RSS feed

Home > Issue38 > Huang article

Relatively Abusive and Relatively Corrupt: An Analytical Framework for the Study of Subtitlers' Visibilities

Boyi Huang, Dublin City University

ABSTRACT

This article attempts to interrogate the complexity of the current subtitling field through the lens of subtitler visibility. Currently, subtitlers adopt diverse subtitling approaches and emerge in various identities. Such multiplicity frustrates any binary and/or definitive preconceptions of subtitler visibility. Scholars have examined subtitlers’ visibilities, however, rarely through subtitles, and more rarely in a systematic manner. This article aims to contribute to knowledge by developing a framework for analysing subtitlers’ visibilities through their subtitles, i.e., for answering the question: how do different sets of subtitles give respective subtitlers different visibilities? Built on Nornes’ (2007) notions of corrupt and abusive subtitling, it proposes a nonbinary, multimodal, and bidimensional comparative framework that contrasts subtitlers’ visibilities as respectively manifested by their approaches towards the represented (non)verbal source and representing technical/verbal target of subtitling. It is argued that this novel framework enables a fuller capture of the dynamics of subtitlers’ visibilities and, as such, it provides a systematic means to move beyond any binary and reductionist preconceptions of subtitler visibility.

KEYWORDS

Subtitler visibility, analytical framework, multimodality, comparative approach, Abé Mark Nornes.

1. Introduction

Long gone are the days when there was only one kind of subtitler, i.e., professional subtitlers who are paid to work in media industries and follow their rules of subtitling. The popularisation of digital technologies has enabled people to subtitle outside the media industry with various motivations, such as fan-initiated subtitlers (fansubbers) and politically activist subtitlers. It has been widely noticed that industrial subtitlers and nonindustrial subtitlers not only demonstrate varied subtitling practices but also selectively learn from each other’s practices (Díaz Cintas 2010; Pérez-González 2014; Orrego-Carmona and Lee 2017). This variegated constitution of subtitlers’ identities and practices challenges any binary and static preconceptions of subtitler visibility as being simply either invisible or visible (Huang 2021). Instead, it would be more appropriate to regard subtitler visibility as various people’s dynamic capacities to be seen and recognised as subtitlers in their diverse social interactions (Huang 2021).

It is also problematic to assume that the mechanisms behind translator visibility in general are the same for subtitlers, as this risks ignoring the idiosyncrasies of subtitling. Subtitling consists of the production of written texts to be technically programmed and shown on screen, while the audiovisual text usually involving multiple modes of signs (e.g., language, image, sound, and music) is being played (Pérez-González 2020: 30). The source-target synchronicity, multimodality, and technical setting of subtitling contrast with traditional print-based media translation that often involves only one mode of sign (e.g., written language) and presents only the target text (TT) but not the source text (ST) on paper. While print-based translators can be said to hide behind their translations, synchronic and contrastive presentation of both ST and TT makes subtitlers already explicit on screen. Moreover, print-based translators can make their presence felt mostly by using words to translate words, whereas subtitlers have more chances of asserting their presence by using technically programmed words to translate both verbal and nonverbal texts. Subtitler visibility, therefore, requires more nuanced examination than translator visibility does in general.

Although such differences distance influential theories on translator visibility, such as Venuti (2008), from the discussion of subtitler visibility, it is still possible to find several studies that have specifically though superficially discussed this issue. Many of these discussions have drawn on prescribed regulations such as a subtitler’s social and legal status and subtitling norms (e.g., Pedersen 2011; Zhang and Mao 2013). Focusing on static and top-down regulations, they end up unsurprisingly with definitive and binary conclusions that subtitlers are either invisible or visible and fail to capture the above-mentioned multiplicities of subtitlers’ practices and identities. More recently, scholars have paid attention to how subtitlers can conduct social interactions through online media to change their visibility (e.g., Orrego-Carmona 2011; Rong 2015; Dwyer 2016), supporting a more complex vision of subtitler visibility. However, it is still rare to see studies that analyse subtitler visibility through subtitles, even though subtitles can be a subtitler’s most typical way of social interaction. Those that have made such attempts in a systematic manner are even rarer, as the few attempts tend to focus on only one or two features of fansubs, such as coloured titles and headnotes (e.g., Díaz Cintas 2010; Pérez-González 2007, 2014; Huang 2021) 1. In contrast, Nornes’ (1999, 2007) notions of abusive and corrupt subtitling have directly theorised subtitling approaches in relation to subtitler visibility, taking into consideration the technicality and synchronicity of subtitling. These notions, however, lack a rigorous integration of a multimodal perspective (Ortabasi 2007) and an in-depth conceptualisation of subtitler visibility (Huang 2021).

This article interrogates the current complexity of subtitlers’ visibilities as manifested by different subtitles from a multimodal perspective. By proposing comparisons between subtitles produced in different social settings, it also problematises preconceived binary understandings of subtitler visibility and assumes that different sets of subtitles, particularly those produced by different subtitlers, give respective subtitlers different visibilities. Tracing calls for attention to subtitler visibility in subtitles from a multimodal perspective made by previous scholarly works, such as Nornes (1999, 2007), Ortabasi (2007) and Pérez-González (2014), the article attempts to contribute to knowledge by proposing a nonbinary and multimodal framework for exploring the question: how do different sets of subtitles give the respective subtitlers relatively different visibilities?

The argument of the article develops as follows. Section 2 gives further details about Nornes’ (1999, 2007) notions of corrupt and abusive subtitling in relation to subtitler visibility. His notions are critiqued with reference to Huang’s (2021) nonbinary conceptualisation of subtitler visibility and Ortabasi’s (2007) proposal for including nonverbal elements in the ST. Based on these notions and critiques, Section 3 introduces a way of looking at subtitling from the perspectives of source and target respectively, which defines the source and target of subtitling and divides the analytical framework into two dimensions. This bidimensional framework allows us, in Section 4, to integrate the subtitling issues discussed in previous scholarly works for identifying units of the comparative analysis of subtitlers’ visibilities. As such, the article concludes by arguing that this new analytical framework develops a nonbinary and multimodal adaptation of Nornes’ (2007) notions, and that it enables a fuller capture of the current complexity of subtitlers’ visibilities.

2. Subtitling Approaches: Corrupt and Abusive Subtitling

Approach means “a way of considering or doing something” (Cambridge Dictionary Online), and a subtitling approach refers to a way of considering the most intrinsic mechanisms of making subtitles (subtitling). Intrinsically, subtitling means to use the available semiotic submodes attached to a subtitle to represent audiovisual texts selectively as the filmic narrative continuously unfolds with the storyline within the frame of the screen (Pérez-González 2007). Subtitles, thus, are somewhat limited by this temporally and spatially confined storytelling. In addition, as a form of translation, subtitling is also compromised by the dissimilarity between the source and target.

