Ethnography, Linguistics, Narrative Inequality

Toward An Understanding Of Voice

Dell Hymes

• Publisher: Routledge
• Year: 2003
• eBook ISBN: 9781135745653
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

This collection of work addresses the contribution that ethnography and linguistics make to education, and the contribution that research in education makes to anthropology and linguistics.; The first section of the book pinpoints characteristics of anthropology that most make a difference to research in education. The second section describes the perspective that is needed if the study of language is to contribute adequately to problems of education and inequality. Finally, the third section takes up discoveries about narrative, which show that young people's narratives may have a depth of form and skill that has gone largely unrecognized.

Part I
Ethnography

Chapter 1

What is Ethnography?¹

Introduction

Ethnography has come to be much discussed in education. Often enough one hears some form of the question, ‘What is ethnography?’ The National Institute of Education commissioned a report to answer the question. All this might be puzzling to an anthropologist, especially to one with an interest in the history of the subject. If one traces the history of ethnography 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 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 fieldwork, both in general and with regard to specific techniques, and more recently, ethics. A book addressed to ethnography in our own society (Spradley and McCurdy, 1972) has been used 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 unified conception of ethnography in relation to the study of institutions of our own society, such as education. And anthropologists 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. 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 anthropologists of avariety of persuasions to criticize such methods. It is often harder for them to state concisely the alternatives. Ethnography cannot be assumed to be something already complete, ready to be inserted as a packaged unit in the practices and purposes 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 them, ‘ethnography’ may be a brief-lived fad in educational research. 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 between ethnographers and sponsors of educational research, a relation that will change both. Because of my conception of ethnography, 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 similar attempt to consider the issues raised here in brief compass. Many others must contribute from their own experience and outlook.

The Ethnographic Tradition

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 experimental design and quantitative measurement should be considered ethnography, but a positive definition is not easy to provide. A major reason for the difficulty is that good ethnography has been produced under a great variety of conditions, by a great variety of persons, some of it before there was a profession to train such people, and professional 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 psychological experience, rather than the discovery of knowledge. It is clear that ethnography involves participation and observation. What should count as ethnography, what kinds of ethnography 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 travellers, missionaries, government officials and the like that may contribute information and insight about the culture of the peoples of the world have been welcomed and gleaned for what they could provide. But when Ibn Khaldun (1967 [1381]) or Father Sahagun (1956 [1840, composed 1547–1569]) 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 or practical, motivated some to seek more adequate knowledge. These guides to inquiry have in common 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 Petty² (1623–87) share in this lineage.) 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 contrastive, 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 Consanguinity 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 distinguish the kind of ethnography organized and guided by John and Bea Whiting, for example, for inquiry into socialization in several contrasting societies. The Whitings had attempted to come to general conclusions, testing hypotheses 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. Therefore 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 socialization, sent into the field in several different societies for an extended period of time, kept in touch with through correspondence throughout the field study, and brought back to write up their results. (Whiting, 1963, is a principal outcome.) Like Morgan, the Whitings had insight into contrasting types of society, a need for specific information, and a general theoretical frame (in this case, psychodynamic) to which contrast and specifics were relevant.

All three types of ethnographic inquiry continue to coexist in anthropology today. There may still be occasional discoveries of unrecognized peoples (as claimed in the Philippines a few years ago), for whom comprehensive information is to be provided from scratch. There are still many peoples, knowledge about whom has never been adequately systematized, and which 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 perspectives 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 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 ethnography 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 ethnography lead to hypothesis-oriented ethnography. Given a substantial 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 begins 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? Are the functions the same? In the ethnography of speaking right now, one is aware of broad contrasts, usually presented as dichotomies (cf. Bernstein’s elaborated 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.

Schooling in Ethnographic Perspective

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 analysed in ways that permit all the insight possible into schooling. If schools were considered from the same standpoint as kinship 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 analysis 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 or 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 ethnographers, it is of the essence of the method that it is a dialectical, 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 discover that the community one is studying actually determines the residence of newly-married couples on the basis of principles one had not foreseen. (The illustration is an actual one. See Goodenough, 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 civilization’s partial view. For many ethnographers, an essential characteristic 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 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 (1965) generalized the experience of linguistic inquiry. In order to discover 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 disclose the system. One had to test the relations among sounds for their functional relevance within the system in question. The result 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 begins analysis of a given case he called etic₁. 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 institutions 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 knowledge 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 questionnaire was prepared by a central-office person generally qualified by training and ethnic background. The purpose of the questionnaire was to find out what parents thought of a community-relations policy and person. The questionnaire was duly administered. The student administering the questionnaire discovered, during informal conversations with parents, that they interpreted the questions differently than the designer and the school. The parents distinguished between a playground (having equipment designed for children to use) and a playyard, but the questionnaire did not. When asked if they had had a chance to meet their School-Community Coordinator, they answered ‘no’, because to them to meet would require having talked, and knowing by name, even first-name, not just having been introduced. In terms of the questionnaire their ‘no’s’ were interpreted as not having met. The student administering the questionnaire was distressed, but the procedure of inquiry had no way for him to take account of what he had learned or to have what he had learned affect the presumed results (Abbot, 1968).

Experiences of this kind make ethnographers distrust questionnaires and quantitative results derived from them, if the meanings of the questions to those asked are taken for granted in advance. Many ethnographers do use questionnaires, but questionnaires devised after sufficient participation and observation to ensure their validity.

The validity of knowledge about persons, families, neighborhoods, schools, and communities in our country depends upon accurate and adequate knowledge of the meanings they find and impute to terms, events, persons, and 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, recreation. 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 of 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 knowledge 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 implicit, 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 they are so fully taken for granted.

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 etic₁), 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 interconnections and conceptions of these relationships of speaking and singing (and music generally) are various and diverse. Modern serious music includes such categories as Sprechstimme introduced by Arnold Schoenberg. The Maori of New Zealand consider the playin...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Series Editor’s Introduction
• Prefaces
• Acknowledgments
• Part I Ethnography
• Part II Linguistics
• Part III Narrative and Inequality
• References