“Oh for Heaven’s Sake, Do I Need to Explain This Really?” Translation Skopoi in Live Art Song Concerts

This quotation is taken from an empirical study into live audience experiences of translations in classical music concerts. Collected at a music festival in Oxford, UK, these remarks describe a participant’s encounter with a specific form of classical music: art song. Art song is a form that sets (often independent) poetry to music in the classical music genre (differentiated from, for example, popular, folk, or traditional music). Due to the form’s language-specific iterations—including but not exclusive to the German Lied, French mélodie, and Spanish canción—art song is often performed in languages other than the vernacular, forcing interlingual translation to be a determinant feature in the genre. As the aforementioned quotation reveals, this phenomenon creates a complex interpretive experience for audiences where the actions of “listening to the music” and “watching the performers” are often accompanied by translating actions, such as “following the foreign language text” and “read[ing]” translations in the vernacular. The challenges associated with art song in translation have attracted modest attention in the literature, which similarly suggests that art song and the way it generates, produces, and propagates meaning through translation leaves the listener “in a most difficult position.” However, translation scholarship to date offers a minimal amount of detail in terms of understanding the nature of these “difficult positions” from the perspectives of translation end-users—the attitudes and actions found in the phenomenological experiences of audience members themselves. This gap in understanding can be attributed to a dearth of academic literature concerned with song in translation, which as a practice requires multidisciplinary approaches, challenging assumptions around authorship, and often blurring theoretical concepts such as translation, adaptation, and creative writing. Within this limited body of research, studies of relevance to the art song genre can be located in functionalist views of song in translation; aligning with trends in translation scholarship that pay greater attention to the reception, social and cultural purposes and effects of translation, and its commercial uses and ethical and political consequences. A key model within this developing field is Peter Low’s functional account of strategies within song translation. Low adopts Vermeer’s concept of skopostheorie, where a translator’s aims are determined by the “skopos” or purpose of a “communication in a given situation.” Applying Low’s version of skopostheorie to the live art song genre reveals the presence of multiple “skopoi,” each requiring varying translation strategies. These strategies are targeted toward performance (word-for-word translations used by performers when learning songs) and consumption: (1) traditional approaches using communicative or semantic translations in printed programs; (2) developing approaches seen in communicative or gist translations used for surtitles and subtitles; (3) gist translations used for spoken introductions by performers; and (4) sung translations. The skopoi of most concern to us in this study are those targeted toward consumption. Although Low’s typology provides a useful framework to analyze the different translation types and strategies identified in live art song, this analysis takes place from a methodological position of some distance from TRANSLATION REVIEW 2023, VOL. 116, NO. 1, 1–12 https://doi.org/10.1080/07374836.2023.2231039


Introduction
I do find that there is a general problem, in that as a listener, my brain is trying to do at least four things: listen to the music; watch the performers; follow the foreign language text; and read the English translation, and that is something I find almost impossible.
[Hugh/RAS6/D] This quotation is taken from an empirical study into live audience experiences of translations in classical music concerts. Collected at a music festival in Oxford, UK, these remarks describe a participant's encounter with a specific form of classical music: art song. Art song is a form that sets (often independent) poetry to music in the classical music genre (differentiated from, for example, popular, folk, or traditional music). Due to the form's language-specific iterations-including but not exclusive to the German Lied, French mélodie, and Spanish canción-art song is often performed in languages other than the vernacular, forcing interlingual translation to be a determinant feature in the genre. As the aforementioned quotation reveals, this phenomenon creates a complex interpretive experience for audiences where the actions of "listening to the music" and "watching the performers" are often accompanied by translating actions, such as "following the foreign language text" and "read[ing]" translations in the vernacular.
