Anglophone culture through the Western Balkan lens: a corpus-based study on the strategies used for rendering extralinguistic elements of culture in Montenegrin subtitling Círculo de Lingüística

. One of the most demanding aspects of intercultural communication is rendering the elements of foreign culture from source to target linguocultural system. As today more people are exposed to audiovisual than any other form of text, cultural representation is an important issue as these texts can be a powerful tool for construction or deconstruction of cultural and sociological stereotypes, dissemination of ideas, facilitating intercultural communication, etc. Hence, this study attempts to shed light on how the Anglophone culture is rendered in subtitling and what could be some underlying reasons for that. To this end, we have constructed a morphosyntactically annotated parallel English – Montenegrin corpus of TV subtitles consisting of 110 episodes of three different TV series that were broadcast on the public service broadcaster of Montenegro and scrutinized it to map the strategies used for rendering extralinguistic elements relying on the analytical framework proposed by Pedersen (2011). results of a corpus-based study of the transfer of extralinguistic elements of from English into Montenegrin language in subtitling. our is the first study of for the region of the Western Balkans as some recent studies focused on the and of towards the transfer strategies used and and not on how the transfer is done (Božović 2019).


Introduction
The process of translation does not entail the mediation between languages only, but cultures as well. This necessity stems from the fact that each language reflects the social realities of communities using it. These include the values, tradition development, needs, symbols, and all the other aspects of the culture whose attributes the language takes, defines, preserves, and, sometimes, attempts to change (Gorjanc, 2012). In such a way, not only does the culture shape the lexicon and grammar of the language (Martínez, 2015), but it creates its specific framework for the interpretation of its elements and translation has the mediating function between these different linguocultural frameworks. Hence, it is no surprise that translation of the elements of a foreign culture and transfer of their denotative, connotative, associative (social, affective, collocational) and other levels of meaning is one of the most challenging tasks for translators.
These issues are especially relevant for the audiovisual translation (AVT) industry which includes such modalities as subtitling, dubbing, voice-over, audio description, free commentary, etc. It is estimated that an average person in the EU watches TV approximately 3 hours a day and that watching a 90-minute subtitled film equals to reading a 200-page long novel each month (Gambier, 2009). If we take into account other TV forms such as series, shows, documentaries, realities, the Internet content, and the accessibility of audiovisual materials through portable devices, it is evident that today more people are exposed to audiovisual texts than any other form of text. Furthermore, scientific studies have identified the influence of the AVT modalities on the perception of foreign cultures (Nikolić, 2011). Hence, all of this makes audiovisual texts, and translations, a powerful tool for the dissemination of messages and can have an important role in constructing or deconstructing cultural or social stereotypes, facilitating intercultural communication and sensitivity, promotion of ideologies, molding attitudes, fostering societal cohesion, and driving language change.
Several studies have been conducted exploring the transfer of elements of culture in audiovisual texts for several regions and different language pairs, among which one of the most influential ones has been a study by Jan Pedersen on subtitling norms in Scandinavia (Pedersen, 2011). Having in mind the importance of the above-mentioned issues, 1 University of Montenegro. E-mail: petarb@ucg.ac.me ARTÍCULOS Círculo de Lingüística Aplicada a la Comunicación ISSN: 1576-4737 http://dx.doi.org/10.5209/clac.73536 this paper presents the results of a corpus-based study of the transfer of extralinguistic elements of culture from English into Montenegrin language in subtitling. To our knowledge, this is the first study of this nature for the region of the Western Balkans as some recent studies focused on the preferences and attitudes of end-users towards the transfer strategies used in subtitling and dubbing and not on how the transfer is done (Božović 2019). In the following sections, first, we will briefly discuss the specifics of audiovisual texts and modality constraints in subtitling, followed by an overview of the audiovisual translation landscape in the Western Balkans. After this short contextualization, in section 3, we will describe the theoretical framework, methodological approach, and the corpus used for our research. Section 4 will outline the results of the study followed by a discussion and conclusion.

