Your translated memory or mine? Re-membering graphic novels in performed audio descriptions for The Cartoon Museum, London

ABSTRACT

This article offers a practical and theoretical exploration of a (highly) specialised type of comics re-mediation, namely, audio described (AD) comics for the Blind and Partially Sighted (BPS). Building on relevant work in comics studies and translation studies, it is argued that translation activity, broadly seen to include interlingual as well as intersemiotic translation (Jakobson 1959), helps rewrite original works. Translation is defined as a blend, or a hybrid mental space combining characteristics of source and target contexts. As such, it entails the imbrication of personal memory, collective memory and the diffusion of a translator’s memory in linguistic codes, social milieus, textual traditions and digital capabilities at play. AD exhibits the same blend logic in that it selectively contains visual source-text information in target-text audio performance. An exploratory comics AD pilot project at The Cartoon Museum (London) serves as an exemplar. The project consists of three phases: a focused interview with the creative team (describer, curator, comic artists); scripting and performance of three AD samples; and collecting feedback from BPS visitors. The project reveals how collective memory started to form in this dialogic process and, ultimately, which aspects of AD practice may be deemed to be effective.

KEYWORDS:

• Accessibility
• audio description
• comics translation
• memory
• museum
• translation technology

1 Audio description for comics

Audio Description (AD) is a verbal commentary that ‘captures and translates’ visual aspects of various services, events or products (Perego 2019:114, my emphasis). It typically addresses the business, public-service and artistic or cultural needs of diverse constituencies around the globe: blind and partially sighted individuals (BPS), new immigrants, neurodiverse individuals, language learners (Fryer 2016, 42). In the creative industries, AD serves a communitarian model of inclusion, currently at different stages of implementation. For example, films and (streamed) audiovisual products have benefited from at least four decades of legislative change, increasing AD quotas and officially recorded practical guidelines (Fryer 2016, 19). A shorter, still evolving tradition exists in static, fine arts (Ginley 2013:online), given the 21^st century impetus for edutainment, multisensorality and personalisation in museum experience (Taylor and Perego 2022, 204–207) and with initial evidence pointing to benefits for both sighted and BPS visitors who use AD (memorability of facts, emotional engagement) (Hutchinson and Eardley 2022, 238–40). On the other hand, AD advertisements are a recent phenomenon (see P&G’s ground-breaking campaign since 2018). Comic art AD in museums suggested here constitutes terra incognita.

Studying comics in AD more closely requires slight re-calibration of research foci and links within, across and beyond Translation Studies. First, a typical agenda in comics translation entails the following: i) translation strategies and identity presentation across versions; ii) comics translation historiography (e.g. publication history); iii) adaptation (Zanettin 2018, 452). It is worth expanding iii) in particular to go beyond the typical afterlife of comics examined by researchers and practitioners so far, such as theatre, film, or video game adaptations. AD may be added as a special type of remediation (Brownlie 2016, 76). Looking across, AD is now a burgeoning field of Audiovisual Translation Studies (AVT). As such, it utilises specific theoretical frameworks (say, linguistics) and systematically examines given sectors, technologies and stakeholders (describers, service providers, users). In the sector of museum AD, for instance, special emphasis is placed on achieving lexical variation, explicating (art) terminology and exploiting the latitude for personal interpretation when describing museum objects (Taylor and Perego 2022, 209–10).

As a new sector in AD, comics AD needs some conceptual anchoring. An orientation concept I suggest here, and which may help focus on the stakeholders, the creative process and function of remediated comics is that of memory, which is discussed in greater detail in section 2. Ahmed and Crucifix (2018) conceptualise comics memory in particular as styles and archives. Archives refer to the physical or digital life and afterlife allowed by institutions, database curators and fans of comics in formal, peripheral or rogue acts of reading, storage and (re)circulation (Ahmed and Crucifix 2018, 9). Archives are the realm of power relations and the constructions of knowledge, as evidenced in the behaviour and financial, legal or cultural agendas of various stakeholders: creators, editors, letterers, translators and publishers (Méon 2018, 195, 203) (Brienza 2016, 78), crowd workers (Tufis and Ganascia 2018), user-generated paratexts (Kraenzle 2020, 128–29) and museums (Daenen 2018, 261).

