When East Meets West: A Comparison of Audio Description Guidelines in China and Europe

Despite China being the country with most disabled people in the world, the provision of media accessibility services and their academic study are still in their infancy. The present study set out with two objectives: to present a sample of three Chinese audio description guidelines that are mainly employed for training purposes, and to compare the Chinese situation in terms of guidelines with that of Europe, to see if there are any relevant differences to point out and any mutual lessons to be learnt. Our results show there are a number of general similarities between the different Chinese and European guidelines and that the main difference is the different approach to objective vs. subjective description. While Europe is moving away from the dichotomy When East Meets West: A Comparison of Audio Description Guidelines... 168 Hikma 19 (1) (2020), 167 186 between objectivity and subjectivity, China still advocates objectivity as being a necessary characteristic of quality audio description.


INTRODUCTION
China is the country with the most people with disabilities in the world (Wu & Xie, 2015) but, paradoxically, media accessibility is still in its infancy. 1 It is a fact that most Chinese TV programmes and films are captioned, but the reason behind this is more of a linguistic rather than an accessibility-related nature: the wide range of Chinese dialects requires captions to guarantee that everybody can follow the content that is broadcast (Li & Looms, 2015;Casas-Tost & Rovira-Esteva, 2018: 33). Zheng (2017) even argues that captions have more of an aesthetic function in many entertainment, cultural and children's programmes in order to attract the viewers' attention. Yet China is not oblivious to the needs of people with hearing loss. Indeed, "[p]ersons with hearing impairments are quite well served at national level with open captions/subtitles with the exception of live content such as news. Sign language interpretation is available to a more limited extent, mainly at local and regional level" (Li & Looms, 2015: 270).
Audio description (AD), however, is not available on Chinese TV. This stands in stark contrast with, for example, Europe, where AD is being offered on TV in various countries. In the UK "most major broadcasters (like the BBC, Channel 4 and Sky) provide AD on 20 per cent of their programmes" (RNIB, 2019), and Germany has been offering AD on the TV since 1993 (Reviers, 2016: 239). 2 In China AD has been offered in films through live sessions and recorded CDs since the early 2000s, and more recently through online platforms such as Ximalaya FM and Youku (Tor-Carroggio & Casas-Tost, 2020). Yet, the provision of the service is still far from ideal and many challenges are yet to be responded to, such as the lack of financial resources and the irregular provision of the service across the country (Li & Pan, 2013).
Similarly, the academic study of AD in China is still in its infancy and the literature mentioning it is very limited. Only a handful of scholars refer to the situation in Mainland China (Yeung, 2007;Li & Pan, 2013;Leung, 2015; Hikma 19 (1) (2020), 167 -186 Wu & Xie, 2015;Feng, 2018), and most of them only do so superficially. One of the reasons for this could be that accessibility-related studies still lack a proper descriptive framework that contextualises the service and provides a clear picture of how it is offered in China. Notwithstanding, there have been some recent efforts to take a more comprehensive snapshot of AD in China. For example, Tor-Carroggio and Casas-Tost (2020) have drawn the profile of Chinese audio describers and Tor-Carroggio (2020) has investigated Chinese users' satisfaction and experience with AD. Yet, the existing Chinese AD guidelines have not been subject to study thus far, although they have taken a special role in improving the quality of the service. A comparison with Europe shows that, again, the situation is completely different there. As Orero (2012: 196) notes: [t]he issue which has attracted the most attention [in media accessibility studies in Europe] is the analysis of existing guidelines, focusing on the research needed to draft new research-based guidelines or standards, with an eye to using them for training at university level.
Another reason that could explain this interest is that standards are a prerequisite for laws, since a law is meaningless if there is no standard to refer to (Matamala & Orero, 2018: 142). Having said that, it would be advisable to make a terminological remark, since standards are different from recommendations or guidelines. In fact, there are Chinese AD guidelines, but no official standard has been agreed on yet.
