Captions, accuracy and corrections

A news photograph without an accurate caption is incomplete and can be dangerous. The caption tells the audience who and what they are seeing, where and when it happened, and why it matters. It also limits how the image can be misused. Many of the worst errors in photojournalism are caption errors: the wrong name, the wrong place, a guess presented as fact. This lesson covers how to write captions to professional standards, how to verify information including images from others, and how news organizations correct mistakes.

What a caption does

A caption, also called a cutline, has three jobs. It identifies: names, places, dates and the event. It explains: what is happening and why it matters, adding information the picture cannot show. And it frames: it signals how the image should be understood, for example by noting that a scene was a re-enactment for a portrait or that an image comes from a handout. Readers often read captions before the story, so a caption should stand alone, giving enough context to understand the picture without reading the article.

Wire-service captions follow a standard structure. The first sentence describes what is happening in the picture, in the present tense, and identifies people from left to right where needed, with the place and date. The second sentence gives background in the past tense. Then come the credit and, in the file's metadata, the copyright and source.

The five Ws and accuracy

Every news caption should answer who, what, where and when, and usually why or how. Who: full names with correct spelling, and titles or descriptions where needed. What: the action shown, described accurately and without exaggeration. Where: a specific place, including city and province or state. When: the date, and in breaking news sometimes the time. Why: the context that makes the picture news.

Accuracy rules are strict. Spell names as the person spells them; ask and check. Identify people only when you are sure; if you are not, say so or leave the identification out. Do not describe emotions or motives you cannot know: a person may look angry, but the caption should say what they did or said, or attribute the description. Do not repeat claims as fact; attribute them to their source. Check numbers against official sources and attribute them.

Style guides

Captions follow the publication's style guide. In Canada, most news organizations use The Canadian Press Stylebook and Caps and Spelling, which use Canadian spelling and conventions such as province abbreviations in datelines. In the United States, most follow the AP Stylebook, with American spelling and state abbreviations in AP style. Both cover titles, numbers, dates, ages and sensitive terminology, and both are updated regularly, for example with guidance on describing gender identity, disability, mental illness, suicide and race. Using the right terms respectfully is part of accuracy.

Gathering information in the field

Collect caption information while you shoot, not afterwards. Carry a notebook or use a voice memo. Ask people for their names, the spelling, their age if relevant, and where they are from. Photograph name tags, signs and documents as aids, but confirm details directly. Record the names and titles of officials. Note the exact location, using a map pin if necessary. At large events, collect information about the event from organizers or press releases, but verify claims.

When people decline to give their names, respect that, and describe them generically or not at all. For children, follow your publication's policy, which usually requires a parent's or guardian's permission for identification in non-news situations.

Scoring a caption

A simple checklist helps photographers and editors check captions quickly before transmission.

Verifying images and information from others

Photojournalists and photo editors increasingly handle images they did not make: user-generated content, handouts from governments, companies and police, and images circulating on social media. Verification asks four questions: Is the image what it claims to be? Is it from the place and time claimed? Who made it, and do they agree to its use? Has it been altered?

Techniques include contacting the creator directly and asking for the original file and other images from the same time; checking metadata when available; reverse image searches to find earlier appearances; comparing landmarks, signs, terrain and street layouts with maps and satellite images; checking weather records; and checking the sun's position and shadows against the claimed time. Organizations such as First Draft, which became part of other projects, and newsroom verification teams have published guides to these methods.

Handout images, from police, governments, militaries, companies and campaigns, should be labelled as such in the credit and caption, for example handout or provided by. They may be selected or staged to serve the provider's purpose. News organizations should state when they could not independently verify an image.

Corrections

Mistakes happen. What matters is how quickly and openly they are fixed. News organizations publish corrections policies and correct errors in captions just as in text: changing the online version with a note explaining the correction, publishing a correction in print, and, for wire services, sending a corrected caption or a kill notice to subscribers. A kill notice tells subscribers to remove an image entirely, used when a picture is found to be manipulated, staged, wrongly identified in a damaging way or legally problematic.

Photographers should report their own mistakes to the desk immediately. Hiding an error, or hoping no one notices, breaks trust far more than the error itself. Keep notes and originals so that errors can be traced and understood. Many publications have ombudspeople or standards editors who handle complaints and corrections.

Captions for different platforms

The same image may appear in print, on a website, in an app and on social media, and each needs a caption that works in its context. A social media post may show only a few words before a reader clicks, so the key facts should come first. Alt text, which describes an image for people who use screen readers, is different from a caption: it describes what the image shows visually, while the caption explains the news. Many newsrooms now write both. Never let a shortened platform caption drop the attribution or context that keeps the image from being misread.

Sensitive language in captions

Words matter as much in captions as in stories. Describe people by what is relevant to the story, not by race, religion, disability or other characteristics unless they are pertinent. Use the terms people use for themselves where possible, and follow your stylebook's current guidance, for example on how to refer to Indigenous peoples, to people who use drugs, or to deaths by suicide, where many guides advise avoiding the phrase committed suicide. Avoid words that judge, such as admitted or claimed, when a neutral word such as said is accurate.

A final check before transmission

Before sending, read each caption against the picture. Check that people are identified in the right order, that the place and date match the metadata, that names are spelled the same way in every caption of the set, and that every number and claim is attributed. A thirty-second check prevents most caption errors.

Captions in long-form stories

In picture stories and essays, captions can be longer and more narrative than wire captions, adding detail that deepens the reader's understanding of each image. They still need the same accuracy: names, places and dates checked, and claims attributed. Some publications place a short introductory text before the images and keep individual captions brief; others write fuller captions that together tell the story. Whatever the approach, each caption should make sense if the image is shared on its own, because that is how many people will encounter it.