According to Nornes (2007), these constraints or characteristics generated by the subtitling apparatus itself make subtitling an inevitably violent activity; the violence is inevitably subjected on both the source and target, the foreign and familiar, the unknown and known. It is the subtitlers’ reactions to such violence that decide their approaches to subtitling. Nornes (2007) contends that there are only two kinds of subtitlers: corrupt or abusive. Corrupt subtitlers disavow the violence of subtitling; thus, they attempt to hide the violence by regulating the erasure of the dissimilarity between source (unknown) and target (known), that is, by domesticating the source with a readily digestible package of target texts (Nornes 2007: 177-178). As the dissimilarity between the source and target is erased, the interaction between the foreign material (unknown) and the viewer is removed, and the subtitler is hidden. On the other hand, abusive subtitlers accept the violence of subtitling; they try to make use of the violence by demonstrating the difference between the source and target, that is, by exposing the translation (Nornes 2007: 178-179). As the difference between the source and target is exposed, the interaction between the foreign material (unknown) and the viewer is augmented, and the subtitler is revealed. Therefore, corrupt subtitlers minimise their visibility, while abusive subtitlers maximise it. These two notions are probably the only ones that have directly theorised how subtitlers make themselves visible or invisible through their subtitles. However, there are still a few limitations with this theorisation, which are critiqued below.

First of all, given that many subtitlers are less restricted now in the digital age and norms are not set in stone, it is the ways of subtitling rather than the subtitlers themselves that can be identified as corrupt or abusive. Subtitling practices used to be stabilised under industrial supervision in the predigital age. Industrial subtitlers consistently followed protocols from their industrial patrons, whereby their identity was often associated with their stabilised practice. Now, as mentioned earlier, subtitling is no longer a purely industrial practice; many subtitlers who make subtitles outside the media industry or nonindustrial subtitlers strive to experiment with subtitling mechanisms. Moreover, subtitlers from different sectors of the field (industrial and nonindustrial) even start to learn from each other’s subtitling practices (Pérez-González 2014: 270). Such mutual adaptations frustrate attempts to identify any subtitlers or groups of subtitlers as definitively corrupt or abusive. Nevertheless, the general trend that a (group of) subtitler(s) follows within their certain subtitling works can still be identified. Thus, it is only the subtitlers’ subtitles and general approaches to subtitling, rather than the subtitlers per se, that should be identified as corrupt or abusive.

Secondly, the theoretical division of corrupt subtitling and abusive subtitling is based on dual standards. According to Nornes’ (2007) descriptions, by hiding the difference between the source and target and, hence, the violence caused by the difference, corrupt subtitling decreases subtitler visibility in subtitles; by exposing the difference and violence, abusive subtitling increases subtitler visibility in subtitles. However, a set of subtitles does not simply either hide or expose the difference, the violence, and the subtitlers; instead, it hides some differences, while exposing other differences. One approach is identified via the difference hidden, whereas another is identified via the difference exposed. This forms dual standards in the analysis. Alternatively, the analysis can be focused only on the exposed differences between the source and target. By focusing on the exposed difference, we focus on subtitler visibility, which is reasonable because with subtitles synchronically superimposed on audiovisual texts, the subtitlers are already explicit onscreen. It should be subtitler visibility rather than invisibility that is under discussion.

Thirdly, to differentiate a subtitling approach as either corrupt or abusive is a binary and reductionist perspective. This binary thinking immediately collapses when more than two sets of subtitles are compared. In a comparison among three subtitling approaches, for example, there must be one that is between the other two in terms of the difference it exposes between the source and target. If the one that exposes most difference is abusive and the one that exposes least is corrupt, the middle one has no place in this theory. A more practical adaptation of this theorisation would be to understand the two notions in relative terms, i.e., the subtitling approach that exposes more difference between the source and target should be relatively abusive, and comparatively the one that exposes less should be relatively corrupt. With more and more sets of subtitles compared, their approaches make a continuum with corrupt and abusive as its poles. This continuum version of the corrupt and abusive subtitling is in accordance with Huang’s (2021) spectrum-like conceptualisation of subtitler visibility. In the digitised mediascape, subtitlers’ practices and identities have vigorously multiplied, whereby subtitlers demonstrate various visibilities and draw a spectrum of visibility, rather than being definitely (in)visible (Huang 2021). Particularly, relatively different subtitling approaches give subtitlers relatively different visibilities. This also explains why the analyses of subtitlers’ subtitling approaches and visibilities are better implemented through a comparative manner.

Finally, Nornes’ (1999, 2007) analyses are all exemplified by the subtitles of dialogues and, at most, written words displayed in a scene, i.e., all verbal samples, leaving nonverbal sources in audiovisual texts unanalysed. Since sound films appeared, filmmakers and audiovisual localisers have been firmly and substantially focusing on the mediation through and of the dialogue in an audiovisual narrative, where language or verbal information is far from the only active channel of the communication (Pérez-González 2014: 183). This unjustified focus, however, has been kept in the audiovisual translation (AVT) sector of the media industry up to now, with the result that the academic discourse of AVT is also prevalently about verbal mediation, including Nornes (2007). It is as if “actually addressing the non-linguistic realm is something of a taboo,” Ortabasi (2007: 280) comments. Nevertheless, Ortabasi (2007) agrees with Nornes (2007) on the ground that the notions of corrupt and abusive subtitling problematise conventional and often inefficient mediation of foreign cinematic experiences; but she argues that, if linguistic “cultural references” are regarded as “untranslatable” or make subtitling inevitably violent, “this holds even more true for visual cultural references” (Ortabasi 2007: 283). This critique calls for rebalancing Nornes’ (2007) notions by giving attention to both verbal and nonverbal elements involved in subtitling.

In 2016, Nornes wrote a piece called Afterthoughts to develop his original arguments and emailed it to other scholars over the next few years. The major change he made to his original notions seems to be to use the two new terms of “sensible” and “sensuous” respectively in place of “corrupt” and “abusive.” While the ‘new’ notion of “sensible subtitling” still corresponds to the original “corrupt subtitling” in terms of its characteristic fidelity to the target, “sensuous subtitling” indicates Nornes’ advocacy for faithfulness to the source, “the foreign of the foreign film,” “otherness,” “the materiality of language,” and ultimately the “sensuousness” of cinema (Nornes 2016: 12), steering away from his original “abusive subtitling” that supports an interventionist approach liberating subtitlers from both source and target. Although he also points out the defect of characterising subtitlers rather than their ways of subtitling (Nornes 2016: 24), he confusingly keeps referring to “sensible subtitlers” and “sensuous subtitlers” throughout his Afterthoughts. Nornes (2016) also critiqued his own original binary thinking and admitted it should be “a matter of degree” between the two approaches (2016: 24). However, he has not seen the dual standards of his theorisation, failing to suggest alternative ways of analysis based on this degree-like understanding of his concepts. Presumably, a degree-like understanding would lead him to adopt a comparative approach to subtitle analysis, as mentioned above, which he has tentatively done yet with only one example – his last example on translating silence where he finally reaches a continuum-like understanding of his concepts (Nornes 2016: 19-23). Besides, he has attempted but failed again to incorporate multimodality into his theory in the Afterthoughts piece by using the new examples of dialects, lyrics, and silence in dialogue (Nornes 2016), which are paraverbal, if not entirely verbal.

Despite the above limitations, Nornes’ (1999, 2007, 2016) notions provide a strong theoretical foundation for the analysis of subtitler visibility. They concretise the abstract issue of visibility into how subtitles manifest different subtitling approaches and how differences in approach give respective subtitlers different visibilities. Therefore, the notions are rebuilt with the help of these critiques into a stronger framework for the comparative analysis of subtitlers’ subtitling approaches and visibilities in the next section2.

