The Discourse of Pictures: Iconicity and Film Studies
Author(s): Stephen Prince
Source: Film Quarterly, Vol. 47, No. 1 (Autumn, 1993), pp. 16-28
Published by: University of California Press
Stable URL: http://www.jstor.org/stable/1213106 .
Accessed: 18/02/2014 15:36
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
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 [email protected].
.
University of California Press is collaborating with JSTOR to digitize, preserve and extend access to Film
Quarterly.
http://www.jstor.org
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
Copyright by the University of California Press. Prince, S. (1993). The discourse of pictures - iconicity and film studies.
Film Quarterly, 47(1), 16-28. doi: 10.1525/fq.1993.47.1.04a00040
Stephen Prince
The
Discourse of
Iconicity
I
and
n its analysisof images, film theorysince
the 1970s has been deeply indebtedto structuralistand
Saussurean-derived
linguisticmodels.Indeed,it would
be difficult to overstatethe depth and importanceof
thisrelationship.As RobertStamhasnoted,"Semiotics
in general, and film semiotics in particular,must be
seen.., as local manifestationsof a more widespread
linguistic turn."' To speak, for example, about"reading"a film, as manyfilm analystsnow do, irrespective
of the critical methodology employed to generatethe
reading,is to index and emphasize this lineage. Like
books, films are regarded as texts for reading by
viewers or critics, with the concomitantimplication
thatsuchreadingactivatessimilarprocessesof semiotic
decoding.
But does it? To what extent are linguistic models
appropriatefor an understandingof how images communicate?Film theory since the 1970s has tended to
place great emphasis upon what is regardedas the
arbitrarynatureof the signifier-signifiedrelationship,
that is, upon the purely conventional and symbolic
aspectof signs. Whatthis focus has tendedto displace
is an appreciationof the iconic and mimetic aspect of
certain categories of signs, namely pictorial signs,
those most relevantto an understandingof the cinema.
This stressuponthe arbitrarynatureof semioticcoding
has had enormous consequences for the way film
Film
Pictures
Studies
studies as a discipline has tended to frame questions
about visual meaningand communication.
Our purpose here is to examine some of these
consequencesandto see how well they squarewiththe
observableevidence abouthow viewers perceive and
comprehendcinematic sequences. We will see that
currentfilm theory,tracingits lineage from Saussure,
Althusser,and Lacan,has constructedaccountsof the
ways in which film transmits meaning that are, in
certain importantrespects, counterto the observable
skills, abilities, and reactions of real-worldviewers.
The viewer, as theorized in these accounts, differs
substantiallyfrom his/herreal-worldcounterpart.We
shallemphasizesome of thesediscrepanciesin orderto
suggest a reorientationof theoreticalfocus. In short,a
renewed attention to the iconic, mimetic nature of
pictorialsigns is warranted,so thatour theoriesmight
become more sensitive to the unique, constitutive
featuresof pictorial-as opposedto linguistic-modes
of communication.(It is importantto note that other
currentsexist in contemporaryfilm theory.PierPaolo
Pasolini and PeterWollen, for example, have offered
analyses of cinematic signs that are inflected rather
differentlythanthe Saussurean-inspiredaccountsthat
this essay examines.2More than20 years ago, Wollen
stressed the importance of paying attention to the
iconic aspectof pictorialsigns.BorrowingfromCharles
16
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
Peirce's triadic model of the sign, he argued that in
cinema, iconic and indexical aspects are more powerful thansymbolic, andhe pointedout thatsemiologists
have neglectedthe subjectof iconic signs becausethey
arebiased in favor of conceptionsof signs as arbitrary
and symbolic.)
Let me state very clearly at the outset that this
suggested reorientationis not intendedas a substitute
for the very real and necessary work on the role that
cultureplays in film spectatorshipand interpretation.
Culture-boundattitudesdo indeed inflect the content
of film narratives,along with their stylistic visualization, at the point of productionand, again, throughthe
inferences viewers draw from those narratives.The
position this essay takes is not intendedas a replacementeitherfor the criticalworkof ideologicalanalysis
and interpretation,or for inquiriesinto how viewers'
judgementsaboutthe natureof a social worlddepicted
in film may be shapedby elements of cinematicstructure. But, in the interest of conducting this kind of
interpretivework, it is importantto have a clear sense
of whereculturalvariablesdo anddo not enterinto the
productionof meaningby the cinematic image. Only
whenwe have a good understandingof thoseaspectsof
visual significationwherecultureplays a less determinativerole are we likely to be able to constructa clear
portraitof where it does play a role. Thus, this essay
certainly does not argue against the role of cultural
analysis in film interpretation,but it does attemptto
suggest some reasons for caution when employing
linguistic categoriesandmodels of culturalrelativism
in film analysis.It is not thatthereareno cases in which
they might profitablybe applied,but ratherthatthere
are some significantaspects of visual communication
to which they probablyshould not be applied.
The Linguistic Turn
in Contemporary
Film Studies
To begin, it will help to briefly trace the
linguisticinfluenceon contemporaryfilm theory,after
whichwe mayquestionwhetherit suppliesanadequate
accountof images or even, in certaincrucialrespects,
of language itself. It is well known that Saussure's
account of the sign as having an arbitrary and
unmotivatedconnectionbetweenits structuralcomponents has been directlytakenover by many film theorists. As Philip Rosen has pointed out, "Oneeffect of
theargumentforthebasicconventionalityof cinematic
images was to openthe way for a utilizationof the ideal
of differencein cinematicsignification."3Stressingthe
signifier as a differential constructionenabled film
theoryto emphasizecommunicationas discourse,as a
culture-boundactivity, relative to and differentially
patternedby theuniquesocial worldsof diversegroups
of interactants.Signs, whetherlinguisticor cinematic,
were viewed as culturallyinstantiated:"Sign systems
don't producemeaning outside of the social and culturalcontext from which they have developed."'
This conceptof the sign enabledtheoriststo explicate many aspects of cinematic coding, from discrete
optical devices like dissolves or wipes to more complex structuressuch as shot/reverse-shotcutting,subjective images, and other aspects of point-of-view
editing. Viewing these devices as symbolic codes
permittedtheorists to emphasize the constructionof
cinematicdiscourse,thatis, the deploymentin film of
an elaboratesemiotic system whose address,and effects, could be comprehensible in AlthusserianLacaniantermsas the interpellationof subjects.Using
a symbol system like language, in this view, entails
being positioned-socially, ideologically-in and by
the categories which that system has helped create.