This paper aims to contribute to the recent wave of interest in Chinese AD by providing a new perspective from which to analyse the current state of the service, namely that of existing guidelines. More specifically, the objective of this study is twofold. First, to present three unpublished Chinese AD guidelines that mainly serve for training purposes. Second, to compare the Chinese situation in terms of guidelines to that of Europe, to see whether there are any relevant differences to point out and any mutual lessons to be learnt. This comparison was deemed necessary and potentially fruitful given the fact that Chinese AD has developed relatively free from European influence, and vice versa. Since both China and Europe share one same final goal in this regardoffering a quality service to people with sight loss -, we believe this comparison is a natural step to see where each region stands and what lessons they can learn from each other. This paper is divided into five sections. The first revises the current situation of European AD guidelines; the second describes the methodology followed in the study; the third presents three different Chinese AD guidelines; the fourth discusses the differences and commonalities between the Chinese and European cases; and the last one draws some conclusions, underlines the limitations of the study and provides some ideas regarding what other paths of research could be explored in the future.

EUROPEAN GUIDELINES AND STANDARDS
Compared to China, Europe has a much longer-standing tradition when it comes to creating guidelines for AD. Dating back to 1997, Benecke and Dosch published Wenn aus Bildern Worte Werden. Ein Handbuch fur Filmbeschreiber, the German guidelines that formed the blueprint for many later documents that offer guidance to audio describers. In the first decade of this century, various other countries created their own guidelines: in 2001, the UK's communications regulator Ofcom published their Guidelines on the provision of TV access services, an updated and reviewed version of which was presented in 2006; in 2005, the Spanish Association for Standardisation and Certification (AENOR) created the UNE 153020 norm Audiodescripción para personas con discapacidad visual. Requisitos para la audiodescripción y elaboración de audioguías, and in 2008 France got its Charte de l'audiodescription (Morisset & Gonant, 2008). 3 Generally speaking, all these guidelines discuss the same categories of information, namely content selection or 'what to describe', AD style or 'how to describe', timing or 'when to describe' and more or less individual specific issues.
In terms of content selection, all guidelines agree that four constituents should be described, namely the actions, the characters performing and the time and place in which they take place. In addition to all this visual narrative information, most of the guidelines point out that some (intradiegetic) sounds, which cannot readily be identified by the target audience, should also be included in the AD, just like opening and ending credits and other types of text on screen, including logos, written messages such as indications on screen of when and/or where a certain movie or scene is set, and subtitles (OFCOM, 2001: 5). 4 One of the few elements where the existing European guidelines do not seem to be in agreement, is as to when characters should be named. The German guidelines suggest that characters should only be named when their name is also mentioned in the film (unless they are historical characters or their name is only given later in the film). The Ofcom guidelines, on the other hand, suggest that for practical reasons the name of a character can be given as early as possible, unless it should remain a secret for the plot (Greening, Petré & Rai, 2010: 6).
When it comes to the style of the AD, they all contain a few specific instructions on how to describe, but none of them are really exhaustive. 5 In summary, the main recommendations advise using fluent and simple language (bearing in mind the AD will have to be voiced later), adapting the description to the genre and the pace of the film, using vivid and varied language including adjectives and adverbs that allow the audience to create its own image, and avoiding patronising and subjective descriptions.
Finally, all the guidelines seem to be in agreement regarding the timing of the AD: generally, descriptions cannot interfere with dialogues and relevant sound effects. The German guidelines go a bit further and state that description should only be added when the production is completely silent (Benecke & Dosch, 2004: 20), but since that never really happens, descriptions can go over (parts of) sound effects that are not essential for the story.
These guidelines have proved to be invaluable for the practice of AD, which would not be at the level it is today without them. However, from a translation studies perspective, they all have some common shortcomings: nearly all of them were drawn up by practitioners who learnt the tricks of the trade on the job. 6 As such, they are fairly intuitive and anecdotic, based on personal experiences, andprobably more problematic from an academic point of viewthey lack a solid methodological foundation. We are well aware of the fact that none of these guidelines were ever designed as an academic publication, so the foregoing is by no means meant to criticise them. Nonetheless, as pointed out by Vercauteren (2016: 74): "if we claim that audio description is a form of translation, it only seems reasonably that we try to apply paradigms used for other types of translation to audio description too". The paradigm suggested by Vercauteren (2016) is that of functionalism. It would go beyond the scope of the present paper to explore this choice in detail, but two elements that are relevant in the light of guideline creation are that functionalism requires translators a) to give a detailed analysis of their source text to determine what to translate and how, and b) to decide what strategies they will use to create their translation. These two steps in the translation, c.q. AD process formed the basis of the guidelines developed by the ADLAB project (Remael, Reviers & Vercauteren, 2015), which differ from earlier guidelines in two essential respects. First, they fit the entire AD creation process within the theoretical framework of narratology, which offers audio describers a generally applicable approach to determine and decide what and how to describe. 7 Indeed, narratology defines the various basic elements or constituents that can be found in any type of narrative (see for example Bal (1997), Herman and Vervaeck (2005) or Herman (2009) for a comprehensive overview of those basic elements). In other words, narratology explains how stories are created and as such offers audio describers a tool to analyse their source text and determine what narrative elements can be included in their description. On the other hand, narratology also explains how audiences process and interpret stories: it offers insights into what elements audiences really need to understand and follow a story and what elements are less crucial. Therefore, it presents audio describers with a tool to decide what elements should be included in the AD and what elements can be left out if there is not enough time to describe everything. This last part in particular is absent from earlier guidelines, which recommend prioritising some information over others, without explaining how to do that.
