?
How does the world appear? What is that world like? It is a
world with color television, suburban-style housing developments,
a resort hotel catering to whites, a.golf course, an organized his-,
torical society, a new administration building, etc. It is also a
world in which a good many of the brightest people become alco-
holics, or so it seems to us; in whit it bright and motivated chil-
dren often leave or are forced from high school, marry, get preg-
nant, go to work, whatever; a world in which some young people
die every year in auto accidents; in which a major bulwark of the
social fabric continues Zo be a number of responsive, responsible
grandmothers; in which the security of assured shares of tribal
income interacts somehow with severely limited opportunities for
work and hopes for responsibility or authority; it is a world
whose every member must at some point decide for himself or her-
self what it means to be an Indian, because there is no way to
avoid or deny the identity.
One anthropologist appears to feel that there is not much more
than tidbits to be gleaned there--the old people who "knew the
culture" are almost all gone. Whether or not there is still at Warm
Springs a "culture" in some of the older ethnological uses of the
term, there is a cultural world. It is a world not wholly or
analytically understood by its members, who have as categories
of understanding mostly only either traditional ones or ones sup-
plied by external institutions and the surrounding rural white
society. What happens to children in sc'o.00ls appears to depend
on how the children interpret their world, given such categories
as they have available. To find out what they see and do, to
convey that knowledge in a way that permitted some of the texture
of their lives and world to come through, would be what I mean by
ethnograpiay.
The level of cultural worlds completes the chain of levels
within which structure is to be discerned through functional
relevance.
Let me try to show how this is so, and in so doing link the
discussion of linguistic methodology with the discussion of ethnog-
raphy.
Language in Education 79

V
The n athodology at the basis of modern linguistics, as has
been said, depends upon the notion of commutation, of demon-
stration of functional relevance through contrast (as against
repetition), showing that a particular change or substitution or
choice counts as a difference within a larger frame of reference.
This methodological principle should be taken into account in any
gon.eral discussion of qualitative methodology, and it is capable
of extension beyond what has been made of it in linguistics
proper. There are indeed two important kinds of extension to
be encouraged by those';apncerned with language as part of
social life. One has to do with the basis of linguistic structure,
the other with the building of it.
Our modern edifice of language structure has employed the
principle of contrastive relevance primarily in the service of cog-
nitive functions, what can be rather simplistically called "refer-
ence." The correlative notions of contrast and repetition have
been used to establish features that enter into the kinds of mean-
ing involved in naming, statement, logical claims, and that illumi-
nate relations of grammatical structure in the service of such
kinds of meaning. To be sure, there has always been some
attention to features and meanings that can be called "stylistic"
or "expressive," but their domain has seemed marginal or second-
ary. In point of fact, the principle of contrastive relevance
applies to both kinds of meanings, and if it is a fundamental goal
of linguistic theory to explain what counts as repetition, what
counts as contrast, then expressive. or (as I prefer) "stylistic"
functions are equally fundamental to linguistic structure.
There is not space to deal with the ramifications of this fact,
but an illustration may show what is involved. The first, and
common, kind of contrastive relevance is illustrated by contrast
between /p/ and /b/ in English, such that "prattle" is something
a baby may do, and "B rattle" a nonhomonymous street in Cam-
bridge. The second kind is illustrated by contrast between a
heavily aspirated and a weakly aspirated /p/. In the first case a
difference in referential structure is conveyed, in the second a
difference in attitudinal structure: emphasis perhaps to make the
word clear, to express disgust or elation, whatever.
ration of a stop such as /p/, precisely because it doesHeavy aspi-
not serve
referential function, can serve stylistic function. It does so as a
conventional device available to speakers of English, a part of
their linguistic competence, Like the referential contrast between
/p/ and /b/, the stylistic contrast between heavily and lightly
aspirated /p / is diacritic. That is, it distinguishes meanings, it
does not embody them. The meaning conveyed depends upon fur-
ther features of the utterance. For this reason, I would refer
to the two, complementary bases of contrastive relevance as estab-
lishing two "elementary diacritic functions." (This point is
80 Qualitative/Quantitative Research Methodologies
elaborated in Hymes 1974a, ch. 8 and 1974b, and taken up in
Hymes 1972.)
The presence of the second, complementary function is im-
plicit in the difference between paradigmatic sets of sentences,
chosen to illustrate points of grammatical structure in the narrow
sense, and paradigmatic sets of sentences, chosen to illustrate
actual choices in the use of sentences in social life. An exti....ple
of the first comes from Postal (1974: 3):
a. I think that he is rich (that clause)
b. That he is rich is thought by me (?) (1st passive)
c. He is thought to be rich by me.-1k (2nd passive)
d. But not: *I think him to be richq (complement)
Leave aside the fact that Postal finds (d)- unacceptable, whereas
it seems perfectly natural to me, a slightly elegant or literary
mode of expression. Leave aside the fact that (c)., which Postal
finds acceptable, and which illustrates his grammatical argument,
seems odd to me, and that the only way I can mike it acceptable
is to introduce stylistic function in supPort, .:'so that one would
be saying (or hearing) a response: "He is thought to be rich by
me," following, perhaps, "No one ever Ihbught him to be rich"
(in my speech, not apparently in Postal's; cf. (d) above). The
main point is that when one asks, what are the alternsaive ways
by which one would express the notion (a) in conversation, people
do not ring changes on the grammatical paradigm exhibited above,
keeping other things constant, but change their utterances in a
variet7 of ...ays. They make use of choices in other sectors of
langUage, lexicon, intonatioii, other types of construction. Lexi-
cal optionts come readily to mind (rich, wealthy, loaded). What
people appear to be doing is to consider the reasons (functions)
for saying the thing differently, that is, they invoke possible
differences of situation, both verbal and social, and consequent
options of style.
One might refer to the kind of relations disclosed by the first
kind of contrastive relevance as having to do with reso.:-ce gram-
mar. The bare bones of grammatical possibilities, preserving
reference and neglecting style, are examined and collated. The
second kind of contrastive relevance brings to light paradigms of
a sort that might be called "natural conversational paradigms" (as
opposed to "analytical grammatical paradigms").
This kind of contrastive relevance has to do with what can
be called discourse grammar. It employs the recognition of
stylistic functions to extend linguistic inquiry beyond the usual
levels of language to the styles and choices involved in use of
language. Even in studies of literary style, the question of con-
trastive relevance, of w hether_ or not an observed feature repre-
sents a choice (for tli4 author, or for the reader), is fundamental.
(Vendler 1975 uses this prirt iple nicely, e.g., pp. 13-6, 22-3) .

9.)
Language in Education 81

The theoretical approach of Michael Halliday makes use of such

an approach in a particularly stimulating way, envisaging gram-
matical means (the discoveries of resource grammar) as organized
according to four generic types of function, ideational, inter-
personal, textual, and logical (see Halliday 1973). The conception
remains to be tested fully in English and across a variety of lan-
guages (I do not think, for example, that pronouns would be
found to o':;cupy ci.uite the same place in all languages). Much
current work by linguists and others, studying texts, conversa-
tion, speech acts, and associated properties of coherence and
conduct, does a great deal to explore language from essentially
thi- standpoint, the standpoint of alternatives in actual uses of
language. Extension at the base and extension at the top, as it
were, are not always integrated, unfortunately. Speech acts,
such as promises and threats, may be analyzed without regard for
the role of stylistic (and communicative) features that enter into
the-:.:t. meaning (cf. Hymes 1974a, ch. 9). Current studies of dis-
course, texts, conversion, speech acts, are doing a great deal
to explore these areas. Social interactional meanings are beginning
to receive their due.
We cannot ad: 4uately evaluate language development and the
uses of language that enter into education without attention to
both these extensions of the principle of contrastive relevance.
Properly pursued, they entail a general conception of language
development and use as a matter of meaningful devices. The
still common use of mean length of utterance as a measure of de-
velopment is not in keeping with this principle. The measure
may helpfully correlate with other things, but it can shed no
light on what is happening, what is being acquired and used.
Again, it is like comparing motors by their size instead of by
their structure. Language, from sound to style, is a complex of
form-meaning covariation. That is another way of putting the
point of contrastive relevance. To discover what is there, what
is happening, one seeks to discover which changes of form have
consequences for meaning, what choices of meaning lead to
changes of form. One works back and forth between form and
meaning in practice to discover the individual devices and the
codes of which they are part.
The limitation of linguistics proper has been that, despite the
potentiality of its methodological principle, it tends to stop short
of the full range of form-meaning covariation, and to stop short
of ethnography.
This is an old story. Modern linguistics advanced decisively
over the popular notion that one could tell something about the
character of a people by the presence or absence of individual
words ("They have no word for 'thank you'," "the Germans have
a word Schddenfreud," as if the absence of the word meant a
posture of ingratitude, the presence of the word a special delight
in the misfortunes of others). Franz Boas made central to his
linguistic work the question of the categories that were not
82 Qualitative/Quantitative Research Methodologies
merely present in a language, but obligatory, that is, unavoidably
involved in verbal expression. (Tense is such a category in Eng-
lish when verbs are used, number when nouns.) Benjamin Lee
Whorf proposed to go beyond registering the obligatory categories
to a study of their articulation with other features in actively em-
ployed "fashions of speaking," but the study of "fashions of
speaking," which would entail ethnographic inquiry into styles,
was not taken up. (This point is discussed in chapter 8 of
Hymes 1974.)
We see the same story today in studies that are called "prag-
matics" or "discourse" in linguistics. From a strictly linguistic
point of view, it is interesting to investigate how it is that a
question may be the answer to a question in certain types of
encounter--how to the query, "Do you have any coffee left?", the
answer may be, "Do you want cream?", presupposing a positive
but unspoken answer to the initial question. These elliptical se-
quences are characteristic of exchanges in stores where the dimen-
sions of the encounter are limited and mutually well known. An
examination of such encounters must necessarily involve field
work, that is, observation of actual cases, to obtain its 'data.
But if the analysis is limited to the consequences of such se-
quences for a theory of language organization, it is not ethnog-
raphy, but field work. A larger frame of reference of contras-
tive choice would be required. When are service encounters of
this type appropriate, when not? When are people insulted by
the restriction of an encounter to such an exchange? What does
it mean to an old store-owner in an ethnic neighborhood that the
new young sales representatives limit their interaction with him to
truncated exchanges of this type? What genre of verbal ex-
change has been replaced? What is the nature of the verbal
ability that now has no occasion? More generally, what is the
range of the truncated service encounter in the society in relation
to the full set of alternative types? And what are the common
styles? One has the impression that the American style is found
brusque to the point of insult in England, tine English style overly
polite to the point of .archness of effeminacy in the United States.
In sum, the full pursuit of form-meaning covariation would not
stop with consequences for linguistic structure.. It would discover
something of the resonance and consequence this instance of a
genre within cultural worlds.
This reasoning of course holds for speech acts and small genres
of all kinds, requests, commands, greetings, teasings, etc. Since
such study unavoidably engages phenomena in change, as well as
choices across a range of settings, quantitative information and
analysis is essential. One expects to find proportions and trends
as much as or more than categorical rules of appropriateness.
The principle of the linguistic ethnography that is needed can
be put in terms of complementary perspectives. If one starts
from social life in one's study, then the linguistic aspect of the
ethnography requires one to ask what are the communicative means,
Language in Education 83

verbal and other, by which this bit of social life is conducted

and interpreted? What is their mode of organization, from the
standpoint of repertoires of codes? Can one speak of appropri-
ate and inappropriate, better and worse uses of these means?
How are the skills entailed by the means acquired, and to whom
are they accessible? These questions lead one into the territory
of the other starting point. If one starts from language in one's
study, then the ethnography of the linguistic work requires one
to ask, who employs these verbal means, to when and
where and how? What organization do they have from the stand-
point of the patterns of social life?
vl
In a critique stressing the use of language and ethnographic
inquiry one should consider one's own uses of language as scholars
and scientists. The discussion has concentrated on qualitative
structures, with recognition of the relevance of quantitative
methods. In mentioning resonance, and using a word like "tex-
ture," one raises the question of narrative reporting as well. To
the best of my knowledge, some of what we learn and know and
should convey can only be expressed through skillful prose. It
is a commonplace in anthropology to admit, or enjoy, the fact that
novels about a country may be a valuable source of understanding.
In recent years a growing number of anthropologists have felt im-
pelled to write a narrative about their field work. Having pub-
lished the scholarly analysis, they write a second book to try to
say what it was really like. This seems to me a healthy impulse.
It has roots in the increasing concern with the reflexive nature
of social science inquiry, but is not to be reduced to that. Much
of what we know,, in anthropology and in personal life, is known
by means of narratives, anecdotes, first-hand reports, telling ob-
servations. In the vital decisions and directions of our lives we
willy-nilly rely on what we know by such means. In our scholarly
chairs we find 'it difficult to acknowledge their validity, though we
may admire their artfulness. There are many purposes and kinds
of verbal art, but some of it, I believe, is a way of getting at _the
truth. One can read poems for fun, sanctity, duty, or a liveli-
hood, but some poems one can read for what they enable one to
experience and know. If we are to extend our understanding of
language to the full, so that we can fully comprehend its role in
schooling, in education, in social life, in our own lives, we have
to find a way to come to terms with the validity of uses of lan-
guages that are aesthetic. Some people are brilliant at numbers
and research design, some excel in discovering and articulating
qualitative structure and pattern, and some are masters of the art
of conveying events and experience and insights in words. To
admit this is not to give way to rampant subjectivity. We can and
must discriminate, establish canons of judgment, make explicit our
criteria for trusting one set- of words, taking _mother under
 Qualitative/Quantitative Research Methodologies
advisement, and distrusting a third. It is a job of verbal criti-
cism and inquiry that has a great deal to contribute to the legiti-
mation of much that anthropologists believe they know through
ethnography.
These considerations bear on the final point to be marle about
the role of ethnography. Ethnography can of course be used for
many purposes, serve different interests. In my conception, its
validity is dependent in part on the knowledge already had of
their ways of life by those whose ways of life one seeks to study.
Behind every classic ethnography, I suspect, stands one or more
members of the culture who were themselves ethnographers with-
out portfolio. Wherever meaning in the third sense discussed
above, having to do with resonance and consequence, is success-
fully conveyed, one suspects a process of inquiry that was col-
laborative. Such a process is one to which the members of the
cultural world bring knowledge of its codes and experiences, and
to which the ethnographer (who may be a member) brings method-
ological skills and comparative perspective. A good part of the
knowledge held by members of the culture is necessarily tacit.
Their languages, their expressions and styles, are indispensable
sources of insight, but never in themselves a complete and ade-
quate metalanguage for their own world. One of the fundamental
questions of anthropology, indeed, or at least of linguistic ethnog-
raphy, has to do with the degree to which a given language is
an adequate metalanguage for the way of life of which it is part.
What concepts and meanings have found explicit linguistic shape
for reporting, discussion, reflection, and which not? And what
is the role of language as such, as a means of communication
more or perhaps less employed, /enjoyed? Some cultural worlds
are permeated with language, others not.
A consequence of this fact for ethnography is that native
documents and testimony, while indispensable, as Boas insisted,
are never sufficient. To a fair extent, subject to ethical choice
and judgment, the process of ethnography can be an exchange of
knowledge. Many linguistic informants have become fair analysts
of their own languages in the course of contributing indispensable
knowledge about it. The same can be true in ethnography
generally. In this passibility lies a possibility for a mode of
ethnography that is not exploitative and that contributes as well
as takes in the world in which it works. Obviously this possi-
bility is surrounded by many complications, not to be gone. into
here. But it is important to note it, especially when we are con-
cerned with ethnography in institutions of our own society, such
as schools. An ethnography that served only higher levels of
government, national institutions, and theory is hardly possible
in any case, as superintendents and principals are quick to tell
us. In the exploration of ways in which ethnographic inquiry in
education can be founded on mutuality, questions of language
themselves, of the sort considered just above, have a part to
play. I would like to think that some of what one learns and
9a
Language in Education 85

knows and has to report is inseparable from uses of language

that are continuous with those of ordinary life. These are the
narrative uses, the uses into which an aesthetic consideration of
apt expression enters. Cultivation and analysis of such uses may
contribute to mutuality between ethnographer and school. And it
may be a healthy thing for the democratic quality of our society
if such uses can be given the justification and legitimacy they
deserve. Indeed, such uses do play a vital part in decisions and
perceptions, so that we handicap our understanding of educational
institutions and the forces that affect them if we do not make
them explicit objects of attention. Our own .language development
is in need of assessment.
ENDNOTES
1. This paper was stimulated by participation in the Workshop
Exploring Qualitative/Quantitative Research Methodologies in Edu-
cation, held in Monterey, California, July, 1976, and sponsored by
the Far West Laboratory for Educational Research and Development,
in cooperation with the National Institute of Education and the
Council on Anthropology and Education. An abstract of the paper,
prepared by members of the Laboratory, appears under the title
"Critique" on pp. 91-3 of this Quarterly (vol. VIII, no. 2). I am
grateful to the editor of the Quarterly for finding space for the
paper, which proved too long for the issue devoted to the Work-
shop as such. I should also like to thank the reviewers of the
paper for the Quarterly, not all of whose cogent comments I have
been able to act upon, since the result might well have been an-
other whole issue. I hope that various references may help fill
out what may be too cursorily treated here.
2. John Dewey used the example of the expulsion of lip-
rounded breath to blow out a candle or to begin an English word,
before Sapir, but I have forgotten the reference. Sapir does not
mention Dewey. Perhaps he had forgotten too.
3. The logic is really much the same as in the discovery of
significant relationships through the assignment of subjects to
experimental conditions. It is just that in the core of language,
viewed as a referential mechanism, the conditions sort out answers
of a "yes" : "no", all or nothing sort, rather than of a "more"
"less" sort. A typical hypothesis, in terms of the simplified
illustration previously given, would be that a given sound is (or
is not) independently relevant. Consider (b) occurring between
vowels. The hypothesis is tested by substitution (commutation).
If (b) cannot be replaced by (p), then in this position it cannot
contrast with (p); the difference between them cannot distinguish
words in the language in question, and the similarity would lead
one to group the two together as members of the same systemic
unit. The difference would be readily explainable in terms of
local conditions and general theory. Voiceless sounds, such as
(p), often become voiced (as is b), when occurring between
86 -Qualitative/Quantitative Research Methodologies
voiced sounds (such as vowels). The same inference and expla-
nation would hold if (b) could be replaced by (p) between vowels,
but the replacement had no concomitant (correlated) difference in
meaning. (The commutation test is an application of the general
principle of form /meaning covariation.) Of course if replacement
by (p) resulted in a different word, the two sounds would belong
to different units. The concomitant change would demonstrate
that one had to do with not one, but two, elements of the minimal
arbitrary code of the language.
4. On this conception of a development, see Hymes 1974,
ch. 8. Of course stylistics has been cultivated for a long time,
and sometimes even seen as fundamental to linguistics, but for
most linguists "style" has been a marginal category, and most in-
vestigations have been of specialized genres. I am arguing that
the heritage of findings and insights into style becomes relevant
now to the central challenge facing linguistics, that the study of
speech styles can be seen, not as additional, but as fundamental.
3. This is a poor place to try to open issues of social theory,
but what I mean by "conditions" can be briefly sketched. The
historically given traditions are one such condition, one means out
of which cultural worlds are constituted. The forms, constraints,
and possibilities of recurrent types of structure, ecological or
economic structure, say, or social structure as a whole, can be
distinguished as another set of conditions. The experiences,
motives, minds of persons are another. There is a recurrent
tendency to take some one of these as the object of study and
theory. In Sapir's day, an impersonal objective "culture" in the
sense of a historical set of traditions was often so taken, and his
writings of the 1930s are in critique of that, for the sake of the
role of third condition, the personal. What we call "social anthro-
pology" often seems to fix upon the second, as if the first and
third were secondary or epiphenomenal; Marxism that derives
"superstructure" from "base" is akin. The fine insights of the
ethnomethodological movement in sociology run the danger of re-
ducing the whole to the third, as if the fact that cultural worlds
are constituted by participants could be enlarged to the proposi-
tion that they are solely or wholly so constituted, or that only
their constitution was worthy of study. All these things received
traditions, environmental and social structures, personal constitu-
tive activity- -seem to me conditions, urigins, of cultural worlds,
jointly, and even all together, not exhaustively. By occasional
use of the word "configuration" I mean to suggest that cultural
worlds, like lives and works of art, come out contingently and
have to be experienced to be known.
6. Ethnological studies of the distribution and diffusion of
traits have shown how permeable mapped boundaries may be. I
recall a Berkeley-trained ethnologist exclaiming that the
Tilbatulabal differed from a neighboring tribe in only two
traits. One, to be sure, was their language. I suspect a
specific structure of feeling would have been found also.
Language in Education 87

REFERENCES
Cazden, Courtney, ,et al. 1977. "Language Assessment: Where,
What and How." Anthropology and Education Quarterly 8(2):
83-91.
Cole, Michael, et al. 1971. The Cultural Context of Learning and
Thinking: An Exploration in Experimental Anthropology. New
York: Basic Books.
Halliday, M. A. K. 1973. Explorations in the Functions of Lan-
guage. London: Edward Arnold.
Hymes, Dell. 1964. "A Perspective for Linguistic Anthropology."
In Horizuns of Anthropology, Sol Tax, ed., pp. 92-107.
Chicago: Aldine.
. 1970. "Linguistic Method of Ethnography." In Method
and Theory in Linguistics, Paul Garvin, ed., pp. 249-325.
The Hague: Mouton.
. 1972. "Introduction." In Functions of Language in the
Classroom, C. Cazden, V. John, and D. Hymes, eds., pp.
xi, Ivii. New York: Teachers College Press.
. 1973. "Introduction." In Reinventing Anthropology, Dell
Hymes, ed., pp. 3-58. New York: Random House.
. 1974a. Foundations in Sociolinguistics. Philadelphia:
University of Pennsylvania Press.
. 1974b. "Ways of Speaking." In Explorations in the
Ethnography of Speaking, R. Bauman and J. F. Sherzer,
eds., pp. 433-451. New York: Cambridge University Press.
. 1976. Toward Educational Linguistics. (Horace Mann
Lecture, University of Pittsburgh School of Education, May
1976). Philadelphia: Graduate School of Education, Working
Paper #1.
Kroeber,, Theodora. 1961. Ishi, Last of a People. Berkeley:
University of California Press.
Philips, Susan. 1972. "Participation Structures and Communi-
cative Competence: Warm Sprivigs Children in Community and
Classroom." in Functions of Language in the Classroom,
Cazden, John, and Hymes, eds., pp. 370-394. New York:
Teachers College Press.
Postal, Paul M. 1974. On Raising: One Rule of English Grammar
and its Theoretical Implications. Cambridge: The MIT Press.
Sapir, E. 1949. "Sound Patterns in Language." Language 1:37-
51, 1925. Reprinted in D. Mandelbaum, ed., ,Selected Writings
of Edward Sapir, pp. 33-45. Berkeley and 4:s Angeles: Uni-
versity of California Press.
Shuy, Roger W. 1977. "Quantitative Language Data: A Case For
and Some Warnings Against." Anthropology and Education
Quarterly 8(2):73-82.
Vendier, Helen. 1975. The Poetry of George Herbert. Cambridge:
Harvard University Press.
Williams, Raymond. 1965. The Long Revolution. London: Peli-
can Books.
what is ethnography?'

