Review: Hermeneutics in Anthropology
Reviewed Work(s): Interpretive Social Science: A Reader by Paul Rabinow and William M.
Sullivan: The Said and the Unsaid: Mind, Meaning and Culture by Stephen R. Tyler
Review by: Michael Agar
Source: Ethos , Autumn, 1980, Vol. 8, No. 3 (Autumn, 1980), pp. 253-272
Published by: Wiley on behalf of the American Anthropological Association
Stable URL: https://www.jstor.org/stable/640129
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
American Anthropological Association and Wiley are collaborating with JSTOR to digitize,
preserve and extend access to Ethos
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
Hermeneutics in Anthropology:
A Review Essay
MICHAEL AGAR
Recently two books appeared that, for me at least, announced the
arrival of something with the mysterious name of "hermeneutics" into
American cultural anthropology. One volume, Interpretive Social
Science: A Reader, (1979) is edited by Paul Rabinow and William
M. Sullivan. After an introductory essay, the reader consists of three
programmatic pieces by Taylor, Ricoeur, and Gadamer. These are
followed by sample applications of hermeneutic ideas in such fields
as literature, political science, history of science, psychoanalysis,
and the sociology of religion. Anthropology is represented in two
essays by Clifford Geertz-one, his classic paper on the Balinese
cockfight; the other, a more recent piece on anthropological
understanding.
The second volume is a single-authored work entitled The Said
and the Unsaid: Mind, Meaning and Culture, (1978) by Stephen R.
Tyler. Tyler is concerned with the everyday use of language. With
this concern in mind, he evaluates a variety of formalisms in anMICHAEL
AGAR is Research Associate at the Language Behavior Research Laboratory,
University of California, 2220 Piedmont Avenue, Berkeley, CA 94720.
Support by RCDA 1 K02 DA00055 is gratefully acknowledged. Helpful comments were offered
by Gelya Frank, Paul Rabinow, Bambi and Buck Schiefflin, Steve Tyler, and the
Editors of Ethos.
ETHOS 8:3 FALL 1980
Copyright ? 1980 by the Society for Psychological Anthro
0091-2131/80/030253-20$2.40/1
253
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
254 ETHOS
thropology, linguistics, and psychology, showing
representations of everyday language. He devel
"meaning as contingency," and draws on hermene
patible framework within which to construct his vie
None of this material is particularly light reading.
the dilemma of wondering whether this is due to lac
the part of the authors or the nature of the materia
nant fit with traditional ideas of social science. Be
material on interpretation carries with it a dizzy r
one is reading applies even as it is being read. I s
some of the illustrative materials in the Rabinow and Sullivan reader
provide good concrete examples of a hermeneutic point of
view-the articles by Geertz, Fish, and Hirschman, in my opinion,
stand out in this regard.
As I worked on my own hermeneutic problem of constructing a
meaning for the two volumes, I decided to focus more on the general
nature of the perspective than on specific instances of its use. For
this reason, my essay draws most heavily on the initial programmatic
chapters in Rabinow and Sullivan, as well as the later portions of the
Tyler volume. This essay is not, strictly speaking, a book review.
Rather, it is an attempt to come to terms with a perspective represented
in two recent books in anthropology.
The term "hermeneutic" comes from ancient Greek. Hermeios
referred to the priest at the Delphic oracle. The verb hermeneuein i
generally translated as "to interpret"; the noun, hermeneia, as "interpretation."
The name of the wing-footed messenger god
Hermes, is also related. Hermes was credited with the discovery o
language and writing, and "is associated with the function o
transmuting what is beyond human understanding into a form that
human intelligence can grasp" (Palmer 1969:13). The term was ap
parently used in the senses of expressing aloud in words, explaining
a situation, and translating.
The field of hermeneutics in modem times has something of an
intricate history. Palmer (1969:33) offers a roughly chronologica
sense of the different uses of the term:
(1) the theory of biblical exegesis; (2) general philological methodology; (3) the
science of all linguistic understanding; (4) the methodological foundation of
Geisteswissenschaften; (5) phenomenology of existence and of existential
understanding; and (6) the systems of interpretation, both recollective and
iconoclastic, used by man to reach the meaning behind myths and symbols.
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
HERMENEUTICS IN ANTHROPOLOGY
He associates Schleiermacher with (3), Dilthey with (4), H
(later, Gadamer) with (5), and Ricoeur with (6).