3. Subtitling Dimensions: The Represented Source and Representing Target

Divided by the inevitable violence that subtitling puts on both the source and target, subtitling retains two dimensions. It unavoidably but acceptably violates some norms of both the source and target by using one to represent another, based on the assumption that no translation fully transfers the meaning of an ST, and nor is it necessary for translation to do so (Catford 1965: 48). Source-oriented researchers would focus on how the source is represented brokenly by the target, whereas target-oriented researchers would examine how the target is disrupted as used to represent the source (Freddi 2013). I argue that the analysis should be focused on both the source and target in order to gain a more holistic and less binary view (cf. Freddi 2013). From the source perspective, the mutilation is seen as explicitly put on the source, as the source is seen to be violently represented (by the target), which I call the represented dimension. From the target perspective, the disruption is seen as explicitly put on the target, because the target is seen to be violently used in representing (the source), which I call the representing dimension. By looking at subtitling (approaches) from both dimensions, we can understand this activity more fully than focusing on only one of them:

▪ In the represented dimension, the violence on the source is seen as explicit.

  • A relatively corrupt approach means being more intended to refuse and hide the mutilation on the source, by disclosing less difference between the source and target, ultimately by erasing the source.
  • A relatively abusive approach means being more inclined to accept and reveal the violence on the source, by disclosing more difference between the source and target, ultimately by exposing the source.

▪ In the representing dimension, the violence on the target is seen as obvious.

  • A relatively corrupt approach means being more intended to refuse and hide the violence on the target, by disclosing less difference between the source and target, ultimately by erasing the target.
  • A relatively abusive approach means being more inclined to accept and reveal the disruption on the target, by disclosing more difference between the source and target, ultimately by exposing the target.
  • In relation to subtitler visibility, in both dimensions, relatively corrupt subtitling gives less visibility to subtitlers in the process of disclosing less difference between the source and target, while relatively abusive subtitling gives more visibility to the subtitlers in the process of disclosing more such difference. In contrast, in the represented dimension the subtitle’s visibility is linked with what subtitlers do to the source, whereas in the representing dimension it is linked with what they do to the target.

I should clarify what source and target respectively mean in this framework. In his description of corrupt and abusive subtitling, Nornes points out that corrupt subtitles smooth the rough edges of foreignness, while “abusive subtitles circulate between the foreign and the familiar, the known and the unknown” (2007: 185). In other words, what differentiates the source and target of subtitling is whether the signs in a text are foreign/unknown or familiar/known to audiences. Albeit Nornes (2007) has been critiqued by Ortabasi (2007) as only drawing on verbal source, his understanding of the foreign/unknown and familiar/known as the source and target of subtitling proves insightful as it can encompass a wide range of semiotic modes that participate in audiovisual mediation. From a multimodal perspective, the source is all the potentially unknown and foreign information (rather than merely dialogue) in the original audiovisual texts, while the target is all the potentially known and familiar in the subtitled audiovisual texts. This is in contrast to the often-used linguistic opposition of foreign and domestic as the source and target of subtitling, which cannot fully examine audiovisual texts. For example, a painting in a domestic film is not necessarily known to all in a domestic audience; thus, it requires translation for some of the audience who deem it foreign, and it is technically the source of subtitling. Indeed, “[e]very film is a foreign film, foreign to some audience somewhere — and not simply in terms of language” (Egoyan and Balfour 2004: 21). In other words, the whole audiovisual text can be foreign/unknown to audiences and, thus, the source of subtitling, while everything about subtitles that is familiar/known to audiences can be the target of subtitling.

The inclusiveness of the definitions for the source and target here does not necessarily lead to an inclusive comparative analysis. Rather, the understanding that everything in an original audiovisual text can be source and everything attached to the subtitles can be target underlines that subtitlers should not problematically assume the homogeneity of a group of audiences. Only based on this understanding can we move on to argue that every subtitle of every portion of an audiovisual text can be analysed in terms of relative corruptness and abusiveness. However, it is practically unnecessary and inefficient to analyse and compare everything. To make the comparative analysis of subtitlers’ subtitling approaches and visibilities more efficient than an inch-by-inch search requires a framework to be able to identify critical and countable analytical points or units. This is explained in the next section.

4. Analytical Points: Subtitling Issues

According to Pederson, certain points in either the source or the target tend to make subtitling particularly more difficult than others do, as they (the former) “stand out from run-of-the-mill translation and require extra special care and the active and conscious employment of subtitling strategies” (2011: 41). Borrowing Díaz Cintas and Remael’s (2007) term, I use subtitling issues to refer to the particularly difficult points in the source and target systems that clarify the difference between the source and target, the violence of subtitling, and thus, subtitlers’ subtitling approaches. As such, subtitling issues are used to identify analytical points of the current framework to make the analysis focused and more efficient. Referring to the explicit difficulties that are from both the source and target, subtitling issues can fall into the two dimensions of represented and representing, respectively. In the represented dimension, it is the certain particularly difficult points in the source that make subtitling explicitly violent on the source, whereby these points are called source issues. In the representing dimension, it is these particularly difficult points in the target that make subtitling explicitly violent on the target, whereby they are called target issues.

Although subtitler visibility has been discussed infrequently, many scholars have studied subtitling issues or similar concepts. Despite their different purposes from that of the current article, these studies have commonly drawn on frameworks that make analyses more efficient also by focusing on certain difficult points or subtitling issues. The subtitling issues provided by previous scholarship are incorporated to break down the comparative analysis proposed by the current framework, which are listed and discussed in the rest of this section.

To explain the framework, I will use the Warner Home Video (WHV) industrial subtitles and the YYeTs fansubs of The Big Bang Theory (TBBT) as examples to showcase the analysis. Both sets of target subtitles are in traditional Chinese characters. WHV is the distributing division of the Warner Bro. Corporation which is the original producing company of TBBT. Given that the WHV subtitles are produced by the subtitlers who are employed by and for the commercial company of WHV, WHV subtitles are recognised as industrial subtitles. YYeTs is a fansub group and its subtitles are voluntarily produced by YYeTs fansubbers outside the media industry and not for commercial purposes. The YYeTs subtitles are considered fansubs.

4.1. Represented Dimension

Previous scholarly works concerning source issues generally follow two directions: those that address the verbal aspect of source (mostly dialogue) by using linguacultural frameworks, and those that do not exclude the nonverbal aspect of source by resorting to multimodal theories. In the represented dimension, source issues, thus, can be divided further into verbal issues and nonverbal issues.

4.1.1. Source Verbal Issues

Source verbal issues refer to the particularly difficult points in the verbal information of the original audiovisual texts that make explicit the mutilation of subtitling on verbal source. They mostly, but not completely, correspond to traditional print-based verbal translation issues. In addition to being linguaculturally confined, subtitling is also spatial-temporally limited by technical matters and synchronised with ST. Certain points in the verbal source that would not be considered to be an issue in traditional print-based translations can be debatable in subtitling, such as punctuation and numbers. Focusing on verbal aspects from the represented dimension, this subsection mainly discusses source verbal issues as analytical points (including examples) prevailing in English language. Current literature that has explored a vast set of source verbal issues of subtitling include: Nedergaard-Larsen (1993), Perego (2004), Díaz Cintas and Remael (2007), Geogakopoulou (2009), Gottlieb (2009), Pettit (2009), and Pedersen (2011). Together, they provide a list of diverse source verbal issues, as shown in Table 13. It should be noted that this list, partly extracted from different scholarly works, is by no means a comprehensive or clearly segmented chart of source verbal issues. Rather, it is presented here to show the many possible issues that can showcase the violence of subtitling on the verbal source and, thus, can be used to identify analytical points for the current framework.