Thus film, like language,could be comprehensibleas
discourse, as the creationof apparentmeaningwhere
only true relations of difference prevail (due to the
arbitrarynatureof the sign andthe consequentneed for
it to receive definitiononly in relationto whatit is not,
i.e., to all othersigns). As AlthusserandLacan(andthe
film theory they inspired) emphasized, these deceptively real constellationsof apparentlyfixed meaning
could be an excellent site for ideology and for imaginary conceptions of the self to take root. Viewed as
discourse, cinema assumed a symbiotic relationship
with ideology, becoming an effective vehicle for its
transmission.The workof film theorybecameincreasingly focused on deciphering the ideology at work
inside the cinema's deceptive and transparentappearance of reality.Thatappearanceof realitywas, furthermore, suspect for having ideological effects (e.g.,
naturalizingthat which is historical or cultural,etc.)
andforcreatingidealandfalse subjectunities.5Viewed
in these terms,film historyis the historyof discourse,
andthe relationbetweenfilm andthe worldis a matter
of representationalconvention.As StephenHeathhas
written,"Thatreality,the match of film and world, is
a matterof representation,andrepresentationis in turn
a matterof discourse.... [I]nthis sense at least, film is
a series of languages,a history of codes."'6
Furthermore,an emphasis upon differentialand
relationalsignifiers has entailed a denial, or, at best a
17
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
suspicion,of reference.If signs arearbitrarilyrelatedto
what they represent,then meaning is at best provisional, at worstillusoryor ideological. Representation
which is non-arbitrarytends to be construedby film
theory in terms of a relationshipof identity between
sign and referent, thus generating a dichotomy of
Semioticrepresentationis either
arbitrariness/identity.
a matterof arbitrarycoding or of identityandtransparence, with the lattercondition being construednegatively in termsof illusionanderror.Identityis regarded
as a deceptively false universal, as a part of what is
sometimescalled the ideology of the visual. "Reading
encounters the text as a relation of difference not
identity."' Taken a step further, this view regards
empirical knowledge about the world as being thoroughly mediated by signs, as itself discourse. "The
empirical . . . is not 'the real' but the productof the
discoursesof the dominantideology."'8
Since the "real"
embodiesthe false universalsof ideology, therelations
among cinematicsignifiers tend to be seen as cultural
(and therefore symbolic), rather than as iconic or
mimetic. The spectator'sunderstandingof the cinema
is thereforeexplainedas a matterof culturalconditioning and learning. "The spectatorchains togetherthe
film's signifierson a culturalgridof intelligibility-an
ensemble of assumptionsand presuppositionsabout
the 'real'-into an accountthatmakesthe film socially
intelligible."'9
Note the assumptionshere. Languageimposes a
system of relationaldistinctionsuponthe world,creating culture from the real. One makes entry into language, or, in Lacanianterms,into the Symbolic order,
learningtheculturallypatterneddistinctionsand,in the
process,being interpellatedas a subject.Film is akinto
languageby virtueof employingrelational,differential
signifiers.Comprehensionof the cinema,then, should
likewise be predicated upon cultural conditioning,
uponthe apprehensionof "culturalgridsof intelligibility."But, as we will see, the spectator'sunderstanding
of cinematic images seems more immediatelyexplicable in terms of mimetic, referentialcoding rather
than via the chains of displaced, arbitrary,and relational meaning in prevailingtheories.In otherwords,
this understandingis morea matterof recognitionthan
translation.We will returnto this point. Moreover,
iconic representationis appropriatelyunderstoodin
termsof degreesof resemblanceratherthanthe all-ornothingtermsof arbitrarinessor identicality.A photograph,forexample,exhibitsa higherdegreeof iconicity
thana line drawing.
Transposedin film theory to units of cinematic
structure,the Saussureanview of the sign as an arbi-
trary and, taken in isolation, meaningless unit has
allowed currenttheoryto constructan extendedanalogy withlanguage.Althoughno oneanylongersearches
for the filmic equivalentof such linguistic featuresas
the period or the comma, the debt to Saussureanlinguistics is apparentin the axioms thatthe connections
betweencinematicrepresentationandthe worldare,in
all importantrespects,a matterof historicalor cultural
coding and convention,thatis, thatfilmic representation is a matterof symbolic ratherthan iconic coding
andthata viewer,ratherthanperceivinga film,"reads"
it.
Problems with
Linguistic Relativity
We shall explore some discrepanciesbetween these axioms andthe observableevidence about
how viewers perceive and process cinematic images.
First, however, it will be helpful to discuss some
reservationsregardingthe alleged arbitrarinessof the
linguistic sign and the uses to which film studies has
put this concept.As Devitt and Sterelnyhave recently
pointedout, emphasizingthe sign as a relationalentity,
defined through relations of difference, presents a
problemfor meaning."1This emphasisfails to specify
how meaningmayariseor even how lexical borrowing
from otherlanguagesystems may occur.Whenspeakers borrowtermsfromanotherlanguage,they areoften
doing so in responseto a perceptionthatsome thingor
conditionexists thatneeds a name,althoughat present
it has none or is insufficientlylabeled. In such a case,
linguistic skills are deployed in response to non- or
extra-linguisticperceptions,a conditionwhich models
of linguisticdeterminismhave a hardtime accounting
for. Furthermore,rejecting reference, as Saussurean
models do, mystifies languageacquisition.
In a Saussureanview, where meaning is defined
throughrelationsof difference(e.g., JonathanCuller's
well-known view that understandingthe color brown
entails graspingthe relationbetween brownand what
it is not, i.e., all othercolors), it is difficult to see how
a child ever learns language. As Devitt and Sterelny
note, "Wewantto say thata child begins by learninga
minimal vocabularyand a few rudimentarysyntactic
rules.The child continuesby extendingtheserules.On
a structuralistpictureof language,we cannotsay this.
Vocabularydoes notremainconstantacrosschangesin
the system. Each time the child changes the system,
everythingchanges.Languagelearningcannotbe represented as a cumulativeprocess.""
18
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
Partof the problemhereis thatto specify meaning,
one at times has to step outside the language system,
and poststructuralmethodologieshave been most unwilling to posit the possibility of doing this. In many
accounts, one cannot get outside representationat all
because it is, by definition, determinativeof human
thought and experience. It is commonly maintained
that"thereis no absolutemomentwithoutsigns, without language,without,in otherwords, a whole host of
mediations between seeing, experience, and knowledge."12 StephenHeathhas noted thatthe referentsof
cinematicimages exist only in discourse,as representationsalreadyconstitutedby historyandculture."The
representeda discourseproducesis grasped,realized,
exists as such in the particulardiscursive process of
representation,and it is this that needs first and foremost to be interrogated."13
This view-that pictorialrepresentationsand,beyond them, patternsof humanand social organization
are the expressions of discursive relations and positions-is a reformulationof the famous Whorf-Sapir
hypothesis which suggested that radically different
languagesystems might organizethe world in unique
ways for theirusers. The Whorf-Sapirhypothesisis a
statementof extreme linguistic relativity,arguingfor
the influence of features of linguistic organization
upon the perceptual habits of a community. It is a
relativisticview becauseit arguesagainstthe possibility of semanticor perceptualuniversals.As languages
vary, so do the realities they construct.The WhorfSapirhypothesishas been a very seductive and influential one for film studies. Bill Nichols has noted the
relevance of this hypothesis for poststructuralfilm
theory. "Post-structuralistwork does not regardlanguage or, by extension, film as the neutralmeans by
which we understandourselves,others,andourworld.