"Ethnography"-is coming to be much discussed in education.

Often enough one hears some form of the question, "What is
ethnography?" The National Institute of Education has commis-
sioned a report to answer the question. All this might be puz-
zling to an anthropologist, especially to one with an interest in
the history of the subject. If one traces the history of ethnog-
raphy where it leads, one goes back centuries, indeed, to the
ancient Mediterranean world, and the temporary rise and fall of
ethnographic inquiry there, Herodotus being its most famous, but
not only, exemplar. With regard just to the Americas, one can
trace a .fairly continuous history of the ethnographic reports,
interacting with the posing of ethnological questions, from the
first discovery of the New World. There is a considerable modern
literature on the practice of field work, both in general and with
regard to,kpecific techniques, -and more recently, ethics. A re-
cent book addressed to ethnography in our own society (Spradley
and McCurdy 1972) is being used, I understand, by teachers of
composition to stimulate topics for their students. If ethnography
is new to some in education, certainly it is not new to the world.
When asked "What is ethnography?", would it not be enough to
provide a short reading list, or to point to the discussion in some
text of what research proposals often refer to as "standard
ethnographic method?"
I fear not_ Anthropologists do not themselves have a unified
conception of ethnography. In particular they do not have a uni-
fied conception of ethnography in relation to the study of insti-
tutions of our own society, such as education. And anthropolo-
gists are far from accepting or perfecting an integration of the
mode of research they would consider ethnography with other
modes of research into a society such as our own. And the
changing intellectual context of the human sciences as a whole
introduces new questions and sources of diversity.
Educational research has been dominated by quantitative and
experimental conceptions of research. It is easy for anthropolo-
gists of a variety of persuasions to criticize such methods. It is
often harder for them to state concisely the alternatives. "Ethnog-
raphy" cannot be assumed to be something already complete,
ready to be inserted as a packaged unit in the practices and pur-
poses of institutions whose conceptions of knowledge and research
have long been different. If there is not careful thinking through
of underlying conceptions, and explicit attention to differences in
88
Language in Education 89
them, "ethnography" may be a brief-lived fad in educational re-
search. Or worse, partial or superficial conceptions may be
taken up.
The true opportunity of the current interest in ethnography
is to enter into a mutual relation of interaction and adaptation be7
tween ethnographers and sponsors of educational research, a re-
lation that will change both. Because of my conception of ethnog-
raphy, I see in this prospect a gain for a democratic way of life.
The following sketch is offered because I do not know of a simi-
lar attempt to consider the issues raised here in brief compass.
Many others must contribute from their own experience and out-
look.
li

One difficulty with the notion of "ethnography" is that it may

seem a residual category. It is associated with the study of
people not ourselves, and with the use of methods other than
those of experimental design and quantitative measurement.
Clearly not everything that is not those two things should be
considered "ethnography," but a positive definition is not easy
to provide. A major reason for the difficulty is that good ethnog-
raphy has been produced under a great variety of persons, some
of it before there was a profession to train such people, and pro-
fessional training has been very much a matter of the transmission
of a craft and of learning by doing, by personal experience.
It has not helped that some people talk as if the key to
ethnography were a personal psychological experience, rather
than the discovery of knowledge.
It is clear that ethnography involves participation and obser-
vation. What should count as ethnography, what kinds of ethnog-
raphy there are, may be more easily seen if we consider what
makes participation and observation systematic--what, in short,
counts as systematic ethnography.
The earliest work that we recognize as important ethnography
has generally the quality of being systematic in the sense of being
comprehensive. To be sure, any and all early accounts of travel-
lers, missionaries, government officials, and the like, that may
contribute information and insight about the culture of the peoples
of the world, has been welcomed and gleaned for what it could
provide. But when Ibn Khaldun or Father- Sahagun are singled
out, it is because they are both early and comprehensive. Their
curiosity was not limited to curiosities; they had an interest in
documenting and interpreting a wide range of a way of life.
Much of the early attempts to make ethnographic inquiry an
explicit procedure reflect the desire to be comprehensive. These
attempts are guides to inquiry, lists of questions, of observations
to make. They bespeak a stage of history when much of the non-
Western world was little known to Europe, and when a variety of
reasons, scientific, religious, practical, motivated some to seek
90 What is Ethnography?
more adequate knowledge. These guides to inquiry have in com-
mon a concern with all of a way of life (although their coverage
may be unequal). What are the people of such and such a place
like?
It was not long before there were explicit procedures that
can be distinguished as topic-oriented. (Indeed, the Domesday
Book and the inquiries of Sir William Petty (1623-87) share in
this lineage.) 2 The great American example is Lewis Henry
Morgan's questionnaires for recording kinship terminologies in the
middle of the nineteenth century. It is worth pausing to consider
the several aspects of Morgan's great work. First,, he had a con-
trastive, or comparative insight: from his experience with the
Iroquois Indians, together with his knowledge of classical Greece,
he realized that there was a principle of kinship organization
sharply contrasting with that familiar 'to contemporary Americans
and Europeans. He sought then to determine the main types of
kinship systems and their locations throughout the world. Second,
he needed systematic information, information not available except
as he sought it himself. Hence his own travels in the Western
United States and his relentless correspondence with those who
could help. Third, he made use of his findings to formulate first
a historical (Systems of Con-sanguinity and Affinity, 1870) and
then evolutionary interpretation (Ancient Society', 1877) of the
most general sort, of human 'development as a whole.
These three aspects of inquiry seem the essential ingredients
of anthropological research proper, as distinct from inquiry that
contributes to anthropology. Each aspect may exist independently
--a contrastive insight, a seeking of specific information, a
general interpretation. Anthropology proper exists insofar as the
three are united in a common enterprise, 'Ethnography is more
than a residual technique, but the name of an essential method,
when all three are united.
With time there has come to be a certain body of ethnographic
inquiry that can be said to be hypothesis-oriented. To be sure,
Morgan had a general hypothesis. But it seems reasonable to dis-
tinguish the kind of ethnography organized and guided by John
and Bea Whiting, for example, to inquiry into socialization in
several contrasting societies. The Whitings had attempted to come
to general conclusions, testing hypothesis on a theoretical base,
from the ethnographic literature that existed at the time. They
found, as so many find, that their questions were more specific
than the literature could answer. The sources were not detailed
enough for their purpose, and not comparable enough. There-
fore they organized a project to provide the detailed, comparable
information they needed. Ethnographic teams (generally, couples)
were trained in terms of a guide to the field study of socializa-
tion, sent into the field in several different societies for an ex-
tended period of time, kept in touch with through correspondence
throughout the field study, and brought back to write up their
Language in Education 91

results. (Six Cultures, by Bea Whiting, is a principal outcome.)

Like Morgan, the Whitings had insiglv into contrasting types of
society, a need for specific information, and a general theoretical
frame (in this case, psychodynamic) to which contrast and spe-
cifics were relevant.
All three types of ethnographic inquiry continue to coexist in
anthropology today. There are still occasional discoveries of un-
known peoples (as in the Philippines a few years ago), for whom
comprehensive information has to be provided from scratch.
There are still many peoples, knowledge about whom has never
been adequately systematized, for knowledge about whom has
serious gaps. Fresh ethnography may be undertaken for first-
hand knowledge :As a basis from which to integrate all that is
known, or to fill a gap in what is already known. .

There are still discoveries of aspects of culture, or of per-

spectives on culture, such that the existing literature fails to
provide much information. The "ethnography of speaking" is a
case in point. The Human Relations Area Files, although rich in
ethnographic data, simply did not contain much information on
cultural patterning of speech, let alone information at all on the
fundamental theoretical question, the functions of speech in the
society. (Anthropological theory had taken for granted that the
functions of speech were everywhere the same.) When such a
discovery or perspective comes to the fore; topic-oriented ethnog-
raphy may be undertaken. One needs to find out something of
the range of cultural patterning, once cultures are investigated
from the new point of view.
Both comprehensive ethnography, and topic-oriented ethnog-
raphy, lead to hypothesis-oriented ethnography. Given a sub-
stantial general knowledge of a culture, precise investigations can
be planned-. Indeed, hypothesis-oriented research depends on the
existence of comprehensive ethnography, and can be fruitfully
pursued only where the latter exists. Again, once something of
the range of patterns for an aspect of culture is known, one be-
gins to formulate more precise questions. Research may show the
recurrence of a contrast in styles of speaking that can be called
"direct" vs. "indirect"--but are the attributes of the two styles
the same in each case? Aire the functions the same? In the
ethnography of speaking right now, one is aware of broad con-
trasts, usually presented as dichotomies (cf. Bernstein's "elabor-
ated" and "restricted" codes). This fact is a sure sign, I think,
of the topic-oriented stage and of the need to proceed to the
hypothesis-oriented stage.
III

What might the subject of schooling in America be like in this

context? Clearly there is a great deal of information already in
hand. It is not so clear that the information is obtained and
92 What is Ethnography?
analyzed in ways that permit all the insight possible into school-
ing. If schools were considered from the same standpoint as kin-
ship systems or languages, the first question might be: what
kinds of schools are there? It would not seem informative enough
to know that test scores were up or down in general across all
American schools, if in fact the country contains schools of many
different types. The point would apply even within a single city
or district. Are the schools of District 1 in Philadelphia all
alike? If they are different, how many different kinds are there?
Probably at any level of consideration, one would not want to say
that all were alike, nor that all were incomparably unique. In
sum, one would recognize a question of typology, as central to
analysis.
A useful typology has to be designed in terms of a particular
purpose. Kinship is important to social life, and central to many
societies, but even so, a classification and analysis of societies
according to kinship is not the same as a classification and analy-
sis according to religion. There are strong connections, but not
invariant bonds, among the various sectors of a way of life, and
so also, among the various sectors of schools. A typology of
schools in District 1 in terms of verbal skills, questions having to
do with literacy, would not necessarily be the right typology for
some other purpose. Conversely, and here is an essential point,
a typology for some other purpose is not necessarily right for a
concern with verbal skills.
This essential point is an example of a general consideration
that divides many ethnographers from an experimental model, at
least as that model is understood by them. For many ethno-
graphers, it is of the essence of the method that it is a dialecti-
cal, or feed-back (or interactive-adaptive) method. It is of the
essence of the method that initial questions may change during
the course of inquiry. One may begin with the assumption that
every community must have a pattern for the residence of newly-
married couples that can be of only one of four types, yet dis-
cover that the community one is studying actually determines the
residence of newly-married couple on the basis 'of principles one
had not foreseen. (The illustration is an actual one. See
W. Goodenough, "Residence rules," 1956.)
The history of anthropology is replete with, experiences of
this sort. The general mission of anthropology in part can be
said to be to help overcome the limitations of the categories and
understandings of human life that are part of a single civiliza-
tion's partial view. For many ethnographers, an essential charac-
teristic of ethnography is that it is open-ended, subject to self-
correction during the process of inquiry itself.
All this is not to say that ethnography is open-minded to the
extent of being empty-minded, that ignorance and naivete are
wanted. The more the ethnographer knows on entering the field,
the better the result is likely to be. Training for ethnography is
only partly a matter of training for getting information and getting
Language in Education 93
along. It is also a matter of providing a systematic knowledge of
what is known so far about the subject. The more adequate this
knowledge, the more likely the ethnographer will be able to avoid
blind alleys and pursue fruitful directions, having a ground sense
of what kinds of things are likely to go together, what kinds of
phenomena need minimal verification, what most.
One conception of this process is that of Kenneth Pike. Pike
generalized the experience of linguistic inquiry. In order to dis-
cover the system of sounds of a language one had to be trained
to record the phenomena in question, and one had to know what
types of sound were in general found in languages. Accurate
observation and recording of the sounds, however, would not dis-
close the system. One had to test the relations among sounds for
their functional relevance within the system in question. The re-
sult of this analysis of the system might in turn modify the
general framework for such inquiry, disclosing a new type of
sound or relation. Pike generalized the endings of the linguistic
terms "phonetic" and "phonemic" to obtain names for these three
moments of inquiry. The general framework with which one be-
gins analysis of a given case he called "etici." The analysis of
the actual system he called "emic." The reconsideration of the
general framework in the light of the analysis he called "etic2."
When ethnographic and linguistic inquiry are described in
such terms, it may be easy to see the connection with general
scientific method, and the exemplification of such method in the
experimental sciences. For many ethnographers and linguists the
spirit of inquiry is indeed the same. The scale and conditions of
inquiry in ethnography, nevertheless, impose essential differences
in tactics. Perhaps the key to these differences is meaning.
For ethnographic inquiry, validity is commonly dependent
upon accurate knowledge of the meanings of behaviors and insti-
tutions to those who participate in them. To say this is not to
reduce the subject matter of ethnography to meaning, let alone to
native views of meaning. It is simply to say that accurate knowl-
edge of meaning is a sine qua non. The problem is obvious
enough in the case of a language and culture we do not know.
It is less obvious in the case of communities around us. Yet
even though one may live nearby, speak the same language, and
be of the same ethnic background, a difference in experience may
lead to misunderstanding the meanings, the terms, and the world
of another community. In Philadelphia, for example, a question-
naire was prepared by a person generally qualified by training
and background. The purpose of the questionnaire was to find
out what parents thought of a community-relations policy and per-
son. The questionnaire was duly administered. The student
administering the questionnaire discovered, by informal conversa-
tions with parents, that they interpreted the questions differ-
ently than the designer and the school. They distinguished be-
tween a playground (having equipment designed for children to,
use) and a playyard, but the questionnaire did not. When asked
94 What is Ethnography?
if they had had a chance to meet their School-Community Coordi-
nator, they answered "no," because to them to "meet" would re-
quire having talked, and knowing by name, even first-name, not
just having been introduced, but in terms of the questionnaire
their "no's" were interpreted as "not having met." The student
administering the questionnaire was distressed, but there was no
way in the procedure of inquiry for him to take account of what
he had learned or to have what he had learned affect the pre-
sumed results (Abbott 1968).
Experiences of this kind make ethnographers distrust question-
naires, and quantitative results derived from them, if the mean-
ings of the questions to those asked to answer are taken for
granted in advance. Many ethnographers do use questionnaires,
but questionnaires devised after sufficient participation and ob-
servation to ensure their validity.
The validity of knowledge about persons, families, neighbor-
hoods, schools, communities in our country depends upon accur-
ate and adequate knowledge of the meanings they find and impute
to terms, events, persons, institutions. To an important extent,
such meanings cannot be taken for granted as uniform, even
within a single city or school district, nor as known in advance.
The overt forms may be familiar -the words, the attire, the
buildings--but the interpretation given to them is subject to
shift, to deepening, to fresh connecting up. (It has been found
that within a single small factory in Pennsylvania, those who
worked in different parts had different terms for the same things
[see Tway (1975)].)
It is in the nature of meanings to be subject to change, re-
interpretation, re-creation. One has to think of people, not as
the intersection of vectors of age, sex, race, class, income, and
occupation alone, but also as beings making sense out of disparate
experiences, using reason to maintain a sphere to integrity in an
immediate world.
All this is not to say that ethnography indulges in an infinite
regress of personal subjectivity and idiosyncratic worlds. It has
to be open to that dimension of social life, because that dimension
affects the reality of social life, and the success or failure of
social programs. The point is to stress the necessity of knowl-
edge that comes from participation and observation, if what one
thinks one knows is to be valid. And all this is not to say that
members of a community themselves have an adequate model of it,
much less an articulated adequate model. All of us are only partly
able to articulate analyses of our lives and their contexts. The
meanings which the ethnographer seeks to discover may be im-
plicit, not explicit. They may not lie in individual items (words,
objects, persons) that can be talked about, but in connections
that can only gradually be discerned. The deepest meanings and
patterns may not be talked about at all, because so fully taken
for granted.
Language in Education 95

Here again the need to discover and validate in the given

case is paramount. Our familiar categories of institutions, modes
of communication, or the like, are an indispensable starting point
(Pike's "etici"), but are never to be equated with an analysis of
the organization of a local way of life. We necessarily distinguish
speech and song, and as polar opposites, there may be speech
with no musical quality, and singing without words. In our own
musical traditions and in the cultures of the world, the inter-
connections and conceptions of these relationships of speaking
and singing (and music generally) are various and diverse.
Modern serious music includes such categories as Sprechstimme.
The Maori of New Zealand consider the playing of the flute a
form of speaking. It is especially these local nodes of connec-
tion, these community-specific ways of putting the encyclopedia
of culture together, that cannot be assumed in advance of in-
quiry, and that can only be discovered through participation and
observation over time.
We know that Philadelphia has newspapers, radio stations,
television stations, libraries, books, comic books, magazines, in-
scriptions and plaques, narrators and joke-tellers. Without eth-
nography we can collect statistics as to production and distri-
bution. Only with ethnography can we discover the connections
among these things in the lives of particular kinds of people.
Even self-report cannot be relied upon--people are notoriously
unable or unwilling to give accurate accounts of the amount of
time, say, they spend on various things. And a key to the sig-
nificance of a type of television program may not be in the
amount of time the family set is on, but in the family pattern of
speaking around it. Is the set on, but ignored? Does someone
insist on and get silence? Is the program essentially a resource
for continuing conversation?
All that I have said is compatible with a generous view of
scientific method. The subject-matter of ethnography--people
and their worlds--imposes conditions such that validity and re-
search design have a complexity and openness at the other end
of the scale from experimental design in many fields. Even so,
there is a similarity to the problems of fields such as astronomy
and geology in their observational aspects. In principle, a
sufficient accumulation of valid knowledge about a particular
society would make possible rather efficient, precise inquiries
of an experimental or quasi-experimental sort. Indeed, there
are some cases of this sort. My own experience is with lan-
guages. Given the accumulated knowledge of the Native Ameri-
can language I have studied, it is possible to address many par-
ticular questions rapidly, systematically, precisely. If a new
word is in question, it is a matter of minutes to establish its
grammatical place in the system. One does know already just
what questions must be asked, to establish that the word is a
noun or verb, say, and just what kind. If a newly discovered
recording of a familiar word is in question, one knows exactly
 96 What is Ethnography ?
what the possibilities of interpretation are, how the facts of the
language constrain the sounds that the letters may represent.
(All of this presupposes a native speaker who has become accus-
tomed to collaboration in work of this kind.) A social anthro-
pologist can look at a newly collected schedule of kinship terms
and place the system approximately quite quickly.
iv
These examples illustrate two points made previously--that
ethnographic training involves training in the accumulated com-
parative knowledge of the subject, and that the existence of com-
prehensive knowledge about a community makes more precise
hypothesis-testing possible.
To leave matters here, however, might suggest that ethnogra-
phy can become almost equivalent to laboratory work. The theo-
retical foundations on which it rests, however, have yet to prove
as certain as that. The social knowledge that ethnography serves
is in the paradoxical position of becoming increasingly certain of
more and more, and yet at the same time vulnerable to dispute
about its very foundation. Our last example, that kind of kinship
terms, is an excellent case in point. No one could deny the strik-
ing progress in our ability to recognize and describe the terms
and behaviors relevant to kinship in the hundred years or so
since Morgan's pioneering work. Descriptive technique, compara-
tive typology, even mathematical modelling are well advanced.
Yet it is possible to dispute the correct interpretation of particu-
lar findings (cf. Blu 1967) and even the correct definition of the
domain of "kinship" itself (cf. Schneider 1972; Geertz and Geertz
1975). Differences in analytic point of view can take different
vantage points within a shared body of data (e. g. , Blu) , and
even prescribe different definitional constructs for master-
concepts such as the "cultural" and "social," such as to require
different placements of findings.
Notice that such disputes presuppose the success of ethnogra-
phy. Ethnography has provided what there is to dispute. To be
sure, new theory can bring out new aspects of old data, or point
up its limitations, or require new kinds of data to develop ade-
quately. Still, such disputes presuppose a great deal of valid
first-order data. And they tacitly assume that valid first-order
data is largely a matter of competence and talent, requiring per-
ceptiveness and imaginative projection, to be sure, but not miracles
of rapport and identification. The native's point of view can be
grasped by someone who does not always like it or them (cf.
Geertz 1976).
In sum, ethnographers have available in many areas a first-
order language of description that permits them to do work that
can be judged for its competence, validity, richness, and the like.
At the same time they work in a discipline whose second-order
language of analysis is contested. A set of native terms may be
Language in Education 97
accepted, but analysts argue as to the priority of ont, or another
part of the data, or as to the scope of the data to be con-
sidered, or as to whether the case does or does not fit a certain
concept or type. Some of these disputes can be resolved with
increasing attention and precision. Others depend on conscious
or unconscious commitments to what it means to be a scholar or
scientist, and to what the world is or should be, such that resolu-
tion is unlikely.
One can look at a schedule of kinship terms, then, and place
it approximately quite quickly. But "approximately" is a crucial
term. Others may agree, that is a Crow system, and add, "But
...". The ethnography of schooling and education no doubt will
have the same experience. We can hope to reach the point at
which anyone can look at a body of data and say quite quickly,
"Ah, that is a Henry Lea type of school ," or the equivalent, or
say, that community or that family has a Hopi type of educational
process. We can probably not hope to reach the point at which
no one will object, but you analyzed the school in isolation from
X, or started the analysis from Y instead of Z.
There is a second question of language which may differentiate
ethnography from the ideal conception of an experimental science.
Some of what we believe we know about cultural patterns and
worlds is interpretable in terms of structure, whether the ingredi-
ents of the structure be lines, graphs, numbers, letters, or ab-
stract terms. Some of what we believe we know resists interpre-
tation in terms of structure. It seems to require, instead, pre-
sentation.
The need for presentation seems to cause no comment when the
presentation is visual. "A picture is worth a thousand words"
and all that. Especially when the object of analysis is material
culture, visual presentation is accepted as indeed essential. Even
with social life, there is an increasing recognition of the value of
visual presentation, through photographs and films. A telling
account of necessity here was made by Frederick Barth, the Nor-
wegian anthropologist, when he reported that in order to make
good sense of a deceased colleague's field notes, he had to go to
the place in question. He had worked ethnographically in the
region (which is why he was asked to interpret the notes for
publication), but still needed to see the land, the distances be-
tween dwelling and well, the heights and contours of the place-
so much of the spatial configuration of life was taken for granted
in the notes on behavior.
The difficulty with presentation seems to arise when the presen-
tation is verbal. What is one to make, for example, of the rela-
tion between the two parts of Clifford Geertz' "The Balinese cock-
fight"--one part narrative, one part analytic (1972). Geertz
thinks that both parts are important. 3 I do also. Through his
narrative skill, he is able to convey a sense (mediated by his
personal involvement) of the quality and texture of Balinese
fascination with cockfighting. Evidence of the fascination is
98 What is Ethnography?
important. It supports taking the activity as a key to something
essential about the Balinese; it helps us understand the analytic
statements. A film might help too, but it would need something
verbal from Geertz to teach us what we should learn from it. The
narrative part of Geertz' article in effect points, as the narrator
of a film might do, and, also, in the absence of a film, shows. It
does so through texture and proportion.
Many anthropologists agree that something of value can be
learned from novels of certain sorts, and even recommend certain
novels. Clearly there is a sense in which narrative can be a
source of knowledge. For some scientists and philosophers of
science, it is a source of knowledge secondary to others, if not
in principle reducible to others. Some ethnographers and philoso-
phers of science hold the contrary. Narrative does not seem to
them in principle entirely reducible to other forms of knowledge,
but fundamental in its own right. Indeed, they may suspect that
narrative accounts play a role in what scientists and administrators
believe themselves to know, even though some of these may not
acknowledge that role. It may be the case that structural forms
of knowledge about social life are usually interpreted, even if
covertly, in terms of images of kinds of person and situation, im-
plicit or remembered narratives. If so (and I think it is so), the
general problem, of social knowledge is two-edged: both to increase
the accumulated structural knowledge of social life, moving from
narrative to structurally precise accounts, as . -we h_ ave commonly
understood the progress of science, and to bring to light the in-
eradicable role of narrative accounts. Instead of thinking of nar-
rative accounts as an early stage that in principle will be replaced,
we may need to think of them as a permanent stage, whose princi-
ples are little understood, and whose role may increase,. How
often, one wonders, are decisions reached on the basis not only
of numbers and experiments, but also on the basis of privileged
personal accounts, fleshing out the data to make it intelligible?
Sometimes these accounts may be provided by the investigator,
sometimes by the audience ("I knew a case once...".) Sometimes
they may not be articulated, yet influential nevertheless.
If narrative accounts have an ineradicable role, this need not
be considered a flaw. The problem is not to try to eliminate them,
but to discover how to assess them. What criteria can we provide
equivalent to the criteria for assessing the significance and
validity and reliability of statistical tests, and experimental de-
signs?
The question of narrative brings us to another aspect of eth-
nography. It is continuous with ordinary life. Much of what we
seek to find out in ethnography is knowledge that others already
have. Our ability to learn ethnographically is an extension of
what every human being must do, that is, learn the meanings,
norms, patterns of a way of life. From a narrow view of science,
this fact may be thought unfortunate. True objectivity may be

