Abstract

This article focuses on visual adaptation strategies and practices in the publication of foreign comics. It first defines visual adaptation in the context of related terms such as audiovisual translation (AVT) and localization. Then, it provides a typology of visual adaptation strategies, illustrated with examples taken from a variety of comics. This is followed by a description of visual adaptation practices in comics translation industry as concerns two comic types which have the largest volume of foreign editions, American superhero comics and Japanese manga. In the case of manga translation, after an initial stage in which Japanese comics were heavily adapted lo local comics visual conventions, market pressure in the form of fan base expectations has met with economic concerns by the publishers. Foreignization strategies (i.e. little visual adaptation) have now superseded domestication strategies (i.e. much visual adaptation) as the general norm for localized manga. In the case of Marvel comics, the technological implementation of adaptation strategies allows for the production of multiple foreign editions tailored to local markets. Foreignization strategies, which privilege non-adaptation over adaptation to target visual conventions, innovate the aesthetic conventions and pictorial repertoires of the target comics polysystem.

Keywords: comics, audiovisual translation, traduzione multimediale, literary translation, comics translation, visual adaptation, comics localization, lettering, graphic editing, images, manga, US mainstream comics, superheroes, translation strategies, translation practices, comics polysystem

1. AVT, localization, and visual adaptation^[1]

Comics are narrative visual texts which usually incorporate verbal content.[2] Since comics are multichannel rather than only verbal texts, the study of translated comics is usually seen as part of audiovisual translation (AVT), the area of translation studies which deals with “the transfer from one language to another of the verbal components contained in audiovisual works and products” (Chiaro 2012: 1050). According to this definition, AVT concerns itself both with screen translation practices such as dubbing, subtitling and respeaking, and with multimodal types of translation such as translation for the theatre and the advertising industry. Comics in translation are  seen as a type of AVT since, it is argued, the translation of comics involves constraints similar to those of dubbing and subtitling (Chiaro 2009: 142).

In the translation studies literature, localization is also often filed under audiovisual translation. However, AVT (including comics translation) could also be seen through the lens of localization, as many of the translation activities covered by the term AVT increasingly involve practices which characterize a localization approach.

Localization, defined as “taking a product and making it linguistically and culturally appropriate to the target locale (country/region and language) where it will be used” (Esselink 2000: 3), is now the term generally used by the language industry to refer to the “translation” of all computer related products, including software and websites, videogames, and mobile applications.

In the localization industry, translation is seen as just one element in a chain of processes, a stage in a larger project carried out by a team, which goes from the managing of the project itself to the adaptation of cultural conventions and hard-coded (audio)visual elements such as, for instance, date and number format in the case of computer programs and soundtrack in the case of videogames. Pym (2010) argues that while not being strictly speaking a theory of translation, localization provides a set of new terms and concepts which can be useful in the study of translated products. First, the  definition of localization stresses the reference to a product rather than a text, and to a locale (a combination of language and regional location) over just language. Second, localization is usually preceded by internationalization which means that an intermediate locale-neutral version of the product to be localized is prepared, while translation directionality goes from one-to-many rather than from one-to-one. The concepts are sometimes subsumed under the term globalization and the full process is referred to by the acronym GILT, Globalization, Internationalization, Localization, Translation.

According to Pym the localization paradigm[3] can arguably be extended to the analysis of other types of text transformation processes. For instance, Pym discusses how in film translation/localization an intermediate script is especially prepared for translators of local versions (dubbed and subtitled). This internationalized script includes instructions about culture specific elements and cross references. Orengo (2005, see also Pym 2004), argues that also the production of international news can be described in terms of localization, since “the process of adapting a text from a news agency release for a newspaper story corresponds to its passage from a global product to a localised version in the local press” (Orengo 2005: 175). O’Hagan & Mangiron (2013: 106) suggest that “the emergence of new media resulting from the convergence of technologies is seeing the previously separate domains of localization and AVT come together to cater for the new type of products needing to be prepared to go global. Whether AVT subsumes localization or viceversa remains to be seen”.

A localization view of translation is, however, not always well received within the translation studies community since the term is perceived as just a fancy synonym used by the computer industry to refer to what is otherwise “simply a commonly accepted definition of translation itself” (Hartley 2009: 107). At the same time the view of translation held by the localization industry is perceived as reductive, in that it tends to consider translation simply as the replacement of decontextualized textual strings (Pym 2010).

In this article, I use the term comics localization as subsuming both “translation” and “visual adaptation”[4], to underline the similarities between the processes and practices involved in the production of a foreign edition of a comic book with those foregrounded in the definition of localization (see also Zanettin 2008b, 2009, 2011). Within comics localization, translation refers to the verbal (written) text which is produced by a translator in order to replace the source language verbal text, while by “visual adaptation” I refer to all changes made to the publication format, layout, pictures (including lettering, see below) and in general all elements of a localized comic book except for the verbal content.

