How to Spot Xenophobia in Someone's Posts

Most of us have a relative who "isn't racist, but." The sentence that follows is where a lot of xenophobia lives. It hardly ever arrives as a slur. It shows up as a worry about the neighbourhood not feeling like home, a crack about a coworker's accent, a reply under a crime story that mentions where the suspect was born and nothing else about them. Any one of those has an escape hatch - I was only asking, it was only a joke. Read two hundred of them from the same account and the escape hatches quietly close.

That is exactly why it helps to read someone's public feed properly before you hand them your trust, a room in your house, or a seat in your group. Hostility online is common, not rare: about 56% of Americans say they have experienced online hate or harassment in their lifetime (ADL, 2024). Below is what the signals look like, where the false alarms hide, and how to tell a real policy argument from plain contempt for people.

What xenophobia actually sounds like

Rarely a burning cross, usually a running theme. The account that only ever mentions immigrants as a cost - never a nurse, a neighbour, a name. The person who answers every bad headline with a guess about the suspect's passport. Watch the small linguistic moves: a whole nationality talked about as one scheming unit ("they come here and..."), migrants and foreigners melted into a single faceless tide, the word "invasion" used for families who queued at a consulate. Nationality becomes a personality, and the personality is always a threat.

Then come the lines built for deniability. "I've got nothing against them, but they need to assimilate." "We're full." "Real locals can't get a look-in anymore." None of these is automatically hateful - people do have genuine worries about housing and wages - but stacked week after week, aimed only ever at the same outsiders, they stop reading as concern and start reading as a verdict on human beings.

How to spot xenophobia in someone's posts

To spot xenophobia in someone's posts, stop scoring individual lines and start reading for direction. A few tells reliably separate a bad-faith account from someone who is simply worried or opinionated:

• Blame by origin. Something goes wrong - a crime, a shortage, a queue - and the reflex is to pin it on where people are from. Guilt gets handed out by birthplace, not by anything the individual actually did.
• Dehumanising framing. People turn into floods, swarms, tides, or "cultures that don't belong." When humans get described the way you would describe pests or weather, that is the tell, whatever the topic sitting underneath it.
• Invasion and replacement talk. The claim, in any dressing, that foreigners are being let in on purpose to swamp or erase the "real" population. It is a named extremist theory, not a passing worry, and it usually travels with other conspiracy content.
• Contempt for the ordinary. Accents, names, languages spoken in the street, food smells from a flat down the hall - mocked not as a gripe but as proof the person is foreign, lesser, or a problem to be managed.
• One-way outrage. Endless fury when an outsider is a suspect; silence or excuses when an outsider is the victim. The asymmetry is doing the arguing.

Any one of these can slip out in a clumsy post from someone who would be genuinely stung to be called a bigot. It is the recurrence, and the fact that it always points the same way, that turns a signal into a pattern you can trust.

Arguing about immigration vs despising immigrants

This is where fair-minded people get twitchy, and they should. Wanting lower net migration is not hate. Immigration is a policy, and policies are meant to be argued over - the numbers, the rules, the trade-offs, a specific decision you think is wrong. Someone can want tighter borders and still treat the migrant in front of them as a full human being. Plenty of immigrants themselves hold hard-line views on the subject.

The line is not the topic; it is the target. An argument aims at policies, institutions and outcomes, and it leaves room for the individual to be a person rather than a category. Contempt aims at people - it treats every foreigner as a stand-in for the worst anecdote in the news, and it strips them of individuality before the conversation even begins. So when you read a run of posts, ask one plain question: is this person angry at a policy, or at a whole category of human? The honest ones can argue hard and still see people. The xenophobic ones only ever see a bloc.

A checklist before you trust someone

1. Read a real stretch of their public posts, not just the pinned one, and scroll back several weeks so a pattern has room to show.
2. Separate criticism of a policy from contempt for people, and note whether immigrants or foreigners get blamed as a group.
3. Watch for the recurring tells: go-back-home talk, they-take-our-jobs blame, mockery of accents or names, and invasion framing.
4. Check who they amplify, because the accounts they quote, boost and defend say as much as their own words.
5. Keep the receipts and screenshot the specific posts, so you are judging what they actually wrote, in context.
6. Weigh the pattern rather than one bad day, and decide what it means for the trust you were about to extend.

The honest limits of a read like this

Be straight about what this can and cannot tell you. It reads public posts only - locked accounts, private messages and deleted posts are out of reach - and it only works if the person actually posts; a near-empty timeline gives you almost nothing to work with. It is a read of what someone chose to make public, not a look inside their head, and it is not a background check or a criminal-records search.

Context cuts both ways, so mind your own false positives. A migrant describing the abuse thrown at them, a reporter quoting a slur to condemn it, a researcher naming an extremist theory to pull it apart - all of that can trip a keyword-level read, which is the whole reason you look at the actual post instead of a number. An AI flag is a nudge to go read the thing yourself, not a verdict. And a clean result means nothing hateful surfaced in what is public - not that a person is safe, and not that you should switch off your own judgement.