ASSESSING CULTURAL

ANTHROPOLOGY

Robert Borofsky
HAWAII PACIFIC UNIVERSITY

New York St. Louis Sm Francisco Aucklmd Bogota Caracas

Lisbon London Madrid Mexico City Milan Montreal New Delhi

Sanjuan Singapore Sydney Tokyo Toronto

 11
Methods Belong to All of Us

Respect the nature of the empirical world

and organize a methodological stance to reflect that respect.
Herbert Blumer (1969:60)

nomena in their natural environment. Sea-

going oceanographers, astronomers, wildlife
The word "method" in anthropology, indeed biologists, and anthropologists are all natural-
in all of social science, has at least three mean- ists. Participant observation is a strategic ap-
mgs. proach to data gathering used by some
First, method refers to epistemology, to sets of naturalists in the social sciences.
assumptions about how we acquire knowl- And third, method refers to techniques or sets
edge. The phrase "scientific method," for ex- of techniques for collecting and analyzing
ample, encapsulates a set of such assumptions data. The survey questionnaire method is a set
(as well as some rules of practice). The as- of techniques (involving sampling, construc-
sumptions are: (1) there is a reality "out there" tion of instruments, interviewing, and other
(or "in there" in the case of ideas and emo- things) for collecting data. Spot observation is
tions); (2) it can be apprehended, more or less, another set of techniques for gathering data
by human beings through direct experience (for example, about how much time people
(or through some proxy for direct experience); spend in various activities). Content analysis,
(3) all natural phenomena can be explained componential analysis, multidimensional scal-
without recourse to mysterious forces beyond ing, and ethnographic decision tree modeling
investigation; and (4) though the truth about are techniques for analyzing (extracting mean-
phenomena is never known, we do better and ing from) data.
better as old explanations are knocked down Anthropologists receive little formal training
and are replaced by better ones. Competing in any of these three kinds of research meth-
epistemologies reject one or more of these ods. Many graduate programs in anthropology
assumptions. offer no required course in ethnographic
Second, method refers to strategic approaches methods (Trotter 1988; Plattner 1989). I be-
to the accumulation of actual data. Experi- lieve that training in all three kinds of meth-
mentalism and naturalism are two strategic ap- ods-epistemology, strategy, and technique-
proaches within the scientific method. Experi- should be a major part of both the undergrad-
mentalism involves the direct manipulation of uate and graduate curriculum in anthropology.
variables under the most controlled conditions How much training in methods is enough?
possible. Naturalism involves observing phe- I'll address this question at the end of the
chapter. First, I discuss method as epistemol- ratio of the degree of generality of their
ogy and method as strategy. conceptions and operations-bankers at the
surrunit, merchants next, then manufacturers,
and agriculturalists at the bottom" (1866:122).
Comte attracted admirers who wanted to
implement the master's plans. Mercifully, they
are gone, but the word "positivist" still carries
the taint of Comte's ego.
One might draw a broad distinction between
two epistemologies in cultural anthropology-
positivist and interpretivist. I will focus on the
former here and leave others (in this volume) The modern version of posItIvIsm was
to describe the latter for, in my view, the developed by members of the Vienna Circle, a
positivist perspective has not always been group of philosophers, mathematicians,
understood. Positivists and interpretivists may physicists, and social scientists who met from
disagree on matters of epistemology. But 1923 to 1936 in a series of seminars. The
when we talk about methods at the level of members of the circle were corrunitted to
strategy and technique, methods belong to all empiricism and against metaphysics. Hunches
of us. could come from anywhere, but scientific
At its birth, positivism was simple enough. knowledge, they said, is based only on experi-
Here is John Stuart Mill in 1866 explaining ence, and scientific explanation is based only
"Ie positivisme" to an English-speaking audi- on logical, mathematical principles. Hence the
ence. label "logical empiricists," by which they were
known for a while.