An example is given below to demonstrate how to comparatively analyse the subtitling approaches of two sets of target subtitles to source verbal issues. This is an example of a geographical reference, Epcot, which is found in the third episode of TBBT Season Nine (S9). It is a source verbal issue because it demonstrates a point in the verbal source that is particularly difficult to be, and thus will be violently, represented by target subtitles. It makes explicit the mutilation of subtitling on verbal sources, in that a limited subtitle cannot fully interpret what can be unknown about this verbal information. It in turn also makes explicit the subtitlers’ approaches to such violence for comparative analysis. The reference Epcot can be foreign or unknown for some audience members. It is the extent to which the foreign or unknown about Epcot is represented in the two groups of target subtitles that demonstrates the differences between the two targets.

In this case, the YYeTs subtitles and the WHV subtitles are apparently different. What WHV subtitlers considered unknown about Epcot and translated as the known in their target subtitles are 迪斯尼 ‘Di Si Ni’ and 未來世界 ‘future world,’ whereas what the YYeTs subtitlers deemed unknown about Epcot and rendered as the known in their target subtitles are 未來世界 ‘future world,’ 主題公園 ‘theme park,’ and its location in 美國佛羅里達州迪士尼世界度假區 ‘the USA, Florida State, Di Shi Ni World Resort 4.’ The same source term — Epcot — has more foreign or unknown information represented by the YYeTs subtitlers in their target subtitles than by the WHV subtitlers in their target subtitles. As more unknown information about Epcot is represented in the YYeTs subtitles — 主題公園 ‘theme park,’ and 美國佛羅里達州世界度假區 ‘the USA, Florida State, World Resort,’ the YYeTs subtitles display more difference between what can be unknown about the source term — Epcot — and what can already be known by the target audience through the Chinese characters. In other words, the YYeTs subtitles display relatively more difference between the source (the unknown) and the target (the known) by disclosing more unknown verbal source than the WHV subtitles do in the analytical case of Epcot, whereby YYeTs subtitles appear relatively abusive and WHV subtitles relatively corrupt.

4.1.2. Source Nonverbal Issues

Source nonverbal issues refer to the particularly difficult points in the nonverbal information of the original audiovisual texts that make explicit the violence of subtitling on the nonverbal source. Nonverbal issues have been used by scholars in order to understand subtitling through semiotics instead of mere linguistics. These scholars have argued that the source of subtitling is not only the languages, but ensembles of various semiotic resources that can be foreign to audiences. Given the long tradition of focusing on the verbal aspect of subtitling, many of these scholars tend to divide the source of subtitling as being verbal and nonverbal (e.g., Delabastita 1989; Zabalbeascoa 2008). Distinctive from verbal elements, source nonverbal issues can only make even more explicit the violence of subtitling compared to source verbal issues, since subtitles per se are verbal. Focusing on source nonverbal issues, this subsection discusses nonverbal issues prevailing in subtitling studies, rather than those in general semiotics. AVT scholars who have done so systematically (e.g., de Linde and Kay 1999; Gambier 2006; Díaz Cintas and Remael 2007; Zabalbeascoa 2008; Georgakopoulou 2009; Perego 2004; Pérez-González 2014) have categorised the signs in audiovisual texts in different ways. Based on Stöckl (2004), Pérez-González’s (2014) categorisation of the nonverbal source of subtitling on the medial level is useful for identifying nonverbal analytical points of the current framework, as shown in Table 2.

To focus the analysis on individual medial variants does not mean that each instance of a medial variant found in an audiovisual text can always form meaning of its own to be perceived by the audience or analysed by AVT scholars. Rather, as entailed by the “resource integration principle” of multimodal communication, the meaning-making process of integrating different modes and variants “cannot be reduced to or explained in terms of the mere sum of its separate parts” (Baldry and Thibault 2006: 18). As each mode requires a different form of cognitive orientation, one mode may allow for faster processing than another, and thus, there are times in the integration as well as the perception that a certain mode plays a dominant role (Pérez-González 2014: 220-221). In such cases, the dominated mode or medial variants “can only be made to mean and communicate specific contents by a combination with other modes” (Stöckl 2004: 18). Therefore, there are certain identified individual nonverbal medial variants that have to be analysed in combination with other modes (e.g., image or language).

For example, in the 19th episode of TBBT (S9), Sheldon (a protagonist of the show) played a piece of performed music, Taps, with which he and his partner Amy stood in silent tribute in front of his condemned computer covered in black canvas (combined imagery signs). Taps is identified as a source nonverbal issue and thus an analytical point, as it makes up an unknown point in the music source (one type of nonverbal source) that is particularly difficult to, and hence will be violently, represented by target subtitles. This point makes explicit the mutilation of subtitling on (nonverbal) music source, since a limited and verbal title cannot fully represent what can be unknown about this nonverbal information. It, therefore, also makes explicit the subtitlers’ approaches to such violence for comparative analysis. Again, with the piece of music being foreign or unknown for some audience members, it is the extent to which the foreign or unknown about Taps is represented in the two groups of target subtitles that contrasts their approaches. This time, the WHV subtitlers did not interpret anything with their subtitles, whereas the YYeTs subtitlers rendered the potentially unknown name and usage of the music with potentially known information to the audience: 該曲名為《Taps》常在軍人葬禮國葬時吹奏‘The music is named Taps. It is often performed at soldiers’ funerals and state funerals.’ In this way, YYeTs subtitles reveal more difference between the source (the unknown) and the target (the known) in their subtitles than the WHV subtitles that uncover no difference, whereby the former approach appears relatively abusive and the latter relatively corrupt.

Similar comparisons can also be made in terms of other nonverbal issues. TBBT is also very rich in its imagery narratives, such as the Gollum statue Sheldon talked to in the 1st episode (S9), the dead rabbit in the 14th episode (S9), the crossed figures Sheldon made in the 16th episode (S9). Take the “crossed fingers” for example. Sheldon made a “crossed fingers” gesture to Leonard, while telling him that if Howard and Bernadette were to have twins, they could do experiments on the kids (combined linguistic signs). This gesture is identified as a source nonverbal issue and thus an analytical point, as it makes up an unknown point in the image source (one type of nonverbal source) that is particularly difficult to, and hence will be violently, represented by target subtitles. Again, with the gesture being foreign or unknown for some audience members, it is the extent to which the foreign or unknown about the image is represented in the two groups of target subtitles that contrasts their approaches to this violence. The WHV subtitlers did not interpret anything with their subtitles, whereas the YYeTs subtitlers rendered the potentially unknown image/gesture with potentially known information to the audience: 雙指交叉為成功祈禱‘Two fingers crossed. Pray for success.’ In this way, YYeTs subtitles reveal more difference between the source (the unknown) and the target (the known) in their subtitles than the WHV subtitles that uncover no difference, whereby the former approach is relatively abusive and the latter relatively corrupt.

