A news story breaks: an attack, a name, a motive still unclear. Somewhere in the comments, someone asks whether the attacker’s religion had anything to do with it.
The replies arrive within minutes. Not all Muslims are terrorists. Not all Christians shoot abortion doctors. Change the topic to street harassment and the same reply turns up in a different costume: not all men.
Every one of those sentences is true. That is what makes them so effective, and so often beside the point.
“Not all” is the correct answer to a claim about all. The trouble is how often it gets fired at a claim about some, where it answers nothing and closes the conversation anyway.
All and some
“All Muslims are terrorists” is a universal claim. Universal claims are fragile: a single counterexample sinks them, and this one sinks under more than a billion of them. “Not all” is exactly the right reply, and it wins.
“Religious belief can move some people toward violence” is a different kind of claim. It is about some, and counterexamples don’t touch it. You could line up every peaceful believer on the planet and the claim would still be standing, because it never said anything about them.
Medicine has lived with this distinction for a long time. Not all smokers get lung cancer. That is true, and nobody concludes from it that smoking has nothing to do with lung cancer. A cause doesn’t have to work on everyone to be a cause. It only has to raise the odds.
So “not all” refutes “all”. It has nothing to say about “some”. When it is aimed at “some”, it isn’t a counterargument. It only sounds like one.
Why the reflex exists
It would be easy to stop there and call the reflex lazy. It isn’t, or not only.
The “all” claim is real, and sometimes it shows up in action. In June 2017, a man drove a van into Muslim worshippers leaving evening prayers during Ramadan near a mosque in Finsbury Park, London. Makram Ali, 51, was killed and twelve others were injured (CBS News, 2018). The court heard that the attacker appeared to be driven by anger at Islamist attacks in London and Manchester earlier that year, and at a grooming scandal in Rochdale: crimes committed by other people entirely (CNN, 2018).
That is the “all” claim in its purest form: the guilt of some, transferred to anyone who shares a label. Anyone who has seen that happen, or lives in fear of it, will reach for “not all” the moment the subject comes up. That fear is not paranoia.
The mistake is narrower. It is using the right answer on the wrong claim.
The question underneath
Strip the “all” away and a real question is left: can religious ideas be one of the things that move some people to violence?
Most people will grant this readily for Christianity. Few would argue that faith had nothing to do with the Crusades, the witch trials, or the murders of abortion providers in the United States. If one tradition’s texts and institutions could do that, there is no principled reason to declare another immune. That consistency cuts both ways, though. It doesn’t say Islam is uniquely dangerous. It says the question is fair to put to every religion, including the ones we are comfortable with.
What it doesn’t give you is the answer, and nobody has a tidy one. Marc Sageman, a forensic psychiatrist and former CIA case officer who spent years mapping jihadist networks, wrote in 2014 that despite more than a decade of funding and thousands of new researchers, the field was no closer to explaining what leads a person to political violence (Sageman, 2014).
The researchers who have tried disagree, sometimes flatly. Robert Pape’s study of 315 suicide attacks concluded that they were mainly a response to foreign occupation, not religious fervour (Dying to Win). Sageman’s own data suggested that people usually joined through friends and family, and that ideology played little part at the start (Penn Press). Scott Atran, an anthropologist who has spent years interviewing fighters and would-be recruits, told the UN Security Council in 2015 that young recruits were searching for identity and significance, and that many felt they had never belonged in the societies they grew up in (UN, 2015).
But notice what even the sceptics concede. Pape, the name most associated with “it isn’t religion”, still concludes that religion plays a role, mainly through the religious difference that makes an occupier feel alien. In Sageman’s account, the friendships that led to violence were formed around mosques and a shared faith. And a group of angry young men still needs a story that tells them their violence is righteous. Ideology may rarely be where it starts. It is often how it gets justified.
That is the honest answer, and it is a modest one: for some people, religious ideas are one cause among several. Not the only one, and often not the first. But not nothing, and the only way to learn how much is to keep asking. “Not all” is how the asking stops.
The same move, closer to home
When women talk about harassment, the reply often comes before the story is finished: not all men. True again. But the claim on the table is rarely that all men harass. It is that enough of them do that most women have a story, and plan their evenings around the chance of another.
When #YesAllWomen spread in May 2014, after the Isla Vista killings, it was a direct answer to that defence: not every man is a threat, but every woman has to reckon with the ones who are (CNN, 2014). The generalisation was never about men’s behaviour. It was about women’s experience, and “not all men” doesn’t touch it.
This is where the Argument earns its keep. If the religion section felt obvious and this one felt irritating, or the other way round, that is worth noticing. The logic doesn’t know which side you are on. “Not all” fails in the same way in both places, for the same reason. (I’ve written more about this from a man’s side in What Feminism Actually Asks of Me.)
Check which claim you were handed.
Before reaching for “not all”, ask one question: did anyone actually say all?
If they did, say it, and say it plainly. It is the right answer, and after Finsbury Park nobody should pretend it’s an overreaction.
If they didn’t, the question in front of you is harder and more useful: what is the link, how strong is it, and for whom? You may well conclude that it’s weak. But you’ll have concluded it, rather than declined to look.
The same discipline applies from the other side. If you think a belief system contributes to violence, say some, and mean it. Name the mechanism you think is at work. A claim stated that precisely is harder to dismiss, and much harder to hear as an accusation against everyone.
In Just Asking Questions, I wrote about the question with a cause hidden inside it: a claim nobody has to own. “Not all” is its mirror image. One smuggles a conclusion in without defending it; the other waves a question away without answering it. Both let everyone skip the part where you look at the evidence.
“Not all” is the right answer to “all”. To “some”, it isn’t an answer. It’s a way of not having to find out.
Written with AI assistance.
Comments