This definition of translation as only one part of the comics localization process is not meant to imply the acceptance of a reductive view of translation, and of translators as only concerned with the verbal aspects of comics in isolation from the visual context. Rather, competent translators of comics should be seen as “semiotic investigators” (Celotti 2008), skilled readers of the medium who are aware that meaning in comics is created by relationships of complementarity and dialogue between verbal and non verbal messages, and for whom the visual context constitutes an opportunity rather than a constraint. However, while translators are perhaps in a position to act as “paratranslators” (Yuste Frías 2010), [5] able to anticipate and perhaps also have an influence in orienting the visual appearance of the localized product, they are not the only agents responsible for the production of a foreign edition, which may involve editing or removing images, adding/removing/altering colors, changing layout, size and pagination, etc., as will be detailed in the next section.

It seems important to adopt a distinction made by foreign edition comic collectors,[6] who point out that a localized version of a comic product first published in a different country is a re-issue rather than a re-print. While a re-print is “a second publication of exactly the same material” in order to meet “capacity demand of distribution channel sales”, a re-issue implies some kind of transformation, as regards packaging but also the visual and sometimes the verbal content of a comic (Zanettin 2008b). Foreign editions are “first publication[s] of any licensed material outside [the] country of origin”, and may be followed by subsequent re-prints or re-issues. Foreign editions may be in the same language as that of the country of original edition, as happens, for example, with US comics published in English in the UK, Canada, India and South Africa, and with Argentinian comics published in Spanish in Mexico and Spain. Foreign editions, however, usually imply translation i.e. that “lettering is translated into [the] national language of [the] region”, as one foreign edition collector put it in a forum message. Collectors of foreign comics insist that foreign editions are not simply reprints of original comics in a foreign language, since foreign editions involve a repackaging of the text, and often changes and alternations to cover and inside drawings, graphic elements, coloring, lettering, inking, format dimensions, etc., or, as another forum user put it, “re-imagining” the text. Furthermore, they point out that licensed (as well as unlicensed) material in which the “original” drawings and story has been extensively changed, redrawn, recolored, rearranged, etc. is often published as “foreign edition”, even though it could well have a claim as an autochthonous product. For instance, in an Brazilian edition of Marvel Super Heroes Secret Wars (1984-85), some characters were expunged or redrawn, possibly for reasons of marketing or copyright,^[7] while many Disney stories and characters are first created and published in Italy and elsewhere in Europe and South America as licensed material (Zanettin 2008b).

A descriptive analysis of comics in translation has to take into consideration all elements which may be subject to change when a comic book is published in foreign edition. As Kaindl (2010) argues

The various techniques involved in designing comics, ranging from various linguistic elements such as text in speech bubbles, narrative texts, onomatopoeia and captions, to typographic elements, pictographic elements such as speedlines, ideograms such as stars, flowers etc., and pictorial representations of persons, objects and situations, are all integral to the constitution of the meaning - and therefore translation-relevant. (Kaindl 2010:36)

Kaindl (1999) distinguishes among three groups of signs in comics: pictorial, typographical and linguistic, this last category including five functional linguistic categories, i.e. titles, dialogue, narration, inscription and onomatopoeia. Drawing on Delabastita’s (1989) work on film translation, he lists six strategies which can be used to translate these signs in comics, i.e. ‘repetitio’, ‘deletio’, ‘detractio’, ‘adiectio’, ‘transmutatio’, and ‘substitutio’ (1999: 275). Celotti (2008) distinguishes among four “loci of translation”, where verbal messages appear with a different function, i.e. balloons, captions, titles and linguistic paratext (verbal signs outside the balloon), and illustrates how translators may use four different strategies i.e. ‘translation’ (Delabastita’s ‘substitutio’, the default option for the first three loci), translation with a footnote, cultural adaptation, non-translation (Delabastita’s ‘repetitio’), deletion, and a mix of all the former.

This article focuses on visual adaptation in translated comics, that is on changes to the appearance of translated comics implemented through visual manipulation strategies, which include both of Kaindl’s pictorial and typographical categories, as well as other elements which characterize the re-packaging of foreign comics. I consider typographic signs only as concerns their visual appearance (the lettering), not as regards their verbal contents. Similarly to Kaindl (1999, 2010) and Celotti (2008), in the next section I describe a series of changes which may take place in foreign editions of comics with respect to first publication.[8] The overview of visual adaptation strategies will show that, though the default option is ‘non-translation,’ or ‘repetitio’, all other visual adaptation strategies can and have been employed.

2. A typology of visual adaptation strategies

Visual adaptation may involve changes in publication format, coloring and the drawings themselves, including the appearance of the verbal content. A change in publication format may involve considerably altering the size, shape and/or page layout of a comic book, as well as the type and quality of paper. It may involve the replacement of the cover page with a new one perhaps featuring a drawing from a different artist, as well as adding, replacing or deleting pages. Changes to the coloring scheme and the pictures may affect single pages or panels, or be applied at a global level.

