There is a supposition behind the title question: when should we ignore evidence in favour of a hypothesis? The supposition is, sometimes we may do so, sometimes we may not. Some philosophers have explicitly rejected this supposition, others have implicitly rejected it. Let me outline the brief history of this point.