Interview: Tessa Dwyer, author of Speaking in Subtitles: Revaluing Screen Translation

Cover of Speaking in Subtitles: Revaluing Screen Translation by Dr. Tessa Dwyer, showeing a film still of a young Asian couple in a dramatic setup, with the subtitle, "There's something I haven't told you yet."

RW: Hi, Tessa! Thank you for agreeing to this interview. I found your book  Speaking in Subtitles: Revaluing Screen Translation very timely, and it provoked many questions and some new thinking for me.

I started my university studies in translation, but I was surprised to learn about “value politics” in translation, which certainly wasn’t something I heard about 35 years ago. Could you provide a sort of elevator-pitch definition for readers?

TD: Perhaps because I come from a different disciplinary background – Film and Screen Studies – the “value politics” of translation immediately stood out to me when I started to engage with intercultural viewing practices and, especially, subtitling and dubbing. In fact, it was an encounter with “value politics” that really sparked my sustained interest in the topic. I was writing about Hong Kong action films in my MA thesis, using some French critical theory. My supervisor suggested I read the French theory in the “original,” yet had no qualms about my viewing of subtitled Hong Kong action films. Obviously, there are hierarchies in place about when and why translation does, or doesn’t, matter. What I found ironic was that a very learned translation of a French theorist by someone with expert knowledge of the field was not deemed worthy of serious analysis, while the less than stringent (to say the least!) subtitling of the Hong Kong film industry flew completely under the radar.

In Film and Screen Studies – especially Anglophone film theory – translation is so undervalued and un-theorised that it is almost entirely invisible. Despite the canonical centrality of European filmmaking, for instance, in the development of film theory and culture, the role of translation and the inter-cultural basis of much theorisation is almost entirely ignored. Translation speaks to reception contexts, over those of production/creation, and for this reason, it is often regarded as utterly inconsequential or, worse, as an affront to the creative process and to authorial vision. In this way, translation threatens the core stakes upon which so much of film and screen culture remains invested. That, I guess, it why I find it so fascinating and why I love how translation can demonstrate in myriad ways how the very distinction between production and reception breaks down. Everyday practices of subtitling and dubbing can really challenge so many assumptions and biases in the way we understand and discuss film and screen.

So much for an elevator pitch!... more like a meandering rumination.

RW: That’s great: all helpful!

You discuss critiques of subtitles which include elitism. Do you think wider access to film and video through prevalent video-on-demand streaming services is reducing this problem, which perhaps was more of an art-house issue for foreign films in the past?

TD: This is certainly something to consider. The disruptive influence of streaming platforms is immense, and as I argue in the book, the global media flows enabled by online networking are affected, at every turn, by language difference and translation. These recent industry shifts really bring issues of translation to the forefront of our changing media landscape. So yes, I think that streaming services are set to impact significantly on attitudes to subtitling and dubbing, yet it is too early to tell how this will play out. In 2014, there were predictions that Netflix would cause the demise of dubbing within Europe by providing timely access to content in its original language. However, by 2018, Netflix was streaming dubbed versions of shows by default, claiming that even when audiences insist they prefer subtitling, dubbing keeps more people watching.

RW: You cover issues around translation studies in your book, and current focus on content accessibility has certainly made this area more important than ever. Do you see audio-visual studies increasing in popularity, either as a result of demands for accessibility or because of the globalization of video content (VODs, gaming, etc.)?

TD: Yes, as I mentioned above, I think that the advent of streaming services is increasing attention exponentially on screen translation and localisation (including fan translation and crowdsourcing) and hence, burgeoning areas of research are emerging within Translation Studies. Content accessibility is definitely on the agenda in terms of industry regulation and policy, while global streaming services are having to prioritise translation and localisation. In 2017, for instance, Netflix launched the custom-built HERMES subtitling and translation test and indexing system, which it claims will allow them to “resource quality at scale” through standardised testing and unique identifiers, enabling it to use “metrics in concert with other innovations to ‘recommend’ the best subtitler for specific work based on their past performance.”

RW: Cultural misappropriation in the arts is a hot topic at the moment. Can you share some advice for young or emerging filmmakers, who might be trying to be more creative in order to get a foothold and visibility in a noisy film climate, about how and why to avoid détourning?

