How to Spot Racism in Someone's Posts

Almost nobody worth worrying about posts a slur under their real name and photo. That is the mistake people make when they go looking - they expect the cartoon version, the profile that announces itself, and they miss the far more common kind: an ordinary-looking account where the contempt is folded into a joke, a question, a link they did not write but keep sharing. The tell is rarely one line. It is the direction everything leans once you read enough of it.

And that is the real skill here - telling a genuinely bad take, the kind anyone can have on a bad day, apart from a habit. One awkward post proves nothing. A month of posts that all bend the same way, aimed at the same people, is a different thing entirely. Spotting racism in someone's feed is mostly the patience to read past the first screenshot.

The blatant and the coded

Racism online comes in two registers. The blatant kind needs no explaining: slurs, open calls to treat a group as lesser, the stuff no one can wave away. It is rarer on accounts tied to a real name precisely because it is so easy to catch.

The coded kind is where most of the reading happens. It leans on dog whistles - phrases that sound neutral until you know the crowd using them - on statistics stripped of context and posted with a knowing caption, on "I'm just asking why nobody talks about..." framing that lets someone gesture at a slur without typing it. None of these is a smoking gun on its own. Stacked up, aimed reliably at the same group, they stop being ambiguous. Hate and harassment are common enough that this is worth doing: about 56% of Americans say they have experienced online hate or harassment in their lifetime (ADL, 2024).

How to spot racism in someone's posts

The method is the same whether the account is loud or careful. You are reading for direction and repetition, not a single quotable line.

• Read the replies and reposts, not just the posts. People are more careful in their own captions than in the accounts they boost and the strangers they pile onto. The amplified stuff often says more than the original.
• Track who gets singled out. Everyone criticises somebody. What matters is whether one group is the recurring target, and whether the criticism is about what they did or about what they are.
• Notice the exits. "Just a joke," "just asking," "just posting the numbers." When the same person reaches for the same escape hatch around the same subject, the hatch is the pattern.
• Read across platforms. The version someone keeps polished on LinkedIn can slip on X or in a Facebook group. Tone rarely stays consistent everywhere.

Joke, opinion, or pattern

This is the line that keeps the whole thing fair. A view you find wrong is not automatically racism, and treating every clumsy opinion as proof cheapens the word and makes you easy to dismiss. What tips a post over is not that it is provocative but that it targets a group for who they are - contempt, mockery, or the suggestion that some people count less because of their race.

The joke question is the same question. Humour that only ever lands on one group, defended with a wink every time someone objects, is not really doing comedy's job. Ask who the punchline is always on. Ask whether the person doubles down or reflects when challenged. And when you genuinely cannot tell whether something is bigotry or a bad joke, treat that uncertainty as a cue to read more of the feed, not to reach for a verdict you would struggle to back up out loud.

A reading checklist

1. Read the last few months of public posts, not the pinned one, so a pattern has room to show itself instead of a single bad day.
2. Separate a strong opinion from contempt aimed at a group for what its members are, which is the line that actually matters.
3. Watch for the coded version - dog whistles, mock-innocent questions, and jokes whose target is always the same group.
4. Look at what they repost and who they boost approvingly, since amplification says as much as an original post.
5. Check each covered platform - X, TikTok, Instagram, Facebook and LinkedIn - because tone kept clean in one place often slips in another.
6. Open the actual post behind any flag before you judge it, since reclaimed language and quotes condemning racism can read as the opposite.

The honest limits of a read

Say plainly what a read like this cannot do. It sees public posts only - a locked account or a barely-used one leaves little to go on, and silence is not a character reference. It is AI flagging content with the receipts attached, and context is exactly where it can stumble: reclaimed language, a screenshot someone posted to condemn, an academic quote, all can trip a flag when no racism was meant. That is the whole reason you get the actual post to weigh, not a verdict. You judge; the tool just surfaces.

And a clean result carries one meaning only: nothing in the public posts stood out. It is not a clearance, not proof that a person holds no such views in private, and not a substitute for how they actually treat people you both know. This is personal due diligence on what someone has already made public - a read of an adult's public posts, not a background check or consumer report, and no part of any hiring, tenancy or credit decision; for those, use a licensed provider. Kept to what it is, though, it does the useful thing a friendly bio never will: it shows you how a person talks when they think only their own crowd is listening.