Whoever regards all events as parts of a constant For these positivists, logic was paramount.
order, each one being the invariable consequent of They understood that the assumptions of
some antecedent condition, or combination of science are assertions. They also understood
conditions, accepts fully the Positive mode of that the assumptions of science are not mere
thought. . . . (15) assumptions. The members of the Vienna
All theories in which the ultimate standard of Circle were corrunitted to a scientific, that is
institutions and rules of adions [IS] the happiness of logical, mode of thinking and to the benefits
mankind, and observation and experience the guides that this would bring to humanity. This was
... are entitled to the name Positive. (69) the component of positivism that they shared
with Comte. It's what motivates many social
Positivism1 in the mid-nineteenth century was scientists today, including me, to count our-
science applied to the study of humanity, and selves as positivists.
done in the exuberant and optimistic spirit of It is clear, of course, just how dangerous this
service to human happiness. corrunitment can be. My ideas about what is
The French social philosopher Auguste good may not be the same as yours. This is
Comte didn't just believe in science as an what makes the study of ethics a critical part
effective means for seeking instrumental of method as epistemology.
knowledge, though. He envisioned a class of
philosophers who, with support from the
state, would direct all education. They would
advise the government, which would be In the early decades of this century, many
composed of capitalists "whose dignity and luminaries of cultural anthropology supported
authority," explained Mill, "are to be in the the collection of both quantitative and qualita-
tive data. They collected texts and counted should be about. An ethnographer had to live
things as necessary in their research. a long time with the people whom he or she
For example, one of Tylor's enduring con- studied, and had to learn the native language
tributions to anthropology was his paper "On fluently. Above all, said Radin, the ethnog-
a Method of Investigating the Development of rapher must provide readers with the original
Institutions" (1889). In that paper, Tylor de- texts, the materials from which he or she
scribed a numerical method for doing sys- makes observations.
tematic cross-cultural comparisons. One of Those texts, and the exegesis of the ethnol-
Boas's great contributions was a monograph in ogist, should show current cultural realities "as
which he demonstrated the relationship of nu- seen through the mirror of an actual man's
trition to body size among immigrants, ren- heart and brain and not through the artificial
dering useless much racist rhetoric of the day heart and brain of the marionettes with which
against allowing Eastern Europeans into the Boas and Sapir and Kroeber operate" (Radin
United States. Kroeber's study of 300 years of 1933:10). Radin didn't mince any words. To
women's fashions is a landmark in the quanti- demonstrate his method, Radin presented a
tative study of cultural trace materials meticulous analysis of a series of texts from
(Richardson and Kroeber 1940). three Winnebagos about the Peyote Cult. In
When I was a graduate student at the Uni- one of those texts, John Rave discussed his
versity of Illinois thirty years ago, the tradition conversion to the Peyote Cult. Radin drew
of using both qualitative and quantitative data on his knowledge oflocal history. He showed
was still strong. Oscar Lewis, for example, ad- that Rave's account reflected acceptance by
vocated the use of survey methods to comple- Rave of particular components of Christian
ment ethnographic field studies.2 Joseph Casa- faith and particular components of earlier
grande taught cross-cultural hypothesis testing Winnebago beliefs. Rave failed to achieve a
using the Human Relations Area Files. Both, vision during his Winnebago puberty rite, and
of course, were ethnographers who had col- was denied membership in the Medicine
lected reams of qualitative data. Dance. This, said Radin, accounted for Rave's
There was, to be sure, another tradition in proselytizing zeal.
anthropology, one opposed to the use of You can object to all this as just so much
quantitative data. The cause against quantifi- pop-psych, but Radin was the consummate
cation in anthropology was eloquently taken fieldworker: he knew the language, he spent
up by one of Boas's students, Paul Radin, in years working with the Winnebagos and with
1933. In The Method and Theory of Ethnology, these particular informants, and he had the
Radin praised his teacher for criticizing Ty- texts to back up everything he said. IfKroeber
lor's and Frazer's evolutionary doctrines. But and Boas and Sapir didn't like his arguments,
he accused Boas of being "naturwissenschaft- Radin said, "there still remains the document
lich eingestellt" or science-minded (Radin for them to interpret better and more pro-
1933:10), of believing that cultural facts could foundly" (1933:238).3
be transformed into physical ones and Radin would not have wanted to be re-
counted. He railed at Boas for training a gen- membered as a scientist, but in my terms he
eration of students like Kroeber, Sapir, Lowie, was-and a gifted one, at that. He was empiri-
Wissler, and Mead who, Radin said, sought to cal in the extreme; he built arguments only
tag cultures and habitats, like animal and plant from observations; he was committed to the
species, and compare them statistically (Radin open-endedness of truth. With all his texts on
1933:10). the Peyote Cult, he had the data on which to
Radin was right about that crowd, but he postulate a theory of revitalization. That he
had a different idea of what anthropology left this nomothetic exercise to others who
would come later diminishes not one whit the lowed up. After decades of experience with
contribution of Radin's effort. the method of participant observation, we are
Many anthropologists today who share not much further along than when we started.