Since hermeneutics has such an elaborate history, one w
where it has been. The idea of "interpretation" is disc
American cultural anthropology, but it is usually presented
its philosophical context. In their introduction, Rabinow a
van suggest that part of the reason for this neglect lies in th
commitment in Western culture to a particular notion of "s
That notion emphasizes timeless, context-free axiomatiz
controlled manipulation of clearly defined variables, and ma
tical measurement in a quest for objective knowledge ab
world. Rabinow and Sullivan argue that operating within thi
sophical context is an epistemological error for the social
It is an error just because it treats, as problems to be elim
the most important aspects of one human trying to underst
others. Rather than trying to eliminate "observer effects," f
ple, one focuses on the interpreter as participant in a t
which guides and is changed by the process of under
another. Rather than striving for "objective knowledge,"
cepts the fact that knowledge is situated in a historical m
some of whose presuppositions are difficult, if not impossib
ticulate. And, at its most radical, one no longer speaks of
tion," but simply assesses an interpretation for its coherenc
ability to make sense of one tradition from the point of
another.
These and other issues will be revisited when the material is
discussed in more detail in a moment. For now, I would like to note
that hermeneutic philosophy provides a more comfortable fit with
the kinds of problems an ethnographer routinely faces than does the
traditional notion of science discussed above. As evaluated by traditional
guidelines, we often have something of a self-conscious preoccupation
about whether or not ethnography is really a "science."
Our tradition emphasizes such difficult notions as holism, context,
and symbol. We have always talked, sometimes with embarrassment,
of the importance of the intuitive apprehension of pattern.
We stress learning rather than control, often using such metaphors
to characterize our work as "student" and "child." Our literature on
field methods is full of concern with the perception of the ethnographer,
the role relations between ethnographer and informant, the
emergent nature of ethnographic fieldwork, and the importance of
255
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
256 ETHOS
the individual ethnographer in the nature of
ethnographic report. Hermeneutic philosophy te
issues belong at center stage; they are not embarrass
be avoided on the way to a good hypothesis.
Hermeneutics also brings with it some serious prob
strong commitment, born of 12 years of interdis
work, to improving ethnographic methodology. In s
tings, ethnography often plays a subversive role, in t
ethnographic report of group X usually complicates,
tradicts, the assumptions about group X that inform
If one expects administrators to act at all on the bas
viant" knowledge, one had better present a credible
lem becomes more complicated when one consider
pient of a report probably has a notion of social scie
ditioned by the classic idea of the testing of h
hermeneutics teaches, in its most radical form, that
possible in the classic scientific sense.
But this position is contradicted by one's intuition
terpretations are better than others, at least from a
point of view. For example, I am now reading M
novel Blood Tie-fortuitously an excellent novel as
hermeneutics. Among many possible examples fro
me mention just one. On a festival day in a Turki
the local women back away and make a discreet gestu
evil eye to protect themselves against the visitin
American women, whom they see as indecently dr
observer, in contrast, comments on the politeness of
as they make way for village guests. As an ethnogra
on the village point of view, which interpretation w
Or, let me use another example I have discu
elsewhere (Agar 1980a). Several years ago, while work
South Indian village, I was about to walk to town.
cook wrapped a lunch of bread and spinach in a
small piece of charcoal on top. I didn't know what to
own sense of charcoal and food had something to do
Only later did I learn that I was traveling at midd
spirits were particularly active. They were especi
lone travelers carrying food, and the charcoal w
package to serve as a spirit repellent.
Or consider another example. When I first bega
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
HERMENEUTICS IN ANTHROPOLOGY
York on the issue of the use of methadone in the streets, I w
prised to hear some addicts talk about going to a methado
as "going to cop." It was among my first clues that, from
point of view, methadone was interpreted quite differently
the medical profession. The clinic was the "dealer," the
was a "junkie," the methadone was the "dope," and g
methadone was "copping."
Any ethnographer could provide numerous examples of his
own to support the intuition that, among all possible interpr
of an action, some of them are ethnographically "better
others. In fact, one of the main contributions of ethnograph
variety of theoretical and geographical areas, has been to sho
a prevailing interpretation of some human action was in fac
rect from the group's own point of view.
These three short examples are offered to ground what is
become a rather abstract discussion. We will return to them
to assess the hermeneutic positions represented in the two v
with a special emphasis on the problem of validation.'
THE PERSPECTIVE OF HERMENEUTICS
We begin with Gadamer's essay on the "principle
productivity." Ethnographers participate in a traditio
in a world of meanings that give a sense of their past, t
situation, and their future possibilities. They cannot
from their world, for there would be no meanings in ter
they could characterize their own existence. They live w
ticular "historical horizon."
However, they are capable of what Gadamer calls "historical selfconsciousness."