13
Language in Education 99

thought to be undermined. But there is no way to avoid the

fact that the ethnographer himself or herself is a factor in the
inquiry. Without the general human capacity to learn culture,
the inquiry would be impossible. The particular characteristics
of the ethnographer are themselves an instrument of the inquiry,
for both good and bad. For good, it is important to stress, be-
cause the age, sex, race, talents of the ethnographer may make
some knowledge accessible that would be difficult of access to
another. For bad, as we all recognize, because of partiality.
Since partiality cannot be avoided, the only solution is to face up
to it, to compensate for it as much as possible, to allow for it in
interpretation. The conditions of trust and confidence that good
ethnography requires (if it is to gain access to valid knowledge
of meanings) make it impossible to take as a goal the role of im-
partial observer. The normal people from whom one has to learn
will not put up with that. In principle, the answer lies in the
view taken by Russell Ackoff, that scientific objectivity resides,
not in the individual scientist, but in the community of scientists.
That community has provided methods which to some degree, often
a great degree, discipline the investigator and overcome partiality;
the rest is a responsibility of critical analysis within the com-
munity.
The fact that good ethnography entails trust and confidence,
that it requires some narrative accounting, and that it is an ex-
tension of a universal form of personal knowledge, make me think
that ethnography is peculiarly appropriate to a democratic society.
It could of course be reduced to a technique for the manipulation
of masses by an elite. As envisioned here, ethnography has the
potentiality for helping to overcome division of society into those
who know and those who are known.
Such a vision of a democratic society would see ethnography
as a general possession, although differentially cultivated. At
one pole would be a certain number of persons trained in ethnogra-
phy as a profession. At the other pole would be the general
population, respected (on this view of ethnography) as having a
knowledge of their worlds, intricate and subtle in many ways
(consider the intricacy and subtlety of any normal person's knowl-
edge of language), and as having necessarily come to this knowl-
edge by a process ethnographic in character. In between--and
one would seek to make this middle group as nearly coextensive
with the whole as possible -would be those able to combine some
disciplined understanding of ethnographic inquiry with the pursuit
of their vocation whatever that might be. From the standpoint of
education, obviously one wants to consider the possibility of add-
ing ethnographic inquiry to the competencies of principals, teach-
ers, and others involved with schools. But on the one hand,
there is no reason not to seek to extend a knowledge of ethno-
graphic inquiry to everyone. And, on the other hand, there is
no reason to think professional ethnographers privileged. In
their own lives they are in the same situation as the rest--needing
100 What is Ethnography?
to make sense out of a family situation, a departmental situation,
a community situation, as best they can.
If this account sounds a little like a form of conscious-raising,
perhaps it is, but it is not the ordinary sort. In this sketch I
have not brought out the sociocultural substantiality of ethno-
graphic inquiry. It is a mode of inquiry that carries with it a
substantial content. Whatever one's focus of inquiry, as a matter
of course one takes into account the local form of general proper-
ties of social life--patterns of role and status, rights and duties,
differential command of resources, transmitted values, environ-
mental constraints. It locates the local situation in space, time,
and kind, and discovers its particular forms and center of gravity,
as it were, for the maintenance of social order and the satisfac-
tion of expressive impulse.
It is for this reason that much observational analysis of class-
rooms does not seem to me to merit the term "ethnography." On
the one hand, there is a kind of work which consists essentially
of recurrent observations according to a pre-established system of
coding. Such work violates the principle of being open to dis-
covering meanings and patterns of behavior not foreseen. There
is no provision for meanings, and patterns are excluded, since
integral stretches of behavior are not observed. A superior kind
of work analyzes intensively integral sequences of behavior. It
contributes greatly to analytic control and to penetration to under-
lying meanings and connections. The limitation of the work is
the lack of a comparative perspective. Thus, Ray McDermott
(1977) interprets the difficulties, and projected failure, of one
group of first-grade readers, in terms of the mode of interaction
among them and the teacher (in contrast to-the mode of inter-
action of another group with the teacher). Since his data is
limited to the classroom, he is not able to consider differences in
mode of interaction that the children may have brought with them
to the classroom. A comprehensive ethnography would consider
all the types of scenes in which the children (and teacher) par-
ticipate, in order to assess validly the meaning of the behavior in
one scene. From a larger point of view, the lack of a compara-
tive ethnological perspective weakens the contribution the research
might make. Suppose it is the case that what happens in the
first-grade classroom is going to determine the success or failure
of the children as readers, and ultimately, as adults. (This is
McDermott's view.) We have to ask whether or not such a circum-
stance is inevitable. Some Third World societies have had great
success with literacy programs, through mobilization of the society
to accomplish it. Is it impossible for the United States to give
that kind of priority to universal literacy among its citizens, and
mobilize to accomplish it? Is the fundamental question not this:
how does it come about that one society, and not another, lets
literacy depend on patterns of interaction in first-grade class-
rooms?
 Language in Education 101
Such an interdependence between general and particular in-
quiry is essential to ethnography
it is essential in my reading of theashistory
a mode of inquiry--at least
of anthropology,
to what I would see as the contribution ethnography ought toand
make to education.
FOOTNOTES
1. I want to thank Peggy Sanday for stimulus to write this
paper, Richard Bauman and Joel Sherzer for including it in the
series of Working Papers in Sociolinguistics, and Perry Gilmore
for inviting its inclusion in the conference on education and eth-
nography that she has coordinated for Research for Better
Schools and the Graduate School of Education of the University
of Pennsylvania.
2. Petty outlined "A method of enquiring into the state of
any country" and in 1686 a series of questions concerning "The
nature of the Indians of Pennsylvania" (Hodgen 1964:190; Slotkin
1965:481, n. 363, where the title is given as "Quaeries concerning
the nature of the natives of Pennsylvania"). Petty's questions
were part demography--he has been called "the greatest exponent
of social statistics in the seventeenth century" (Slotkin: 139) , part
ethnology. He was a man who both recommended that the Royal
Society admit only words that mark number, weight, or measure,
and who was concerned with clearer definitions of ethnological
entities. Altogether a worthy forefather for any effort to inte-
grate quantitative and qualitative methods today. His program of
ethnological investigation
had actually done some 40was similar to what Roger Williams (1643)
years earlier among Indians, but still
quite unusual for its time. By the eighteenth century serious
frames of reference for the collection and interpretation of cultural
facts ("manners and
lished tradition, as incustoms") were to become part of an estab-
the Scottish work of Adam Ferguson, Lord
Kanes, James Millar, and William Robertson.
3. Geertz (1976) expresses superbly the dialectic between the
two orders of analysis, the descriptive and the generalizing, that
requires concern with adequacy of "presentation." Thus, he
writes (p. 223):
Confinement to experience-near concepts leaves an eth-
nographer awash in immediacies, as well as entangled in
vernacular. Confinement to experience-distant ones
leaves him stranded in abstractions and smothered in
jargon. The real question...is...how, in each case,
ought one to- deploy [the two sorts of concepts]...so
to produce an interpretation of the way a people livesas
which is neither imprisoned within their mental horizons
...nor systematically deaf to the distinctive tonalities of
their existence.

116
102 What is Ethnography?
REFERENCES
Abbott, Lloyd M., Jr. 1968. "Considerations of language and
culture for a social survey." Paper for Anthropology 528,
University of Pennsylvania.
Btu, Karen I. 1967. "Kinship and culture: Affinity and the
role of the father in the Trobriands." Southwestern Journal
of Anthropology 23:90-109. [Reprinted in Symbolic anthro-
pology, ed. by J. L. Do lgin, D. S. Kemnitzer, pp. 47-62.
New York: Columbia University Press, 1977. ]
Geertz, Clifford. 1972. "Deep play: Notes on the Balinese
cockfight. " Daedalus 101: 1-37.
. 1976. "'From the native's point of view': On the nature
of anthropolOgical understanding." In Meaning in anthropology,
ed. by K. H. Basso and H. A. Selby, pp. 221-237. Albu-
querque: University of New Mexico Press.
Geertz, Hildred and Clifford Geertz. 1975. Kinship in Bali.
Chicago: University of Chicago Press.
Goodenough, W. H. 1956. "Residence rules." Southwestern
Journal of Anthropology 12:22 -37.
Hodgen, Margaret T. 1964. Early anthropology in the sixteenth
and seventeenth centuries. Philadelphia: University of
Pennsylvania Press.
McDermott, R. 1977. "The cultural context of learning to read."
In Issues in evaluating reading, ed. by S. F. Wanat,
18. (Linguistics and Reading Series, 1). Arlington, Va.:
Center for Applied Linguistics.
Morgan, L. H. 1870. Systems of consanguinity and affinity of
the human family. Washington, D.C.: Smithsonian Institution.
. 1877. "Ancient society." Researches in the lines of
human progress from savagery through barbarism to civiliza-
tion. New York: Henry Holt.
Petty, Sir William. [1686] "Queries concerning the nature of the
natives of Pennsylvania." In The Petty papers: Some un-
published writings of Sir William Petty, ed. from the Bowood
Papers by the Marquis of Lansdowne. London: Constable,
1927. Vol. II, pp. 115-9.
Pike, Kenneth L. 1965. Language in relation to a unified theory
of the structure of human behavior. The Hague: Mouton.
Schneider, David M. 1972. "What is kinship all about?" In
Kinship studies in the Morgan centennial year, ed. by Priscilla
Reining, pp. 32-63. Washington, D.C.: Anthropological
Society of Washington.
Slotkin, J. S. 1965. Readings in early anthropology. (Viking
Fund Publications in Anthropology, 40). Chicago: Aldine.
Spradley, James P. and David W. McCurdy, eds. 1972. The
cultural experience: Ethnography in complex society.
Chicago: Science Research Associates.
Tway, Patricia. 1975. "Workplace isoglosses: Lexical variation and
change in a factory setting." Language in Society 4(2):171-183.
1I
Language in Education 103
Whiting, Beatrice B., ed. 1963. Six cultures: Studies of child
rearing. New York: John Wiley.
ethnographic monitoring

I want to consider the contribution of ethnography to bilingual

education, and to argue, indeed, that ethnography is essential
to the success of bilingual education.
One contribution of ethnography has to do with the planning
of programs and the need for knowledge of the initial state of
affairs. This contribution of ethnography (initial knowledge) is
perhaps familiar, though neglected. There are two other kinds
of contributions as well. One has to do with the conduct of pro-
grams, the need to recognize and understand patterns and mean-
ings that may emerge during the course of a program, perhaps
outside the classroom. The other has to do with the evaluation
and justification of programs, and ultimately, the evaluation and
justification of bilingual education itself. The second and third
contributions can be thought of as ethnographic monitorinc-, the
one of an ongoing operation, the other of effects and conse-
quences. Attention should be paid to both.
Ethnography might be thought of as something purely descrip-
tive and objective, done by someone who comes from outside.
This is not the view I hold. Ethnography must be descriptive
and objective, yes, but not only that. It must be conscious of
values and goals; it must relate description to analysis and ob-
jectivity to critical evaluation. Bilingual education involves social
change in the light of certain goals. As a matter of law, it is
defined in terms of the goal of equality of educational opportunity.
As a matter of social change, it involves much more--personal
goals and commitments, what people consider their life chances
and identities to be, what they want them to become. Not every-
one may agree on the goals of change, or how much and what
kind of change is desirable. An ethnographer must come to
understand the values involved and the validity of those values
to those who hold them, as well as come to understand his/her
own attitudes (perhaps attitudes that emerge in the course of his
or her work) and the reasons for them. Only explicit concern
with values, in short, will allow ethnography to overcome hidden
sources of bias. To be truly useful, ethnography must relate
what is described to goals. It may be that certain goals and cer-
tain situations are not compatible, or that certain goals and cer-
tain means are not, or that an unwitting inconsistency exists
among goals or means. Ethnography is an essential way of dis-
covering what the case is, and social programs that ignore it are
blind, but ethnography that ignores values and goals is sterile.

104

ha
 Language in Education 105
Ethnography might be thought of as something done by some-
one from without, a hired professional. It does require training
and talent; not just anyone can do it. But of all forms of scien-
tific knowledge, ethnography is the most open, the most compati-
ble with a democratic way of life, the least likely to produce a
world in which experts control knowledge at the expense
who are studied. The skills of ethnography consist of theofen- those
hancement of skills all normal persons employ in everyday life; its
discoveries can usually be conveyed in forms of language that
non-specialists can read. It comes to know more of a way of life
than those that live it are consciously aware of, but must take
crucial account of what they consciously, and unconsciously, know.
A crucial ingredient of ethnography is what
known to members of a community, what theyinmust a sense is already
know , consci-
ously and unconsciously, in order to be normal members
community. As a discipline, ethnography adds a body ofof the
concepts
and techniques that directs attention and relates observations
more systematically than community members would normally have
occasion to do, that provides for making explicit relationships and
patterns that members leave implicit, and that provides for inter-
preting patterns in the light of a comparative knowledge of other
ways of life to which a community member would not usually have
access. Ethnography, in short, is a disciplined way of looking,
asking, recording, reflecting, comparing, and reporting. It
mediates between an understanding of what members of a given
community know and do, and an accumulated comparative under-
standing of what members of communities generally have known
and done.
A member of a given
source of data, an object community,
at the other then,
end
need not be merely a
of a scientific instru-
ment. He/she already possesses some of the local knowledge and
has access to knowledge that is essential to successful ethnogra-
phy; he/she may have a talent for sifting and synthesizing
special insight into some part of it. What the member needs it, is
a
the other part of disciplined ethnography, the comparative insight
distilled over the decades. This can come in a variety of idioms
and does not require a graduate degree. Indeed, one might argue
that an educational system devoted to a democratic
would provide this other part to every student, as way of life
a basis for citizenship. Not to do so is to withhold afrom right and as
citizens
the best that we have to offer for the understanding of social
experience and for coming to terms with it or changing it.
When I refer to ethnography,
doing the ethnography may be fromthen, I assume that the person
the community in question.
Indeed, I think it is highly desirable that this be the case in a
large proportion of cases (Hymes, 1974a; Hymes, 1972).
The- contribution of ethnography to initial knowledge may be
familiar. Still, when I suggested ethnography to the director of
a language program in a large city, the response was, "What
would you want to know?" The only elaboration of the answer
106 Ethnographic Monitoring

was that the people in the community would not want it 'to be
known that so many of them were illiterate.
Such a response suggests that the idea of sociolinguistic de-
scription, of ethnography of speaking, has not gone very far be-
yond academic halls. Indeed, it has not gone very far within
them, so far as educational settings are concerned. One repeat-
edly cites Philips' study (1972, 1974) of the relation between
Madras, Oregon classrooms and the Warm Springs Indian reserva-
tion culture, not only because it is good but also because it re-
mains unique. -
Most educators would agree with the principle that teaching
should start where the child is. Few appear to recognize that to
do so requires knowledge of the community from which the child
comes. Many teachers would agree with the Office for Civil Rights
that formal tests do not adequately show the abilities and needs of
children. These teachers may not recognize that their own obser-
vations may be skewed, confounding impressions of intellectual
ability with impressions of voice and visual appearance. Observa-
tion needs to be systematic across a range of settings and activi-
ties: in class, on the playground, and at home. The interdepen-
dence between specific settings and a display of abilities and
skills is coming to be recognized as a crucial focus for research.
To start where the child is, then, one needs systematic
knowledge of the verbal repertoire of the child in relation to that
of his/her community: the range of varieties of language, the
circumstances, purposes, and meanings of their use. These can
differ from one community or district to another, and local knowl-
edge is needed. One needs to know the role of speaking, hear-
ing, writing, and reading, in a given language variety and what
it means to do each of these activities. These can differ from
one place to another, too, and local knowledge again is needed.
One needs to know, in short, the locally relevant ways of speak-
ing (using "speaking" as a shorthand expression for all modes of
language use). The organization of language use in a classroom
is but part of a systematic whole, from the vantage point of the
student, and from that vantage point, classroom norms may take
on a meaning not intended or comprehended by school personnel
(Hymes, 1974b).
It is common to think of a choice of one language or another,
one variety or style or another, one genre or another, a mode or
occasion of use of language, as appropriate or inappropriate,
right or wrong. Certainly teachers in classrooms often seem to
think in this way. An essential ethnographic point is to remem-
ber that a choice may have to do with elements in a system of
signs. The language used has to do with more than right and
wrong. It is not only the elements within a language that are to
be understood as signs, as uniting form and meaning or as show-
ing their status by contrast with other signs. Every choice
within a way of speaking--of language, variety, style, genre,
mode, occasion--conveys meaning through contrast with other
Language in Education 107

choices not made quite in addition to the meaning conveyed by

what is written or said. The same is true for the use of lan-
guage; Goffman (1956) has called it deference (what one conveys
about one's attitude toward oneself). Too often one thinks that
one particular choice represents order, anything else, lack of
order. This assumption can keep one from discoveringathe true
order that is present, the system of sociolinguistic signs implicit
in students' communicative conduct. One may wish to change
that conduct--to change that system. One has to recognize it,
be able to interpret accurately what is communicated, in order to
know what one wishes to change. (This is an example of the
relation between the descriptive and critical aspects of ethnogra-
phy.)
The point applies both to rules of language and to rules of
the use of language. Let me say a little more about the latter
first in order to highlight their importance.
A teacher or curriculum may make an assumption--about the
role of language in learning, about the etiquette of speaking,
listening, writing, reading, about getting and giving information,
getting and giving attention in talk -that is at variance with what
students experience elsewhere. Variance in itself of course is
simply a fact. Whether or not it is a problem depends on the
situation. If a' classroom pattern is accepted and respected by
all concerned, it may succeed, at least as success is defined by
those concerned. (Their goals of course may not include equality
or social change.) The pattern may also stamp what is learned in
the classroom as appropriate only in similar settings.
Schools have long been aware of cultural differences, and in
recent years have attempted to address them, rather than punish
them. Too often the differences of which the school is aware, of
which even the community is aware, are only the most visible,
"high" culture symbols and the most stereotyped conventions.
What may be slighted is the "invisible" culture (to use Philips'
title), the culture of everyday etiquette and interaction, and its
expression of rights and duties, values and aspirations, through
norms of communication. Classrooms may respect religious belief
and national custom, yet profane an implicit ceremonial order hav-
ing to do with relations between persons. One can honor cultural
pride on the walls of a room yet inhibit learning within them.
One may find children fitting classroom expectations, but in a
way that defeats the purpose of their being there. Some Anglo
teachers in Philadelphia schools have been delighted to have
Spanish-speaking children in their classes because the children
are so well behaved--that is, quiet. The reason for the quietness
is that the children do not understand what is being said. When
they do understand (after being placed in a bilingual classroom,
for example), they participate actively. The equation of being
good with being quiet implies a further equation with being dumb
(both senses of the word). Communities, of course, may differ
108 Ethnographic Monitoring
in the meanings and normal occasions of silence, something that
needs to be known in an individual case.
Knowledge of the local repertoire of varieties of language is
obviously essential. In terms of the descriptive contribution of
ethnography, a salient concern is the local meaning and interpre-
tation of the joint use of both Spanish and English forms. Of
course, instances of the occurrence of elements of one language
in the context of another may be quite ad hoc, due to the fa-
miliarity or the forgetting of particular words. But some may
consider mixing of any sort reprehensible, and especially if it is
extensive. Others may recognize in extensive mixing a special
style of speech, appropriate to certain people and situations.
They may find not a failure to keep languages apart but rather a
skill in mingling them, one that has to be learned. Someone who
knows Spanish and knows English may not know how to mingle
them in the special style. From one standpoint, then, error and
eonfusion are something to be stamped out; from another, a skill
with social meaning to be enjoyed.
The significance of mixing cannot be judged without knowl-
edge of the community and individual norms, but the knowledge
is not an end to the matter. The temptation of descriptive eth-
nography is to let understanding imply acceptance, but ethnogra-
phy can be used critically. In a given case, a community may
not have been conscious of some aspects of its pattern of language
use, and it might wish to reject some part when brought to its
attention. Or a community may decide that a change of pattern
is desirable, even necessary. Or it may decide to accept and
value a pattern previously little noted. Whatever the case, the
goals of bilingual education should be informed by ethnography
but set by those affected. The most difficult issue may be to
analyze and assess information bearing on choice of linguistic
norm and on its implementation.
It is clear that there is an issue. There have been class-
rooms in which the native speakers of Spanish were seated in the
back of the Spanish class and the Anglo students in front. There
is a, school in West Chester, Pennsylvania, whose Spanish-speaking
Colombian teachers are certain that their Spanish-speaking Puerto
Rican students need to be taught correct (Colombian-flavored)
Spanish, ignoring a demonstration that the students are able to
read newspapers published in Madrid. In both cases, a different
classroom norm would yield different results.
There well may be a genuine problem of inadequacy of compe-
tence in certain cases. Puerto Rican children raised in New York
City may be disadvantaged in schools in Puerto Rico. Children
growing up in a particular community may not acquire the full
range of varieties and levels of Spanish, especially if Spanish has
been part of a stable multilingual situation as a language of the
home rather than of education. Ethnography can help discover
the facts of local situations in this regard. Whatever the facts,
 Language in Education 109
difficult matters of analysis and assessment remain. At this point
the critical, comparative use of ethnographic knowledge becomes
essential.
Let me address the issue of linguistic norms--of which to use,
accept, and reject. Some leaders in bilingual education have been
heard to say there might be a danger of perpetuating through
Spanish the failures of schools conducted in English. We are in-
deed familiar with the kinds of misperception and misconception of
ability that can be fostered by prejudicial attitudes towards varie-
ties of English different from the variety assumed in the classroom.
Is the problem of attaining equality tiien simply one of such differ-
ences? Could the problem be solved by eliminating the differences,
either by stamping out all but the preferred norm, or by substi-
tuting or adding one common norm in the repertoires of everyone?
I think that such an approach fails, and fails necessarily, in
the United States, if the problem is indeed defined as one of elimi-
nating linguistically-defined inequality. If every user of English
in the United States used certifiably standard English, in recog-
nizably middle-class ways, little would change except for the cul-
tural impoverishment, the loss of diversity, and interest in Ameri-
can ways of life. Fault would continue to be found. People who
do not use double negatives may be found redundant in their ad-
jectives. People who do not misuse tenses can be faulted for their
use of adverbs and conjunctions.
Variability and evaluation of usage are indeed universal in
human life, but the issue here is not one of individual differences
--of the variation in personal ability inevitable in any group. If
such were the case, it would be a matter of talent; as it stands,
it is a matter of shibboleths. If linguistic discrimination is a cul-
turally deep-seated way of maintaining social distinctions, then
discrimination, is likely to continue. If not through English against
other languages, then through one dialect of English against
others, or one style of standard English against others. What
Barth (1969) has shown for ethnic boundaries holds for class
boundarieS as well: even slight and infrequent features will serve
as boundary markers, if there must be boundaries.
One can further suggest that the United States is culturally
organized to produce continually the appearance of a "falling rate
of correctness," of a "law of increasing illiteracy." There is not
only the constant reproduction of linguistic inferiority, but also
the constant renewal, of markers of it. There is also the constant
projection of decline. This last one draws on a disposition to
interpret change in language (itself inevitable) as inevitably for
the worse. From this standpoint, necessary distinctions are al-
ways being lost, never gained; etymology condemns vitality.
Intelligibility is so often found wanting that one must infer that
for the vast majority of people, talk is nothing more than a verbal
blind man's bluff.
It is a curious thing that a country whose civic ideology has
been so committed to "progress" should so despair of language.

