Violence isn’t the answer
The phrase does three jobs at once and they are not the same job.
The first is a moral claim: violence is wrong, full stop, regardless of whether it works. That is a defensible position, but it stands on its own and does not need evidence.
The second is an empirical claim: violence does not achieve what its users intend. That is a testable claim, and it is not simply true. Wars have settled things. Revolutions have replaced regimes. Self-defense stops attacks. Police force ends some situations that would otherwise continue. Any honest version of the argument has to say what “working” means and over what timescale.
The third is a strategic claim, and this is the one most people actually mean: violence is a poor instrument for the goals the person using it claims to hold, because it is slow, expensive, imprecise, and produces consequences that outlast the original grievance.
The slogan collapses all three into one line, which is why it is easy to say and hard to defend. Pressed on the empirical claim, the speaker usually retreats to the moral one. That switch is what makes the argument feel slippery to the person hearing it and feel obvious to the person making it.
What I would keep: the strategic claim is strong and underused. What I would lose: the implication that the empirical claim is settled, because it is not, and pretending otherwise hands the argument to whoever can name a counterexample.
In any case, let us look at the actual evidence, because it is more interesting than the slogan.
The research that exists on this splits violence into two modes, and the distinction does most of the explanatory work.
Instrumental violence is used as a means, with a target and a theory of how the target responds. It gets evaluated on whether the theory was right.
Expressive violence is used to discharge a feeling or assert an identity. It is not really aimed at an outcome, so the question of whether it works does not quite apply, which is why arguments about it tend to go nowhere.
On the instrumental side, the honest summary of the historical record is that force succeeds under narrow conditions and fails broadly outside them. It tends to work when the target is clearly identified, the force is proportionate to the aim, the campaign can be sustained, and the objective is limited enough that there is a point at which it stops. It tends to fail when the objective is open-ended, when the target is a population rather than an actor, when the resolve of the other side is underestimated, and when there is no plan for the day after.
Get Rhi Gladney’s stories in your inbox
Join Medium for free to get updates from this writer.
The counterinsurgency literature is the clearest case, because it has been studied and against itself for decades. The finding that keeps reproducing is that killing the wrong people creates more combatants than it removes. Coercion generates the resistance it is meant to suppress. The mechanism is not mysterious: an act of violence recruits, it legitimizes the enemy’s narrative, and it closes the political options of the moderates who might otherwise have settled.
On deterrence, the evidence is genuinely mixed and heavily dependent on context. Deterrence works best with clear signalling and a credible commitment, and it is not obviously separable from other causes. It also gets a lot of credit that belongs to the absence of the conditions for conflict in the first place.
On states and revolutions, I would not put a number on this. Almost everything quantitative in that literature is contested, and it is a field where the definition of success is doing an enormous amount of the work.
The one case where the “violence does not work” claim is weakest is immediate self-defense, where the moral and empirical cases align: the alternative is worse now, and the timescale is short enough that long-term consequences are not the deciding question. Any honest version of the article has to carve this out, because the person hearing it will.
Which brings the argument back to the strategic claim, and that one holds up much better.
If you grant that force sometimes achieves limited aims, the case against it stops being about impossibility and becomes about cost, and cost is where it actually loses.
The costs are asymmetric in time. The action happens once. The consequences keep arriving. A person who uses violence to settle something is usually still paying for it years later in relationships, in reputation, in legal exposure, in the set of options they no longer have. This is true of a state as well: the war ends and the reconstruction does not.
The costs are also invisible at the moment of decision. The counterfactual is never tested. The person who resolves a problem by force never learns what the other routes would have done, so the belief that force was necessary is never corrected. You get a confident person with a small sample and no control group.
And the costs land on people who did not choose them. This is the part the strategic framing understates: it is not only that violence is inefficient for the user, it is that the inefficiency is paid by third parties who had nothing to do with the decision.
Which is why the strongest version of the argument is not “violence never works.” It is narrower and much harder to answer:
You are probably wrong about how this ends, you will not be around to be told so, and the bill does not come to you.
What to do with that instead. Slow the decision down, because violence is almost always chosen under time pressure that is itself manufactured. Name the actual objective, because most of the time the objective is a feeling rather than an outcome. And separate the grievance from the response, because the strongest predictor of whether a conflict escalates is not the size of the grievance but whether the parties have a route to being heard. Institutions, mediators, courts and process are slow and frequently do not deliver justice. They still beat the alternative on cost, which is a lower bar than people want and a higher one than is often met.
Where this leaves the slogan: it is a good instinct, badly worded. “Violence is the last resort and it is usually not the last resort” would be more accurate and much less quotable.