The complexity of AVT and an overview of practice in the Western Balkans
The specifics of the AVT are closely connected to the type of text that is being translated and modality constraints. Unlike many other types of texts, audiovisual ones are polysemiotic in nature. This means that the meaning is created through a synergy of different code channels which include the visual and acoustic component. Chiaro (2009) notes that the complex visual component can consist of its non-verbal and verbal elements. The non-verbal element could include scenery, gestures, costumes, and verbal could include street signs, advertisements, written messages, etc. The acoustic component could also consist of its verbal and non-verbal elements (see Figure 1).  The verbal element could include the dialogues, songs, etc., while the non-verbal could include background music, sound effects, noise, as well as other paralinguistic features (pitch and tone). This shows that it is necessary to take into account and decode the meaning of all these contextual factors that make up the audiovisual situation while translating the verbal component, otherwise it could result in an unsuccessful communicative transfer. For instance, oftentimes the background music, facial expression, tone, and pitch can indicate whether an utterance is to be translated with the elements of humor, sarcasm, irony, etc. This will ensure "a successful triangular marriage between words, acoustic and kinetic information" (Díaz Cintas, 2009).
As one of AVT's most prominent modalities, subtitling is commonly described as "a vulnerable modality" (Díaz Cintas & Remael, 2007). There are multiple reasons for this. Unlike other forms of translating, subtitles are unique in that the target readers, or rather viewers, are simultaneously exposed to both the source and the target text. Because of this, they tend to be scrutinized by the audiences who have some knowledge about the source language and are, thus, occasionally rather judgmental about the translator's decisions. Another factor that contributes to subtitles' "vulnerability" is the space and time constraints. These result from the necessary synchrony of subtitles with other code channels (sound, image). Ideally, a subtitle should appear at the precise moment a character starts speaking and disappear when the utterance is over. Depending on the audiovisual situation, a subtitle normally consists of 32 -41 characters per line and in a maximum of two lines with the maximum recommended exposure time of 6 seconds (Diaz Cintas & Remael, 2007). This makes it possible for average viewers to read, process the information, and follow the action on the screen. Additionally, unlike dubbing and voice-over that are examples of isosemiotic translations (using the same semiotic channel as the original, in this case, acoustic), subtitling is a form of diasemiotic translation (using different semiotic channels, in this case, speech in the source language and written text in the target language) (Gottlieb, 2005). This semiotic shift, as well as the already mentioned technical, constraints, calls for the use of omission and/ or condensation of some lexical items that are used in the source text and original information in general. Arguably, these translation strategies would be considered illegitimate in some other forms of translations. This, however, is not the case in subtitling.
All of the above mentioned creates some translation challenges that are not typical of some other forms of translations. For instance, while in literary translation there is ample space for the annotation, and additional clarifications of culture-specific elements, subtitling, in its most typically used form, gives no room for that. Additionally, while in the case of literary translations readers can take their time while reading and processing the information, in subtitling the information normally needs to be processed under limited time and at the same time be informative enough to reach the translation skopos, i.e. entertain, amuse, etc.
The European audiovisual translation landscape has traditionally been divided into two major blocks. Depending on the most preferred predominant modality used in AVT, these include the subtitling countries (e.g. the Scandinavian countries, Benelux, Portugal) and the dubbing countries (the so-called "FIGS" countries including France, Italy, Germany, Spain). When it comes to the Western Balkans, the audiovisual translation culture would place them in the former group as subtitling is, by far, the most common AVT modality. Dubbing is used in cases of cartoons and shows made for children (but not all), and voice-over typically in the cases of some documentaries or news reporting. However, regarding all the other media content, it is subtitled in most cases. The languages spoken in the region include, among others, the former Serbo-Croatian polycentric language's recodified standards: Bosnian, Croatian, Montenegrin, and Serbian and, thus, the audiovisual content is translated in one of these standards. National broadcasting companies translate the AVT content in the official language of the country. However, there is a specific situation in Montenegro where Radio and Television of Montenegro (RTCG), the national public service broadcaster, has been using the translations in three different standards: Serbian, Croatian and Montenegrin, as the media content is often bought together with the translation in Serbian or Croatian. Due to a high level of intelligibility between these different standards, and a tradition of use on the AVT market, this was probably seen as a convenient way of reducing the costs. However, in recent years, especially after Montenegro became independent in 2006 and the language standard underwent the recodification and standardization (which remains an on-going process in Montenegro), an increasing number of media content to be televised on the RTCG has been subtitled into Montenegrin. These translations were used for the corpus that was scrutinized for the present study, as we will present later.