Rather,it draws on versions of the Whorf-Sapirhypothesis, which describes a world constructedin, by,
and throughlanguage."'14
BenjaminLee Whorf's study of Indianlanguages
led him to emphasize how different languages may
segment the world for theircommunityof speakersin
divergent ways.15There are some data to supporta
limited view of this hypothesis.16 Languageundoubtedly does modulateour experience of the world. Yet
thereis little evidence to supportthe extremerelativity
of the Whorfianview (i.e., that language determines
boththoughtandperception)and some clearevidence
to counter it.17 Advocates of linguistic relativity, or
discourse relativity, often supporttheirposition with
reference to color terminology, i.e., to the fact that
differentcultureshave varyingnumbersof color terms
to designatelocally importantattributes,such as snow
for an arcticcommunity.In suggestingthatrepresentational images, like other texts, rely upon culturally
determined codes, Nichols, for example, has employed this notion of linguistic relativity,writingthat
"to not know the perceptualcodes maintainedby a
given cultureis tantamountto being an illiterateinfant
wanderingthrough an unintelligible world. (An examplewould be the utterinabilityof most membersof
non-Eskimoculturesto distinguishthe dozens of differentkindsof snow for whichthe Eskimohas separate
words).""8Analogously, Mas'ud Zavarzadehhas asserted that "we understandcolors not because we
respondto them directlythroughour sensory organs,
but because the responses of our sensory organs are
made meaningful for us by the language. Different
languages make sense of this physical continuumin
"
startlinglydifferentand dissimilarways .... 19
In contrastto the Whorf-Sapirhypothesis, however, a classic crossculturalstudyof colorterminology
found respondentsplotting the same range of basic
colors on a Munsell color chart irrespective of the
labels furnishedby their culture.20Experimentaldata
wereobtainedfromspeakersof 20 languagesandwere
supplementedwith historicalinformationon 78 additional languages. The researchersfound a universal
inventoryof 11 basic color categories across all languages studied and a fixed evolutionary sequence
governing the order in which languages added new
basic color categories. Writing that the allegation of
arbitrarinessin the way languages segment the color
spectrumis "a gross overstatement,"the researchers
concludedthat"thereferentsfor the basic color terms
of all languagesappearto be drawnfroma set of eleven
universalperceptualcategories, and these categories
become encoded in the historyof a given languagein
a partiallyfixed order.Thereappearsto be no evidence
to indicatethatdifferencesin complexityof basic color
lexicons between one language and another reflect
perceptualdifferencesbetween the speakersof those
languages."21 Different languages can make more or
fewer distinctionsaccordingto environmentalor culturalneeds but cannotoverridethe biological basis of
color perception.
The evidencefurnishedby BerlinandKayon color
terminology fails to support the extreme linguistic
relativity hypothesis that has been so influential for
film studies, and suggests instead the importanceof
retaininga conceptof referentialityoutsidelanguageto
which languagemay respond.This example can be a
heuristic one for film theory. Reopening a space for
principlesof referentialityandiconicitycouldbe a very
19
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
useful development, enabling our theories to move
closer to observable real-world evidence about how
viewers understandthe cinema. Thus, with respectto
pictures,film theorymightaskif thereis a nonlinguistic,
even biological, basis on which visual communication
might rest. Investigation of this question may help
provide a more secure sense of the methodological
limits constrainingthe equation of cinema with discourse. JulianHochberghas recentlymade this point,
noting, "only when we know where the line lies that
separates ...
lower, mandatory cognitive processes
Despite these objections,Metz neverthelesswenton to
explorefilm structureasparole, in termsof a taxonomy
of segment types.
If cinematic meaning is to be developed from a
theoreticalbasisoriginatingwitha Saussureanconception of the relativityof the sign, film theorysometimes
seems a little imprecise about the functional and at
times invariantnatureof communicationalrules. As
noted,film theorytendsto view cuttingpatterns,camera positions, even perspectivally based images as
culturally relative yet syntactically precise conventions. But, with respect to language (and perceptual
images), Noam Chomsky points out that "havingacquired the system of language, the person can (in
principle)choose to use it or not, as he can choose to
keep to or disregardhis judgements concerning the
position of objects in space. He cannotchoose to have
sentencesmeanotherthanwhatthey do, anymorethan
he can choose to have objectsdistributedin perceptual
space otherwise than the way they are."25 In other
words, if the assumptionsof the Saussurianview, as
applied to pictures,are correct,we should find more
crossculturalvariation on basic picture recognition
tasks thanwe in fact do. We will explore this point in
more detail shortly.
from the higher, elective functions, can we sensibly
formulate explanations-linguistic, psychoanalytic,
and so on-that rest on culturaldetermination."22
One final point to note about the concepts of
linguistic relativityand arbitrarinessoperativein currentfilm theoryis the extent to which accountsof the
cinema as discourse, analogized with language, constructthe analogybasedon limitedfeaturesratherthan
seeing language in more comprehensiveterms. The
linguist CharlesHockett,for example, seeking to distinguishhumanlanguagefromanimalsystemsof communication, formulated a set of 13 design features
uniquely characterizinghumanlanguage.23 Arbitrariness of the sign was only one of them, andmanyof the
others, such as prevarication,displacement,and abstractness,pointto cleardifferencesbetweenthe comVisual Grammar or
municational capabilities of language and pictures,
suggestingthatanalogieswith languagemaynotbe the
Narrative Context?
best way of explicatingpictorialcommunication.The
design featureof displacement,for example, denotes
Chomsky's observationraises the issue of
the capability of the language-systemto create mesgrammaticality,which should be centralto questions
sages that refer to a time and space outside of the
aboutwhetherfilm structureoperateslike a language.
immediate communicational situation. Pictures, by
As noted,Metz's presentationis somewhatambiguous
contrast,lack tense andotheraspectsof syntaxthatcan
with respectto grammaticality,butthis is an issue that
be used to establishremotetemporalor spatialcondicannot easily be avoided for accounts that seek to
tions (the dissolves, wipes, and odd music used to
analogize film and language. Film theory has identisignal flashbacks in Hollywood narrativesare a less
fied a number of syntactic elements (e.g., point-ofpowerful and flexible means of approximatingthis
view editing employing perspectivallybased images,
ability).Furthermore,as severalscholarshave pointed suture) which are thought to operate as codes with
out, pictures cannot express negatives.24 These and discernibleeffects across a range of individualfilms,
otherdifferencesbetweenpictorialandlinguisticmesthusconstitutinga kindof cinematicgrammar.But the
sages point toward distinctly different communicational modalities. Interestingly, Christian Metz also
noted some of these fundamental distinctions between
linguistic and pictorial modes of communication. He
pointed out that cinema lacks the double level of
structure in language (i.e., the morphemic and phonemic levels) and is a one-way system of communication,
unlike language, where senders and receivers are interchangeable, and he concluded, therefore, that the concept of a language is probably inapplicable to film.
test of ungrammaticality is not often applied (that is,
judgements about the syntactic correctness of given
filmic constructions). The most thorough and complete
exploration of the application of linguistic principles to
problems of film structure is found in the work of John
Carroll, who has sought to apply Chomskian transformational-generative grammar to the cinema.26
Carroll identifies a series of principles underlying
cinema grammar violations of which ostensibly constitute cinematic equivalents of grammatical errors and
20
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
allegedly give rise to sequencesthatviewers willjudge
to be confusing or unfilmic. These are essentially
continuityeditingrules,suchas theinjunctionthatif an
actor looks and reacts to anythingoff-frame,the next
shotwill be interpretedby viewers as a subjectiveshot.
Carroll describes a sequence from The Birds where
Hitchcock seems to violate this principle,and he remarks that viewers will find this confusing. It is the
sceneshowingTippiHedrenwaitingoutsidetheschoolhouse while, behindher, the birdsgatheren masse on
the playgroundjungle gym. Hitchcockcuts between a
shot of Hedren looking off-frame and a shot of the
gatheringflock, as if to imply (by virtueof the cutting
pattern)that she must see it.