A second element that sets the ADLAB guidelines apart from earlier documents, is that it steers clear of the discussion between objective and subjective description, by pointing out that "no one ever sees the same film […] and this is no different for the blind and visually impaired audience since it is just as heterogeneous as the sighted one." (Remael, Reviers & Vercauteren, 2015: 16). Therefore "AD too is always subjective to some extent since it is based by [sic] the interpretation of the audio describer." (Remael, Reviers & Vercauteren, 2015: 16). So rather than prescribing that descriptions have to be objective or subjective, the guidelines advise that the audio describer has to try and find: "a balance between a personal interpretation and personal phrasing (subjectivity) and more text-based interpretation and phrasing (objectivity) that leaves room for further interpretation by the blind and visually impaired users." (Remael, Reviers & Vercauteren, 2015: 16).
To achieve that aim, the ADLAB guidelines offer various strategies that audio describers can choose from, ranging from the more objective "her eyes open wide" to the more subjective "she is amazed", or a combination of these: "her eyes open wide in amazement" (Remael, Reviers & Vercauteren, 2015: 17). This principle also brings the AD process closer to that of other forms of Hikma 19 (1) (2020), 167 -186 translation, that are never 'or -or' but always a matter of degree depending on the specific strategies that are chosen.

METHODOLOGY
A descriptive analysis of the Chinese guidelines that were collected served as the methodological backbone of our study. This analysis was complemented by data gathered in interviews with the people who draftedalone or within a team-the guidelines under study: The comparison between Chinese and European guidelines and standards was carried out following a list of items that had previously been agreed on and that allowed us to see what information was included or omitted: • What to describe (i.e. content selection); • How to describe (i.e. formulation or AD style); • When to describe (i.e. timing of the AD); • Voicing

CHINESE AD GUIDELINES
The lack of unified Chinese guidelines comes as no surprise since AD is still at its outset in the country. In fact, not even the most widespread audiovisual translation modality, i.e. subtitling, can be evaluated by comparing the output to any national standard (Casas-Tost & Rovira-Esteva, 2018). Matamala and Orero (2018: 150) highlight that it is of the utmost importance that the end users' voice is heard and that their opinion is duly taken into account. Unfortunately, China is still far from defining any unified guidelines related to media accessibility partly because the Chinese users' voice is weak or fragmented. In fact, China lacks unified AD guidelines because AD providers have just started to discuss this issue and have not considered it to Hikma 19 (1) (2020), 167 -186 be a priority so far. Yet, a proposal of empirically tested AD guidelines has recently been tabled. Drafted by Leung (2018) as part of her PhD thesis, it is based on empirical research carried out in Hong Kong. Although Leung's guidelines offer a response to the criticism that some media accessibility guidelines and standards are not supported by experimental research involving users (see, for example, Orero, 2005), they are still largely unknown in the Mainland, where most of the time, non-tested in-house guidelines are used to train volunteer audio describers.
This section will present three unpublished AD guidelines. Although they were drafted in different Chinese cities, they are not exclusively representative of these, since it is possible that more than one set of good practices coexist in one same city.

