Audiovisual translation

13

social agents in the global political space of the the terms used to designate this field of enquiry
asylum system. They contribute consciously and during the period in question. Considering that
unconsciously to both inclusionary and exclu- the mainstream forms of audiovisual translation
sionary practices as they negotiate linguistic and – i.e. subtitling and dubbing – were born on the
cultural meanings in the context of institutional back of sound motion pictures, it is only natural
and political agendas and national and interna- that the terms ‘film dubbing’ and ‘film translation’
tional relations of power. came to feature prominently in early scholarly
work (Fodor 1976; Snell-Hornby 1988). The
See also: subsequent emergence of television as a mass
community interpreting; globalization; medium of communication and entertainment
institu­tional translation; minority; provided new avenues for the dissemination of
mobility. translated audiovisual texts, with labels such as
‘film and TV translation’ (Delabastita 1989) and
Further reading ‘media translation’ (Eguíluz et al. 1994) gaining
Barsky 1996, 2001; Blommaert 2001; Inghilleri visibility in the literature. The most recent devel-
2003; Pöllabauer 2004; Bot 2005; Inghilleri opments relate to the exponential growth in the
2005a; Jacquemet 2005; Maryns 2006; Pöllabauer volume of audiovisual texts produced by and for
2006; Inghilleri 2007a, 2007b. electronic and digital media. Terms like ‘screen
translation’ (Mason 1989; O’Connell 2007) and
moira inghilleri ‘multimedia translation’ (Gambier and Gottlieb
2001) illustrate the extent to which audiovisual
translation has outgrown its core domain of

Audiovisual
enquiry and annexed neighbouring fields under
an all-inclusive research agenda.

translation The genealogy of audiovisual

Audiovisual translation is a branch of trans- translation
lation studies concerned with the transfer of
multimodal and multimedial texts into another Even during the silent film era, exporting films
language and/or culture. Audiovisual texts are to foreign markets involved some form of inter-
multimodal inasmuch as their production lingual mediation. The turn of the twentieth
and interpretation relies on the combined century witnessed the incorporation of written
deployment of a wide range of semiotic resources language into the conglomerate of film semiotics
or ‘modes’ (Baldry and Thibault 2006). Major in the form of intertitles (Ivarsson 2002). The use
meaning-making modes in audiovisual texts of these texts placed between film frames grew
include language, image, music, colour and in parallel with the emergence of increasingly
perspective. Audiovisual texts are multimedial complex filmic narratives. Intertitles situated
in so far as this panoply of semiotic modes is the action in a specific temporal and spatial
delivered to the viewer through various media setting, provided viewers with insights into the
in a synchronized manner, with the screen characters’ inner thoughts and helped them
playing a coordinating role in the presentation negotiate the discrepancies between screen time
process (Negroponte 1991). and real time, during a period when filmic
Since the 1970s, screen-based texts have techniques were rudimentary (Dick 1990).
become increasingly ubiquitous. Scholars have Removing the original intertitles and inserting
been quick to bring the investigation of new a new set of target language texts back into
textual manifestations – ranging from software the film was all that was required to exploit it
to videogames – into the remit of audio- commercially in a foreign market. But intertitles
visual translation research, thus extending the also served as the springboard for the devel-
boundaries of this area of study (see also local- opment of new forms of audiovisual translation.
ization). Chaume (2004) has documented the In-house commentators were employed to fill
successive stages of this expansion by looking at the same gaps as the intertitles (Dreyer-Sfard

