Philosophy

How Much Evidence Should a Claim Require?

The evidence required should rise with the claim's improbability, consequences, precision and resistance to ordinary explanation

Not every claim needs the same amount of evidence. Everyday low-risk claims may reasonably be accepted provisionally, while extraordinary, consequential or highly specific claims require stronger and more independent support.

Ordinary claims often require modest evidence

Claims consistent with common experience may be accepted provisionally when the cost of error is low and correction is easy.

Improbable claims require stronger support

A claim that conflicts with well-established knowledge must overcome the evidence already supporting the existing explanation.

Consequences affect the standard

Medical, legal, financial and public-policy decisions require stronger evidence because mistaken conclusions can seriously harm people.

Specific claims require specific evidence

Evidence that supports a general possibility may not establish a detailed allegation about a particular person, event or cause.

Independent confirmation increases reliability

Several reports derived from the same original source do not provide the same support as genuinely independent observations.

Evidence quality matters more than quantity

A large amount of weak, biased or irrelevant material does not necessarily outweigh a smaller amount of direct and reliable evidence.

Standards should be consistent

People should not demand impossible proof from opposing claims while accepting preferred claims on anecdote or authority.

Evidence notes

The appropriate standard should consider prior probability, source reliability, independence, directness, alternative explanations, risk of error and the practical consequences of acceptance.

Ethical questions

What would be the cost of accepting this claim incorrectly?

Is the evidence independent or repeatedly copied from one source?

Am I applying the same standard to claims I favour and claims I oppose?

Conclusion

A claim should require evidence proportionate to its improbability, specificity and consequences. The aim is not universal scepticism, but consistent confidence matched to the quality of available support.