China Braille Library (Beijing)
In 2011, the China Braille Library created a specific centre for AD that is responsible for AD research, production, training and activity-planning. It offers weekly sessions of films with live AD in Beijing and it also sells-or often donates-CDs with recorded AD to other AD providers in China (Tor-Carroggio & Casas-Tost, 2020). The China Braille Library audio describes approximately 50 movies every year and records the AD of 20-30 movies on CD. Unlike what happens in many Chinese user associations, the people working in this centre are not all volunteers, since some were employed for their background in film studies or broadcasting. Although its activity mainly focuses on audio describing films, they also claim to have offered AD in art exhibitions, magic shows, variety shows and TV series.
Their guidelines are used in their training sessions, which are organised around ten times every year and are offered to audio describers from all around China for free. These guidelines defend three basic principles, which, and also according to our interviewee, are all rather unspecific: first, AD must be produced in Standard Chinese; second, not all kinds of audio-visual products can be described; and third, audio describers should not take excessively long pauses. As for the materials that are "audio describable", our interviewee claimed that those programmes or movies that are too quick or that do not have long enough pauses (such as the news) are not suitable to be described.
The China Braille Library's guidelines also specify some language requirements that need to be met. For example, language needs to be objective, concise, accurate and formal, but specialised terminology and slang need to be avoided. Furthermore, a clear distinction of what is essential and what is secondary must be made so that not all details are described. Moreover, audio describers must construct complete sentences and avoid AD Hikma 19 (1) (2020), 167 -186 overlapping with the original dialogues. The guidelines also offer an AD script sample.

Sun Yat-sen Library (Guangzhou)
The AD guidelines used in the Sun Yat-sen Library were also drafted for training purposes. Mrs. Li not only organises training sessions at the Sun Yat-sen Library, where 15 described movies are offered live every year, but also in other libraries and user associations in the province of Guangdong.
The author of the guidelines started taking notes on her experience as an audio describer in 2014, and in 2017 she combined all her observations in a Power Point presentation, which she now uses to teach and present her guidelines. She is currently drafting the guidelines in a more conventional format so that they can be shared more easily among the audio describers. The guidelines are updated every time she holds a training session, after which more examples are usually added. Mrs. Li was trained by a Hong Kong AD expert who, in turn, had been trained in the USA, and she has read foreign guidelines, mainly Pictures Painted in Words. ADLAB Audio Description Guidelines (Remael et al.). What Mrs. Li believes must be exported to China from Western guidelines is the concept of objectivity. Yet, and according to our interviewee, Chinese people still seem to be used to audio describers acting more as commentators. As for other domestic guidelines, our interviewee admits having read those of the Beijing Hong Dan Dan volunteer organisation. Her guidelines were not drafted with the help of people with sight loss but every time she gets feedback from users she incorporates it into her materials.
These guidelines begin by defining the object of study, their target audience (both primary and secondary) and the various delivery modes that are possible. Yet, it must be stated that the only one carried out on their premises is live due to copyright restrictions. The presentation also enumerates the steps to produce AD, the first of which is to select the material to be audio described. It is recommended that this be carried out taking into account the age and the educational level of the audience. It is also suggested that, in addition to new and popular films, suggestions made by users are adopted as well. At this point, the guidelines offer specific advice for AD beginners and advise against selecting martial arts and science fiction films, if possible. This is suggested so that the intrinsic difficulty of these genres does not discourage volunteers from continuing to collaborate with them. Beginners are also advised not to select movies in which dialogues are scarce and those in which the image does not have excessive relevance. Instead, those films that bear similarities with real life and those in which the story is what stands out the most are recommended. Once the movie is chosen, the Hikma 19 (1) (2020), 167 -186 source text needs to be analysed and only after that can the script be drafted. The guidelines stress the importance of a quality assurance step after the script is drafted, preferably including end users. Yet, Mrs. Li, admits that, unfortunately, this is not always possible.
After that initial introduction, the guidelines move on to analyse the "when", the "what" and the "wording and style of AD". Regarding the "when", it is indicated that AD can be inserted under the following circumstances: when it does not overlap with the dialogues, when characters appear, when the place or time changes or when action happens. With regards to the "what", what they call the "principles" of objectivity and neutrality are put forward and their importance is stressed. Emphasis is put on the fact that AD is different from a personal interpretation or a guided appreciation. Indeed, describing subjectively and even venturing to suggest what a character is thinking is still one of the most frequent mistakes for beginners, according to these guidelines. The importance of describing the essence of what happens in the image is also stressed as opposed to providing superficial or secondary details. As for the wording and style of AD, it has to be "clear, simple, direct but vivid and diverse". The guidelines also point out that the time constraints must be respected, and they recommend the use of short sentences that contain action verbs. They also recommend the description of elements based on a tactile perspective (e.g. "soft as a pillow", "thin as a sheet of paper"), and the appropriate use of film terms such as "close-up", "slow motion" and "flashback". Word repetitions must be avoided, and the description must suit the film's style. Finally, it is important not to fill every single gap between dialogues with description.