TD: Well, I think cultural misappropriation is an ongoing (perhaps necessary) risk attached to many forms of intercultural communication and creation. Détournement was a radical, activist strategy that sought to upset boundaries and challenge modes of thought and politics. It didn’t shy away from cultural misappropriation, but rather, confronted it head-on. It set out to offend and to shock. My take on all this is that intercultural modes of production and reception are vital, essential elements of mediatisation – no matter how risky. We need to recognise this and consider the complexities of translation involved in everyday practices and modes of engagement. I would rather that misappropriation continue to surface as an issue, than that creatives simply avoid engaging beyond their own safe cultural borders and boundaries.

RW: You talk about abusive and corruptive translation and quote Derrida about translation: “... it necessarily violates even as it devotedly follows or respects the original.” As a copy editor of books, I find my profession needs to walk a fine line between being “at once violent and faithful” in helping but also maintaining the author’s voice. “Nornes locates translation abuse within populist practices like anime fansubbing.” I feel the same way about self-publishers who think Grammarly can replace professional editing or who just want to ignore all writing conventions in the name of creativity. But your book seems to make a reasonable, unemotional examination of fansubbing. You changed my black-and-white thinking about it—well, brought my righteous indignation down a notch or two! Just as editors should not encourage grammar policing, what can you say to people who really bristle at fansubbing?

Let’s start with a provocation: maybe translation is, at heart, a fan activity?

TD: Let’s start with a provocation: maybe translation is, at heart, a fan activity? What motivates someone to labour so intensively and minutely with another’s text or creative work, if not some form of respect, devotion or fandom? Of course, the professionalization of the industry means that naturally many translators now routinely labour on works they do not love in any sense, but if we try to think about the origins of the practice, in scholarly and religious contexts say, the fan sense of investment holds.

Speaking from outside the field of professional translation – without the need to defend my own territory – I think it’s easier for me to appreciate the creative and sometimes subversive nature of fansubbing. Also, I’m interested in what fansubbing shows us about global media industries broadly. Fansubbing alerts us to very interesting things that are happening within global media flows, articulating gaps and loopholes, challenging politics, re-purposing technologies and, in some ways, helping to shape the future of global media industries.

Fansubbing is thought to have begun in the US when TV networks stopped broadcasting anime titles like Astroboy and Gigantor. Fans simply went in search of content themselves (sourcing video tapes directly from Japan or Hawaii), which then needed to be translated. As they set about translating for themselves, they discovered the extent of cultural adaptation/appropriation and reworking involved in the US TV broadcasts, and came to see their own translations as more faithful and authentic, and ultimately as safe-guarding the texts. This history is important as it shows how professionalism is by no means a guarantee of quality, due to corporate agendas, industry conventions, cultural attitudes and others factors.

Also – I should mention that many professional audiovisual translators are themselves very interested in fansubbing, and feel that there are many lesson to be learnt. Minako O’Hagen, for instance, notes the benefits of collaborative, peer-to-peer working environments with in-built feedback and mentoring mechanisms. O’Hagen and others also point to the value of expert genre knowledge as something that the industry is learning from the example of fansubbing. Netflix’s Hermes tool is a case in point: the aim is to match the right translator with the right content.

... we should value, not fear, fansubbing...

One of the major reasons why we should value, not fear, fansubbing is due to the fact that many language communities around the world are underserved by online offerings and by professional translation. Collaborative fansubbing provides a means to do something about the inequalities that persist in online modes of screen media access. While Netflix has expanded into 190 (out of 195) countries across the world, it only supports around 20 languages. The Netflix Vietnam service, for instance, offers a very limited range of Vietnamese-subtitled content, and so, once again, viewers resort to fansubs, using websites like subscene.com.

