Writing for Accessibility: Don’t Use Alt Text for Memes, Please

by Kelli Hughes

If you’re like me and 59% of Gen Z, watching TV must include subtitles. Sometimes the audio mixing for a show is especially poor, muddling the dialogue and making it difficult to parse. Sometimes I can’t tell who is talking because of an especially large cast to keep track of. And sometimes, after a long day, I just need the clarity subtitles offer.

Photo by Monica Flores on Unsplash

As a hearing person, while I appreciate the affordances of subtitles, I do not need them to experience media. But that isn’t the case for everyone. Nearly 15% of adults in America report difficulty hearing, and 8% of Americans report some degree of vision loss. Where one person listens to the TikTok their buddy sent them for a laugh, another person requires closed captioning so they can read the joke on their screen. Without the proper use of accessibility tools, several users are excluded from the same experience.

Fortunately, most social media platforms, such as X (formerly known as Twitter) and TikTok, have instituted community accessibility tools in recent years. These tools, including captioning and image description, are meant to give users the ability to make their own content more accessible. Users can caption their videos for d/Deaf and hard of hearing viewers and write alternative text (alt text) to describe images for viewers with low vision.

What a wonderful move, one might think. An increase in accessibility tools is one of many technological triumphs this past year. Unfortunately, simply having those features available does not mean they will be used correctly. Without proper education and advocacy for the disabled community, these vitally inclusive tools will not serve their intended purpose. Worse yet, the misuse of them can actually impair disabled users. Two particular grievances plaguing online accessibility are misuse of alt text and censorship in captioning, which turns this accessibility triumph into a conundrum and prevents the internet from truly being a space for everyone.

Alt text: why the alternative?

Photo by dole777 on Unsplash

Alt text is text read by a screen reader which describes an image, including any important visual elements and textual information provided within it. Alt text should be concise, descriptive, and relate to the context the image is placed in. The Washington Post design system is one of many design systems which asserts alt text should be “objective.”

An increase in accessibility tools is one of many technological triumphs this past year. Unfortunately, simply having those features does not mean they will be used correctly. ​

Striving for objectivity seems like a logical and responsible move. And for the most part, it is! Screen readers should not be inundated with opinions and misinformation. But that doesn’t mean alt text has to be straight-to-business and boring. Accessibility engineer and Director of tech accessibility company TetraLogical Léonie Watson uses screen readers. She argues alt text should include emotionally impactful language when describing emotion rich images. Users with low vision should not be deprived of the amusement, shock, or awe that visuals evoke. Alt text can do more than describe the bare bones; it can paint a picture through words. 

Those words must connect to its respective image, however. The “alternative” in alternative text is vital: the text is meant to be an alternative way for users with low vision to receive information. Alt text is not a space for unrelated text like secondary captions, subtitles, or jokes and memes, but this is often what’s happening. It’s bad enough when the average Twitter user is abusing this feature, but when verified accounts such as British Prime Minister Rishi Sumak’s are blatantly misusing alt text to promote their agendas, it’s obvious greater media accessibility education is a priority. And when accessibility bots—whose purpose is to catch and notify accounts that their content is inaccessible—are run off the platform, it becomes the duty of every user to ensure that their posts are inclusive.

Censorship in captions: cut the [bleep]

The terms “subtitles” and “captions” are often used interchangeably to refer to the text that appears on screen for video content. However, there is an important distinction to make. Subtitles simply transcribe and translate dialogue, allowing viewers to read what they hear.

Photo by Immo Wegmann on Unsplash

Captions, meanwhile, are specifically for d/Deaf and hard of hearing viewers because in addition to transcribing dialogue, they also include cues and descriptions which identify speakers and other audio elements, such as music and ambient sound; think “wings flapping” or “indistinct conversations.”

Captions come in two forms: open and closed. Open captions cannot be toggled off and are a permanent fixture of video content. Closed captions—the form most users are likely familiar with, visible with a little “CC” icon—can be toggled on and off, allowing users to watch with or without them. Social media platforms like X/Twitter, TikTok, and Instagram utilize closed captions.

Seems simple enough, right? Well, not exactly. ​

Omitting information in captions is a form of ableism: disabled users are prevented from understanding and engaging with media when they are not given the same information as their able-bodied peers.

There are a litany of captioning issues that trouble viewers. We’ve likely all come across supremely incorrect auto-generated captions on live TV, YouTube, or social media platforms. Maybe you have a hearty laugh when an actor says “brushing” but the captions read “bloodshed” instead. Or maybe you’re aggravated when a character speaks a language unfamiliar to you and all the subtitles offer is “[speaks in Spanish].” As funny or as frustrating as this may be, inaccurate captions become a serious problem for both d/Deaf and hard of hearing viewers and English Language Learners, who now have to decipher the original meaning.

An even bigger problem arises with the dreaded “c” word: censorship.

Context is equally important to captions as it is to alt text. Omitting information in captions is a form of ableism: disabled users are prevented from understanding and engaging with media when they are not given the same information as their able-bodied peers. Exacerbating the issue is when captioners intentionally censor explicit language: while a hearing person hears a spoken profanity without any bleeping, the captions have replaced the word with a sanitized alternative—like “crap” or “darn”—or have redacted the word entirely, leaving a gap in dialogue.

Many platforms have been criticized for this censorship, such as Netflix in 2018 for its censored captions in its hit show Queer Eye. Despite acknowledging this issue, several other shows and programs on the platform still feature censored captions. TikTok user @demtxt82 is deaf and advocates for the equal treatment of the community: “They shouldn’t be deciding for us what we should read. We’re not children.” @demtxt82 is speaking to the larger issue of infantilization faced by the disabled community. Censorship assumes d/Deaf and hard of hearing viewers cannot handle mature content. They are not only denied the same experience as able-bodied viewers, they are coddled and patronized.

What can we do about it?

Unless you are a professional captioner or an accessibility engineer, you may be thinking “What am I supposed to do about it?” But that’s the wonderful thing about advancements in accessibility tools like captioning and image description on social media: anyone can use them. Take the extra thirty seconds to caption the video you post on Instagram. Use the alt text feature to describe images within your Tweet. Likewise, anyone can misuse these features. When you notice another user blatantly misusing alt text, call attention to it. ​

Subtitles may not be make or break for you. But social media is for everyone, and for everyone to be able to enjoy it, accessibility must be a priority.

---

I refer to the deaf/hard of hearing community as “d/Deaf” and/or “hard of hearing” per the explanation provided by the National Association of the Deaf. I refer to the blind/low vision community as “blind” and/or “low vision” per the explanation provided by the National Center on Disability and Journalism “Disability Language Style Guide.”