The challenges associated with art song in translation have attracted modest attention in the literature, which similarly suggests that art song and the way it generates, produces, and propagates meaning through translation leaves the listener "in a most difficult position." 1 However, translation scholarship to date offers a minimal amount of detail in terms of understanding the nature of these "difficult positions" from the perspectives of translation end-users-the attitudes and actions found in the phenomenological experiences of audience members themselves. This gap in understanding can be attributed to a dearth of academic literature concerned with song in translation, which as a practice requires multidisciplinary approaches, challenging assumptions around authorship, and often blurring theoretical concepts such as translation, adaptation, and creative writing. 2 Within this limited body of research, studies of relevance to the art song genre can be located in functionalist views of song in translation; aligning with trends in translation scholarship that pay greater attention to the reception, social and cultural purposes and effects of translation, and its commercial uses and ethical and political consequences. 3 A key model within this developing field is Peter Low's functional account of strategies within song translation. 4 Low adopts Vermeer's concept of skopostheorie, where a translator's aims are determined by the "skopos" or purpose of a "communication in a given situation." 5 Applying Low's version of skopostheorie to the live art song genre reveals the presence of multiple "skopoi," each requiring varying translation strategies. These strategies are targeted toward performance (word-for-word translations used by performers when learning songs) and consumption: (1) traditional approaches using communicative or semantic translations in printed programs; (2) developing approaches seen in communicative or gist translations used for surtitles and subtitles; (3) gist translations used for spoken introductions by performers; and (4) sung translations.
The skopoi of most concern to us in this study are those targeted toward consumption. Although Low's typology provides a useful framework to analyze the different translation types and strategies identified in live art song, this analysis takes place from a methodological position of some distance from audience members themselves. The lack of empirical research into the practical application of translation skopoi within musical settings-and the demands the varying skopoi place on audience members-leaves gaps in knowledge, in both scholarship terms, and (as the participant remarks above suggest) in practical terms too, providing areas for further inquiry. In that sense, this study responds to calls for further research into the reception of translation and music, looking beyond individual works and composers to consider wider contexts, changing audience tastes, expectations, and ideologies. 6 While it is apparent that critical positions exist toward art song in translation within the literature, key questions remain as to how translation types, strategies and behavioral norms interact and shape audience members' experiences of the genre: questions that are addressed in this study.

Methods
An aim of this study was to produce in-depth first-person accounts of audience members' experiences of live art song, through undertaking rich qualitative inquiry. The present study was not limited to investigating audience members' use of translation resources alone, and this study's focus on translation represents one feature of a much larger research project to examine audience members' experience of live art song concerts in their totality. That being said, translation played a prominent role in the accounts of those participants who took part. Data for this study was collected at Oxford Lieder's music festival "Tales of Beyond," held in Oxford, UK, in 2019. Oxford Lieder is one of the UK's leading promoters of the art song genre, with a mission to "re-establish an appreciation of song; the meeting of words, music, languages and artistry that can be so powerful, yet which has been neglected in recent decades." 7 At the heart of Oxford Lieder's work is the annual Oxford Lieder Festival, featuring high-quality performances from world leading artists. The Festival is delivered to in-person audiences in venues across the city of Oxford, as well as being livestreamed to audiences internationally.
The study used a mixed method approach (questionnaire, diary methods and guided interviews). Statistical and thematic analysis, and Interpretative Phenomenological Analysis were applied to a data set that utilized eighty-two individual participants' experiences of live art song, featuring a regular attendee sample and a sample of individuals who were experiencing art song for the first time. A key assumption underpinning this study was one that is grounded in audience research scholarship: namely, that audience engagement, be that viewing, reading, listening-or within the context of this study, engaging with types of translation-is a "motivated experience." 8 Consequently, audience experiences are shaped not only by the experience audience members have in the immediacy of consumption, but also by how these individuals prepare for their experiences (why people attend, their motivations, expectations, and cultural capital) and how experiences impact upon audience satisfaction, acceptance, rejection, and their longer-term integration into feelings, lives, and future encounters. As such, the data collection activity in this study intersected with participants' experience before, during, and after the performances themselves, gathered at specific points identified in a twelve-month timeline around the Festival.
A cross-sectional questionnaire elicited responses from audience members who regularly attended live art song and was distributed and completed before the Festival took place. Seventy-seven questionnaires were submitted in total (n = 77). The questionnaire targeted the choices participants made in deciding how to engage with the genre, and their previous experiences with the genre more broadly. The questionnaire was distributed at a time when the participants were making decisions about how they would engage with the 80 upcoming festival events (if at all) and it provided an opportunity for individuals to reflect on tastes and patterns in their choices and engagement to date. Demographic data was collected to benchmark against UK-wide data on cultural engagement, 9 and foreign-language competencies, 10 and this data provided a method to profile the audience more broadly.