RW: Some people might be surprised to learn about subversive and spontaneous translation of films by audience members; online, I recently learned about lektoring. These brought to mind my days watching shadow-cast performances at The Rocky Horror Picture Show! You also talk about the “participatory” nature of today’s popular and public realms in the area of media consumption. Recently, an article I had posted, about the “good enough” attitude to captions being unacceptable particularly in terms of accessibility, was criticized by a competitor as being too simplistic. I know your book focuses on debates around translation in subtitles, but what’s your opinion on accepting a “good enough” level of captioning? (And you don’t have to agree with me. )

TD: I think it’s always important to advocate for high standards in captioning and other forms of media translation – especially in relation to policy guidelines and regulations. Yes, good enough is not an attitude that industry bodies should take on board, nor translation professionals. And yet, I would never want to dismiss the efforts of amateur, volunteer and community translators, who largely labour at the task of translation in response to industry gaps. I agree whole-heartedly that machine translation can never substitute for human translation and perhaps streaming platforms like YouTube that offer automatic captioning tools are creating such a misconception. The fact that captioning is often unedited is indeed a sign of discrimination and shows a lack of commitment by governments and media industries. It’s an important issue, and one that I think fansubbing and DIY captioning can actually aid. The battle isn’t against amateurs lending a hand where they can – it’s about governments and corporations avoiding their responsibilities and obligations. This is largely what fansubbers are also battling against: lack of access. So why not join forces and get fansubbers to champion the cause and help advocate for change? (n.b. Viki did this when it joined with deaf actor Marlee Matlin in the Billion Words March campaign.)

RW: If I ever teach a course on caption and subtitle editing, Speaking in Subtitles is going to be one of the books on my required reading list, and it really should be a staple on cinema studies intro courses. Although it’s academic, it’s packed with interesting information for general readers that will open their eyes to subtitling and captioning issues that go way beyond craptions and typos: literacy, ethics, politics, media piracy and guerilla efforts, cultural capital, interactivity, quality control, “thick translation” and User Generated Content, massively open translation, CT3—community, collaborative and crowdsourced translation, and Viki. Even the term animé is demystified. And thank you for setting us straight on the word for @#$%&! to represent prohibited expletives: grawlixes or “obscenicons” (Dwyer, pg 120; Díaz Cintas, pg 13). Finally, you’ve provided me with the terminology I needed for a future article I’ll be posting about more creative applications of captions: “integrated subtitles.”

Is there anything else you’d like to share? Perhaps something that didn’t make it into this book?

TD: I’ve published recently on barrage cinema in China (where viewers text comment onto the movie screen) – which relates tangentially to subtitling as a text-on-image mode! I’m also developing a fansubbing project around an in-production Spanish-language web-series called Distancia (watch the trailer here).

RW: I love the discussions around language and vocabulary in the barrage cinema article (“assault,” “bullet subtitles,” “hecklevision”!), and I'll keep an eye open about Distancia. Thank you again, Tessa!

TD: Thanks so much for this positive feedback. It’s truly gratifying to hear that you have found something of value in my book (despite its occasional forays into academic abstraction), and that it even has use for someone working in the industry. I really appreciate your thoughtful comments and enquiries and look forward to catching your next post. So, the pleasure is all mine – thank you!

 

Headshot of Dr. Tessa DwyerDr. Tessa Dwyer is a Lecturer in Film and Screen Studies. Prior to arriving at Monash, she taught Screen and Cultural Studies at the University of Melbourne, and worked as a researcher at the Swinburne Institute for Social Research, Swinburne University. Tessa is a member of the inter-disciplinary research group Eye Tracking the Moving Image (ETMI) and president of the journal Senses of Cinema (www.sensesofcinema.org).

Tessa’s research focuses on screen translation, language difference and transnational reception and distribution practices. She holds an Honours and MA degree in Fine Arts (Film) from the University of Melbourne, and a PhD in Screen and Media Culture, also from the University of Melbourne.

Dear VODs: Stop Blindfolding Deaf People!

Close-up of CC #NoMoreCraptions button. CC logo in black and white on button, pinned to black leather bag. VWells image of Rikki Poynter button.

I often can’t hear sound effects in shows and movies, so I use captions so as not to miss anything. Because I mostly use Netflix to stream (I don’t have a TV; I sometimes rent/borrow DVDs), I was curious to know how other VODs (video on demand [streaming] services) fared in the application of and care taken with subtitles and captions. It’s not just the D/deaf/hard of hearing who use captions for same-language film and video, and access to global programming is making good subtitling a must more than ever.

This is my survey of captions and subtitles on some common VODs. As in my survey of cinema access for the D/deaf in Toronto, when I contacted these companies about problems I was transparent, with Reel Words info in the email signature. I also approached this as an everyday user. Where necessary, I paid out of pocket to get the services. I’ve assigned a star rating system for overall application and treatment of captions/subtitles.