124
110 Ethnographic Monitoring
I suspect two complementary attitudes and interests to be at work:
a widespread popular distrust of verbal skill and an elite's defi-
nition of verbal skill as something only it can have and so con-
trol. It is perhaps an interaction between these two forces that
produces a phenomenon such as a President careful to explain to
an audience that he did not know a word he used and had to look
it up in a dictionary, while expecting to be trusted to manage a
vast bureaucracy that lives and breathes with the manufacture and
manipulation of esoteric discourse.
The role of language in the maintenance of cultural hegemony
in the United States has been little explored. The main point-
and this brings us back to the role of schools--is that the United
States would seem to have a culture in which discrimination on the
basis of language is endemic. To achieve equality within a given
language, it would never be enough to change the way people
speak. One would have to change what the way people speak is
taken to mean.
In this regard, one can hardly avoid the thought that a latent
function of schools has been to define a certain proportion of peo-
ple as inferior, even to convince them that they are so, and to
do this on the seemingly neutral ground of language. Language
seems a neutral ground, so long as one can maintain that there is
just one proper norm, and that the schools do their duty if they
provide everyone access to that norm. The language of the norm
is necessary, and everyone has a chance to acquire it (so one can
imagine the reasoning). Any inequality of outcome cannot be the
fault of the school or system, but must be fair and reflect differ-
ences in ability, effort, or desire on the part of students. If it
is pointed out that some students begin unequally, relative to the
norm assumed in the school, the responsibility is assigned to the
student or student's community, for lack of proper language or
even a virtual lack of language at all.
Centers to stimulate verbal communication in infants in disad-
vantaged homes are even now being newly established. From what
I have said, you can see that questions about such centers would
arise. What are the norms of communication, including use of lan-
guage, in such fathilies? What is the evidence that they are in-
adequate? Are they judged to be inadequate intrinsically, or in-
adequate in relation to the assumptions made in local schools as
to the role of language? Is the set of assumptions made in local
schools the only possible set? Is the difference between schools
and homes a difference between normalcy and impoverishment, or
a difference between two ways of doing things, two ways of speak-
ing, each normal in its own setting? Is the program of such a
center to change the culture of the home something we would call
cultural imperialism if it were reported from the Soviet Union? Is
change of the culture of the homes the only option? Could the
differences be tolerated, or the schools change? These questions
are not rhetorical. Reasonable people might arrive at different
answers, given different situations. The essential thing is that
 Language in Education 111

such questions be asked. Given the many differences among

societies in the role of language in child-raising, yet the unfail-
ing success of children in acquiring language, together with the
norms of use appropriate to their society, it is doubtful that any
viable community needs a program of verbal stimulation for its
children for its own successful continuation. A program, of ---
course, may be a way of changing the community or of diverting
the children from it.
Bilingual education challenges the very fabric of schooling
insofar as it adheres to the goal of overcoming linguistic inequal-
ity, by changing what happens in schools themselves. But if
linguistic discrimination is a culturally deep-seated way of main-
taining social distinctions deeply embedded in educational institu-
tions, is bilingual education likely to escape its influence? I have
suggested that the form of attention to language in schools serves
to maintain social stratification, and as long as' the society re-
quires such stratification, it is likely_to find ways to reproduce
it linguistically. The society is defined as one of opportunity,
yet the relative distribution of wealth and class position hardly
changes year after year, decade after decade; language plays
some part in accomplishing and legitimizing that result. Is suc-
cess for bilingual education then to mean that the accusation,
"That's not Spanish," will be heard as widely as the accusation,
"That's not English?" Or that children who know varieties of
Spanish other than the norm adopted for a classroom will bear
the stigma of not knowing two languages? (One hears of teachers
saying to a child, "I thought your problem was that your lan-
guage was Spanish instead of English; now I find out that you
have no language at all.")
The issue of linguistic norms is inescapable within the level
of an established standard itself, because of the distinctness of
Cuban, Puerto Rican, Mexican, and other national standards
represented in the United States. It is inescapable as well with
regard to the relation between national and regional standards,
on the one hand, and the other components of the verbal reper-
toires of Spanish speakers in this country. The issue of norms
involves the verbal repertoire as a whole. What is the desired
role of each component of the repertoire, and what is the attitude
toward each? As I have indicated, ethnography can assist in ob-
taining the initial knowledge needed to determine the present
state of affairs. Clearly, ethnography is needed instead of
questionnaires and surveys. What to make of the knowledge
provided by ethnography is a matter for the bilingual
to decide. Perhaps in a given case it may be decided community
that an
insistence on standard Spanish is necessary in order to maintain
the language. It is also entirely possible that some may decide
to reject bilingual education and Spanish as necessary ingredients
of, say, Puerto Rican identity, spurred in their decision perhaps
by elitist decisions as to the norm within Spanish (see Language
Policy Task Force, 1978).
112 Ethnographic Monitoring
Let me make clear that I do not mean to imply that all evalu-
ation of language and usage is merely social bias. The point is
that social bias infects evaluation. It is not the case that "any-
thing goes," but it is also not the case that there is a single,
homogeneous, unquestionable norm. The existence of a norm is
a social fact, but not a fact beyond critical analysis in the light
of knowledge of other norms, of the effects of the norm in ques-
tion as it is implemented, of alternative relations between linguis-
tic norms and ways of life.
There are normative criteria that apply to languages and
their use, e.g., criteria of clarity, elegance, pithiness, musical-
ity, simplicity, and vigor. The difficulty is that people differ
in the criteria to which they give most weight, even within the
same community, let alone between communities. People may differ
in what they count as satisfying a norm on whose importance they
agree. Much of the history of language policy and attitude, much
of the history of .linguistic research itself, can be related to alter-
nate attitudes towards the existence and character of two broad
classes of norm, "standard" and "vernacular." In general, social
bias affects willingness to recognize the presence and legitimacy
of a norm in the first place and the interpretation placed upon
meeting or failing to meet it. Is adherence to a "standard" ele-
gance or pretentiousness? Is it logical regularity or empty form?
Is adherence to a "vernacular" revitalizing or corrupting?
Natural or uncouth? Is it an expression of the spirit of the folk
or of the spawn of the uneducated?
Differences in pronunciation are stigmatized as stupidity the
world over; absence of features of grammar is taken as absence
of logic; propriety of diction identified with virtue. Such inter-
pretations of the speech of others are frequently arbitrary. The
association between a feature of language and a feature of intelli-
gence or character is generally not inherent and universal, but
local, secondary, and projected.
Within one's own linguistic tradition, one may be on surer
ground in assessing the speech of those who share it, but it is
ground that cannot be made more secure than the tradition itself.
One can judge others (and oneself) in relation to known norms
but not withhold the norms themselves from scrutiny, if their
consequences cause them to come into question. Despite their
pervasiveness and familiarity, the norms may be secondary and
projective. We may honor them because through them we have
experienced so much that is inseparable from our own naming and
knowing of life: satisfactions and illuminations even that mastery
of a norm may sometimes permit. Even so, we have to ar,..cept that
similar experiences may occur in relation to norms we can hardly
recognize as such. There must be norms, if there is to be
mastery, whether of interactional wit or composed art. But the
norms themselves are not fundamental, I think. What is funda-
mental is that which the norms make possible, the functions
served in creative, resourceful, adaptive, and expressive uses
Language in Education 113

of language. Many norms can serve those functions, and a given

norm can be made into an enemy of them.
We want to ask, then, not if the norm is observed but what
is accomplished through observance of the norm? Is it desirable
to spend a term insisting that a child be perfect in a minor gram-
matical feature, if the result is to teach that child that the norm
of which the feature is a part is a torture chamber? Or if the
child is inhibited from ever attempting to use that norm resource-
fully?
We recognize that there are universal capacities for the struc-
tures and functions of language shared by all normal human be-
ings; that a degree of individual variation in ability is inescapable;
that a degree of normative stability is essential to the possibility
of reliable communication and expressive mastery; that mastery of
features may facilitate their resourceful use, but that the same
features, treated as shibboleths, may inhibit resourceful use of
language. The functions of language are fundamental, the forms
instrumental. Quite literally; the letter killeth, but the spirit
giveth life.
The goals of a community of course may not be to encourage
creative and resourceful use of language. The goals may be to
ensure that persons can be placed by the way they speak and
write. Or to ensure that persons can perform useful work, can
read instructions, newspapers, and other communications from
those who direct things. (It is perhaps instructive that our
society defines reading, a receptive ability, as its main concern,
not the productive ability to write.) Insofar as a community both
says and means that resourceful, creative use of language is a
goal for all its citizens, then questions of norms, and questions
of pedagogy too, must be decided in favor of an emphasis on
function as primary;- form as instrumental. This is not to ignore
the one in favor of the other but to recognize which of the two
will bring the other in its train.
If this view is accepted, then the task of ethnography is
both indispensable and difficult. It is not enough to discover
what varieties of language are in use, when and where and by
whom, what features of language vary according to what param-
eters. One has to discover what varieties of language, features
of language, are being used for, and to what effect. Is the
choice of one norm over another the choice as well of certain
functional possibi7 ties as against others? Let me cite the circum-
stances of many Native American communities, which have ac-
quired English but not the literary glories that English depart-
ments like to cite, while having lost rich literary traditions of
their own. These communities have English instead of some Native
American language, insofar as it is a question of language alone.
They have been impoverished insofar as it is a question of the
functions of language.
The issues and choices are difficult. I only hope to have
shown that the knowledge one needs in order to deal with them
114 Ethnographic Monitoring
is ethnographic in nature. This is true, not only with regard
to initial knowledge, but with regard to the monitoring of ongoing
programs and of outcomes.
Whatever the strategy of a program, those who direct it obvi-
ously benefit from feedback during its course. Test scores and
other classroom results may give some indication of the progress
of students. Even with regard to what is learned alone, obser-
vation is desirable as well. Students may show abilities in peer-
group interaction and other settings that do not appear on tests.
Insofar as the program is concerned with the general development
of the students, and with the success of bilingualism itself, eth-
nographic observation is essential. A central question will be:
what does it come to mean to succeed, or to fall, in the program?
What does it come to mean to do well or poorly?
Perhaps some of the meaning will have been clarified through
the assignment of students to locations in the classroom or to
other groupings. Studies by Rist (1970), McDermott (1974), and
others have shown the importance for success and failure, and
for social meaning of success and failure, of teacher-assigned
groupings.
Some of the meaning of the program to its students, and to
the community from which they come to school, will emerge in
interacts; -,n outside of school. Peer-group discussions and judg-
ments, family discussions and judgments, community perception
of the purposes and consequences of the program, need to be
known and taken into account. While the program is teaching
language, it will also be creating social definitions and judgments.
These definitions and judgments may be as important to success
for bilingualism in the country as formal instruction.
A great deal can be accomplished by establishing regular
community participation in the guidance_of_the program. Still ,
-

no one is a perfect or even adequate ethnographer of him/herself,

if engaged in observation, comparison, and inference only ad hoc.
It would be a valuable element of the monitoring of a program
during its operation to have one or more persons formally re-
sponsible for ethnographic observation and inquiry.! What group-
ings emerge in classrooms, playgrounds, or elsewhere? Does use
of language change outside of class during the course of the pro-
gram, and if so, in what ways? Does conversation about language
change? What is said about the program, about those who suc-
ceed better than others, about those who do less well? Even more
meaningful is discovery of what is presupposed in what is said- -
what comes to be taken as shared assumption in terms of which
specific remarks are to be understood, e.g., that a student who
does well is a teacher's pet, that a student who does poorly is
stupid, that only students from a certain class or neighborhood
or kind of family do well, or do poorly.
Ethnographic monitoring need not be conceived as an isolated
task. The staff of the program and representatives of the com-
munity could participate valuably, if one or a few people were
 Language in Education 115
responsible for coordinating information, for providing initial
orientation as to the kinds of observation needed, and, indeed,
for listening to learn the kinds of observation that might not have
been initially thought of. A much higher degree of validity might
be possible through cooperation. Since the purpose of the eth-
nography is to aid the program, its result must be communicated
to the participants in the program in any case. It is far better
to have the communication as an ongoing process throughout the
program. An additional benefit may be to share ethnographic
skills that participants in the program will be able to use in other
circumstances.
The greatest value of cooperative ethnographic monitoring is
that the participants in the program will have the firmest grasp
possible of the working of the program, of its successes and
failures, strengths and weaknesses,
it. They will not be-in-the position in relation to their hopes for
of being confronted by an
outside evaluator's charts and tables, and told a rating for their
program, with nothing to say, or nothing, at least, that such an
evaluator feels required to heed. The participants will not have
been bystanders. They will have concrete knowledge of the pro-
cess of the program, and be 617.-..,:=4 to address the processes that
have produced whatever statistics and graphs a formal evaluation
process may yield. An evaluation in terms of
only guess at what produced the numbers, andgross numbers can
indeed, can only
guess as to whether its numbers were obtained with measures
appropriate to what is being evaluated. The participants in co-
operative ethnography may benefit from having their cumulative
observations and interpretations compared with independently ob-
tained measures. Both kinds of information could be combined to
provide a deeper understanding. But if measures are to mean
anything, especially in relation to bilingual education as a process
of social change, the ethnography is essential.
All this is the more important, if we look ahead, and think of
the monitoring and assessment of individual programs as contribut-
ing to judgments likely to be made a few years from now as to the
success or failure of bilingual education as a national policy.
A few years from now the charge is likely to be made that
bilingual education has failed. Money was spent, little was accom-
plished--it is easy enough to predict what will be said if bilingual
classrooms join busing and poverty programs as targets of resent-
ment.
The political strength of those who support bilingual education
may be great enough to offset the pressure of those who will make
such charges, once the first wave of support and funding has
crested. And much can be said to deflate the prejudice that may
lie behind such charges. From the standpoint of what is known
about languages and their uses, it is clear that bilingualism can
be an entirely taken-for-granted aspect of a society, something
entirely within the normal capacity of individuals. The list of
flourishing bilingual, even multilingual, situations throughout the
116 Ethnographic Monitoring
world is long indeed. The evidence that human beings can
readily acquire a range of varieties of language is so clear that
the question must be, not, is it possible, but where it does not
happen, what prevents it?
Arguments from suspicion of bilingualism in general, then,
can be won. Arguments from the situation of bilingualism in the
United States may be more difficult. Arguments will likely raise
two issues: educational success, and political consequences. As
to educational success, it can be pointed out that a few years is
hardly enough to overcome the consequences of generations of
effort to impose monolingualism in schools. And insofar as suc-
cessful programs require research, there has been little accumu-
lated knowledge on which to build. Most linguists have been as
blind to the importance of the linguistic diversity of the country
as anyone. They too have proceeded as if knowledge of English
alone would be sufficient. Far too few scholars of other disci-
plines have been helpful. Only in recent years has any substan-
tial number of anthropologists thought ethnographic research in
their own country legitimate. As we know, bilingualism has been
made a vital issue through social, political, and legal processes.
These have led the way. Research, by and large, has only be-
gun to follow. Insofar as successful programs require accurate
initial knowledge of the situations in which they operate, and
appropriate methods for assessing the communicative competence
of students, they have had little on which to draw. Bilingual
education may be accused of having failed before it has been
fairly tried, if to be fairly tried means to have the support of
the kinds of knowledge and methods indicated.
In this regard, the ethnographic monitoring of programs can
be of great importance. The circumstances and characteristics
of successful results can be documented in ways that carry con-
viction. Unsuccessful efforts can be interpreted in the light of
the conditions found with success. Attempts to argue that bi-
lingual education as a whole has been a failure in the United
States can be countered by getting down to cases and knowing
well what the cases are. To do this requires confidence in the
kind of knowledge that ethnography provides, a willingness to
accept the legitimacy of the conclusions arrived at by cooperative
ethnographic observations and analysis, if such conclusions differ
from formal tests and measurements. 'I think we frequently accept
the legitimacy of understandings of our own that are ethnographic
in nature, as against statistics that run counter to our personal
knowledge. I think we should do so; to do so is essential to a
democratic way of life. But it is necessary to admit that we do;
only by admitting that we do can we proceed to go beyond im-
pressions and attain the validity of which an ethnographic ap-
proach is capable.
Some will argue against the political consequences of bilingual
education, claiming that it is divisive. There is a general answer
to this, of course; the social meaning of languages is not inherent
Language in Education 117

in them, but a consequence of the uses to which they are put.

Where languages are symbols of division, it is because of social
forces that divide and pit people against each other along lines
that coincide with language boundaries. Difference of languages
is hardly necessary; a single sound will suffice, as the Biblical
example of the killing of those who said shibboleth instead of
sibboleth indicates. The greatest internal conflict in the history
of the United States, the Civil War, was not fought in terms of
language boundaries. On the other hand, there are many areas
in which multiplicity of language is in no way a part of social
mobilization and conflict. In sum, small differences can become
symbols of hostility and large differences can be accepted and
ignored. The causes are outside of language.
To be sure, a given language policy may favor some interests
as against others. Bilingualism may be experienced as a burden
by people who have been able to assume that theirs was the only
language that counted, that their convenience and the public inter-
est were the same. But to argue that bilingualism is divisive is
really to argue that it makes visible what one had preferred to
ignore, an unequal distribution of rights and benefits. It is
common to call "political" and "divisive" the raising of an issue
that one had been able to ignore, and to ignore the political and
oppressive implications of ignoring it. In this regard, the ethno-
graphic monitoring of programs can also be of great importance.
The ethnographic approach can go beyond tests and surveys to
document and interpret the social meaning of success and failure
to bilingual education.
It may be that some years from now those who work for bi-
lingual education will not themselves be of one mind about its
role. One view of the relation between such movements and
general social processes is that they represent a phase of the
interdependence between an expanding world economy and locally
exploited groups. It may not be presently clear to what extent
the movement for bilingual education is a recognition of sheer edu-
cational necessity, an expression of a phase in the relation be-
tween a minority group and forces dominant in the society, an
expression of a growing commitment to the ideal of a multilingual/
multicultural society. It may be that some sectors of the Spanish-
speaking community will argue for intensive English training as a
preferable route to economic opportunity while others argue for
Spanish maintenance programs on the grounds of cultural identity.
Class differences may appear in this regard.
My own belief is that a multilingual society is something to be
desired and maintained, but it is for others to decide their own
interests. Whatever the ultimate policies decided upon, the wis-
dom of those choices will be greatly enhanced if ethnographic
monitoring has been an integral part of bilingual education.
118 Ethnographic Monitoring
REFERENCES
Barth, Frederick. "Ethnic Groups and Boundaries." The Social
Organization of Culture Difference, ed. Frederick Barth.
Boston: Little, Brown & Co., 1969.
Goffman, Erving. "The Nature of Deference-and- Demeanor."
American Anthropologist LVIII (June, 1956), 473-502.
Hymes, Dell. "Introduction." Functions of Language in the
Classroom, eds. Courtney B. Cazden, Vera P. John, and
Dell Hymes. New York: Teachers College Press, 1972,
pp. xi-lvii.
, ed. Reinventing Anthropology. New York: Vintage
Tess, 1974a.
. "Ways of Speaking." Explorations in the Ethnography
of Speaking, eds. Richard Bauman and Joel Sherzer. New
York: Cambridge University Press, 1974b, pp. 433-451..
Language Policy Task Force. "Language Policy and the Puerto
Rican Community." Bilingual Review (February, 1978). (An
earlier version was circulated as "Toward a Language Policy
for Puerto Ricans in the United States," by E. Gonzalez
Atiles, P. Pedraza, and A.C. Zentella.)
McDermott, Ray P. "Achieving School Failure: An Anthropologi-
cal Approach to Literary and Social Stratification." Educa-
tion and Cultural Process: Toward an Anthropology of Edu-
cation, ed. George D. Spindler. New York: Holt, Rinehart
and Winston, 1974, pp. 82-118.
Philips, Susan U. "Participant Structures and Communicative
Competence: Warm Springs Children in Community and Class-
room." Functions of Language in the Classroom, eds. Court-
ney B. Cazden, Vera P. John, and Dell Hymes. New York:
Teachers College Press, 1972, pp. 370-394.
. "The Invisible Culture." Unpublished PhD dissertation,
irversity of Pennsylvania, 1974.
Rist, Ray C. "Student Social Class and Teacher Expectations:
The Self-Fulfilling Prophedy in Ghetto Education." Harvard
Educational Review XL (August, 1970), 411-451.

