When a submission names someone — a decision guide for admins
How to tell whether a person is identifiable, when it matters, and whether to publish, edit or decline. With worked examples from the kinds of submissions confession pages actually receive.
Also in: Bahasa Melayu · 简体中文 · 繁體中文
Short answer. If your readers could work out who a post is about, and it says something that person would not want said, do not publish it. Decline damaging statements of fact and private information about identifiable people. Crush posts and opinions about someone's work go out only with the identifying details removed. Identifiable does not mean named: a course, a year and a society role together can point to one person.
Of all the decisions an admin makes, this is the one with the most at stake. A crude joke annoys people. An off-topic post bores them. A post that identifies a real person and says something damaging about them can end a friendship, cost someone a job, bring a lawyer's letter — and it is the admin who pressed Publish.
It is also the hardest kind of decision to hand to software, because whether someone is identifiable depends on what your particular audience already knows. This guide gives a way of thinking it through.
Identifiable does not mean named
The test is not "does the post contain a name?" The test is whether people who know the person would recognise them from the post. That can happen with no name at all.
Direct identifiers are the obvious ones: a full name, a username or handle, a photograph, a phone number, a student or staff number, a car plate.
Indirect identifiers are details that narrow the field: a course and year, a hostel block and floor, a job title, a role in a society, a distinctive physical description, a timetable slot, a relationship to someone else who is named.
A single indirect detail rarely identifies anyone. Several together often do. "A second-year in the engineering faculty" describes hundreds of people. "The second-year engineering student who is treasurer of the photography club" describes one. This assembling of harmless-looking details is called jigsaw identification, and it is how most people on confession pages are in fact identified.
Initials, nicknames and "you know who" are not protection. If the comments are going to fill with "this is obviously about J", the post identified J.
Four questions, in order
When a submission refers to a person, work through these in sequence.
1. Could your audience work out who it is?
Read it as a member of the community would. If the answer is clearly no — "to the stranger who returned my wallet at the bus stop" — you can stop here and treat it as any other post.
If the answer is yes or maybe, continue.
2. Is the person a public figure, and is the post about their public role?
Politicians, celebrities and senior office-holders are legitimately discussed in public, and criticism of what they do in that capacity is ordinary comment. That does not extend to their private lives, their families, or invented claims about them.
Within a campus, be careful about who you treat as "public". A vice-chancellor's decision on fees is a public matter. A lecturer's teaching is somewhere in between — students have a real interest in discussing it, but the lecturer is not a public figure in any wider sense. A classmate who happens to be well known is not a public figure at all.
3. What does the post say about them?
Sort it into one of these:
- A statement of fact that would damage them if believed — they cheated, stole, harassed someone, slept with someone, are incompetent, are ill. This is the dangerous category. You cannot verify it, the person cannot respond, and if it is false or cannot be proved, it may be defamatory.
- An opinion or experience, honestly held — "I found his lectures impossible to follow." Much lower risk, particularly when it is about a professional role rather than a person's character.
- Attention the person may not want — a crush post, a compliment on someone's appearance, a "who is the girl in the red jacket?" These feel harmless to the writer and to most readers. To the person singled out, they can be unsettling, and they frequently end with someone posting the name in the comments.
- Private information — where they live, who they are seeing, their health, their sexuality, their family circumstances, their results. Publishing this is harmful whether or not it is true.
4. What is lost if you do not publish it?
Usually very little. The page will have another post within the hour. Compare that with what is risked if you do publish. In almost every identifiable-person case, the two are not close.
A decision table
| The post… | Private individual | Public figure, public role |
|---|---|---|
| States a damaging fact | Decline | Decline unless it is already reliably and publicly reported |
| Gives an opinion on their work or role | Edit to remove identifiers, or decline | Usually fine |
| Gives an opinion on their character or private life | Decline | Decline |
| Is a crush post or compliment | Publish only with identifying details removed | Usually fine |
| Reveals private information | Decline | Decline |
| Is a thank-you | Usually fine; consider removing the surname | Fine |
"Decline" is not a judgement that the submitter is lying. It is a recognition that an anonymous page is the wrong place for a claim nobody can test.