In a recent and thoroughreview of the issue of
visual literacy in film and television, Messarispoints
outthat,while Carrollis technicallycorrect,the salient
point is thatthe confusion (not sharedby all viewers)
is only momentary,because the surroundingcontextual-narrativeinformationmakes it clear thatthe subsequentshot is not a subjectiveone.27In otherwords,
given the narrativecontext, it would be implausibleto
assume the charactersees the birds without reacting
because their deadly nature has already been established in the story.Messarispoints out thatone reason
why notionsof grammaticalityaredifficultto applyto
film is thatnarrativecontextoftenoverrulescode. "The
role of formal conventions in conveying a movie's
meaningis generallysubordinateto conventionalstandardsof plausibilityor probability('conventional' in
the sense that the film-makermust be able to assume
themof her/hisaudience).... [T]heviewer's interpretationof edited sequences is largely a matterof crossreferencingpossible interpretationsagainsta broader
context (i.e., the larger story in the movie itself, togetherwith correspondingsituationsfromreal life and
other movies), ratherthan a matterof 'decoding' formal devices (e.g., an off-screen look followed by a cut
to a new shot). . . . Interpretationis driven by the
narrativecontext, not the code."28
Anotherexampleshouldmakethis clear,although
this instance does not involve an alleged violation of
grammar.In The Searchers, after Ethan and Marty
discoverthatLucy hasbeen killed by herIndianabductors and Brad, her lover, is killed while charginginto
the Indian camp, the searchers ride into a snowy
landscapewhereMartydespairsof everfindingDebbie.
Ethan,in reply, makes his famous speech aboutbeing
a critterthatjust keeps coming. Fadeout andfade in on
the Jorgensens'ranchas Ethanand Martyride up. As
a symbolicpictorialdevice thathas no clearanalogyin
real-life visual experience,the fade operatesin accord
with its conventional usage to bracket and separate
sectionsof the narrative,assistingthe viewer's parsing
operations.But it is the contextualnarrativeinformation that provides the salient cues about the precise
natureof the narrativeshift. At the Jorgensens',Ethan
and Martyinhabita differentlandscape.The absence
of snow indicatesa seasonal shift, andJorgensentells
Ethan that his letter about Brad's death arrivedthe
previousyear.The viewer's extra-filmicknowledgeof
seasonal patternsand, most importantly,the dialogue
between Ethan and Jorgensen establish the precise
natureof the time shift in a way thatthe purelyvisual
code-the fade-cannot. The dominanceof narrative
context over code explains why it would be equally
permissible-though not in a film of the 1950s-to
signal the temporal shift with a straight cut. With
redundantinformationalcues available in the narrative, the viewer should be able to follow the flow of
events, permittinga degreeof interchangeabilityin the
use of codes. Thus, image patternsmay not need to be
as rigidly sequenced as the concept of a "grammar"
might imply, andthe ostensive violationof grammatical rules may carryless significance (for the viewer's
abilities)thanis sometimesthought,as
interpretational
the violations of the 180-degreerule in contemporary
productionsshould indicate.
EmpiricalEvidence about
Image Comprehension
Theforegoingdiscussionhassuggestedthat
the linguisticorientationof film theorymay not be the
most efficacious for dealing with visual meaning or
cinemastructure.Currently,the disciplineis undergoing some profoundshiftsas the SaussureanandLacanian accountsarecontestedby alternativeformulations
advocating perceptually or cognitively based approaches.The cognitive turnin contemporarytheory
has openeda spacefor renewedappraisalof perceptual
evidence that runs counterto notions of film as language and their logical consequence-that viewers
must learn to interpretpictorial displays. Reviewing
the resultsof empiricalresearchon the perceptualand
cognitive processing of visual images and narratives
will help clarifysome of theproblemsinherentin using
concepts of arbitraryor relationalsignifiersto explain
principlesof cinematicmeaning and communication.
Contemporarynotions that film is analogousto a
languageare consistent with the speculationsof film
21
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
theoristsdatingfromvirtuallythe beginningof cinema
history. Boris Eichenbaum,for example, noted that
"Filmlanguageis no less conventionalthanany other
language.... Cinemahas not only its 'language,'but
also its 'jargon,'ratherinaccessibleto theuninitiated."29
B61laBalkizssuggested that, to be comprehended,the
new "form-language"
of cinemarequiredof its viewers
a new sensibility and understanding,without which
novice viewers would be baffled.30 However, to posit
language-basedmodes of cinematiccommunicationis
to implicitlyraisethe issue of visual literacyby implying thata periodof tutoringwould be logically necessaryinorderto gaininterpretivemasteryof thecinematic
vocabulary (learning, for example, that a subjective
shot representsthe view of the absentcharacteror that
the portions of an event elided by continuityediting
nevertheless still occurredwithin the narrative).The
assumptionthat untutoredor inexperiencedviewers
would encounterdifficultiesmakingsense of unfamiliar images finds some apparentconfirmationin the
existing anecdotal reports of viewers' bewildering
first-timeencounterswith motion and still pictures.31
However, these difficulties are often due less to an
inherentinabilityof naive viewers to see the pictorial
objects than to a first-time encounter with a novel
recordingsurface (e.g., paper in the case of still pictures) or a culturally unfamiliar depth cue used to
renderan objectin abstractterms(e.g., linearperspective suggesting a straight road in a line drawing).
Moreover,even whereanthropologistsandfield workers reportdifficulties with pictureperception,respondents typically are able to integrate figure-ground
relations ratherquickly, resulting in correct picture
perception.32Viewerswho findpictureperceptioncompletely impossible underany condition are quite rare
and probablyanomalous.
Convincingevidence aboutinherenthumanabilities to perceive pictures is available from a classic
experimentin which the researcherspreventeda child
(their own) from birth from seeing pictures until the
age of 19 months,when it began to actively seek them
out.33 The child learnedhis vocabularysolely
through
the use of objects and received no training regarding
pictorial meaning or content. He was nevertheless able
to recognize, when tested with a series of 21 twodimensional line drawings and photographs, the series
of pictured objects (people and things familiar in his
environment). The researchers concluded that the results indicate the existence of innate picture perception
abilities and that, if there are allegations of cultures or
viewers lacking these, it cannot be a matter of not yet
having learned the language of pictures.
Empirical research indicates that one basis for
these innateabilitiesprobablylies in the activationby
picturesof perceptualskills (objectrecognition,depth
perspective,etc.) developed in real-worldexperience,
skills which are then transferredto pictures.Scholars
have commonly emphasizedthat picturesreplicate a
series of monocular depth cues used in real-world
experience for inferringinformationabout the positioning of objects in physical space (e.g., overlap,
texturedensitygradients,shading,and, in motionpicPictureperception,then,seems
tures,motionparallax).34
based
on
3D
clearly
spatialskills transferredto the 2D
representation,and one would expect that the more
complete the set of cues employed by the picture
regardingthe spatiallayout of the depictedscene, and
the more culturallyfamiliarthe depicted objects, the
greaterthe ease of recognition. If naive viewers can
perceive still pictures, there is no logical reason for
inferringthatthey would have difficultydoing so with
moving pictures,especially since movies will supply
the additionalreal-worlddepthcue of motionparallax.
But whataboutthe role of culturein visualperception? As we have seen, culture-bounddeterminations
are very importantin language-basedfilm theories
whichrely on arbitrary,relationalsignifiers.The move
to empirical, perceptuallybased evidence might be
unwelcomefor linguisticallybased accountsof pictorial meaning because it can be seen as replacingcultural categories with biological (and therefore
ideologically suspect) ones, and because it may suggest thatpictorialandlinguisticmodalitiesaredistinct
from each other.With respectto the role of culturein
pictureperception,some evidence does exist for varying culturalsusceptibilitiesto differentvisual illusions
(the Mueller-Lyerillusion and the Sanderparallelogram),perhapsdue to variationsin the physical environmentsof differentcultures,butthese do not seem to
impactbasic pictureperceptionabilities.35
In a recent summaryof the existing crosscultural
researchon pictureperception,Deregowski suggests
that relationshipsof culture,perception,and viewing
abilitiesbe conceptualizedin termsof sets of contiguous and, at times, overlapping skills.36 Real-world
spatial experience may utilize 3D skills that 2D representations do not exploit, such as binocular disparity
(using the differences in the images recorded by each
eye as a means of inferring information about depth
and distance). Many real-world perceptual skills, by
contrast, do overlap with visual skills relevant for
picture perception, and some may be influenced by
varying patterns of cultural organization (e.g., as Segall
et al. point out, respondents from plains or dense jungle
22
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
environmentsmay demonstrateless sensitivity to linear perspectivewhen used as a pictorialdepth code).