133
 educational ethnology'

One hears a good deal about "ethnography" in education today,

but not about "ethnology." I use "ethnology" in my title to call
attention to issues that the
leave obscure. These issuesuse of "ethnography" alone might
have in common the theme: What
would knowledge of schooling in the United States be like, if
anthropologists regarded it more like knowledge of kinship,
chieftainship, religion, technology, and the like, in areas of the
world in which anthropologists have worked long and intensively?
Two caveats: I do not wish to idealize anthropological study
of Native American kinship, African chieftainship, South Asian
religion, Oceanic economic life, and the like. In focusing on
schooling, I do not _wish to forget that an anthropological per-
spective on education is broader than schooling, and necessarily
seeks to understand schooling as one mode of learning and insti-
tutionalization among others.
It remains that educational research in the United States does
focus on schooling, and that it would be different were an
anthropological perspective to be thoroughly established. The
difference would be a benefit, I think, to schools, to anthropology,
and potentially, to democratization of knowledge and to the rela-
tions between academic centers of research and the communities
of which they are part. The difference to research can be sum-
marized in three words: cumulative, comparative, cooperative.
Understanding of individual schools is not now cumulative,
but for the most part a matter of in and out. An individual
school does not seem to count as a legitimate object of long-term
study. 2 Boas is known for life-long study of the Kwakiutl, others
are associated with other groups, but anthropologists do not seem
to be associated with long-term involvement with particular schools
or school systems. That makes the anthropology of schooling
odd. One way to describe anthropology is to say that it has seem
divided the world into names--names of peoples, languages, cul-
tures- -that it has made legitimate objects of knowledge. There
are bibliographies of such knowledge, organized in terms of such
names. If I discover an additional fact about such a unit, I can
publish it as a legitimate addition to knowledge. My first publi-
cation, indeed, resulting from a first summer of field experience
as a graduate student, was such.: "Two Wasco motifs." I had
been able to record incidents missing from the texts collected a
half-century earlier by Sapir, and a helpful professor encouraged
me to write them up and send them off.

119
120 Educational Ethnology
It is hard to imagine publishing "Two Longstreth classrooms"
on the grounds that an earlier study had overlooked the two.
(For "Longstreth," the name of a school in West Philadelphia,
substitute the name of any school near where you live.) Schools
do not seem to be thought of as objects that it might take a long
time, many hands, and even more than one generation, to come to
understand. Individual schools do not seem to be thought of as
individual in character. They are thought of, perhaps, as
"urban," "inner-city," or the like, but mostly they seem to be
thought of as equivalent settings for the interaction of certain re-
current variables--principals, teachers, pupils, curricula, and
methods of instruction. Comparative perspective seems to be a
matter of differences on such variables, together with demographic
data, test scores, and the like. There does not seem to be a
comparative perspective in terms of any, say, of the integrative
approaches to the little community that Redfield sketched years
ago, or in terms of any other dimensions found in the anthropology
of other areas. 3 Such comparative perspective, of course, would
depend on knowledge of sociocultural context.
I do not claim to know educational research well, and apolo-
gize to those of you who do, but it is my impression, so far not
corrected by any to whom it has been mentioned, that educational
research does focus on the testing of relation among variables
without much regard to sociocultural context. Specialists in edu-
cational research have been heard to lament its inconclusiveness
on various points: ten studies, say, show a positive relationship
between two variables, and 17 studies show a negative relation-
ship. The obvious implication, it would seem, is that there may
be two types of school, and that the variables under study were
interacting in two different types of systematic contexts. That
is a familiar anthropological point, from speech sounds to avoid-
ance behavior and presence of belief in a supreme being. But, I
am told, it is usually difficult to recover from educational studies
the information about context that would enable one to character-
ize it.
If this impression is correct, then knowledge of schools in the
United States is about one hundred years behind knowledge of
American Indian kinship. The pioneer work of Lewis Henry Mor-
gan brought together systematic information and proposed a broad
typological dichotomy in 1871. Since that time, a galaxy of names
--Kroeber, Lowie, Leslie Spier, George Peter Murdock, Driver,
Lounsbury--have contributed to the more precise identification of
relevant features, dimensions, processes, and types. We seem
nowhere near the identification of what would correspond to Omaha,
Crow, Dakota, Iroquois, etc.. types of school. Note that such
types are not pigeon-holes, but bases for systematic analysis of
particular structures and processes of change.
We do not have such bases for analysis of structure and
change in American schools. Yet it cannot be the case that all
schools in the United States, or even in a single city, are the
133
 Language in Education 121
same. Nor, on the other hand, can it be the case that each is
entirely unique. In a city, or the country, it seems a reasonable
able, fundamental question: What kinds of schools are there?
I suggest that we do not know because there has not been
enough anthropology in educational research, and because educa-
tional anthropology has understandably been concerned with estab-
lishing its ethnographic mode of work. In education by and
large, "anthropology" has come to mean "ethnography," and
"ethnography" has come to mean field work: participant obser-
vation, narrative description, and the like. I believe strongly
in this mode of work, and will take it up briefly at the end.
Here I want to stress the danger of letting the
perspective on education become equated in otheranthropological
minds with just
a mode of field work. The result will be dozens of people called
"ethnographers" because they have observed, although with little
or no training in cultural analysis; attempts to insert "ethno-
graphic components" in helter-skelter research designs; a brief
vogue for the name "ethnography";
immunity to the true challenge of anand at the end a heightened
ethnographic, anthropologi-
cal mode of thought.
Ethnography, as we know, is in fact an interface between
specific inquiry and comparative generalization. It will
well, I think, to make prominent the term, "ethnology," serve
that
us
plicitly invokes comparative generalization. And it will serve ex-
schooling in America well. An emphasis on the ethnological di-
mension takes one away from immediate problems and from attempts
to offer immediate remedies, but it serves constructive change
better in the long run. Emphasis on the ethnological dimension
links anthropology of education with social history, through_ the
ways in which larger forces for socialization, institutionalization,
reproduction of an existing order, are expressed and interpreted
in specific settings. The longer view seems a surer footing.
Let me come at the matter in terms of a map. If Harold Driver
had tried to map North America in terms of what is known about
its schools, would not most of the map be empty? If so vacant a
map dealt with kinship, would there not be a demand for studies,
an unwillingness to talk about "Indians" in the abstract until more
were known in the concrete? The need for cumulative, compara-
tive research would be obvious, as it is obvious today in a less
explored region such as New Guinea.
Some may object that such
is based on a mistaken analogy.a call for an ethnological approach
Schools and cultures, say,
Longstreth in West Philadelphia, and Wasco in Oregon, are not
alike. Schools are less autonomous and more subject to change,
even manipulation, by external forces.
an elementary school are not comparableAn Indian community and
units. I think the
analogy holds, despite the differences. First, our sense of each
named Indian community as a distinct entity is partly an artifact
of our own profession. The academic requirement of
to knowledge through research, and the need to have contribution
a
I

122 Educational Ethnology

contribution of one's own, contributes to a tendency to differ-
entiate the anthropological world into named entities that anthro-
pologists can claim. Second, the similarities among differently
named Indian communities may be great; their contemporary cir-
cumstances on a reservation may be greatly shaped by external
forces. Common ecological base, diffusion, pan-tribal movements,
retention of elements of ancient tradition, and orientation common
to many, all make the autonomy and distinctiveness of named
Indian groups an empirical question, just as is the case with
named schools. If one school has changed drastically in a decade,
through change in the population served, another a mile or two
away may have not. An essential dimension of comparative re-
search might be the continuities that schools have, and are felt
to have. Third, the issue of access to the research site may be
very much the same today. With both schools and reservations
one has to address suspicion rooted in past experience; concern
about exposure and embarrassment; demands that research and
the researcher be useful to those studied.
In regard to access, anthropology encounters the same prob-
lems with American schools that it encounters throughout the
world, and has, perhaps, a better chance of solving them.
Sustained cooperation can serve both parties, academics, and
schools. To say this is not to overlook the conflicts of interest
that are latent; not everyone wants everything known, or even
to know certain things at all. But the problems seem intrinsic
to research, not to schools.
A leading element of sustained cooperation is the involvement
of others in the research. Here a major anthropological tradition
can be an essential asset. In the study of a language, a kinship
system, or the like, one is to a great extent seeking to make ex-
plicit in a comparable framework what others, in a certain sense,
already know. Speakers of a language, participants in a network
of kinship, are not merely objects, but, as sources of information,
party - in inquiry as well. This tradition suggests that the
;.vi,..opriate strategy for school personnel who seek advanced de-
grees is to capitalize on what they know where they are. Often
enough they have been made to believe that a legitimate contribu-
tion to knowledge, and advanced degree, requires methodology
and subject-matter disconnected from their experience. Anthro-
pological tradition suggests that they can capitalize on their ex-
perience, and make a far more valuable contribution to knowledge
by doing so.
If the map of American schooling is to be made less empty,
after all, it will require more than the anthropologists and anthro-
pological funding available. Just so, knowledge of Native Ameri-
can traditions would be far less rich, given the small number of
anthropological investigators, were it not for members of those
cultures George Hunt, William Beynon, and others--who became
contributors of knowledge themselves.
 Language in Education 123
At the Graduate School of Education at the University of
Pennsylvania we have such a program, now in its third year,
involving a group of principals mainly, together with a few other
school personnel. It is still too early to judge the research out-
come, but indications are encouraging. It is clear that the pro-
.

gram could not have begun, let alone prospered, had the nature
of degree requirements not been rethought in terms of principles
of cooperation, indeed, partnership. Past experience, distrust,
uncertainty of mutual benefit had to be overcome gradually. The
anthropological principles were heard but only gradually believed.
Had there been insistence on the model previously familiar to the
principals, the isolated researcher carrying out an experimental
design, the program would never have been possible. One essen-
tial aspect of the program has been training in observation and
narrative reporting, concerned with better perceiving and express-
ing the process of implementing a reading program in each person's
school. Another essential aspect has been grouping researchers in
teams. The grouping provides moral support among people who
are pursuing a degree above and beyond a full-time job. Ideally,
rather than "cheapening" the degree, it provides for deeper in-
sight. The grouping builds controlled comparison into the pro-
cess. Finer perception of similarities and differences among
schools results.
These principals, and members of other school districts in the
region, perceive this approach as taking their circumstances and
needs into account. They see such a program as evidence that
an "Ivy League" university, reputedly disinterested in less than
elite affairs, is in fact responsive. We in the faculty see such a
program as a golden opportunity. There is little chance that we
could find funding to study fifteen or thirty, schools simultane-
ously. Even given funding, .there is little chance that the schools
would let researchers in just because they knocked. The princi-
ple of cooperative ethnographic research leads to the schools and
the University having investments in each other.
This short account highlights the positive, ignoring the
tremors, misunderstandings, partial understandings; it ignores
the fragile dependence of the effort in the University itself on
adjunct faculty. Yet it is fair to say that the effort has led to
an atmosphere in which an anthropological approach is welcomed,
indeed so welcome that interest outruns supply, an atmosphere
in which the long-term questions of the role of anthropology in
educational research can be addressed.
Two long-term questions seem to me of especial importance:
What will be the structure of knowledge of schooling? What will
be its form? The importance of the questions lies in their impli-
cations for a democratic way of life. A mode of research that
focuses on experimental design, quantitative techniques, and the
impersonality of the investigator has its place, but, carried to its
perfection, as the exclusive mode, it would tend to divide society
into those who know and those who are known. The

138
124 Educational Ethnology
anthropological recognition of the contribution of the practitioner
as one who also knows counteracts that tendency. So does the
legitimacy, indeed necessity, in ethnographic research of narra-
tive. Good narrative accounting is not easy, and may be harder
sometimes than quantitative analysis, but it is more accessible to
the citizens of society. Moreover, it can be argued that even
quantitative analysis invokes narrative models of social life at
some point, and that institutional decision making certainly does.
Explicit attention to narrative accounts and models can make an
essential contribution. It can legitimu.e the form in which the
knowledge of most citizens as to their circumstances is cast, and
it can make apparent Ei hidden form of cultural hegemony. Of
all the disciplines interested in schools, anthropology is best
equipped to make that contribution. le
To sum up: I have argued that Native America and School
America pose anthropological problems of the same kind. The
question, "What kinds of schools are there?", is naive, yet
natural to an anthropological perspective. An answer draws on
those aspects of anthropological tradition that regard research
as cumulative, comparative, cooperative. A strategy that draws
on such an answer, while looking to the long run, can hope to
serve change and even immediate advantage.
FOOTNOTES
1. This is the text of my address as outgoing presidential
officer of the Council on Anthropology and Education, during the
business meeting of the Council, at the annual meetings of. the
American Anthropological Association, December 1, 1979, in
Cincinnati.
2. An exception is the ten-year involvement in three State
of New York high schools on the part of Francis Ianni and others.
3. Gastil (1975) calls attention to regional cultural differences
insofar as statistical indications permit, but notes the lack of
explanatory power in such a preliminary approach. See Hs dis-
cussion of "The relationship of regional cultures to educational
performance," pp. 116-127.
4. See discussion of narrative in Hymes (1977) and in Caz-
den and Hymes (1978). On cognitive analysis of norms of inter-
action, embodied in an expressive genre, see Hymes (1979) and
the book to which it is a foreword.
REFERENCES
Cazden, Courtney and Dell Hymes. 1978. "Narrative thinking
and story-telling rights: A folklorist's clue to a critique of
education." Keystone Folklore 22(1-2):21-35.
Gastil, Raymond D. 1975. Cultural Regions of the United States.
Seattle: University of Washington Press.
Language in Education 125
Hymes , Dell. 1977. "Qualitative /quantitative research method-
ologies in education: A linguistic perspective." Anthro-
pology and Education Quarterly 8(3):165-176.
. 1978. "What is Ethnography?" Working Papers in
Sociolinguistics 45. Austin: Southwest Educational Develop-
ment Laboratory. To appear in a volume edited by Perry
Gilmore and Allan Glatthorn, and published by the University
of Pennsylvania Press.
. 1979. "Foreword." In Portraits of "The Whiteman,"
by Keith H. Basso. New York: Cambridge University
Press.
narrative thinking and story-
telling rights: a folklorist's clue
to a critique of education
INTRODUCTION: Dell Hymes
This article has its origin in instances of its own subject--use
of narratives to explore and convey knowledge. In the course of
a conversation with Courtney Cazden, I mentioned material re-
corded by Joanne Bromberg-Ross. She had recorded conscious-
ness-raising sessions of a women's group, and presented a por-
tion to my seminar. One session in particular contained
two
a mar-
different
velous demonstration of interdependence between
modes of clarifying meaning. The topic was what was meant by
"strength" in men and *women. Discussion began with discussion
of terms. An unresolved back and forth about terms was followed
by a series of personal narratives. Suddenly definitional discus-
clear that there had
sion returned, stated in a way that made it
been no break in metalinguistic focus. Narrative had (one
solved the
good, one
problem of differentiating two kinds of "strength"
bad), when direct definition had floundered. The second mode
of language use continued the purpose of the first, coming suc-
cessfully to its rescue.
These two foci, terms and stories, often appear to contrast,
rather than to complement each other, as here. My telling of the
example from Bromberg-Ross reminded Cazden of instances of
contrast from her experience at Harvard, which she recounted.
I urged her to write them up, for they highlighted the possibility
that one form of inequality of opportunity in our society has to
do with rights to use narrative, with whose narratives are ad-
mitted to have a cognitive function. Cazden's writtenand account
suggest
follows next. After it, I will cite other observations
some general implications. The most pertinent and obvious impli-
cation for folklore can be stated right off. If differential " at-
ment of narrative experience plays an important role in prt.....ant
educational practice, then folklore can claim a special place in
the study and change of education in this regard.
WAYS OF SPEAKING IN A UNIVERSITY: Courtney Cazden
We who work in universities may find contrasts in ways ..f speak-
ing in our own classrooms. Two personal reports from -graduate
students and one case study of changes in language use over an
undergraduate's four college years point to a particular contrast
between narrative and non-narrative ways of clarifying meaning
(exemplified, I realize, in the following account).
126
 Language in Education 127
One fall recently I gave my class in "Child Language" to two
different student groups: two mornings a week to a class of
graduate students (master's and doctoral level) at Harvard
ate School of Education, and one evening later in the week Gradu-
double lecture to a class in Harvard University Extension. as a
The
latter is a low-tuition, adult education program whose older than
college-age students are either working for a college degree
through part-time evening study or taking single courses for per-
sonal or professional interest. My Extension class had a mixture
of the two groups--degree candidates like the tuna fisherman from
San Diego who works as a bartender while progressing
toward a BA and then law school, and teachers in local slowly
day care
centers, bilingual programs, and Perkins Institute of Helen Keller
fame. Each class knows of the other's existence, and students
have been encouraged to switch when convenient -as an evening
make-up for the morning class, or the chance to experience
"real" Harvard atmosphere for the Extension students.
One evening, I noted two Black students from the Graduate
School in the Extension class. Instead of sitting in a far corner,
they were near the front. Instead of remaining silent, they par-
ticipated frequently in the evening's discussion. Finally, the man
spoke publicly about his perceptions of the difference in the two
classes. I paraphrase his unrecorded comments:
In the morning class, people who raise their hand talk
about some article that the rest of us haven't read.
That shuts us out. Here people talk from their
sonal experience. It's a more human environment.per- 1
I remember a similar contrast described to me two years ago
by a Tlingit woman graduate student from a small village in
Alaska. She spoke about discussion in another course during
her first semester at Harvard. Here the contrast was not only
between ways of speaking, but how these ways were
acknowledged by the professor. Again I paraphrase: differentially
When someone, even an undergraduate, raises a question
that is based on what some authority says, Prof. X says
"That's a great question!", expands on it, and incor-
porates it into her following comments. But when people
like me talk from our personal experience, our ideas
not acknowledged. The professor may say, "Hm-hm,"are
and then proceed as if we hadn't been heard.
In Philips' (1974) sense, contributions to class discussion based
on narratives of personal experience did not "get the floor."
"Michael Koff" came to Harvard College from a working-class
community in Boston. Yearly interviews with him had been con-
ducted at the Bureau of Study Council as part of a, study of the
impact of college experience. Some years later, for a Graduate
School term paper, Bissex (1968) analyzed the transcripts for
128 A Folklorist's Clue to a Critique of Education
linguistic indicators of what she called "the Harvardization of
Michael." She found a cluster of co-occurring shifts between
Michael's sophomore and junior years, including one from for
instance to I mean.
In Michael's sophomore interview, there are twenty-five occur-
ences of for instance and other words used to introduce examples,
compared with ten, three, and four in his freshman, junior, and
senior years. His language as a sophomore is, as he says, "con-
crete": every page of the transcript includes at least one illus-
trative incident, and the last half of the interview is almost en-
tirely anecdotal. These incidents always function to clarify
points. Michael does not trust the "big, vague general words
that do not mean anything"; he trusts the meaning that resides
in concrete experience (Bissex 1968: 11-12) .
One of the things that I developed an interest in over
this past year is some young high school people who
live in a housing project.... Somehow if I wanted to
talk about life in the project, I either said, "Life is
terrible!" or "Life is not too bad." It didn't mean any-
thing. It's easier to, I mean, for instance, just talking
here, it would be easier if I could think of some, some-
thing--some specific instance. (pause) For example,
this family in Larchwood Heights...
Michael Koff, sophomore
Michael's junior interview is marked, in contrast, by twenty-
four occurrences of I mean compared with nine, seven, and four
in his freshman, sophomore, and senior years. "I mean has re-
placed for instance in its function of introducing an intended
clarification of a previous statement. The interesting difference
is in the nature of the clarification during his sophomore and
junior years; the shift from concrete illustration to restatement,
generally on the same level of abstraction as the original state-
ment" (Bissex 1968:16) .
I mean, you just Zook at things differently. I--ah--it's
hard to say what. It's hard, I mean, because you can
only put your finger on some of them. You feel you're
growing up. I mean, certain things become less im-
portant, certain things become more important.... I
mean, the things that you think are important drop out
and new things take their place.
Michael Koff, junior
Although narratives have an honorable history as "the tem-
porizing of essence" (Burke 1945:430), they are often denigrated,
particularly by social scientists, as "mere anecdotes." Evidently
there is a press in at least some speech situations in this uni-
versity to substitute other modes of explanation and justification.
 Language in Education 129
A NARRATIVE VIEW OF THE WORLD: Dell Hymes
Let me try to generalize, or at least extend, Cazden's observa-
tions. We tend to depreciate narrative as a form of knowledge,
and personal narrative particularly, in contrast to other forms of
discourse considered scholarly, scientific, technical, or the like.
This seems to me part of a general predisposition in our culture
to dichotomize forms and functions of language use, and to treat
one side of the dichotomy as superior, the other side as some-
thing to be disdained, discouraged, diagnosed as evidence or
cause of subordinate status. Different dichotomies tend to be
conflated, so that standard:non-standard, written: spoken,
abstract : concrete, context-independent :context-free, technical /
formal : narrative, tend to be equated.
When we think of differences in verbal ability, for example,
many of us think in terms of command of standard varieties of
English, command of the vocabulary, syntax, and written genres
associated with standard varieties. We tend to group standard
norms and verbal acuity together. William Laboves widely re-
printed essay, "The logic of non-standard English," has done
something to change that situation,
of discourse, the cogent flow of onebywith
contrasting two examples
the stumbling of the
other--the cogent discourse being in a non-standard variety, the
stumbling in a standard. Still, it is probably hard for narrative
to get a hearing or approval in our schools, however apt its inner
form of idea, if its outer form pronunciation, or spelling, word-
form and sentence-form, is not ofapproved.
There is a connection here with Bernstein's well known con-
trast between "elaborated" and "restricted" codes. (More re-
cently Bernstein speaks of contrasting coding orientations, each
with its "elaborated" and "restricted" variants.)
that Bernstein calls "elaborated" is associated withThe orientation
as independence of context, objectivication of experience,things
such
analysis
of experience, a kind of metalinguistic potentiality. The orienta-
tion called "restricted" is associated, among other things, with
dependence on context and a taking of pre-established meanings
and values for granted. One suspects that the contrast is in
some respects a version of the older contrast between "abstract"
("elaborated") and "concrete" ("restricted") modes of thought.
Certainly it is Bernstein's view that an "elaborated" orientation
is necessary in order to go beyond the socially given. This is
part of his defense against charges of favoring the middle-class
and putting down the working-class: an elaborated orientation is
necessary for the kind of analysis that could lead to a transfor-
mation of the condition of the working class.
have taken up the notion of a link between an Other sociologists
"elaborated" code
or orientation, and a radical social perspective, taking the one
to be a condition of the other (Mueller 1973 and Gouldner 1975-6).
Now, if one applies Bernstein's contrast to everyday genres,
then one is likely to take written communication as "elaborated,"
130 A Folklorist's Clue to a Critique of Education
as against spoken. (Various writers have done so.) A main
basis is the assumption that written communication is ipso facto
context-independent. That assumption, of course, is false. Our
traditional stereotypes about the functions of writing perpetuate
it, but an empirical examination of the uses and interpretation of
writing would falsify it. A written document may be dependent
on knowledge of non-linguistic context for its interpretation just
as speech may be. One may need to be present, or privy to a
description of the scene, in order to know the referents of pro-
nouns in spoken narratives (this kind of example is typical of
work in the Bernstein vein). One may equally well need to be in
privy to an implicit scene to know the true referents of norms
a written narrative or document. 2 Personal letters afford many
instances. Even written documents in the most formal style may
be deceptively explicit. A diplomat, a bureaucrat, a college ad-
ministrator, has to learn to interpret written communications as if
present to a drama in which they are context-dependent utter-
ances. In other words, it would work against adequate under-
standing of the cognitive uses of language to treat difference of
channel as a fundamental difference. Actual uses of writing may
not have the properties conventionally attributed to them. To
think of spoken narrative as cognitively inferior to written state-
ment, because less independent of context, is to rely unreflect-
ingly on a stereotype.
Again, if one applies Bernstein's contrast to everyday genres,
then one is likely to take discourse employing abstract terms,
definitions, numbers, and statistics as self-evident examples of a
cognitively superior ("elaborated") orientation. But the form is
not a necessary evidence of the function. Abstract terms, defi-
nitions, numbers, and statistics may be present as a consequence
of rote learning, rather than complex creative thought. One may
find abstract, analytic forms that are bound to their immediate
context, unable to transcend it, and one may find concrete narra-
tive uses of language that leap toward alternative futures.
In sum, our cultural stereotypes predispose us to dichotomize
forms and functions of language use. Bernstein's contrast of
codes, distinctions between spoken and written, between narrative
and non-narrative, tend to be absorbed by this predisposition.
And one side of the dichotomy tends to be identified with cogni-
tive superiority. In point of fact, however, none of the usual
elements of conventional dichotomies are certain guides to level of
cognitive activity. In particular, narrative may be a complemen-
tary, or alternative mode of thinking.
Even if dichotomous prejudices were overcome, so that narra-
tive, even oral narrative in non-standard speech, were given its
cognitive due,' the greater equality that resulted would be an
equality of modes and genres, not of persons. The stratification
of our society, including its institutions, such as schools, would
favor the telling of some stories over others, because of the posi-
tion of the teller. The structures of relationships and settings