Theoretical framework and methodology of research
The present research falls within the descriptive translation studies (DTS) theoretical framework which is focused on the quantitative and qualitative empirical inquiry within the translation studies field. The DTS aim to describe and explain the general norms or tendencies in translation practice and can be focused on the translation product, and translation process (Toury 1995).
The methodology that we used in this study is a corpus-based one. By using the methods and resources of corpus linguistics, corpus-based translation studies (CBTS) have been a thriving approach within the discipline ever since the methodology was officially introduced back in 1993 by Mona Baker in her seminal paper Corpus Linguistics and Translation Studies: Implications and Applications. The CBTS within the AVT is mainly focused on regularities, irregularities, and variability, as well as innovation in the translation process and products for various AVT modalities (e.g. the authenticity of the language of dubbing, the sociolinguistic and pragmatic parameters of the language of dubbing, the narrative aspects of audio description) (Romero-Fresco, 2006Freddi, 2013;Hurtado and Soler, 2013;Špela, 2017). At the applied end, the CBTS contribute greatly to developing machine translation systems, translation memories, and other translation resources, translation quality assessment, translation training, etc. The most commonly used resources within this methodological approach are parallel and comparable corpora. The former consist of source texts and their translations, and the latter of texts in source and target languages that belong to the same domain or register (legal documents, journal articles, literature, etc.) but are not translations of the source texts.
The corpus that was scrutinized for this study is Opus-MontenegrinSubs 1.0. (Božović et al., 2018;Božović 2020). This is the first electronic English-Montenegrin parallel corpus of subtitles that were broadcast on the RTCG. It consists of the English source text and subtitles in Montenegrin of three TV series: House of Cards, Damages, and Tudors. This is 110 episodes in total, with the duration of app. 5,563 minutes. The Montenegrin part consists of 463,298 tokens, 365,698 words, and the English part consists of 595,391 tokens and 468,337 words. The corpus is fully morphosyntactically annotated and lemmatized. It was aligned by using the time method (Tiedemann, 2007). It is available on online concordancers via CLARIN.SI platform on the KonText and (no)SketchEngine, as well as the Sketch Engine platform.
The focus of the present study was to analyze the transfer of the extralinguistic elements of culture from English into the Montenegrin linguocultural system. These are the elements that refer to realia that are outside the language system and include the personal names and names of institutions, agencies, geographical names, events, real-world, historical or fictional characters, artistic items, folklore, religion and customs, the world of entertainment, etc. (Pedersen, 2011). We did not focus on the intralinguistic elements, those pertaining to the language system itself (idioms, slang, dialectal variations, proverbs, and sayings, etc.). For mapping the transfer strategies and the classification of elements, we relied on the analytical framework proposed by Jan Pedersen (2011). It consists of the following strategies: • Low-intervention strategy: retention (oriented towards the source text and preserving its foreignness) • Mid-intervention strategies: specification and direct translation (oriented towards the source language), generalization (oriented towards the target language) • High-intervention strategies, substitution, and omission (oriented towards the target language and adapting it to the target linguocultural norms) • Non-intervention strategy: official equivalent (a pre-fabricated solution that does not pose a translation challenge) According to the parameter of transculturality, the elements were classified into transcultural and monocultural (Pedersen, 2011). The former refer to those elements that are part of the encyclopedic knowledge of both the source and target language viewers. Examples of such elements include the White House, the Pentagon, FBI, CNN, Clinton, Last Judgement, Wall Street, Coca Cola, etc. As these elements do not pose a translation challenge they were not the primary focus of our study. The latter elements refer to those which are known in the source but not in the target language and culture and as such are most likely to pose a translation challenge. The examples of such elements include the Sentinel, Walmart, June Cleaver, Webelos, PBS, ICU, VA, GLAAD, the Crimson, Hudson General, Pebble beach, US Weekly, AT&T, etc. We will provide further information about these elements in the section below, but they would require an additional explanation as they are not part of the common encyclopedic linguocultural knowledge. It goes without saying that this classification could vary depending on how big the intercultural asymmetry is between the source and target culture. Determining the degree of transculturality is often done intuitively, or with the help of online or some other type of research. (Pedersen, 2011) The analysis was carried out in the following way: 1. The extralinguistic elements were identified in the source text and classified according to the level of transculturality; 2. The translations of the identified elements were found in the target text through KonText and (no)Sketch Engine concordancers. This enabled a quick search and visualization of all translations used for one element in the whole corpus and to easily spot any variations; 3. The translation strategy was identified according to the suggested taxonomy.
It is important to mention that all the elements were extracted as types and not tokens. This means that if one particular element reappears multiple times in the corpus, and is rendered with the same translation strategy, it was extracted only once. In this way, we extracted 797 extralinguistic elements, out of which 549 are monocultural and 248 are transcultural. These could further be divided into extratextual and intratextual (Pedersen, 2011). The former include the realia that are part of the real-world outside the text, and the latter the realia constructed within the text itself and non-existent in the world outside the text (e.g. characters, fictitious places, etc.).