Editing instead of declining
Sometimes a submission is worth publishing once the identifying details are gone. If you edit:
- Remove enough. Taking out the name but leaving the course, year and society has changed nothing. Read the edited version as someone who knows the person would.
- Do not alter the meaning. You may remove details; you should not rewrite someone's views.
- Mark it. A short "[edited by admin to remove identifying details]" keeps you honest with readers and with the submitter.
- Decline if it no longer makes sense. If the post was only interesting because of who it was about, it was not a confession. It was gossip.
Worked examples
"Dr [name] failed half our class because he can't teach." Names a private individual and makes a factual claim about results and competence. Decline as written. A version without the name — "half our class failed a core module and nobody feels the teaching prepared us" — raises the same concern and points nobody toward one person, provided the module is not identifiable either.
"To the girl in the blue hijab who sits by the window in the library every morning — you have the kindest smile." Kindly meant, and it identifies a specific person to everyone who uses that library. She did not ask for it, and she will be recognised. Publish without the description ("to someone I see in the library every morning") or not at all.
"The treasurer of [society] has been taking money from the club account." An accusation of a crime against a person who is identifiable by role. Decline. If it is true, the society's committee, the student affairs office and the police can act on it. A confession page can only spread it.
"My ex cheated on me with my best friend and now they're together." No names, but the people involved — and anyone who knows them — will recognise the situation at once. This is the commonest hard case. The more specific the circumstances, the more identifiable it is. Published with the specifics removed, it becomes a post about being betrayed, which is something many readers will recognise in their own lives.
"The café in block C gave me food poisoning." A business rather than a person, and a statement of personal experience. Businesses can be defamed too, but honest accounts of a customer's experience are ordinarily treated as legitimate comment. The risk is lower. It rises if the post claims as fact something the writer cannot know ("they use expired meat").
"[Full name] is gay." Private information about an identifiable person. Decline, whatever the submitter's intention and whether or not it is true. Publishing it could put that person in real danger.
The comments are part of the post
A carefully anonymised post can be undone in ten minutes by a comment saying "this is about so-and-so from 2B". If you publish something that refers to a person, even obliquely, watch its comments, remove guesses and tags promptly, and be ready to close comments altogether. See Managing comments on confession posts.
When you get it wrong
You will, eventually. When someone tells you a post is about them:
- Take it down first. Do not wait to be sure they are right.
- Do not ask them to prove it publicly or to explain why they object.
- Thank them, confirm that it has been removed, and apologise.
- Note in your decision log what was missed.
The full process is in How to handle a takedown request. The legal background is in Defamation and confession pages.
The short version
If your readers could work out who it is, and the post says something that person would not want said, do not publish it. Almost every serious problem a confession page has ever had began with an admin thinking, "It's probably fine."
Common questions
Should a confession page publish a post that names someone?
Not if the post states a damaging fact about them or reveals private information: decline it. An honest opinion about a person's work or role can be published once the identifying details are removed, and a thank-you is usually fine. If readers could work out who it is and the post says something that person would not want said, do not publish it.
What makes a person identifiable in a post?
The test is whether people who know the person would recognise them. A name, handle, photograph, phone number or car plate does it directly. So do several indirect details together — a course and year, a hostel block, a role in a society — even though each one on its own fits hundreds of people. This is called jigsaw identification.
Are initials or a nickname enough to protect someone?
No. Initials, nicknames and "you know who" are not protection. If the comments are going to fill with "this is obviously about J", the post identified J.
Can a confession page post about a lecturer?
With care. Students have a real interest in discussing teaching, and an honestly held opinion about lectures carries much lower risk than a claim of fact. A post that names a lecturer and states as fact that they are incompetent or failed a class should be declined as written; a version without the name, where the module is not identifiable either, raises the same concern without pointing at one person.
Is it all right to publish a crush post that describes someone?
Only with the identifying details removed. A crush post feels harmless to the writer, but the person singled out did not ask for it and will be recognised, and such posts frequently end with someone putting the name in the comments.