TF17062.indb 13 26/9/08 11:06:08

 14 Audiovisual translation

1965; Díaz Cintas 2003), although these enter- and their acceptance by European audiences
tainers often sought to enhance the viewer’s ended the moratorium on American control
experience by spreading gossip about the film of European markets (Forbes and Street 2000),
stars or even explaining how the projector with American films regaining a market share of
worked (Cazdyn 2004). The national industries 70 per cent in Europe and Latin America by 1937
of the USA and a number of European countries (Chaves 2000). This second wave of domination
thrived on this absence of linguistic barriers was regarded as a threat not only to the sustain-
to film exports until the aftermath of World ability of Europe’s national film industries, but
War I took its toll on the financial capability also to their respective languages, cultures and
of European industry to fund new projects. By political regimes – in the mid-1930s, the latter
the early 1920s, American films had come to ranged from democratic systems to fascist dicta-
secure a dominant market share throughout torships. The multiplicity of European interests
Europe, pushing some national film industries and ideologies would soon lead each country
(e.g. British and Italian) close to the brink of to adopt its own protectionist measures and/
collapse (Nowell-Smith and Ricci 1998). or censorship mechanisms (Nowell-Smith and
According to Forbes and Street (2000), Ricci 1998), which were, in many cases, enforced
the advent of sound in the late 1920s put a through the choice of specific policies and forms
temporary end to the American domination of audiovisual translation. Despite these efforts,
of European film industries, as the big studios and except for brief exceptional periods like
became suddenly unable to satisfy the demand World War II, these dynamics of domination
of European audiences for films spoken in their were to remain unchanged.
native languages. Experimental attempts to
appeal to local European sensibilities – e.g. the
‘multilingual filming method’ and the ‘dunning
Subtitling
process’ (Ballester 1995) – failed to earn the
American industry its lost market share back,
A typology of subtitling procedures
and it soon became obvious that new forms of
audiovisual translation were required to reassert Subtitling consists of the production of
its former dominance. During the second half snippets of written text (subtitles, or captions
of the 1920s, technological developments in American English) to be superimposed on
made it possible to ‘revoice’ certain fragments visual footage – normally near the bottom of the
of dialogue or edit the sound of scenes that frame – while an audiovisual text is projected,
had been shot in noisy environments through played or broadcast. In so far as it involves
a process known as ‘post-synchronization’ a shift from a spoken to a written medium,
(Whitman-Linsen 1992; Chaves 2000). Despite subtitling has been defined as a ‘diasemiotic’
being conceived as a means of improving the or ‘intermodal’ form of audiovisual translation
quality of an original recording, post-synchro- (Gottlieb 1997). Interlingual subtitles provide
nized revoicing was soon used to replace the viewers with a written rendition of the source
source dialogue with a translated version, and text speech, whether dialogue or narration, in
is therefore acknowledged as the immediate their own language. In communities where at
forerunner of dubbing as we know it today. least two languages co-exist, bilingual subtitles
Concurrent advances in the manipulation of deliver two language versions of the same source
celluloid films during the 1920s allowed distrib- fragment, one in each of the two constitutive
utors to superimpose titles straight onto the film lines of the subtitle (Gambier 2003a).
strip images through optical and mechanical Each of the fragments into which subtitlers
means (Ivarsson 2002). By the late 1920s it had divide the speech for the purposes of trans-
become customary to use this evolved version of lation must be delivered concurrently with its
the primitive intertitles to provide a translation written rendition in the target language via the
of the source dialogue in synchrony with the subtitle. And given that ‘people generally speak
relevant fragment of speech, thus paving the much faster than they read, subtitling inevitably
way for the development of modern subtitling. involves . . . technical constraints of shortage
The perfection of these new techniques of screen space and lack of time’ (O’Connell

