Not all atheists are supercilious, of course. Many are content to live and let live, and some even grant that religion (which, in America, basically means Christianity) does some good. But atheism as an organized, evangelizing movement has been on the offensive lately. Witness the "New Atheists" such as Richard Dawkins, Sam Harris and Christopher Hitchens, with their aggressive stance against God and their bestselling books attempting to debunk religion. So, assuming you are a theist, what do you say to the atheist who asks, "You don't (chuckle) actually believe in God, do you (snicker)?"
The natural response would be to start giving evidence for God: the origin of the universe in the Big Bang requires a cause that is beyond matter, energy, space and time, the design of life requires an intelligence to account for the information that it contains, the many accounts of miracles and the supernatural cannot all be fabrications, and so on. Entire libraries have been written on the evidence and arguments for God.
But before you start showing them the evidence, consider: Most aggressive atheists say "I would be willing to believe in God if there were any evidence that He exists, but no such evidence exists, so I don't believe." No matter what evidence you give, the supercilious atheist finds a way to dismiss it. To him, it is not the case that your evidence for God is valid but nevertheless is cancelled out by his superior evidence against God. No, in the atheist's mind your evidence does not even count as evidence. And therefore your reasoning has no effect on his thinking, other than to confirm to him that you are irrational.
What's going on here? Does the atheist have superior insight that allows him to see the errors that invalidate the arguments for God that seem valid to us theists? Or is it the atheist who is missing something?
I would argue that it's the latter. Consider: the theist believes in the real existence of everything that the atheist believes in: matter, energy, space, time. The theist believes that the physical world really exists, just as the atheist does. And the theist believes that the scientific description of nature is fundamentally correct, as far as it goes.
But the atheist refuses to expand his mental universe by also believing in the transcendent things that the theist believes in: God, souls, angels and demons, for example. The atheist restricts himself to a sort of tunnel vision.
And this is where atheism becomes vulnerable. The atheist does not disbelieve in God because he has neutrally examined all the evidence, and drawn the proper conclusion that there is no God. On the contrary, the atheist radically misconstrues the plentiful evidence for God, and he does this because of his false worldview, which tells him that only the physical really exists. Before he has examined the evidence, the atheist thinks he knows that nothing non-physical actually exists, and this assumption governs how he responds to the evidence.
About that word "worldview:" It means a comprehensive system of thought that describes the nature of reality, answers the big questions of life, and provides man with a code of conduct. Most Western atheists have a worldview that the philosophers call "naturalism," the basic elements of which include atheism, empiricism (the doctrine that all knowledge is obtained inductively, based on our sense perceptions), and materialism (the doctrine that only matter and its properties exist).
But is it true that only the physical really exists? How would one test this belief? Certainly not by using one's senses to try to detect non-physical entities; that would be rather like the old joke:
Secretary: "Sir, the Invisible Man is here".
Boss: "Tell him I can't see him!"
Of course you can't see the non-physical; it's invisible. Imagine a man, blind from birth, who is skeptical of the existence of color even though he frequently hears other people talking about it. Just as it would be foolish for the blind man to conclude that color does not exist simply because he is unable to detect it with his senses, it is foolish for the atheist to dismiss God because he cannot detect God with his senses.
There is only one effective way to respond to the supercilious atheist's question: Speak his language, the language of evidence and reasoning, of logic and proof. Challenge him to give his reasons for disbelieving.
But that's not enough. You have to challenge his assumptions, which are the real impediment  to his believing. Say something like the following:
"I believe in God because that's what the evidence shows. But before you try to debunk my evidence, we have to ask, what are your criteria for deciding whether a God exists, and how do you know that these criteria are correct? Until you can have confidence that you have the correct criteria, it is useless to begin investigating God's existence."
So when the atheist asserts that there is no evidence that any miracle has occurred, ask him: "What sort of evidence for a miracle would you regard as being valid? And how do you know ahead of time that any miracle not validated by this type of evidence must not have occurred?"
And when he says that naturalistic (that is, atheistic) science provides the more plausible explanation for the entire history of the universe since the Big Bang, ask him: "How do you know that a super-naturalistic explanation, involving a God who intervenes from time to time, cannot be the correct explanation? Wouldn't one have to be, for all intents and purposes, omniscient in order to know that God could not have been involved?"
At this point, the atheist may respond by asserting that it would be a logical contradiction for the God of the Bible to exist, and since contradictions do not occur, there must not be such a God. For example, the atheist may declare that God cannot be omnipotent, because then He could create a rock so big that even He cannot lift it, which would contradict His alleged omnipotence: If God is omnipotent, then He is not omnipotent; therefore no God. Q.E.D.