Through an awareness of their historicality, they are
capable of bringing some aspect of their world into consciousness
and reflecting on it. With this consciousness comes not only
awareness, but a sense of other possibilities. A piece of their world,
previously implicit, stands revealed as a "prejudice" which can be
changed. Through historical self-consciousness, then, there is the
possibility of altering a world, of changing the horizons of a history.
1 Undated references in the following discussion refer to the Tyler volume or to articles included
in the Rabinow and Sullivan reader.
257
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
258 ETHOS
There are many ways that one's world can partially
consciousness. Unifying them would be a confrontat
person and that person's world-something goes w
problem; expectations are violated. Such "breakdo
consciousness a part of the world which previously
awareness. Awareness of the problem's existence is th
a hermeneutical situation. The interpreter is tor
belongingness to a tradition and his distance from th
are the theme of his investigation" (p. 155).
To deal with the problem, now bracketed as an "obj
pretation, the person attempts to fuse the horizons
the object. One "appropriates" the object, brings it
world. One begins by reflecting on his or her prec
"preunderstanding" brought to the encounter. One
"anticipation of perfect coherence" (p. 153). One m
forth in a dialectic between one's world and the obje
meanings, altering one's horizons, until the object an
unified into a coherent whole. One thus "understa
and one's own convictions within the same experient
This process of understanding is characterized as "a
cular relations between the whole and its parts" (p. 1
preter's world and the meanings for the object ac
other dialectically-one cannot be understood in s
the other. This is one form of the "hermeneutical circle." Rather
than seeing this as a problem, though, Gadamer discusses it as a
"fundament of positive and productive possibilities for understanding"
(p. 156). This is his principle of historical productivity.
How does one know when he or she has understood? "Coherence"
is one clue -we understand when the object disappears as a problem
and becomes something we can make sense of in terms of the meanings
in our (now altered) world. Gadamer also states that we have
understood when "we do not lose any ground in the face of any possible
arguments advanced by one's adversaries" (p. 131). By his own
argument, though, it is difficult to see how the "possible arguments"
could be enumerated.
Once one becomes accustomed to the jargon, Gadamer's disc
sion provides a framework within which to talk about t
ethnographic examples given earlier. In the example from t
novel, there is no problem, no breakdown. Rather, we see two di
ferent traditions used to make sense of the same behavior. The
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
HERMENEUTICS IN ANTHROPOLOGY
Englishman and the villagers both come up with coherent in
pretations. Though the Englishman does make sense of the s
tion, we somehow want to show eventually that, in spite o
coherence, it is ethnographically weak.
In my own examples from India and New York, a breakd
does occur. It occurs partly because, as an ethnographer,
trained under the charters of "culture" and "symbol" to look fo
In the methadone example, I have an interpretation, like
Englishman, provided by my own tradition. The person I see
methadone patient." However, I also have a background
stresses attention to informants' meanings, so that when I hear
ing to cop," it produces a breakdown in my understanding. B
ning with the difference between my own tradition's definition o
clinic" and the informant's use of a term that applies to de
heroin, I move back and forth between observations and conv
tions until I understand why the informant said "going to cop."
the process, my own tradition is changed, for I will never again
look at methadone clinics from the medical point of view.
In the case of the charcoal included with my food in South In
the breakdown occurs even before I know anything about g
meanings. The stream of behavior-placing the charcoal on to
some food -is simply something my tradition does not enable m
interpret. I am confronted with actions that make no sense. Ag
move back and forth until I understand it, and again my tradit
altered in the process. For these examples, then, Gadamer w
show the similarities and differences as follows: In the first case
interpretations exist side by side, both perfectly coherent. In the
ond, two interpretations exist, but the ethnographer moves to u
derstand one after beginning with the other. In the third case,
action is initially uninterpretable--a breakdown in the classic se
of the term -and the ethnographer moves to appropriate it into
tradition.
Not that there aren't problems with Gadamer's argument. All
understanding is historical - therefore, "objective" knowledge,
knowledge free of context, is not possible. But this leaves us in
something of a lurch in terms of methodology, since in his article
Gadamer offers only "coherence" and "ability to defend" as the
main sources for an evaluation of an interpretation. Palmer
(1969:67) argues that this is not a fair criticism-he writes that
Gadamer recognizes the need for more methodological concern, but
259
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
260 ETHOS
that his prime interest, explicitly stated, is in an anal
tological nature of understanding.
