Navigation path

Left navigation

Additional tools

Tag ‘dubbing’

What’s in the cards for AVT – prognoses on the future of dubbing

Thursday, September 15th, 2022

By Gabriela Flis and Tomas Senda, Applied Linguistics, Institute of Applied Linguistics, University of Warsaw

Not that long ago, many papers concerning audiovisual translation started with similar sentiments, stating how it is a new and uncharted territory. While this might have been true quarter of a century ago, the AVT landscape has changed drastically, especially in the last decade or so. And with this rapid change academics, so used to neat divisions and classifications, may once again find themselves entering unknown waters. However, before going into what the future holds, one must understand the foundation on which our prognoses were built.

The history of audiovisual translation is, of course, closely intertwined with the history of cinematography itself. Although many people believe Lumières’ 1895 film Workers Leaving the Lumière Factory to be the first motion picture ever shot, the truth is it came to be seven years after Louis Le Prince shot Roundhay Garden Scene – a two second film that pictures his family walking in a garden, believed to be the oldest surviving film ever made (Guinness World Records 2022, 2021). Although both pictures were but seconds long, the rapid development of the technology during the end of the XIX century soon enabled artists to shot longer movies that contained plot. As those movies were silent, plot elements and dialogues were conveyed using intertitles – simple text boards inserted in between shots on a film roll (Ivarsson, 2004). Of course, during the silent films era, the issue of translation had a rather simple solution – it sufficed to cut out the original boards and replace them with intertitles in the target language. The AVT landscape became more complex at the end of 1920s, when first sound films began to enter cinemas and inserting intertitles ceased to be a viable translation method. Of course since sound films saw the beginning of synchronizing recorded music and speech, the idea of what we now know as dubbing quickly appeared. Although many considered it to be too complex and pricy, dubbed productions were shown in cinemas as soon as 1930s (O’Brien, 2019). Many also experimented with various methods of either displaying the text with a separate projector or later adding text to the film roll itself, which finally resulted in the creation of what we now know as subtitles. The history of voice-over is harder to trace, as it is almost exclusively treated as a mode of audiovisual translation by post-communist countries, Poland included. In Poland it came to existence in 1960s, when the only state-approved dubbing studio found it increasingly hard to meet tight dubbing deadlines and therefore turned to voice-over as an easier-to-produce alternative (Plewa, 2015). As it was also significantly cheaper, it became the default AVT mode for TV productions during 1980s, when Poland entered an economic crisis (Plewa, 2015).

Photo by Jeremy Yap on Unsplash

Different modes of audiovisual translation became popular in different countries and their development was influenced by both political and economic reasons. Based on their preferred type of AVT, countries were then classified into dubbing, subtitling and voice-over countries (Gottlieb, 1998). Dubbing countries are usually wealthy nations with big potential audiences. Perhaps the best known example are FIGS – France, Italy, Germany and Spain (Gottlieb, 1998). It is worth mentioning that dubbing countries can have a history of totalitarian regime and censorship, since dubbing, as the original voice track is completely replaced, is most prone to control and propaganda (Danan, 2002). Economic aspects may also shed a light on why voice-over, although the original is still audible, became popular in less affluent countries of the Eastern Bloc, as it was much cheaper to produce than dubbing. However, financial reasons are not to be used as a rule of thumb, as some of the widely recognized subtitling countries may be found in Scandinavia, which contains arguably wealthy states.

Although scholars became used to Gottlieb’s (1998) classification, it may soon become obsolete, if it hasn’t already, as the growing popularity of VOD (video on demand) services has changed the market forever. Although Gottlieb’s categorization was based on preference, it was truly more about what was available to the viewers. Before streaming platforms such as Netflix, Amazon Prime, HBO Go and the likes of it entered the film and series market, viewers’ choices were limited in two main ways. Setting the issue of censorship aside, audiences could only watch what was previously selected by channel officials and had no say in choosing the mode of audiovisual translation. Although VOD services lured their clients with promises of their extensive catalogues, they usually also offer a wide range of AVT modes accessibility services, from subtitles for the deaf and hard of hearing, through dubbing, to audiodescription and voice-over, depending on the country. One can also easily change the language, meaning viewers can now chose not only the content itself, but also their (truly) preferred mode of audiovisual translation and the language they want to consume the film in. A recent study shows that this abundance of choice is starting to change viewers’ habits, with younger people being four times more likely to choose subtitles, even though they have fewer hearing issues than their older counterparts (Youngs, 2021).