In relation to crafting comics AD, pre-existing museological policies, the general interpretive context or organising narrative of a specialised comics museum, budget pressures, community outreach programmes and a general openness vis-à-vis issues of accessibility may all serve as enabling or constraining factors. Audio describers are also key stakeholders here. They typically work for broadcasters or video-on-demand platforms and, more occasionally perhaps, for museums, theatres or other event organisers. Describers are well-versed in (creative) writing and have broad world knowledge and technological skills (AD software use, recording, voicing). The added complication in comics is that a describer would need to have the genre knowledge (or visual literacy) required to judiciously relay non-verbal signs into verbal signs (intersemiotic translation), following (Jakobson 1959); the resulting aural text should, ideally, help listeners form expectations and fill in gaps in their knowledge while the story is re-told in this mode. Some latitude for interpretation would presumably be available, if a plain description of characters, colours, material (paper quality), actions and location is to be avoided. With no pre-existing, codified guidelines and because this type of AD has, to my knowledge, not been attempted before, a question arises as to how information from the visual source may be prioritised, so that ‘something important’ of that original is remembered and presented in a coherent way.

A preliminary answer could be the need for describers to be in tune with styles (Ahmed and Crucifix 2018, 3–4): affordances of visual narrative and language, plus genre conventions, all of which serve as indices of creative process. Some relevant examples may be found in linkages between the broader themes, more specific topics and micro-level affordances deployed by creators: narrative switches in (documentary-testimonial) and alternating perspectives of characters in serious war narratives (Ricketts 2013, 177–180), or more playful verbo-pictorial allusions in playful narratives on the Cold War (Kraenzle 2020, 125–26) and stylistic tensions and manipulation of layout in migration stories (MacLeod 2021, 191–95) – the latter also appearing in dedicated museum exhibitions. Such examples may be context-specific, yet they are typical of how events and social groups are remembered or forgotten in comic art and its consumption. In principle, an audio describer would need to adjust their background knowledge as well as their agency in mediating some of the individual stories, voices, visual techniques and social discourses mentioned above through their framing verbal narrative.

This article will answer the question of how AD may re-mediate comics in the museum, what practical challenges AD may pose in this context and how it may be received by yet another key group of stakeholders, the users. This will be done by retracing the steps of an exploratory pilot study conducted in the summer of 2022. The study entailed pitching the idea of comics AD to a museum specialising in comic art, The Cartoon Museum (London). Once the curator agreed and research ethics approval was obtained, groups of participants were recruited in two phases: first, a creative team, consisting of the curator, four artists and a describer; then BPS visitors. The study entailed qualitative research, that is, a focused interview with the creative team and individual interviews with four BPS visitors, observed creative process (drafts) leading up to a final output of 21 min of AD and critical reading of selected AD excerpts.

In what follows, I will delve into some further theoretical anchoring (section 2) by offering a more holistic overview of memory via translation, thus teasing out the complexity of creating comics memory, also taking a broader perspective and offering a critical overview of current technological developments. This is followed by a discussion of interactions among the creative team in the focused interview, observation of AD scripting and close reading of samples, which instantiate translated comics memory in this project (section 3). Then, feedback from BPS users is distilled (section 4); although the conditions of a fully fledged exhibition were not recreated, visitor views highlight areas of strength and some limitations in AD output. Finally, the article presents concluding remarks (section 5).

2 Memory via translation: procedural and declarative memory of mediators of source texts

The Language Services Industry is currently evolving in leaps and bounds. Distinctions between spoken and written communication become increasingly blurred (e.g. speech-to-text transfer in live subtitling). ‘’Traditional’ translation between languages occurs in conjunction with more peripheral types such as intralingual translation (paraphrasing, conversion into easy-to-read language) and intersemiotic translation (descriptions) with various modalities of cross-fertilisation. For example, original sentences may be simplified through automatic paraphrasing systems, so that they are more readily translatable into several languages (Mehta et al. 2020), while AD scripts can be translated interlingually, a more efficient way to cover requirements for several locales than creating scripts from scratch for each locale separately (Vercauteren, Reviers, and Steyaert 2021). Indeed, borders between types of translation are constantly interrogated, at a time when companies and professionals tend to offer diverse multilingual services, rather than just traditional ‘document’ translation. The above developments have two important implications. First, with the omni-presence of technologies and automation (machine translation), human agency, and the meaning of creativity, are cast in sharp relief. Second, when seen from an angle of diverse technology-mediated solutions, the main purpose of translation in all its forms is to offer recipients access to individual and collective memories inscribed in texts. This allows recipients of translation services to participate in social and cultural life.

In the light of the above, any translation may entail remembering (or potentially forgetting) aspects of its inaccessible (chrono)logical starting point, the source text. Translations of official, technical documents prioritise resemblance to that source. By contrast, creative texts such as poems may go beyond resemblance to fulfil a function in the target culture, thus prioritising the transfer of style (Boase-Beier 2011, 81–82). Boase-Beier describes creative translations as blends. In linguistics, a blend is the combination of mental spaces, or cognitive structures that facilitate thinking about events and states, in way that the combination of these mental spaces both inherits and goes beyond the characteristics of their input (Boase-Beier 2011, 67).