Radin's views about primary data also want to We haven't studied participant observation
contribute to the development of knowledge systematically; we haven't improved it.
about regularities in human behavior. They At least five things affect the kind and qual-
may have some well-founded doubts about ity of data we can collect as participant ob-
whether all the phenomena in which they are servers: (1) personal characteristics, such as age
interested can be quantified accurately. But I and gender; (2) language fluency; (3) objec-
believe that what keeps many cultural anthro- tivity; (4) informant accuracy; and (5) infor-
pologists (I haven't counted them, but I'm mant representativeness.
guessing hundreds, not dozens) from conduct-
ing quantitative research is that they are just Personal Characteristics. By the 1930s, Margaret
unequipped to do it when they leave graduate Mead had already made clear the importance
school. of gender as a variable in data collection.
Gender has at least two consequences: it influ-
ences how you perceive others; it limits your
METHOD AS STRATEGY: access to certain information.
PARTICIPANT OBSERVATION In all cultures, you can't ask people certain
questions because you're a [woman][man].
Anthropologists are divided on epistemolog- You can't go into certain areas and situations
ical issues, but almost all of us use the strategic because you're a [woman] [man]. You can't
method of participant observation to collect watch this or report on that because you're a
our primary data. Even anthropologists who [womanHman]. Even the culture of anthro-
use questionnaires in their research engage first pologists is affected: your credibility is dimin-
in participant observation in order to build the ished or enhanced with your colleagues when
questionnaire instrument. you talk about a certain subject because you're
Participant observation is what makes it pos- a [womanHman]. (See Golde 1970; White-
sible for interpretivists and positivists alike to head and Conaway 1986; Scheper-Hughes
collect life-history documents, attend sacred 1983b; Altorki and EI-Solh 1988.)
festivals, talk to people about sensitive topics, When she worked at the Thule relocation
map the landholdings of informants, trek with camp for Japanese-Americans during World
a hunter to count the kill, and interview War II, Rosalie Wax did not join any of the
women traders formally and informally about women's groups or organizations. Looking
how they cover their losses in the daily mar- back after more than forty years, Wax con-
ket. Among other things, participant observa- eluded that this was just poor judgment.
tion helps us build rapport. Rapport is what
makes it possible for us to observe and talk to I was a university student and a researcher. I was
people and record information about their not yet ready to accept myself as a total person, and
lives. 4 this limited my perspective and my understanding.
Those of us who instruct future field workers should
encourage them to understand and value their full
range of being, because only then can they cope
intelligently with the range of experience they will
Anthropologists developed the method of par- encounter in thefield. (1986:148)
ticipant observation, and we have seen with
satisfaction its acceptance by researchers in Besides gender, we have also learned that
many other disciplines. But we haven't fol- being old lets you into certain things and shuts
you out of others. Being a parent helps you way to become alert to and avoid this prob-
talk to people about certain areas of life and lem.
get more information than if you were not a If participant observers run the risk of being
parent. Being wealthy lets you talk to certain suckered, then research is called for to deter-
people about certain subjects and makes others mine (1) the conditions under which that's
avoid you. Gregarious anthropologists may be most/least likely to happen and (2) how to
unable to talk to shy people. recognize and avoid it. Research is called for
Being divorced has its costs. Nancie on the kinds of data available to "indigenous
Gonzalez found that being a divorced mother anthropologists" that are not available to out-
of two young sons in the Dominican Repub- siders, and vice versa (see Messerschmidt 1981
lic was just too much. "Had I to do it again," for papers on doing fieldwork in your own
she says, "I would invent widowhood with culture). Research is called for on making par-
appropriate rings and photographs" (1986:92). ticipant observation the best strategic method
Even height may make a difference: Alan it can be.
Jacobs once told me he thought he did better
fieldwork with the Maasai because he's almost Objectivity. No one I've ever known in the
6 1/2 feet tall than he would have if he'd social sciences has seriously thought that
been, say, an average-sized 5 feet 10 inches. humans could become robotic, completely
objective field researchers. But just because
Language Fluency. Thirty years ago, Raoul perfect objectivity is impossible, this does not
Naroll (1962:89-90) found suggestive statis- let us off the hook. The economist Robert
tical evidence that anthropologists who speak Sarlow is reported to have observed that,
the local language are more likely to report while a perfectly aseptic environment is im-
witchcraft than those who don't. His inter- possible, this doesn't mean we might as well
pretation was that local language fluency im- conduct surgery in a sewer (cited by Geertz
proves your rapport, and this, in turn increases 1973:30). Objectivity clearly varies from per-
the probability that people will tell you about son to person. Some people achieve more of
witchcraft. it, others less. More is better.