4.2. Representing Dimension

In the representing dimension, as said above, target issues are those particularly difficult points in the target that make subtitling explicitly violent on the target. Target issues can be different from source issues. Arguably, subtitled audiovisual texts are overall a target text, as subtitles are synchronic with the original audiovisual texts. However, only the target subtitles have corresponding parts in the source that can form a translational analysis. Therefore, regarding approaches to the disruption on the target, only subtitles are analysed. Subtitles are written verbal texts that are technically programmed, whereby the target verbal and technical systems are violently used to represent the source. In the representing dimension, target issues, hence, can be further divided into technical issues and verbal issues.

4.2.1. Target Technical Issues

Target technical issues refer to the particularly difficult points in the technical system of target subtitles that make explicit the disruption of subtitling on the target technical system, and subtitlers’ approaches to such violence, for comparative analysis. Similar to the grammars of verbal systems, the technical system of subtitles has limitations in terms of representing another (Karamitroglou 1998). Different from verbal grammars, these technical limits of subtitles have been invented or conventionalised by scholars much more recently. Conventionalising subtitling, many scholars, such as Luyken et al. (1991), Ivarsson and Carroll (1998), Karamitroglou (1998), Díaz Cintas and Remael (2007), Díaz Cintas (2008), and Georgakopoulou (2009), have proposed sets of technical limits to be followed by subtitlers in their practices, and those lists of limits are known now as technical conventions or settings. Every time these conventions are followed, or settings used, it grows deeper in people’s mind that these are the ways subtitles are supposed to be done. However, many of these technical settings are now found to be constantly experimented upon and violated by various subtitlers’ increasingly diverse and limit-breaking subtitling practices. Therefore, these technical settings have become debatable points in the technical system, or issues of how the technical appearance of subtitles should manifest. These technical issues make explicit the mutilation of subtitling on the conventional settings of target technical systems. Table 3 provides a list of target technical issues that are based on Díaz Cintas and Remael’s (2007) technical settings and Georgakopoulou’s (2009) categorisation of technical aspects. It serves efficiently the purpose of identifying these technical issues in subtitles for comparatively analysing subtitling approaches.

Take WHV subtitles and YYeTs subtitles for TBBT(S9) for example. Conventionally, subtitles are set as monolingual in most media markets (Díaz Cintas and Remael 2007), which is a language character setting. The WHV subtitlers followed the convention to consistently use monolingual subtitles, while the YYeTs subtitlers innovatively used bilingual subtitles throughout the set of subtitles. With the target language (Chinese) in the subtitle, both monolingual and bilingual subtitles can show some difference between the source and the target, as the source (in audio) is kept synchronic with the target subtitles. In contrast, bilingual subtitles display both the source language (English) and the target language (Chinese) in written forms, which can show additionally the difference between the source and target languages both in written forms. Therefore, the bilingual subtitles (YYeTs subtitles) reveal more difference between the source and target than the monolingual subtitles (WHV subtitles) do.

Also, some technical issues are only observable occasionally rather than continuously. Such issues have to be analysed in terms of both whether a convention is followed and how it is followed. In terms of spatial settings, notations, for instance, are not used in WHV subtitles in adherence to industrial conventions, but they are adopted in YYeTs subtitles and innovatively used in the form of headnotes for a total of 97 times for TBBT (S9). These headnotes as part of subtitles are positioned near the top of the screen usually to explain what cannot be fully represented about the source in the common bottom space of subtitles. Instead of translating the filmic diegesis, they provide extra information about certain elements in the diegesis. They open up an “extra-diegetic dimension” (Pérez-González 2012: 345) to help the audience connect what the subtitlers think they do not know with what the subtitlers think they already know, displaying difference between the source (the unknown) and target (the known). In other words, the YYeTs subtitles that have headnotes reveal more difference between the source and target than do the WHV subtitles having no notations at all.

In addition, some technical issues are not directly observable on screen and thus have to be analysed by reading the subtitle file (not always available). Shot changes, as part of the temporal settings, can be a good example. According to Díaz Cintas and Remael, a subtitle “should leave the screen just before the cut occurs and a new subtitle spotted after the cut” (2007: 91). This convention is set based on reception studies on eye movement to prevent viewers from reading the same subtitles repetitively. However, this conventional setting also de facto keeps the target synchronic with the source audiovisual texts as if the subtitles are part of the source texts, erasing the temporal difference between the source and target. For TBBT (S9), the WHV subtitlers always keep their subtitles showing within two consecutive cuts (following the convention), whereas the YYeTs subtitlers have their subtitles showing as long as possible and constantly exceeding the duration of at least one shot cut. Comparatively, the WHV subtitles show less temporal difference between the source and target, whereas the YYeTs subtitles show more temporal distance from source texts. All three analytical cases have shown that YYeTs subtitles are relatively abusive and WHV subtitles relatively corrupt regarding target technical issues.

4.2.2. Target Verbal (Issues)

Target verbal issues refer to the particularly difficult points in the language of target subtitles that make explicit the disruption of subtitling on the target verbal system. Like target technical issues that encourage the comparison of different sets of target subtitles with the target technical system, target verbal issues should stimulate the comparison of different sets of targets with the target verbal system(s). The analytical cases based on a list of target verbal issues should manifest how the target language(s) has/have been unconventionally used in subtitles to represent the source. This list of target verbal issues, thus, would largely depend on the target language and its linguistic idiosyncrasies, such as the gendered adjectives of Spanish, the attributive clause of English, the four-character idioms of Chinese. Similar to the target technical issue of notations, many target verbal issues, like those linguistic markers, only occur occasionally but possibly more frequently, and therefore a comparable corpus would be required to search out all occurring examples and analyse how they, as part of a verbal system, are violently used in different sets of targets.

Monolingual comparable corpora usually consist of a set of texts originally in one language and a set of translation texts in the same language. The comparable corpora are designated to find “patterns that are distinctive of translated texts as opposed to texts produced in a nontranslational environment” (Saldanha and O’Brien 2014: 67). Comparable corpora have also been applied in AVT, where they are used to find out the similarities and differences between AVT and original fictional verbal content (Pérez-González 2014: 165), for example Freddi’s (2013) comparison between dubbed and original fictional dialogue in Italian. According to Pavesi (2019), there have been only a few fully described AVT-focused corpora with copyright clearance, such as TIWO (Television in Words), PCFD (Pavia Corpus of Film Dialogue), OpenSubtitle parallel corpora, and CORSUBIL (Corpus de Subtítulos Bilingües Inglés-Español). While, among these corpora, PCFD has a comparable component composed of original dialogue of 24 Italian films (Pavesi 2014), the comparison was made between original fictional dialogue and, again, dubbing language rather than subtitles. See Pavesi (2019) for a more detailed account of how to apply corpus in AVT.