Results and discussion
The results suggest that the extralinguistic elements of culture (both transcultural and monocultural) are mostly rendered through low-intervention or non-intervention strategies which include retention which was used in 397 cases (49,8%), and official equivalent which was used in 110 cases (13,8%). The mid-intervention strategies are used in a rather equable way. This is especially the case with the direct translation which was used in 98 cases (12,2%) and generalization, which was used in 71 cases (8,9%). After that, the most frequently used strategy is substitution (5,8%), omission (4,7%), and specification (4,5%). This is represented in graph 1 below which shows that the overall results seem to suggest that translators mostly use source text-oriented strategies. In translation studies, this approach is commonly referred to as foreignization, while the orientation towards the target-language is commonly referred to as domestication (Venuti, 1995). The ratio between these two orientations in subtitling for the present corpus is given in graph 2.  A closer look at the monocultural elements which were the primary focus of this study also seems to suggest that low-intervention strategies are the ones that are most frequently used in the transfer. Retention is dominant here as well (53,8%), and direct translation (10,6%) and generalization (10,6%) were equally used. These are followed by the substitution (6,6%), omission (5,3%), specification (5,1%), official translation (3%). We will now give a closer look at the transfer of monocultural elements and each strategy of transfer with illustrative examples and a brief discussion. When giving examples, the following convention is used: first we will give the source text in English, then the target text in Montenegrin, followed by a back-translation in English. In cases of literal translation (mostly retention and direct translation), the back translation will not be provided. The element which is in the focus of our discussion will be in bold. After each example, the source will be given in the following fashion: in the right corner, we will give an abbreviated title of the series in English with the information about the season and the episode. For instance, the abbreviation HoC 3: 18 means that the given example is taken from the series House of Cards, season 3, episode 18. The abbreviation for the series Damages will be Dam, and for Tudors Tud. Similar abbreviations are used for the metadata in the electronic corpus that was used as the basis of this study. Sometimes, it will be clear that more than one strategy is used when rendering certain elements.