However, it is worth noting that the ever-growing number of productions being made seemingly all the time can also negatively impact viewers, as well as the AVT market itself. Even though there has never been so much job opportunities, with deadlines being short and rates low, skilled translators and subtitlers are fleeing the industry (Bryant, 2021). This can, in turn, result in the quality of AVT services being impacted in a negative way. The situation got so dire that it became noticed by mainstream media (see Bisset, 2022; Motamayor, 2022).

Another industry practice that may completely change the landscape is the use of artificial intelligence to create dubbing. Since the dawn of dubbing, one of the biggest challenges when it came to translating a movie, was not only to convey the original meaning (provided one was not trying to censor or obscure the original) but also to synchronize the lip movements of characters with the translated utterance. In recent years, new technology of deep fakes and neural networks allowed AI experts to manipulate already existing videos. Aside from funny videos of presidents, heads of state and other respected officials stating things well below their station, it also resulted in companies using this technology to produce “perfectly” synchronized dubbing. In 2021, Flawless AI published a short fragment from Forrest Gump dubbed into Spanish and Japanese, where Tom Hanks’s lip movements have been carefully manipulated to match the audible translation, giving the impression that the actor is at least trilingual.

Although compared to literary studies or even SLA, audiovisual translation remains a fairly young research field, the rapid technological development over the last one hundred years has brought many changes. As of 2022, thanks to the VOD platforms taking over the market, viewers are being offered more choice when it comes not only to the content itself, but also their preferred method of audiovisual translation, resulting in once clear boundaries between dubbing, subtitling and voice-over countries becoming more blurred. Unfortunately, the abundance of film and TV series, combined with short deadlines and low wages, may impact the quality of AVT services in a negative way. What is more, as the artificial intelligence and video synthesis are on the rise, the issue of synchronization may soon become a thing of the past, with dubbing translators being able to concentrate less on lip sync and more on the content, making their work much easier, but arguably also less creative and challenging, taking the craftsmanship away from the job description.


Bisset, J. (2022, February 11). Inside the dying art of subtitling. CNET.

Bryant, M. (2021, October 14). Where have all the translators gone? The Guardian.

Danan, M. (2002). Dubbing as an Expression of Nationalism. Meta, 36(4), 606–614.

Gottlieb, H. (1998). Subtitling. In M. Baker (Ed.), Routledge encyclopedia of translation studies (First edition). Routledge.

Guinness world records 2022. (2021). Guinness World Records Limited.

Ivarsson, J. (2004). A Short Technical History of Subtitles in Europe. Transedit.

Motamayor, R. (2022, May 31). Every Language Everywhere All at Once. Vulture.

O’Brien, C. (2019). Dubbing in the early 1930s: An improbable policy. In C. O’Brien, The Translation of Films, 1900-1950 (pp. 177–190). British Academy.

Plewa, E. (2015). Układy translacji audiowizualnych. Wydawnictwo Naukowe Instytutu Komunikacji Specjalistycznej i Interkulturowej Uniwersytet Warszawski. Youngs, I. (2021, November 15). Young viewers prefer TV subtitles, research suggests. BBC.

Gained in translation: why mist can sometimes turn into rubbish

Tuesday, September 22nd, 2020

By Pr Oliver Czulo, EMT representative M.A. Translatologie, University of Leipzig

Video excerpts: Star Trek: The Next Generation, Season 4, Episode 5 (English/German), 1990, (c) Paramount/CBS

Dr Beverly Crusher is not to be envied: First, one by one, the crew members of the starship Enterprise D are disappearing, and then, it’s not even possible to call out to other ships for help. When Dr Crusher orders the ‘viewscreen’ to be switched on, all she can see is a dimly lit … misty something. She calls out: ‘Computer, what is that mist I am seeing?’