D/deaf people who sign have the right not to have their hands and arms restrained because it prevents them from communicating. What I discovered is that these providers might as well be blindfolding the D/deaf/hard of hearing. They can’t see the content that isn’t provided for them.

Google Play 

I had heard that The Silent Child (2017) was on Google Play for $2.99, and I obviously wanted to see it for the storyline and use of ASL. It turned out the film could only be placed on a wish list for when it was made available on Google Play in the future. YouTube Movies says it is not available there.

I searched for some free movies and tripped upon the sociologically fascinating (although perhaps not intentionally…)The Creators, made “in conjunction with YouTube” and boy, did that ever show!

It was a sort of advertorial documentary about young YouTube phenoms in the UK making their living through that platform, all with different…talents. Here are two I’d never heard of, Niki and Sammy, branded online as NikiNSammy. Not quite sure what their talent was aside from having sprung from the same egg, but let’s focus on the captions as they were used throughout the doc. Here’s an example of YouTube’s idea of accessibility:

Young adult twins side by side, incomplete captions are And that's really; Which is amazing
Screenshot from The Creators on Google Play

 

NikiNSammy’s captions were split, both left-aligned (which is not helpful for twins…) and these two titles containing 36 characters was only up for one second. Now, the most current Netflix guideline is 20 characters per second. While it was a UK-filmed short doc, the spelling used was both British and American. The worst offence was that it seemed they really did use YouTube automatic captioning because there were constant errors, such as captioning react for interact and real for raw. The caption bands jumped all over the screen, as if placed for the cool factor—with absolutely no understanding of what captions (CCs) are intended for. A professional titler was not employed, and no QC person could have reviewed it. Clearly the producers didn’t give a hoot about accessibility. Thoroughly appalled at Google/YouTube.

Amazon Prime Video ★★

The Silent Child wasn’t listed on amazon.ca’s Prime Video.

So I decided to take the 30-day free trial to watch another example and promptly cancelled to avoid the $79CAD/yr fee.

I decided to go with Robin Hood (Ridley Scott, 2010), a choice influenced rather arbitrarily by my recent participation at an axe-, knife- and archery-target place. I wasn’t sure if the captions used were made for this stream or were from a previous production and distribution. As is common, the captioner was not credited.

The font was crisp white on black below the screen or with a shaded dark band for a top line placed above that on the film. But there were great inconsistencies: [ALL SCREAMING] vs only [CHEERING] when all were cheering; poor and inconsistently styled offscreen speaker IDs; poor choices in deciding on the most relevant sound effects, such as dogs barking but not men yelling in attacks or falling off walls; the King of France (should be king if not using a personal name); Jimmy boy as a name (should be Jimmy-boy or Jimmy Boy). In short, the usual problems that I see constantly on Netflix. Visually accessible but not textually accessible, which is a huge part of the game! As I often comment, unedited titles cause reader stumbles, which cause the user to lose concentration and thus comprehension.

Group shot of warriors and brigands under a tent at night. Caption: When you had us herd 2,500 Muslim men, women, and children together, (incomplete)
Screenshot from Robin Hood (Ridley Scott version) on Amazon Prime Video (amazon.ca)

 

Here, two and a half thousand should have been used, but space was tight; however, space could have been made by forcing an earlier and split title. Not great, but Russell Crowe did not say “twenty-five hundred.”

Good points: the ability to rewind OR fast forward by 10 seconds, and three subtitle format options (the fourth was black on black…).

Netflix ★★

Netflix’s captions and subtitles were the raison d’être for this survey. I was curious to know if other VODs were as ineffective and disregarding of Deaf/HoH needs as Netflix. (You can read a lot about my feelings on Netflix on my website on the blog tab and see various #CraptionFails on the Gallery of Fails tab.) No point in repeating it all here, but suffice it to say that despite (or perhaps because of) having the monopoly on VOD, this machine has grown too big to have serious quality control of captions and subtitles. The failure of the Hermes Test, the lack of qualified QC (“Master-Level Quality Control”) editors, and the low pay leave it just a step above YouTube craptions. Really and inconsistently poor. I’ve only had about four or five examples in my portfolio of well-done captions with this provider. See my website and blog posts for many illustrated examples of the problems and how they should have been avoided.

