Science (ie, rational thinking) mustn’t be confused with belief, because it can’t tell what something is or was. It can just model reality and manipulate it. The reason is that it can’t tell what is true, but just what isn’t true (ie, falsify). It can thereby falsify belief, but not replace it.
Science’s hope is that it ultimately will be able to tell what something is and was by peeling off everything that isn’t and wasn’t, but this is a contradictory hope, since if it could, then it would actually be able to tell what is true, which it thus isn’t. Instead, consistency tells us that this “peeling off” lies is endless. Never will science be able to tell what is and was (and thus be able to replace belief).
So, our choice between science and belief is thus not a choice about what we shall believe in, but rather about whether we shall believe or not. The scientific alternative is thus neither another belief, nor belief in disbelief (eg, atheism), but non-belief. Consistent science (rationalism) simply denies all forms of belief.
But, since science is a method, not a belief, a scientist may well believe in something as long as he doesn’t confuse this belief with his method, because as soon as he does, he’s not a scientist. A scientist is thus a scientist as long as he doesn’t confuse his belief with his method. It may be problematical to entertain a belief at the same time as falsifying beliefs, but it is not contradictory as long as one doesn’t confuse them, just ambiguous. The problem is contradictions, not ambiguities. We can’t avoid both of them.