If applied under the current framework, comparable corpora can be used to find out how two or more sets of target subtitles deviate from the target languages originally in fictional settings like films, and thus to shed light on how the target verbal systems are violently used in different target subtitles and on the subtitlers’ approaches (relatively corrupt or abusive) to such disruption. The corpora have the advantages of easily identifying linguistic markers used in two or more TTs and of generating statistics of such usages if from a quantitative angle. In light of such merits, it is hoped that future scholarship could engage in such corpus-based investigations, under this framework, on subtitlers’ relatively different approaches and visibilities in the target verbal aspect. See Figure 1 for a graphic display of the framework described above.

4.3 Patterns and Interpretations

The framed comparative analysis generates patterns of subtitles in relation to subtitlers’ approaches and visibilities and, as stated, the patterns can vary among dimensions and aspects of subtitling. There can be generally two patterns in each dimension (represented and representing), and these are respectively called R-ed(1), R-ed(2), R-ing(1), and R-ing(2).

In the represented dimension, a comparative analysis focused on both source verbal issues and nonverbal issues should lead to two general patterns: R-ed(1) the analytical cases where subtitle set A reveals more difference between the source and target than subtitle set B or vice versa (as exemplified above for source verbal and nonverbal issues), because one set has comparatively more unknown source represented to audience than the other; and R-ed(2) those where sets A and B reveal such difference to the same extent, because they have the unknown source represented to the same extent.

In the representing dimension, the comparative analysis focused on both target technical issues and verbal issues should also result in two general patterns: R-ing(1) the analytical cases where subtitle set A reveals more difference between the source and target than subtitle set B or vice versa (as exemplified for target technical issues), because one set demonstrates more innovative ways (by not following the conventions) of employing target systems that reveal more difference between the source and target, while the other is adherent to conventional ways of using target systems that hide such difference; and R-ing(2) the analytical cases where set A and set B reveal such difference to the same extent, because they indicate the same ways of applying target systems.

The patterns are interpreted in two stages. Firstly, the patterns described above tell the result of which set of subtitles eventually reveal more difference between the source and target, indicating which set is relatively abusive and/or relatively corrupt. The relative abusiveness and corruptness can vary across different dimensions and aspects. For example, based on results of the comparison between the WHV and YYeTs traditional Chinese subtitles for TBBT (S9), YYeTs subtitles are relatively abusive and WHV subtitles are relatively corrupt; the relative difference between their approaches is more drastically manifested by source verbal issues (446 R-ed1s out of 1167 total analytical cases) than by source nonverbal issues (50 R-ed1s out of 422 total analytical cases); such divergence is even more palpably shown by target technical issues (3 R-ing1s out of 3 total analytical cases). This part of the data is used to interpret the subtitler’s physical visibilities that indicate how much one group of subtitlers can be more seen than the other in their social interactions (Huang 2021: 53) and in their subtitles in this instance. In the example, the fansubs that are relatively abusive reveal more difference between the source and target and, thus, give more physical visibility to the subtitlers than the industrial subtitles that appear relatively corrupt do.

Secondly, the patterns also tell how and why one subtitle set reveals more difference between the source and target than another and manifests relatively abusive and corrupt approaches. Such reasons are different between dimensions. Again, take WHV and YYeTs traditional Chinese subtitles for the TBBT (S9) for instance. In the represented dimension, YYeTs fansubbers’ relatively abusive approach, which reveals more difference between the source and target, is shown by having more unknown source represented, whereby the fansubbers play the role of educators “to educate a curious audience” with unknown information (Condry 2010: 203); in contrast, WHV industrial subtitlers’ relatively corrupt approach, which reveals less such difference, is shown by having less unknown information represented, whereby the industrial subtitlers confirm their position as gatekeepers helping the industry safeguard domestic cultural values throughout the institutional history of films and other media products (Nornes 1999). Both educators and gatekeepers are subtitlers’ symbolic visibilities that indicate what (different) roles or positions subtitlers are recognised as (Huang 2021: 53). The other dimension might show nuances of the subtitler’s symbolic visibilities. In the representing dimension, YYeTs fansubbers’ relatively abusive approach, which reveals more difference between the source and target, is shown by experimenting with the target conventions and innovatively exploiting the target systems, whereby the fansubbers present themselves as innovators who constantly exhibit “creativity to the use of subtitles” (Díaz Cintas 2010: 124); differently, WHV industrial subtitlers’ relatively corrupt approach, which reveals less such difference, is shown by following institutional conventions, whereby the industrial subtitlers act as adherents of institutional conventions that encourage a “self-effacing presentational style” (Pérez-González 2014: 51-52).

Such interpretations claimed should be based on the research cases chosen. The research cases used as examples here are sets of subtitles produced in different working contexts or social settings. In the fansubbing setting, fansubbers’ role as educator is validated via their motivation to avoid being accused of copyright infringement or other legal disputes and thus via ascertaining the educational nature of their practice (Tian 2011: 94), whereas in the industrial setting, industrial subtitlers’ role as gatekeepers is both restricted and supported by the institutional agenda of “naturaliz[ing] a dominant, hierarchically unified worldview” (Nornes 1999: 18).

This difference across dimensions and aspects explains why it should be ensured that subtitling issues to be analysed should be selected respectively from the two dimensions (represented and representing) and four aspects (source verbal, source nonverbal, target technical, and target verbal). If necessary, the numbers of analytical cases in each of these patterns can be collected, collated, and interpreted. In terms of research cases, due to limited space, I have used only two sets of subtitles to illustrate the analyses, which for some readers may appear as another binary contrast. However, the idea is that this comparative framework encourages future research to map out more subtitlers’ relative positions on the ever-expanding subtitler visibility spectrum (e.g., fansubbers hired by media corporates, subtitlers appointed by directors, politically activist subtitlers, and other subtitlers whose working contexts are neither entirely industrial nor fan-based). Depending on the research cases chosen, this framework can also be used to investigate other factors, in addition to working contexts, that might shape subtitlers’ visibilities, such as source dialogue and target languages, audiovisual genres, directing styles, and geopolitical locations. The more cases that are interrogated, the clearer and more complex the subtitler visibility spectrum gets.

4.4. What it is (not)

Before conclusion, I wish to articulate on a few points that may appear unclear.

First, one may argue that this framework is still a dichotomy divided by the abusive-corrupt binary. In this paper, the word ‘binary’ is understood to describe an absolute either-or structure, whereby a continuum stretched by two poles is not a binary in this article. The either/or binary understanding of abusive/corrupt subtitling may not be how Nornes has originally intended them to be understood or used. As mentioned, Nornes has hinted in his later works that he also sees, if not applies, fluidity between the two extremes. However, the two terms have rarely been applied relatively. This relative application is what this framework proposes, and it provides a systematic means to do so.

Second, in Section 3, I state that one assumption of this framework is that no translation fully transfers the meaning of an ST, and nor is it necessary for translation to do so (Catford 1965: 48). I shall expand on this. Human communication including translation involves the use of (verbal and nonverbal) symbols or texts that are loaded with information. However, meaning making is a complex cognitive process that is dynamic, subjective, and context-based. That means a translator, as a communicator, is concerned not only with the texts but also the cognitive environment of their target audiences. The use of known symbols to represent unknown information is shown by the (translation) texts, whereas the meaning-making or cognitive process is happening in the mind. An ST consists of symbols loaded with possibly countless information. A translator communicates only certain pieces of the information that they consider are unknown and relevant to the TT readers with the symbols that they consider are known and mostly likely to be inferred as the intended meaning by these readers. Therefore, it is neither possible nor necessary for a translator to transfer all the information of an ST, let alone its meanings that are made in the readers’ mind.

