Vous êtes sur la page 1sur 26

Hutchins Center for African and African American Research at Harvard University

Author(s): Lucien Taylor
Source: Transition, No. 69 (1996), pp. 64-88
Published by: Indiana University Press on behalf of the Hutchins Center for African and
African American Research at Harvard University
Stable URL: https://www.jstor.org/stable/2935240
Accessed: 18-09-2019 19:45 UTC

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

Indiana University Press, Hutchins Center for African and African American
Research at Harvard University are collaborating with JSTOR to digitize, preserve and
extend access to Transition

This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
( Position

How anthropology lost it at the movies

Lucien Taylor

lens for our two human eyes, it is impe-

Ethnographers considerfilm to be like a book,
rious and monocular; its beauty is dis-
and a book on ethnology appears no different
from an ordinary book. torting; it tries to simplify and disarm, as
-Jean Rouch well as to impose. By implication, text,
and anthropological text in particular, is
The least advanced of men can convey infor-
none of these things-neither imperi-
mation, that is, they can write by means of
ous or monocular, nor simplifying, dis-
arming, or imposing. Thus, anthropolo-
-Alfred C. Haddon
gists search for complex connections
between disparate particularities, while
In I977, the acclaimed Africanist an-
filmmakers, rather like development
thropologist P. T. W. Baxter reviewed a
planners-the preeminent put-down in
film about an East African people called
Africanist anthropology of the period-
the Rendille for the British Royal An-
suppose that life is simple, and the issues
thropological Institute Newsletter (RAIN).
clear. Baxter "resent[s]" films; he is "re-
It was a film for which he felt instinc-
luctant to submit" to them.
tive ambivalence, and he set out to say
A decade later (I988), in terms a little
just why it was that he was so suspicious
less bellicose but equally ardent, the em-
of ethnographic films. He decided that
inent Marxist anthropologist and distin-
anthropology and film are fundamentally
guished theorist of ritual Maurice Bloch
incompatible, distinct in "aims" and meth-
echoed this distrust. In RAINs succes-
ods. Each, said Baxter, "seeks quite dif-
From The Nuer sor, Anthropology Today, he tells us that
ferent aspects of truth and utilises differ-
(1970), dir. Robert not only is he "not very interested" in
Gardner, Hilary ent means of stitching scraps of culture
Harris, and George ethnographic films, he can "hardly bear
together creatively." Whereas anthropol-
Breidenbach. to watch them" at all. Contemporary
© Film Study Center, ogy is open-minded and detached, film
Harvard University
ethnographic filmmakers, he says, imag-
is anything but. Substituting a single glass


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
f :fi

This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
ine they can learn something about peo- other." Watching Mursi spit at each other
and theorist
ple simply by "star[ing]" at them and lis- provides him with precious litde context
Trinh T. Minh-ha
tening to their words out of "context." for their "words," or, presumably, for the
Moreover, whereas writing anthropolo- rest of their actions. For Bloch, then,
gists are beginning to consider how "context" is not something outside a text,
ethnographies are "constructed," ethno- something that puts a text in its place.
graphic filmmakers are becoming ever Textuality itself, and textuality alone (a
more naive about the nature of repre- "thesis"), is the condition of possibility
sentation. If, he says, ethnographic films of a legitimate ("discussive, intellectual")
must be made at all, they should be made visual anthropology.Visuality itself be-
with a "thesis," but without any anthro- comes merely ancillary, illustrative rath-
pological collaboration. "I think there is er than constitutive of anthropological
great scope for anthropology on televi- knowledge.
sion," says Bloch, "but for a discussive in- Lest these two cases seem isolated,
tellectual form of anthropology; I want anecdotal, and outdated, a similar fear of
less staring at Mursi spitting at each the filmic, indeed a suspicion of the visual

This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
tout court, is evident in the leading article way of "maps" that "totalise" observa-
of a I992 collection of essays entitled, tions, whereas ethnographic texts offer
oddly enough, Film As Ethnography. In guided tours through a "discursive series
the keynote paper, "Anthropological vi- of oppositions." Filmmakers thus commit
sions," Kirsten Hastrup-Scandinavian the "sin" of separating words from things,
anthropologist and leading authority on a sin for which postmodern textual an-
(in a word) "experience"-sets out to thropologists (presumably including Has-
combat what she perceives as "a burst of trup herself) atone by returning their
interest in visual anthropology" that reca- readers to the hearth and home of lived