Finally, some picturesemploy purelyrepresentational
codes with no overlap in real-life experience (e.g.,
wipes in film or streakylines used to representspeedin
comics). Obviously, with this last categoryone might
expect greater interpretationaldifficulties for naive
viewers, but not inevitably.As our example from The
Searchersindicated,narrativecontextualinformation
may be usedto connectseparateimages or events even
when the optical transitionemployed is a more symbolic one.
What,then,does all of this implyabouttheneedfor
a period of tutoringimplicit in the film-as-language
model? Quite simply, such a period should not be
necessaryfor inferringnarrativerelationsin standard
movies (i.e., movies thatdo not create deliberatenarrativeenigmas, in contrastto movies like Last Yearat
Marienbad).Empiricalresearchwithnaiveviewers(in
most cases young childrenand, in one unique study,
inexperiencedadults) offers evidence that the use of
specifically cinematic devices (or, as the film-as-language paradigmmight term them, symbolic codes),
such as montage, camera movement, or subjective
obstacles
shots,do notpose substantialinterpretational
for naive viewers provided developmentalrequisites
aremet (i.e., thatthe viewershave developedsufficient
real-worldconceptualand cognitive skills thatcan be
applied to the film or television medium). A brief
review of some of this evidence will help clarifythese
pointsandwill providefurtherevidence fromwhichto
question the theoreticaland methodologicalefficacy
of groundingfilm theory in conceptions of cinematic
relationshipsinvolving arbitrary,unmotivatedsigns.
Young children's comprehensionof an arrayof
cinematictechniqueswas exploredin a recentstudyby
Smith,Anderson,andFisher.37Inone experiment,they
showed three- and five-year-old children brief stopanimationvideo sequencesthatemployedpans,zooms,
and cuts to present a simple story. Anothergroup of
children saw the same stories in sequences with no
editing or cameramovement.When the childrenwere
invited to recreate the stories using the dolls and sets
seen in the films, no differences in story comprehension were observed between the two groups, indicating
that these film techniques seemed to pose few cognitive or interpretational problems for these young viewers. These were viewers who had seen television before.
It is, however, the absence of age-related differences in
their performance that is striking, especially in light of
the hypothesis that these techniques should require
medium-specific learning in order to be understood.
Because these sequenceswere relativelysimple in
structure(featuringonly one pan, zoom, and cut), the
researchersran a second experimentwith more complex visual presentationsand longer visual narratives.
Cinematicmanipulationsthis time consisted of parallel editing to imply simultaneityof action, subjective
shots representinga character'sliteral viewpoint, ellipses achievedthrougheditingthatdeletedportionsof
a continuous event or action, and editing to imply a
layout of contiguous spaces (e.g., cutting from an
establishingshotof two buildingsto a shotof a character looking out of a window, then to a reverse-angle
interiorshotof the characterat the window).This time
the childrenwere four and seven years old.
As before, they were asked to recreatecharacter
perspectivesandspatiallayoutsusingthedolls andsets
thathadappearedin the films. Comparisonsacross all
classes of montage indicated that clear majoritiesin
both age groups made correct inferences of space,
ellipsis, and characterperspective. Inferences of simultaneityproveddifficult for the four-year-olds,but
not for the seven-year-olds.Both groupsprovedespecially skilled at reconstructingimpliedactionsomitted
in the edited narratives.
This studyindicatesgood comprehensionby young
childrenof basic montage techniques,which is what
one wouldexpectif interpretation
of thevisualdisplays
drawson a child's developingreal-worldvisual skills
and experience. Emphasizingthe status of cinematic
images as iconic signs having a clear referentialbasis
andinvitinga transferof real-worldvisual skills to the
pictorial display does not deny the existence of medium-specific skills relevant for making sense of a
motionpicture,butit does reservea moremodestspace
for them than would the arbitrary-relational
signifier
model. Clearly, some film techniquesare more symbolic thaniconic andmay be coded in a morearbitrary
fashion. Such techniques would invite the viewer's
applicationof specific, medium-basedcompetencies.
For example, the Smithet. al study employed parallel
editing as one categoryof montage.Drawingthe correct inferencefrom parallelediting is arguablya medium-specific skill, and, as one would expect, the
largest age differences in the average percentageof
correct responses showed up there. But, ratherthan
needingto become proficientat manipulatingan arbitrarilysymbolic set of cinematic signifiers, children
cangenerallyinterpretvisualdisplaysinvolvingpointof-view editing by transferringto the display their
developing real-world visual skills and experience.
Evidence supporting a developmental view of the
applicationof real-worldcognitiveandperceptualskills
23
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
to film andtelevision displaysinvolving point-of-view
editing is found in Comuntzis-Page's study of the
relationshipbetween children's evolving real-world
perspective-takingskills andtheirabilityto infermeaning from an edited video sequence.38She found that
childrenwho successfully demonstrateda knowledge
of visual perspectives in actual 3-dimensional situations (e.g., childrenwho knew thatobserverssituated
in differentplaces can see differentsides of a common
object) were better able to understandthe changing
cameraviewpointsof a television presentation.In fact,
she found thatskill at inferringcharacterrelationships
in a 3-dimensionallayout seemed to be a prerequisite
for makingtheproperanalogousinferencesfromthe 2dimensionalvideo display.Only childrenwho did well
at the formertask also did well on the latter,and clear
age-relateddifferencesbetweenperspectivetakersand
nonperspectivetakerswere found, supportinga developmentalview.
The researchcited thus far has used children to
representa populationof relativelyunskilledviewers.
However, one uniquestudy of motion pictureperception and interpretationexists using adultviewers who
had little familiaritywith any mass media.39Working
witha seminomadic,nonliterate,pastoraltribein Kenya,
a community whose aesthetic expressions concentratedon personaladornmentandperformancearts,the
researchersshowed adultvillagerstwo videotapesof a
culturallyfamiliarstory.One versionwas unedited,the
other featured 14 cuts with frequent alterations of
close-ups, medium shots, long shots, and zooms. No
significant differences in ability to recall story information were found between respondentswho viewed
the editedanduneditedversions.Fragmentationof the
visual scene through point-of-view editing did not
hindercomprehension,nor did it seem to requirethe
use of any medium-specific skills. The researchers
concluded that, on the contrary, continuity editing
codes thatmanipulatepointof view seem to functionas
"analogsof perceptualprocesses."40
Levels of Iconicity
The evidence reviewed thus far bears on the
capability of pictures to furnish cues regarding the
spatial layout of a scene or situation that are analogous
to sources of information commonly found in realworld visual experience. It also bears mentioning,
however, that another level of iconic information typically exists in pictures which is central to a consider-
ationof the comprehensibilityandemotionaleffects of
thecinema.Photographicimages of peoplenecessarily
reproducethatinformationaboutfacialexpressionand
gesturewhich Ray Birdwhistellhas called "kinesics."