Hirsch, a critic of Gadamer, argues that the method
lems result from a failure to separate "meaning" from
A text (this is the type of object that interests Hirsch
meaning which is "changeless, reproducible, and d
(Palmer 1969:81). This meaning has any number of "
depending on the historical context in which it is inte
ever, meaning, not significance, is the problem for
Furthermore, Hirsch argues that the meaning can b
appeal to the author's intention, about which ev
gathered. Palmer dismisses Hirsch's arguments, feeling
reduced hermeneutics to the "logic of validation" on
Ricoeur, in his article, takes an intermediate posit
Hirsch's idea of the "dialectic of guess and validation."
gests approaches from the "logic of subjective pro
"legal interpretation" as methodological sources. He
validation carries with it the possibility of invalidation
per's discussion of the importance of "falsification"
does not fully develop these ideas, though. A possibl
these different positions will be offered later. The iss
notes, that there are methodological problems "to m
quake."
HERMENEUTICS AND SOCIAL LIFE
In their essays, Taylor and Ricoeur extend hermeneutics
ly into the study of human life. Though they differ somew
I think, characterize behavior in ways that fit in with G
description. Behavior has a context; it is situated in space a
and connected to other aspects of the situation; it is intend
human agent for another; and it has "a sense distinguish
its expression" (Taylor) or "refers to a world" (Ricoeur).
Now the complications increase. It is one thing to talk of
tology of understanding when the object is not conscious for
example, discusses Heidegger's frequent use of the exam
"hammer." It is quite another thing to talk about under
when the object involves consciousness of others. And i
complex still when the object becomes a "text" in an as yet
metaphorical sense.
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
HERMENEUTICS IN ANTHROPOLOGY
Ricoeur devotes much of his chapter to a discussion of the
of text. In a text, discourse is "inscribed" and detached from its
"moorings" to the psychology of its author. It is freed from its ostensive
reference to the situation of its doing, and addressed to a potentially
unknown and large audience. Obviously a text also has some
natural edges as a phenomenon-it is a finite, neatly bounded entity
which guarantees that separate interpreters will agree that it is the
common object of interpretation.
He then extends the text metaphorically to "action," noting that
it has a propositional content which can be identified in other actions;
it is inscribed in the sense of such social metaphors as "record"
and "reputation"; it can have consequences for other actions different
from it; and it is open to multiple interpretations by others
who can "read" it. Action, in other words, does have textual
characteristics.
Taylor, with a critical eye cast towards the field of political
science, takes a different line. Action, he notes, is obviously for and
by a subject. He goes on: "We make sense of action when there is a
coherence between the actions of the agent and the meanings of his
situation for him" (p. 35). He adds that "making sense" does not
necessarily imply a rational account.
But then, he notes that we are dealing with both "text analogue"
(behavior) and "text" (accounts). He discusses how the agent experiences
a situation in terms of meanings, which are then "shaped"
by the language through which the agent lives, which are in turn interpreted
by the outsider. He adds the possible complication of the
outsider's interpretation in turn feeding back, altering the world
within which the agent does his or her own interpretations.
He goes on to characterize language and situation as coconstitutive-practice
"has to do in part with the vocabulary
established in a society as appropriate for engaging in it or describing
it" (p. 47). In fact, he distinguishes what he calls "experiential
meaning" from linguistic meaning just because linguistic meaning is
the meaning of signifiers about a world of referents. There is a fine
line between language as constitutive through "description" and
reference, but let that pass for now.
He characterizes experiential meaning as "not just in the minds of
the actors but out there in the practices themselves" (p. 48). In other
words, the meanings that the outsider is after are intersubjectively
261
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
262 ETHOS
available. He briefly develops the idea of "com
which are the "basis of community." These comm
"interwoven" with a "powerful net of intersubjectiv
51). This discussion begins to sound somethin
thropological idea of "themes" (Opler 1959), but
pursue it in this chapter.
With Taylor and Ricoeur, than, we get the feelin
possibilities, and equal number of accompanying h
strong sense that we've been there before. The prim
terpretation for an ethnographer is a group o
themselves interpret. The horizon we are trying
ethnographers is in constant motion, and it consists
ferent individual horizons which are unlikely to be ni
With Taylor and Ricoeur, we begin to see the problem
sion of philosophical and literary hermeneutics in
At the same time, the problems often have a familia
them represent issues, explicit and implicit, that
ethnography itself.
While discussing the problems introduced by exten
eutics into the study of human action, let me bring
material from Tyler's book. He is interested in th
works in everyday life. With a continual view towar
he takes an impressive tour through a variety of per
linguistics, anthropology, psychology, and phe
sociology. His critique throughout is an attack on
disembodied, mechanical representations of human
be reconnected to the world by means of "awkwar
rules" (p. 461). I was sometimes derailed by what I sa
reflexive denial, with his own formalized representa
tique on formalism. But for the most part, it is an i
thesis, with some delightful nose-thumbing at a
malisms along the way.