A comic book typically consists in a sequence of pages each subdivided into a sequence of panels, arranged in a more or less regular grid. Each panel is a “frame”, an image from the story, in which visual and verbal elements coexist and interact. Groensteen (2007: 9) argues that in comics “the apparent irreducibility of the image and the story is dialectically resolved through the play of successive images and through their coexistence, through their diegetic connections, and through their panoptic display”. This dialectic applies first at the level of the page, and then at the level of the panels within the page: before looking at the panels in their narrative sequence and before reading the verbal and non verbal signs within each panel, the reader experiences each page as a whole image and in relation to the preceding and following pages. As comics are seen before they are read, the interpretation of verbal signs is intimately tied to understanding their wider visual context and co-text: words not only co-exist with drawings, but are also part of them, as the size, shape and arrangement of the verbal text, both within balloons and boxes (dialogues and narrations) and outside them (“stage scripts”, visual representations of sounds, graphic use of characters and words) is, together with its content, part of the message conveyed by the words.

Within a page, adaptation may involve resizing, deleting or adding panels; within a panel balloons and boxes may also be resized, deleted or added, and the size, shape and arrangement of the letters in which the verbal text is written may be changed. The following examples, mainly referring to the Italian context, provide an overview of visual adaptation strategies, starting from publication format and concluding with lettering.  

2.1 Changing publication format

Foreign editions often involve changes at more than one level, as can be seen in translated comics going back a century. Some of Winsor McCay’s Little Nemo in Slumberland Sunday pages, originally published in the US from 1905 to 1914, appeared in Italy in the children’s weekly Il corriere dei piccoli (Zanettin 2007). The new format implied by the change of publication type (magazine rather than newspaper) and target readers (children rather than both adults and children) brought with it the adaptation to local cultural and genre conventions: a change in page layout, the addition of panels, the deletion of balloons, as well as changes in lettering, coloring and drawings. Figure 1 shows the first strip of a Sunday story as published in The New York Herald, while Figure 2 shows how the last two panels of the original story were adapted to fit the layout of the children’s magazine. Most notably, the title and the balloons were deleted and the background was redrawn and recolored. The dialogues in the balloons were replaced by rhymed captions underneath the panels, following Italian cultural conventions for ‘drawn stories’.

Figure 1: Little Nemo in Slumberland, by Winsor McCay

Figure 2: Little Nemo in Slumberland, Italian adapation, Il Corriere dei Piccoli.

When comic strips are published in magazines, journals and books as collections, or when French albums, American comic books, Italian pocket book series are republished in a different format in translation (Zanettin 2008b, Rota 2008), a change in publication format is sometimes accompanied by a change in page size and layout. Even a simple change in paper size may affect the translation, not only because of the different visual perception of the images, but also because when the size of the page is shrunk, the size of balloons and boxes is reduced as well. Thus, when comics are republished in translation in a smaller format the textual material in the translation may need to be less than that in the source text (or, conversely, the balloons need to be enlarged), in order to allow readability.

2.2 Changing the colors

Sometimes a comic book in color is printed in black and white in translation, and viceversa. For instance, when Alan Moore’s The Swamp Thing saga was published in Italy as a “cult” product rather than as part of the regular DC series, it was published in black and white rather than in color (D’Arcangelo 2008). On the other hand, the first American and European editions of Kazushiro Otomo’s Akira were published in color rather than in the original black and white (Jüngst 2008; Malone, 2010: 319).

Colors can also be considerably altered in foreign editions. For instance, in an Italian 1970 edition of Charlier and Giraud’s La piste del Navajos (Figure 3), which targeted a younger readership, the bright colors used for the background are strikingly different from the softer tones of the French publication (Zanettin 2008b). Figure 4 shows a later Italian edition which reproduces the colors of the original French one.

Figure 3: La piste des Navajos, Italian edition, Crespi
© Charlier et Giraud, Dargaud 2006

Figure 4: La piste des Navajos, Italian edition, Nuova Frontiera
© Charlier et Giraud, Dargaud 2006

2.3 Changing the drawings

Figure 5 shows a page from an episode of the Italian popular monthly series Dylan Dog, while Figure 6 shows the translation published in 1999 by Dark Horse Press, a major player in the US comic publishing industry. Two differences are noticeable at first sight: first, the amount of written text in the American edition is less than that in the Italian version, leaving more white space in the balloons. This can be seen as an adaptation to US comics reading pace and conventions, which privilege action over dialogue. Second, the character sitting at the centre of the second panel in Figure 6 does not wear moustaches, and this is because some images of Groucho Marx are protected by copyright in the US. Accordingly, the name of the character was changed from Groucho into Felix.

Figure 5: Dylan Dog, by Tiziano Sclavi
© Sergio Bonelli Editore

Figure 6: Dylan Dog, US edition, Dark Horse
© Sergio Bonelli Editore

This difference in the way Groucho/Felix is drawn may at times lead to some confusion. For instance, while in the Italian sequence in Figure 7 it is apparent that the three mouths in close-up belong to three different characters, the identity of the characters is less clear in the American translation in Figure 8.

