The basic assertion which I propose is that, epistemologically speaking (i.e. in terms of what and how we can know anything), the act of believing precedes all knowledge – not in a temporal sense, but in a logical sense. In other words, one cannot "know" something (as a proposition) without "believing" the very same thing (as a proposition) for whatever reasons, although it is possible to "believe" something which is not true and therefore it cannot constitute knowledge in such a case.
But if this claim is true, then the common notion that a strict dualism between science (as the embodiment of "knowledge") and religion (as the embodiment of "belief") exists, cannot be true. The differences have to be sought elsewhere.
My assertion is based on an old and basic (although perhaps simplistic) definition of "knowledge": knowledge is justified true belief. Unless knowledge can objectively and consistently be defined otherwise, so as to not include reference to belief in the definition, the notion that science deals with "knowledge" as strictly opposed to religion dealing with "faith" is not true – by definition.
What is true when a comparison between the natural sciences and religious beliefs is made, in the light of this definition, is that the "justification" part of the definition generally relates to different kinds of ways in which justification is brought about.
Not everything we believe to be true finds its justification in empirical observation, and not everything we believe finds its justification in the principles of logical reasoning. (Reason and observation are the hallmark of the method employed by the natural sciences).
Much of what we believe to be true also finds justification in the testimony of other people, and may justly be considered "knowledge". For example, I can know that Napoleon Bonaparte once lived on the face of the earth, not because reason leads me to that conclusion, neither because I have seen him, but because I believe the testimony of many people to serve as justification for why I believe it.
If it is true that Napoleon Bonaparte did live on this earth, then my belief that it is true constitutes "knowledge"; if it is not true, then my belief is false (even if justified) and it cannot constitute "knowledge".
Whenever the idea of "evidence" is mentioned, I would suppose that a call is made to some kind of justification that would render knowledge out of what seems to otherwise be an unjustified belief or a myth. In my opinion, the problem that often occurs when religious beliefs are judged is that people insist on the kind of justification required in natural sciences in all respects, when in its very nature the epistemological problem differs in many respects from those in the natural sciences. (It may be true that one might find merit in the classical proofs of God's existence, hence employing the required reason used by the natural sciences, and it might have its place, but ultimately most theists do not consciously base their beliefs on these proofs and most probably don't even know too much about these.) The scientific method employed by the natural sciences can only validly function in its own domain, and has therefore little, if not nothing at all, to say about matters relating to ethics, politics, historical events (unless by extrapolation based on some assumptions) and metaphysics.
Another problem is that the "justification" behind propositions often functions within a broader spectrum of beliefs, which constitute what is known as a paradigm or a belief system, amongst which it has to be consistent, for otherwise either the paradigm is wrong or the proposition lacks inherent justification. Even the justification required within the scientific method, which many purport to be "objective", ultimately rests upon "unprovable" and not necessarily indispensable propositions within paradigms.
For example, whenever a scientist declares that nothing exists outside the realm of the natural world, the very statement cannot be proved but has to be believed without any external justification. All empirical data, all observations, will be interpreted within this naturalistic paradigm. Thus, no piece of "evidence" comes without interpretation that relates to a particular paradigm (or worldview), implying that whatever "evidence" may be given to the atheist if it is true that God indeed exists, could be explained away to be consistent with the paradigm that he adheres to.
One atheist might require that God strike him with a lightning bolt as "evidence" of His existence, and another will claim that it is mere coincidence, a statistical event that is not impossible by chance, if it does in fact happen. Ultimately, an atheist will have to decide what kind of "evidence" would be required and be indubitable, objective for God to use if He indeed exists; if no such indubitable, objective criterion exists, but if God exists, may He not choose what kind of evidence He would provide? (Of course, Christians would claim that such evidence exists, and atheists and agnostics would claim that it does not count for evidence). It therefore seems to be impossible to acquire completely objective principles by which all propositions may be judged as having warrant or not. The principles underlying the natural sciences most assuredly do not suffice for all attainable knowledge.
Applying this to the Christian faith, the following could be said:
In terms of the historical accounts of Scripture, Christians believe these accounts as the testimony of eyewitnesses (for the most). As far as the metaphysical or ontological interpretations are concerned, the justification for the Christian belief is brought about by different means: the internal, self-authenticating witness of the Holy Spirit.
Although the Bible declares that God has given all people the evidence that is required to believe in the existence of God, so that no one is without excuse, and I believe that this evidence is rejected and interpreted in such a way that it is not deemed evidence, much of what Christians believe otherwise – especially the Gospel message as such – finds personal justification (or warrant) in this witnessing work of the Holy Spirit. (For this reason, I believe the FSM argument falls apart, for neither has any flying spaghetti monster revealed itself to be the source of all reality, nor is there an internal witness justifying a belief in the FSM.
If anyone has received such revelation and there indeed is some self-authenticating FSM witness involved, then they would be justified to believe in the FSM, and if it indeed be true, it would constitute knowledge.)
The unbeliever does not have access to this witness, and therefore the Christian faith will always be unjustifiable and therefore foolish to him. Yet neither reason nor empirical observation can contradict the message of the Gospel, because both of these epistemological tools are dependent on preceding ontological premises that do not originate from either observation or reason.
The Christian is also not required to prove his faith merely by these sources of knowledge, for he cannot provide objective, indubitable proof to the unbelieving world that the most fundamental propositions pertaining to his faith are all true, and neither can others provide objective indubitable prove that they are false. If what the Christian believes is indeed true and therefore constitutes knowledge, his warrant rests subjectively in a sphere inaccessible by the world at large: in the self-authenticating work of the Holy Spirit.
Yet the Christian can know the truth about what matters most – if what he believes is indeed true – and he can have this knowledge without the conclusions of any faction amongst the scientific community, without contradicting any indubitable "evidence" within the realm of the natural, observable world and without contradicting the fundamental principles of reason.