Tyler explicitly uses hermeneutics to make a transi
study of meaning as necessity to meaning as cont
comes to hermeneutics as a compatible framework w
further develop what he sees as the two fundamenta
issues in formalist approaches-will and representa
tique, he repeatedly notes, by appeal to anecdotal ins
day life, that use of language is situational and pu
stresses the constitutive nature of language, showing
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
HERMENEUTICS IN ANTHROPOLOGY
talk is an unfolding, emergent process-meaning, in other w
is contingent. Each step is "reflexive, confirming both itsel
prior understandings and serving as a basis for further interpr
tion" (p. 364).
To get to this point, he first describes the various ways
speakers tie discourse together. He travels across such well-k
concepts as presupposition, entailment, equivalence, paraph
and metaphor, anaphora, synonymy, ellipsis, and several oth
His discussion varies from the usual formalist approaches, howe
For example, he notes that synonymy is formally defined as
synonymous to Y if they are mutually substitutable in all conte
with no change in meaning. Through examples he shows that
more sensible to assume that in everyday use the context usually
change with such substitution, but that nonetheless, hearers wi
cept the change as "close enough." He then shows how a spe
can, in fact, slowly shift meanings so that by the end of a disc
the meaning of the "synonym" is actually quite different from
initial way of talking. The various shades of difference, unnotic
they gradually occur over extended discourse, add up to a sub
tial difference by the end.
He then considers, as part of the hermeneutic circle, the prob
of arriving at a representation of the whole meaning of a text in
tion to its parts. This circle is a bit different from Gadamer's u
the idea, discussed earlier. For Gadamer, the circle encompasses
interpreter and the object. For Tyler, now that the object is a s
ject who also interprets, the focus shifts to how the situated ag
coordinates meaning between part and whole of the expression h
she is producing or attending to. Same circle, different scope
A guideline for successful interpretation is to judge if that in
pretation is reasonable in a given context. Two of the problems
everyday reasoning, though, are that it rests on tacit understan
that are only made explicit in problematic cases, and that the in
preter must have the capability to treat utterances as an "index
general pattern" (the so-called documentary method of interp
tion).
With these problems in mind, then, an interpretation is
reasonable if it fits "into an unfolding scenario, plan, or script
characterized by causality, likelihood, moral necessity, typicality, or
means-ends efficacy. They are implicated in a linguistic and extralinguistic
whole. These normal schemata, in other words, contain or
263
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
264 ETHOS
exemplify means of inference" (p. 399). In other wor
stitute "the causal texture of the world." He adds that
yield determinate intelrpetations, for there is obviousl
for perspectival relativity, but the interpreter looks for
between intentions and underlying patterns" (p. 402
Tyler finishes with the general point that "we un
terances and texts by providing them a context of
perience" (p. 454), which contains "time, place, perso
and acts corresponding to a schema of possible experien
This process is "immediate, tentative, and whole." Perh
always that tentative, for elsewhere he notes that "as th
terpretive edifice grows, fewer and fewer are the thing
firm it, so much so that even the speaker's overt attem
firm it may be taken instead as confirmation" (p. 41
Tyler, Taylor, and Ricoeur, in their different ways, sh
from the hermeneutical situation of interpreter and
hermeneutical properties of the "object" itself when
other historically situated persons. Ricoeur argues,
social action has textual properties and is therefore
hermeneutic approach. Taylor makes a similar argum
hermeneutic aspects of everyday life, though he does s
tention of criticizing the concern in political scienc
data." Tyler, on the other hand, takes after a range
that claim to represent different aspects of the hu
language. In comparison with the Gadamer essay, w
began this discussion, the outside interpreter is curi
That leads into a set of problems that I would now li
ETHNOGRAPHIC HORIZONS
First of all, in my mind there is some confusion over th
hermeneutic principles as a guide for ethnographic work v
use of hermeneutics as a guide to the properties of the
tivities that the ethnographer confronts. Clearly, they ar
in both contexts. In fact, the two Geertz contributions illus
In the one, Geertz shows how the Balinese interpret the coc
the other, he discusses "anthropological understanding"
erally, showing how the ethnographer works within a her
circle that articulates details, the whole of which they are
tion with reference to each other.
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
HERMENEUTICS IN ANTHROPOLOGY
One interesting possibility, argued by Gadamer with h
on the ontological unity of all understanding, would
two uses are but different instances of the same pro
they? On the one hand, Spradley, in his recent book on
graphic interview (1979), introduces the idea of eth
discussing the situation of a person who has just mov
neighborhood. In my own introductory treatment of
(1980b), I also use an everyday problem-explaining a
tially rich trip to a friend -as a similar introduction to
of ethnographic method. Along other lines, Hill-Burn
argued that development of ethnographic method is i
tribution to cultural learning theory in general. Wi
modification, her argument could be translated into
terms to support Gadamer's position.