Figure 7: Dylan Dog, by Tiziano Sclavi
© Sergio Bonelli Editore

Figure 8: Dylan Dog, US edition, Dark Horse
© Sergio Bonelli Editore

Figure 9 shows a panel from a Disney story published in Kuwait in which it can be noticed that the thighs of the Tarzan character were covered with culturally appropriate garments.

Figure 9: Brad of the Jungle
© Disney

Figure 10 shows a panel from a Donald Duck story published in the United Arab Emirates. As can be seen in, the body of a woman wearing a bikini suit has been almost completely blackened in the adapted edition, in compliance with the guidelines of the Saudi Ministry of Information (Zitawi 2004, 2008a, 2008b).[9]

Figure 10: a panel from a Donald Duck story
© Disney

Another example of a change to the drawing is the transformation of a swastika into a less politically loaded symbol in the German edition of an American comic book (see Figure 36).

2.4 Changing page layout

Hugo Pratt’s La ballata del mare salato (Ballad of the Salt Sea), was first published in Italy in black and white in magazine instalments in 1967, and in the course of the years as a volume in various editions and different formats, in colors and in black and white. In the US, it was published by The Harvill Press in 1996 in black and white in “album” format (slightly smaller than A4 paper size), and in 2012 by Universe, a division of the Italian publishing house Rizzoli. The 2012 US edition was based on an Italian edition in the smaller “graphic novel” format[10] published by Rizzoli–Lizard (Migliori 1999), but as opposed to the Italian edition, it is in color. Because of the change in publication format, the page layout was redesigned so that each page contains three rather than four strips, as originally conceived by Pratt. To this end, on the one hand the images were blown up, on the other the length of the strips was shortened. To make them fit the new grid some panels were repositioned, split, cut out or enlarged. For instance, the first panel in the opening page of the story (Figure 11) has been reshaped by chopping out part of the image on both sides, and adding a stripe of white background at the top (Figure 12). The written text is accordingly adapted to the new shape. The remaining two strips are broken down, so that three panels slide to the following page. The panel which concluded the first page is now found in the middle of the second page. A third of it has been cut out and it is juxtaposed to a panel which was originally the first panel of the second page. Incidentally, this panel contains a reverse angle shot, showing the face of the character which was seen from the back before turning the page, thus spoiling the visual climax.

Figure 11: A ballad of the salt sea, by Hugo Pratt, Hanvill
© 1967 CONG sa - Switzerland

Figure 12: A ballad of the salt sea, by Hugo Pratt, Universe
© 1967 CONG sa - Switzerland

Other examples of this “destruction of a comics classic” (Jared 2012) include the breaking down of a sequence representing a fight inside a hut (Figure 13 and Figure 14) so that “instead of the natural end to the fight at the bottom of the page, the fight seems to continue too long, dragging onto the next page. The final two panels seem a waste of page space and reading time.” (Jared, 2012: online).

Figure 13: A ballad of the salt sea, by Hugo Pratt, Hanville
© Cong SA

Figure 14: A ballad of the salt sea, by Hugo Pratt, Universe
© Cong SA

A change in page layout is also involved in the publication of manga in Western left-to-right reading direction (see below), whereas by flipping the pages not only most characters notoriously become left handed, but the composition of the page is altogether altered (see Barbieri 2004, Zanettin 2008a). In some cases, a change in reading direction may be effected by flipping individual panels rather than the whole page (Rota 2008, Zitawi 2008a).

2.5 Replacing, deleting or adding pages

An example of page replacement comes from a comparison between page 56 of the Tintin in Congo by Hergé in French[11] and two Italian translations, both published in the last decade: while the French edition is still as it was when it was first published in 1946, the Italian edition (Figure 16) is derived from the Swedish edition of 1978. This happened because, when the album was to be published in Sweden, the publishers objected to the scene depicting Tintin blowing up a rhinoceros with a stick of dynamite (Figure 15). They asked that the page be redrawn, and Hergé complied.

Figure 15: Tintin au Congo, by Hergé
© Hergé/Moulinsart 2014

Figure 16: Tintin in Congo, Lizard Edizioni
© Hergé/Moulinsart 2014

Since then there are two versions, the French one and the “international” one. A more recent republication of the adventures of Tintin in Italy by Rizzoli-Lizard (2011) reinstates the original page. This edition reprints all the original 23 French albums in the smaller graphic novel format, putting together three albums in each volume, and is framed as a more ‘philological’ publication for adult readers through a prefatory note by the popular art critic and TV presenter Philippe Daverio.

A rather different strategy was followed in a 1970 Italian translation of La Piste des Navajos, an episode of the French western Blueberry saga by Charlier and Giraud (see Zanettin 2008b). The original 48 page story was in fact shortened to 46 pages, to make room for advertising (see below).