143
 Language in Education 131

would discourage some displays of narrative skill, inasmuch as

true performance of narrative depends on conditions of shared
background, similarity of identity, and the like (cf. Wolfson
1978). Some evidence and thinking in narrative form would not
be admitted, or not counted. If reasons were to be asked, or
given, very likely they would draw on the dichotomizing stereo-
types just sketched. Narrative forms of evidence would be dis-
missed as "anecdotal," even where narrative might be the only
form in which the evidence, or voice, was available. But the
dismissal would be an application to others of a principle the user
would not consistently apply to himself or herself--a principle,
indeed, that no one could consistently apply, if I am right in
thinking that narrative forms of thinking are inescapably funda-
mental in human life. The truth of the matter would be that only
the "anecdotes" of some would count. Even if overt performance
of anecdote (narrative) were to be excluded, there would still be
covert appeal to narrative forms of understanding.
lae, data, statistics, would be interpreted silently inTerms, formu-
terms of
"representative cases," and representative cases inevitably embody
representative stories, what Kenneth Burke (1945:59, 324) has
Called "representative anecdotes." From Burke's point of view,
every pattern of thought and terms must appeal to such anecdotes.
One's choice is not to exclude them, but to chose ones that are
appropriate and adequate. To exclude the anecdotes of others
by a rule against anecdotes in general is in effect to privilege
one's own anecdotes without seeming to do so.
In sum, if one considers that narrative may be a mode of
thought, and indeed, that narrative may be an inescapable mode
of thought, then its differential distribution in a society may be
a clue to the distribution of other things as well--rights and
privileges, having to do with power and money, to be sure, but
also rights and privileges having to do with fundamental functions
of language itself, its cognitive and expressive uses in narrative
form.
Cazden's account, and the uses to which Bernstein's cate-
gories have been put, suggest that we do indeed tend to think
of our society, and our educational institutions, as stratified in
ways that define certain kinds of narrative
to whom such kinds of narrative are naturalas inferior, and people
as inferior as well.
Certainly the students at Harvard that Cazden discusses are be-
ing encouraged to repress or abandon personal narrative in cer-
tain settings and roles. Very likely something similar happens in
many schools at many levels of education. The student or child
is told in effect that his or her own experiences do not have
weight (except perhaps as diversion). Not that there is not an
essential purpose to going beyond individual experiences. But if,
as the Bromberg-Ross recordings indicate, narrative of individual
experience is a complementary mode of solution of cognitive ques-
tions, then a pattern of discouraging it is a pattern of
132 A Folklorist's Clue to a Critique of Education
systematically discouraging what is at least a valid starting
point, and may be an essential means of thought.
The irony, or better, contradiction, is of course that aca-
demics are not themselves like that. Consider graduate work, or
teacher training. When a student is considered a candidate or
initiate for a profession, he or she becomes the recipient of
gossip and lore of the field, of insight and orientation passed
down in narrative form, of personal experiences that were mean-
ingful to those who tell them, that have shaped understanding of
the field. What many of us know about our subject comes in part
from conversations with colleagues, from the stories they have
told us, not from reading and evaluating published works. And
from those accepted as co-members of the profession we do not
discount verbal interest and effect. Indeed, we may relish it,
if the result is a good story that makes a point with which we
agree. We pay it the compliment of introducing it into our lec-
tures.
The implication of such observations is that the narrative use
of language is not a property of subordinate cultures, whether
folk, or working-class, or the like, but a universal function.
The great restriction on its use in a society such as ours has to
do with when it is considered appropriate and legitimate. Gener-
ally speaking, it is considered legitimate, a valid use of language
in the service of knowiedge, when it is used among co-members of
a group.
If the narrative function is excluded in an institutional setting,
such as a college or school, the implication is that the students
are defined as NOT co-members of a group with those who teach
them.
Perhaps some of the decline in education in the country is
connected with this suppression of the narrative function. Cer-
tainly it is more and more the case that teachers come from dis-
tricts outside the district in which they teach, even from outside
the city. Possibly schools worked better in the past when staff
and students shared more of the same world of experience, and
narrative use of language was more acceptable between them.
This factor could only be a partial one, but it may nevertheless
be significant.
Students may come from homes in which narrative is an im-
portant way of communicating knowledge. They may take part in
peer groups in which experience and insight is shared through
exchange of narrative. A classroom that excludes narratives may
be attempting to teach them both new subject matter and a new
mode of learning, perhaps without fully realizing it. Again,
difference between the culture of the teacher and the classroom,
on the one hand, and the culture of the children outside the
classroom, on the other, may be a problem. If so, a teacher may
not be able to be an ethnographer in the community, but she or
he can be an ethnographer of what is present in the classroom
itself. Giving children turns at narrative may allow them to bring

14
Language in Education 133

the outside culture inside. Finally, a teacher who permits her-

self or himself personal narrative, but not the children, may not
be b-inging children closer but underscoring the barrier (as well
as being perceived perhaps as wasting time).
Consider graduate studies again. Success on the part of a
graduate student, in the eyes of the faculty, is in part a matter
of socialization into the profession. That socialization is a matter
in part of acquiring the lore and outlook of the profession, an
informal education. A student who had mastered facts and
theories and methods, and who had no stories, and no interest
in stories, would trouble a faculty. On the one hand, the already
initiated want to be considered entertaining or at least useful
sources of lore that is of interest; on the other hand, the initi-
ates at appropriate steps should show themselves to be entertain-
ing, or at least useful, sources of lore in turn (as when having
returned from field work). There is a desire on both sides per-
haps for the link between generations to be more than names of
documents and in bibliographies.
(Fame can be defined simply as the case in which a larger,
non-professional circle knows some of the names and is interested
in some of the stories. Others not themselves the object of inter-
est find audiences for whatever narratives they themselves can
tell that involve the name. Stories could be studied in terms of
their range of distribution: department-wide, campus-wide, pro-
fession-wide, general intellectual circles.)
This argument goes somewhat against the grain of a major
thrust of our society for generations. That thrust has been to
transcend the parochial, the local, the rural, in the interest of
the opportunities and accomplishments of a general public sphere.
The often told journey to the city, or the larger city. (Though
even in the city one finds the successful able to indulge their
sentiment for their starting point and the events along the way,
others wanting or required to listen.) But perhaps this argu-
ment also helps to point up a major dilemma of our society: suc-
cess in technical, professional fields is defined in such a way
that someone cannot both stay at home, or return there, to serve,
and feel successful. This is a major problem for persons with
strong ties to their communities of origin, such as Native Ameri-
cans. One needs advanced training in order to be competent, to
be able to cope with problems faced by the community of origin,
but the advanced training embodies a message of on, upward,
and away. Perhaps the fundamental failing of higher education
in the United States is to educate for status and not for service.
Or to define service without regard to considerations of locality,
so that local is inevitably seen as lesser. Perhaps the treatment
of personal narrative in educational settings plays a part in all
this.

148
134 A Folklorist's Clue to a Critique of Education
WARM SPRINGS INTERLUDE

Much of my sense of this problem comes from experience over the

years (over the summers, mostly) at Warm Springs in Oregon.
Let me try to convey something of this experience. In doing so,
I draw on a letter written (24 August 1976) to Dennis Ted lock,
responding to questions about the directions of the journal he
edits, Aicheringa, and leading into general questions about the
role of language in poetry, ethnography, and social life. Just
before writing, my wife, Virginia, and I had been reflecting on
a quality of the use of language in the life of Indian people we
know. It is a quality one comes to have a sense of through be-
ing around them over a period of time. In one way it is a sense
of a weighted quality to incident in personal lives: as when one
friend, Hazel Suppah, told us that her son had been out to look
at a root cellar her family had built many years ago. He came
back to say, "You know, it's still good. I think we could use
it." All this in the context of a visit off the road to where an
old man had lived years ago, the house now fallen in, and the
barn, nothing disturbed but only gradually reassimilated to the
land. Hazel had lived nearby when young; the old man had come
over to their place when lonesome. One bike lay prone against a
slight rise, now a magnificent red bronze, green growing around
and through the lines of its structure, the lowest and nearest
point, a pedal, already partly within the soil. Hazel was looking
for an old-style wooden trough (resembling a canoe) that the old
man had had out for watering horses; it was gone, she realized
it must be the very one that the Tribe had installed in the re-
sort at the other end of the reservation, with flowers planted
it. If she'd known the land had been sold to the Tribe, she in
would have come to get it herself. We rummaged all around
the land, nothing to be heard but insects, the white peak of Mt.
Hood just visible from certain points behind the high hills across
the highway from which we had come. All those old places are
vacant now and most everything in them taken, years ago, by
men who built power lines across. The Indians themselves didn't
take an interest then; Hazel said they all had the same things
themselves then. Now these old places, the isolated homesteads
allotted to families in the founding of the reservation more than
a century ago, to make Christian farmers of them, are another
world and time to the Indians themselves, who cluster mostly
around the end of the reservation where the Agency, the Tribal
administration, the mill, the restaurant, the housing projects are.
Places that one can go out to in order to find and pick up things,
memories, like berry patches. We brought back an intact old
kerosene can for Hazel; she was sure her daughter would want to
go out and get the two others there. A weather-polished twin-
pronged grey piece of wood, having nothing to do with the farm,
was found by Ginny, and now shows between two trees just
Language in Education 135

outside the window of our cabin back across the mountain. Two
matching bronze sections of a broken harness, metal, a few links
of chain on each, I carried about in each hand as we walked all
round the rises on which the buildings half-stood, up to the
fences, down to the run-off creek, and finally put in the back
of the car. Well, I got carried away with trying to convey some-
thing. And forgot to put in the sunlight, along with the stillness.
Back to functions of language. 3
Virginia pointed out that in going around with a friend from
Warm Springs one often ,,,aw a bit of experience becoming an event
to be told, being told and being retold until it took shape as a
narrative, one that might become a narrative told by others.
Hazel had such stories about the old man who'd had that place,
Dan Walker, stories I had heard from others. Her son's remarks
had the weight of a theme, a kernel, of a story, the first act
perhaps. Perhaps we'll hear the rest after it has come about.
My oldest friend, Hiram Smith, once did this to me. We had
wandered about twenty minutes in a store in The Danes, drifting
out at last; later, to his daughter, Hiram reported, "Oh, that
young guy in there, he didn't know nothin' about fishing equip-
ment, Dell and I just turned around and walked right out." No-
nonsense partners, us.
Many must have had experiences of this kind. Such experi-
ences seem to point to something a bit beyond our current con-
cerns. There is a current movement to go beyond collection and
analysis of texts to observation and analysis of performance.
That is essential, but perhaps only the second moment of three.
The third is what Hazel Suppah often did, what Hiram Smith did,
what members of cultures world-wide often do, I suspect. Con-
tinuous with the others, this third is the process in which per-
formance and text live, the inner substance to which performance
is the cambium, as it were, and crystallized text the bark. It is
the grounding of performance and text in a narrative view of life.
That is to say, a view of life as a potential source of narrative.
Incidents, even apparently slight incidents, have pervasively the
potentiality of an interest that is worth retelling. The quality of
this is different from gossip, or the flow of talk from people who
have nothing but themselves to talk about--their illnesses, their
marriages, their children, their jobs, etc. Not that the difference
is in the topics. The difference is in the silences. There is a
certain focusing, a certain weighting. A certain potentiality, of
shared narrative form, on the one hand, of consequentiality, on
the other.
If such a view and practice is the grounding of an essential
texture of certain ways of life, then it needs to be experienced
and conveyed if others are to understand and appreciate the way
of life. Indians do not themselves think of such a thing as their
"culture." They use "culture" as we do popularly, for "high cul-
ture," dances, fabricated material objects, things that can go in
a museum and on a stage. Norms for speaking and performance

1.)
136 A Folklorist's Clue to a Critique of Education
go further into general norms of etiquette and interaction that are
at the heart of certain qualities and problems, yet not explicitly
acknowledged.
Ethnography is the only way in which one can find out and
know this aspect of a way of life. Of course one could ask in an
interview, or on a questionnaire: "Do you ever make up little
stories about things you see or do?" ("Oh, I guess so.")
"Could you tell me one?" ("Well, let me see, once....") Even if
successful in getting little texts--texts almost certain not to be
truly performed (see Wolfson 1978)- -such an approach would not
discover the texture of the text, the way in which it is embodied
in the rhythm of continuing life and observation and reflection of
life. One has to go around and be around to come to see how the
world is a world closely observed.
CONCLUSION
All this offers folklore, as a discipline with a special interest and
knowledge of narrative, an opportunity for both many empirical
studies and a principled critique of present society. The narra-
tive use of language seems universal, potentially available to
everyone, and to some degree inescapable. Humanity was born
telling stories, so to speak, but when we look about us, we find
much of humanity mute or awkward much of the time. The right
to think and express thought in narrative comes to be taken as
a privilege, .as a resource that is restricted, as a scarce good, so
that the right to unite position and personal experience in public
is a badge of status and rank. My account is to be listened to
because I am an x; yours is of no interest because you are only
y. All this in independence of narrative ability. The one who is
y may be an excellent raconteur, x a bore. To be sure, the ex-
cellent raconteur may be enjoyed if he or she chooses time and
topic with discretion. But very likely we hear narrative as much
these days, and enjoy it less. The decline in narrative perform-
ance among ethnic groups assimilating to the mainstream of life in
the United States has been deplored often enough. On the thesis
of this paper, the result is not a decline in quantity, but perhaps
in quality. If the Michael of Cazden's account enters the security
of an established profession, and gains standing in it, probably
he will find that his narrative accounts of his professionally rele-
vant personal experiences are considered appropriate, count for
something. Whether or not he tells them well. Successful people,
interviewed on TV shows, are recurrently asked to tell "how they
got their start." No doubt many develop a moderately interesting
narrative, if only because it is needed and they have opportunity
to practice it. But sheer narrative ability, apart from success,
seldom finds a place. Orson Bean is a superb narrator, and some
times Johnny Carson gives him his head, but on other shows, he
has gotten short shrift from MCs looking only for short repartee,
and embarrassed by the presence of small works of art.
Language in Education 137

Study of the interaction between ability and opportunity with

respect to narrative experience is very much needed. The find-
ings have a special bearing on education. Folklore has a special
role to play in providing such findings.
NOTES

1. More than the students are different in the two classes.

During the double three-hour Extension lecture, we take a break
for coffee and informal talk. The evening hour, and the second
presentation of the same content, probably made me more relaxed
as well. Participants and situation are thus confounded in their
influence.
2. Thus, from White's essay on Metahistory (1973a):
...it can be argued that interpretation in history consists
of the provisions of a plot-structure for a sequence of
events so that their nature as a comprehensible process
is revealed by their figuration as a story of a particular
kind. What one history may emplot as a tragedy, another
may emplot as a comedy or romance. As thus envisaged,
the "story" which the historian purports to "find" in the
historical record is proleptic to the "plot" by which the
events are finally revealed to figure a recognizable struc-
ture of relationships of a specifically mythic sort. (291)
In other words, the historian must draw upon a fund
of culturalLy provided "mythoi" in order to constitute the
facts as figuring a story of a particular kind, just as he
must appeal to that same fund of "mythoi" in the minds of
his readers to endow his account of the past with the
odor of meaning or significance. (294) One can argue,
in fact, that just as there can be no explanation in his-
tory without a story, so too there can be no story with-
out a plot by which to make of it a story of a particular
kind. (297)
This perspective is developed in detailed analyses in the 1973
book, but these elements of the perspective are stated in clearer,
more quotable form in the article.
3. Some- readers may be embarrassed by this bit of personal
narrative. If so, the embarrassment helps me make my point.
REFERENCES CITED
Bernstein, Basil, ed. 1973. Class, Codes and Control II: Ap-
plied Studies Towards a Sociology of Language. London:
Rout ledge & Kegan Paul.
Bissex, G. 1968. The Harvardization of Michael. Unpublished
term paper, Harvard University.
Burke, Kenneth. 1962. The Grammar of Motives. Cleveland:

152
138 A Folklorist's Clue to a Critique of. Education
World. (Originally published Englewood Cliffs, N.J.:
Prentice Hall, 1945.)
Gouldner, Alvin W. 1975-76. "Prologue to a Theory. of Revolu-
tionary Intellectuals." T elos 26:3-36.
Horton, R. and R. Finnegan. 1973. Modes of Thought: Essays
on Thinking in Western and Non-Western Societies. London:
Faber & Faber.
Hymes, Dell. 1976. "Sapir, Competence, Voices." Ms.
Labov, William. 1972. "The Logic of Non-Standard English."
In- Language in the Inner City: Studies in the Black English
Vernacular. Philadelphia: University of Pennsylvania Press,
pp. 201-240.
Mueller, Clauss. 1973. The Politics of Communication. A Study
in the Political Sociology of Language, Socialization and
Legitimation. New York: Oxford.
Philips, S. 1974. The Invisible Culture: Communication in the
Classroom and Community on the Warm Springs Indian Reser-
vation. Unpublished doctoral dissertation, University of
Pennsylvania.
Ricks, Christopher. 1967. Milton's Grand Style. London: Ox-
ford University Press. (First published, Clarendon Press,
1963,)
White, Hayden. 1973a. "Interpretation in History." New Liter-
ary History 4(2):281-314.
. 1973b. Metahistory. The historical imagination in
nineteenth-century Europe. Baltimore: Johns Hopkins Press.
Wolfson, Nessa. 1976. "Speech Events and Natural Speech:
Some Implications for Sociolinguistic Methodology." Language
in Speech 5(2):211-218.
. 1978. "A Feature of Performed Narrative: The Con-
versational Historical Present." Language in Society 7
(forthcoming).
 language in education: forward
to fundamentals