On the other hand, in Spradley's and my discussion, th
were introductions. The hermeneutic problem confr
ethnographer is in fact similar to that confronted by an
newcomer to a group. But that does not necessarily m
process by which understanding is achieved is the sam
impossible issue to resolve, since the nature of "
"ethnographic" understanding (if indeed they are dis
well understood at any rate. It is difficult to compare u
Perhaps ethnographic understanding will turn out t
"historically self-conscious" version of natural understan
I had to wager on this particular issue, I think I would b
nographic understanding will include everyday under
will be broader and more self-conscious. As a side bet
wager that everyday understanding itself, under furthe
will be shown to vary in some interesting ways--a concl
reached from a reading of Tyler's book. We now approac
tigo of reflexivity by adding that a study of understand
the worlds of its occurrence may in fact be a violation of
tological characteristics. Understanding, like the subje
that unite within it, may be bound by the horizons of it
Now to a second problem: the object of interpretat
hermeneutic circle. Ethnographers are primarily con
neither hammers nor texts. The groups that they a
themselves interpreting in worlds that are both muc
much less neatly bounded than texts. The general statem
hermeneutic circle is that the whole is interpreted in
265
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
266 ETHOS
parts, but the parts are in turn interpreted in terns
But "whole" and "parts" may take on different mean
in the discussion of the two Geertz articles, and in t
earlier in the paper between Gadamer and Tyler.
Since the group activities in themselves consist p
pretations, we can look at the hermeneutical circle as
tic of these activities. This we see emphasized in Geer
cockfight and in Tyler's book. On the other hand, w
the circle with reference to the ethnographer en
group. In this sense, the "whole" is the encou
ethnographer in his or her world with a group foreig
The "part" is the new meanings that allow this world
to be fused or appropriated. This we see emphasi
essay on anthropological understanding (thoug
epistemological sense) and in Gadamer's article.
Again we are back to the first problem. Are these
the concept of the hermeneutical circle? On the one
ferent hermeneutic circles suggest different questio
ethnography. As a characteristic process of group life
gests ways to evaluate the ethnographer's characteriz
article and Tyler's book, for example, are critiques o
just because they ignore such characteristics.
But as a problem in ethnographic research, the
more of an account of the ethnographer's world as a
of understanding the interpretations offered. There
in these two volumes to illustate this point, intere
There are, however, some recent personal account
written from an explicitly acknowledged hermene
(Rabinow 1977; Dumont 1978). In fact, in support of
between the two circles. Dumont wrote two books, o
Panare" and the second one, cited here, "directed t
ing, apprehending, and interpreting the 'and' of
which my fieldwork built between an 'I' and a 'th
Now on the other hand, if we accept the ontologica
understanding, then the apparent confusion results f
cial distinction imposed on a dialectic reflexivity.
hopefully more lucidly, an increased understanding
eutical circle of ethnographer and group is als
understanding of the hermeneutical circle for the pe
stitute that group, and conversely. To put it yet anot
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
HERMENEUTICS IN ANTHROPOLOGY
hermeneutic perspective, as ethnographers get better at thei
(in the sense of how they do it), they will get better at their
the sense of how they describe how groups get things do
conversely.
There is a risk here. A while ago I wagered that eventually
graphic understanding would be seen as different from
understanding. If I win the bet, and if we don't treat the
plications of the hermeneutical circle as distinct, then w
adopted a universal model of human life that may lead us
the central issues of a group's life as the group lives it. Peop
been seen as, among other things, large rats and ambulato
puters. I would like to think they might appear more com
and interesting as inarticulate hermeneutic philosopher
trained ethnographers, but I am not sure this would be the
am I sure it would be accurate.
Now, how can one write of accuracy after this extended discussion
of the plurivocality of the interpreted object, the relativity of the
historical situation of the interpreter, the inescapability of th
hermeneutical circle, and so on? First of all, recall the discussion of
Settle's novel and the two anecdotes I offered from my own fieldwork
in South India and among urban American narcotics users.
the examples, different interpretations were suggested for one action.
Both were acceptable from a hermeneutic point of view- they
made sense of the action - and given the notion of plurivocality, it
no surprise that more than one interpretation was possible.
However, I don't think there's an ethnographer alive who, give
the information that the study was about Turkish villagers, Sout
Indian villagers, or urban junkies, would not pick one of the interpretations
as ethnographically better than the other. This is not an
argument for a single interpretation, but an argument that some ar
better than others. It is crucial that ethnographers have a way t
document this kind of judgment in a way that can be publicly
displayed in a credible manner.