The concept can be usefully extended to translated comics, which come with specific affordances (styles) and afterlives (archives) in different locales, simultaneously expressing and negating source- and target-culture norms. A notable explanation of this phenomenon has come from the sociology of comics (Brienza 2016), via observations of how specific profiles of field actors, (often adversarial) social positions, processes and functions shape overlapping (yet distinct) transnational fields of cultural production. Translation scholars have suggested the same, resorting to systems and cultural field theories (Kaindl 1999; Zanettin 2008). However, the latter have additionally grappled with concrete rhetorical/transfer strategies: size, format and reading direction may (not) be preserved; culturally-filtering introductions and glossaries may (not) be added; panels may be retouched and visual elements may be expunged/retouched; lettering may be transferred (less) creatively; textual meaning may (not) resemble the original, or may be conveyed through ‘translationese’ or code-switching between languages.

Ultimately, and to shift focus to the process of translation, comics as blends will be shaped by a translator’s personal memory. This consists of declarative memory of acquired (conscious) knowledge of symbols, facts, concepts or subject matter relevant to a translation task (Brownlie 2016, 27; Dimitroulia 2019: 94). For instance, knowledge of the visual and verbal style of artists of 300 (Frank Miller) and Three (Kieron Gillen) or knowledge of factual information relevant to the ancient Spartan universe of discourse they both tackle is useful declarative memory. At the same time, the translator may resort to procedural memory, relying on ready-to-use action plans: problem-solving in (small) chunks of text, deciding on degrees of deviation from the source, adopting an overall approach (Brownlie 2016, 29; Dimitroulia 2019: 94). For comics translators in particular, procedural memory hinges upon previous training and skills, inter alia, writing well, writing concisely and adopting an ethical approach towards reader needs, source-language creators or the norms of professional translation conduct (Kadric and Kaindl 1997:136–140).

The imprint of declarative and procedural memory may be seen in styles. Take the most studied comics translations, those of the Astérix series. English, German, Korean and Arabic versions display unique, idiosyncratic linguistic features from both the source and target culture, as evidenced in names, puns, accents, cultural references and educational or broader (nationalist) comments/subtext (Harvey 1995: 69; Kadric and Kaindl 1997:141–143; McElduff 2016: 152–154). Over time, styles may be rethought in order to be innovative, or in order to purposefully (de)select collective memories of social or linguistic groups. For instance, academic translators of Astérix in Greece added to the body of extant translations by creating versions in ancient Greek and in the Cypriot, Cretan and Pontian Greek dialects (Missiou 2010: 120–123).

The styles utilised a comics translation project will depend on the genre of each text, its overall rhetorical purpose, and the client’s/publisher’s translation brief. The priority in Astérix would perhaps be to deploy casual, spoken style and culture-specific references that convey the emotions, thoughts and perceptions of funny characters so that their intentions and behaviour appear equally incongruous, and therefore funny, in another language; this may require research on the art of Goscinny and Uderzo, and some careful thinking on how far to deviate from the source to elicit laughter. In the example of more serious comics AD commissioned for this project, the brief is to convey comic art in an innovative format, so that BPS communities may benefit. Priorities need to be set in terms of how much of visual source to include in the description; this includes orientation for each piece, main visual qualities of each work (colour, découpage, materiality), actions, character qualities (real or type-like) and mood (framing, colour). Most importantly, and not unlike a specialist writing descriptions for museum objects, describers will resort to a creative mix of plain description with interpretation; the describer’s framing narrative will contain some (but not all) mise-en-scène elements, with contextualising statements where necessary, based on background knowledge and research undertaken (declarative memory). The resulting narrative will also be characterised by a certain level of proximity to/distance from the listener, both in terms of how many words are used to describe art and, say, in the kind of adjectives or figurative language used. All of the above-mentioned priorities may be set without losing sight of the need to convey a coherent story that satisfies curiosity, suspense and (possibly) surprise. Finally, as the final product is voiced performance, paralinguistic features (intonation, rhythm) may facilitate characterisation, mood creation and audience enjoyment.

Here lies an opportunity to further highlight the complexity of translation as a type of remediation. A holistic focus on translation practice, be it in terms of product styles, creative process function and translator profile, may be useful in this respect. Contextualising what translators do can show the workings of personal memory, and how it taps into the collective memory of professional groups.