Does the credibility of our data depend on Objectivity means becoming aware of one's
control of the local language? In his 1933 dia- biases, and transcending them, not the lack of
tribe against his science-minded peers, Radin any biases. Some people do better at tran-
complained that Mead's work on Samoa was scending, some do worse. Striving for objec-
superficial because she wasn't fluent in tivity is important even if perfect objectivity is
Samoan (1933:179). Fifty years later, Derek unobtainable.
Freeman (1983) raised questions about Laurie Krieger, an American woman doing
whether Mead had been duped by her in- fieldwork in Cairo, studied physical punish-
formants, perhaps because she didn't know the ment against women. She learned that wife
local language well enough.5 beatings were less violent than she had
According to Brislin, Lonner, and Thorn- imagined, and that the act still sickened her.
dike (1973:70) Samoa is one of those cultures Her reaction brought out a lot of information
where "it is considered acceptable to deceive from women who were recent recipients of
and to 'put on' outsiders. Interviewers are their husbands' wrath. "I found out," she says,
likely to hear ridiculous answers, not given in "that the biased outlook of an American
a spirit of hostility but rather sport." Brislin et woman and a trained anthropologist was not
al. call this the "sucker bias," and warn field- always disadvantageous, as long as I was aware
workers to watch out for it. Presumably, of and able to control the expression of my
knowing the local language fluently is one biases" (1986:120).
 The need for objectivity is recognized by that scholars from many disciplines had read
even the most qualitative fieldworkers. Colin La Pierre's study and had recognized its
Turnbull says that the key to good fieldwork importance.
is "to know ourselves more deeply by con- My colleagues and I conducted a series of
scious subjectivity" because in this way "the studies during the 1970s to find out if people
ultimate goal of objectivity is much more could accurately report their social interac-
likely to be reached and our understanding of tions. A lot of social science, after all, is based
other cultures that much more profound" on data collected by asking people "who did
(1986:27). you [talk to] [exchange memos with] [interact
Objectivity does not mean (and has never with] [call on the telephone] in the last [24
meant) value neutrality. No one asks Cultural hours] [week] [month]?" In 1984, my col-
Survival, Inc., to be neutral in documenting leagues and I reviewed the literature on in-
the violent obscenities against indigenous formant accuracy (see Bernard and Killworth
peoples of the world. No one asks Amnesty 1977, for example). A long list of studies, in-
International to be neutral in its effort to cluding our own, showed that a fourth to a
document state-sanctioned torture. We recog- half of what informants say about their behav-
nize that the power of the documentation is in ior is inaccurate.
its objectivity, in its irrefutability, not in its This finding shows up in studies of what
neutrality . people say they eat, how often they claim to
have gone to the doctor in a given period of
Informant Accuracy. In 1940, C. Wright Mills time, and in how much money they say they
wrote that "the central methodological prob- have in their savings accounts. It shows up in
lem of the social sciences" is the fact that what the most unlikely (we would have thought)
people say and what they do is so different places: in the 1961 census of Addis Abapa, 23
(1940:329). He was right, and the problem percent of the women underreported the
remains central today. number of their children. Apparently, people
Informants lie, of course. But lying appears don't "count" babies that die before reaching
to be a small part of the problem. Informants, the age of two (Pausewang 1973:65).
like all of us, make honest errors of Progress is being made in explaining in-
commission (they say they did things that they formant inaccuracy. Thirty years ago, Cancian
didn't do) and of omission (they neglect to (1963) showed how informant errors con-
report things that they did do) in reporting formed to expected patterns in prestige rank-
their own behavior. ings in a Mexican village. D'Andrade showed
The problem was articulated clearly by in 1974 that there is a general pressure to
Richard La Pierre in 1934. Accompanied by a think in terms of "what goes with what" even
young Chinese couple, he crisscrossed the if this creates errors in reporting factual events.
United States by car. They stayed at and ate at (Also see Shweder and D'Andrade 1980.)
66 hotels and 184 restaurants. Six months after More recently, Freeman, Romney, and Free-
the trip was over, La Pierre wrote a letter to man (1987) found that, for some behaviors at
all those hotels and restaurants, telling the least, informants report typical behavior to
managers that he was planning a trip and anthropologists rather than specific behavior
would they mind serving Chinese people? when asked to dredge up what they actually
More than 90 percent said they would not did and who they actually interacted with (see
serve Chinese people. Irwin Deutscher picked also Freeman and Romney 1987; McNabb
up the theme in 1972, in his book What We 1990).