Profiling the participants in this way revealed that the audiences in this study were older and more educated than national average classical music attendees, and conformed to national average classical music trends regarding gender and ethnicity. The audiences were mostly retired and had received a musical education in terms of learning to play an instrument. With particular relevance to translation, participants had significantly higher levels of foreign-language competencies than the UK population as a whole. 11 Furthermore, audiences brought a significant amount of knowledge and cultural capital to their art song encounters, which had been developed not only through previous musical, language and other educational training, but through attendance at the Festival over a long period of time, and high levels of engagement with other cultural activities too.
A second phase of the study sought to expand on the data acquired from the questionnaire phase, providing opportunities for the participants to reflect on actual art song encounters, instead of relying on recall and memories of the past. Diary methods provided an opportunity to "probe deeply to uncover new clues, open up new dimensions of a problem and to secure vivid, accurate, inclusive accounts that are based on personal experience." 12 15 members of the original questionnaire sample participated in this phase of the study, alongside an additional sample of five participants who had never attended a live art song performance prior to their participation.
In the third and final phase of this study, interviews were conducted with participants from both the regular and new attendee samples. Using a subset of the unstructured interview method, the guided interview, 13 a general order of questions was drawn up from each participant's diary, providing an opportunity to clarify meaning from the diaries and to garner additional reflections.
Questionnaire, diary, and interview data were extracted and compiled into a case file for each participant. Analysis primarily drew upon Interpretative Phenomenological Analysis (IPA). 14 IPA has its theoretical grounding in hermeneutics (the philosophical exploration of interpretation), phenomenology (the philosophical exploration of experience) and idiography (a focus on the particular or specific). In other words, the methodology sought to produce interpretations of a specific audience member's experience of live art song engagement, at a specific place and time. Line-byline coding was applied initially for each individual case file, identifying broader patterns (themes) and locating elements of convergence, commonality and divergence within each case. This was done iteratively for each individual case file, using dialogue between the researcher, the data, and knowledge of theory, with a view toward generating an interpretive account of audience encounters. The material was then organized into overarching coding and themes, and was clustered to develop a series of subordinate themes that could be applied across multiple cases. To ensure participant confidentiality, all participants were assigned a pseudonym and an ID code. Direct quotes lifted from the dataset are therefore attributable only to ID codes and not to individuals themselves. 15

Results and Discussion
A total of 91.3% of participants stated that it was "very important" or "somewhat important" to understand the meanings of song texts, evidencing the value placed on translations by the participants in this study. Examples of the rationale behind this positionality include: "good setting[s] will emphasise words (or phrases) which the composer considers important and you need to know which they are" [Hugh/RAS6/Q]; that a "song is rather meaningless if one doesn't understand the content" [Matilda/RAS8/Q]; and "If you don't understand the words, some of the potential magic is being lost" [Maxwell/RAS7/Q]. For others, a text's "meaning would have informed the composer's creative process" [Colin/RAS14/I] and, therefore. it was considered essential to access this meaning through translation. Only a minority of the regular attendee sample deviated from this position. For these participants "during the performance it's the music and the presence of the performers and what they are doing which demands [their] attention" [Douglas/RAS1/Q]; and they "like to hear the words even if [they] don't understand them' suggesting that 'the music is much more important to [them]," [Lydia/RAS10/Q]. In these examples, participants explicitly placed greater value on musical elements over lexical ones, making translation less of a necessity. The rationale for adopting these alternative positions was often due to constraints in attention and concentration, where a need to focus on the experiential features brought about by event "liveness" took precedence over the need to use translation materials. Regardless, no participant sought to completely devalue or ignore translations entirely. Understanding the meaning of texts was still considered "somewhat" important and, in the entire study, not a single participant stated that understanding the meanings of texts through translation was "unimportant." Textbased translations (printed and surtitles) and sung translations were most prominent in the participants' accounts, due in part to the prevalence of the former within live art song today and the relative novelty of the latter (which was particularly pertinent across this study as three Schubert song cycles were performed in English for the first time at the Festival). The text translations used were predominantly by art song translators Richard Stokes and Richard Wigmore; whereas the sung translations used were by Jeremy Sams, a renowned translator of opera and theatre. 16 It is beyond the scope of this article to scrutinize and discuss in detail the quality of these translations. However, these translators are all established translators in their respective fields. Instead, the remainder of this discussion explores audience attitudes toward the different text based and sung translation skopoi in more detail.