pitulates an obsolete anthropological dis- experience. Texts can move freely be-
course. Hastrup rehearses Baxter's and tween the past, present, and future, im-
Bloch's apprehensions, adding to them a plying "meanwhileness" and "conjunc-
whole series of oppositions between films tion," but the "knowledge" contained in
and texts, as ideal types. Film, she says, is ethnographic films is irreducibly icono-
capable of producing no more than a thin graphic. Ethnographic writing alone can
description of a "happening." Text, on the be reflexive, and thereby transform know-
other hard, can articulate a thick de- ledge into consciousness. In the same
scription of an "event," a happening in-vein, only anthropologists (not filmmak-
v sted with cultural significance. The idea ers) have admitted that the person of the
would seem to be that a happening is an ethnographer is part of the plot. In sum,
objective occurrence, represented indif- there is no conflict between ethnographic
ferently, while an event is an incident films and anthropological texts. Not be-
witnessed firsthand, invested with first- cause they are complementary, but be-
person subjectivity. A happening is some- cause films are, quite simply, logically
thing viewed from afar, dispassionately, inferior.
more or less from nowhere, while an
event is narrated perspectivally-that is,
from the point of view of a human par- How you respond to such generalized
ticipant, evoking that participant's per-iconophobia depends, surely, on where
sonal experience. Film, Hastrup goes onyou come from, and on whether you are
to say, consists of no more than concrete a writer or a filmmaker. To me the most
images of what-once-was, while text striking quality of these examples is the
transcends the particular and conveys aextraordinary anxiety the academic au-
more comprehensive truth, the truth of thors evince toward images, especially
the "ethnographic present." Although a film images. The filmic detachment of
picture of a happening can, at a later date, words and things (if indeed that is what
invoke the memory of its "space" for a films do) is characterized in a quasi-reli-
firsthand participant, only writing cangious idiom as sinful. The fear that films
evoke the existential texture of the will somehow destroy or discredit their
"place" to someone who wasn't there. anthropological makers and viewers-as
Borrowing from the French thinkerBloch puts it, "when anthropologists be-
gin to dedicate a large part of their time
Michel de Certeau, Hastrup argues that
to ethnographic films it is usually be-
ethnographic films represent reality by


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
cause they have lost confidence in their the process of editing, film disincarnates
own ideas"-is surely part and parcel of this transcendent subject's glass eye and
an abhorrence of imagery in general, a lets it roam pretty much wherever it will.
sentiment that, together with an array of Far from the ensuing shots, with their
attendant anti-iconic prohibitions, has different angles and focal lengths, pro-
existed from time immemorial. The fear ducing a multitude of conflicting and
of icons and graven imagery, profound inembodied perspectives, these potentially
the monotheisms ofJudaism and Islam,diverse subjectivities are collapsed into
is neither novel nor restricted to anthro- an artificially harmonious and singular
pologists. But what about this apprehen- subject through a process of ideological
sion is peculiarly anthropological? suture. By means of editorial conven-
tions that simulate space-time continu-
Even in 1970s Paris, film-going ity, the spectators are obliged to identify
with the superhuman gazes of the ap-
was hardly the antisocial experience
parently unified subjects on the screen.
apparatus theorists imagined Under the influence of Lacanian psy-
choanalysis, apparatus theorists likened
spectators to young children. Films, the
One of the more interesting attributes
argument went, force spectators to "mis-
of this anthropological aversion is its re-
capitulation, apparently unawares, ofrecognize"
a their specular identity in
large body of critical work-most of it
much the same way that children do
during the "mirror" stage. The experi-
French and published in the late sixties
and early seventies, in magazines likeence of children and filmgoers alike is,
as Metz memorably suggested, one of
Cahiers du Cine'm, Cinethique, and Tel
Quel-that sought to elaborate, and then
"under-motricity" and "over-perception."
supersede, a semiology of cinema. Much Both are characterized by a hypertrophy
of this work coalesced under the rubric of the visual. Stuck in their seats, in a
of "apparatus" theory. To oversimplify,dark and antisocial cinema, spectators
film critics like Jean-Louis Comolli,Jeancannot help but renounce all voluntary
Narboni,Jean-Andre Fieschi, Jean-Louiscontrol, regress into an infantile, dream-
Baudry, Christian Metz, Jean-Pierrelike state, and give themselves up to the
Oudart, and Marcelin Pleynet argued spectacle unfolding before their eyes on
that film is an ideological instrument the two-dimensional screen-a condi-
that is as coded as any other symbol sys- tion in which they identify primarily
tem, that it has inherited the scientific with the Archimedean camera eye, but
perspective of the Quattrocento, and thatalso, if to a lesser degree, with the char-
its vision is monocular, ideal, and tran- acters up on the screen. The severance of
scendent-and by implication, omni- the subject of desire (the viewing self)
present, omniscient, and omnipotent. As from the object of that desire (the screen
Christian Metz would later describe it, or screen subject) transforms the film-
this is a seeing "which has no features or goer, no less than the infant, from a par-
position, as vicarious as the narrator- ticipant into a voyeur.
God or the spectator-God." Through The iconophobia of Baxter, Bloch,


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
W From Moana,
: dir. Robert Flaherty.
: Courtesy of the
Museum of Modern
Art/Film Stills
:--, Division


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
From Forest of Bliss

(1985), dir. Robert

Photo: NedJohnston.
C Film Study Center,
Harvard University


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
and Hastrup has many affinities with ap- of its discursive basis. As Bloch implies,
paratus theory, at least in its broader filmmakers seem not to recognize that
strokes. For Baxter, film has none of the their works are "constructed." Ethno-

virtues of text. It is not tentative, de- graphic filmmakers, says Hastrup, deny
tached, open-minded, or uncertain; on -or worse, don't even realize that
the contrary, it is bossy, one-eyed, dis- they are part of the plot.
tortingly beautiful, simplifying, and dis-
arming. Film imposes itself "through
the temporary suspension of disbelief," Baxter, Bloch, and Hastrup are an appa-
which would seem to be what Metz ratus theorist's dream come true. How-