These guidelines conclude by offering several interesting materials. First, an AD script sample is provided. The script is divided into three columns: the first provides information on what happens in the scene, the second specifies, for example, the speed of the description and the last one includes the description itself. Second, the three "don'ts" of AD according to a user called Zhu Junyi are listed: don't give too much information, don't be subjective and don't disturb the viewing experience. Third, some suggestions on what has to be checked in case of live AD in films are put forward. The document ends with the following message: "People are of the most importance and we need to start from rights. Audio description provides equal opportunities and rights to people with sight loss."

Sound of Light Barrier-free Film & TV Culture Development Centre (Shanghai)
Sound of Light Barrier-free Film & TV Culture Development Centre (Guanying zhi Sheng 光影之声, hereinafter SoL) is a Shanghainese Hikma 19 (1) (2020), 167 -186 association set up in 2016. As Tor-Carroggio and Casas-Tost (2020) reported, it produces about 50 audio described movies every year, which are recorded on their premises. These movies can be accessed at specific facilities, which are all located in Shanghai. Their AD is created by volunteers, who are trained with the help of Mr. Yu Jiang's guidelines. These training sessions usually last for half a day and end with a test. The guidelines were drafted in 2018 and their main objective is to ensure the quality of their AD scripts. 8 As a basis for the guidelines only the Netflix style guide was consulted, since it was thought to be quite simple. 9 Yet, Netflix's stress on objectivity had a great influence on this association. Indeed, SoL had been debating at length whether AD should be objective or subjective at that time, and Netflix made the final decision for them. No other domestic or international document was used as a reference. Nonetheless, they are aware of the existence of other guidelines, such as the ones drawn up by the Beijing Film Institute.
Given that SoL is led by a person with total sight loss herself, user needs, and requirements are faithfully reflected in the guidelines. Also, SoL organises AD screening activities every year to which many users are invited so that they can give feedback on the AD SoL delivers.
Yu Jiang's guidelines start off by defining basic concepts and explain what elements should be included in the AD. The second section presents what they call "the principles of AD". The importance of objectivity is pointed out because users need to be able to draw their own conclusions based on what is described to them. However, given the ever-present time constraints in AD, the guidelines state that the use of qualitative adjectives such as "handsome" or "adorable", can be used. In terms of when to describe, the guidelines specify that the AD needs to follow the image closely and in an orderly manner but, in general, when there is a pause that lasts between one or two seconds, no description should be added. However, in some cases, if a relevant character appears or something important happens, no matter how short the pause is, it should be used to ensure users follow what is going on. As everything else in these guidelines, this point is illustrated with some descriptions that are only a few characters long (e.g. "in the library", "at KTV with some friends") that are indispensable for the comprehension of the examples provided.
The third section focuses on the requirements for the script. Apart from some layout-related issues, the number of Chinese characters per second (including punctuation) is specified: a maximum of five and a minimum of Hikma 19 (1) (2020), 167 -186 three. Audio describers also need to be careful not to make the AD too long so that it does not overlap with the dialogue and not too short so that it does not leave the audience unattended and confused. In fact, the guidelines state that audio describers should use their time to the full to depict a scene in the most vivid way. Also, the AD must be specific and correct; therefore, audio describers need to do the necessary research to learn about what they are describing. The guidelines suggest that audio describers should look for the appropriate research materials or consult others.