2.6 Resizing, deleting, replacing or adding panels

The Blueberry story just mentioned could not of course be simply depleted of two random pages. Thus, a few panels were deleted from two consecutive pages, and the remaining panels were re-combined into a new page. However, while this adjustment prevented a complete disruption of the narrative flaw such as would have resulted from simply taking out two full pages, the adaptation still generates an incongruous transition in the translated narrative (Zanettin 2008b).

An example involving the replacement of panels comes from the Disney story published in Kuwait, in which not only the drawings were re-touched (see Figure 9 and  Figure10), but also sometimes created anew. Figure 17 shows a page from the source story, while Figure 18 shows how one panel was replaced with a new one created by merging drawings from two adjacent panels, to avoid showing a kiss between Clarabelle Cow and a monkey. Some Arabic text was also added in the new panel.

Figure 17: Brad of the Jungle
© Disney

Figure 18: Brad of the Jungle, Kwaiti edition
© Disney

2.7 Resizing, deleting or adding balloons and boxes

Changing the appearance of the physical spaces which contain dialogues and narratives implies a change in the drawings, which can be partly covered by the resized balloons (sees Figure 25 and 26). Balloons or boxes may be added to insert additional text provided by the translator, for instance a note explaining a narrative link to an episode of a series yet unpublished in translation. Balloons may also be deleted, for instance when no direct “equivalent” exists in the translated comics culture. This happens, for instance, with the Japanese visual rendering of the sound of “silence”, which is sometimes omitted in foreign editions together with the balloon.

An example or resizing a balloon comes from an American comic book translated into Italian, an episode from the epic fantasy Bone by Jeff Smith (Figure19). In the Italian publication (Figure 20), the smaller monster creature was completely erased to make room for a larger balloon. Accordingly the words in the balloon are uttered by the bigger monster creature. By looking at the larger narrative context it is not clear why this change was effected, as both versions work. The change appears to have been made only to accommodate for a longer text in Italian.

Figure 19: Bone, by Jeff Smith
© Jeff Smith

Figure 20: Bone, Italian edition
© Jeff Smith

2.8 Changing the characters (lettering)

Celotti’s (2008) four loci of translation -balloons, captions, titles and the linguistic paratext- are the physical spaces where the content, size and shape of characters change in translation. The placing of verbal signs in these loci requires graphical skills and is materially carried out by a letterer/graphic editor. A distinction can be made between lettering, which involves the replacement of text in regular type font inside balloons and boxes, and other types of visual adaptation which involve retouching the drawings to replace titles and verbal signs which are part of the visual paratext, such as inscriptions, road signs, newspapers, sound effects, onomatopoeic and unarticulated sounds, etc.

Calligraphy and typography play an important role in the visual reading experience, and comics creators make use of both conventional techniques such as using a bigger character size to represent shouting, or stereotyped fonts to represent national provenance (e.g. gothic font type for Germans), and of more unconventional and creative effects. In translation, lettering can be carried out manually or electronically, and the resulting lettered text can approximate to a larger or smaller extent the visual impact of the source image. Thus, lettering can have a considerable effect on how a translated comic is read and received.

Figures 21 and 22 illustrate how lettering affects two different Italian translations of a page from Will Eisner’s graphic novel A contract with God. The two editions, which also differ as regards the verbal content, use a different font size and shape. Figure 21 contains the text hand-lettered by Will Eisner, in which the characters are drawn with raindrops dripping from them. Figure 22 shows one Italian edition (Punto Zero 2001) which has been lettered using regular font types. Figure 23 shows a later Italian edition (Fandango 2009) which is hand-lettered and thus much closer to the American original.

Figure 21: A contract with God, by Will Eisner
© Will Eisner Studios Inc.

Figure 22: A contract with God, Italian edition, Punto Zero
© Will Eisner Studios Inc.

Figure 23: A contract with God, Italian edition, Fandango
© Will Eisner Studios Inc.

3. Visual adaptation practices

Generally speaking, we can distinguish several stages in the production of translated comics, after the negotiation of publication rights with the foreign publisher and before a book is sent to the printer. First, a translator (in-house or free-lance) provides the verbal text which is meant to replace the content of balloons and boxes. The translation is then reviewed by a text editor and passed on to a letterer/graphic editor, who is responsible for visual adaptation and who, if needed, may make further changes to the text. Depending on the publisher (and on the size and budget of a translation project), the distinct aspects of visual adaptation (lettering, graphic design, pagination, page layout, typesetting) can be carried out by one or more people, in-house or outsourcing them to a graphic studio. The process may differ from country to country, with some countries still resorting to manual lettering and graphic editing to a more or less extent.

Comics localization practices and processes have changed considerably around the turn of the century, and all translated comics are now processed using desktop publishing and photo editing software. However, the type and origin of the product translated largely determines which visual adaptation practices are implemented. The largest volume of translation in Europe is generated by Japanese comics (manga), and by US mainstream comics, mostly of the superhero adventure type by publishers such Marvel/Disney and DC.