To illustrate, interviews with translators of Greek Aristophanic comics (which I conducted between September and December 2018), showed that none of them had training in comics translation. Instead, they relied on their multicultural backgrounds and on materials they knew would be relevant to the task: songs, TV discourse, political references, everyday conversations (i.e. declarative memory). They also noted that creative use of language in their professional roles (education and/or the creative industries) informed their practice. Interviews are thus indirect, subjective means of accessing out-of-awareness, procedural memory, which is based on personal, past experience and on habits instilled by participating in interconnected social networks. Direct evidence may be available in personal archives of translators and consecutive drafts leading up to final output. To return to Aristophanic comics, I was able to glean additional information from the diary of one of the English translators, Bella Spiropoulou. Her diary entry underscored fidelity towards the source-text intention but with a measured sense of freedom on how to promote it. In a rare glimpse into the writing process, the same translator gave access to translation drafts. With each draft, she increased the speakability of the text, as well as her distance from the source, often reproducing speech balloons on paper in order to gauge visual and spatial arrangement of translated text (see translation technology below) and to ‘see and hear’ the text on the page, as she explained. In the final drafts she involved the letterer and editor, who made valuable contributions, but whose contributions she always handled with a modest sense of confidence in her own solutions. To gain further direct access to a translator’s procedural memory, a researcher would need real-time video recordings of their behaviour and detailed drafts as they (re)write the text. This is a promising, untested (to my knowledge) methodological approach for comics. In this study, interviews with creators, drafts and communication between the describer and creators have been employed as tools for eliciting information on declarative and procedural memory.

Scrutinising translation practice shows the dissipation of memory in a web of actions and agents. Yet there is one more seat of memory: technology. Dedicated academic studies, professional surveys, blogs and translation fora (themselves collective archives) record a recent positive shift (since 2020) among professional translators towards using translation technology in creative texts (Youdale and Rothwell 2022, 383). Traditionally used for repetitive, specialised texts, Computer-Assisted Translation (CAT) tools can help compile terminology bases and customised glossaries. More importantly, they suggest partial or full matches to segments translated previously, retrieved from the tool’s translation memory (TM), they allow text searches within the same or several translated documents and retain (track-change) versions, a useful mnemonic tool when translators make final revisions (Youdale and Rothwell 2022, 387). CAT tools can be used in conjunction (or as a default) with neural machine translation (NMT), such as Google Translate, to retrieve solutions where TM is not effective (Youdale and Rothwell 2022, 386). The disadvantages of translation technology are more or less known, pertaining to such dimensions as user-friendliness and, crucially, translation quality, which may affect time and effort investment. Thus, NMT may omit text, leave words untranslated, add words that were not there in the original, or employ unnatural word order and unidiomatic solutions. Both NMT and CAT tools also force translators to work sentence-by-sentence, which may be less effective for creative texts where co-text is important for creativity. The same disadvantages have been observed in machine-translated AD scripts for films, one of the few areas where it has been experimentally piloted (Vercauteren, Reviers, and Steyaert 2021).

Despite its drawbacks, translation technology allows human creative memory and collective digital memory to complement each other. Indeed, machine translation for large-scale comics translation projects has been used since the early 2010s; see Marvel’s contract with digital content distributors for mobile platforms (<https://goodereader.com/blog/e-book-news/iverse-attains-international-language-rights-to-marvel-comics>). In the last 3 years, various startups have sought to develop CAT tools for comics. The Geo Comix platform, for instance, uses AI-powered transcription in a desktop publishing environment, which allows translators, proof-readers and publishers to annotate and work collaboratively. Imagetrans offers similar transcription, with integrated NMT and CAT features (TM and retrieval of lines from large corpora of texts). More recently, Natural Language Processing (NLP) experts have developed a fully automated manga translation system (Hinami et al. 2021). The system relies on state-of-the-art visual and text processing of Japanese manga, working on text and context simultaneously: textual links across speech balloons, reading direction, scene make-up and semantic tags for visual elements. The system comes with custom-made NMT, which has been trained with data from large corpora of film subtitles and a ‘gold-standard’ bilingual samples of published manga translations (258 pages) and human translations commissioned especially for the development of the system (214 pages). The output was assessed with an NLP quality metric, arguably less relevant or even problematic for comics texts. Bilingual manga readers also rated it for accuracy (sic). This is a highly interesting development, albeit with drawbacks. I looked at a randomly selected English translation from the publicly available dataset and detected minor problems (capitalisation, alphanumeric values, spelling) as well as more distracting ones (missing definite articles, obscurity of expression, lack of consistency). A critical linguist might also say that translation quality may be more complex an issue than suggested (consider creative ‘knots’ such as irony, conversational moves, dialect and im/politeness).

Further technological solutions are on the horizon, which are bound to change human agency in translation decisions. Large-language models (LLMs), such as those of ChatGPT, are already being pressed to the service of translation by AI specialists (Jiao et al. 2023). AI-powered writing assistants are also developed (Ippolito et al. 2022). Whilst situated further in the future, the scenario of such tools affecting workflows of those who create verbal translations or descriptions of images is not inconceivable. On a minimal level, tools may generate initial suggestions for a first draft, thus shifting creativity norms towards an art of harnessing technological suggestions, to paraphrase Assmann (Assmann 2011, 130, 202). The Comics AD project here is perhaps the most incongruous scenario to this future reality as a human audio describer was asked to take on a novel task. She consequently transformed content in an original way. Her creative task will be discussed in context in the following section.