Sundance Now 1/2★*

This  check started with a series of emails to customer service because I couldn’t find the CC/ST button on the interface for some movies I wanted to review. Also, the overall platform is clumsy and annoying to navigate; the only good point is the ability to rewind 10 seconds but—unlike Amazon Prime—there’s no option to fast forward by increments.

The customer service rep kept insisting they were there if only I would use a supporting device and look in the right place, but they weren’t. His final email said in part: “I checked Off the Rails and Julia, and unfortunately we don’t have captions available on those two titles right now—- [sic] while the majority of our content is now captioned, we’re still working on updating our catalog.” Now this may be true. I didn’t check all of their catalogue. But my random selections of movies did not have captions.

Incidentally, I used the search box to find films with the keyword deaf and was provided with 13 suggestions on death content.

The usual problems with missing words, wrong words (Yeah for Yes) and letters, speaker IDs, punctuation. Overall it was very sluggish. Read: inaccessible. Note these captions, all of which lagged, from Broken Flowers (2005) with a terrible line break.

Exterior shot, two men speaking seriously. Captioned A few years ago now, mate; Yeah, well, you'd hope so; She's dead you (incomplete)
Screenshot from Broken Flowers on Sundance Now

 

The worst part was that, no, I didn’t read all 17 pages of the Terms and Conditions. I had stopped at the part that says [sic]: “7 day free trial, you can cancel anytime” which, due to the lack of copy editing, I read as you had seven days free and you could cancel anytime. NOPE. $59.00 down the tube. You can “cancel” your subscription but you’re free to watch for the rest of the year. In other words, you’ve cancelled a renewal next year. The only reason I’m not freaking out about the $60 is that I thought This Close, even at six episodes (and six “discussions” post-watch) was worth the money, especially if season two comes out within my subscription year.

So overall, the captions were terrible. But here’s the thing: for This Close, they were flawless. Not only that, they dealt with a bilingual show creatively and effectively. This show saved them from getting no stars.

Sub-survey: This Close as an example of captioning as it should be.

Here the captions are placed according to the speaker:

Michael and Kate in bookstore, talking, with ASL interpreter to the right, bookstore manager to the left. Captions: Can we get out of here? Okay, I need you to deal with the emergency.
Screenshot from season 1 of This Close on Sundance Now

Bookstore manager in background, Kate and Michael blurred in foreground. Captions: You can't tell me what to do. They're talking business, right?
Screenshot from season 1 of This Close on Sundance Now

 

White letters and, in fact, a different font are used for the ASL translation, and white on dark grey bands is for the oral dialogue. This is particularly helpful if you’re like me and following both sets of captions, listening to the hearing speaker and trying to follow the ASL or watching the fictional ASL interpreter sign and interpret! It was actually doable with this thoughtful crafting.

And when they do have to include an offscreen line, it’s correctly IDed and formatted to the side.

Kate in front of a bookstore's shelves, a finger pointing at her. Left caption: You can't tell me what to do. Right captions: [Morgan] Yeah.
Screenshot from season 1 of This Close on Sundance Now
 

And sidebar, I loved this scene where Michael’s hands are restrained as he is removed from a plane, and Kate says they can’t do that to a deaf person:

Kate yelling on an airplane: He's just trying to communicate, you fucking audist!
Screenshot from season 1 of This Close on Sundance Now

 

So, really clear, almost always perfect renditions of the audio (except for the odd Yes for Yeah). Although, I’m not sure what’s up with the three lines, especially when they are so short. Perhaps to ensure the titles are large enough to be visible for low-vision users, since Sundance Now doesn’t allow viewers different caption-display options?

This tells me one thing. VODs basically don’t care about Deaf/HoH access unless the executive producers (and guest stars like Nyle DiMarco in this case) of the show are Deaf and get the need! So, yay them, but that’s a fairly easy win, and boo Sundance overall.

