Many atheists say that atheism is the default position, so theists have the burden of proof. I will consider the best argument I can come up with that theism is the default position and the best argument I can come up with that argument that atheism is the default position. I believe that the argument that atheism is the default position seems a little more plausible, but I am not yet convinced that either of the arguments are rationally compelling. Even so, the arguments I will present are merely food for thought and could be considered to be a starting point when considering whether atheism or theism is the default position (or perhaps neither).
What is the default position? What I mean by a default position in this context is that there’s a better reason to think something is true than the alternative when considered in isolation of our knowledge of the world. To think that atheism is the default position would then mean a person who has no knowledge of the world would have more reason to think gods don’t exist than to think at least one god exists. We would then say that theists have the burden of proof in a debate to give people at least some reason to think that at least one god really does exist—otherwise everyone will continue to have more reason to think gods don’t exist than to think they do.
What I call “atheism” in this context is literally the view that gods don’t exist. Many atheists say they are in some sense undecided and have no opinion about whether gods really exist or not. That is not what I am talking about.
For any given belief, it is often reasonable to think that the default position is simply being undecided—to be skeptical that particular belief is true. We would then say that both sides of the debate have the burden of proof. Both sides should give arguments.
In everyday life the default position is not necessarily relevant to us because we already know a lot about the world and we might already be familiar with various arguments. A theist could very well have a good reason to believe in God and an atheist could very well have a good reason to believe no gods exist. In that case the default position will do little to change their mind. We would have to debunk the arguments they have to give them a good reason to change their mind.
A reason to think theism is the default position
The reason to think something is the default position is based on mathematical or logical principle of probability rather than because of what we know about the world. The assumption is that we can at least imagine that we know nothing about the world while still giving arguments involving the probability of something being true. Why think theism is the default position? Because there’s an unlimited number of possible gods and one could argue that to claim that none of them exist is a stronger position than saying that at least one of them exists.
Consider some gods that might exist: Yahweh, Zeus, Thor, Quetzalcoatl, and the Force.
Each god has at least somewhat different characteristics. Some of them are much different. Yahweh is an all-powerful person, gods like Zeus are much more powerful than humans are, and the Force is not a person at all with no thoughts or intentions.
We will represent each god with a letter (A) for Yahweh, (B) for Zeus, (C) for Thor, (D) Quetzalcoatl, and (E) the Force. “Not-A” will translate to mean “Yahweh does not exist” and so on.
We could then say that theism only requires that at least one of these gods exist:
(1) A or B or C or D or E…
Atheism requires that none of them exist:
(2) Not-A, not-B, not-C, not-D, and not-E…
The odds of (1) seems higher than (2), so theism appears to be the default position.
There’s a rule of probability that states that “A or B” is more likely true than “A.” For example, the odds of an alien species from another planet being a silicon-based life form or a carbon-based life form is higher than the odds that the alien will be a silicon-based life form.
Also, assuming that we know nothing about the odds of “A or B” being true, it seems plausible to assume that “A or B” is more likely true than “not-A.” For example, it is plausible that the odds that an alien species from another planet being a silicon-based life form or a carbon-based life form is higher than the odds that the alien isn’t a silicon-based life form.
Also consider an undiscovered species of mammal and an undiscovered species of fish. It seems plausible to say that it’s more probable that there is either an undiscovered species of mammal or an undiscovered species of fish than the odds that there is no undiscovered species of mammal.
I believe that this simplistic formulation of the alternatives misrepresents the actual alternatives and the odds of them being true. The problem is that saying a god exists is actually several assertions. It’s not clear what exactly counts as a “god” because they are all so different, but a plausible list of godlike characteristics are the following:
It has thoughts.
It is much smarter than human beings.
It is much more powerful than human beings.
It is immortal or eternal.
It is a person.
It is always or often without a physical body.
It exists everywhere.
It exists outside of space and time.
It can interact with the world.
It is often or always invisible.
I don’t think all people will agree that god(s) have all ten of these characteristics, but there is a “family resemblance.” To count as a god, perhaps at least five of these characteristics have to be involved. Otherwise we might just be dealing with a powerful alien, a ghost, or an energy-based life form.
Many seem to think that Yahweh has all ten characteristics. He has thoughts, is much smarter than human beings, is much more powerful than human beings, is eternal, is a person, is often without a physical body, exists everywhere, exists outside space and time, can interact with the world, and is often invisible.
Zeus has five of the characteristics. He has thoughts, he is more powerful than human beings, he is immortal, he is a person, and he can interact with the world.