TF17062.indb 14 26/9/08 11:06:08

 Audiovisual translation 15

1998:67). Subtitles composed according to teletext (Neves 2007). They are known as ‘closed
widely accepted spatial parameters contain a subtitles’ because they are accessible only to
maximum of two lines of text, each accom- viewers whose television sets are equipped
modating up to 35 characters (Karamitrouglou with the relevant decoder and who choose to
1998). The actual number of characters that display them on the screen while watching the
can be used in each subtitle then depends on programme. The advent of DVD and digital
the duration of the corresponding speech unit television represents a departure from this
(Titford 1982). tradition as both media provide viewers with
Since the 1970s, we have witnessed the prolif- closed intralingual and interlingual subtitles,
eration of intralingual subtitles, which are normally in more than one language.
composed in the same language as the source
text speech. Intralingual subtitles were tradi-
The subtitling process
tionally addressed at minority audiences, such
as immigrants wishing to develop their profi- The subtitler’s basic working materials have
ciency in the language of the host community traditionally included a time-coded VHS copy
or viewers requiring written support to fully of the source film or programme and a ‘dialogue
understand certain audiovisual texts shot in list’; i.e. an enhanced post-production script
non-standard dialects of their native language containing a transcription of the dialogue, a
(Díaz Cintas 2003). However, intralingual description of relevant visual information
subtitling has now become almost synonymous and sometimes notes for the translator (Díaz
with subtitling for the deaf and hard of Cintas 2001). The text is typically subjected to
hearing in the audiovisual marketplace, where a ‘spotting’ process, during which the dialogue
accessibility-friendly initiatives are receiving is divided into segments that are time-cued
increasing attention. Subtitles for the hard of individually. Each dialogue segment is then
hearing provide a text display of the speech but translated or transcribed in compliance with
also incorporate descriptions of sound features certain segmentation and editing conventions
which are not accessible to this audience. To (Karamitrouglou 1998), including time–space
compensate for their higher density (Wurm correlation standards. The output of this process,
2007), this type of subtitle complies with specific normally an electronic list of spotted subtitles,
conventions in terms of timing, text positioning is then returned to the commissioner of the
and use of colours (Neves 2005). Although translation. In recent years, increased circu-
subtitles for the deaf were for a long time lation of audiovisual texts in digital format and
restricted to films and programmes recorded the development of dedicated software applica-
in advance, the development of real time or tions have brought about important changes
live subtitling technologies, ranging from the in the subtitling process. Although these new
stenograph and stenotyping methods to speech technologies are not necessarily available to all
recognition systems (Lambourne 2006), has freelance professionals, they now allow subti-
increased the accessibility of live news, live chat tlers to complete a project – including the actual
shows and reality TV to the deaf community transference of subtitles onto the text – using a
(see also signed language interpreting). standard computer.
Historically, the terms ‘interlingual’ and
‘intralingual subtitles’ correlated with open
and closed subtitles, respectively. Interlingual
subtitles have tended to be printed on the Advantages and limitations of subtitling
actual film, thus becoming part of the audio- Empirical evidence suggests that subtitles can
visual text itself. Given that they are visually deliver 43 per cent less text than the spoken
present throughout the screening and univer- dialogue they derive from (de Linde and Kay
sally accessible to all viewers (except for the 1999). Given the constraints arising from the
visually impaired), interlingual subtitles are synchronous alignment between spoken sound
said to be open. Intralingual subtitles, however, and written subtitles that the industry requires
have tended to be encoded in the broadcast (Naficy 2001), subtitlers are expected to prior-
signal using a number of technologies, mainly itize the overall communicative intention of an

TF17062.indb 15 26/9/08 11:06:08

 16 Audiovisual translation

utterance over the semantics of its individual these procedures, except for lip-synchronized
lexical constituents (Gottlieb 1998). Deleting, dubbing, which is commonly referred to as
condensing and adapting the source speech are ‘dubbing’. Although all these methods involve
thus some of the most common subtitling strat- a greater or lesser degree of synchronization
egies deployed by professionals. Under such between soundtrack and on-screen images,
tight medium-related constraints, subtitling is the need for synchronization is particularly
claimed to foster cultural and linguistic stand- important in the case of dubbing.
ardization (Fawcett 2003; Díaz Cintas 2005) Voice-over or ‘half-dubbing’ (Gambier
by ironing non-mainstream identities – and 2003a) is a method that involves pre-recorded
their individual speech styles – out of the trans- revoicing: after a few seconds in which the
lated narrative. Pragmatically, this streamlining original sound is fully audible, the volume is
process can affect, for instance, the impression lowered and the voice reading the translation
that viewers form of characters in terms of becomes prominent. This combination of
friendliness (Remael 2003). In terms of Venuti’s realism (the original sound remains available
‘domestication/foreignization’ dichotomy, the in the acoustic background throughout) and
subtitling process typically leads to the domesti- almost full translation of the original text
cation of the source dialogue and the effacement (Luyken et al. 1991) makes voice-over particu-
of the translator (Ulrych 2000). larly suitable for interviews, documentaries and
Subtitling can be viewed as a form of ‘overt other programmes which do not require lip
translation’ (Battarbee 1986) since it allows synchronization. Voice-over is also used today
viewers to access the original speech (see to translate feature films for some small markets
quality). Effectively, this empowers viewers in Europe and Asia because it is substantially
who have some knowledge of the source cheaper than dubbing (O’Connell 2007).
language but are unaware of how the subti- Although it is not always pre-recorded,
tler’s work is conditioned by media-related narration has been defined as ‘an extended
constraints to monitor and criticize the trans- voice-over’ (Luyken et al. 1991: 80). This form
lation. Criticisms are often levelled at subtitling of oral transfer aims to provide a summarized
because it represents an intrusion on the image but faithful and carefully scripted rendition of
and its processing requires a relatively intensive the original speech, and its delivery is carefully
cognitive effort on the part of the viewer, thus timed to avoid any clash with the visual
detracting from the overall viewing experience. syntax of the programme. In recent years, a
On the positive side, advocates of subtitling very specific form of pre-recorded, mostly
highlight the fact that it respects the aesthetic intralingual narration has become increasingly
and artistic integrity of the original text. The important to ensure the accessibility of audio-
viewer’s exposure to a foreign language has also visual products to the visually impaired: this
been found to promote the target audience’s is known as audio description. An audio
interest in other cultures (Danan 1991). And description is a spoken account of those visual
finally, subtitling is a comparatively cheap and aspects of a film which play a role in conveying
fast form of audiovisual translation (Dries its plot, rather than a translation of linguistic
1995). content. The voice of an audio describer delivers
this additional narrative between stretches of
dialogue, hence the importance of engaging
Revoicing in a delicate balancing exercise to establish
what the needs of the spectator may be, and to
Although there is a lack of consensus on the ensure the audience is not overburdened with
scope of the term ‘revoicing’ (Luyken et al. excessive information.
1991; Baker and Hochel 1998), it technically As opposed to these pre-recorded transfer
designates a range of oral language transfer methods, other forms of revoicing are performed
procedures: voice-over, narration, audio on the spot by interpreters, presenters or
description, free commentary, simultaneous commentators by superimposing their voices
interpreting and lip-synchronized dubbing. over the original sound. Free commentary, for
In practice, ‘revoicing’ tends to encompass all example, involves adapting the source speech