Retention
As noted by Pedersen (2011), retention can have two of its subcategories: complete (the source element is retained without any adjustments) and target language adjusted (the source element is adjusted slightly to meet the requirements of the target language in terms of spelling, case forms, pronunciation, articles, etc.).
The most frequently used form of retention in this corpus was the adjusted one, especially in cases of proper names of people, stores, companies, magazines, geographical names, etc. (examples 1, 2, 3 below). This does not come as a surprise because it is a norm in Montenegrin to translate the names phonetically, adjusting them to the local pronunciation. This is not the case in Croatian, for instance, where the original spelling of names is preserved very often. Furthermore, Montenegrin renditions are adjusted to the language norms as well, especially in terms of case endings. Unadjusted retention is to be found in cases of certain abbreviations (example 4 below).
Dame i gospodo, moj muž, predsjednik Frenk Andervud! Ladies and gentlemen, my husband, President Frank Underwood! (HoC 3:13) (2) The starting salary for an employee at Walmart is below the poverty line.
The starting salary at Wallmart is below the border of poverty. In the example (4), the medical acronym is used which stands for an automatic external defibrillator, a medical device used for diagnosing cardiac arrhythmias. Even though it is a specialized term, it is retained as the same acronym is used in Montenegrin. The use of retention, in cases such as the above mentioned, that include the names of supermarket chains, sights, etc. unknown to the most of the target text readers, seems to be justified because the nature of these elements can be deduced from contextual factors. Even though these elements are devoid of connotative meaning and not much is given to make them fully transparent, this doesn't seem to pose a problem as these elements are of peripheral importance for appreciating the audiovisual text. The use of retention might be problematic in cases of those elements that are central in the audiovisual text and its meaning cannot be deduced from the context and polysemiotic interplay.

Specification
The specification can be realized in two ways: through addition (by adding a piece of information to an element which makes it more accessible), and explicitation (by spelling out the abbreviations, acronyms, etc.) (Pedersen, 2011). The results show that this is the least frequently used of all the strategies in subtitling. This could be explained by the modality constraints which limit the availability of space that could be used for providing additional information. It is most frequently used in cases of abbreviations, acronyms, names of institutions, associations, agencies, etc. And GLAAD wants you to be a keynote speaker at their Media. Gej udruženje GLAAD poziva Vas da govorite na dodjeli nagrada. Gay association GLAAD invites you to speak at the awards ceremony.
(HoC 3:7) The example (5) illustrates the use of the explicitation as it spells out the acronym AFL which stands for the American Federation of Labour. It could be argued, however, that explicitation is achieved here through the use of another strategy -generalization as the precise wording of the element was not central to the scene. Examples (6) and (7) illustrate the use of addition as the information is added to make the element more accessible. In both cases, this is achieved in combination with retention.

Direct translation
Direct translation has an equal frequency of use along with other mid-intervention strategies. It could be argued that it is a strategy that can find an easy application in subtitling because of the modality constraints. It can be found in the form of calques or shifts resulting from the differences between language systems and it presents a link between foreignization and domestication-oriented strategies (Pedersen, 2011).
It is most frequently used in translations of proper names, or names of legal documents, geographical names, associations. It is sometimes used in combinations with other strategies, especially omission.
(8) The Stafford Act gives me full authority to define an emergency.
Stafordov zakon mi daje ovlašćenja da definišem vanredno stanje. The Stafford Act to me gives authority to define an emergency. (HoC 3:5) (9) There is precedent with the Uniting for Peace Resolution.
There is precedent with resolution United for Peace. (HoC 3:3) (10) Three years ago, tried in Munich Regional Superior Court.
The case three years ago in front of the Munich Regional Court.
(Dam 5:8) The example (9) illustrates the use of a shift as the word order is adjusted to the target language syntactic norms. The example (10) illustrates the use of omission in combination with the direct translation as in this case the qualificator superior is left out. This practice could be legitimate because of the modality constraints, especially in cases of a transparent element of peripheral importance.