means by spectatorial "disavowal":ever,

film-as one detractor has since gibed, film

goers are of course (at least since thedoes not mystify all of its spectators with
a "delirium
screening of L'arrive d'un train en gare in of clinical perfection." For
Paris in I895) fully aware that they are are by no means-or, rather,
watching a representation of reality onall due respect for the anthropolo-
the screen, rather than reality itself, andnot always-the wretched little
yet they are obliged to pretend that it is that apparatus theorists imag-
ined, alienated from their true selves,
reality if it is to have the desired effect.
In a realistic film, there is a complicity of captured or captivated" before
an almighty
disavowal between the filmgoer and the screen. Even in classical nar-
filmmaker, a refusal of reciprocity be-cinema, and certainly in ethno-
tween viewer and viewed. Metz has ex- graphic and documentary film, the dis-
pressed this in an intentionalist idiom, cursive underpinning, or authorial voice,
is not uniformly disavowed. Above all,
"The film is not exhibitionist. I watch it,
but it doesn't watch me watching it. cinematic production and reception is
Nevertheless, it knows that I am watch-not some transhistorical, transcultural
given. Spectatorship is a "total social fact"
ing it. But it doesn't want to know. This
if anything is, embedded in a cultural
fundamental disavowal is what has guided
the whole of classical cinema into the context and historical moment, and thus
susceptible to sociological as well as psy-
paths of'story' relentlessly erasing its dis-
cursive basis, and making it (at best)chological
a interpretation. Even in 1970s
beautiful close object." Baxter is as Paris,
re- filmgoing was hardly the antisocial
luctant as Metz to "submit" to such a experience apparatus theorists imagined.
It's not only in Jamaica that spectators
filmic regime. Whereas a filmgoer is im-
sometimes shoot at the characters on the
prisoned in the temporal order of the
Boatman with
film, a book-reader has the freedom to
screen. After all, film is not a purely visual
dead child, medium. It has always-but especially
pause or stop, as well as to flip back and
Benares, India
since the advent of talkies in the I920S
forth through the pages. Freethinking
From Forest of Bliss
and I930s-involved a complex inter-
and freewheeling adult that he is, Bax-
(1985), dir Robert
Gardner. ter resents "not being able to pause,play
to of picture and sound. Sounds, im
Photo Christopher turn back, to recheck and to compare ages, and words gush around (and into)
( Film Study Center,
each other continually. Indeed, many
statements and pieces of data." Bloch and
Harvard University ethnographic films accord a particularly
Hastrup also underline film's disavowal


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
I.A-- ^ .
. li5 .z,_s _aB B
I - - - ---------


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
:i. :!i

© Napoleon Chagnon
From Yanamamo:

The Last Days of

Eden series.


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
elevated place to dialogue. The picture as of this indexical "excess"- one that is la-

a whole is transformed by the simultane- tent within shots as much as it is gener-

ous sound track, which is in turn modi- ated by their juxtaposition-one could
fied by the adjacent picture. Film is a sen- argue that film is not more, but less bossy,
sory medium, nearly as much as the one-eyed, distorting, simplifying, dis-
human subject is a sensory being, and it arming, imposing, and so on, than text.
is more often than not made up of both Indeed, one could make the further ar-
images and words. As W. J. T. Mitchell gument that since an observational aes-
has so eloquently argued, language and thetic has for some time now enjoyed
imagery continually contaminate one pride of place among ethnographic film-
another. makers-an aesthetic that favors long
As spectators, Baxter, Bloch, and Has- takes, synchronous speech, and a tempo
trup all seem equally insensible to the
properties of the medium, especially to Film can no more be transposed
the relationship of documentaries to the
wholesale into text than poetry can be
real (or, as film critics like to say, "pro-
filmic") world of which they try to pro- transposed into prose
vide a record. (Realist fiction films, of
course, absolutely elide the pro-filmic in faithful to the rhythms of real life, and
telling their stories.) They seem also, that discourages cutting, directing, reen-
more generally, to be unaware of how acting, interviewing-their films are
films are fabricated. One could, in fact, unusually open to multiple interpreta-
just as plausibly make a case for the very tions. In particular, the aesthetic of long
opposite of all their main propositions. takes is more realistic than the "psycho-
Let's, for a moment, do that. In semioti- logical montage" of continuity cutting,
cian C. S. Peirce's terms, anthropological which fragments events in such a way as
prose, like any other, is a succession of to simulate the shifts in our attention if
pure symbols. It is arbitrary and artificial, we were present, because (so the neore-
completely conventional. Film, by con- alist argument goes) it does not suppose
trast, consists not only of symbols and of that events have a singular meaning and
icons (since it resembles what it refers to dictate the attention of viewers accord-
in some way or another), but also of a se- ingly. On the contrary, this "technical re-
ries of indices. For it has a "motivated" alism," as Andre Bazin put it, restores to
or materially causative relationship withthe viewers some of the autonomy they
what it refers to; as semioticians would have in interpreting reality when they
say, there is a natural bond between theare confronted with it as witnesses in
signifier and signiignified. Film is photo- real life. It allows action to develop with-
chemically permeated by the world, andin a single shot, over an extended pe-
analog video electrically infused with it. riod, and on several spatial planes; it
The indexicality of ethnographic filmconstructs relationships within frames as
makes it open-ended, and thus suscepti-much as between them; and it honors
ble to differing interpretations in a waythe homogeneity of space by preserving
anthropological writing is not. In viewthe relationships between objects rather