The Force has five of the characteristics as well. The Force is immortal, is always without a physical body, exists everywhere, can interact with the world, and it’s always invisible.
Rather than claiming that theists are saying “A or B or C or D or E” it would be more accurate to say that they are saying something more like the following:
(A and B and C and D and F) or (A and B and C and F and G) or (A and B and F and G and H) or (A and F and G and H and I) or (A and F and G and H and J)
Because the theistic position was misrepresented, it is no longer clear that it is more probable than atheism prior to empirical investigation. The odds of each god existing is much less probable once we take them to require multiple assertions to be true. The odds of “A” being true is higher than the odds of “A and B and C and D and E” being true.
The fact that we can formulate what initially looks like a single assertion as multiple assertions isn’t enough to say that it should be done. However, I believe it should be done in this case because we need to know that everything we consider to be a god really is a god. Only by realizing that the concept of gods is ambiguous because it’s based on a family resemblance can we determine what atheists need to reject. Some things that have been said to be gods, such as the Sun, Earth or Universe, should be excluded from the list of gods that atheists must reject (unless such things are also said to have certain godlike characteristics from this list).
Moreover, it’s not entirely clear how we should define gods for the purpose of the debate over their existence. I believe it can be fair for an atheist to assume that all gods must have at least certain godlike characteristics to be considered relevant to the debate. Various non-god supernatural beings, such as ghosts, energy-based life forms, angels, and demons should be excluded. In that case being an atheist could be compatible with belief in the Sun, the Earth, the Universe, ghosts, energy-based life forms, angels, and demons.
A reason to think atheism the default position
Let’s say that we define “god” in a non-arbitrary way. For example, the following five characteristics are plausibly characteristics that every god has in order to count as a god:
It is immortal or eternal.
It is always or often without a physical body.
It can interact with the world.
It is often or always invisible.
It is not a ghost.
In that case we can characterize theism as stating that something has all five of these characteristics:
(A and B and C and D and E)
The atheist will then deny that something has all four of the characteristics:
Not-(A and B and C and D and E)
In this case atheism seems more probable than theism when taken apart from our knowledge of the world. In isolation, the odds of five assertions being true is lower than that at least one of those assertions is false. In fact, in that context the odds of two assertions being true is lower than at least one of them being false. At least prior to empirical investigation, (A and B) is less likely true than not-(A and B). For example, the odds of a woman being both a lawyer and a feminist is less likely true than a woman not being both a lawyer and a feminist.
Atheists often say that they don’t believe in gods just like theists don’t believe in faeries. It seems plausible to at least say that the default position is that faeries don’t exist because it too requires multiple assertions (i.e. there is a person that is (1) very small and (2) can fly.) It seems rational for theists and atheists to assume faeries don’t exist (until good enough arguments are presented). It also seems unreasonable to demand that a person who rejects the existence of faeries prove that faeries don’t exist. There are hundreds of fictional types of creatures and it would be silly to think everyone has a duty to disprove that they all exist.
Some potential objections
Not everyone will be convinced by my argument, and I will discuss six objections that can be raised.
- How do we know the correct way to formulate these beliefs? I have suggested that the correct way to formulate the assertion that a god exists is as actually having multiple assertions. The mere fact that we can give a god a single name and say simply “Thor exists” does not mean that it really should only count as a single assertion to determine the odds of it being true. As was stated earlier, it’s important that we know the characteristics that we consider gods to have and how we can differentiate gods from other things. The characteristics that the argument states gods have is meant to help us understand why we should think of a being as a god and to differentiate it from non-god things.
Aren’t there many types of gods? Doesn’t the atheist have to deny that any of them exist? The argument that atheism is the default position only argues against one conception of god, but that conception is compatible with several different kinds of gods. Since the conception of god used by the argument is compatible with so many types of gods, it literally does tell us why we should reject several different kinds of gods. The gods it does not tell us to reject are assumed not to be gods at all. Perhaps the argument did not quite define gods properly by giving the correct list of characteristics, but the argument can easily be changed by defining gods by using a set of different characteristics instead.
Doesn’t the fact that the atheist must reject multiple types of gods increase the odds that theism is true? The atheist rejects everything that has every necessary godlike characteristic, but there’s perhaps infinite different gods must be rejected. Doesn’t that mean that the existence of at least one of the gods is more probable? I think not. The odds of a women existing is not more likely just because a women can exist and have long hair, and a woman can exist and have short hair, etc. Adding more characteristics to a list does not increase the probability of the entire list of characteristics actually describing something, even though there are many different ways we can add more characteristics to a list.