The fourth section puts forward the language-related requirements. To begin with, everyday language should be used so that people of all ages can understand the AD. Thus, both overly formal and informal language must be avoided. Moreover, language needs to be formulated in such a way that it can be voiced easily. In order to achieve said objective, the guidelines suggest that audio describers should choose the synonym that makes the text as easy to read as possible and also reminds them to be aware of the tone changes that may arise when combining characters. Given that, for instance, the "s" and the "sh" sound is not always easy to differentiate in Chinese, instead of saying 环视四周 (huan shi si zhou, to look around), describers should choose 环顾四周 (huan gu si zhou, to look around). Similarly, and since "he" (他, ta) and "she" (她, ta) are pronounced the same exact way, the use of these pronouns should be avoided in the same sentence when referring to different people. Anachronisms need to be avoided as well. Furthermore, adjectives cannot be used to interpret what is actually shown on the screen. Therefore, if a woman whose face is covered in tears is shown, saying "tears roll down her cheeks" would be more appropriate than saying "she is devastated". In addition, in some cases it will be necessary to specify what Chinese character we are referring to, since some characters share the same pronunciation. For example, in Chen Jialuo's Transfer Student from Taipei (2012) professor Wang writes a character that is pronounced han on the blackboard. Audio describers should not just say "he writes a han character", since this does not say much, but "he writes a han character, the han from the word Han chao [literally, 'Han dinasty']". The explicit reference to a known word will facilitate users making an immediate mental association. Finally, cinematic terms are only allowed when they are essential, when one is completely certain that a word is used in a particular way and, obviously, when the time for the AD allows for that extra information.
The fifth and last section clarifies how to deal with more specific issues, such as: • What information to include during the opening credits. In this case, the producer, the director, the main actors should be Hikma 19 (1) (2020), 167 -186 mentioned. Were the describer to have more time, the name of the scriptwriter, for example, can also be provided. The movie can also be presented with a short sentence, always following a similar pattern: "讲述了……的故事" ("the movie tells the story of…"). The time for this depends on how long the initial credits last. • When to name characters: the guidelines recommend waiting until a character's name is mentioned in the film to reveal their identity. In order to avoid using a character's name, some strategies are shared. For example, short phrases like "a middle-aged woman" or a "bald man" are suggested. Yet, in case there are too many characters, their names can be anticipated. • How to deal with subtitles: they should either be integrated in the AD or included in the formula "subtitles appear: XXX". If the subtitles contain the title of the movie, it should be voiced as "The title of the movie appears: XXX". Also, other text on screen, such as historical background information at the beginning of the film or narrative explanations at the end, should be included in the AD.

•
How to deal with foreign languages and Chinese dialects: the guidelines recommend that voice over is used to read subtitles in Standard Chinese. Yet, there is no need to translate English words like "yes", "no" and "okay" into Standard Chinese because they are generally known. The guidelines also recommend not to choose movies in which the presence of Chinese dialects or foreign languages is too strong to avoid that AD becomes voice-over. • How to overlap with the dialogues when this cannot be avoided: describers need to look for moments when the dialogue does not provide essential information or when sounds or words are repeated, so that they gain a few extra seconds to include description. • How to handle music and sounds. Since music and sounds are seen as essential components of the movie, describers should make sure that these can be heard so the audience can appreciate them.
The SoL guidelines are clearly the most extensive ones in China, not only because they touch upon the widest range of issues, but also because they are accompanied by clear examples that illustrate how the suggestions can be implemented. Moreover, the fact that ample input from end users was integrated and foreign guidelines were studied and used as background materials contribute to them being a comprehensive training document.

DISCUSSION
From the presentation in the previous two sections of this paper, it is clear that there are a lot of similarities between the Chinese guidelines and the ones created in various European countries. In terms of content selection or what to describe, all guidelines are in agreement that the main narrative constituents, i.e. the characters, their (re-)actions and the spatio-temporal setting in which these take place, shouldif possiblebe included in the description. In terms of timing or when to describe, again there is general agreement: descriptions should not interfere with dialogues and/or sound effects that are hard to interpret without any visual context. Also in terms of AD style or how to describe, there is some resemblance: both in China and Europe guidelines advise audio describers to use generally accepted language that is at the same time vivid and varied, yet easy to voice and understand. In addition, it is recommended audio describers use short, simple sentences that preferably contain action verbs, and to adapt their descriptions to the genre they are describing.