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
© Napoleon Chagnon.than substituting the abstract time and odds with the maker's. It is not exactly
From Yanamamo:
The Last Days of synthetic space of montage. Long takes, that observational films permit "aber-
Eden series.
by exhibiting a deficiency of authorial rant" or "alternative" readings, for there
intelligence (for which they have been may be no correct, dominant or intended
taken to task by nearly everyone since writing to which they may be counter-
Sergei Eisenstein), reflect an ambiguity posed: the metaphor of reading/writing,
of meaning that is at the heart of human with its connotations of scientific rhet-
experience itself. oric and decipherment, is inappropriate.
An observational aesthetic, then, does But certainly observational films are open
not relinquish authorial control entirely, in the sense and to the extent that they
but it does so differently from other doc- permit multiple viewings.
umentary forms. Observational films are Baxter is assuredly right that film will
still authored, but less authoritatively. not let him pause or go back, as he might
They are still reductive, but watching with a text. In certain respects it is a very
observational films is a more digressive domineering medium indeed. Unlike
experience than watching other docu- still images and text, the temporal order
mentaries. In these regards they empower of projected film precludes what Peter
the film's subjects and the spectators Wollen has called a "free rewriting time."
alike: the subjects are less mutilated by (Video, as well as film on an editing
the montage, and the spectators may gar- table, are different matters.) However, as
ner meanings or simply come away with a spectator, Baxter is at liberty to take
sensations and impressions that are at from the images meanings that were

This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
never attached to them, perhaps never ofjuridical discourse here intimates that
even imagined by the filmmaker, to a far film is on trial. But why? Why does it
greater degree than he is with the lines pose such a threat?
of an ethnographic monograph. Thus Hastrup's inattention to the experien-
Baxter's problem with film may not in tial realism of movies is all the more un-

fact be that it is too bossy, but that it usual in that her writings argue that the
isn't nearly bossy enough. It doesn't give vocation of contemporary anthropology
him the answer; it demands too much is to restore us to the sensuous flow of

work from the viewer. (Reality doesn't what phenomenologists like to call the
give up the answer either, and it also can flesh-of-the-world-a calling that seems
be obtuse and intransigent.) This is ex- almost inherently cinematic. Although
actly what Bloch resents about the me- she claims film is bound to present only
dium too; it doesn't hand him his thesis "real-time" sequences, one of the me-
on a plate. If he wants more "discussive, dium's signal features is, in fact, its abil-
intellectual" films and less staring at spit- ity to manipulate time and space. Ob-
ting Mursi, it may be that he doesn't servational and verite films, in particular,
want a film at all, or rather that he sim- offer embodied "itineraries" through
ply doesn't want to go to the bother of space, and tell perspectival stories, in a
looking. way that academic monographs rarely
At the very least he doesn't want films do. Storytelling has been at the heart of
that require him to be engaged in ac- cinema since its inception, of course, but
tively generating meaning out of the it is only recently that ethnographic
scenes that pass before his eyes and ears monographs have tried to move beyond
-a form of engagement closer to the abstract, synchronic, and synthetic classi-
experience of an onlooker at the event fication-beyond, in Hastrup's terms,
than to a reader of an ethnographic maps. She claims the "more compre-
monograph. Discussive, intellectual films, hensive truth of the ethnographic pres-
in Bloch's book, are those that are pre-
textualized, that elaborate a thesis, that Filmic ethnography, whether about Mursi
have already done his work for him
spitting at each other, an Icelandic ram
films, in short, that mimic anthropolog-
ical prose. No wonder they always fall exhibition, or anything else, requires as much
short! "The idea that ethnographic film
speaks for itself is wrong," he writes.
"local knowledge" as written ethnography

Most anthropological writers would

agree. Peter Loizos, a scholar of Greek ent" is the exclusive preserve of writing,
gender relations, is one of the few to ap- on Hastrup's account. Yet the ethno-
preciate the divergent capacities of films graphic present has never been in greater
and texts. But even he, in the final analy- disrepute, as a comprehensive and mys-
sis, wants to insist that "as anthropologists tifying totalization that removes one's
we most fruitfully admit films in evi- subjects from the entanglements of his-
dence when we can relate them ... to tory-indeed, asJohannes Fabian would
sources outside the film itself." The hint say, takes them out of time altogether.


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
:*: | - ' :L%:i:

_U)s^U)tBr s-
o ,C s.x.

