Believes are very important within NLP. First of all one must make a distinction between believes and convictions. Second of all, one needs to take into account that a belief is a nominalization.

That a beliefs strength alone guarantees nothing whatsoever about its truth, in fact is even capable of slowly, slowly distilling out of the most reasonable thing a concentrate of folly: this is our real European insight in this, if in anything, we have become experienced, been made cautious, shrewd, wise, apparently through much injury … ‘He that believeth shall be saved’: fine! Now and again, at least! But he that believeth shall most certainly be made stupid, even in the rarer case that the belief is not already stupid, that it was an intelligent one in the first place. Every long-held belief finally becomes stupid, which means (to express it with the clarity of our modern psychologists) that its reasons sink ‘into the unconscious’, disappear there – from then on it no longer rests upon reasons but upon affects (that is, whenever it needs help it gets the affects, and no longer the reasons, to fight its cause). Supposing one could discover which was the most strongly believed, longest held, least disputed, most honest belief that exists among men: it would then be highly justified to conjecture that this belief could also be the most profound, most stupid, ‘most unconscious’, the most thoroughly defended against reasons, the longest abandoned by reasons. – Agreed: but which is that belief? – Oh, you’re curious! But since I’ve started setting you riddles, I’ll be fair and come out quickly with the answer and solution – they won’t be easily anticipated. Man is above all a judging animal; but in judgment lies concealed our oldest and most constant belief. Every judgement rests on a holding-to-be-true and an asserting, on a certainty that something is thus and not otherwise, that in it man has really come to ‘know’: what is it that, in every judgment, is unconsciously believed to be true? – That we have a right to distinguish between subject and predicate, between cause and effect – that is our strongest belief; in fact, at bottom even the belief in cause and effect itself, in condition and conditioned is merely an individual case of the first and general belief, our primeval belief in subject and predicate (as the assertion that every effect is a doing and that every conditioned presupposes something that conditions, every doing a doer, in short a subject). Might not this belief in the concept of subject and predicate be a great stupidity?

Notebook 4, beginning of 1886 – spring 1886 paragraph 8