INTRODUCTION
I have been asked to speak about imperatives for change-
change in university settings such as this, and change in school-
ing generally. This I am glad to do. I believe profoundly in the
need for change in the way we understand language, and in what
we do with language in schools. I agreed to become Dean of the
School of Education because of that belief. But let me pause.
Some of you must suspect that you are about to hear another
lecture from a self-appointed bearer of light to the benighted.
Not so. Part of what we need to know in order to change is not
now known to anyone; teachers are closer to part of it than most
linguists. No one who gives priority in the study of language to
the needs of education could consider present linguistics a region
of the already saved, toward which educators must look for mis-
sionaries and redemption. I have argued against the mainstream
in linguistics for years, precisely because it has been inadequate
to study of the role of language in human life. It has made
assumptions, adopted methods, accepted priorities that prevent
the contribution to education that serious study of language
should make.
There are serious scholarly reasons for critique of the main-
stream in linguistics, reasons that draw on traditions of thought
with roots in the anthropology of Sapir, the sociology of Marx,
the linguistics and poetics of Jakobson, the literary criticism and
rhetoric of Burke.1 There are scientific problems internal to lin-
guistics that cannot be solved without change in the foundations
from which they are approached. But there are civic reasons for
critique as well One by one some of us find it intolerable to
continue a linguistics defined in a way that divorces it from the
needs of the society which supports us. The number of students
of language sharing this outlook grows. The time is ripe for a
relation between the study of language and the study of education
that is one of partnership, not preaching.
Please do not misunderstand. To criticize linguistics is not
to absolve education. The ability of schools to deal with the lin-
guistic situation in the United States is severely limited. One
often says, start where the child is, develop the child's full po-
tential. To do that, linguistically, one must have knowledge of
the ways of speaking of the community of which the child is part.
Very little knowledge of this sort is available. Each of us has
some insight into these things--some command of the ways of
1 39
140 Language in Education: Forward to Fundamentals
speaking, but each cZ us is a poor judge as well. Just because
language is basic to so many other things, so presupposed, much
of our speaking is out of awareness; we may be ignorant of much
of it, or-even in good faith confidently misreport it. Things we
are sure we never say may turn up on someone's tape; matters
of more or less may be assimilated to a sense of all-or-nothing.
Our impressions of the speech of others may be remarkably accur-
ate for placing them, without our being aware that our own speech
may contain some of the same features. Recently a linguistic and
anthropologist in Montreal recorded the speech of two friends, a
man and a woman, each speaking sometimes in formal situations,
sometimes in informal situations in which the colloquial French
known as Joual was appropriate. She played samples to a dis-
tinguished Montreal audience. The audience heard four people,
not two. It could not be convinced that there were only two, so
strong were its preconceptions as to the categorical difference.
If we are to know objectively what speaking is like, there
must be ethnographies of speaking, open to discovery of facts
that are inconvenient for one's grammar, pedagogy, or social
assumptions. Educators and linguists alike have been remiss in
not thinking of such knowledge as needed. Where linguists have
pursued_intuitions_and_universal models that ignore the realities
of speech communities and language use, many educators have
pursued notions of language and correctness that have had the
same effect. Why want to know more about something one already
knew was not really 'language'? that one knew was 'wrong'?
All this leads me to believe that there are three primary im-
peratives for change.
First, to see the need for knowledge of the language situa-
tions of our country.
Second, to support training and research to obtain such
knowledge.
Third, to change the relations between linguistics and edu-
cation.
Let me elaborate on this last imperative for a moment. A new
relation between linguistics and education may be basic to all the
rest. The essential point is that the nature of the change that
is needed is not one-sided or one way. Linguists and educators
should work together and change together. Only thus can re-
search on language be relevant to the situations faced by schools.
I shall return to the relation between linguistics and educa-
tion. Now let me try to give substance to the need for knowl-
edge.
WANTING TO KNOW
Certain goals on which we would probably agree should govern
imperatives for change. The treatment of language in schools
should help, not harm. It should help children, and through
them their families and communities, to maintain and foster
i
Language in Education 141

self-respect. It should be consonant with respect for diversity

of background and aspirations. It should contribute to equality
rather than inequality.
It is probably hard to keep from nodding to words like these.
Such words are familiar and accepted. Yet we face assumptions
and ignorance about language that contradict and work against
such goals.
Consider a school in a community. What would you want to
know, were you responsible for the linguistic aspect of the school-
ing there? There are many who would not think that there was
much that they needed to know, even how to use the language of
the children. Recently I was asked to a meeting at the Philadel-
phia School Board to help resist pressure to remove the require-
ment that a teacher in a TESOL class know the language of the
children being taught English, that is, be able to communicate
with them. Most teachers at the school in Madras, Oregon, to
which children from Warm Springs Indian Reservation go, do not
think they need to know anything about the Indian languages in
the homes from which the children come, or the etiquette of speak-
ing there. By and large, indeed, knowing languages and knowing
about language is little valued in our country, if it involves ac-
ceptance of diversity. You and I may have no difficulty in under-
standing standard West Indian English, may even admire it--I
think it myself the most lovely English I have heard. But the
daughter of a family from Jamaica was just admitted to a state-
affiliated university in Philadelphia on the condition that she take
a course in English for foreign students. Have you not often
heard a proper middle-class white say in exasperation to a cab-
driver or voice on the phone, "Oh, I can't understand you", al-
though the black or Spanish accent was entirely intelligible?
Identification of the difference having closed the listener's ear?
When educated, concerned people want to know about lan-
guage, what is it they are likely to want to know? A graduate
student at my university reports that when she spoke recently to
the group that supports her studies, their serious, well inten-
tioned questions made assumptions about languages and their re-
lations to human groups that a linguistics student could not even
have imagined entertaining. Recently I was asked by a cultured
voice on the phone to help with a program being planned for the
Canadian Broadcasting System, to view French in Montreal in the
light of similar situations in the United States and the Caribbean.
I began helpfully naming friends who know about such things,
when it came out that the premise of the program was that the
French-speaking lower classes of Montreal could not think right
because they could not speak right. (You can imagine the haste
and confusion with which I withdrew the names and tried to dis-
associate myself from the whole thing.)
These are merely recent instances that have impinged upon
me in the course of a month or two. It is almost too painful to
be a student of language attentive to such things--examples
142 Language in Education: Forward to Fundamentals
accumulate so readily of prejudice, discrimination, ignorance
bound up with language. It makes one wonder if discrimination
connected with language is not so pervasive as to be almost im-
pervious to change, so deeply rooted as to almost preclude sup-
port for the asking of questions that might lead to change. To
be sure, some may be sure that children would be fine if left
alone, and be glad to learn what is wrong with schools. Others
may be sure that schools are doing what is right, and be glad
to learn what is wrong with children and the homes from which
they come. It is hard to find people who sense a need to under-
stand objectively the school child's communicative world, a world
seriated into a multiplicity of contexts of situation and ways of
speaking suitable to each, a world of a plurality of norms for
selecting and grouping together features of a verbal style, of a
plurality of situation-sensitive ways of interacting and interpret-
ing meaning in terms of styles, such that a type of situation such
as classroom interaction with a teacher or formal test-making has
meaning in terms of its relation to the rest. Such that each in-
volves a spoken or written genre that has a place in a series of
such, a possibility of performance dependent on particular rules
for commitment to performance. So that to understand the part
of a verbal repertoire that appears in educational settings, one
needs to be able to compare choices of communicative device and
meaning, displays of communicative and cognitive ability, across
a range of settings. So that to understand the part of a child's
(or teacher's) ways of speaking one sees in school, one needs to
understand the whole. One needs to do or to draw on linguistic
ethnography. 2
As you know, there is little done and little on which to draw.
What I have sketched in general terms is what one might reason-
ably ask about if concerned with the role of language in school-
ing in another culture or country; expecting things to be strange,
one wants to know. In a sense, we need to be able to stand
back from our own situations so as to see them as strange and
as needing to be known.
BLACK AMERICANS, NATIVE AMERICANS, SPANISH-SPEAKING
AMERICANS
I should not suggest that nothing at all is known or being done.
Certainly there has been a good deal of attention in recent years
to patterns of speech associated with some of the major groups
that make up this diverse country. Yet the research is scattered
and spotty with regard to both geography and class. The case
of 'Black English' is instructive. In the 1960's the ways of speak-
ing of Black Americans attracted attention. The research has
been important in demonstrating the systematic, rule-governed
nature of the vernacular spoken by many Blacks, as against
notions of it as an incoherent corruption. Notions of Black chil-
dren practically without language were shown to be functions of

15
 Language in Education 143

intimidating formal situations in schools, to be situational, not

general. Some of this work helped as well to highlight the re-
spects in which distinctive features of the vernacular point to the
wider spectrum of Caribbean Creoles and their West African ele-
ments. Awareness grew of the place of the vernacular in peer
group interaction against the background of Caribbean and Afri-
can traditions of spoken artistry. Still, research focused mostly
on the variety of speech most strikingly different from the public
standard, the vernacular of adolescent urban males. Much less
analytic attention was given to the speech of Black women, of
preachers and ministers, of established upper-class families, or
to the Caribbean and African background of elaborated 'talking
sweet' and public oratory. And some explanations of what became
known were so partial as to be false. Some linguists wished to
treat the vernacular as only superficially different and formally
derivative from standard English, for reasons having to do in
part with convenient simplicity of a grammatical model. Others
wished to treat the vernacular as so distinct that it might require
its own textbooks. There are indeed places where people want
their variety maintained independently in print from a closely re-
lated one (in Czechoslovakia Slovaks feel this way about the rela-
tion of their variety to Czech). In the United States such a con-
ception fails to take into account the actual attitudes of many
Black people who want the variety of English in the classroom,
especially the written variety, to be the common standard. Still
others drew from this isolated fact the inference that Blacks de-
preciated the vernacular, even speaking of 'self-hatred'. In point
of fact, there is widespread acceptance of the vernacular variety
at home and in informal situations generally; 3 it retains a special
place even among Black students at a university such as Prince-
ton. Yet sympathetic interpretations of Black speech can be in-
adequate too. Many come to know Black terms for uses of lan-
guage, such as 'shucking' and 'jiving', and regard them entirely
as an Afro-American ethnic heritage. Yet analogous genres of
language use can be found among lower class white youths, and
such ways of coping verbally may have their origin in subordi-
nate social status as well as in ethnic tradition.
The relation between varieties and uses of English, on the
one hand, and being Black, on the other, is complex and only
beginning to be adequately known. The situation is little better
with regard to other major groups. We think of Native Americans
in terms of the many languages lost, and of efforts to maintain
or revive those that remain. The relation of schools to these
efforts is of the greatest importance. My own anger and passion
about the treatment of language in schools comes largely from
experience of local schools and educational research institutions
that affect Indian people at Warm Springs Reservation in Oregon.
But these situations must not be oversimplified. Indian Americans
themselves may differ in their views as to what is best in terms
of language. And aspects of language that are crucial to the

156
144 Language in Education: Forward to Fundamentals
success of Indian children may not involve the traditional Indian
language at all. Where the Indian children are, linguistically, may
not be an Indian language, but an Indian variety of English.
There are probably several dozen such Indian varieties of English
in the United States. They play a significant social role. Some-
one who has been away, and who returns to a local community,
must take up the local variety of English or be judged snobbish.
Features of children's speech that seem individual errors may in
fact reflect a community norm. They may reflect a carryover into
English of patterns from an Indian language. In the English of
Indians at Isleta pueblo, south of Albuquerque, New Mexico, a
double' negative contrasts with a single negative as a carryover of
a contrast between two types of negation in the Isleta language.
There are doubtless other such examples, but the fact is that
Isleta English is the only form of Indian English carefully studied
and reported on in print, and that only in the last few years.4
The language situation of an Indian community will be still
more complex, in having standard as well as local vernacular Eng-
lish present, and a vernacular, even reduced, variety of an Indian
language as well as or even instead of its 'classic' form. In the
Southwest Spanish may be a factor as well. Yet we have hardly
more than a few sketches of such cases. With Indians as with
Blacks, research has not attempted to provide systematic knowl-
edge of the language situation of the communities experienced by
children. Research has focused not on social reality, but on the
exotic. To say this is not to condemn study of traditional Indian
languages. Much of my work continues to be devoted to the study
of one group of languages, now nearing extinction. I and a col-
league are the last to work intensively with fully fluent speakers,
and like others in such a situation, we have obligations both to
those who have shared their knowledge with us and to those who
later will want access to it. The work has its contribution to
make to respect and self-respect for Indian people. The dispro-
portion between what most linguists do and what most needs to be
done is not here. There have never been trained scholars enough,
and much has been lost unrecorded in consequence. With all its
wealth our country has sparsely supported knowledge of the lan-
guage that first named the continent. The fact is telling. We
have barely managed to study languages that fit our image of the
noble Redman, let alone begun to notice the actual linguistic make-
up of Indian communities.
The knowledge one needs to start where Indian children, any
children, are goes beyond varieties of language, of course, to
patterns of the use of language--customary community ways of
answering questions, calling upon others, taking turns in conver-
sation, speaking or remaining silent, giving instruction by verbal
precept or observed example, all the ways in which etiquette of
speaking and value of language may take distinctive shape. Many
Indian children come to school, speaking only English, yet

aa
Language in Education 145

encounter difficulty, not because of language difference, but be-

cause of difference in patterns for the use of language. Children
found 'shy' and non-talkative in class may be as talkative as any,
if observed in situations where the rights and duties of speaking
are those of the community from which they come. In such a
case one needs to know not a language, but a community way of
speaking.
The issue and language most prominent today are bilingual
education and Spanish. I cannot attempt to treat this complex
situation here, except to note that the general difficulty is the
same. Too little is known as a basis for policy and practice in
schools. The widespread resistence to such a thing indeed may
cause bilingual education to be attacked as having failed before
it will have had a chance to be understood and fairly tried. Ef-
forts to provide equal educational opportunity to Spanish-speaking
children must proceed with a mirOmum of information as to the
Spanish the children speak, in relation to the varieties and uses
of Spanish in the community from which they come. No simple
general answer can be laid down in advance. There are several
national and regional standards, Cuban, Puerto Rican, northern
Mexican, Colombian, etc.; in many communities there is a range
of varieties from a standard to colloquial vernacular and an argot,
as well as a way of mingling Spanish and English in conversation
that can count as a special variety among intimates. The attitudes
of Spanish speakers toward the elements of this complex language
situation are themselves complex. Clearly it is not enough to
advocate "Spanish." It is possible to have Anglo children doing
well Spanish-speaking children doing poorly in a Spanish class
in a school. There are problems of the fit or conflict between
the Spanish spoken by children and the Spanish taught, between
community and teacher attitudes, between the language-linked
aspirations of cultural traditionalists and the job-linked aspirations
of some of the working class; the desire of some speakers to insti-
tutionalize Spanish as a language of higher education and profes-
sional activity, versus the needs of children for whom Spanish is
primarily a vernacular of the home and community; problems of
children educated in Puerto Rico corning to the mainland with in-
adequate English and children educated on the mainland going to
Puerto Rico with inadequate Spanish.
There are problems of assessing the language abilities of chil-
dren both for assignment to classes and for evaluation in pro-
grams. Assignment to classes is sometimes being done under man-
date of law in a begrudging rough-and-ready fashion, minimizing
the number of children to be assigned. Sometimes the availability
of monies to a district prompts forced assignment to special
classes of bilingual children who have no English problem at all.
Valid assessments of language ability require naturalistic observa-
tion across a range of settings, but such methods have been little
developed in explicit form. Formative evaluation of programs in
bilingual situations needs ethnographic knowledge of the community
146 Language in Education: Forward to Fundamentals
language situation, and summative evaluation needs ethnographic
monitoring of the process by which a program comes to have
particular meanings and outcomes for participants and community.
Such success as bilingual programs have will be best attested in
the debates ahead, not by test scores, but by case-history ac-
counts that show convincingly the benefits to children and com-
munities, and how they were achieved. 5
ETHNIC HERITAGE AND USAGES OF LANGUAGE
The situations of Black Americans, Native Americans, Spanish-
speaking Americans are salient but not unique. Bilingual educa-
tion is an issue for communities of Chinese, Japanese, Filipino,
and others. Many European languages in addition to Spanish are
maintained to a significant extent. Immigration renews some of
these communities. All of them participate in a climate of opinion
that is world-wide. The general truth would seem to be that
about twenty years ago, when those who spoke in the limelight
foresaw an end to ideology, and an endless technocratic future
whose chief problem would be leisure, many ordinary people
around the world were drawing a different lesson from their
experience. They had been caught up in such a vision of the
post-war future for a while, only to begin to find that their
place in it was not worth the giving up of all that they had been.
'Progress' came more and more to seem the 'dirty word' that
Kenneth Burke has called it--less an engine carrying them onward
and upward, more a juggernaut about to run over them, their
place, their customs, their speech.
This general revival of concern with ethnic heritage is not
merely a part of the annual tourist laundry ring around the
world, each countryside emptying out in summer to take in some-
one else's carefully staged culture while on vacation. It is a
shift in outlook that has to do with what one is for oneself, as
a member of a family with a certain name, a certain history, a
knowledge of certain places, certain ways of meeting sorrow and
sharing joy. Many of you may know personally the price that can
be exacted in acquiring a lingua franca at tho cost of a language
of the home.
Some repudiate concern of this kind as nostalgia and senti-
mentality, even as a dangerous refusal to face present realities.
1 think that something profound is involved. Any one concern
may seem particularistic and limiting; when all such concerns are
considered together, one sees something general, a deep-running
tide. It is a vision limited to a national lingua franca that begins
to appear old-fashioned, limited, sectarian.
The deep-running tide seems to me a shift in what is regarded
as the dominant obstacle to a way of life in balance with human
needs. A century, even a generation ago, it was common to think
that the dominant obstacle consisted of traditional ideas and cus-
toms. Except when compartmentalized in diminished form, as
 Language in Education 147

objects of intermittent piety and curiosity, specific cultural tradi-

tions, beliefs, conventions, identities seemed brakes from the past
on progress. The future lay with a science and mode of produc-
tion that could realize the control of nature, and the plenty, of
which mankind was capable. Now we are far lcss sure. Some
critics of contemporary society consider the. very idea of incessant
technological change to be itself the dominant obstacle to a way of
life in balance with human needs. Not that material progress is
irrelevant, but that the quality of life is seen more clearly to de-
pend on other things as well. What seemed a policy in the inter-
est of all has come to seem an instrument of profit to some at the
expense of others in many cases. Uncontrolled, it threatens com-
munity today and even sustenance tomorrow.
There is an essential linguistic dimention to this. It is hard
to specify, but necessary to address. Let me try to suggest
something of its nature.
The internal structures of language and the structures of use
to which languages are shaped athce show two fundamental, com-
plementary general kinds of function, of meaning, at work. They
are intertwined in reality, but our way of thinking about language
has separated and opposed them. 6 One can be roughly indicated
as concerned with naming, reference, sheer statement, the techni-
cal, analytic, logical uses of language. Modern linguistics has
built its models on this aspect of language. Modern science,
technology, and rationalized bureaucracy give it pre-eminence.
For a time the uses of language characteristic of literature, reli-
gion, personal expression were neglected and on the defensive.
For a time the pinnacle of knowledge appeared to many to be a
single logical language to which all science and legitimate knowl-
edge might be reduced. That ideal has been largely given up and
replaced by recognition of a plurality of legitimate uses of lan-
guage. The seminal figures in philosophy of course were Cassirer
and Wittgenstein, and there have been related developments in
poetics, anthropology, sociology. Interpersonal, expressive,
aesthetic uses of language come more to the fore. In part it is
because an ideal of language that seemed the touchstone of pro-
gress, of the advance of reason, has been too often traduced.
The idioms of objective knowledge, of science, mathematics, logic,
experiment, statistics, contracts, regulation and control were once
seen as common bases for progress for us all. We have too often
seen claims to authority, couched in such idioms, turn out to be
rationalizations of special interests, elite excuses, outright decep-
tions, as with the Vietnamese war. Idioms of moral concern and
personal knowledge that had at first no standing came to be seen
as more accurate guides than the trappings of elaborate studies
and reports. A little later it was general discovery of the per-
sonal "roice through transcripts of tapes that decided, I think,
the public verdict on a president. I could not prove the point,
but I think these two experiences have had complementary, de-
cisive effect on our sense of validity in the use of language.
148 Language in Education: Forward to Fundamentals
I sense a more general drift as well. Increasingly we are
concerned to have a place for things that cannot be said without
distortion, or even said at all, in the idioms of elaborated, formal,
purportedly rational and referential speech that take pride of
place in public science, public government, linguistic and peda-
gogical grammars. There are things we know and need to be
that have no standing there. A sense of this is a reflection of
the central problem of the role of language in modern society, the
crisis of language, namely, what the balance is to be between
modes of use of language. The old dichotomies--correct vs. in-
correct, rational vs. emotional, referential vs. expressive -fail
to capture the nature and complexity of the problem, for it is
not a matter of mutually exclusive opposites, but of the inter-
weaving of mutually indispensable functions.
EDUCATION AND LINGUISTIC FOUNDATIONS
I am sorry not to be precise, concrete, and clear about this. It
would require far more than one lecture to try to explain the
ramifications of this point for the study of language, to trace
the implications at different levels of the organization of language,
to appraise the efforts that are being made now to devise an ade-
quate general model. I can try to say clearly what this complex
situation means for the future of language in education. It is
this. Linguistics developed out of a situation in which the study
of language was loosely distributed across a variety of .disciplines.
It became the central discipline by development of general methods
for the formal study of language structure. The methods and the
associated conception of language structure focused on an essen-
tial, but partial, aspect of the organization of language. Other
aspects remained secondary or eschewed. The focus of attention,
having started with phonology, and proceeded through morphology
and syntax, has now reached semantics and even 'pragmatics'
(that is, the interpretation of meaning in context of use). From
every side it begins to be recognized that linguistics as we have
known it is inevitably part of a larger field.7
At the first, language structure was divorced from language
use. Now language use is included along with language structure
by most. Eventually it will be generally recognized that it is not
use that is a derivative of structure, but structure that is de-
pendent on use. That one can never solve the problems of the
organization of language in social life without starting from social
life, from the patterns of activity and meaning within which lin-
guistic features are organized into styles and ways of speaking.
A 'Linguistics that is truly the science of language, liuguistics that
is truly a foundation for education, will be a linguistics that is
part of the study of communicative interaction. It will understand
linguistic competence as part of communicative competence. It will
understand the character of competence in relation to the social
history and social structure that shape it in a given case.
 Language in Education 149

Such a linguistics, should the day arrive, will have an essen-

tial property. Its practice and theory will be adequate to all the
means employed in speech and all the meanings that speaking (or
another use of language) has. Its theory of English phonology
will attend not only to the features that make a consonant /p /
instead of /b/, but also to the aspiration that can make the word
angry. Its theory of syntax will attend to isolated grammatical
sentences as but a special case among the intelligible, acceptable
sequences of discourse. Its theory of meaning will attend not
only to words and constructions, but also to the meanings inher-
ent in choice of dialect of variety, of conversational or narrative
genre, of occasion to speak or be silent. Its theory of compe-
tence will go beyond innate and universal abilities to the kinds of
competence valued and permitted in a given society, to opportuni-
ties and obstacles of access to kinds of competence. It will recog-
nize that the very role of speaking, of language and use of lan-
guage, is not the same in every society; that societies differ in
their ideals of language and ability in language; that use of lan-
guage, like sex and eating, is a universal possibility and neces-
sity of society, but without power to determine its place or mean-
ing. Its relative importance among other modes of communication,
its role as resource or danger, art or tool, depends on what is
made of it. 8 Two things follow. First, the relation between edu-
cation and linguistics cannot be a matter simply of joining the two
as they are now. We do not yet have the kind of linguistics just
described. Second, we are not likely to get it if linguistics is
left to itself. The prestige of formal models as against empirical
inquiry remains strong. The pull to continue to concentrate on
familiar ground will be great. To get the linguistics we need will
take pushing by others. Educators ought to be in the forefront.
If you should remember just one thing from this occasion, please
remember this: Do ask yourself what linguistics can do for you,
but even more, demand of linguistics that what it can do be done.
And do not apologize for the demand, or assume that it diverts
the study of language from pure science to murky application.
The fact is that the study of language does not now have the
knowledge on which much of application should be based, and
cannot get it without new theoretical, methodological, and empiri-
cal work. To demand attention to the needs of education is not
just a demand for applied linguistics. It is a demand for change
in the foundations of linguistics.- The struggle for educational
change with regard to language, and the struggle for scientific
adequacy in the study of language, are interdependent.
I have used the word 'struggle' advisedly. It would be mis-
leading to suggest that the kind of linguistics we need is an apple
almost ripe, ready to drop at a tweak of the stem. There is in-
deed a diffused slow drift in the right direction, such that work
entwined with practical problems has low status. Still, work
more abstract and remote from practical problems, the higher the
status. Some leading linguists, such as William Labov, want to
6
150 Language in Education: Forward to Fundamentals
reverse this polarity. Educators can help, and may have some
leverage these days when conventional positions for linguists are
hard to find. The fact that linguistics itself is evolving in a
direction that makes work in educational settings germane is a
help, as is the fact, just mentioned, that new theory is part of
what is needed. Still, a second great difficulty remains. This
is the difficulty of seeing language in education in the context
of American society, steadily and whole.
SEEING OUR LANGUAGE SITUATION
The history of attention to language situations within the country
points up the difficulty. Black uses of English have been evolv-
ing in the United States since before the Revolution, but have
begun to be adequately studied only as a consequence of the Civil
Rights Movement and the federal attention and funding that re-
sponded to it. Spanish has been here for centuries as well, but
Spanish bilingualism and language situations have begun to be
studied adequately only as a result of the socio-political mobiliza-
tion of Spanish-speakers. American Indian communities have had
multilingual situations and distinctive ways of speaking for gener-
ations without much attention. The interest of many Indian peo-
ple in maintaining and reviving traditional languages fits into the
traditional approach to the study of Indian languages, but it has
taken the Native American mobilization of recent years to make
academic scholars think of the- preparation of materials useful in
education as something they shOuld do. Indian English and ways
of speaking still remain relatively little studied.
In general, educationally significant aspects of a language
situation have come into focus only after the community in ques-
tion has been defined as a social problem, and more especially,
as a social force. Previous attention to the languages involved
focused upon what seemed most exotic and remote. Immigrant and
Indian languages alike have been viewed mostly as something
lingering from the past.
We need to begin to think of the linguistic heterogeneity of
our country as continuously present. The United States is a
multilingual country, with great numbers of users of many lan-
guages. American multilingualism is not an aberration or a resi-
due. If anything, it has increased in recent years, especially
with regard to Spanish, Vietnamese, and perhaps a few other
languages. We need to address the linguistic heterogeneity of
our country as a permanent feature of it, discuss what shape it
will and should have, anticipate the future. To do so, we have
to address the linguistic ethnography of the United States as a
sustained, central scientific task. Ad hoc responses after the
-fact- of-social-mobilization-connected-with-language-come-too--late
and provide too little help. And ad hoc responses are too easily
distorted by the immediate terms of social and political issues.
 Language in Education 151