3.1 Manga

Though some manga were also published in foreign edition in the 1980s, the market for Japanese comics begun to take off only in the early 1990s, becoming increasingly more important in Western comics publishing.[12] Currently, translated manga are the single largest sector, representing between one-third and one-half of all comics titles published in Europe, the US and Australia (Rampant 2010, Bouissou et al. 2010, Goldberg 2010, Malone 2010). As opposed to superhero comics for which US publishers provide licensed products in electronic format (see below), foreign editions of Japanese comics are often published by working from images scanned from a copy of an original Japanese printed copy. Thus the letterer/graphic editor, after receiving the file with the translation (with additional notes and explanations by the translator or publisher if necessary and envisaged) has to first delete the content of balloons, then replace it with transparent boxes containing the translation pasted from the translation file received. All other types of editing, including retouching balloons, drawings and the linguistic paratext are carried out by manually editing the images, at considerable expense of time and effort, as well as production costs.

At first, translated manga were heavily adapted to target cultural conventions, in order to make them as similar as possible to Western comic books. The reading direction, right to left in Japan, was reversed, balloons, sound effects, onomatopoeia and inscriptions were redrawn and the stories were colored (Jüngst 2008, Rampant 2010). Figures 24, 25 and 26 illustrate an example from the late 1990s, when this norm was prevailing in manga localization practices. The example comes from a Japanese version of an American comic, namely Spider-Man: The Manga, published in Japan from January 1970 to September 1971.[13] Both the American and Italian translations, published almost simultaneously more than 25 years later (in 1997 and 1998, respectively), in turn adapt the Japanese comic to the prevailing Western reading conventions for manga of the time, that is printed left to right and editing the drawings in order to replace Japanese paratextual verbal signs. The onomatopoeia in the third and fourth panel have been translated differently and rendered with different lettering style in the English and in Italian editions. The Japanese text in the panels in the centre strip has been replaced by narration in text boxes on a white background, which cover part of the drawings. In the American translation, the white text boxes clutter the view and hide significant pictorial details. The Italian adaptation is somewhat less obtrusive, with smaller boxes, and some partial reconstruction of the background.

Figure 24: Spider-Man: The Manga, Japanese version
© Kodansha

Figure 25: Spider-Man: The Manga, American edition
© Kodansha

Figure 26: Spider-Man: The Manga, Italian edition
© Kodansha

Later on, manga started to appear in the original black and white format, but still in reversed reading format and with considerable visual adaptations.

At the turn of the millennium, manga started to be printed in the original right to left reading direction. Currently most manga are printed right to left and with very little visual adaptation. Japanese onomatopoeia is usually left untouched, while Japanese characters are accompanied by a translation in small print, as can be seen in Figure 27 and Figure 28.

Figure 27: Hokusai, Italian edition
© Kadokawa Shoten

Figure 28: Pluto, Italian edition
© Tezuka Productions, Noaki Urasawa and Takashi Nagasaki

A variety of reasons can be brought to explain this shift in visual adaptation norms. According to a number of scholars (Jüngst 2008, O’Hagan 2009, Rampant 2010), the most important factor was the pressure exerted by the market through the unofficial, non-commercial translation practice known as scanlation. Since the advancement of the Internet in the 1990s, manga fan groups started to get hold of unpublished (in the West) Japanese series from their counterparts in Japan. Some of them organized into teams of “translators, editors, photomanipulators (who place the text onto the image), and scanners who digitally scan the original comics” (Rampant 2010: 236) and distributed the translated manga in digital format. Manga readers liked the experience of the foreign, they wanted their manga “to look as ‘Japanese’ as possible” or even “more Japanese than the original” (Jüngst 2008: 74). Thus, scanlation practices favouring ‘formal equivalence’, i.e. unchanged visual appearance, created market expectations as to this type of localization strategy. According to Rampant (2010: 231) “the advent of scanlation has shown how the advent of a translation group outside the publisher’s market ... can change the norms of manga translators within that market”. Whereas initially the prevailing norms favoured domestication, manga publishers have adopted a foreignizing translation strategy “because of scanlators... fan and consumer pressure” (ibid.). However, technical and commercial considerations may also have been at a premium, as suggested by Andrea Baricordi (2012).[14] One of the first manga to lead the change was Akira Toriyama’s Dragon Ball, at the explicit request of the Japanese rights holders, who required it to be printed in the original reading direction (Bouissou et al. 2010). Dragon Ball was an unsurpassed best seller in the mid-nineties and it is thus certain to have exerted a considerable influence on how manga titles were published in its wake. Certainly publishers did not object to saving the cost of flipping images.

3.2 US mainstream comics

According to Andrea Baricordi (2012), Japanese publishers had, at least initially, little interest in foreign editions, as translations represented for them a rather small sector of the market. US mainstream publishers such as DC and Marvel adopted quite a different approach to foreign markets.