Generalization
Among the domestication-oriented strategies, generalization is the most frequently used one in the present corpus. The primary reason seems to be the need to make the element transparent, and thus, accessible to the audience, while the secondary reason seems to be the need for condensation due to time and space constraints. It can appear in two of its sub-categories: by using a hyperonym and paraphrase. It was identified in cases of names of state agencies, institutions, events, magazines, etc.
(11) The Maryland State Police and Federal authorities have yet to locate the car.
Policija i FBI nijesu našli auto. The police and the FBI haven't found the car. (HoC 3:1) (12) You think I would have turned over all the AT&T records?
Misliš da sam predao telefonske podatke? You think that I turned over the telephone data? (HoC 3:1) (13) Sir, we're going to ask that you wear your Kevlar and helmet.
Zamolićemo Vas da nosite pancir i šljem. We will ask you to wear a vest and a helmet. (HoC 3:10) (14) Eight-year-old Gl Joes with assault weapons, walking around.
Mislila sam da ćete ga Vi zvati iz aviona. I thought that you would call him from the plane.
U restoranu nijesmo. At the restaurant, we didn't.
(Dam 1:1) Some of the above-mentioned elements require brief comment. Example (12) mentions the records of the American telecommunication company, AT&T. As most of the viewers are not familiar with the company in the target linguocultural context, the translator used a more general term telephone data. Example (13) mentions the Kevlar, a brand of aramid fiber that is not damaged by high temperatures and which is used for making bulletproof vests. For the same reasons mentioned above, it is rendered simply as a vest. Example (14) makes a reference to GI Joe, the line of action figures representing different branches of the U.S. armed forces, which is rendered in Montenegrin as a soldier. Example (15) mentions the aircraft of the President of the United States which is also known as Air Force One. This element is rendered as a plane. Although there were cases in the same corpus where the same element was rendered with a qualificator predsjednički avion (president's plane) the context of the statement makes the element transparent even without the qualificator. Finally, the American restaurant chain Quiznos is rendered as a restaurant in the example (16). These and similar examples from the corpus suggest that hyperonymy is the most frequently used sub-category of generalization in subtitling, which, due to the modality constraints, enables faster processing of information and the use of fewer characters.

Substitution
Substitution can be realized through two of its sub-categories. As Pedersen (2011) pointed out, it can be cultural, in which case a transcultural or target-language element might be used to replace a monocultural one. Furthermore, it can also be situational in which case the source monocultural element might be replaced with any element that fits the given situation. Example (17) illustrates the use of a colloquial expression, Elephants, for the United States Republican party whose traditional mascot is an elephant in red, blue, and white color. As this expression is unknown among most of the target language viewers, the translator rightly substituted it with a transcultural expression. Another unknown element among the target culture viewers is June Cleaver (18), a character from the American sitcom Leave it to Beaver which was televised from 1957 to 1963. She has become an archetypical symbol of a housewife and a mother in the source American culture. The translator used a situational substitution and rendered it with a housewife. In the third given example (19), one of the characters uses the exclamation Jesus, Mary, and Joe Cocker which creates the effect of humor in the source text, especially because of the mentioning of British rock and blues singer John Robert Cocker. The translator used a situational substitution and rendered it as Lord god, which comes at the expense of the skopos, i.e. recreating the effect of humor in the target text. The last example (20) mentions the North American emergency phone number which is used for many different services, including the police. In line with the context of the scene, it was rendered as police.
Substitution is among the strategies that possibly highlight individual translator's creativity the most. Its use, however, requires a careful examination of possible advantages and drawbacks, especially with regards to the reception of audiovisual texts.

Omission
As pointed out before, while in some other forms of translating leaving out of textual elements might be problematic, the omission is a legitimate strategy in subtitling. Depending on the centrality of the element, a translator will most frequently decide to leave out a peripheral element or the one that has become such due to frequent repetitions in the audiovisual text.