Attitudes Toward Text based Skopoi
Throughout the data set audience members' translation habits were largely self-consistent, and there was evidence of participants adopting a default mode in their approach to using types of translations across different concerts. The preferred type of translation for the regular attendee sample was printed text translations. Since the late nineteenth century, program booklets featuring texts and translations have been routine in art song concerts in the UK, as seen in historiographical research into UK musical centers. 17 Within this study, the prominence of comments on handling printed translated texts suggests that these translations occupy a significant place in audience experience: For art song I normally try to ensure I have copies of texts and translations in a form which I can (hopefully) follow during the concert. Many concert organisers don't provide them, and, even when I think they probably will, I try to ensure that I am not caught out! [. . .] I normally read them through beforehand. Many promoters turn the lights down during the performance (sometimes even when they have provided the texts and translations themselves!) so that I can't read them anyway! [Hugh/RAS6/D] Our arrival just before the start meant that I couldn't do my usual thing with the program, which is to go through what is to be presented and then spend a few minutes skimming through the words of the various songs to see what they are about. Doing that enables me to concentrate on looking at the singer and listening to what he or she is singing without needing to follow the words in the program [Brian/RAS9/D].
In these comments, participants describe the value of printed translations to their experience (to the extent of bringing one's own translations so as not to be "caught out") alongside their default "mode" (arriving early to produce their own mental gist translations). This approach was not an uncommon one: preparatory gist-translation work enabled the focus of a participant's attention to shift to other features presented to them within art song environments, and to mitigate limitations of the printed skopos caused by extraneous production factors (for example dimmed lighting). Furthermore, this preparatory work enabled participants to filter out other actions to aid their navigation of complex interpretive art song experiences (e.g., the "almost impossible" nature of combined "listening," "reading," "watching," and "following" actions found in the audience member comments that precede this manuscript). For the majority of participants, translation activity more often took place around the music itself, both pre-and postconcert: If I'm really bored and don't really like someone I'll read [the printed texts] assiduously, and I'll know every word. But when I really like someone, I might just glance and go . . . I'd much rather go away and read the text afterwards. [Harriet/RAS3/D] [There was] one very long song, which I thought I knew, and there was loads more to it than that and I didn't know it at all. So yes, that's a wonderful thing. It's happening, it's going, it's slipping away from you, and you want to go home and dig out your records, or whatever. And do the work there. [Barnabas/RAS12/I] These comments suggest that extending translating activity post-concert is a valued part of the experience, and the participant explicitly uses the term "work" to describe this activity. Importantly, this activity is located in the home and not in the concert hall. For many participants, their sustained engagement with live art song as a genre (with some participants attending Oxford Lieder concerts for around 20 years) meant they had acquired their own collections of printed translations to refer to. These participants continued to translate at home by using these materials and listening to recordings as knowledge-formation and reinforcement techniques. As such, participants developed a sophisticated appreciation of not only translations, but also translators, and the differing types of translation strategies afforded to them as part of their live art song experiences: As for Richard Stokes' translations . . . between the two extremes of a literal prose account and a translation, he tends to fall closer to the former but in Des Antonius von Padua Fischpredigt why do the lines: "Der Karpfen so gfallen"/"Der Stockfisch so gfallen"/"Den Krebsen so gfallen" all get translated as "so pleasing to fish"? [Maxwell/RAS7/D] Participants scrutinized the communicative and semantic translation strategies often employed in live art song. In this example, Maxwell also references the translator by name, a recurring feature in the participants' accounts. In doing so, he works to counter the habit observed in other forms of literary translation, as in Venuti's work, who argued that the relative invisibility of a translator impacts their power and influence. 18 As art song translating is a rarefied phenomenon, the relative visibility and reliance on a small number of translators plays a part in shaping audience experience: I don't know how far this is going, but I'd say I far prefer Richard Stokes' translations to Richard Wigmore's. Richard Wigmore keeps very close to the original text and there's an argument for that, whoever reads the translation knows what's going on. Richard Stokes tries to include little bits of the poetry, and I prefer that by far. Because, if you don't know German and you're only reading the English translation, some of the translations can come across rather pedestrian. [Matilda/RAS8/I] Not only does Matilda reference the names of translators, but her initial reservation in sharing these remarks emphasizes the connectedness amongst the field in showing concern that her comments might be shared more broadly. Matilda's preference for additional "little bits of poetry"-and the inference from Maxwell's account that simplified translations were unsatisfactory-suggest a desire for more elaborate translations. However, these views counter those of a minority of the regular attendee sample who found text translations distracting: I would really never sit and read texts, I find it hard. I find it hard to understand how people can sit and read the texts and listen to the music. To me that's, I just couldn't! If I was reading the words I wouldn't be letting myself go to the music. [Lydia/RAS10/I] In these participant remarks, translated texts have a tendency to become subordinate to the voice and obstruct the voice, hampering the potential for meaning-making derived from the music. Lydia seems to privilege the materiality of sound, suggesting that a focus on translated text as a visual signifier would block her body's capacity to receive the music. Although this positionality contradicted most of the regular attendees, it did align with the majority of those attending live art song for the first time: I listened more than read. And sometimes read more than listened. I think it was, when it was sung in German, it was just one step too far to follow the translation and locate it in the German, and then respond to it in some way as well. So, I tended to just stop and listen instead. [Dale/NA3/I] These interpretive challenges, which the participant characterizes as a "step too far," resonate closely with the regular attendee's remarks that precede this article ("almost impossible"). This comparison is particularly striking in consideration of the evidenced cultural capital the regular attendees brought to their art song encounters (which include foreign-language competencies). Although both samples observed the challenges with printed text translations, new attendees explicitly stated that printed translations created a barrier to engagement. Instead, these new attendees preferred an alternative form of text based skopoi used in several performances, surtitle technology: The best thing about the production I felt was the screens behind the musicians with the text and translation. I got the sense that some of the more seasoned lieder goers weren't so keen, I found it a useful, non-intrusive aide that meant I was able to keep my focus on the performers whilst following the poetry. [Zac/NAS1/D] In recent years art song promoters in the UK have adopted surtitle and subtitle technology, although this practice is relatively rare, departing from other vocal art forms such as opera and musical theatre where surtitles enjoy widespread use. The findings in this study align with a small number of studies in opera translation that reveal the benefits of surtitle technology include greater audience accessibility. 19 However, the use of surtitles was often at the expense of other challenges, which included a detrimental impact on the visual aesthetic of the action on stage, and specific problems for spectators with impaired eyesight, due to them needing to adjust their focus between the surtitles and the performers. Despite these challenges-and contrary to Zac's observations of "seasoned lieder goers" not being "so keen"-the regular attendee sample was largely appreciative of surtitles too: The timing of the surtitle operator was impeccable-an audience and performance are sometimes better united when there are no printed words to shuffle, drop, plus the bobbing heads etc. [Barnabas/RAS12/D] Surtitle technology was praised for its ability to enable the audience to focus on the performers, although some participants noted this created "a slightly false impression" as the screens became the focus rather than the performer themselves [Brian/ RAS9/I]. As per studies in opera translation that show how surtitles mitigate the "attendant risk of intrusive noise as hundreds of pages [are] turned simultaneously," 20 Barnabas highlights that printed translations can break the sense of connection with the performer, through unintended behaviors as audiences interact with these materials ("shuffling," "drop"-ping materials, or "bobbing heads"). Barnabas's remarks therefore illustrate the impact of translation types on extraneous auditory and visual actions, which can impact an audience member's focus and attention at any given time.
Although surtitles were positively received by most participants, these were not deemed to be a substitute for printed text translations, but instead were used as a supplementary tool: [In the surtitle screens] you are given both languages, there is only usually room for at most a couple of lines, whereas, if you have a printed text, you can see the whole "shape" of the poem and know where you currently are within it-both, for me, very important to appreciating the quality of the setting. [Hugh/RAS6/D] Using printed texts is distinct from if you just look at the words on the screen, (for example, behind the performer), you don't know what's coming. So, you can listen to the singer singing about it and not be much wiser in the end. [Brian/RAS9/I] Translation actions in these accounts are not isolated ones aimed at understanding the meanings of individual words, but are used to navigate spatially through the songs themselves as a device to "[know] where you currently are" within the poem. Hugh's comments that he needs to see "what's coming" and "the whole poem on the page" suggests that an element of anticipation and wholeness is required, which is lacking in surtitles and made possible only through printed translations. As such, within live art song encounters, rather than segmenting individual skopoi to translate in specific moments in time, what we actually see is evidence of a blended approach to the use of printed and surtitle text skopoi, used variously before, during, and after specific song encounters.