3 Prefiguring textual memory in audio-described comics

Accessible comics constitute a fairly recent field of investigation. The website of academic and comics author Nick Sousanis offers an excellent overview of accessible comics for the truly overlooked group of blind and partially sighted (BPS) users (<https://spinweaveandcut.com/blind-accessible-comics/>). Accessible comics include, inter alia, tactile comics, web comics read by screen readers, alt text (text embedded in code for images) adapted for screen readers, and AI-enabled reader systems that incorporate audio books or e-books – the latter being supported by cloud-based speech synthesis (Lee et al. 2021). The plethora of artistic ideas and developer initiatives indicate a gap in available options for BPS users.

To my knowledge, there has been no attempt to use ‘traditional’, human-voice AD, a form of blend displaying characteristics from a visual source and using narrated output to remediate comics memory - with one exception: Sousanis and Beitiks' Comics Beyond Sight completed in June 2023 (https://www.technologyreview.com/2023/06/28/1074341/comics-beyond-sight/). Similarly, no study has focused on institutions of memory par excellence, museums, which thrive on translation in three ways. First, experience may be translated into an exhibition, thus informing collective and individual memories of certain events (cf. migration-themed comics in section 1) (Neather 2022, 157). Second, a museum imparts knowledge by linking past experience (emotions, minds, bodies) to facts and new knowledge (Neather 2022, 161). Third, they are multilingual spaces, with originals and translations contributing slightly or markedly different perspectives (Neather 2022, 162), as AD might do.

The pilot study (May-July 2022) consisted of recruiting participants and conducting a focused interview with the creative team: four comic artists, an audio describer and the museum curator. This was followed by scripting, voicing and mixing the AD. A focused discussion among the ‘creatives’ was deemed to be a fast and effective way of bridging perspectives and background information, that it, for sharing declarative memory. Short bios and examples of film/cinema AD from the Royal National Institute for the Blind (<https://www.rnib.org.uk/2018-audio-description-campaign>) were shared beforehand. Artists were asked to bring along one work they thought represented them best. Three of these works, two graphic memoirs (a, b) and a comic book with complex materiality (see c below), were later audio described:

1. Coma (Slatery, 2021, book format, 278 pages): tackles the theme of surviving a serious infection and a drug-induced coma whilst hospitalised;

2. Satin and Tat (Trirstram, forthc., digital comic, 1 page): presents a teen Goth girl growing up (and dating an older man) in Devon in the 1980s (and returning to Devon decades later);

3. Threadbare (Brookes, 2019, embroidered, then photographed and printed on paper, 56 pages): overhearing a train conversation among two old women remembering the last time they were in love, while another passenger coldly watches porn on their mobile.

All three memorialise the embodiment of experience (Miller 2018, 103) in distinct styles: fantasy-evoking colour drawings juxtaposed to black-and-white découpage (a), vivid colour-pencil feel in a digital format (b), negated tactility, or the appearance of embroidery in print format (c).

I asked the creative team to respond to the following scenario:

The Cartoon Museum has the resources to prepare a special exhibition on comic art for BPS visitors, and you have been invited to contribute. What would make this exhibition successful and effective?

A discussion (1 h 40 min) followed, with participants taking the lead and responding to each other’s comments. This was transcribed, tagging each contribution – DE/audio describer; CUR/curator; CA#/comic artist – and analysed qualitatively first with descriptive, in-vivo, emotion, process and values coding (524 coded segments in total) and then with focused coding (Saldaña, 2012, pp. 111, 216) revealing how participants understood mediation of art in this context. Emerging codes coalesced under two main themes: institutional context and artistic habitus.

Institutional context covers the staging of samples for public, non-iterative reading, multimodal reinforcement (large-size samples, tactility, ambient sounds), explaining the artistic process (judiciously) and knowing what audiences need (in light of BPS diversity and comics literacy). Artistic habitus covers the aesthetic complexity of comic art (plot, discourse, mediating perspectives) and the balance needed between story and graphic style, especially in autobiographical work. As Brookes put it:

[…] you’re in the subjectivity of the artist, every mark is a mark of the artist’s body so there’s like this sense of being with the artist in comics all the time and sometimes you withhold that […] and sometimes you accentuate that (CA4, Focus Group Interview, Threadbare)

Given the resources and the exploratory nature of the project, some ideas did not materialise (ambient sounds, co-designing with BPS users). Yet others stimulated new procedural memory for the task: writing informative introductions (see below), printing large-size samples for low-vision visitors and selecting sequences with coherent narrative arcs – sample descriptions here defy the typical 3-min norm in museum guides as it takes time to establish a scene (Threadbare, 15 panels in 6 mins; Satin and Tat, 5 panels in 5 mins; Coma, 39 panels in 15 mins).