Say, What We Do. He made it clear that the People round off, in other words, and report
problem had not been solved, despite the fact behavior according to rules of central ten-
dency. This may go for informants' reports of for selecting informants who are experts in
their own behavior as well as for their reports particular cultural domains (like ethnozoology,
of the behavior of others. (Cognitive psychol- local medical practice, or whatever). Romney
ogists, of course, have done careful investiga- et al. demonstrate that, when assumptions of
tions of how people store and retrieve infor- their model are met, just six informants may
mation.) be enough to achieve representative, valid in-
Still, the question remains: When people formation about a cultural domain. It is an
report their behavior to us, do we get inac- intriguing line of research.
curate answers, or answers that are accur-
ate proxies for some other question about
norms? METHODS TRAINING
IN ANTHROPOLOGY
Informant Representativeness. Even if infor-
mants tell us accurately what they know, there I believe that training in methods should be a
remains the question of whether what they key part of both the undergraduate and the
know is representative of the population we graduate curriculum in anthropology. How
are studying. much is enough? Well, given the three mean-
Since the 1930s, cultural anthropology and ings of the word "method" (epistemology,
sociology have gone their separate ways strategy, and technique), it takes a lot to be
largely over this issue. Sociologists have enough.
focused on developing systematic instruments
(like questionnaires) and applying those ques-
tionnaires to representative samples of re-
spondents in a population. This places empha- All anthropologists need a thorough ground-
sis on reliability (whatever the instrument ing in the various approaches to knowledge
measures today it will measure tomorrow, that have characterized our discipline. This
more or less) and on external validity (what- means exposure of all students, whatever their
ever we find out from the sample can be ex- initial predilections, to the philosophical foun-
tended to the rest of the population, within dations of structuralism, symbolism, interpre-
some known limits of error). tivism, hermeneutics, phenomenology, posi-
Anthropologists have focused on internal tivism, and empiricism.
validity. What good is it to know that data are Training in ethics should be part of the
reliable or generalizable if they are inaccurate? training in epistemology. Plattner (1989:33)
Every social scientist knows horror stories of makes the case:
questionnaires that force people to produce
nonsensical answers (because the questions are Training in the ethics of research is as important as
culturally inappropriate, for example). training in the techniques of research design, data
On the other hand, why would we want to collection and data analysis. Research consists of a
know what a handful of informants think or series of decisions (where to study, what to study,
do, if the information cannot be generalized who to interview, what questions to ask, when to
beyond those informants? This is one area leave the field), and every decision ... has an
where considerable progress has been made in ethical component.
anthropology. Jeffrey Johnson's recent book
(1990) on selecting ethnographic informants Every anthropologist has stories to tell about
summarizes much useful information in this how his or her ethics were tested by circum-
area. Romney, Weller, and Batchelder (1986) stances in the field. We have accumulated a
offer a technique called consensus modeling large collection of these cases, in the Newsletter
of the American Anthropological Association, for multivariate analysis). Students who want to
example, and in Cassell and Jacobs (1987). We work with a large corpus of field notes, or
ought systematically to use the wisdom they with life history material, need training in
provide and not rely on the telling of stories computer-based text management.
informally to disseminate that wisdom. The All anthropologists need formal training in
case-study method is used in teaching law and the collection and analysis of both qualitative and
business administration. A case-study course quantitative data. Photography, videography,
on ethics should be part of every graduate cur- audio taping, stenography, and sketching are
riculum in anthropology. methods for collecting qualitative data. Cod-
ing, counting, and measuring produce quanti-
tative data. People's responses to open-ended
questionnaire items are qualitative. Coding
All anthropologists need formal trammg in and attaching numbers to the responses create
participant observation. This means familiarity quantitative data. Observing people and
with the accumulating literature on the sub- recording their behavior in words produce
ject. It means studying what we already know qualitative data. Counting the behaviors
about how good and bad participant observa- r
creates quantitativ data. Transcribed texts are
tion is done-about culture shock, about gen- qualitative data. Counting the number of
der roles in the field, about violence, disease, times people use a particular word or theme in
and other hazards of field research, about a text creates quantitative data.
maintaining objectivity without the pretense Pawing through field notes is qualitative
of value neutrality. data processing. Putting the notes on a com-
This kind of training will give anthropolo- puter and using a text management program
gists the skills they need to do better field- to paw through them is still data processing.
work. It will also give them the means to Doing a chi-square test with a hand-held
contribute concretely to improving participant calculator is quantitative data processing. So is
observation. putting thousands of numbers into a computer
and having the machine calculate the chi-
square value.
