Should you include race in alt text?
Best practices for describing people when writing alt text.
In April 2024, I co-presented “Alt text for fun and profit” with Jenn Czeck at Minnebar. During the Q&A section, Jenn and I were asked just how much detail needs to go into alt text when describing stock images.
This question prompted an excellent discussion around the intention of alt text for images. The purpose of alt text is to equalize user experience across all members of your audience. If the image conveys something visually, it must also convey it non-visually through alt text or image description metadata. Placeholder images generally don’t provide meaning to a user, so they don’t require detailed descriptions.
But what if the placeholder image includes a person?
I’m a firm believer that if the focus of an image is a person, the identities of that person matter.
Why do identities matter?
In Krystal Jagoo’s article “The Importance of Representation in Books”, she discusses the negative impacts of omitting identity markers. (She defines identity markers as things like gender; race; disability; sexuality/queerness; language; family structure; and “topics of social significance” like “homelessness, incarceration, immigrant/refugee status”.)
While her focus is primarily on classroom literature, the lasting negative impacts from lack of representation extend past schools and into kids’ daily lives. One of the studies she cites showed BIPOC toddlers developed low senses of self-worth when faced with board books without diverse characters.
A lack of visually diverse images can harm companies, too. Dr Rebecca Swift wrote an article for The Undercover Recruiter about research done by the creative research team at Getty Images regarding cultural diversity in professional visual representations. In their study, they found potential job applicants valued diversity: “if they can’t see themselves represented by your brand, they are less likely to look at you as a potential employer.”
On the opposite end, John D Saunders from blackillustrations wrote an article describing the positive impacts of representation: lessening alienation, building community, fostering understanding and acceptance, expanding our ideas of identities and social roles, and reducing stigma and bias.
Shutterstock has an entire blog series called “How We Show It”, where they describe their efforts in creating more inclusive visual representations for different identities. Each article goes into detail about how each identity connects to imagery and provides great context around building inclusive content. (The “How We Show It: Black Elders” article is pretty rad and you should read it.)
Why do identities matter in alt text?
Even if a user can’t see the image, the user has an identity. If the user is marginalized, knowing what kind of person is represented in the image still matters.
Let’s look some examples.
Example #1
Shutterstock’s photo description is “image of a black health professional wearing face mask under chin and stethoscope around neck, hands raised up for support and strength in covid-19 pandemic”.
If I were to include this photo as filler on a web page, I’d add alt text: “Photo of a Black health professional in scrubs, smiling at camera, with a tablet and stethoscope”
(Note: it’s “Black” when describing people, and “black” when describing a color.)
The photo description doesn’t add any gender or age information, so I don’t either. I don’t personally know the model, so I can’t say anything more specific.
For Black users, it’s important to know that the subject of the photo is Black. Without that descriptor, it’s easy to erase the identity of the person photographed, and to alienate the user.
Example #2:
Shutterstock’s photo description is “Middle-aged Hispanic woman with disabilities on a wheelchair against a colorful background”.
If I were to include this photo as filler on a web page, I’d add alt text: “Photo of a middle-aged Hispanic woman in a wheelchair, smiling at camera”
Again, I don’t know this model, and I can’t say anything more specific than “in a wheelchair”, “middle-aged”, “Hispanic”, and “woman”. It isn’t immediately clear this model has disabilities, but I know she uses a wheelchair.
The alt text for this photo could read “Photo of a middle-aged Hispanic woman, smiling at the camera”. It could also read “Photo of a middle-aged woman in a wheelchair, smiling at the camera”. While both of these descriptions are true, they aren’t complete.
In the first, disabled people don’t know the subject of the photo is disabled. In the second, Hispanic women don’t know the subject of the photo is a Hispanic woman. The goal of alt text is to fully represent what a user needs to know, and to achieve that we have to consider all the parts of the subject’s identity.
Why is alt text not enough?
I want to be clear that I’m advocating for visible representation as only a part of larger efforts to make tech more inclusive and accountable. Representation without action is shallow at best, and harmful at worst.
Kevin Leo Yabut Nadal, Ph.D. wrote an article for Psychology Today describing his experiences with Asian American and queer representation growing up and how we can improve in the future. Sutheshna Mani wrote a great Medium article about representation in media and where it frequently falls short. I’ve personally been hurt by autism representation in media. I could pay off my student loans if I got a dollar every time someone tells me I don’t “act autistic” because I don’t match what they’ve seen in movies or TV.
TikTok and other social media platforms provide an incredible opportunity for representation to improve, but that representation isn’t reflected in accessibility measures yet. Lots of incredible organizations exist that fight to improve diversity in imagery, but accessibility lacks pretty far behind. Designers, writers, UX practitioners, marketers, and content creators need to consider how we refer to people so that our audience isn’t left out. Adding identity markers to our alt text is a perfect first step.
About the Author
Jess Schalz (she/they) is a software engineer-turned-technical writer. She has a terrible cat named Sudo, and the two of them live in Minneapolis, MN.