Birdwhistellandhis colleagues studiedthe systematic
patterningof body motion within cultures and designed anelaboratenotationalsystemto transcribeand
studythedistinctivelypatternedstringsof bodymotion
cues (called "kines")that are situationallyarticulated
as a communicativeformin everydaylife. Birdwhistell
explicitly rejected the notion that language was the
most importantchannel of communication,arguing
insteadfor a conceptionof communicationas a multimodal, multi-sensoryprocess which included visual
andgesturalas well as auditorychannels.41He demonstratedhow body motions are culturallypatternedin
symbolically significant ways that are understoodby
membersof a given community whose socialization
processes have sensitized them to coded kinesic displays. While Birdwhistellemphasizedthe importance
of cultural context in determining the encoding of
specific kinesicpatterns,otherresearchershaveargued
thatsome gesturalexpressions-those on the face, for
example-may function as biologically based panculturalsignals for emotion. While all culturesmandate display rules governing who may display which
emotions and when, experimentalevidence furnished
by Paul Ekmanand his colleagues has indicatedthat
certainfacial displays seem correlatedwith basic categories of human emotion and are capable of being
recognized crossculturally.42
The obvious point to be made from this brief
discussion of the work on facial and body motion
communicationis thatthe motion picturecamerafurnishes an excellent means of recordingthe expressive
meanings carriedby these channels. Birdwhistell,in
fact, employed still photographsand motion picture
footageto recordandstorethedatausedin his analyses.
It shouldbe anassertionof the obviousto pointout that
a majorsource of the appealandpower of the movies
lies here, in film's ability to capturethe subtletiesand
nuancesof socially resonantstreamsof kinesicexpressions, and not just to passively capture them but, via
close-ups and other expressive devices, to intensify
and emphasize the most salient cues for the viewer' s
understanding in cognitive and affective terms of the
meaning of the scenes depicted on screen. (Recent
studies of acting by James Naremore and Roberta
Pearson have emphasized the way that different performance traditions code expressive behavior and how
film technique may be used to frame and emphasize
this coding.43) The relational-arbitrary signifier hy-
24
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
pothesis has tended to draw attentionaway from this
clear source of iconic meaning in motion pictures,
despite its obvious importancein helping to explicate
the crossculturalappealand power of the movies.
be cautiousin speakingaboutthe discourseof pictures:
"The analyst must recognize that the very fact of
talkingaboutnarrationrequiresa re-presentationof it
by verbalor othermeanswhichmay captureonly some
of its features."46
The concept of the relational,arbitrarysignifierwould seem most applicableto pictures
here, in the languageof analysiswhich is employedto
The Discourse of Pictures
Furintellectuallymanipulatepictorialrepresentations.
thermore,whileempiricalevidenceindicatesthatviewWhat do these studies and avenues of reers easily and readily make the requiredinferences
searchtell us aboutcinematiccodes and the appropri- necessaryto sequencea series of picturesinto a coherateness of defining them, as currentfilm theorydoes,
ent narrative,even to the extent of inferringinformaas a series of relationaldifferences among arbitrary tion about situations not directly pictured,47several
signs? The empirical evidence clearly suggests that scholars have argued that these inferences are orgapictorialidentificationskills do not develop from an nized on a conceptualandpropositional-but not necextendedperiodof exposureto significationand conessarily a linguistic-basis. Pylyshyn, for example,
sequentlearning,as do languageskills, andthatthis is
suggestedthat "Suchconcepts and predicatesmay be
due
to
the
fact
that
most
realistic
are
probably
pictures
perceptuallywell defined withouthaving any explicit
naturallanguagelabel.
isomorphicwith correspondingreal-worldvisual disThus we may have a concept correspondingto the
plays, unlike symbolic signs, which have a more arbitraryrelationshipto whatthey represent.Furthermore, equivalenceclass of certainsounds or visual patterns
via the technologies of motion picturerecording,the
without an explicit verbal label for it. Such a view
camerais ableto reproducein clearlyrecognizableand implies that we can have mental concepts or ways of
even intensifiedformthe familiarstreamsof facial and abstractingfrom our sense datawhich are beyond the
body motion cues which symbolically encode the reachof our currentstock of words, but for which we
meaningsof the social situationsportrayedon screen. could develop a vocabulary if communicatingsuch
These cues should be readily understoodby cinema concepts became important."48Linguistic models, in
viewers, just as they are in real-worldvisual experi- other words, are not requisitefor explaining how we
ence. If the distinctionsbetween iconic and symbolic respondto andmakesense out of pictorialinformation,
modes that we have been emphasizingare really relnoreven for describinghow we encodevisualinformaevant to differences between pictures and language, tion for longtermmemory storage such that it can be
then one would expect iconic modes to be processed subsequentlyreaccessed.49
morereadilythansymbolic ones, especially by young
As noted, arbitrary-relational
signifiermodels are
children, and existing research lends supportto this
mostapplicablewherethe discoursebeingexaminedis
idea.44
language-basedratherthan strictly pictorial. In this
Thus, it is not clear how the concept of the rela- respect, it follows that theories about ideology and
tional, arbitrarysignifier might apply to cinematic ideological effects arebest formulatednot with regard
images or, for that matter,to any iconic image. Cer- to pictures sui generis (e.g., as has been done with
tainly, as Hockett indicated, few communicational perspective-basedimages) or with such machinesfor
signs are completely iconic, since a state of total seeing as the camera and projector,but ratheras a
iconicity would imply that the sign was completely matterof thecontentof a given film andits modulation
via techniquesof cinematicstyle. As Noe1 Carrollhas
indistinguishablefromits referent.45Nevertheless,pictorialsigns do bearclear structuralsimilaritiesto their remarked,ideological design needs to be examined
referents,with the attendantconsequencesfor percep- film by film, on an individual case-by-case basis.
tion and comprehensionthat we have just reviewed. Becausetheoriesaboutlinguisticstructuredo not seem
The linguisticmodel has provento be a very powerful to be directly applicable to
pictures, they therefore
paradigmfor the analysis of images and larger film
form an insufficientbasis for groundinga critiqueof
structures(e.g., dramaticscenes) in partbecause that cinematicdiscourseortheplaceof
ideology in pictorial
analysis is typically conductedvia words. When film
images.
scholars analyze images, visual informationis transProponentsof the applicationof the Saussurean
lated into verbal description,one modality is substi- lineage to cinema studies
might reply that the
tutedfor another.But, as Braniganhas noted,one must poststructuralattentionto
textualityand discourse as
25
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
culturallyrelativeconstructsprovidesa frameworkfor
investigating ideology and the cultural or political
polyvalence of a given text. That the poststructural
methodology works in this way is undoubtedlytrue.
But, unfortunately,what tends to be displaced are
issues of how cinema is able to communicate
crossculturally(i.e., attain global popularity)and the
even more basic questions of what makes the cinema
intelligible to its viewers. Not all cultures organize
libidinalandpsychic energy in the same way, andone
would not thereforewish to posit Lacaniancategories
as basic mechanisms explicating the means of cinematiccommunication.Yet all culturescurrentlystudied do demonstrate clear pictorial and cinematic
perceptionabilities.Moreover,theseabilitiesareshared
with a varietyof animals.5Picturerecognitionabilities
have been demonstrated across a wide range of
nonhuman subjects-primates, birds, fish, reptiles,
even insects. These include unlearned,spontaneous
responses to still and motion pictures as well as responsesthataretheresultof conditioning.Recognition
of objects by nonhuman subjects has been demonstratedacross different classes of pictorial mediablack-and-whiteandcolor photographs,high-contrast
photographs,film and videotape,even line drawings.