Moreover, one may find the analysis of the known and unknown problematic at first glance. I also found them problematic when I first read them in Nornes’ work because Nornes did not seem to have defined them clearly. What is considered known or unknown by a translator depends on their context and the positioning of their target audiences. That is, what is (un)known for one translator/audience might not be the same for another. This is a problem for the translator, but not for the analyst. Based on the understanding that the unknown is the source, while the known is the target, the TT explicitly demonstrates both the known and the unknown. The TT is the known representing the unknown, or is the unknown represented by the known. This identification of the (un)known was a decision made not by the analyst but by the translator when they translated. While what is considered as (un)known may vary from translator to translator, the completed translations clearly demonstrate what the translators have decided as (un)known. This framework does not enable the analyst to assume the translators’ positions, but allows fair comparisons between translators’ decisions by subjecting all TTs to the same comparison with the same ST.

I should add that this framework does not, and is not designed to, enable the researcher to claim what subtitlers consciously want with their subtitles, or to analyse how audiences may see or perceive the subtitlers. Rather, the framework provides a tool for the researcher to interpret the subtitles and texts in terms of the visibilities they give to the subtitlers. It is more useful when applied in accordance with findings of subtitler visibility that are evidenced or claimed otherwise, for example findings from interviews or eye tracking. If it is applied before the researcher knows the subtitlers’ motivations or the audiences’ perceptions of the subtitlers, the framework serves as a means to generate hypotheses in such regards. If it is applied after the researcher’s awareness of the subtitler’s motivations or the audiences’ perceptions, the framework functions to substantiate such awareness/assumptions/claims by providing evidence from the subtitles: as discussed before, how different sets of subtitles give different visibilities to the respective subtitlers. As the framework is built under a definition of subtitler visibility that includes both physical and symbolic values, it enables the interpretation to go beyond the texts analysed. For example, the arguments that fansubbers are more visible than industrial subtitlers and that the former appear as educators whereas the latter perform as industrial gatekeepers are assumptions that are often claimed but rarely substantiated. These assumptions can be and have now been supported with evidence from example subtitles using the current framework.

5. Conclusion

This article has proposed a novel framework for the analysis of subtitler visibility through their subtitles, which has rarely been done previously. It moves beyond the few previous analyses that focus on only one or two subtitle features by suggesting analysing subtitler visibility more systematically (i.e., from two dimensions and four aspects) in order to examine subtitler visibility in their subtitles to a fuller extent. Theoretically, it has adapted some of Nornes’ (2007) notions relevant to subtitler visibility, such as corrupt and abusive, foreign and familiar, unknown and known. Different from Nornes (2007), the concepts of corrupt and abusive in this framework are (1) used in a more dynamic manner by referring to subtitling approaches rather than subtitlers themselves, (2) evaluated in a more consistent manner by the single standard of the difference between the source and target revealed by the subtitles instead of the difference both revealed and hidden, (3) applied in a more relative/nonbinary manner by understanding them as a continuum stretched by the sets of subtitles subjected to comparison, (4) integrated with multimodality by including nonverbal issues in the analysis. These notions have also proved compatible with the working definition of subtitling, based on which ‘new’ concepts have been proposed to build the framework:

  • Subtitling is divided into the represented and representing dimensions, which not only helps avoid bias or preference towards either source or target, but also generates nuanced findings regarding subtitlers’ approaches and visibilities.

  • The source is considered as all the possibly unknown rather than merely the foreign dialogue, whereas the target is regarded as all the possibly known rather than only the domestic words, which enable the framework not only to minimise the impact of researchers’ potentially biased assumptions about what is linguaculturally foreign and domestic, but also to allow assessment of all semiotic resources involved in subtitling.

  • A list of (potential) subtitling issues in terms of the four aspects of verbal source, nonverbal source, technical target (and verbal target) separately in the two dimensions is garnered for efficient identification of analytical points, focused on which the analysis is initially qualitative but can also mount to quantitative examination.

To conclude, this bidimensional analytical framework provides a systematic means to move beyond any binary and reductionist preconceptions of subtitler visibility. It testifies and, hence, further contributes to a complex understanding of subtitler visibility (Huang 2021) that acknowledges the contextualised multiplicity of subtitler’s practices and identities. Future projects applying this framework should, with their various scopes, patterns, and interpretations, reveal more complexity of subtitlers’ visibilities that is ever more magnified in this increasingly digital and interactive world.

Acknowledgement

The author would like to thank Luis Pérez-González and Patrick Zabalbeascoa for their invaluable theoretical insights. The author also thanks Wayne Wen Chun Liang, Yinran Wu, Yuxiang Wei, Patrick Cadwell and the two anonymous reviewers for their constructive comments.