The switch to digitally applied color and digital lettering began in the 1990s, affecting the production of original as well as of foreign editions. Superhero comics are traditionally produced by a team which includes a script writer, a penciller, an inker, a colorist and a letterer. With the advent of computers, the inked story was digitally colored and lettered, and this brought about a noticeable change in the characters of comics “by adding a wider range of color in finer gradations, thereby allowing artists to model forms through color rather than through cross-hatching in black” (Petersen 2010: 228).

According to Peterson, however,

[t]he largest significant improvement with digital production of comics has been the way it is now possible to digitally move and adjust speech bubbles. This capability has significantly enhanced the reading quality of digital comics and become extremely useful in translating comics because the size of the emanata can be changed to suit the needs of the language. Such flexibility has greatly reduced the cost of producing translations and dramatically expanded the diversity of titles in different languages.

When preparing a foreign edition, the letterer/graphic editor receives a set of files which include the translation already edited, the source files to be localized, and a copy of the original comic book, either in print or in electronic format. The comic book contains numbered references to the textual items to be replaced in the translated publication, which serve as signposts for the letterer/graphic editor. Figure 29 shows an example from an electronic copy of a page of a Spider-Man story.

Figure 29: Ultimate Comics: Spider-Man
© Marvel Comics

Source files are effectively processed as three different layers. A first layer contains only the drawing, and is stored as a digital image in bitmap file format (e.g. TIFF or BMP). Figure 30 shows the bitmap image of a page from a PunisherMAX story.

Figure 30: PunisherMAX, the image
© Marvel Comics

A second layer contains empty balloons and boxes, paratextual verbal signs (onomatopoeia and sounds effects) and titles (Figure 31).

Figure 31: PunisherMAX: the balloons
© Marvel Comics

The empty balloons are overlaid on the underlying image. Figure 32 shows the two layers combined in the Punisher story.

Figure 32: PunisherMAX: balloons and drawings
© Marvel Comics

This second layer is stored in graphic files in a vector format called EPS, which allows to resize the image without loss of definition. This allows the letterer/graphic editor to move or resize balloons and boxes in order to accommodate the text. This layer can also be used to add balloons or boxes containing, for instance, translator’s notes. Titles and paratextual verbal signs such as sound effects and onomatopoeia can also be modified as necessary. To save time and efforts pre-designed word art can sometimes be used to replace verbal signs, but often titles and other translated information are re-created from scratch. Figures 33 and 34 show the American edition of a DC Aquaman story and its German translation.

Figure 33: Aquaman, titles
© DC Comics

Figure 34: Aquaman, German titles
© DC Comics

A third layer contains the written text, and is in its turn overlaid on the underlying image containing balloons and boxes. The letterer/graphic editor deletes the text in the original language from the text layer, and replaces it with the text from the file provided by the translator, placing it over the balloons and boxes in the second layer. The text is assigned a font type and size, centered in the balloon, hyphenated, and then adapted to the size and shape of the balloons and boxes. If the length of the translated text exceeds the space of the balloon, the letterer can either reduce the size of the characters or enlarge the balloons (or, as it sometimes happens, shorten the translation).

Typographic fonts are either selected from those commercially available or created anew by the letterer him or herself. In some cases, for instance for the Italian translation of Will Eisner’s graphic novels, lettering may involve the sampling and re-using of characters from the original work, or even the creation of new characters from scratch (Figure 15). Some font types specifically designed for comics lettering aim at replicating as far as possible the irregular style of hand-written characters by providing a set of different realizations for each letter, based on letter combination (Ficarra 2012).

American superhero comics are usually printed using the CMYK color definition, which is based on the combination of four colors, Cyan, Magenta, Yellow and Key black. In printing the first three basic colors are aligned, or keyed, to the black key plate. In the CMYK subtractive color model, white is the natural color of the paper background, while black results from a full combination of colored inks. Black is used to save money on ink, and to produce deeper black tones.

Lettering proper (i.e. inside the balloons) affects only the black color plate, so that in the case of co-productions, that is, the simultaneous production of different language versions of the same comic book, production costs may be considerably lowered by using the same cyan, magenta and yellow plates for all editions, while only the black color plate containing the lettered translations is changed for the different language editions. Figure 35 shows the German translation text layer of the page from the Punisher story, currently being produced in Italy for the Italian, French, and German markets. Figure 36 shows the corresponding page in the German edition.[15]

As can be seen in Figure 36, the letterer/graphic editor has not only lettered the balloons, but also retouched the drawing. Since the representation of swastikas is legally forbidden in Germany, the symbol worn on the forehead by the character in the third panel has been changed into a parallelogram with two crossing lines. In order to change the image only in the German edition, and to preserve the swastika sign in the Italian and French ones, the letterer/graphic editor has edited the layer of the German text, as can be seen in Figure 35.