Wouldn’t the default position be that human beings and rocks don’t exist? At this point it is not clear that faeries or gods need to be singled out. I think we should agree that humans, dogs, rocks, and chairs are all unlikely to exist when taken in isolation of our knowledge of the world. The default is then that none of these things exist. Even so, we certainly can give good arguments that these things exist. The belief that humans, dogs, rocks, and chairs exist is highly justified when we consider our knowledge of the world. Additionally, no one will debate over whether humans, dogs, rocks, and chairs exist insofar as everyone already agrees that they do. When it’s time to actually debate shared assumptions are not required to be proven, at least for practical reasons.
What if the default position is that nothing exists? Everything that can exist has more than one characteristic, so the probability I discuss would indicate that the default position is that they don’t exist. This could be taken to be a reason to think that we shouldn’t believe anything about the world unless it’s proven, but in that case perhaps we can’t believe anything counts as evidence for a belief about the world at all. We couldn’t accept that anyone has experiences or observations that could be used as a reason to believe anything about the world. My response would be that what is discussed as a default position here is incomplete and certain assumptions will be required for any debate to exist. In fact, during actual debates anything both sides agree with will not need to be proven. Perhaps there is something “nonrational” about having assumptions, but it would be impractical to demand that no one ever has assumptions. It seems important that we can hypothesize about the world and test our hypotheses while tentatively assuming the hypothesis is true.
Two objections to default positions
Perhaps the strongest objection to the arguments for default positions is against default positions as I have defined them in general. Both of the arguments for default positions that I presented require us to accept that we can assess the odds of certain beliefs being true prior to any knowledge of the world, but perhaps arguments given in that situation about such probabilities are not meaningful after all. Consider the following two objections to default positions:
What good does a default position do us when we never argue in a bubble where we know nothing about the world? Perhaps the default position does indicate that those who argue against the default position has more hurdles to face in order to give the skeptic good reasons to think they don’t exist. Characteristics that are non-controversial (such as having a mind) should be accepted by atheists, but certain characteristics are controversial (such as having a mind without a body). It’s those characteristics that will actually be relevant to the debate over theism.
How can we discuss probabilities in isolation of our knowledge of the world? The probabilities I discuss are not set in stone. Sometimes “A and B” has the same probability as “B.” For example, when A has a 0% chance of being true. We need to know how we can actually compare the probabilities of various things in order to take them seriously. This is probably the strongest objection to the arguments, but I think probabilities can count for something if we can find a way to agree about what is likely true about probabilities. Perhaps the atheist should give the theist the benefit of the doubt and say that each controversial characteristic of a god can be initially assumed to be 99% probable until the relevant data is discussed. Giving the theist the benefit of the doubt would be a good reason to reject some of the probabilities that I gave, so the probabilities and default position would have to be re-evaluated.
I have presented an argument that theism is the default position, and I explained why I don’t think it is a good argument. I also presented an argument that atheism is the default position and considered five objections to it, but it is unclear that any of these objections are serious. However, there are also two objections I discussed to any probabilistic argument involving the default position, and I believe that the second of those objections is serious. I can’t think of a good way to respond to that objection, but perhaps arguments involving the default position can be better developed in the future.
If a good argument can be presented that atheism is the default position, then atheism could be said to be more probable than theism when considered in isolation from our knowledge of the world. This could be taken to be an a priori argument for atheism—an argument that does not require us to know anything about the world. However, this argument would merely be a reason to prefer atheism based on probability when taken in isolation of what we know about the world, so it would not be the final word. It would not absolutely prove that gods don’t exist. The implication would be that atheism can be rational until theism is proven to be more likely true.
The default position has little to no bearing once arguments are given. If there is a good enough argument to believe in gods, then atheists should change their mind. Until then we should expect atheists and theists to continue to debate. There can be arguments for and against the existence of gods, and those not persuaded by those arguments will provide objections to the arguments given by the opposing side.
Update (1/30/2013): I added three new objections to the argument that atheism is the default position. Other parts of the essay were changed to reflect my increasing doubt based on those objections.
Update (1/31/2013): I added more information about the first objection I raised against the argument that theism is the default position, I improved the discussion of the first two objections to the argument that atheism is the default position, I added the third discussed objection against the argument that atheism is the default position, and I made it clear that there are two objections to arguments about default positions in general.
- Do Default Positions Exist?
- What is the Burden of Proof?
- An Argument Against God, a Teapot, and Garvey’s Objection (Part 1)