This argument, unlike the previous ones, actually has a chance of being valid. If God were contradictory, then He would not exist. But unfortunately for the atheist, this argument only proves that a God who can create a rock so big that He cannot move it is not literally omnipotent in the sense that he can do anything that can be conceived of. But once we admit that there are some things God cannot do (e.g., create a square circle), the contradiction disappears. A God who is omnipotent in the sense that He can do anything that can be done may indeed exist.
Similar comments pertain to the atheist's argument that God cannot be both omnipotent and all-loving, because evil does occur. After all, says the atheist, if God were all-loving, He would want to eliminate all evil, and if He were all-powerful, then He would be able to do so. And since evil persists, then God must be either not all-powerful, or else not all-loving. But once again, this argument only proves that a God who eliminates all evil does not exist. Some other sort of God, one who for some reason allows evil, may very well exist.
It is not the purpose of this essay to give an exhaustive account of the arguments against God; that would take an entire library. Instead, my goal has been to indicate to the believer that the atheist's arguments are, in fact, much weaker than they may appear to someone who is untrained in the art of defending the faith. These arguments are not the result of dispassionately and scientifically examining the evidence, and discovering that it clearly shows the absence of God. On the contrary, the atheist's position is the result of several highly questionable assumptions about how reality operates; assumptions that cannot be proved without violating the atheist's way of thinking.
Consider: all knowledge  must ultimately be based on other knowledge that is not proved, in the ordinary sense of the word "proved." For to prove a proposition,  we must appeal to other propositions that are already known to be true: X is true because Y and Z are true. And therefore if we demand that every proposition we know must be proved, then we will have an infinite regress of proof: Y and Z are true because A, B and C, are true, and A, B and C are true because etc...
But such an infinite regress, combined with the demand that everything must be proved in order to be known, means that we know literally nothing. Some postmodernists may be content to say that they know nothing, but normal people will recognize the necessity that some knowledge be unproved if we are to know anything.
In mathematics, the fundamental propositions that are accepted as true in order to begin the business of validating knowledge are called either "axioms" or "postulates." Although these axioms are not always explicitly articulated in discussions of mathematics, they are always there, and sometimes making progress in a particular field of mathematics will require clarifying the axioms.
But the same is true about any field that claims to be knowledge rather than just speculation or ego assertion. If anything is to be regarded as known, it will have to be based ultimately on true propositions that are not proved, because the process of proof must always terminate somewhere. Taking these axioms as the starting point, we then combine them to prove the other propositions that make up the particular field of knowledge we are working with. In most cases, and taking science as an example, we will also make use of empirical data that will give us additional truths that we can include in our system of knowledge.
To say that we do not prove the most basic propositions does not mean that we must choose what to believe by a purely arbitrary "leap of faith" that is unaccompanied by any evidence or reasons. There are, in fact, two basic ways that we test the system of axioms that create the particular field of knowledge we are studying. First, does the axiom, at least as far as we can understand what it asserts, agree with what we know by intuition? "Intuition" means the mind's ability to know something immediately, without engaging in a process of calculation or reasoning.
In a relatively (logically) simple field such as mathematics, intuition is often enough to validate the axioms. For example, the axiom "through any two distinct points there is only one (straight, infinitely long) line" is self-evident, once we understand what the words mean, and as long as these words have their everyday meaning.
But outside of well-defined fields such as mathematics, two people do not always have the same intuitions. Fortunately, there is a second way, more important and fundamental than the first, of testing a system of axioms: Is the system logically, morally, and existentially consistent?
Logically consistent means that the axioms, and the propositions we can prove from the axioms, do not contradict each other. For example, the naturalistic worldview of atheism holds that all knowledge must be based on empirical data. But this belief about the nature of knowledge cannot be validated empirically, and therefore naturalism contradicts itself.
The requirement of moral consistency only applies to systems (unlike mathematics) that make moral demands on individuals and groups. This is basically the requirement that moral precepts not contradict each other. Naturalists, for example, generally assert that morality is made up by the group. But if morality is whatever the group says it is, then nobody has the right to say that the group is wrong, and so reformers such as those who fought against slavery are always wrong: If our ancestors said that slavery was acceptable, then it was acceptable. This sort of naturalism is thus morally contradictory, at least over time.
Finally, a system such as atheistic naturalism also can be judged existentially, that is, by judging whether individuals and groups can live within its dictates. As an example, naturalism, strictly speaking, declares that there is no such thing as "the meaning of life." To the strict naturalist, your life means whatever you say it does.
If so, then your life can mean X, and my life can mean non-X, and both our answers are equally valid. (X could be, for example, "Always putting my desires above the desires of all others.") But this means that both our answers are also equally invalid, in which case I have no reason to believe that my life really should mean non-X. And if there is no reason why my life should mean non-X, it will be impossible for me to keep to my professed non-X ideal: if it's all made up, why stick to it when the going gets tough? Naturalism is unlivable.