Much of what is called statistical analysis is,
This is where it gets tough. College graduates in my vocabulary, data processing, not data
in the social sciences these days simply must analysis. Thinking about the themes in a text,
know how to use a computer-not just for like field notes, is data analysis. Thinking
word processing but for data entry, data man- about what a chi-square (or any statistical)
agement, data retrieval, and data analysis. Stu- value means is also data analysis. All analysis is
dents of anthropology who manage to escape ultimately qualitative. But we can't even get
getting these skills before going to graduate to that stage unless we've collected solid
school will need compensatory education (valid, credible) data.
before they can study formal data collection Besides learning to use research methods,
and analysis. we should be improving them. Every field trip
Once the basics are out of the way, anthro- provides information on how personal charac-
pologists need training in research design, in teristics and language fluency affect data col-
data collection (structured and unstructured lection. Every field trip is an opportunity to
interviewing, for example), and in data test the relationship between what people say
analysis. Students who hope to do any serious and what they do. We need to monitor and
work with survey data, for example, need at publish the results of all these naturally occur-
least two courses in applied statistics (through ring experiments.6
 As I said at the beginning of this essay,
anthropologists may disagree, even vehe-
In the past, cultural anthropologists were more mently, about method as epistemology. But
concerned with description than with explan- they share a common interest in and concern
ation and prediction. A good description is still for method at the level of strategy and tech-
its own analysis; being right about what causes nique. Controlling many different methods for
a problem is still the best contribution anyone collecting and analyzing qualitative and quan-
can make to solving it; and good writing- titative data liberates us to investigate what-
telling a good story-is still the best method ever theoretical questions attract us. And con-
any anthropologist can acquire. versely: limited training in methods limits our
Many anthropologists today, however, are ability to think about and solve many impor-
interested in research questions that demand tant theoretical questions. Knowing more is
explanation and prediction, questions like: better than knowing less.
Why are women in nearly all industrial
societies, socialist and capitalist alike, paid less
than men for the same work? Why is medical
care so hard to get in some(societies that pro-
duce plenty of it? Does cultural pluralism sup- 1. Mill felt that the French word positillisme was not
port or undermine the stability of states (as in suited to English, but decided in the end to use Comte's
word, properly anglicized to refer to Comte's concept.
Canada, Belgium, India, Azerbaijan, Yugo-
I've long wished Mill had decided otherwise.
slavia, or Kenya, for example)?
Answering such questions demands greater 2.. See Lewis's contribution to the 1953 volume Anthro-
sophistication in research methods than is now pology Today, edited by Alfred Kroeber. See also Lewis's
customary in anthropology. It requires skills in article on family studies in the 1950 volume of the
comparative statistical research and in gather- American Journal of Sociology. In those days, family studies
ing data from published sources including data was still the province of sociologists. When Lewis wrote
for anthropologists, he stressed the need for survey
that are only available on computer tapes. research as a complement to ethnography. When he
Anthropologists who are involved in multi- wrote for sociologists, he stressed the need for ethno-
disciplinary team research on complex human graphic research as a complement to survey data.
problems-in agricultural development, fer-
tility control, health care delivery, and educa- 3. I was inspired as a graduate by Radin's work on the
tion-need more methods training than has production and presentation of native texts. Radin
followed Boas's lead in native ethnography of indigenous
been the norm to date.
North American peoples. Boas trained George Hunt, a
Knowing about methods for collecting Kwakiutl, to write ethnographic descriptions of
qualitative and quantitative data is more diffi- Kwakiutl culture. Hunt produced over 5,000 pages of
cult in some ways today than it was forty years notes that served as the basis for much of Boas's pub-
ago. There are simply more methods now lished work on Kwakiutl life. Radin trained Crashing
Thunder to write in Winnebago, and Crashing Thunder
than there were then? In some ways, though,
wrote his classicautobiography of the same name.
learning research techniques is easier now. Native ethnography has long held great promise as a
Microcomputers and efficient software make means to develop an authentic emic data base about
text management, statistics, and cognitive do- Indian and other indigenous cultures. I can only surmise
main analysis less intimidating.8 No one can that the physical difficulties involved must have kept
native ethnography from becoming a major method in
control all research methods. But those who
cultural anthropology.