Figure 35: PunisherMAX: the text layer
© Marvel Comics

Figure 36: PunisherMAX, German edition
© Marvel Comics

Figure 37 illustrates a still different case, in which the verbal signs in the original comic were part of the drawing in the first layer. The Spider-Man comic book was also a multilingual co-production, so that only the black color plate would differ from one edition to the other. To this end, using photo editing software, the letterer/graphic editor first erased the newspaper sub headline and byline, then reconstructed the colored background and then, on the text layer overlaid a transparent text box with the translations (one for each language). The final result for the Italian edition can be seen in Figure 37.

Figure 37: Ultimate comics: Spider-Man, Italian edition
© Marvel Comics

The Aquaman comic book in Figure 34 is produced only for the German market, allowing for the change of all four color plates. In multilingual productions, on the other hand, titles are in either black or white. If the background does not allow for a stark contrast with either one of these two colors, the background of the text box containing the title is filled up with the one of the two not used for the title. In this way the comic book can be printed in all languages by changing just the black color plate. The alternative strategy of leaving the title in the original language and provide a translation in a footnote seems to be less used for this type of comic books.

4. Conclusions

This article has provided an overview of visual adaptation strategies and practices, showing how the latter are implemented in the localization of two of the main types of comics published in translation, Japanese manga and US mainstream comics. To conclude, I offer a tentative overview of the main causes of visual adaptation strategies and practices, together with a comment on the effect of visual adaptation norms for manga on the polysystem of Western comics.

Visual adaptations may be the result of cultural or commercial factors, of the publisher abiding by overt state regulations concerning translated (as well as non translated) comics, or be prompted by the verbal content. Adaptation strategies and practices also differ according to the type and format of the comics translated, and to the form in which the source material is transmitted to the target publisher, as shown by the differences in the making of Italian editions of Japanese and US comics.

Publishers are driven by commercial motivations, and do not publish comic books which would either be censored, lead to legal disputes (Figures 5, 6, 7 and 8), or simply not sell. Furthermore, visual adaptation may considerably affect productions costs. However, the costs involved in the visual adaptation of foreign comics may well be compensated by high sales, as originally envisaged by Western publishers when importing manga. The shift toward less intrusive visual adaptation in foreign editions of manga provide an example of how translation norms developed within a strong users community have influenced commercial publication standards. In its turn, the move toward less visual adaptation has been welcomed by the comics industry as a way to save on production costs. When first appearing in Western markets manga where more heavily edited than now, but the extra cost was justified by the preoccupation of entering the market with a product which would otherwise look too foreign. Now, saving on the production budget is justified by a readership which is ready if not willing to accept what was until not long ago a highly unfamiliar visual presentation. The visual adaptation of American superhero comics, on the other hand, is streamlined in a process which follows the internationalization-localization pattern. Technological developments have made it possible for the files used for original publication to be used by letterers/graphic editors as a basis for multiple foreign editions, allowing for the manipulation of different layers of visual content, similarly to what happens when US film companies prepare their DVDs for multiple foreign versions.

Good examples of visual censorship are the Disney comics published in the Arab peninsula (Figures 9, 10 and 18) and the Marvel comic published in Germany (Figure 31). In some cases censorship is not sanctioned by an official body or law, but still cultural conventions and taboos may persuade the publisher to carry out significant adaptations. In Egypt, while no official censoring body existed, both pictures and verbal content are often deleted or changed in Disney comics published in the 1990s because of the expectation that they be otherwise received as ‘inappropriate’ by the target readership (Zitawi, 2004; 2008a; 2008b). Another good example of self-censorship is the ‘international’ edition of Tintin au Congo (Figure 16). Finally, visual editing may be prompted by the need to accommodate for the written text (Figures 19 and 20).

To conclude, it could be argued that, in the polysystem of Western comics, the norms for manga localization have developed in a way that is similar to what first happened with European editions of American comics.[16] While foreign editions of American comics were first adapted to existing visual formats (see Figure 2), when they acquired a more central role in the European comics polysystem in the 1930s, visual conventions such as balloons and paratextual signs were ‘repeated’ in translation, along with onomatopoeia, action lines, and other visual features specific to American comics. European comics, while developing their own national traditions, incorporated these conventions both through translation and through indigenous production.[17]

Similarly, when in the 1980s and early 1990s manga had a peripheral role in the polysystem of Western comics literature, the specific conventions of Japanese comics were adapted to Western visual conventions. As the number of titles and sale figures increased, manga acquired a more central role in the polysystem, and Manga visual and narrative conventions entered the mainstream. Again, this happened both through translation and through indigenous production by authors who have grown up reading manga, and who have increasingly adopted manga drawing style and conventions (Rommens 2000, Jüngst 2008, Bainbridge and Norris 2010). By privileging foreignizing strategies (i.e. resemblance to the original images) over domesticating strategies (i.e. resemblance to target visual conventions), manga visual adaptation practices have favoured innovation of the aesthetic conventions and pictorial repertoires of the target comics polysystems.

About the author(s)

Federico Zanettin is Full professor of English Language and Translation at the University of Venice, Italy.

Email: [please login or register to view author's email address]

•  Next article