Christopher Hitchens's phrasing — what can be asserted without evidence can be dismissed without evidence — is a restatement of a rule that goes back at least to the Latin lawyers: the burden of proof lies on the one who asserts, not on the one who denies.

It is a rule about procedure, not about truth, and almost every misuse comes from forgetting that.

What it settles

It settles who has to speak next.

In an exchange where one party makes a positive claim with nothing behind it, the other party is under no obligation to disprove it. Declining to accept an unsupported claim is not itself a claim; it is the absence of one, and demanding evidence for a non-acceptance is a move that reverses the burden and can be repeated indefinitely.

This matters because the alternative is unworkable. If every assertion had to be refuted before it could be set aside, anyone could occupy any amount of anyone else's time by generating claims — which is exactly the asymmetry that Brandolini's law describes and the reason this razor is a practical necessity rather than a debating preference.

What it does not settle

Whether the claim is true. An unsupported assertion can be perfectly true; the person making it may simply have offered no reason. Dismissal is a decision about what to do in the absence of a reason, not a verdict on the content, and a razor used as though it delivered a verdict is being used as a rhetorical instrument.

Anything about claims that do have evidence. Once some evidence is offered, however weak, the razor is finished. What follows is an argument about the quality of the evidence, and that argument cannot be shortcut by re-invoking the rule.

Where the burden sits in a genuine disagreement. Both sides in most real disputes are asserting something. When two parties hold competing positive claims, the rule does not adjudicate — it just says each owes support for their own, which is usually where the conversation should have started.

The relation to the other razors

Occam's razor is about choosing among explanations that all fit the evidence: prefer the one with fewer assumptions. Hitchens's is about what to do when there is no evidence at all. They answer different questions and are frequently quoted as though they were versions of each other.

The connection to falsifiability is closer. A claim constructed so that no possible observation would count against it cannot generate evidence in its own favor either, which puts it permanently on the wrong side of this rule.

The honest limit

Used well, it is a defense against being drawn into refuting an endless supply of unsupported claims. Used badly, it is a way of not engaging with anything, and the failure is hard to see from the inside because the move is identical in both cases.

The test is whether the claim has been offered with a reason. If it has, the razor does not apply and reaching for it is an evasion. If it has not, nothing is owed — and the correct thing to say is that no reason has been given, which is a smaller and more accurate statement than declaring the matter closed.