control a certain fraction will be unintimi- In recent years, I have found that microcomputers
dated about learning new methods that be- make native ethnography much easier to do. I simply
come available and that seem appropriate to program a word processor to handle whatever special
particular research projects. characters are needed in a particular language, teach
native speakers of previously nonwritten languages to use tion cross-culturally. Those tests were discredited and
the equipment, and hand the equipment over to them. abandoned by most researchers thirty years ago (Brislin,
For many years, my partner in this effort has Lonner, and Thorndike 1973:109).
been Jesus Salinas Pedraza, a Nahfiu (Otom!) Indian
from the Mezquital Valley of Mexico. Salinashas written 8. For handling relatively small amounts of text material
an extensive ethnography of the Nahfiu, which I (up to say, a thousand pages of field notes, transcribed
translated and annotated (Bernard and Salinas 1989). interviews, etc.) GoFer is an excellent product. Write to
Since 1987, Salinas and Josefa Gonza lez Ventura, a Microlytics, Inc., Two Tobey Village Office Park,
Nuu Saavi (Mixtec) Indian from Oaxaca, Mexico, have Pittsford, NY 14534. For handling large amounts of text
run the Native Literacy Center at the Centro de Inves- material, ZylNDEX is highly recommended. Write to
tigaciones y Estudios Superiores en Antropologia Social ZyLAB, Information Dimensions, Inc., 100 Lexington
in Oaxaca City. They have now trained 75 other Indians Drive, Buffalo Grove, IL 60089. For coding and analyz-
from Mexico, Peru, Bolivia, Argentina, Ecuador, and ing text, consider the following: (1) The Text Handler,
Chile to use microcomputers to write books in their by Gery Ryan. This package works with WordPerfect.
respective native languages. Write to Gery Ryan, Dept. of Anthropology, University
There are many social and ethical issues raised by this of Florida, Gainesville, FL 32611. (2) TALLY 3.0, by
effort. I deal with these, and the history of the project, in Jeffrey W. Bowyer. This program works with ASCII
Bernard 1992. ftIes. Write to Wm. C Brown Publishers, 2460 Kerper
Boulevard, Dubuque, IA 52001. (3) DtSEARCH, by
4. In this sense, participant observation, like everyday David Thede. DtSEARCH works with all of the top
life, is also an exercise in impression management (see word processors, as well as with ASCII flies. Write to
Berreman 1962:11). DT Software, Inc., 2101 Crystal Plaza Arcade, Suite
231 Arlington, VA 22202. (4) THE ETHNOGRPH,
5. Was Mead duped? Whether she was or not has
by John Seidel. This program works with text from
nothing to do with her overwhelming contributions to
major word processors. Write to Qualis Research
anthropology. All ethnographers run the same risk.
Associates, P.O. Box 2240, Corvalis, OR 97339. (5)
6. How much, for example, can we rely on informants ANTHROPAC, by Stephen Borgatti. This program
to tell us about their income? The size of their land- helps researchers collect and analyze numerical data. It is
holdings? Where their grown children are? Data on the particularly useful for multidimensional scaling, hier-
nature and causes of informant inaccuracy would help archical clustering, consensus modeling and other
illuminate the relation between cognition and behavior, methods that operate on similarity matrices. Write to
between the internal and the external worlds of human Analytic Technologies, 306 South Walker Street,
beings around the world. Columbia, SC 29205.
All of the programs mentioned here have been
7. Methods do come and go, however. No one uses so- reviewed in Cultural Anthropology Methods from 1989 to
called "culture free" tests anymore to investigate cogni- 1993.

H. Russell Bernard is Professor of Anthro- Jesus Salinas Pedraza, 1989); Technology and
pology at the University of Florida. He has Social Change (edited with Pertti Pelto, 2nd
done field research in Greece, Mexico, and edition, 1987); and a series of articles on social
the United States and on ships at sea. Bernard network analysis (with Peter Killworth and
was editor of Human Organization (1976-1981) others).
and of the American Anthropologist (1981-
1989). Recently Bernard has been working
In the summer of 1959, as a junior at Queens Col-
with Jesus Salinas and other Indian colleagues
lege, I went to Mexico to study Spanish and came
in Oaxaca, Mexico, to establish a center back knowing that I wanted to be an anthropolo-
where native peoples can publish books in gist. As an undergraduate, I studied with Ernestine
their own languages. Bernard's best-known Friedl, Hortense Powdermaker, and Mariam Slater.
contributions are Research Methods in Cultural Late in my senior year, Powdermaker told me
Anthropology (1988); Native Ethnography (with about a new Ph.D. program just opening at the
University of Illinois. Perhaps I could get in there, perience how important it is for students to be-
she said. come involved in research projects early and often.