Let's consider these three issues -the scope of the circle, the properties
of the studied group, and the issue of validation-by returning
to the three ethnographic examples in more detail. The firs
thing to notice is that the examples are all situational. Lik
ethnography itself hermeneutic philosophy emphasizes what Goffman
(1964) long ago called "the neglected situation." The second
thing is that the situations are constituted through the words and ac-
267
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
268 ETHOS
tions of informants and an ethnographer. In the Turk
course, there is difference. For the Englishman, the a
with "politeness"; for the villagers, it is seen as "avoida
eye."
I don't mean to say that the Englishman's sense of the action is irrelevant,
especially, say, to an ethnographer in the village interested
in the impact of tourism. But, to address the issue of validation, I do
want to argue that, for an ethnography of Turkish village life, the
Englishman's interpretation is wrong in some important ways. The
problem is to figure out explicitly how this assessment can be made
by an outsider reading an ethnographic report. While the issue will
not be resolved here, I would like to suggest at least a couple of considerations
that might point towards some solutions.
Let me now shift to my own two examples. In both cases, the
natives constitute their world in one way; the ethnographer constitutes
it in another. How does the ethnographer know when the job
is done? When has the ethnographer, through the classic work of observation
and conversation, understood the situation in an ethnographically
adequate way? Put another way, how can we begin to unearth
the presuppositions that guide the evaluation of ethnographic
accounts?
First of all, recall that the two examples are just that-fragments
of long-term ethnographic studies. Surrendered back into the flow
of ethnographic work, two important possibilities for validation
emerge-possibilities that occur just because we work with living
human groups rather than disembodied texts. The first possibility is
indicated by our commitment to make sense of informant interpretations.
In the Turkish village, the Englishman is not even aware
that a different interpretation exists. He lacks what we might call an
ethnographic attitude. In my two examples, I not only am trained to
look for different interpretations by group members, but I am supposed
to try to make them happen. When I think I understand
something, I should, through further observation and conversation,
try to find something that doesn't fit.
In this commitment to group meanings as the ultimate authority
of our ethnography, we find one important validation anchor. Of
course the ethnographer brings to the work his or her tradition
within which similarities and differences are assessed, an issue amply
documented by the differences in restudies. But even so, given a
commitment to report group situations and group inteirpetations of
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
HERMENEUTICS IN ANTHROPOLOGY
those situations, it would seem that ethnographers nevertheless
tend to similar phenomena.
There is the counter-argument that the process of interpr
alters what is interpreted (see Rabinow 1977). At this point, my
bias is to feel that our impact is more often than not overrated
of the reason that ethnographers routinely think in terms of a
for their work is so that community activities will return to n
after their entry. Few people, I think, will maintain a special se
response to an ethnographer in all situations for such a long pe
of time. Besides, in my own experience in the drug field i
1960s, four ethnographers trained in three different d
plines-two in different northeastern cities, one on the west
and one (myself) at a Kentucky treatment center-independ
came up with descriptions of the heroin addict's world that
remarkable both for their consistency and for their shared
ferences from the then dominant view of the addict as a socialpsychological
failure. Later, in the 1970s, another ethnographe
and I, working independently in two northeastern cities, made som
similar statements about methadone use from the street point o
view. Something is going on here, and I think it is our shared im
plicit commitment to make sense of group meanings, whatever else
we then do. This issue points to an area in need of methodologic
development.
A second methodological issue is raised by the continuity of group
life. As we make sense of situation 1, 2, ... N, we do not treat each
de novo. Rather, we look for connections, for threads or themes that
recur across them. I think this quest is connected with Taylor's ideas
about "common meanings" and Ricoeur's discussion of "depth
semantics." As discussed elsewhere (Agar 1980a), the charcoal example
from South India is eventually shown to connect with other
situations in the village-situations related to spirits, patron-client
relations, social isolation, time, and space. The point is that, as illustrated
in Geertz's articles, the ethnographic "whole" that we construct
in order to make sense of the "parts" we participate in, leads
us to have expectations about future situations. These expectations
can be systematically checked.
There are some other methodological implications in the hermeneutic
literature reviewed here that are also related to ethnographic
work. For example, there is often an emphasis on "breakdown."
Either through the conscious discipline of the ethnographer or
269
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
270 ETHOS
through encounters with the group, problems arisemanifest
themselves and suggest new possibilities. I am
here of several scattered discussions. For example, there i
(1957) use of conflict cases to reveal implicit principles th
lie in the background of group life. Or there is Garfinke
monition to test one's understanding of implicit social ru
triving to violate them and note group members' reaction
is a suggestion by Nadel (1939) that one can frame specifi
to push interviewees to contradict one's growing expectat
they see a particular issue.