Comprehensionof filmed events has been elicited, as
well as the formationand pickup of object-classconcepts from pictures (e.g., as in the ability of pigeons,
trainedto recognizehumanfiguresin slides of various
environments, to transfer this ability and the class
concept of "humanbeing" to sets of slides they had
never seen before).
Positing arbitrarysignifiers and unconsciousprocesses of the mind seems to deflect the ability of
cinema studiesto grapplewith some of the most basic
questions about cinema, namely, how are viewers
crossculturallyable to make sense of the mediumand
what is the necessary biological basis of pictorial
communicationthat makes it so effective for human
and nonhuman subjects alike. As Noe1 Carroll has
recently suggested, emphasizing the way that film
structureworks to secure the cognitive clarity of the
experience for the viewer can give us a basis for
explainingthe attractivenessand appeal,and perhaps
even the emotionalpower, of the medium.51
Furthermore,emphasizing the cinematic code as
an arbitrary,unmotivatedsign places film theoryin the
uncomfortableposition of implying certain consequences for the real-worldviewer thatare contraryto
the observable evidence, namely, that cinematic images should not function in ways that are isomorphic
with theirreal-worldcounterparts,thatviewers should
haveto learnthe symbolicmeaningsof basiccinematic
structure, and that, logically, because discourse is
culture-bound,the appealof the mediumshouldbe too.
Film theoryhas dealt with these problemsby pointing
to the transparencyeffect of the cinema or to the
illusionism of the photographicimage based on what
are regardedas the culture-boundand symbolic conventions of perspective. But, as this article has suggested, some importantgaps and contradictionsare
found between these formulations, the Saussurean
lineage, andthe empiricalevidence regardingpictorial
recognition and visual communication.Unless film
theory can reconceptualizethe cinema as an iconic,
ratherthanas a purelysymbolic, mode of communication, it is difficult to see how these serioustheoretical
gaps may be closed. We need to recovera recognition
of the analogicalcomponentof pictorialsigns. Rather
than dealing with this component in metaphorical
terms(e.g., via transparencyeffects, subjectpositioning in the discourseof the Imaginary,or illusionism),
an appreciationof the isomorphic relations between
pictorialsigns and theirreferents,and attentionto the
differencesbetween pictorial and linguistic modes of
communication,can help invigorate film theory by
reconnectingit to the observable experiences of real
viewers.
As noted at the beginningof this essay, the intention hereis not to remove culturalconsiderationsfrom
questions aboutfilm viewing and interpretation.Culture clearly enters into the inferences viewers draw
from cinematicimages and narrativeswheremeaning
maybe constructedin termsof theintellectualhorizons
providedby class, race, gender,and similarvariables.
But it has been the contentionof this essay thatculture
plays a less decisive role at the level of comprehension
discussedherein.Furthermore,only by knowingwhere
cultural considerationsshade off into what may be
moreproperlytermedphysiologicalor cognitivecapabilities are we likely to be able to apply cultural
analyses in fruitfuland relevantways. This approach
can help theoryexplicate the intelligibilityandpower
of the movies in ways that are responsive to, and
grounded in, the observable experiences of actual
viewers.
* StephenPrinceteaches in the
CommunicationStudies Department
of VirginiaTech. He is the Book
Review Editorof Film Quarterly.
26
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
NOTES
My thanks to Carl Plantinga for reading an earlier version
of this essay and offering helpful suggestions.
1. RobertStam, "FilmandLanguage:FromMetz to Bakhtin,"
in The Cinematic Text: Methods and Approaches, ed. R.
Barton Palmer (New York: AMS Press, 1989), p. 277.
2. See Pier Paolo Pasolini, Heretical Empiricism, ed. Louise
K. Barnett, trans. Ben Lawton and Louise K. Barnett
(Bloomington: Indiana University Press, 1988) and Peter
Wollen, Signs and Meaning in the Cinema (Bloomington:
Indiana University Press, 1976), especially pp. 116-54.
3. Philip Rosen, ed., Narrative, Apparatus, Ideology (New
York: Columbia University Press, 1986), p. 3.
4. "Introduction,"in Explorations in Film Theory: Selected
Essays from Cine-Tracts, ed. Ron Burnett (Bloomington:
Indiana University Press, 1991), p. xv.
5. In the influential analyses of Jean-Louis Baudry, these
functions are a consequence of the machinery of camera
and projector, irrespective of the content and subjects that
are filmed. See "Ideological Effects of the Basic Cinematographic Apparatus"and "The Apparatus:Metapsychological Approaches to the Impression of Reality in the
Cinema," in Rosen, ed., Narrative, Apparatus, Ideology,
pp. 286-318.
6. Stephen Heath, Questions of Cinema (New York:
Macmillan, 1981), p. 26.
7. D. N. Rodowick, The Difficulty of Difference: Psychoanalysis, Sexual Difference and Film Theory (New York:
Routledge, 1991), p. 138.
8. Mas'ud Zavarzadeh, Seeing Films Politically (Albany:
State University of New York Press, 1991), p. x.
9. Ibid, p. 11.
10. Michael Devitt and Kim Sterelny, Language and Reality
(Cambridge, MA: MIT Press, 1987), pp. 215-18.
11. Ibid, p. 216.
12. Burnett, Explorations in Film Theory, p. xvii.
13. Heath, Questions of Cinema, p. 191.
14. "Introduction,"in Movies and Methods, vol. 2, ed. Bill
Nichols (Berkeley, CA: University of California Press,
1986), p. 16.
15. The relevant essays explicating this position can be found
in Language, Thought and Reality: Selected Writings of
Benjamin Lee Whorf, ed. John B. Carroll (Cambridge,
MA: MIT Press, 1984).
16. See, forexample, JohnB. CarrollandJosephB. Casagrande,
"The Function of Language Classifications in Behavior,"
in Readings in Social Psychology, ed. EleanorE. Maccoby,
Theodore M. Newcomb, and Eugene L. Hartley (New
York: Holt, Rinehart and Winston, 1958), pp. 18-31. It
should be noted, however, that the results of this study,
while offering some support of the Whorfian view, are
nevertheless mixed and somewhat inconsistent.
17. For a review of evidence bearing on strong and weak
versions of Whorfianhypothesis, see EarlHunt and Franca
Agnoli, "TheWhorfian Hypothesis: A Cognitive Psychology Perspective," Psychological Review 98, no. 3 (1991),
pp. 377-89.
18. Bill Nichols, Ideology and the Image (Bloomington: Indiana University Press, 1981), p. 26.
19. Zavarzadeh, Seeing Films Politically, p. 102.
20. Brent Berlin and Paul Kay, Basic Color Terms: Their
Universality and Evolution (Berkeley, CA: University of
California Press, 1969).
21. Ibid., pp. 4-5.
22. Julian Hochberg, "The Perception of Moving Images,"
Iris, no. 9 (Spring, 1989), p. 41.
23. See Charles F. Hockett, "Logical Considerations in the
Study of Animal Communication," in The View from
Language: Selected Essays 1948-1974 (Athens, GA: University of Georgia Press, 1977), pp. 124-62.
24. Sol Worth, "Pictures Can't Say Ain't," in Film/Culture,
ed. Sari Thomas (Metuchen, N.J.: Scarecrow Press, 1982),
pp. 97-109, andEdwardBranigan,"HereIs a Pictureof No
Revolver," WideAngle 8, no. 34 (1986), pp. 8-17.