During the discussion, the describer mentioned professional norms in her field, such as using recognisable vocabulary and writing economically, but she also noted the unique nature of each work, asking for the creators’ help at drafting stage. After the focused discussion she produced two drafts, with 113 minor revisions between drafts, mainly syntactic reversals, spelling and lexical adjustments. The very few changes were not only due to her pre-existing artistic habitus rooted in 30 years of film AD experience but also due to collaborative input. Artists provided information on theme, topic, setting, characterisation, artistic process and format. They were available to verify factual and visual information too.

For example, during the focused interview, Zara Slattery explained synaesthetic interconnections (sound, pain sensations and image composition). Unresolved questions were followed up in draft feedback, as shown in Example 1. Here, Slattery corrects ‘monster’ to ‘creature’, a more accurate signal of the topic of suspended existence. This continues to disambiguate visual information. Similarly, Myfanwy Tristram highlighted the differences between mother and daughter (topic, character voice) during the interview. She dwelled on setting specifics in particular, including story time, furniture and objects in the kitchen (where the action unfolds), an instance of which is shown in Example 2. Two of the perennial issues in AD are what visual information to include and how to avoid over-/under-interpreting (Fryer 2016, 56, 166). The artists assisted in both areas, guiding the describer’s gaze and allowing access to their own procedural memory. They directly influenced her interpretation of given panels, as well as the selection and number of words used. Thus, Tristram’s focus on cultural references resulted in a detailed 223 words-long AD of the setting in Satin and Tat, including the Voyager 1 space probe photo (Pale Blue Dot). Slattery’s explanation of visual metaphors allowed privileged access to artistic intentions and a richer description of style.

Example 1: Page 111: The creature is the coma itself, I’m carried in the belly of it and it’s through the cracks of its door (head) that I view the world [TOPIC]. The infection is the serpent [TOPIC]; it has a tight hold of me within the coma [TOPIC]. (CA2, draft 1 feedback via email, Coma)

Example 2: There’s the Pale Blue Dot in the calendar which is just to indicate that mum’s a bit of a hippy [ASPECT OF SETTING/CHARACTER] (CA1, Focus Group Interview, Satin and Tat)

Arguably, the most impactful contributions were made by artists when they introduced their art at the focused discussion. As they aired their views, they were able to unpack aspects of individual artistic habitus. Their contributions were recorded in the interview transcripts, which were available to the describer as mnemonic devices soon after the meeting. Thus, the describer tapped into personal and collective memory to frame the interpretation of each piece, as in Example 3. The example contains the introduction to Threadbare (paragraphs a-e), a transition (paragraph f) and storyworld description (g-i) in this museum sample. Paragraphs correspond to the ‘takes’ the describer created in her script (original formatting and emphasis) (see also Figure 1):

Figure 1. Threadbare, pages 29 and 39–31 © [Gareth Brookes]. Reproduced by permission of Gareth Brookes.

Example 3:

a. […] The tactility of the book […] heightens the idea of desire and regret [THEME] […] You look at images that create desire but which you can’t touch [THEME] […]

b. Threadbare is made using hand embroidery on calico [PROCESS] […] the back of the embroideries get tangled into the embroideries opposite, and the turning of the page allows the reader to see the backs and fronts of each embroidery. [PROCESS]

c. To Gareth, the backs are a metaphor for the messy feelings we keep hidden beneath the much neater surface which we present to the world. [TOPIC]

d. In order to be able to show this work in a book format, Gareth had to use a scanner to scan in the embroideries and then assembled them in photoshop. [PROCESS]

e. There are two distinct aspects to Gareth’s telling of the conversation he overheard [TOPIC]. Firstly the verbatim snippets of dialogue, which appear like short blocks of text as you would have them on a mobile phone, but this is embroidery, so texts appeared embroidered onto a small grey wool strip, in back stitch, no bigger than two inches by one inch on a plain, pale, brown background. [FORMAT]

f. There are two stories to Threadbare, but we’re only going to be looking at one of them, only one of the women’s reminiscences.

g. Oh! I only loved one person, a boy. We were fifteen [SPEECH BALLOON]

h. In the next two panels, a wool text block says So long… [SPEECH BALLOON] the panel unravelling at its edges and threads of white wool wrapping themselves around a naked young woman with long auburn hair, flawless white skin, bright button eyes and a carefree smile. Small breasted with just a hint of pubic hair, she seems to be dancing. Arms outstretched, one leg bent at the knee. [CHARACTER]

i. On the next panel, the text continues: Ago Now [SPEECH BALLOON] and below it, we see the back side of that same embroidered girl, with a tangle of white, orange and brown threads, reaching up to the text [CHARACTER]. [AD script, Threadbare, original emphasis]