However, the comparison makes it very clear that the two continents are at different stages in the guideline development process. This is reflected by the fact that the Chinese guidelines are of a more prescriptive nature than the (more recent) European ones. One possible explanation could be that AD is still relatively new in mainland China and is still a service that is almost exclusively provided and taught by practitioners and user organisations. In Europe, AD is increasingly being offered as an academic course at universities and other higher education institutions, which also started seeing it as an object of research. Since AD has from its onset been considered as a type of audio-visual translation, this meant that paradigms and principles from translation studies have been applied to AD too. The development echoes Jankowska's (2015) remark that "the guidelines or standards existing in many countries […] are often established with reference to the personal experience of their creators or to similar guidelines used on foreign markets which are often derivative of other sources" (24). It can also explain the shift that can be witnessed, particularly in the latest documents such as the ADLAB guide, from prescriptive guidelines to more descriptive strategies that offer the describer more freedom.
The most important consequence of this development and arguably the biggest difference between the European and Chines guidelines, is that Europe is moving away from the dichotomy between objectivity and Hikma 19 (1) (2020), 167 -186 subjectivity. Rather than presenting a black and white "either-or" choice, 10 current European guidelines offer audio describers the choice between various strategies on a continuous scale, ranging from highly objective to highly subjective. This approach acknowledges the very complex nature that is inherent to the AD process, and allows audio describers to create and formulate their descriptions depending on the genre of the programme or film, the specific moment or scene at hand, the narrative context, the target audience, etc. This does not mean that Europe advocates an "everything goes" approach to AD. In that respect, Leung's observation (personal communication, 9 August 2018) that AD in mainland China is sometimes more of a spoken commentary than a description, is valid. AD has to follow the narrative that is being developed and should contain (all) the elements that allow the visually impaired audience to reconstruct that narrative just as the sighted audience does. But, as with any form of translation, there are various ways to achieve that goal.
An interesting step in the development process of the Chinese guidelinesand one that has often been neglected in Europeis that they are heavily based on end user input. Indeed, in the Chinese case user satisfaction and opinion are taken into account and, subsequently, the guidelines are constantly modified. In fact, Tor-Carroggio's (2020) paper reporting the results of SoL's latest questionnaire to gather data on users' views on the current AD provided exemplifies the interest in seeking users' feedback that will later be reflected in the guidelines.
Finally, just as in Europe, the importance of guidelines for Chinese audio describers is out of the question. As Tor-Carroggio (2018) showed, 80% of the respondents to her study admitted that guidelines are important to them, which suggests that most Chinese describers have a great awareness of the AD process and find it important to follow some recommendations. In any case, there was a point on which all of them agreed: they were all willing to collaborate with other associations and institutions and share their experience.

CONCLUSIONS AND IMPLICATIONS OF THE STUDY
The aim of the present paper was to present different AD guidelines that are being used in Mainland China and to compare them both to each other and to similar documents existing elsewhere, particularly in Europe. As became clear from section 4, there are a lot of general similarities between the different Chinese and European guidelines, particularly in terms of content selection and timing, and to a lesser extent also in terms of AD style. The main difference between Chinese guidelines on the one hand and the most recent European documents on the other, is the different approach to objective vs. subjective description, a difference that can be ascribed to the fact that the former guidelines are created by practitioners with input from end users, whereas the latter are the product of academic research.
Our study is not exempt from limitations, the main one concerning both the small sample of written Chinese guidelines analysed and the limited number of AD providers interviewed. This is considered to be a limitation due to the size of the country and the several groups of volunteers offering AD independently. Nonetheless, and in light of what was outlined above, interesting avenues for further research open up, such as the comparison of existing ADs to see to what extent different guidelines lead to different types of AD, and/or the testing of different types of ADbased on different AD strategiesto gauge audience preferences and mental effort. Indeed, recent research (Fresno, N., Castellà, J., & Soler Vilageliu, O., 2014;Fresno, N., Castellà, J., & Soler Vilageliu, O., 2016) seems to indicate that different AD strategies lead to a different mental workload induced in the audience, and experimental studies testing different types of AD could shed some initial light on these hypotheses. The results from these tests can then be used to train audio describers and hence further improve AD both on a qualitative and quantitative level. It would also be interesting to compare the way AD is conducted on the Mainland as opposed to Hong Kong, where they already have experiment-based guidelines. In sum, we hope this paper will lead to further professionalisation and both qualitative and quantitative improvement of AD services in the country. Finally, through the comparison of the Chinese and European traditions, it is likely that the paper will generate discussions that may lead to further adaptations and optimisations of AD as is currently offered in Europe.