TF17062.indb 16 26/9/08 11:06:08

 Audiovisual translation 17

to meet the needs of the target audience, rather process. Although access to a working copy
than attempting to convey its content faithfully of the film is crucial for translators to verify
(Gambier 2003a). Commentaries are commonly non-visual information and make appro-
used to broadcast high-profile events with a priate decisions on aspects such as register or
spontaneous tone. Simultaneous interpreting pragmatic intention, this is not always made
is typically carried out in the context of film available to them (Hensel 1987). The translators’
festivals when time and budget constraints do participation in the dubbing process often ends
not allow for a more elaborate form of oral with the production of a dialogue list in the
or written language transfer (see conference target language; in practice, translators do not
interpreting). Interpreters may translate with concern themselves with lip movements as they
or without scripts and dub the voices of the usually lack experience in dialogue adaptation
whole cast of characters featuring in the film and adjustment techniques (Luyken et al. 1991).
(Lecuona 1994). A ‘dubbing writer’ who is adept at lip reading
(Myers 1973) but not always familiar with the
source language takes over at this point to
Lip-synchronized dubbing
‘detect’ the text. This involves identifying those
Lip-synchronized (or lip-sync) dubbing is one sounds delivered by screen actors in close-up
of the two dominant forms of film translation, shots that will require maximum synchrony on
the other being interlingual subtitling. In the the part of dubbing actors and marking their
field of audiovisual translation, dubbing denotes presence on the relevant frames of the film strip
the re-recording of the original voice track (Paquin 2001). Once the adaptation is ready,
in the target language using dubbing actors’ the film dialogue is divided into passages of
voices; the dubbed dialogue aims to recreate the dialogue, called ‘loops’ (Myers 1973) or ‘takes’
dynamics of the original, particularly in terms (Whitman-Linsen 1992), whose length depends
of delivery pace and lip movements (Luyken et on the country where the dubbed version is
al. 1991). Regarded by some as the supreme and produced. These takes become the working
most comprehensive form of translation (Cary units during the revoicing of the dialogue track,
1969), dubbing ‘requires a complex juggling which is carried out under the supervision of
of semantic content, cadence of language and a dubbing director and a sound engineer. The
technical prosody . . . while bowing to the prosaic involvement of so many professionals in the
constraints of the medium itself ’ (Whitman- dubbing process explains why this form of
Linsen 1992: 103–4). In the last three decades, audiovisual translation is up to fifteen times
there have been several attempts to map out more expensive than subtitling (Luyken et al.
the set of variables moulding this transfer 1991). The actual translation and adaptation of
method, mainly by diluting the importance of the dialogue amounts to only 10 per cent of the
lip synchrony proper within a wider range of overall cost (Dries 1995), although this depends
synchrony requirements. These new and more on the genre – with action and humour films
elaborate models of dubbing synchrony advocate being the cheapest and most expensive, respec-
the need to match other features of the original tively (Muntefering 2002).
film which contribute to characterization or
artistic idiosyncrasy (Fodor 1976; Whitman-
Advantages and limitations of lip-sync
Linsen 1992; Herbst 1994; Chaume 2004). At
dubbing
any rate, the relative weighting of lip matching
vis-à-vis other types of synchrony depends on Dubbing allows viewers to watch a film or
the target market, with American audiences, for programme without dividing their attention
example, being more demanding than Italians between the images and the written trans-
in this respect (Gambier 2003a). lation (Goris 1993). This reduces the amount
of processing effort required on the part of the
audience and makes dubbing the most effective
The lip-sync dubbing process
method to translate programmes addressed at
The translation of a source language dialogue children or viewers with a restricted degree
list is one of the earliest stages in the dubbing of literacy. In so far as dubbing is a spoken