Back to the show’s “Now The Discussion” segments. These featured cool young people sitting around a lovely studio drinking beer and chatting about the previous episode. Thankfully they had a Deaf/HoH participant in one I watched, and they also interviewed Shoshannah Stern and Josh Feldman after the last one. BUT but but: they applied post-production dubbing of an interpreter to overlay on the signers’ speech, rather than using the real-time interpretation they’d have had to facilitate the convo with the hearing people. So it struck me as a bit disingenuous: a bit like sim com, trying to create a false syncopation or an aversion to causing unwanted pauses in group dialogue while the signers’ words were finished being interpreted. Ew. As if waiting for a Deaf/HoH person to be orally interpreted was a problem or awkward. This struck me as ironically ableist/audist.

Also, these chats allowed expletives to be kept in the audio track but then captioned them with two hyphens, which is not only bad practice but also reinforces that the Deaf/HoH audience is not deserving of full and correct content. When characters swore/used vulgarisms in the show, the f-bombs etc. are shown fully spelled out. I thought this was another unfortunate dichotomy that spoke volumes.

As for the show itself, who doesn’t appreciate it when #DeafTalent is used, such as RJ Mitte, who has cerebral palsy? And the occasional use of loud static and other noises to obscure dialogue reinforces the challenges faced by some deaf people, in case we hearies get complacent in our viewing.

*So This Close gets ★★★★1/2  but Sundance Now itself gets only a half star.

Hulu (star rating not applicable, but see Reel Words home page image)

I’ve wanted to watch The Handmaid’s Tale since the get-go, as it is one of my favourite books. This time I did check the 17-page Terms which do allow for cancellation at any time, for reelz… I wanted to sign up for the $7.99 Limited Commercial plan ($11.99 for no commercials: isn’t that the flipping point of VOD—no commercials?)

I had trouble signing up due to not having a US zip code but a support chat told me “Our services are actually not available outside of the US unless you're a US military member living on base. Would you happen to fall in this category?”

Okay, over to:

CRAVE TV ★★★

Sub-survey: THE HANDMAID’S TALE

I went with the $7.99 monthly after 30-day free trial because I really, really wanted to see the show and didn’t mind paying for month two if I didn’t get through it all. (Who was I kidding? I binged it!)

Sidebar: as I’m sure you know, this was filmed in Toronto and around/near the GTA, so if you’re interested, here are links to the locations details.

https://torontoist.com/2017/06/where-the-handmaids-tale-was-filmed-in-toronto-part-one/

https://torontoist.com/2017/07/where-the-handmaids-tale-was-filmed-in-toronto-part-2/

If you go to my website home page, you’ll see a screenshot from Hulu from a trailer for the show. I discovered that they hadn’t done anything to address the issues of captioning since it was bought by Crave, and here are some of the more problematic captions as they appeared in Episode 1.

The Handmaid’s Tale has some parts of captioning done right, but then there are the usual inconsistencies that creep in. Sometimes [indistinct radio chatter] is heard and captioned as a very important part of setting and mood for the show, but at other times it is not captioned when clearly heard and significant. This does not provide full and complete access. And that’s supposed to be a standard in this country, but it is not enforced.

The show starts with a bang, with sirens blaring before the first visible frame, but [SIRENS BLARING] only shows for one second (nonstandard practice) and does not continue with the action, even though this is important to set up the mood and story.

This show is often based on the unshared thoughts of the heroine, Offred, and what she does dare to utter aloud. But these are not differentiated. (That’s Captioning 101, by the way.) In this scene, she was mocking another character internally but vocally answered Yes to her.

Offred leaving sumputuous grounds of house; captions: You wanna come along? Yes.
The Handmaid's Tale exterior

Below, a character who is losing it says I want my mom. and is comforted by Offred with a gentle Okay. We can barely see this dark scene, so a caption-dependent person might well be confused by the lack of speaker ID.

Dim shot of woman's head. Captions: I want my mom. Okay.
Screenshot from season 1 of The Handmaid's Tale on Crave TV

 

Netflix’s style guide determines that vulgarisms are to be spelled out (see my discussion of this around the 45-second mark on my Craption Eyerolls series on YouTube), and the same problem arises here. I’m not sure if the captioner wasn’t savvy or was applying their own personal values to not using the spelling cum. Either way, QC should have caught this. Crave certainly lets the f-bombs drop in captions, so I don’t think it was an issue of conservatism.