m r w .
_0). a .
(0 ~ ~ ~ aWS. .:V

This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
Little wonder, then, that Hastrup sug- that makes it so expressive of lived-body © Napoleon Chagnon.
From Yanamamo'
gests that texts alone can convey the experience. In addition, documentary The Last Days of
Eden series.
"timelessness" that is part of human ex- films that foreground the active engage-
perience when most of us would have ment between filmmaker and filmed in

thought that experience was distin- the production of cinematic meaning

guished precisely by its timeliness-its predate by half a century the current
concrete and contingent coordinates in vogue for reflexivity among ethno-
time and space. Cinema, of all the me- graphic writers. Moreover, quite apart
dia of human expression, has long been from any self-conscious baring of the
praised for its ability to simulate a world device on the part of the filmmaker, the
of living flux, what Andre Bazin fa- indexicality of the medium, and partic-
mously called "objectivity in time." The ularly its use of experience, make it in-
cumulation of successive film frames herently reflexive-that is, at once sub-
evokes the sensation of movement overject and object to itself-in a way that
time quite literally through movementhas no precise parallel in other media or
over time, and captures the experiencearts.
of animate presence in a way that neither Hence Hastrup's presumption that
photographs nor text can. As Metz putfilms alone separate words and things,
it, "Film gives back to the dead a sem- and that only postmodern ethnographic
blance of life." By contrast, one could ar-texts may recover the originary vitality
gue as credibly as Hastrup's claim to the of prereflective existence-that only
contrary that the prosaic text of the an-writers may disclose something of what it
thropologists, and not least the denial ofis feels like, in any particular local setting,
"coevality" between observer and ob- to be-in-the-world-is eccentric in the
served, has clear affinities, not only withextreme. As it happens, only since the in-
timelessness, but also with lifelessness. troduction of magnetic sound stock in
Hastrup reserves another property forthe late I95os have documentary editors
texts: the capacity to transcend "the in-been able to afford to separate sounds
stance of fieldwork." But the moment an from pictures, and so words from their
editor makes (or imagines) the first splice, speakers (Hastrup's "things"). But with
a film has already embarked on the slip- texts, as with noniconic symbols in gen-
pery road to abstraction, synthesis, and eral, taking words away from their utter-
transcendence. If finished films still bear ers is absolutely free, a penstroke or
scars of the encounters that produced touch of the keyboard away. It is as old
them-indexical "stigmata" of their his- as (written) history itself! What is the dis-
tories-might that be a virtue, and not tinguishing hallmark of literacy if it is
the vice she takes it to be? Hastrup be- not its radical disjunction of the utter-
lieves that texts, and texts alone, can cap- ance (the enonce) from the moment of
ture the existential space of cultural ex- utterance (the enonciation)? A sin, if you
perience, and yet it is the motivated, like, but hardly one that postmodern
existential, "real relation," as Peirce put ethnographic texts can manage to atone
it, between the cinematic signifier and for.
signified, the filmic and the pro-filmic, But this is all academic. For film is es-


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
sentially a sensory medium, fusing "words tures, excellent at arousing empathetic
and things," in a way that writing, or at identification with an exotic people or
least expository academic writing, is not. an alien way of life-useful for popular-
As film theoristVivian Sobchack has re- izing anthropological knowledge with
cently reminded us in The Address of thethe help of some well-chosen voice-
Eye, "More than any other medium ofover, but little more. "Ethnographic film,"
human communication, the moving writes Asen Balikci, "is characteristically
picture makes itself sensuously and sen- descriptive to the point of largely ex-
sibly manifest as the expression of expe- cluding analysis.... film is not an appro-
rience by experience." Film, unlike any priate medium for sophisticated analy-
other art form, thus depends upon ex-sis." Timothy Asch, the filmmaker of the
perience twice over: as form and con- canonicalYanomamo series, hoped against
tent, discourse and representation, sub-hope that one day anthropologists would
ject and object-in short, as signifier stop conceiving of ethnographic film as
and signified. Acts of moving, hearing, "entertainment" and start thinking of it
and seeing are at once presented and as (guess what?) "data." And ethnogra-
represented as the originary structures of phic "hypermedia" expert Peter Biella
embodied existence and the mediating has argued that the "observational style
structures of discourse. It is the double ... cannot present theory." (This, despite
duty, as Sobchack calls it, that experiencethe fact that the etymology of "theory"
performs in the cinema that would seemis "to look" or "to gaze," and that if there's
to make the medium so fit for exploringone thing that observational filmmakers
existence in all its ambiguity, fit for ex-do, it's that.) Comments like these are a
pressing the undifferentiated significance dime a dozen in almost every issue of the
of the human condition; fit, that is, for si-various international journals devoted to
multaneously embodying and evoking ethnographic filmmaking.
the intuitive lived experience of what There are others, though, who see
Husserl and later Heidegger would call continuities between ethnographic films
the Lebenswelt, the lifeworld. and anthropological monographs, who
feel that ethnographic films should not
so much illustrate as actually embody
If anthropological writers, naturally anthropological knowledge. This posi-
enough, have only their own best inter-tion is often traced to SolWorth, collab-
ests at heart in their depreciation of film,orator withJohn Adair on the celebrated
what do ethnographic filmmakers and "Navajo Project." (Wondering whether
specifically visual anthropologists have tothe Navajo might have a "film gram-
say on the subject? mar" of their own, one related to their
Surprisingly, many ethnographic film- language and worldview, they handed
makers seem to accept the aspersions out i6 mm triple-turret Bell and How-
cast on their trade. They concede with-ells to neophyte Navajo filmmakers to
out protest that ethnographic films aresee what they would do with them. Not
marginal to the evolution of anthropo-a lot, they found out, unless the filming
logical knowledge. Films are pretty pic-could be shown to be beneficial to their


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
© Napoleon Chagnon.
From Yanamamo:

The Last Days of

Eden series.