Attitudes Toward Sung Translation Skopoi
A central feature in the discussion thus far has been the role of audience member agency in their translating activities: that is to say, the "choices" participants make to translate (which consequently aid or hinder their entire live art song experiences) are genuine choices. However, not all translation skopoi permits this degree of agency. This is most obviously demonstrated in the use of "sung translations," which enforces a translation approach on audiences instead.
In comparison to the text based skopoi, the use of sung translations within art song contexts has been historically more sporadic. Recent attention has been afforded to these translation types through a modest proliferation of art song performed in new English translations since 2018. 21 Sung translation approaches transfer skopostheorie's "translational actions" from the domain of production to performance, in making changes to art song objects themselves. Like the text translations in printed programmes (but unlike surtitle technology), the phenomenon of art songs performed in English translation is not a new one. Literature examining historical performance practice reminds us that UK attitudes toward art song in translation were transformed by the political upheavals during World War I, when a reluctance to sing or hear the German language resulted in German art song being sung almost always in translation, a trend that would continue until the early 1920s. 22 In the study, 88.4% of participants believed it was very important or important for songs to be sung in their original language, showing resistance to sung translations that counter the norms of text-based translation types that are present within live art song practices today. In fact, there was an expectation by some participants that these translations would be flatly rejected by those who would not accept the use of alternative skopoi: We had wondered if the ticket sales would be low because the cycle was being sung in English. My friend [name redacted] told us, in a polite and friendly way, that he wouldn't be attending for that reason. I haven't told him this, why should I, he's a nice man, but I think he's putting his own nose out of joint by being a bit precious about that. [Hugh/RAS6/I] Hugh's awareness of the potential impact of translations on the wider audience community relates to Toury's conceptualisation of translation norms: unspoken or hidden rules that are applied when evaluating translations. 23 This concept was later extended by Chesterman to highlight the presence of forbidden and permitted aspects of translation that include "expectancy norms." 24 Expectancy norms allow for evaluative judgements on translations, where users have a sense of what is an "appropriate" or "acceptable" translation within a specific genre. Translations-as shown in these remarks-are therefore approved or "validated in terms of their very existence in the target language community": Question: How important or unimportant is it for you for songs to be sung in their original language? Participant: Oh for heaven's sake, do I need to explain this really? What a question from a Lieder organisation! The music is written for the text, the music "fitting the language like a glove." This can never be achieved with a language, for which the music was not composed.
[Dorothy/QS62/Q] The frustration toward the question evident in this participant's response is indicative of the deeply-held nature of attitudes toward translation. As such, it suggests a perceived inviolability of art song in its supposedly originally-conceived construction, that is seemingly challenged by a sung translation. Despite music "fitting the language like a glove," a feat that "can never be achieved with a language, for which the music was not composed," as discussed, historical performance practices remind us that there are, in fact, recent histories of different conventions surrounding sung translation. However, participants were not necessarily aware of the previous widespread use of sung translations. Comments that describe sung translations as being "trendy for trendiness sake" [Lydia/RAS10/I] suggest unfamiliarity with previous prevalent practices. Furthermore, these historical examples signal the types of conditions that prompted the acceptance of changes to translation norms, which include the political and socioeconomic implications from the fallout of international conflict, and favorable or ill feeling toward different cultures in different national contexts at various moments in time. These examples serve as a reminder that translation approaches within musical practice are not shielded or segmented from extraneous factors found outside the concert hall and in wider society.
Participants' resistance to hearing sung translations coalesced around a series of distinct rationale. Some suggested that sung translations were deemed inappropriate to the culture of the live art song genre itself, as they were considered to "trivialize" the form [Matilda/RAS8/I]. Others felt that the need for sung translations was redundant due to the existence (and preference toward) text based skopoi. As seen from the participant comments above, there was also a sense that sung translations had a negative effect on the songs themselves, and the music has been designed to "fit" specific technical features of a poem. Elsewhere, some participants highlighted specific technical differences: vowel sounds, rhythm, cadences, stresses, syllables, alliteration and other verbal devices, all of which made sung translations sound "false" or "artificial" [Harriet/RAS3/D].