Official equivalent
The official equivalent may represent any of the previously mentioned strategies the result of which has become the official translation of the given element which, as such, does not represent a challenge for translators. Most frequently, it is a form of retention, direct translation, or cultural substitution. However, the number of official equivalents for monocultural elements is not large as by definition they are unknown in the target culture. They appear more frequently among the transcultural elements. The examples can be found in areas such as health care (official translations of medicines), titles, institutions, geographical names, etc.

Conclusion
This corpus-based study of the transfer of extralinguistic elements of Anglophone culture from English into Montenegrin suggests that there is a general tendency to use the foreignizing, low-intervention, strategies for the transfer of all, and especially monocultural, extralinguistic elements from a source into target linguocultural system in subtitling. Among these strategies, it is especially evident the frequent use of retention. Among the mid-intervention strategies, the use of direct translation and generalization is equable. At the same time, generalization is the most frequently used domestication-oriented strategy. The underlying reasons for such a state of affairs are to be found among different influencing factors. Among the most prominent ones seems to be the centrality of the element in question regarding the audiovisual text. The more peripheral the position of the element is, i.e. the less important for understanding and following the plot and appreciate the humor it is, the more frequently will it be left out or rendered with some of the foreignization-oriented strategies. This seems to be an especially convenient solution in situations that require condensation because of the high density of subtitles and modality constraints. On the other hand, the more central the position of the element is, i.e. the more significant its role is in understanding and following the plot and appreciating the humor, the more frequently will the mid-intervention (direct translation, generalization) and high-intervention (substitution) translation strategies oriented towards domestication be used.
Another influencing factor seems to be the dialogue density which calls for the use of a greater number of subtitles and characters, and less time of visibility on screen. The more demanding the text segment in this regard is, the more frequently will omission be used in cases of peripheral elements. On the other hand, in case of a lesser density, a peripheral element will more frequently be preserved and transfer (most frequently) made by retention. This is shown in Figure 2, where the arrow shows the direction of the increase of the dialogue density and complexity of time and space constraints of the audiovisual situation. The "P" stands for the peripheral position of the element within the audiovisual text, while "C" stands for its centrality. Below that, translation strategies are given as well as the general tendency of their use according to the parameter of centrality and density. The official equivalent is not given as a strategy because the elements which are rendered in this way do not pose a challenge for translators.

P
C P

Re
Spec DP G Sup Om Figure 2. The use of strategies for rendering the extralinguistic elements of culture concerning the parameter of centrality and dialogue density Another influencing factor in the treatment of the elements of culture is the degree of transculturality. The greater the degree, i.e. the more the element is shared by two liguocultural systems, the more frequently will official translations be made by using retention, direct translations, or, in some cases, substitution. The "vulnerability of the modality" is to be taken into consideration as well (Diaz Cintas & Remael 2007) as one of the factors that most certainly contribute to the domination of foreignization-oriented strategies in the transfer. As the recipients are simultaneously exposed to the source and target text, they are more prone to judging the quality of translation based on the fidelity to the source text principle, which most probably influences the degree of freedom a translator might feel to have during the translation process. This could account for such a high frequency of use of retention.
Finally, the translator profile should be mentioned as one of the key influencing factors for the treatment of the elements of culture. Experience, the level of education, expertise, working conditions, deadlines, sensitivity, etc. are just some of the factors that should not be overlooked when analyzing the translator's decision-making process, and especially the ones which were the focus of the present study.
The results of the study carried out in Montenegro have confirmed some of the conclusions reached by Pedersen in his study for the Scandinavian region (Pedersen, 2011). As a way forward, it would be useful to extend the study to other parts of the Balkan region and other languages (Bosnian, Croatian and Serbian), make a comparative analysis to find possible norms or even a wider cross-cultural study to unveil possible translation universals in the field of AVT. Moreover, the present study has been focused on a specific genre of the audiovisual text (TV series). As translation approaches and the cognitive operations underlying them might vary from genre to genre (Peña-Cervel 2016), studies focussing on different genres could yield different results.