Religious friend was asking me: what make u believe in things ? e.g. do u only believe in things u see? ...etc.
I wondered few seconds what would be the best way to answer that question. It's actually a very deep question which to be answered correctly would need me to talk about the nature of reality and the inner functioning of brains (not only human brains, but also animal brains). This said, I have maybe found a way to answer. My answer presents things in a slightly simplified way, but I think that it will correctly answer the question.
Part 1. Trust levels
When it comes to reasoning and thinking, you can consider that there are essentially two objects in my mind and my memory. Statements and Implications. An example of statement is "The Earth is a sphere (well, mostly a sphere)", and an implication is "If there is a powercut in my flat, then I won't be able to switch the lights on".
Every single statement in my mind comes with an attached number called its "trust level", ranging from 0% to 100%. If the statement refers to a phenomenon that I can observe myself (for instance the statement "It is sunny in this moment "), then its trust level is 100%. If I have not been out today and could not observe the weather myself, and a friend living nearby (say a floor above mine) tells me on the phone "It is sunny in this moment ", then the trust level will be computed depending on a number of factors, one of them is how much I trust my friend. If the evaluation of this number gives a 50% trust level to the statement, then this will be the "trust level" of this statement in my mind. If later on I can observe the weather myself and see that it is indeed sunny, then the trust level of the statement will jump up to 100%. But if just after the phone conversation I see that it is not sunny, for instance it is raining, then the trust level will move to 0%. I will then know that the correct statement was "it is not sunny in this moment" (with a trust level of 100%), and I will also ask my mind to remember the fact that information coming from this friend will from now on need to be given lower trusts level. Of course, if then my friend calls back and say "I am sorry I told you something wrong on the phone few minutes ago. It is not sunny today". Then my friend will be given back his/her original trust credentials in my mind.
The algorithm for the evaluation of a trust level is relatively complex because I take account of hundred of factors, but once the number is computed, then I remember it. The algorithm is always active as to allow the number to be corrected if a new information comes in.
There are only a couple of people I give a trust level of 100% and usually the stuff they tell me comes with that level. Incidentally, religious people are automatically given a lower trust level than non religious. Everything religious people say is stored in my mind with lots of warning of being potentially untrue. This come from my own observation that "in general religious people are either brainwashed and therefore not able to use their mind correctly, or have a fundamental problem with the real world" (a statement which will be having a high trust level in my mind until I observe a significant occurence of the opposite).
There are also retroactive re-evaluations. If I suddenly discover that one of my friends is religious (without me having known it before), then everything this person said in the past, that I haven't checked myself, is automatically downgraded.
While being at it, I recently downgraded Wikipedia: wikipedia.com and mathematics, the case of L^p spaces embeddings [weblog.alseyn.net].
Part 2 : Necessity
Now, in my mind, the set of statements that have a relatively high trust level is larger than the set of statements I checked myself or that might have come from highly trusted source, how is that possible ? This were the notion of necessity come in. If for instance the statement "it is sunny today" has high trust level (and here the question of considering how it got that level doesn't matter), and since the implication "when it is sunny people then go to the park near my flat" is also true for me (here again how this implication got this trust level is irrelevant), then I will consider that the statement "there are people in the park near my flat" is true with a high level of confidence, say 70%. If later on a trusted friend tells me "I have just seen people in the park", then the trust level of "there are people in the park near my flat" jumps to 90%, and if later on I happen to go to near the park and see the people myself, then the level moves up to 100%.
Necessity also creates impossibilities. For instance, if I know that "If it rains then the ground outside is wet" is true (with high confidence), and I know that the statement "the ground outside in not wet" is also true (with high confidence), then I will consider that the statement "it is raining" is not true (trust level 0%).
Due to positive and negative necessity every new observation I do, has the consequence that everything that somehow relied on this information (or was related to it) have their trust level recomputed automatically (in a cascading manner, like dominos falling).
I also need to mention that the fact that many people might make the same statement has no effect on the trust level of that statement. In particular supernatural claims are treated the same in me regarless of how many people make them. The only way this claim can increase in trust level is if I trust somebody making it or I could check it myself.
On a related note, the couple-of-years-ago, mis measurement of neutrinos moving faster than the speed of light, is an example of supernatural claim (relatively to the Standard Model), made by trusted people (relatively to me), that could be true, but we were not sure. In such a case I simply freeze the computation of the trust level in my mind until the end of the investigation. In this case we discovered that it was a mis-measurement. If it had turned out that those neutrinos were actually moving faster than the speed of light, then the entire standard model and special relativity section of my mind would be frozen until somebody comes from a compelling extension of those theories (which I would have to check myself), which would explain the new observation while being compatible with all the precious observation that we did over the past 200 years.
