How to write an image description

I wrote this how-to guide with the immensely helpful counsel and insights from Bex Leon and Robin Fanning, as well as through an online survey of Blind / low vision / visually impaired people.

What is an image description?

An image description is a written caption that describes the essential information in an image.

Image descriptions can define photos, graphics, gifs, and video — basically anything containing visual information. Providing descriptions for imagery and video are required as part of WCAG 2.1 (for digital ADA compliance).

It’s sometimes referred to as alt text since the alt attribute is a common place to store them. The alt attribute appears in the HTML code like this:

<img src=”link” alt=”text goes here”>

While alt text and image descriptions are sometimes used synonymously, they’re not actually the same thing. Alt text refers to the text specifically added to the alt attribute, and is often short and brief. Image descriptions can be found in the alt text, caption, or body of the webpage and are often more detailed. For more about alt text and image descriptions, check out @higher_priestess on instagram.

Additionally, image descriptions are a gesture of care and an essential part of accessibility. Without them, content would be completely unavailable to Blind/low vision folks. By writing image descriptions, we show support of cross-disability solidarity and cross-movement solidarity.

How to write a good image description

Object-action-context

Something that I learned from talking to Bex is that there is a storytelling aspect to writing descriptions. It doesn’t necessarily make sense to go from left to right describe everything in an image because that might lose the central message or create a disorienting feeling. For that reason, I came up with a framework that I recommend called object-action-context.

The object is the main focus. The action describes what’s happening, usually what the object is doing. The context describes the surrounding environment.

I recommend this format because it keeps the description objective, concise, and descriptive.

It should be objective so that people using the description can form their own opinions about what the image means. It should be concise so that it doesn’t take too long for people to absorb all the content, especially if there are multiple images. And it should be descriptive enough that it describes all the essential aspects of the image.

What counts as “essential” can sometimes be subjective. In my research, some people preferred very brief descriptions while others preferred a great deal of visual detail (like Robin and Bex). My best advice in this area is to be as descriptive as your audience needs. You know your audience and your content best. If your work is highly visual and your audience is interested, then add more detail. If you and your audience care more about the non-visual content, then keep it on the shorter side.

Example:

Description: Black Lives Matter sign being held in a crowd.

Object: Black Lives Matter sign Action: being held Context: in a crowd

We can add more detail to this description to paint a more vivid picture. In most cases, I recommend adding the detail with the object-action-context format. This makes the description more concise.

A concise way to add detail: A painted Black Lives Matter cardboard sign being held anonymously in a blurred out crowd in front of a stone building.

A redundant way to add detail: Black Lives Matter sign being held in a crowd. Behind them, there is a building made of stone. The crowd is blurred out. The sign is painted on cardboard. The person holding the sign is anonymous.

You can see that the redundant example repeats a lot of words and skips from one thing to another, making it longer and more disorienting. The object-action-context format helps fixate on the primary focus, and then slowly expands to get the full picture.

However, sometimes there is so much detail that adding it will cause one big run-on sentence. In this case, it’s better to keep the first sentence short and in an object-action-context format. This way, people can get a basic understanding and skip through the rest of the details if they want.

Example:

Description: A tattooed person holding a sign that says, “Teach your children well,” in a crowd of people. In the middle of the sign, there’s an illustration of the earth with 2 raised fists on either side that have a rainbow pride square background, a trans pride circle background, and brown skin tone stripes on the fists.

Object: person Action: holding a sign Context: crowd

To be even more descriptive, you can go into the meaning of the visual imagery. This is where image descriptions are useful to sighted people too because they provide an additional explanation for things that might not be obvious otherwise.

Example of adding meaning: The raised fist is a symbol of solidarity and specifically Black power as popularized by the Black Panther Party in the 1960s. The rainbow pride flag has rainbow stripes and symbolizes general LGBTQ pride. The trans pride flag has pink, blue, and white stripes and celebrates pride for the trans and gender non-conforming umbrella.

This is obviously a longer commitment and might not be necessary if you have written other content in a surrounding article, or if you’d rather reserve your energy from explaining and encourage people to look it up themselves. And obviously, if you choose to explain symbolism, then make sure to do the proper research.

How context impacts the description

(April 2021 update)

Digital images don’t just exist in a vacuum, they have surrounding context on whatever app or website they are found on. We often come across images when we read news articles, scroll through our friends’ social media, go online shopping, etc.

Example:

Description: dog sitting in a field of autumn leaves

For an image on a news article, that description works reasonably well. It’s objective, concise, and descriptive while also quite brief. However, if the photo is on a dog breeder website that has images of several different breeds of dogs, it doesn’t actually communicate much useful information (especially if every description just says “dog”). For that reason, I’d include the breed and more visual descriptions of the dog.

Dog breeder website description: light brown pitbull dog breed with a black nose and short, pointed ears

If I am an artist who took the photo of the dog and want to showcase my skills and evoke some emotion, I’d probably want to add more detailed descriptions. This is of course subjective based on the type of artist I am and how I want the audience to receive my work.