The nature of this criticism targeted toward the technical features of sung translations aligns with theoretical considerations found in the literature too. Jack Stein highlights the common practice of composers taking liberties with the poems they set to music, in making adjustments to form, meter and rhythm, duration, intonation, and emotional and conceptual values, which make "the juxtaposition of lyric poem and music" within an art song "full of problems." 25 With reference to the earlytwentieth-century proliferation of sung translations, songs were often not translated by professionals, but instead by singers and their acquaintances, accounting for varying degrees of quality. 26 The music critic and translator AH Fox Strangways commented on the sometimes inferior quality of these translations, stating how "few have any idea of the gift and, failing that, the effort required to make a good [sung translation]." 27 Strangways further highlights technical obstacles in sung translations, citing tessitura and appropriateness of vowel sounds, rhyme scheme, syllabic accentuation, and modification of note lengths as critical issues. These specific challenges map onto the contemporary perspectives found within functionalist translation literature too. A more recent study by Low proposes a pentathlon model of sung translations that seeks to balance sense, naturalness, rhythm, rhyme, and singability, as discrete criteria to appraise a sung translation's success. 28 Johan Franzon-although cautioning the ambiguity of singability as a concept-further suggests a triad of prosodic, poetic, and semantic-reflexive match, functions that "must come together if the translation is to be perceived as fully functional, i.e. singable." 29 Although some participants lacked the vocabulary to articulate their resistance to sung translations in these terms, it is clear that these participants' critiques can be mapped onto those found in the theoretical literature.
The resistance to sung translations was not limited to the perceived technical differences between language systems but was also a result of ideological concerns. For some, sung translations represented a perceived fracture in the composeraudience psychological contract, as it was important for audiences not only to preserve, but to connect in their mind with what the composer envisioned and intended. For many participants a song represented "a marriage of music and text" [Maxwell/RAS7/Q]; it consequently was "the best combination" and therefore it was considered "vital" to present music as the composer imagined it [Sarah/Q23/Q]. Without doing so, audience members could not "make sure that [they] understand what the composer/poet wants to say" [Michael/Q46/Q]. This critique highlights the craft of art song, which is different from other forms of song making. Within the art song genre composers rarely write the poems themselves. 30 Instead, the collaborative process between a poet and composer is often conducted at a physical and temporal distance; the result is a "love-child" of two artists, where the "onus is on the composer to treat the defenceless text with courtesy and honour." 31 It is questionable of course whether there is such a thing as a "best combination," not least because participants also expressed the desire to hear different musical settings of familiar texts. Moreover, as art song typically draws upon existing independent poems, these views establish a paradoxical position that suggests the original artistic output (the poem) was not so sacrosanct in itself, to permit it being adapted or "forced" "to fit" into the music in the first place; even if participants then confer a level of inviolability on the song itself as a result of a "matrimonial" bond of lexical and musical pairing. In essence, for many of the participants in this study, it is acceptable for a composer to violate a poem; but for other creative actors, be that a performer, promoter, or-in this instance-a translator to violate a song (for example through the production of sung translations), amounts to "heresy" [Brian/RAS9/I].
Although the majority of the regular attendee sample rejected sung translations outright, there was evidence of some audience members-all the while being sceptical-adopting more accommodating positions. For participants who were more sympathetic toward the use of sung translations, these translations enabled an enjoyable slant on familiar songs and an unexpected benefit was the ability to more closely relate or connect familiar works to the everyday. Responding to a sung English translation of Schubert's Winterreise, some participants made comparisons between the mental disintegration of its protagonist in the closing song and the challenges of mental health and homelessness found on Oxford's streets. Resonating with Linda Hutcheon's view in the related practice of adaptations, which enable "stories [to] evolve and mutate to fit new times and different places," 32  Another recurring feature in these accounts was audience members' consideration of a wider stakeholder set beyond the audience-composer relationship. Sung translations were sometimes cited as an unexpected success due to a triumvirate of the performer and translator, combined with audience willingness. Crucially, the performer was cited as the primary reason the sceptical participants took a risk in the first place: stating "if he thinks it's a good idea in English, there must be a lot of merit in it. Because I like him, respect him, sort of trust him" [Hugh/RAS6/I]. The singer here might be considered what Chesterman describes as a "norm authority," 33 who sanctions the validity of the alternative skopos in this community. Even though it was the pianist who devised the project and commissioned the translator (who himself is highly respected and visible in the field), participants tried sung translations because they "trusted" and "respected" the singer, in this instance the baritone Roderick Williams. Participants attributed the success of sung translations as a tribute to the singer's "talent, art, and enthusiasm" and admitted "[They're] not sure anyone else doing it would get away with it" [Hugh/RAS6/I].