References
  • Baldry, Anthony and Paul Thibault (2006). Multimodal Transcription and Text Analysis. London: Equinox.
  • Cambridge Dictionary Online. https://dictionary.cambridge.org/ (consulted 27.04.2022).
  • Catford, John (1965). A Linguistic Theory of Translation. London: Oxford University Press.
  • Condry, Ian (2010). “Dark energy: What fansubs reveal about the copyright wars.” Frenchy Lunning (ed.) (2010). Mechademia 5: Fanthropologies. Minneapolis: University of Minnesota, 193-209.
  • Delabastita, Dirk (1989). “Translation and mass-communication: Film and T.V. translation as evidence of cultural dynamics.” Babel 35(4), 193-218. DOI: https://doi.org/10.1075/babel.35.4.02del.
  • De Linde, Zoé and Neil Kay (1999). The Semiotics of Subtitling. Manchester: St. Jerome.
  • Díaz Cintas, Jorge (2008). “Teaching and learning to subtitle in an academic environment.” Jorge Díaz Cintas (ed.) (2008). The Didactics of Audiovisual Translation. Amsterdam: John Benjamins, 21-38.
  • Díaz Cintas, Jorge (2010). “The highs and lows of digital subtitles.” Lew Zybatow (ed.) (2010). Translationswissenschaft – Stand und Perspektiven, Innsbrucker Ringvorlesungen zur Translationswissenschaft VI. Frankfurt: Peter Lang, 105-130.
  • Díaz Cintas, Jorge and Aline Remael (2007). Audiovisual Translation: Subtitling. Manchester: St. Jerome.
  • Dwyer, Tessa (2016). “Multilingual publics fansubbing global TV.” Katja Lee et al. (eds) (2016). Contemporary Publics. Basingstoke: Palgrave Macmillan, 145-162.
  • Egoyan, Atom and Ian Balfour (eds) (2004). Subtitles: On the Foreignness of Film. Cambridge: MIT Press.
  • Freddi, Maria (2013). “Constructing a corpus of translated films: A corpus view of dubbing.” Perspectives: Studies in Translatology 21(4), 491-503. DOI: https://doi.org/10.1080/0907676X.2013.831925.
  • Gambier, Yves (2006). “Multimodality and subtitling.” Mary Carroll et al. (eds) (2006). Proceedingsof the Marie Curie Euroconferences MuTra: Audiovisual Translation Scenarios (Copenhagen, 1−5 May 2006). https://www.euroconferences.info/proceedings/2006_Proceedings/2006_proceedings.html (consulted 27.04.2022).
  • Georgakopoulou, Panayota (2009). “Subtitling for the DVD industry.” Jorge Díaz Cintas and Gunilla Anderman (eds) (2009). Audiovisual Translation: Language Transfer on Screen. New York: Palgrave Macmillan, 21-35.
  • Gottlieb, Henrik (2009). “Subtitling against the current: Danish concepts, English minds.” Jorge Díaz Cintas (ed.) (2009). New Trends in Audiovisual Translation. Bristol: Multilingual Matters, 21-43.
  • Huang, Boyi (2021). “Subtitlers’ visibilities on a spectrum in the digital age: A comparison of different Chinese translations of The Big Bang Theory.” Renée Desjardins, Philippe Lacour and Claire Larsonneur (eds) (2021). When Translation Goes Digital: Case Studies and Critical Reflections. New York: Palgrave Macmillan, 29-44.
  • Ivarsson, Jan and Mary Carroll (1998). Subtitling. Simrishamn: TransEdit.
  • Karamitroglou, Fotio (1998). “A proposed set of subtitling standard in Europe.” Translation Journal 2(2), 1-15. https://translationjournal.net/journal/04stndrd.htm (consulted 27.04.2022).
  • Luyken, Georg-Michael et al. (1991). Overcoming Language Barriers in Television: Dubbing and Subtitling for the European Audience. Manchester: European Institute for the Media.
  • Nedergaard-Larsen, Birgit (1993). “Culture-bound problems in subtitling.” Perspectives: Studies in Translatology 1(2), 207-240. DOI: https://doi.org/10.1080/0907676X.1993.9961214.
  • Nornes, Abé Mark (1999). “For an abusive subtitling.” Film Quarterly 52(3), 17-33. http://www.jstor.org/stable/1213822?origin=JSTOR-pdf (consulted 27.04.2022).
  • Nornes, Abé Mark (2007). Cinema Babel: Translating Global Cinema. Minneapolis: University of Minnesota Press.
  • Nornes, Abé Mark (2016). Afterthoughts on ‘For an Abusive Subtitling,’ or the Multimodality of Debt in Audio-Visual Translation. Unpublished lecture notes. Stanford University.
  • Orrego-Carmona, David (2011). The Empirical Study of Non-professional Subtitling: A Descriptive Approach. Master’s minor dissertation. Universitat Rovira i Virgili.
  • Orrego-Carmona, David and Yvonne Lee (eds) (2017). Non-professional Subtitling. Newcastle upon Tyne: Cambridge Scholars Publishing.
  • Ortabasi, Melek (2007). “Indexing the past: Visual language and translatability in Kon Satoshi’s Millennium Actress.” Perspectives: Studies in Translatology 14(4), 278-291. DOI: https://doi.org/10.1080/09076760708669044.
  • Pavesi, Maria (2014). “The Pavia Corpus of Film Dialogue: A means to several ends.” Maria Pavesi et al. (eds) (2014). The Languages of Dubbing. Mainstream Audiovisual Translation in Italy. Bern: Peter Lang, 29-55.
  • Pavesi, Maria (2019). “Corpus-based Audiovisual Translation Studies: Ample room for development.” Luis Pérez-González (ed.) (2019). The Routledge Handbook of Audiovisual Translation. Abingdon: Routledge, 315-333.
  • Pederson, Jan (2011). Subtitling Norms for Television. Amsterdam: John Benjamins.
  • Perego, Elisa (2004). “Subtitling ‘culture’ by means of explicitation: Different ways of facing otherness.” Maria Sidiropoulou and Anastasia Papaconstantinou (eds) (2004). Choice and Difference in Translation: The Specifics of Transfer. Athens: The National and Kapodistrian University of Athens, 145-169.
  • Pérez-González, Luis (2007). “Intervention in new amateur subtitling cultures: A multimodal account.” Linguistica Antverpiensia 6, 67-80. DOI: https://doi.org/10.52034/lanstts.v6i.180.
  • Pérez-González, Luis (2012). “Amateur subtitling and the pragmatics of spectatorial subjectivity.” Language and Intercultural Communication 12(4), 335-352. DOI: https://doi.org/10.1080/14708477.2012.722100.
  • Pérez-González, Luis (2014). Audiovisual Translation: Theories, Methods, and Issues. Abingdon: Routledge.
  • Pérez-González, Luis (2020). “Audiovisual translation.” Mona Baker and Gabriela Saldanha (eds) (2020). Routledge Encyclopedia of Translation Studies. 3rd Edition. Abingdon: Routledge, 30-34.
  • Pettit, Zoë (2009). “Connecting cultures: Cultural transfer in subtitling and dubbing.” Jorge Díaz Cintas (ed.) (2009). New Trends in Audiovisual Translation. Bristol: Multilingual Matters, 44-57.
  • Rong, Zongxiao (2015). Hybridity within Peer Production: The Power Negotiation of Chinese Fansub Groups. MSc diss. London School of Economics and Political Science.
  • Saldanha, Gabriela and Sharon O’Brien (2014). Research Methodology in Translation Studies. London: Routledge.
  • Stöckl, Hartmut (2004). “In between modes: Language and image in printed media.” Eija Ventola et al. (eds) (2004). Perspectives on Multimodality. Amsterdam: John Benjamins, 9-30.
  • Tian, Yuan (2011). Fansub Cyber Culture in China. MA diss. Georgetown University.
  • Venuti, Lawrence (2008). The Translator's (In)Visibility: A History of Translation. 2nd edition. Abingdon: Routledge.
  • Zabalbeascoa, Patrick (2008). “The nature of the audiovisual text and its parameters.” Jorge Díaz Cintas (ed.) (2008). The Didactics of Audiovisual Translation. Amsterdam: John Benjamins, 21-38.
  • Zhang, Weiyu and Chengting Mao (2013). “Fan activism sustained and challenged participatory culture in Chinese online translation communities.” Chinese Journal of Communication 6(1), 45-61. DOI: https://doi.org/10.1080/17544750.2013.753499.
Biography

Boyi Huang is currently a PhD researcher of Translation Studies at the School of Applied Language and Intercultural Studies in Dublin City University. His current research centres on the role of translation in online LGBT+ subtitling communities. He is particularly interested in identity, motivation, and community translation and interpreting.

huang_portrait


ORCID: 0000-0002-6306-877X

E-mail: huangb3@mail.dcu.ie

Notes

Note 1:
Headnote is a subtitle notation technique to add notes near the head/top of the screen in addition to the ‘normal’ subtitles that are usually at the bottom space.
Return to this point in the text

Note 2:
In this paper, the two concepts of corrupt and abusive as part of the framework are only used to theoretically describe subtitling approaches and analyse subtitler visibility, without any intention to judge or criticise any subtitlers with values that might have been originally attached to these terms. In the Afterthoughts, Nornes (2016) attempts to disavow and hide the violent connotation of his original concepts by using two new terms, which seems, in his own word, to be a “corrupt” act.
Return to this point in the text

Note 3:
All the examples in charts are provided by the author.
Return to this point in the text

Note 4:
All English descriptions in brackets following Chinese characters are the author’s back translations.
Return to this point in the text