In the Spring of 2009, a reporter for the Associated Press published a news feature about rape in the Democratic Republic of Congo. Much of the piece focused on the taboo that silences rape survivors there. But it began like this: “Zamuda Sikujuwa shuffles to a bench in the sunshine, pushes apart her thighs with a grimace of pain and pumps her fist up and down in a lewd-looking gesture to show how the militiamen shoved an automatic rifle inside her.”

If you ignore the content for a moment, this is a textbook-perfect feature lead. It’s full of “color.” It gives us a “character” to follow for nearly 200 words. It’s loaded with “action verbs”—shuffles, pushes, pumps. It creates a miniature “scene,” set outdoors, somewhere, on a well-lit bench.

But if you bring the content back, the first sentence horrifies—and not just because what happened to Zamuda Sikujuwa is horrifying. It horrifies because we, as readers, have no idea what’s going on: In a country where rape and sex are taboo, why did a woman, sitting outside, “push apart her thighs” and mimic the violation she experienced? Did the reporter ask her to? Are there other women around? Men? And why, after putting Sikujuwa the woman through all of this, does Sikujuwa the character disappear after fewer than 200 words?

As a general piece of journalism, this lead does everything right. But the unanswered questions raised by these narrative choices make me doubt the writer. Her article fails to adequately address the most important question facing a journalist covering rape and other violent traumas: How do we make readers ethically comfortable with our storytelling choices and morally uncomfortable with what the story depicts?

Trauma stories require the writer to consider the reader, listener, or viewer as a partner in the creation of ethical journalism. Our choices as craftsmen—about identity and attribution, about detail, about writer’s voice, about structure and style, and even about medium—do more than simply tell the story. They tell readers about our values.

Most journalism seeks to convey information objectively, but trauma stories have an agenda: they call to the reader to witness, to agree with the writer that “This should not have been.” If there is no agreement between reader and writer, or if the writer fails, the story is an exercise in voyeurism. In rape stories, we are publicly exposing the personal suffering of survivors. If we do this with any other intention than that rape should not happen—or if we do this without any clear intention at all—we are indulging in a kind of storytelling that critics do not hesitate to call pornography.

What separates good trauma journalism from voyeurism are two elements of the job: the way we report and the way we write. The best practices for reporting trauma have been getting more attention. There are tools, like Roger Simpson and William Cote’s textbook, Covering Violence: A Guide to Ethical Reporting about Victims & Trauma, and guidelines from the Columbia University Graduate School of Journalism’s Dart Center for Journalism & Trauma, that help us think about whether and how to identify our subjects; what effects post-traumatic stress disorder may have on the memories and storytelling capacities of our subjects; how to interview with sensitivity to avoid re-traumatizing survivors. All of these choices shape the ethics of our reporting.

Jina Moore was a 2013 New Media Fellow of the International Reporting Project