Those who were more accepting of sung translations also had a tendency to consider the impact of translating on others. These participants suggested that the performers might be happier singing in their vernacular, and expressed admiration for the translator, acknowledging the craft of translation is a "skilful job where there is lots to admire" [Brian/ RAS9/Q]. These participants also questioned if sung translations improved access for new audiences, in observing the practical challenges of using the text based skopoi which necessitated a focus away from the performers, and the limitations in audience members' own foreign-language capital. This observation was often stated out of concern for the audience members who might be new to the genre: a view that is substantiated by the previous comments of the new attendees, who suggested that text translations were "a step too far." Unlike the regular attendee sample, the new attendees were overwhelmingly supportive of sung translations in this study. The presence of these alternative views show that although the majority of participants rejected sung translations, differing expectancy norms did exist among these audiences. In that sense, musical audiences that use translations are not homogeneous entities: how art song promoters and performers reconcile competing audience translation needs is therefore a challenge.

Conclusion
Translation remains a determining feature in live art song both for seasoned attendees and for those who are new to the genre. Taking a skopos-based view, the majority of regular attendees spoke negatively about sung translations and positively about text translations, and demonstrated a clear preference for printed translations. Surtitle technology was praised but was viewed as a supplementary translation type rather than an alternative to printed translations. Because of this position, participants had a tendency to fuse available skopoi simultaneously to create their own blended skopos, where the strengths of one translation type mitigated the limitations of another. These findings remind us that existing theoretical approaches can only take us so far in analysing art song in translation. Furthermore, theoretical approaches do not take into account the temporal dimensions at play on the part of audience members' agency in approaching translation. As shown in these findings, the boundaries of when and where translation actions took place exceeded the single event of the music, and often took place before and after performances. In that sense, there are practical questions and opportunities for promoters, performers, and translators in cultivating translation practices both inside and outside the concert hall. Addressing these challenges and opportunities could increase engagement within the genre and propagate changes in norm conventions. One example in practice could be experimenting with printed gist translations, given that so many participants endeavored to arrive early to produce these (mentally) themselves. That being said, attitudes toward sung translations demonstrate the potential resistance to change with regard to language preferences and translation types within the genre. Yet, these findings show us that there are evidently dividends for the producing, performing, and audiencing actors who embrace change. In suggesting that printed text translations were a 'step too far' [Dale/NAS3/D] (while enjoying sung translations), views from the new attendee sample show that alternative translation approaches could attract interest from new audiences. Even for existing audiences, sung English translations reinvigorated the poetry for a number of participants, enabling them to connect to familiar material in new ways. These alternative perspectives suggest that more experimentation may be possible and indeed welcomed by audiences. Perhaps in line with other song genres (such as pop songs that often receive translation and coverversion treatment), art song too could foster versions that can co-exist with traditional presentations of translation materials. In that sense, to respond to the quotation that appears in the title of this manuscript-'for heaven's sake, we do need to explain this really,' as successfully embedding refreshed translation approaches will need to draw upon an understanding that songs are "violable" objects, which composers, poets, translators, performers, and producers have had-and continue to have-agency to change. However, if the introduction of alternative approaches to translation are to be successfully introduced into live art song audience communities, these approaches will require a reframing of audience members' own expectations, in order to shift expectancy norms. As this study has revealed, the intervention of traditional gatekeepers of the art song genre (particularly trusted singers) will remain key to any future changes. Respected performers and producers and other norm authorities will be needed to help validate and legitimize alternative forms of translation to encourage audiences to experiment with fresh approaches.