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
sheep.) In I976,Worth proposed an "an- formed by an explicit or implicit "the-
thropology of visual communication," ory" of culture, that they be articulated
urging scholars to explore how anthro- within an "anthropological argot," that
pological knowledge can be inscribed in they contain explicit reflexive "state-
film and other iconic symbol-systems. ments" revealing the author's "method-
Implicitly arguing against naive realist ology," and that they thus furnish a "sci-
theories of visual representation, Worth entific" "justification" for every selection
insisted that ethnographic film offers not made, including the framing and length
a "copy" or a "magic mirror" of the world of every shot, the film stock, the lens, the
"out there" but "someone's statement about "type" of sound, and all editing deci-
the world." Worth wanted to take us be- sions. But for all his insistence on hard-

yond "our deeply held and largely unex- core ethnography and an authentic "an-
amined notion that ... motion pictures thropological argot," Ruby never pauses
are a mirror of the people, objects, and to provide a definition of either. Eth-
events that these media record photo- nography is invoked, almost fetishisti-
chemically," just as he wished to ques- cally, as a magical elixir of anthropolog-
tion "the jump we make when we say ical truth.
that the resultant photographic image Of course, even for written texts, many
could be, should be, and most often is of Ruby's prescriptions would be a tall
something called'real,' 'reality,' or'truth."' order. Anthropological monographs no
In the hands of well-trained anthropol- more provide a "scientific justification"
ogists, film may be not only a record of for the "multitude of decisions" involved
culture (which it invariably is) but also in their production than do ethnographic
an analytic record about culture. Once films-be it shot length or word choice,
one allows the possibility that film could sequence or sentence, film form or liter-
make a scientific statement about the ary style, "type of field sound" or choice
world, we can step outside the seductive
of informants. Rather than introducing a
myth system that would have us believe
critical distance by foregrounding a text's
constructedness, Ruby's "reflexivity" is
that it is only a crass copy of it. This frees
supposed to produce an absolute trans-
us, says Worth, from "the impossible po-
parency, a state of complete self-con-
sition of asking whether [they] are true."
sciousness-a state that is logically im-
And some of us, he says, "are arguing that
it is as silly to ask whether a film is true
possible in the human sciences just as it
or false as it is to ask whether a gram-
is in the arts.While it is, of course, possi-
mar is true or false. Or whether a per-
ble for me to dream up a reflexive hall of
formance of a Bach sonata or a Beatles mirrors in which I could represent (my
song is true or false." representing) myself representing my
original representation, there is no Archi-
A year earlier, in 1975, fellow anthro-
pologist of visual communication Jay
medean Prime Representer at the end (or
Ruby enumerated a series of specifica-
beginning) of the line. What the Welsh
writer and documentary editor Dai
tions for ethnographic films-that they
"describe" a "whole culture" or a "de- Vaughan says about film is true of repre-
fineable" unit thereof, that they be sentation
in- in general: "Events must be con-


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
trived for the camera; and to make the would somehow liberate us from asking
audience aware of the contrivance is to whether it's true or false, when precisely
the opposite is the case. It is not the
fall into the absurdity of an endless re-
gression.... Once we have accepted that
world itself, or even our experience in it,
that has a truth-value, but rather our
there is no purely technical criterion for
realism-no gimmick of presentation representation of it. If film were noth-
which can guarantee authenticity-then ing but a magical mirror held up to the
we are forced to recognize that we must world, we would not have to ask if it
rely upon the integrity of the artist for its were true at all. It is because it aspires to
creation and upon the judgement of the be, or cannot resist being, discourse that
viewer for its proof." Is it any different for we are still obliged to ask such ques-
written ethnography? As Vaughan sug- tions of it. It is no coincidence that
gests elsewhere, "Wilfully or by oversight, films are criticized as biased or subjec-
some materials may be wrongly labelled. tive far more frequently than photo-
Some things may have been less rehearsed graphs.
or more rehearsed, less spontaneous, less Worth compared textuality with his
calculated, less uninfluenced by the cam- proposed mode of anthropology, which
era's presence than we-as-viewers suppose he called pictorial-visual. But rather than
them to have been. But there is no sharp expanding anthropology to include the
demarcation between the misunder- distinct properties of this "pictorial-vi-
standings of documentary and thesual,"
mis- he smothered it with metaphors of
understandings of life." prosaic textuality. In his attempt to rela-
Ruby's hope that films and texts tivize the role of language, he in fact en-
might one day be virtually identical can shrined language as paradigmatic for
only be maintained by downplaying meaning by reducing anthropological
what distinguishes them. Although he films to "statements about" and "records
sees himself as making a case for a truly of." But film can no more be transposed
filmic ethnography, his terminology (de- wholesale into text than poetry can be
scriptions, definitions, methodologies, transposed into prose. His problem, in
statements, and justifications) reveals that short, is the problem at the core of semi-
in his conception, visuality is entirely otics, for the paradigm of semiotics has
absorbed by the "logos" of anthropol- always been linguistics.
ogy-by, that is, Margaret Mead's "disci- Worth was half right to distinguish
pline of words." between a record "about" and a record

Ruby's domestication of the visual "of" culture, even if the distinction could
and Worth's proposed shift from a visual be articulated more accurately as one
anthropology to an anthropology of vi- between discourse about and record of. But
sual communication go hand in hand. Worth failed to recognize that every film
The degree to which Worth sought, de- is by definition both of these things at
spite himself, to linguify film is quite re- once: it is not that film is not linguistic
markable. He claimed that conceiving of at all, nor even that it is a language un-
film as a statement about rather than a like any other, but that it is, incongru-
copy of or "magic mirror" to the worldously and oxymoronically, at once both


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
lls; 0,t¢''Wt .................. .^.l . ....