Offred lying on a pillow in moonlight in a dark room; caption: I can feel the Commander's come [sic] running out of me.
Screenshot from season 1 of The Handmaid's Tale on Crave TV
 

The Handmaid’s Tale takes place in a new-order society which has retro values. Church bells toll frequently, driven by plot and mood. But I kind of think using thrice is a bit over the top. Since it’s not the first time that church bells toll in the episode, space could have been saved by shortening the subject thus: Bells tolling three times. We would get that they are church bells from episode history, context and connotation.

Looking down from window to SUV in a driveway by a garden. Caption: (CHURCH BELLS TOLL THRICE)
Screenshot from season 1 of The Handmaid's Tale on Crave TV

 

Sometimes, captions are moved up to avoid covering a speaker’s mouth to make it clear that the person on screen is the speaker. But here Offred is not speaking; Aunt Lydia’s words should be italicized both as offscreen and over an amplification system. It also just looks weird.

Offred looking down, distressed. Caption: As you know, the penalty for rape is death.
Screenshot from season 1 of The Handmaid's Tale on Crave TV

 

Just as we need to see the speaker’s mouth moving, we need to watch Offred’s eyes carefully, as they reflect so much of what she may not say aloud. The cinematography in the series includes a lot of close-ups, particularly of Offred, to make us feel sympatico with her. Here, she is appalled by the speech, but we’re distracted by the titles actually touching her eyes proper rather than pondering her reaction. I talk about the need to facilitate audience immersion, rather than distract, frequently on my website. Breaking or preventing that immersion is one of the main ways to fail the caption or subtitle user, and it’s a key focus in my posts. This scene was made less evocative by careless captioning. And again, this show has been bought and the captions could have been improved by the new provider, but they either didn’t check or didn’t care.

Close-up. Offred looking up, pensively. Caption: But we cannot wish that ugliness away. Caption coversher eyebrows and upper parts of her eyes.
Screenshot from season 1 of The Handmaid's Tale on Crave TV

 

I really dislike the way VODs treat music, songs and lyrics. The presentations are not helpful, the rules don’t make sense, and they’re inconsistent. And they need to change.

Credit for director Reed Morano. Caption (YOU DON'T OWN ME BY LESLEY GORING PLAYING)
Screenshot from season 1 of The Handmaid's Tale on Crave TV

 

Here, for clarity, I would have styled the caption like this:

[♫ You Don’t Own Me ♫ by Lesley Gore]

It’s obvious that it’s playing—by the fact that the caption is there—and it’s obviously a title of a song, as indicated by the addition of customary musical notes. Had the house style not been to use ALL CAPS for sound effects, styling this like I have would have made the text clearer as a title, too.

And now, the pièce de résistance, the type of reason this show is on my home page:

Credit for Samira Wiley. Caption: alt code gibberish Don't say I can't go with other boys alt code gibberish.
Screenshot from season 1 of The Handmaid's Tale on Crave TV

 

Every musical caption has these—what I can only guess are Unicode cut-and-paste issues. This says We don’t care enough about accessibility to create acceptable captioning for our users.

And this is the crux of the matter. Communication is a right, and bad communication is a breach of that right.

Kate pointing angrily offscreen; caption: You can't hold a deaf person's hands like that.
Screenshot from season 1 of This Close on Sundance Now

 

As people in the cultural-content and entertainment industry, people who use captions and subtitles (basically everyone at some point or another) are our everything. The reason we have jobs. The reason we will always have work. And if there is one theme that is prevalent on my website, it is that the audience deserves better, and we should be ashamed of delivering less than excellent. Sure, human errors happen. We do our best. But when we deliver a product at subpar quality because it doesn’t matter to us personally or we are ignorant of issues of accessibility, we fail our fellow viewers.

I’m not the only one who thinks like this. I noted this Gamasutra post for its candour, and it reminds me to try my best to make my work accessible:

“I never have forgot the feeling of of depriving someone of an experience just because I didn't think to add a button”

Ian Holstead, Ubisoft

By providing craptions, VODs are preventing all but primarily D/deaf/hard of hearing viewers from accessing content—analogously blindfolding them.

I think we can do better than an average of 1.7★ out of ★★★★★ in caption and subtitle content on VODs. Industry standards are 95–98% accuracy, and in these five services I have found a 34% rate of success.

Please share this widely. And please leave a comment about your experiences in other countries.

Coming soon: Apps for ASL Learners; Creative Applications of Captions and Subtitles: Yay or Nay?