TF17062.indb 17 26/9/08 11:06:08

 18 Audiovisual translation

translation of an oral source text, it is possible return on the investment. In some cases (e.g.
for the target text to convey more of the infor- France), the dissemination of a single dubbed
mation contained in its source counterpart. version across the length and breadth of the
Also, dubbing allows for the reproduction of national territory has been instrumental in
the original dialogue’s interactional dynamics, achieving linguistic uniformity, to the detriment
including stretches of overlapping speech and of regional dialects or minority languages
most other prosodic features. On the negative (Ballester 1995). On the other hand, the
side, dubbing is expensive and time-consuming. predominance of dubbing in Germany, Italy and
Furthermore, it tends to draw on a restricted Spain in the 1930s and 1940s was fostered by
range of voices to which viewers may become fascist regimes. Revoicing a whole film became
over-exposed over a number of years, which an effective instrument of censorship, enabling
detracts from the authenticity of the dubbed the removal of inconvenient references to facts
film. In relation to the translation process itself, and values that clashed with the official doctrine
the concern of dubbing practitioners with (Agost 1999). Voice-over, on the other hand,
synchronization and the take-based approach became the transfer method of choice in most
to the revoicing process has often resulted in a Soviet bloc countries and other Asian markets
‘compartmentalization’ of the source text. This (e.g. Thailand), either because the national
adherence to the constraints of micro-equiva- language was unchallenged (Danan 1991) or
lence often proves detrimental to the ‘naturalness’ because budget constraints made the cost of
and ‘contextual appropriateness’ of the translated lip-sync dubbing simply prohibitive (Gottlieb
dialogue (Herbst 1997; Pérez González 2007). It 1998). Subtitling, on the other hand, thrived
is also held accountable for most of the so-called in a group of rich and highly literate countries
‘universals’ of dubbed language, including its with small audiovisual markets (Scandinavian
failure to portray sociolinguistic variation and its countries) and bilingual communities (the
overall tendency towards cultural neutralization Netherlands and Belgium), as well as in other
(Pavesi 2005). The transmission of culture- states with lower literacy rates but much poorer
specific terms and values in dubbed audiovisual economies (Portugal, Greece, Iran and most
texts remains a highly problematic issue. In Arab countries), for whom other forms of
principle, the revoicing of the dialogue allows audiovisual translation were unaffordable.
for an easy domestication of the original text, Until the mid-1990s, the audiovisual market-
including the replacement of source cultural place remained divided into two major clusters:
references by their naturalizing counterparts, subtitling versus dubbing countries (Luyken
i.e. their functional equivalents in the target et al. 1991). Since then, however, we have
viewer’s cognitive environment (Chiaro 1992). witnessed a series of changes in the audio-
However, these attempts to maintain the illusion visual landscape, including the ever growing
of authenticity may backfire and damage the volume of programmes and broadcast outlets,
commercial success of the dubbed product when the development of digitization techniques
the foreign language and culture draw attention and the emergence of new patterns in the
to themselves, e.g. through poor synchroni- distribution and consumption of audiovisual
zation of mouth movements or the reliance on products (Pérez González 2006b). This has
culturally idiosyncratic visuals (Fawcett 1996). contributed to blurring the lines between the
formerly opposing camps: in any given market,
‘dominant’ or traditional forms of audiovisual
Translation in the audiovisual transfer now co-exist with other ‘challenging’
marketplace or less widespread types (Gambier 2003a). The
combined use of several established methods
Lip-synchronized dubbing, the most expensive within a single programme constitutes devel-
method of audiovisual translation, has tradi- opments that continue to contribute to the
tionally been the preferred option in countries hybridization of the media industry worldwide
with a single linguistic community – and hence (ibid.).
a large potential market to secure a sizeable

TF17062.indb 18 26/9/08 11:06:08