The level of detail in the introduction (paragraphs a-e) can be traced to dialogue with the artist, whilst the physicality in character description (paragraphs h-i) harks back to the describer’s long film AD experience. The introduction offers a gradual progression from an overarching theme, to topics, to creative process and format (highlighted above). It therefore helps build a mental model (Fryer 2016, 45) of the world created in a new piece of art, in this particular style. The introduction creates cognitive shortcuts (schemata) (Fryer 2016, 46) that listeners may carry over to their experience of the rest of the story. This experience is supported by external input from the describer, combined with what listeners already know about objects (embroidery, texture), technology (mobile phone, texting) and human behaviour (growing up/old, having sex, dancing, smiling). The next section will offer an overview of visitor reactions to these descriptions.

4 User responses

Visitors to The Cartoon Museum self-selected after a recruitment message was sent in newsletters of two London-based BPS charities social media groups. Eight responders expressed an interest and four agreed to a mutually convenient time for a visit to a hired studio at The Cartoon Museum. Each visitor spent circa 44 min in the studio. Half of this time was dedicated to sample listening, with the other half dedicated to feedback. Once the interview was over, museum staff gave visitors a general introduction to the current exhibition. One of the visitors (female, P1) identified herself as blind, one as ‘closer to blind’ and with Charles Bonnet syndrome that causes visual hallucinations (male, P3), and two (one male P2, one female, P4) as adventitiously blind.

Three themes emerged at these sessions, cultural consumption habits, AD performance, and AD comprehension.

In terms of cultural consumption, P3 and P4 said they ‘rarely’ visited museums due to lack of audio description or individual docent tours, while P1 and P4’s responses were ‘sometimes’ and ‘frequently’, respectively. They all agreed that increased AD provision is important in all areas of the creative industries, including museums. P2 had several suggestions on physical access (easy-access stairs, lighting) and on multisensorial input in a museum context. Visitors referred to their comics literacy too, which ranged from familiarity with and a preference for funny cartoon strips (P1, P4), to adult humour comics and satire (P3), to adventure, superhero genres (P2). References to these genres elicited memories of reading with family members (P2, P4) and fond appreciation of funny drawing styles (P3, P4). Allegiance to genre was stronger for P2, who expressed his disappointment with genre selection in this project; he had indeed expected sci-fi or adventure before visiting The Cartoon Museum. This expectation negatively affected his level of engagement and comprehension (see below). Two visitors, on the other hand, expressed a pleasant surprise in finding out that AD samples represented serious topics that is, in juxtaposition to cartoon strips they were used to. As they noted

Example 4:

[I]t’s a little bit more accessible when it’s more realistic, so you can relate to it better. You don’t feel like you’re reading like, you know, your baby cousin’s comic or something (P3, visitor transcript)

Example 5:

I just felt there was a lot more detail in there than if I was just looking at it in book form […] giving a sense of what people were going through and trying to deal with the situation (P4, visitor transcript)

Predictably, responses confirmed diversity of individual or collective memories of reading. They also indicated the potential for broadening visitor horizons, both in terms of genre knowledge and the possibilities of different formats, with AD offering a more informative and engaging experience of a story (as opposed to audio books, for instance). Comments suggest further possibilities for a tailored approach to individual tastes and needs. P3, for instance, who was able to catch a glimpse of the printout before he listened to the AD, praised comics AD as it allowed him to concentrate on the story without running the risk of having visual hallucinations (Charles Bonnet syndrome). It is conceivable that AD itself may be adapted further to cater for the needs of visitors with different memories of reading and visual art or sensory and neurodiversity-related requirements.

In terms of comprehension, the AD was overall characterised as ‘clear’ (all participants), including the presentation of characters and their thoughts (P1, P3, P4). On the other hand, the need for more detail in the description of secondary characters (P1) and of the main character in a particular scene (P3) in Coma was highlighted. One participant, P2, generally found all narratives confusing. Lack of comprehension here may be attributed to thwarted genre expectations (see above). This is a visitor who overtly stated a dislike for personal narratives or memoirs, adding that this dramatically affected his level of concentration and enjoyment. Overall, visitors praised the AD for its ‘simplicity’ (all participants, including P2) and for being ‘detailed’ (P1, P4). Detail on the art was appreciated, be it in the overall connotative values of colours (P1, P3, P4), the physical appearance of characters (all participants), their behaviour (P2, P3) and their emotions (P1, P3, P4). As P4 noted, ‘I think one was sea green, and then there was darkness representing chaos. Those were the things that […] wouldn’t have occurred to me [… it] almost reminded me of, you know, when one goes to the theatre and one listens to the audio description’. The amount and sequencing of detail, however, was occasionally a double-edged sword. For example, P3 and P4 would have preferred more description of the characters, less on the setting of Satin and Tat (see section 3). P1 would have liked more explanation for art terminology (e.g. ‘brushstrokes’, ‘watercolour’) and more detail in the art of dialogue-rich panels in Coma. The piece where opinions converged was Threadbare. Here, visitors do not always understand how threads in this embroidered format interlocked in given panels, as shown in this comment:

Example 6:

[…] when we talk about the image of the woman and we talk about threads, including at that point the brown thread, we understand, I understand what that probably looks like. I wouldn’t stake my life on it, but I feel I know what it is. And I feel the power later on in the panel when that same image is replicated but without the brown and I understand the artistic significance of that. But the later panels that follow where they’re talking about his dark hair and the threads from that over her body and the threads of her hair, I’m lost at that point (P1, visitor transcript)

As was shown in Example 3 (section 3), introductions to each piece were intended to integrate comics memory and, ultimately, to help listeners create a mental model of the storyworld through cognitive shortcuts for theme, topic and the (subjective) telling of the story. All four visitors consistently appreciated the usefulness of introductions to each AD sample. This included P1, who pointed out how effectively Threadbare had been introduced; in her words, ‘you can feel the effort in the art and the tactile artistry of seeing the neatness of the front and the chaos of the back is […] such a powerful visual image and so instantly understood against what’s being said’. Lack of comprehension in Example 6 must therefore stem not from the type or amount of information provided but from signposting. Brief bridging statements could have further explained the links between old information (mental model triggered in the introduction), recent information, such as an outstanding stylistic feature in a given découpage, and new information on the current panel. As listeners proceed in linear fashion through the story, it is important for the describer to use unobtrusive devices for jogging their memory and for creating unity across panels.

The third theme focuses on AD performance, which was described as ‘very good’ by all participants. Visitors appreciated how a human voice stood out (synthetic voices in screen reading technology being the prevailing norm), how the tone of delivery was aligned with the storyline (P1, P3, P4), and how slow the pace became enjoyable (conversely, lapses to fast tempo of delivery in some instances were criticised; P1, P2, P4). At close inspection, delivery (see Bosseaux 2015) in the introduction (full text in Example 3) is done in soft voice, mid-to-low pitch, slightly nasal, in narrow range (but wide range for speech balloons, giving them a sing-song intonation to reflect older age of the characters) and slow-to-mid tempo. These voice qualities may help explain the positive feedback, as visitors seemed to have expected a voice that was ‘tonally interesting [in] the way things were emphasised’ (P1), indicating ‘change of impression, delivering style’ (P2), a voice that can ‘actually put you in the moment’ (P3) and ‘put expression in words’ (P4).

5 Concluding remarks

When the call for participants for this project was disseminated, one comic artist responded negatively, predicting AD to be little more than incoherent action description, as it is arguably impossible to convey the artist’s vision, the painting-like (sic) complexity of the product or the story’s pace through someone else’s words. Such objections may ignore the embodied performance inherent in reading comic books (Hague 2014), the tradition of audible comics (see call by Comicalités <https://journals.openedition.org/comicalites/7593>), or the alleged origins of the AD industry in comics descriptions over the radio (3Playmedia: online 2022). The above-mentioned objection does, however, cast in sharp relief the challenge of conceptualising, as a blend, and then delivering comics AD for a new type of audience and setting.

The pilot study showed the feasibility of comics AD and followed exploratory approaches in museum practice in that was object-led; in other words, it veered off any pre-conceived ideas on art and focused on the possibilities and questions created by objects, on art as a way of knowing the world (Bjerregaard 2020, 4). By bringing together a creative team, the project stimulated new procedural memory for the task, one that is based on an ad hoc collective memory in two professional fields with no point of contact before – to my knowledge, such collaboration between describers and artists has not been attempted before. As was shown, such dialogue allowed privileged access to artistic intentions, one that sighted users do not have, arguably creating positive bias for BPS users. A future study could test further personalisation, dialling up/down aspects of the visual source, as per feedback collected in this study; for instance, more links between local aesthetic elements to the overarching theme may be deployed and a different balance between setting and characters may be achieved for different users. The output could then be combined with other modalities of translation (see section 2), say, paraphrasing to tailor texts to neurodiversity needs, or with interlingual translation, thus tapping into textual traditions beyond a given language (and English, as a lingua franca). Overall, the approach can serve as the basis for opening more comic art collections to BPS visitors and sighted visitors alike.

In terms of limitations, a knowledge-in-the-making approach (Neather 2022, 158) was not possible in this small exploratory study, even though this possibility was mentioned in the focus group with the creative team. A larger, follow-up project could concentrate attention to synchronous co-creation with BPS users instead. This would yield complex and highly valuable data on individual creativity as well as the creativity underpinning a model of community inclusion.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by The Enhancing Excellence in England project (AN1999).