:w ;' < 'V ., *' F^ r

-,ps *:, 7 0 T'"
' v . . --g
v E .

- :. , r

^'.ly ^

>*- ' t W -^ g , -* '" s.

,C '«/ * ', :: j'" y -

, i 9'W '; E j. , wi.
S~,._~~ ~i~r , F0w v
^ t W0^ v 0s0 tv
',8; F-fE~~~~~~~~~~~~~~~~~~~~~~~,
Xfv rm' *:s,,<
a r > ^:" '<. ::' -., : ^»»e
r , , '-^'^r,
I':<i.- iX i * ' r 1 **. 5

This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
From Ika Hands
a language and not a language. If film its overall meaning. A dense image is also
(1981), dir.
does not provide a mimetic copy of the "continuous"; the various features that Robert Gardner.

world, it does very definitely throw up a make up the whole defy reduction into (© Film Study Center,
Harvard University
"magical mirror" to it. It is anything but isolated, unique characters, each with its
reducible to someone's statement about own singular referent. In this regard, pic-
tures are dense in a way that texts are
it. As Roland Barthes put it in Camera
not. Film, of course, as an ongoing fis-
Lucida (1981), quite possibly with Worth
in mind: sion-fusion of words, sounds, and mov-
ing pictures, all flowing into and through
one another, is both dense and differen-
It is the fashion, nowadays, among Photogra-
phy's commentators (sociologists and semiol- tiated, continuous and discontinuous, all
ogists), to seize upon a semantic relativity: no the same time.

'reality' (great scornfor the 'realists' who do Goodman's notion of density, in itself,
does not directly address the indexicality
not see that the photograph is always coded).
of film, which is what sets it apart from
... the photograph, they say, is not an analo-
the larger class of icons, nor indeed the
gan of the world; what it represents isfabri-
mobility that distinguishes it from still
cated.... [However] the realists do not take
photos. But at least it does not assume
the photographfor a 'copy' of reality, but for
that language is paradigmatic for mean-
an emanation of past reality: a magic, not an
art. ing, and so does not criticize film for
lacking qualities that are essentially lin-
guistic. As the heydey of structuralism
and semiotics has passed, both Lacan's
Worth described Nelson Goodman's claim that the unconscious is structured
seminal Languages of Art as a catalyst
like for
a language and Levi-Strauss's con-
his own work; a curious claim, for viction
Good- that kinship systems display
grammars as intricate as those of lan-
man's writing is notable for its treatment
of what sets images and texts apart. Hehave been discredited. Few peo-
suggests that nonlinguistic systemsple
nowadays believe that language of-
fer from languages ... primarily through
fers an apposite analogy for culture or
lack of differentiation-indeed through
society. But so long as anthropologists
density (and consequent total absence of to hold that language is para-
articulation) -of the symbol system."
digmatic for anthropology, then a "pic-
While Baxter, Bloch, and Hastruptorial-visual"
asso- mode of anthropology can
ciate this undifferentiated, unarticulated
only come into being by divesting itself
quality with deficiency-.with (anthro-
of its distinguishing features. And if that
pological) absence-Goodman pro- is the case, then why bother?
poses on the contrary that a symbol
system's degree of differentiation is in-
versely proportional to its density. A Because we humans express ourselves
symbol system is dense, its symbols "re-through images as well as through lan-
plete," to the extent that the various guage, and because anthropology consti-
properties of its symbols are important to tutes an exploration of the human con-


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
Filming the constitutes discourse about the world but
dition, it seems needlessly delimiting to
conceive of the form of anthropology it- also (re)presents experience of it? What if
Longole, from
self as exclusively linguistic. If anthro- film does not say but show? What if film
A Wife Among
pology is to create a space for the vi- does not just describe, but depict? What,
Wives (1982),
sual-in this case, film-it must seek then, if it offers not only "thin descrip-
pictured shooting
neither to disavow discontinuities be- tions" but also "thick depictions"?
David and Judith

tween the two media nor to transform If film critics and visual anthropolo-
one into another. For a start, this would gists have had an equally hard time com-
Fieldwork Films
entail a shift from the attempt to convey piling an inventory of the rules and reg-
"anthropological knowledge" on film- ulations of film, it may be because these
the attempt to linguify film-to the idea rules are not half as hard and fast as those
that ethnography can itself be conducted of plain prose, and because they're partly
filmically. Filmic ethnography, whether improvised as filmmakers go along. If, as
about Mursi spitting at each other, an Barthes claimed, one of the connotations
Icelandic ram exhibition, or anything of film, or photography, is that it has a
else, requires as much "local knowledge" capacity to offer a "message without a
as written ethnography. Bloch declared, code," then, try as analysts might, this
"The idea that ethnographic film speaks record or trace of the world will never
for itself is wrong." But what if film wholly submit to semiotic decoding. In
doesn't speak at all? What if film not only other words, if the rules of film resist for-