Artist’s website description: a photo of a light brown dog with short ears and a pink tongue sticking out, sitting in a field of autumn leaves. The dog is in focus while the sunlight reflects off the light orange and yellow leaves, fading into the background.

If the image is on a site or app with several hundred images, perhaps for an animal science textbook or a flashcards app, then it’s better to optimize for quick and easy scanning of content. Perhaps “dog in field” or even just “dog” will suffice in that context.

This section is new and very much in progress. If you have additional thoughts about how context impacts image descriptions, feel free to reach out! (Contact info at the bottom.)

Describing race and gender

In general, I recommend describing race and gender if (1) it’s relevant to the image, (2) if you know what their identity is, and (3) if the way you’re describing it is consistent with other descriptions.

It’s relevant if it’s an essential part of the message — such as a joke specifically about identity or celebrating the radical visibility of marginalized people. It’s known if the people in the photo have confirmed their identity. It’s consistent if you’re describing everyone’s race and not just people of color (rather assuming white as the default).

For example, the stock photo project Disabled and Here has created a library specifically of disabled BIPOC (Black and Indigenous People of Color). They have also kindly provided image descriptions! This is an example of describing identity that is relevant, known, and consistent.

Description: Three Black and disabled folx (a non-binary person holding a cane, a woman sitting in a power wheelchair, and a woman sitting in a chair) partially smiling at the camera while a rainbow pride flag drapes on the wall behind them.

Other great diverse stock photo libraries include The Gender Spectrum, Nappy, and CreateHer Stock.

In many situations, mentioning race or gender might not be relevant at all. And while it might be tempting to label folks as “women” or “men” there is the risk of misgendering them. It’s actually better to say “person” or “people” and talk about physical characteristics more literally instead of assigning gender to every presentation (e.g. say “person with long hair” instead of “woman” or “femme-presenting person”).

Example:

Description: Four people sitting on a sidewalk bench, most wearing sunglasses and heels and one with a big brown dog, in a city on a sunny day.

Keep in mind that this article is written by a nonbinary person! Because I’ve experienced so much misgendering, I’m inclined to avoid gendering people in most of my descriptions. However, I recognize this isn’t everyone’s experience and there are many situations where gendering people is important for visibility and representation. For that reason, I encourage you to write in a way that feels most genuine to you while also being respectful of others.

Describing complicated diagrams

In the event that you’re describing charts, infographics, or diagrams with more complexity, I recommend dividing it into sections and focusing on one section at a time in a logical order.

Example:

Description: Graphic titled “Protesting safely” at the top in lavender text on dark purple background. It includes 3 sections: “What to wear,” “What to bring,” and “Don’t bring” all with graphic illustrations. In the light purple “What to wear” section, it says, “nondescript, solid color, layered clothing; cover identifying tattoos, goggles & mask, emergency contacts written down, heat resistant gloves, and tie your hair up.” In the medium purple “What to bring” section, it says, “water for drinking & tear gas, snacks, cash/change & ID, washcloth, bandages & first aid supplies, ear plugs, and protest signs. In the light red “Don’t bring” section, it says, “cell phone without first turning off Face/Touch ID, going on airplane mode, and disabling data, jewelry, anything you don’t want to be arrested with, and contact lenses.”

I choose not to go into detail describing visuals if the text essentially already describes them and if the description is already pretty lengthy. In this case, the goal of the image is definitely to convey information, therefore I felt describing the graphics would be distracting.

A few last tips

Because images are just a collection of pixels and vectors, text is not available to screen readers (sometimes screen readers can pick up text, but this doesn’t appear to be consistent). Make sure to transcribe all the text in the image if you’re providing a description.

There are a few screen reader limitations to keep in mind. Be sure to write hashtags in title case #ForExampleLikeThis so that screen readers can decipher individual words. Also, avoid all caps because screen readers sometimes interpret those one letter at a time.

Most social media platforms these days offer an alt text option. From what I have heard from the community, it’s actually better to put the description in the caption instead. This way, it’s also visible to screen magnifier users and sighted people who might need or benefit from the description. (It’s actually unclear to me whether we should also add a full description or abbreviated version in the alt tag if we are adding to the caption. If you have thoughts or ideas about this, please reach out! Contact info at the bottom).

If you’re describing something that requires some subject matter expertise, it’s helpful to add a short explanation for additional context. For example, me and my opponent training BJJ (Brazilian Jiu-Jitsu, a grappling martial art).

Of course, people have limited time and energy, and image descriptions require time and energy. In my opinion, it’s better to have something than nothing. Better write a description, even if it’s super brief, so that the image is accessible rather than nothing at all.

There are community access groups out there where you can crowd-source image descriptions, e.g. People’s Image Descriptions run by Bex. Depending on your own position of power and access to resources, you can also ask a volunteer or pay a writer for their time. Writing image descriptions does not have to be a solo activity and can be a community effort.

I welcome feedback!

I wrote this guide because I had such a hard time finding a coherent image description how-to when I was starting out. I hope this is helpful! I welcome any feedback you have on making this information understandable, accessible, and respectful.

Say hi to me on twitter and instagram as @access_guide_ :)

If you like my work, you can support me on ko-fi.