Illinois in 1961 was an intense, intellectual envi- During my Ph.D. studies, Julian Steward,
ronment. I studied with Kenneth Hale and Duane Dimitri Shimkin, Joe Casagrande, and Kris Leh-
Metzger for my M.A. in anthropological linguis- man encouraged me to pursue my interests in
tics, and then with Edward Bruner, Oscar Lewis, quantitative data analysis. In Casagrande's seminar
Julian Steward, Dimitri Shimkin, and Joseph on cross-cultural research, I first learned to use the
Casagrande for the Ph.D. Human Relations Area Files and to test hypotheses
Metzger was part of the (then) new ethnoscience using cultures as units of analysis.
camp. The goal was to write the grammar of a cul- I don't recall anyone labeling all this "positivism"
ture--to learn what a native speaker of a language in those days, or worrying about whether my in-
knows about, say, ordering a drink and to lay that terest in scientific, quantitative research was un-
knowledge out clearly. healthy. I read works by Tylor, Boas, Kroeber,
Making cultural grammars turned out to be Driver, Wissler, Murdock, and Roberts and
harder than anyone imagined. Metzger offered a noticed that all of them did quantitative work and
hands-on seminar. With a few other students, I published reams of ethnographic work as well. I
spent a semester working with one Japanese house- found this mix of qualitative and quantitative
wife, learning and mapping the implicit rules she methods to be very sensible.
used for deciding how to cut and arrange vegeta- My major doctoral professor was Ed Bruner. Ed
bles on a plate. became identified with symbolic anthropology and
It was an enormous effort just to keep track of I went in a different direction. But Ed taught me
the data. One of the other students got the com- to write, and to understand that seeking knowledge
puter to sort and print the whole corpus every time was only half the battle. You have to be able to tell
we learned a new rule. The people over at the others what you have learned, to engage their
computer center thought this was pretty quaint, attention, and to keep them from closing the book
but this systematic approach to data gathering and before you have finished your argument. This may
the idea of using computers to make light work of be one of the few things that positivists and inter-
complex data-management tasks have stayed with pretivists fully agree on; but for my money, it's the
me ever smce. most important thing of all.
Ken Hale was Carl Voegelin's student. Like Carl In 1972, I spent a year at the Scripps Institution
(and like ~oas and his early students before him), of Oceanography, where I met Peter Killworth, an
Ken worked closely and collaboratively with ocean physicist. We decided to study problems
Indian colleagues. The model was to help Indian together that (1) neither of us could tackle alone,
colleagues produce their own texts, in their own (2) both of use agreed were sheer fun, and (3) were
languages, and then to use the texts for linguistic not in the mainstream of research in either of our
analysis and for cultural exegesis. Ken's example, disciplines. We also agreed that we would not let
and the tradition it represented, led to my lifelong our joint projects get in the way of our separate
collaboration with Jesus Salinas, a Nahiiu Indian research careers. (He is in ocean modeling.)
from the Mezquital Valley in Mexico. We did, in fact, have a great time doing a series
Jesus was my informant in 1962 when I did the of papers on informant accuracy, and we are hav-
research for my M.A. thesis on the tone patterns of ing just as much fun now testing a network model
Otom! (called Nahiiu in those days). In 1971, I for estimating the size of populations that you can-
became Jesus's informant, teaching him to write in not count (like the number of rape victims in a
Nahiiu and, in the 1980s, to use a Nahiiu word city). Peter has taught me a lot about data analysis.
processor. I'm still working with Jesus, who now I have also benefited greatly from my association
heads the Native Literacy Center in Oaxaca, where with Pertti Pelto. We began the National Science
Indians from around Latin America train in using Foundation Summer Institute on Research Meth-
computers to write and print books in their own ods in Cultural Anthropology in 1987. Stephen
languages. Borgatti joined the teaching team of the summer
Much of my career, then, was shaped by my institute in 1988, and I have learned a lot from him
work at the M.A. level. I learned from that ex- about new analytic methods.
 My intellectual biography is still being written. I of contemporaries, of junior colleagues, and of stu-
can look back and see the influences of my profes- dents. This is what makes anthropology so exciting
sors clearly. But just as clearly, I see the influence for me. The learning never has to slow down.