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
mulation, this may not be because film- tems of signification as f they were lan-
makers are even more unconscious about guages. At a certain point the analogies
break down, and semiotics (and semi-
the form they manipulate than everyday
language speakers are about their syntax.otics-derived communication theory)
It may be that the relative syntactic loses its purchase. What makes film so
poverty of the medium is precisely its se-captivating is that it is something other,
mantic strength, that which allows it toor more, than just language. Indeed,
respond to the diversity and density of
given the apparent affinity of film with
human experience as flexibly as it does.
life itself, moving images evoking mov-
In Jean Mitry's words: ing life, hearing evoking hearing, and
seeing seeing; given the centrality of the
[Cinematic]forms . .. are . . . as varied aslifeworld to anthropology; given the ex-
life itself and, just as one doesn't have theemplary open-endedness of ethnogra-
phy, whose wealth of detail is always sup-
knowledge to regulate life, so too one hasn't
the knowledge to regulate an art of which life to transcend the theoretical
is at once the subject and the object. services to which it may be put; and
Whereas the classical arts sought to signifygiven the attention anthropologists have
devoted lately to representations of the
movement with the immobile, life with the
body and to the embodiment of expe-
inanimate, the cinema must express life with
life itself It takes up there where the othersrience, the backlash against film no less

leave off It thus escapes all their rules as itthan the ongoing desire to linguify it
does all their principles. seem all the more unlikely.
Or do they? Jean-Francois Lyotard,
Of course, at a certain point this be-for one, has argued that the ambiguity
comes mumbo jumbo. If films were in-and opacity of the perceptual medium
will always upset orders of prosaic tex-
deed to forgo all rules, they would soon
be incomprehensible, all noise and no tual representation, with their yearning
signal. And film is usually verbal as wellfor clarity and lucidity. This may be true.
as visual, and as such ethnographic film-It is a curious irony that of the anthro-
makers have to confront thorny prob-
pologists who are so fearful of film, one
lems of verbal and visual representation,(Bloch) is an expert on "ritual" and the
both. Semiotics is not all wrong: films are
other (Hastrup) a specialist in "experi-
constructed sequentially, they narrateence." Bloch has long made a convinc-
ing case for the non-propositional, per-
stories, and so have syntagmatic features;
formative, "illocutionary" quality of
in these and other respects they are in-
ritual, but with the constraint of indi-
deed imbued with at least-paralinguistic
qualities. Films can be studied for their
vidual freedom at its core. Perhaps what
records of and about the world, and in irks him most about film is, paradoxi-
anthropological film reviews that it ex-
cally, the qualities it shares with ritual-
actly what is done. But the foundationalits illocutionary aspects and its temporal
metaphor of semiotics remains language,coercion: it permits no free rewriting
and semiotics continues to derive its time. (What could be more liminal than
sitting silently in a dark cinema, eyes
force from looking at nonlinguistic sys-


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms
transfixed on an illusionistic screen seem- Bad object of film (any more than it is to
ingly teetering between two- and three- make a Bad object out of the written
dimensionality?) And maybe what so word and a Good object of film), but to
galls Hastrup is precisely film's simula- recognize that the textual and the filmic
tion of lived experience. As ethno- are both multiple rather than mono-
graphic filmmaker David MacDougall lithic, and culturally and historically vari-
has remarked, "the truth is that anthro-able in their imbrications rather than
pologists were made anxious by this cin-God-given in their differences. Through
ema which eluded them, which was nei- dialogue and narration, subtitles and in-
ther science nor mere exoticism, but tertitles, end credits and opening cred-
which trespassed upon their dreams andits, film is shot through with language,just
memories of fieldwork." as imagery ineluctably infuses language.
However, contrary to Lyotard and a Anthropological writers seem to have
lot of postmodernist hype, this doesn'tturned their backs on film because they
mean that the discursive is inherently in- begrudge documentary its unique affin-
ferior to the figural, or the textual to theity with the human experience they too
visual. (How dispiriting it would be to take as their (missing) object. Films have
have to resort to writing to make thata way of exceeding theoretical bounds,
case.) For if there is an intrinsic impov-and of showing anthropologists'purchase
erishment to the image in knowledge, ason the lived experience of their subjects
surely there is, then there is equally evi-to be rather more precarious than they
dently an impoverishment to knowledgewould like to believe. In its plenty, film
in the image. Density is diminished bycaptures something of the lyricism of
being articulated, as is differentiation iflived experience that probably attracts
all the pieces are put back together again.many anthropologists in the first place.
Moreover, the phenomenologists' and DaiVaughan has argued that film's plen-
itude "defies its reduction ... into a sim-
neorealists' hopes that film would reunite
viewer and viewed in the sensuous in- ple linear statement approximating the
tersubjective flesh-of-the-world have
condition of prose." Might it be that an-
thropologists resent documentary's re-
clearly been dashed. The answer is not,
semblance-insofar as it may be said to
pace Bloch and Hastrup, to wax lyrical
resemble literary forms at all-not to
about the Good object of the written
word and to hoot and holler about the their own plain prose, but to poetry?


This content downloaded from on Wed, 18 Sep 2019 19:45:28 UTC
All use subject to https://about.jstor.org/terms

Centres d'intérêt liés