The Ultimate Test of Truth
According to what has been said so far, a system of axioms must be judged collectively, as a system, and not individually, on an axiom-by-axiom case. This is one of the key points of this essay, so I will explain it carefully:
To prove that an individual statement is true will require that we already know the truth of other statements that, taken together, imply the truth of the statement we are trying to prove. (That's what the word "prove" means: giving reasons.) But these statements that we already know to be true, like all statements that we know to be true, will ultimately be true because of the system of axioms that we accept without proof. And therefore our proof will never be any stronger than the system of axioms that grounds our worldview. These axioms, by definition, cannot be proved by appealing to other statements that we know to be true. Since they are axioms, they are not proved.
Furthermore, we are not referring to a relatively straightforward field of knowledge such as mathematics. We're talking about a comprehensive worldview, a system of that will describe the nature of all reality, answer the big questions (such as the existence of God), and provide a code of conduct. Therefore, our axioms will not be known simply "by intuition." Intuition cannot be counted on to answer correctly subtle and cosmic questions.
Since none of these individual axioms can be proved, our only recourse is to judge the system of axioms as a whole. This act of "judgment" is less precisely defined than the act of proving a statement by deduction. If, for example, we know that the proposition "If A, then B" is true, and also that A is true, we can deduce with 100% certainty that B is true. But the act of judging a system of axioms by checking it for the forms of consistency mentioned above is less certain than formal logical deduction. However, it is the only means available for us to make what is in fact the supreme test for truth, namely, to test our worldview.
Is it true that the supernatural does not exist? We can only answer this question by judging the system of axioms of which it is a part, because intuition is not a reliable guide to that which goes beyond our senses and our ordinary experience.
And therefore the proposition "God exists" cannot be judged except from within a complete worldview. Everyone has some sort of comprehensive system of thought that they use to judge true and false, right and wrong, and beautiful and ugly, and this system always seems right to the one who holds it. But if the atheist says, "The supernatural does not exist," his statement is not true just because he says it. It must be judged true or false in light of what we already know to be true, and what we already know always boils down to the overall worldview we possess.
When we judge the atheist's naturalistic worldview, we find it wanting. Some of the reasons have been indicated above, and more details can be found in my online essay The Scientific Leftists of the Center for Inquiry. Therefore there is no good reason to believe the atheist's assumption that the supernatural does not exist. This assumption is part of a faulty system of thought, which must be abandoned.
It is always possible to have a belief that is true, that is, in accordance with the way things really are, even though your reasons are faulty or even nonexistent. Furthermore, some people are content to believe without having any proof of their beliefs, and you can't argue with someone like that. But this essay is intended for those who care about giving, and having, reasons. And if you are such a person, the conclusion is inescapable:
Naturalism fails. Some other worldview, one that is consistent, must be found. Although it would take another essay to even outline the argument, a theistic worldview is more consistent and does a better job of accounting for the facts of reality. And in any case, the supercilious atheist has no ground to stand on.
So where does the rubber meet the road? You probably cannot deliver an off-the-cuff lesson on the necessity of judging worldviews by their consistency when the supercilious atheist challenges you. So do the next best thing: ask him how he knows that atheism is true (or probably true), and then ask him how he knows that his criteria for deciding the question are the correct criteria. He has probably never been asked these questions, and in any case, it is impossible to prove that his naturalistic criteria are correct without violating naturalism itself. In any case, then, you've defeated his challenge.
Will your strong response to his challenge cause the atheist to rethink his position? Probably not. But making a strong case against atheism will nevertheless be valuable for at least two reasons. For one thing, the uncommitted onlookers will probably be impressed; like the atheist, they may very well have never encountered an intellectually rigorous defense of theism. A strong argument will often impress people far more than they let on, so don't be dismayed if nobody seems to respond positively.
But let's not overlook the benefit to you of vigorously defending theism. By learning the arguments against atheism, and using them (even if only in your mind) against its strongest defenders, you increase your confidence and trust in God. And this isn't just self-esteem gobbledygook; "trust and confidence in God" is another name for what the Bible describes as the indispensable requirement for being saved: faith. Of course, just believing a God exists is not enough; you also need to know about, and have faith in, the correct God. But that's a subject for another day.
Alan Roebuck is a Reformed (that is Calvinistic) Christian.
 Intellectual impediment, that is. The more fundamental reason he disbelieves is probably emotional: he hates God. But one cannot deal with emotions intellectually, and one's intellect does influence one's beliefs; man is not a prisoner of his emotions.
 "Knowledge," by definition, is true belief that has been justified in some logically valid manner.
 "Proposition" is the technical name for a statement, that is, a sentence that asserts that something is the case, as opposed to a question, a command, or an interjection.