In an ideal world, I should be checking by myself the validity of every statement I consider true. This is not realistic (I don't always have the time), so I sometimes rely on the trust I have in the peer review process of academic publications etc. This said, things I never contaminate myself with are statements for which a checking process in not possible. For instance, a statements such that "Angels exists, but you cannot check this by yourself, you have to believe it" or "Jesus rose from the dead" or "Mohammed went to heaven on a flying horse", I would give them negative-infinite trust levels, if such thing made sense :-) In practice, I simply consider them untrue (0% trust level). Now, there is the fact that those statements do not exists in isolation, but are stated in religious books. Since I do not trust those statements, and since the books self-claim to be totally trustworthy, then by domino effect I distrust anything coming from those books. This said, I do not take this alone as a proof of the non exitence of God. God could still logically exists; but those books are then necessarily not His word (unless God was lying, which I think is impossible due to God's own properties -- at least the properties states by people who think they know God).
Note that I consider many claims of historical events as being true (with a high level of confidence). For instance, I consider the statement "King Leonidas and his men kicked some Persian butts at the Battle of Thermopylae" as being true (with a high level of confidence -- "high level" and not "totally true", because we all know that history was written by the winners of battles, not losers; those latter having died before having had a chance to share their version of the story). The problem that I have with something like "Jesus rose from the dead" is that to be true this last statement would have to conflict badly with some basic other statements that I know are true. It would be irrational for me to then force myself to consider it true. This would be like starting to believe that my best friend is an alien lizariod from Andromeda galaxy, on the simple basis that I cannot prove it false.
As Carl Sagan (I think) said, "Extraordinary claims require extraordinary evidences, and the burden of the proof rely on whom are making them". Unfortunately for Christianity, the fact that one billion idiots believe that Jesus rose from the dead, doesn't count as evidence. What is easier for me to believe is that one of his friends sacrified himself and took Jesus place on the cross... DNA identification had not been invented yet at that time, so the Roman police --none of whom knew Jesus well enough as to avoid mistaken identities-- could have easily fallen into the trap. The people who knew Jesus the most, would have then done what all of us would do in the same situation: shut up to prevent him being captured. It would then be perfectly easy for Jesus to show up few days later, since he wasn't captured nor dead in the first place.
Part 3 : Possibility
A statement I consider true with a high level of trust is "There is (biological) life outside Earth in the universe". I do not consider it true with high confidence because I observed it, nor because somebody I trust told me, but because this statement (and the related fact) seems to be a natural consequence of my biology and cosmology knowledge. Where "natural" means that "the probability that there is no other life in the universe is as much as a person winning national lottery 100 weeks in a row... (without this person cheating and without bug in the lottery engine)". What I am trying to say is that it's easier for extraterestrial life to exist than for extraterestrial life not to exist.
Part 4 : Mathematical and Computational universes
There are statements that I consider true because of mathematical necessity. For instance the statement "there are (at least) two humans with the same number of hair on their head" is true (100%) simply because the number of hair on somebody's head is between 0 and 1,000,000 (actually between 0 and 500,000 if I well remember) and we are several billion humans. (If you don't get it, never mind, just know that the claim is very easy to prove mathematically.)
Part 5 : Conflicts
If I suddenly realise (remember that my mind updates and maintains my network of statements truth levels automatically) that two opposite statements are simultaneously given high trust levels, then this is an relatively serious emergency mental event to me. In such a case, the most important thing in my life becomes to investigate until the conflict is resolved, no matter how long it takes. Often the investigation results in the fact that somebody had made a grammar mistake or a typo (inverting the sentence they wanted to make).
Part 6 : Conclusion
In conclusion, the set of statements I consider true comes from a network of statements and their trust level (and constant re-evaluation of the latter). The core statements are what I could check, compute, see myself etc, together with all the other statements that I could find that are true due to basic necessity, basic probability, the fact that the universe is unsurprising (in other words that the physical laws do not fluctuate randomly), together with mathematical or computational necessity.
Anything outside this set is considered religion to me, meaning that I should stay away from it like a contagious disease. And of course, to me religion is the ultimate form of mental laziness. A form of reasoning where people make statements and consider them true, most of the time because somebody they like made it, often because it makes them feel good or reinforce whatever previous unfounded beliefs they already had, without ever asking themselves whether they are actually true or not.