25. Noam Chomsky, Reflections on Language (New York:
Pantheon, 1975), p. 71.
26. See John M. Carroll, Toward a Structural Psychology of
Cinema (New York: Mouton, 1980).
27. Paul Messaris, Visual Literacy: How Images Make Sense
(Boulder, CO: Westview Press, forthcoming).
28. Ibid.
29. Boris Eichenbaum, "Cinema Stylistics," in Russian Formalist Film Theory, ed. Herbert Eagle (Ann Arbor, MI:
University of Michigan Press, 1981), p. 77.
30. B6la Balhzs, Theoryofthe Film (New York: Dover, 1970),
p. 34.
31. See the accounts in Balhizs,Theoryof the Film, pp. 34-35,
andJ. B. Deregowski, "RealSpace and RepresentedSpace:
Cross-Cultural Perspectives," in Behavioral and Brain
Sciences 12 (1989), pp. 56-59.
32. Messaris also makes this point.
33. JulianHochberg and Virginia Brooks, "PicturePerception
as an Unlearned Ability: A Study of One Child's Performance," American Journal of Psychology 74, no. 4 (Dec.
1962), pp. 624-28.
34. Discussing these cues, David Bordwell points out that
cinematic images also routinely distort or alter them (e.g.,
through the effects of different focal length lenses, etc.).
David Bordwell, Narration in the Fiction Film (Madison,
WI: University of Wisconsin Press, 1985), pp. 107-10.
35. See MarshallH. Segall, Donald T. Campbell, and Melville
J. Herskovits, The Influence of Culture on Visual Perception (Indianapolis, IN: Bobbs-Merrill, 1966).
36. Deregowski, "Real Space," pp. 69-72.
37. Robin Smith, Daniel R. Anderson, and CatherineFischer,
"Young Children's Comprehension of Montage," Child
Development 56 (1985), pp. 962-71.
38. Georgette Comuntzis-Page, "Changing Viewpoints in
Visual Presentations: A Young Child's Point of View,"
paper presented at Eastern Communication Association
Convention, Baltimore, 1988.
39. Renee Hobbs, Richard Frost, Arthur Davis, and John
Stauffer, "How First-Time Viewers Comprehend Editing
Conventions," Journal of Communication 38 (1988), pp.
50-60.
40. Ibid., p. 58.
41. See Ray L. Birdwhistell, Kinesics and Context (Philadelphia: University of Pennsylvania Press, 1970).
42. See Paul Ekman and Wallace V. Friesen, "Constants
Across Cultures in the Face and Emotion," Journal of
Personality and Social Psychology 17, no. 2 (1971), pp.
27
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions
43.
44.
45.
46.
47.
48.
49.
50.
124-29, and Paul Ekman, "Expression and the Nature of
Emotion," inApproaches to Emotion, ed. Klaus R. Scherer
andPaul Ekman (Hillsdale, N.J.: Lawrence ErlbaumAssociation, 1984), pp. 319-43. The latter volume is especially
useful for containing a range of theoretical positions, often
competing, on the nature and expression of emotion and
the ways it may be culturally coded.
See James Naremore,Acting in the Cinema (Berkeley, CA:
University of California Press, 1988) and RobertaPearson,
Eloquent Gestures: The Transformation of Performance
Style in the Griffith Biograph Films (Berkeley, CA: University of California Press, 1992).
See S. L. Calvert, A. C. Huston, B. A. Watkins, and J. C.
Wright, "The Effects of Selective Attention to Television
Forms on Children's Comprehension of Content," Child
Development 53 (1982), pp. 601-10, and D. S. Hayes and
"Preschoolers' Retention of Televised
D. W. Bimrnbaum,
Events: Is a Picture Worth a Thousand Words?,"Developmental Psychology 16 (1980), pp. 410-16.
Hockett, "Logical Considerations in the Study of Animal
Communication," p. 143.
Branigan, "Here Is a Picture of No Revolver," p. 11.
See Patricia Baggett, "Memory for Explicit and Implicit
Information in Picture Stories," Journal of Verbal Learning and Verbal Behavior 14 (1975), pp. 538-48.
Zenon W. Pylyshyn, "What the Mind's Eye Tells the
Mind's Brain:A Critiqueof Mental Imagery,"Psychological Bulletin 80, no. 1 (July 1973), p. 7.
However, with regard to certain categories of picturesnamely, relatively ambiguous ones-some evidence exists to indicate that linguistic cues may influence the
recollection of details of pictorial form. See the classic
study by L. Carmichael, H. P. Hogan, and A. A. Walter,
"An Experimental Study of the Effect of Language on the
Reproduction of Visually Perceived Form," Journal of
Experimental Psychology 15 (1932), pp. 73-86, and the
replication by W. C. H. Prentice, "Visual Recognition of
Verbally Labelled Figures,"American Journal ofPsychology 67 (1954), pp. 315-20.
For a review of picture-perception research involving
animals, see Patrick A. Cabe, "Picture Perception in
Nonhuman Subjects," in The Perception of Pictures, vol.
2, ed. Margaret A. Hagen (New York: Academic Press,
1980), pp. 305-43.
51. Noel Carroll, Mystifying Movies (New York: Columbia
University Press, 1988).
Expanded 2nd Edition!
FilmStyle& Technology:
History& Analysis
BarrySalt
* The first and only history of film style
* A new practical approach to film theory and
film analysis
* 30% bigger! New discoveries and insights!
Thousands more films examined! Twice as
much on the 1907-13 period!
* Still the only source for what really happened
in the first 25 years.
* Also now full chapters on the 'seventies and
'eighties.
Ernest
"marvellous"Ray Durgnat;"capitalimportance"Callenbach;"indispensable"JonathanRosenbaum.
FRENCH
$34.95 in the USA from SAMUEL
7623 Sunset Blvd., Hollywood, CA 90046
phone: (213) 876-0570
STARWORD
Writers
"FromaFitzgeraldto
Dos Passos to
of the
ingway, a crewHem-ly
majorwritersof our
Copying notice: Authorization to photocopy items for
internalor personal use, or the internalor personal use of
specific clients, is granted by the Regents of the University of California for libraries and other users registered
with the Copyright Clearance Center (CCC) Transactional Reporting Service, provided that the base fee of
$1.00 per copy is paid directly to CCC, 27 Congress
Street, Salem, MA 01970. For permission to photocopy
materials for classroom use, call CCC's APS Customer
Service Department at (508) 744-3350 or fax them at
(508) 745-9379.
*
*
H
ollyw
in
ood,
time lived and worked
inHolly- 1928-1940
foraspace
They all
had somethingto say Richard Fine
and RichardFine puts Witha new preface
in a
it all
wood....
together
book that capturesthe Alice-in-Wonderland
persona
of this never-neverland. -Stages
By placingthe legendof Nollywood as the
destroyerof writersin a culturaland historiccontext,
Fine showsthat it was not the talentof thesewriters
that was underattack,but ratherthe professionof
as they understoodit.
authorship
216
b&w
1-56098-263-2P
40
photographs pp. Paper:
$14.95
j] SMITHSONIANINSTITUTIONPRESS
Smithsonian Studies in the History ofFilm and Television
Dept.900 * BlueRidgeSummit,PA 17294-0900* 800/782-4612
28
This content downloaded from 128.173.125.76 on Tue, 18 Feb 2014 15:36:20 PM
All use subject to JSTOR Terms and Conditions