Members of language communities themselves may have a partial

view. We need sustained work that provides both knowledge of
language situations and an independent, critical assessment of
language problems.
.... Educators-have a stake-in the mounting of-such a program of
study, since mobilization around issues of language so commonly
turns attention to schools. Educators have a special stake in
making sure that a sustained program of study includes inde-
pendent, critical attention to the nature of language problems.
That attention should include study of the process by which
something having to do with language does (or does not) become
defined as a problem in our country in the first place. It is not
to be assumed that there is a fit between public recognition of
problems and actual language situation. (To repeat, teacher
failure to recognize the structure and role of Black English Ver-
nacular still handicaps many Black children, and did even more
before it became recognized as a 'problem' in the 1960's.
of those who resist such recognition continue to be Black.)Some
I suspect there are four kinds of case. That is, there are
indeed situations recognized as problems that are genuinely prob-
lems (bilingual education, for example); there may well be situ-
ations not defined as problems that can be left alone. But I sus-
pect that there are also situations not now defined as problems
that ought to be so defined--situations taken for granted but at
possible cost. For example, very little has been done to study
communication in medical settings, especially between professional
personnel and patients. 9 What are the effects of difference in
idiom, terminology, semantic system? or even of difference in
native language, there being so many medical personnel of foreign
origin? and in some regions so many patients with little command
of English? Peihaps there is no recognized problem because those
affected have little visibility or consciousness of common concern.
Yet a series of articles in the New York Times might make this
situation, itself unchanged, suddenly a 'problem'. Finally, there
may be str.uations defined as problems that ought not to be, the
issue being falsely or superficially posed--e.g., the supposed
problem of children with practically 'no language'. Any of us
may be subject to cultural blinders and public fashions. We need
comparative, critical, historical perspective to transcend them.
We need, in short, to be able to see our country in terms of
language, steadily and whole. To do so is to go beyond questions
of diversity of languages and language-varieties. Black English,
Navajo, Hopi, Zuni, Spanish, Italian, German, Slovenian, the
many, many languages of this country are salient and important.
The diversity they comprise is so great, so neglected, as to be
almost overwhelming by itself. Yet there is something further.
There is a unity that_has_also_escaped-us.--1-do-not-mean--political-----
and social unity. That ik; not in question. To be sure, the drive
for homogeneity has been so great that even today the thought of
diversity being accepted can frighten some. Street signs in

16
 152 Language in Education: Forward to Fundamentals
Spanish, even in a Spanish-speaking neighborhood, can attract
ire. A telephone company may refuse to hire a Spanish-speaking
operator, to answer emergency calls, in an area with many Span-
ish dominant speakers. To argue for recognition of ethnolinguis-
tic diversity seems troubling to many, as if the ties between us
were so fragile as to break beneath a crumb of difference. But
the forces making for integration, the economic and communicative
ties of the country, are irreversibly dominant. To argue the
right to diversity is to argue only for breathing space within the
hive.
The unity in terms of which we need to see our country is
the unity in its dominant. groups and institutions that gives it a
certain cut and pattern, regarding language, regarding the value
on language, the way in which language enters into life. We need
to be able to imagine the United States sociologically as if it were
a small country, a Belgium or Switzerland, a single entity of
which one could ask, as one can ask of any society: what are the
basic patterns of the use of language? what are the values, rights,
responsibilities associated with language? what is the outlook of
the culture with regard to language? how did it come to be that
way? how does it seem likely to change?
We are able to think of the Navajo or the French in this way.
We need to be able to imagine ourselves in this way as well, to
find, through comparative, historical, and descriptive study, a
mirror in which to see the United States as possessed throughout
its history of language polic:s, of predominant attitudes towards
language and its role, that give it one place among many possible
places in the roster of the world's cultures.
Even if .there were only English the unity to be seen is not
simple. Imagine that the only language in the country was Eng-
lish, even standard English. Situations, roles, activities, per-
sonal characteristics such as age and sex would still affect and
shape ways of using language. The occupational and class struc-
ture of the society would still be there as a source of hetero-
geneity, on the one hand, and hegemony, on the other. Let us
consider heterogeneity first.
Inherent Heterogeneity. Even if everyone used some form of
standard English, all the manifold ways of talking as a person of
a certain kind, of using language to do a certain kind of thing,
would be present, needing to be discerned and described and
their consequences considered. Many of the judgments made of
persons in everyday life, many of the opportunities one has or
does not have, involve command or lack of command of these
styles and genres, of being able to talk like an X, or being able
to use language to do Y. Such diversity is inherent in social
life. Research has barely begun to address it adequately, relat-
------ing-linguistic -devices-and-patterns -to social -meanings_and_roles
It is the same here as with differences of whole language or
language-variety. Research mostly follows the flag of social
mobilization. Sex-related differences in language hardly appeared
16
 Language in Education 153
for the first time a few years ago. Yet until recently one would
have had to conclude that men and women talked alike in every
society except for a few American Indian tribes, the Chukchee of
Siberia, and some scattered others, so far as the published liter-
ature could show. Again, status-related differences in language
are hardly the monopoly of the Japanese,
yet until recently linguistic theory treated Koreans,
them as
and Javanese,
fascinatingly
special.
There is a general lesson to be
starts from grammar can see sociallydrawn. A linguistics that
relevant features only when
they intrude within the grammar. If the very units of phonology
or morphology cannot be stated without reference to the sex or
status of a participant in
into account; indeed, the speech,
case may
then the social fact is taken
become celebrated as an in-
stance of "men's and women's speech" or special
expression of status. Yet sexual roles and statusconcern with the
universal in society, and assuredly come into play when peopleare
differences
speak to each other. Starting from grammar,
how they come into play; one has to start fromone does not see
the social feature
itself, and look at the use of language from its vantage point.
Then the features of language that are selected and grouped to-
gether as characteristics of speaking like a woman, speaking like
an elder, and the like can be seen.
A final example: many are aware of the interesting ways in
which choice of second person pronoun in French (tu: vous),
German (Du: Sie), Russian (ty: vy), etc., can signal lesser or
greater social distance. Many is the paper written
nouns and related forms of salutation and greeting. onYet such pro-
it is a
safe assumption that variation in social distance is universal, and
universally expressed in one or another way in use of language.
Management of social distance may well be one of the most per-
vasive dimensions of language use. One has to start from recog-
nition of social distance to begin to see thoroughly and accurately
how it is accomplished as a function of language.
Even if only standard English were found in the United
States, then, there would be many socially shaped patterns of
language use to discover and consider. Still, the diversity would
have a certain unity. Not "English", but the history, values,
and social structure of the United States would give a character-
istic configuration to them.
Hidden Hegemony. Schools would not find their problems of
language resolved in the situations we are imagining now. Con-
cern to develop the full potential of each child would lead to
recognition of language as involving more than command of a
standard. For example, I suspect that there is a pervasive domi-
nant attitude that discourages verbal fluency and expressiveness
in _white males.._It_ought_to-be-food-for-thought-that-in most
known societies it is men who are considered the masters of verbal
style, and indeed often trained in its ways, whereas women are
subordinated and even disparaged. In our own country, as we
1 us
154 Language in Education: Forward to Fundamentals
know, it is commonly girls who show most verbal ability, who
learn to retain foreign languages, etc. Men in public life whose
work depends on use of language may be heard to disclaim any
special knowledge or command of it. The hint of homosexuality
seems not to be far from aesthetic mastery of language in a man.
Again, I suspect that many persons spend much of their lives in
what might be called 'verbal passing', the maintenance of the
public verbal face that is not chosen, but imposed. And what is
the fate of narrative skill in our society? There seems some rea-
son to think that the expressivity of traditional narrative styles
has often been disapproved by the upwardly mobile and middle-
class. One sees a loss between generations of a vital narrative
style in some people of Indian communities. People continue to
relate accounts and narratives, of course; are we storying more
and enjoying it less?
Most serious of all, and most difficult for schools perhaps to
accept: I suspect that our culture is so oriented toward discrimi-
nation among persons on the basis of language that even a society
of 200 million speakers of standard English would show a class
and occupational structure much like the present one, matched by
a hierarchy of fine verbal discriminations. In other words, we
must consider the possibility that schools, along with other insti-
tutions, have as a latent function the reproduction of the present
social order on the apparently impartial ground of language.
Given the inherent' variability in language and language use, even
a society of standard English speakers would show detectible
differences in pronunciation, diction, preferred constructions,
and the like. Are we so convinced that language change is lan-
guage decline (as many of our educated elite appear to be), so
predisposed to correctness and correction, that most of that
society of standard English speakers would still leave school with
a feeling of linguistic insecurity and inferiority?
Perhaps not, but in order to see our society, and the place
of language in it, especially the place of language in education in
it, we have to ask such questions. To what extent are the in-
adequacies and senses of inadequacy about language in the society
to be explained by the backgrounds and characteristics of those
who pass through schools? To what extent are they unwittingly
produced by schools themselves?
EQUALITY-IMPLEMENTATION
Perhaps our society can never come closer to equality of oppor-
tunity, to a treatment of language in schooling that starts where
the child is, that develops the fullest linguistic potential of the
child. Still, those are the goals in terms of which one often
speaks. It is only that the change required to come closer to
them is so pervasive--change in knowledge, change in attitude,
ultimately change in social structure itself. Change in what we
It:now can never be enough, yet without it the other changes are
163
Language in Education 155
impossible. One sees some change in the treatment of Black Eng-
lish Vernacular that would not have come about without the re-
search of the past decade or so. Knowledge of other situations
can have effect too, especially in the context of a view of the
history and direction of the role of language in the society as a
whole.
My call for such knowledge in relation to schools amounts to
a call for an educational linguistics, as a major thrust of schools
of education, departments of linguistics, and all concerned with
language and with education. Let me add that it should be shaped
not only by educators and linguistics, but also by members of the
communities concerned, teachers and parents both. It is inherent
in adequate study of language that one must draw on the knowl-
edge that members of a community already tacitly have, and the
same is true for ethnography, for knowledge of ways of speaking
in relation to cultural contexts. And insofar as the work to be
done involves policies and goals, members of the communities af-
fected must necessarily play a part. The educational linguistics
envisioned here is in part a community science.
Such an educational linguistics entails change in both lin-
guistics and education. In a sense, its goal must be to fill what
might be called a 'competency' gap. There is a gap in the sense
of a lack- of persons able to do the kind of research that is
needed. The gap exists because the need to fill it has not been
recognized, and recognition of the need depends on overcoming a
'competency' gap in another, theoretical sense. Both linguists
and educators may use the term 'competence'; the gap between
their uses is at the heart of what needs to be changed.
In linguistics the term 'competency' was introduced by Chom-
sky a decade or so ago. Its ordinary meaning suggested a lin-
guistics that would go beyond language structure to the linguistic
abilities of people. The promise proved a bit of hyperbole. The
term was used in a reduced sense as equivalent to just that por-
tion of competence involving knowledge of a grammar, and gram-
mar itself was defined in terms of an ideal potentiality, cut off
from any actual ability or person. Grammar was to explain the
potential knowledge of an amalgamated everyone in general, and
of no one in particular. Social considerations were wholly absent
from such a 'competency'. The result has been conceptual con-
fusion that has led some to abandon the term altogether; others
to tinker with it; still others to denounce its use as partisan
apologetics ('that's not 'competence' was used to mean 'what you
are interested in is not linguistics'.). In Chomskyan linguistics,
in short, 'competence' has meant an abstract grammatical potential,
whose true character and whose relation to realized alike remain
quite uncertain. The image of the language-acquiring child has
been one of an immaculate innate schemata, capable of generating
anything, unconstrained and unshaped by social life.
In education the terms 'competence' and 'competency-based'
have become associated with a quite different conception. The
156 Language in Education: Forward to Fundamentals
emphasis is upon specific, demonstrable, socially relevant skills.
No one can be against demonstrable skills, but there is fear that
the notion reduces education to a very limited conception of
ability and potential. It suggests an image of an externally
shaped repertoire of traits that does not allow for going beyond
what is already given. It suggests that success in transmitting
basic skills is something that was once in hand, lost, and now to
be gone back to.
Each polar notion of 'competence' treats as basic something
that is derivative. The simple linguistic notion treats formal
grammar as basic, and use of language as unconnected, or de-
pendent, whereas in fact the opposite is the case. What we con-
ceive as grammar is a precipitate of a normative selection from
among the ways of speaking, the true verbal repertoire, the full
organization of means of speech. Grammar began that way in the
service of Hellenistic cultural hegemony and continues that way in
the service of a certain conception of science. A valid notion of
verbal competence reaches out to include the full organization of
means and meaning of speech, and becomes part of a notion of
communicative competence.
The notion of 'competence' that has gained currency in edu-
cation treats distinguishable skills as elementary, underived,
whereas any prescribed set of skills is a precipitate of a complex
of assumptions and understandings as to the nature of society,
its present and future opportunities, and the probable or pre-
scribed relation of a group of students to it. There is a tendency
to focus on instrumental, vocational ingredients of verbal skills,
perhaps at the expense of the full range of verbal abilities valued
and possible.
In both cases the limited notion of competence is bound up, I
think, with a limited ability to see the nature of the language
situations in the United States. That limitation is academic. I
want to suggest that the problem of language in education is not
to go back to basics, whether in the grammar of the linguist or
the grammar of the schoolbook, but to go forward to fundamentals.
How does language come organized for use in the communities from
which children come to schools? What are the meaning and values
associated with use of language in the many different sectors and
strata of the society? What are the actual verbal abilities of chil-
dren and others across the range of settings they naturally en-
gage? What is the fit, what is the frustration, between abilities
and settings--where is an ability fru'strated for lack of a setting,
a setting unentered for lack of an ability, in what ways are pat-
terns of personal verbal ability shaped by restrictions of access
to settings, on the one hand, culttirally supported aspirations, on
the other?
When we consider where a child what its potential is, we
are coiiiiderink-abilities for which 'competence' is an excellent
word, if we can understand it aright, in something close to its
Language in Education 157

ordinary sense, as mastery of the use of language. To use the

notion in education, we need to know the shapes in which mastery
comes in the many communities of speaking that make up the
country, and we need to be able to relate those shapes to the
larger historical and social factors that constrain them. Ethno-
linguistic description can at least enable us to see where we truly
stand with regard to linguistic competence in the United States.
The knowledge it provides is indispensable for those who wish to
change where we stand.
To see the need for knowledge of the language situations of
our country, to support training and research to obtain such
knowledge, to change the relations between linguistics and educa-
tion, so as to bring into being an educational linguistics that can
foster all this--these are the imperatives for change, the funda-
mentals to which we must move forward.
The key to implementing such changes, I think, is in the
hands of Schools of Education. There is little chance of success,
little chance of results relevant to schools, if educators do not
play a principal role in shaping the growing concern of students
of language with the social aspects of language. At the Univer-
sity of Pennsylvania we are expanding a Reading and Language
in Education,
Arts program into a general program of Language Linguistics
and including in it a specialization in Educational as a
foundational field. The purpose is both to train researchers and
to influence the training and outlook of those in other parts of
the School. The new program is possible partly because of the
cooperation and support of some linguists outside the School.
Each School of Education may find its own particular pattern, but
a successful pattern ought to have these three ingredients men-
tioned: training of research specialists, influence on the training
and outlook of others, cooperation between educators
research
and linguists.
of greatest
The greatest challenge to research, the
benefit to schools now, will be to domesticate and direct the skills
of ethnography and descriptive linguistics, of sociolinguistics or
ethnolinguistics in broad senses of those terms. We need programs
of research that can function within a limited frame of time, say a
year, and provide through linguistic ethnography
district
a usable sketch
served by a
of the ways of speaking of a community or
school. For the most part linguistic ethnography has flourished
abroad with studies of cultural uses of language in Mexico, Africa,
Panama, the Philippines. We need to bring it home to Pittsburgh
a-3 Philadelphia. The support of Schools of Education will be
essential for this. The models or research that are needed are
not wholly ready to hand: practical, relevance and research develop7
ment must grow together, in the sort of environment that a School
of Education can provide.
It is not too much to imagine, indeed, that language
of a
in educa-
School.
tion can be an integrating focus for many aspects
The ties with Reading and Language Arts, with developmental
 158 Language in Education: Forward to Fundamentals
psychology, with English studies are obvious. When one con-
siders the way in which problems of language are shaped, by cul-
tural assumptions and attitudes, it becomes apparent that there
are ties with the historical, sociological, and anthropological foun-
dations of education. There is a complex of spurious and genuine
problems of language diversity in relation to special education and
school counselling. Issues of curriculum and instruction
well. With a bit of luck and a lot of initiative, education arise as
find itself a major force in shaping the study of language might
in the
United States.
NOTES
1. The contribution of
my debt to them, is indicatedeach of these men, and something of
in Hymes 1970, 1974, ch. 8, for
Sapir; Hymes 1974:85-86, 121-122, 204, for Marx; 1975, for
Jakobson; 1974, ch. 7, for Burke.
2. This point is developed more fully in my Introduction
Cazden et al. (1972). In introducing to
my lecture, Donald
son quoted a very apt passage from that essay, framing what I Hender-
had to say perfectly, and I am grateful to him for it.
3. See Hoover (1975).
4. The pioneer in this work William Leap. See his article
(1974); a book-length collection ofis studies of Indian English is now
being edited by Leap.
5. I try to address these issues in some detail in a paper
called 'Ethnographic Monitoring', written for a symposium on
'Language Development in a Bilingual Setting', March 19-21, 1976,
organized by Eugene Briere for the Multilingual/Multicultural Ma-
terials Development Center of California State Polytechnic Univer-
sity. Plans for publication are not yet definite. (See pp. 104-
118 in this volume.)
6. Let me stress that I do not suggest
language structure and use can simply be assigned that every aspect of
other of the two generalized types of function. They to one or the
either-or catch -ails. They are interdependent; their nature are not
not quite the same at one level of language as at another; their is
manifestations enter
levels of language. into a variety of relationships as between
of language cannot beThe essential point that an adequate study
built on attention is
to just one of them. I
speak of generalized types of function because
ment on the specific set required in a model of therelanguage
is no agree-
structure,
and a good many specific functions may need to be recognized,
some universal, some local. I do think that at any one level there
are fundamentally just two kinds of
means, roughly a 'what' and a 'how'.means, The
and organization of
principle of contrastive
relevance within a frame that is basic to linguistics' applies to
both: the 'same thing' can be said in a set of contrasting ways,
and the 'same way' can be used for a set of contrasting 'things'.
A key to the organization of language in a particular culture or

1
Language in Education 159
period is restriction on free combination of 'what's' and 'hows',
the things that must be said in certain ways, the ways that can
be used only for certain things. The admissible relations com-
prise the admissible styles. In effect, the study of language is
fundamentally a study of styles. There is further discussion in
my Introduction to Cazden et al. (1972) and my essay, 'Ways of
Speaking', in Bauman and Sherzer (1974).
7. See Hymes (1968).
8. This point should be obvious, yet seems hard to grasp,
so deeply ingrained is a contrary assumption. I have been try-
ing to make the point for almost twenty years. See Hymes 1961a,
1961b, 1964a, 1964b, 1974, ch. 6.
9. Roger Shuy has pioneered in this regard. For discussion
of the general issue of language problem, I am indebted to mem-
bers of the Committee'on Sociolinguistics of the Social Science Re-
search Council, especially Rolf Kjolseth.
REFERENCES
Bauman, Richard and Joel Sherzer (eds.). The Ethnography of
Speaking. New York: Cambridge University Press, 1974.
Cazden, Courtney; Vera John-Steiner; Dell Hymes (eds.).
Functions of Language in the Classroom. New York: Teach-
ers College Press, 1972.
Hoover, Mary E. R. "Appropriate Uses of Black English as Rated
by Parents." Stanford University, School of Education, Stan-
ford Center for Research and Development in Teaching,
Technical Report No. 46, 1975.
Hymes, Dell. "Functions of Speech: An Evolutionary Approach."
In Fred C. Gruber (ed.), Anthropology and Education.
Philadelphia: University of Pennsylvania, pp. 05-83.
printed in Yehudi A. Cohen (ed.), Man in Adaptation.Re-
Chicago: Aldine, 1961, pp. 247-259.
. "Linguistic Aspects of Cross-Cultural Personality
Study." In B. Kaplan (ed.), Studying Personality Cross-
Culturally. Evanston: Row, Peterson, 1961 (later: New York:
Harper and Row), pp. 313-359.
"The Ethnography of Speaking." In T. Gladwin and
W. Sturtevant (eds.), Anthropology and Human Behavior.
Washington, D.C.: Anthropological Society of Washington,
1962, pp. 15-53.
"Directions in (Ethno-) Linguistic Theory." In A. K.
Romney and R. G. D'Andrade (eds.), Transcultural Studies
of Cognition. Washington, D.C.: American Anthropological
Association, 1964, pp. 6-56.
(ed.). "Introduction: Toward Ethnographies of
munication." In J.J. Gumperz and D. Hymes (eds.), Com- The
Ethnography of Communication. Washington, D.C.: American
Anthropological Association, 1964, pp. 1-34.
160 Language in Education: Forward to Fundamentals
Hymes, Dell. "Two Types of Linguistic Relativity." In W.
Bright (ed.), Sociolinguistics. The Hague: Mouton, 1966,
pp. 114-158.
"Linguistics--The Field." International Encyclopedia
of the Social Sciences. New York: Macmillan, 9:351 -371,
1968.
. "Linguistic Method of Ethnography." In Paul Garvin
,;ed.), Method and Theory in Linguistics. The Hague:
Mouton, 1970, pp. 249-325.
. Foundations in Sociolinguistics. Philadelphia: Uni-
versity of Pennsylvania Press, 1974.
. "Pre-War Prague School and Post-War American Anthro-
pological Linguistics." In E.F.K. Koerner (ed.), The Trans-
formational-- Generative Paradigm and Modern Linguistic Theory.
(Amsterdam Studies in the Theory and History of Linguistic
Science, IV; Current Issues in Linguistic Theory, I.)
Amsterdam: John Benjamins B.V., 1975.