There are no doubt other examples as well. The point
are dealing here with a key theme shared by ethnograph
meneutic (and perhaps all) worlds; differences, conflic
lems are where the information is. For ethnography, t
pervasive, as a theme should be, running from a general
to the problem of interpretation down to the level of fo
specific questions to ask. With further development, it is
sibilities for validation strategies at all levels of ethnogra
There is another contribution of a hermeneutic pers
methodological issues in ethnography. There is abundan
on ethnographic methods, but it is widely scattered and
with no overall context serving to organize the discuss
cludes such issues as restudies; critiques of Euroameric
raphy by "native" anthropologists; the "personal equation
tions and changes in roles of fieldworkers, field assistants
mants; the emergent nature of ethnographic models; p
the "scientific" treatment of voluminous field note/informal interview
material; informant responses to ethnographic reports; and
many, many more (see Gutkind and Sankoff 1967). One of the contributions
that I think a hermeneutic perspective will make to
ethnography is to suggest a context that includes such issues as interrelated
problems.
Related to this, another reason I find the area promising is as a
solution to a problem in applied anthropology. I have worked in the
area of U.S. drug research for several years. Needless to say, the
standard of science that guides evaluation of research in that area is
a traditional positivist one. As an ethnographer interested in having
my type of research supported and attended to, I have often been
frustrated by demands that ethnography be made to fit a
philosophical context that in many ways contradicts its very nature.
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
HERMENEUTICS IN ANTHROPOLOGY 271
At the same time, I found myself in the embarrassing position of
having no clear statement to articulate as an alternative position. I
think that hermeneutics will contribute to the development of just
such a statement. For its rhetorical uses in such applied interdisciplinary
settings alone I intend to study it further.
However, it is more than just a rhetorical advice. My own experience
of working on this review discussion has led me to think
through some ethnographic issues in new ways, and to see some
issues as related that previously I had seen as distinct. The discussion
also shows, I hope, that hermeneutics is also full of problems, both
as a separate field and especially in its attempted connection with
the study of human life. Hermeneutics is not a source of solutions,
but I think it will offer a clearer sense of the problems.
REFERENCES
AGAR, MICHAEL. 1980a. Ethnography as an Interdisciplinary Cam
Cognitive, Social and Environmental Psychology (John Harvey, ed
NJ: Lawrence Erlbaum.
1980b. The Professinal Stranger. New York: Academic Pr
DUMONT, JEAN-PAUL. 1978. The Headman and I. Austin: Un
Texas Press.
GARFINKEL, HAROLD. 1967. Studies in Ethnomethodology. Englew
Cliffs, NJ: Prentice-Hall.
GOFFMAN, ERVING. 1964. The Neglected Situation. American Anthropo
gist 66(3, part 2):122.
GUTKIND, P. C. W., and G. SANKOFF. 1967. Annotated Bibliograph
Anthropological Field Work Methods, Anthropologists in the Field (D
Jougmans and P. C. W. Gutkind, eds.). New York: Humanities Press.
HILL-BURNETT, JACQUETTA. 1974. On the Analog between Cultu
Acquisition and Ethnographic Method. Council on Anthropology and Educa
Quarterly 5:25-29.
NADEL, S. F. 1939. The Interview Technique in Social Anthropology
Study of Society (F. C. Bartlett et al., eds.). London: Routledge and Kegan P
OPLER, MORRIS. 1959. Component, Assemblage and Theme in Cult
Integration and Differentiation. American Anthropologist 61:273-281.
PALMER, RICHARD E. 1969. Hermeneutics. Evanston, IL: Northwestern
University Press.
RABINOW, PAUL. 1977. Reflections on Fieldwork in Morocco. Berkeley:
University of California Press.
RABINOW, PAUL, and WILLIAM M. SULLIVAN. 1979. Interpretive Social
Science: A Reader. Berkeley: University of California Press.
SETTLE, MARY LEE. 1977. Blood Tie. New York: Ballantine Books.
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms
272 ETHOS
SPRADLEY, JAMES. 1979. The Ethnographic Interview. New
Rinehart & Winston.
TURNER, VICTOR W. 1957. Schism and Continuity in an African Society.
Manchester, England: Manchester University Press.
TYLER, STEPHEN R. 1978. The Said and the Unsaid: Mind, Meaning and
Culture. New York: Academic Press.
This content downloaded from
162.255.45.238 on Tue, 13 Sep 2022 09:47:05 UTC
All use subject to https://about.jstor.org/terms