On the surface, a very simple task for a translator: no complex linguistic structures, no ambiguities, no unanswered questions. But, wait! Apparently, it all went wrong, at least according to numerous comments from the German fan community. [Tip: The web pages are in German; the comments mentioned can be found by searching for ‘Mist’.] So, what went wrong? Well, here is the German dubbed version:

‘Computer, was ist das für ein Mist, den ich da sehe!’

She says something like mist here, doesn’t she? She absolutely does – but Mist in German means ‘dung’, or, in this context, rather ‘rubbish, garbage’. Mist should be Nebel in German, shouldn’t it? It’s clear to everyone: The translator must have made a grave mistake!


The following thoughts have arisen out of various discussions with colleagues and students. The purpose in laying them out is not to sort out what really happened, but to explore the kind of reasoning that could have led to the ultimate decision for the German dubbed version.

Formal factors

One thing not to lose out of sight when translating for dubbing is the need for at least a certain degree of lip synchronicity. When a pairing such as mist/Mist presents itself, involving two (almost) identically pronounced words formed with pretty much the same lip movement, then this is a strong formal argument for using the paired word, providing that the choice will be semantically plausible in the given context.

In the short YouTube video linked above, lip movement may not seem important, given the camera perspective, but in a dubbing studio with a large screen and high resolution, the situation may look quite different.

Emotional factors

The emotional state of the character is also relevant in the situation. Dr Crusher is clearly frustrated. All the people she works with on a daily basis have disappeared, and the universe is about to collapse. She cannot explain what is happening and the viewscreen is not providing any useful information as to the whereabouts of the ship.

In such a scene, mist can be an inspiration for the German Mist, which is not only lip-synchronous, but which better transports the emotional state of Dr Crusher than would the use of the default translation, Nebel. And, as the ‘mist’ has no further relevance for the development of the story, the parameter space shifts towards other translation options.

Relationship between text and image

Not only does the ‘mist’ play no role in the rest of the story, but it is not even easily discernible as such. Whatever the viewscreen is showing, it would be hard to identify it if it were not clearly named: You could just as well ask what that strange light is that is coming from that one corner.

The English variant shifts the focus (away from the light) and pre-empts an interpretation (the pale light is a nebula), whereas in the German variant, no perspective is construed. But what becomes more apparent in German is that whatever the viewscreen is showing, it is not what Dr Crusher expected – after all, when the viewscreen allows a glimpse outside the Enterprise, both she and the audience are usually looking at either a field of stars or the crew of some other starship.

Translation or dubbing decision?

Last but not least, we have to ask whether the shift from mist to Mist happened during translation or maybe somewhere else: in the dubbing studio. Did the translation initially contain the variant Nebel instead of Mist? Was there a problem with the timing, was a shorter variant (one vs. two syllables) necessary? Or, did the translated sentence just not fit the character, who had been dubbed by the same person for years, during which time the concept for the character had evolved? Was the English original an inspiration and mist sounded so good that, after a brief discussion, Mist was chosen instead of Nebel? We may never find out, but the possibility exists.

All but normative

The image many people have of translation is that it simply involves the advanced use of dictionaries. Of course, dictionaries are indeed an important tool for translators, but unfortunately, this narrow view results in a very normative stance on translation, as witnessed by the never-ending topos of the ‘correct’ translation. While there may in fact be misleading, incorrect and simply wrong translations out there, the spectrum extending in the other direction is very broad. After all, what is a correct translation? One that is based on the first option given in some dictionary, or one that is (at the least) just as entertaining, informative, etc. as the original? If it had not been for the existence of the original, would anyone have noticed a ‘problem’ with the Mist variant in German?

‘Lost in translation’ is a common phrase, but ‘gained in translation’ should become just as common. Translation enriches our culture, technology and economy in many ways, by bringing in new ideas, by enabling collaboration, etc. And, from a less lofty perspective, translators are sometimes able to reveal previously unrecognized nuances. The example discussed here should provide for an excellent case-in-point.

My thanks go out to Jonatan Jalle Steller for creating the video excerpts and to Cynthia Dyre for revising the English version.

This blog entry is largely based on a German entry which first appeared in the LingDrafts blog:


EMT map 2019-2024