Sobel on Presumptive Atheism

I think two claims should be distinguished:

(1) The absence of evidence for the existence of X doesn't give us any reason to think X doesn't exist.

(2) The absence of evidence for the existence of X doesn't give us any reason to think X doesn't exist even when we would expect there to be evidence for the existence of X.

(2) is false. But I think (1) is true -- the mere absence of evidence for the existence of X doesn't give us reason to think there's no X. I suspect that if you find (1) implausible, it's because you're focusing either on Xs which (2) applies to, or you're thinking of Xs which there really is evidence against (e.g., Santa Claus, or the existence of a perfectly round stone with a 10" diameter at some specific location on the moon). Whether (2) is true when X=God strikes me as a good question, which I don't know the answer to. If it doesn't, then I say that a lack of evidence for the existence of God doesn't by itself give us any reason to think God doesn't exist.

"In the absence of good reasons to believe that God exists, it seems odd to say that I cannot rationally believe that he doesn’t."

I suppose one might have some theory about the function and history of our cognitive faculties on which it is not odd to say that one cannot rationally believe that God does not exist in the absence of good reasons. It would be odd to make the claim with a quadruple negative though.

"why is it more difficult to show that something instantiates all of the relevant properties than it is to show that some subset of those properties is inconsistent?"

Is Sobel claiming that the atheist would have to show that a subset of the properties is inconsistent, or merely that all the atheist has to show is that the properties in the subset are not instantiated or coinstantiated?

Hey gang, your blog is just awesome ! Thanks for this post Mike.

Why is there any more presumption against the god of love than there is against most other gods? I guess I don’t see why it follows from the assumption that G is a god of love, that G is any more likely to make itself known to us than the god of peace or justice would.

Hi Mike,

I guess I find it plausible to think that a god of love would be at least somewhat more likely to reveal itself than a god of (merely) justice.

A god of justice would desire a really just world; but it might be able to get this without revealing itself to the inhabitants. But we might think that love requires personal interaction or involvement or relationship (or something like this); so that a maximally loving god who creates a world would be strongly disposed to make such personal relationship between itself and creatures possible by revealing itself in certain ways.

Mike:

So, let me probe your intuitions. Say that x is an aeene iff the absence of evidence for an x is evidence of non-existence. Are the following aeenes:
- mammal with prehensile nose
- mammal with prehensile ears
- mammal with four legs
- alien which naturally has organic rotating wheels as its mode of locomation
- alien which has wheels, wings, and breathes fire
- persons other than oneself
- a maximally evil being
- a maximally good being
- a maximally great being

Absence of evidence is evidence of absence IF the object in question would have an impact if it was real. For example, there are certainly no dragons on Earth because we'd notice them controlling the weather, shaking down peasents/eating maidens, hording gold.

"- alien which naturally has organic rotating wheels as its mode of locomation"

The problem with this is simple- how does it get blood to its wheels?

"alien which has wheels, wings, and breathes fire"

I don't think fire breathing a a possible evolutionary trait due to the whole burning your throat raw for starters.

"a maximally evil being"
"a maximally good being"
"a maximally great being"

What do you mean by these? If you mean an individual who always seeks to maximize the amount of good they do than the second could exist. If you mean an individual who always seeks to maximize their own glory, than the third would be Napolean. If you mean an individual who seeks to cause as much evil as possible... I can't think of any.

Mike,

My point above was put confusingly. Let me say it like I should have.

(1) and (2) are universal claims. (2) is false, I claim (1) is true. The reason it's true is that the MERE ABSENCE OF EVIDENCE FOR X provides no reason to think that X doesn't exist, UNLESS WE WOULD EXPECT THERE TO BE EVIDENCE FOR X IF X DID EXIST. (That's also what Samuel Skinner's first sentence amounts to, I think.)

Now, here's an interesting question: if there was no evidence for the existence of God, would that fact by itself give us a reason to think God doesn't exist. I say no. On the other hand, maybe if God did exist, we would expect there to be evidence for His existence. If that's the case, then of course the absence of evidence for God's existence DOES provide us with evidence that God doesn't exist. But then we believe that God doesn't exist NOT because there's a presumption against his existence, but because the hypothesis that God exists predicts certain evidence to be there, and it isn't there, just as our inability to find a white raven provides us with evidence that there aren't any.

So Mike, I'm not sure what you're saying. Are you saying "God's an extraordinary being. There's a presumption against the existence of extraordinary beings. It's not that extraordinary beings are such that if they existed we'd expect to find evidence for them. There's a presumption against their existence that doesn't reduce to that. [Presumably this is an a priori presumption against them.]"? Or are you saying "God's an extraordinary being. So if he existed, we'd expect to find evidence that he existed. So if we find no evidence that he exists, we have good reason to think he doesn't."?

You bring up the example of dragons. The thing is, that we definitely would expect to find evidence for their existence if they existed. But we don't. We also have pretty direct evidence that they don't exist. We've never discovered any animal that breathes fire, for instance. And given what we know about animals, it's very unlikely that there has ever been an animal with such a capacity. In short, we don't need to appeal to an a priori presumption against the existence of extraordinary beings to have good evidence that there are no dragons.

Now here's a question: If God existed, would we expect to find evidence that he existed? Not sure. I suspect it depends on what properties we ascribe to God.

Elliot Sober's new paper "Absence of Evidence and Evidence of Absence:
Evidential Transitivity in connection with Fossils, Fishing, Fine-Tuning, and Firing Squads" might be of interest and relevance here, and is available online at:

http://philosophy.wisc.edu/sober/Absence%20of%20Evidence%20and%20Ev%20of%20Abs%20dec%201%202008.pdf

We seem now to have a hidden God, or a God that is not present in obvious ways. That does not prevent large numbers of theists--almost all, really--reporting that they have a more or less strong, personal relationship with God.

OK, but hiddenness also (you might think) inhibits a large number from believing in God; many just don't see or feel God in any real way, where it's also plausible to think that if they did, they'd respond positively.

I don't think hiddennes is a particularly worrying problem; but I do think there's something to it. And it seems to me that hiddenness would be at least slightly less of a problem, slightly more expected, if God isn't a God of love. But maybe you don't think hiddenness is a problem at all.

"Now here's a question: If God existed, would we expect to find evidence that he existed? Not sure. I suspect it depends on what properties we ascribe to God."

Yes we would. Went through all the trouble to make the universe, didn't he?

"I don't see it. We seem now to have a hidden God, or a God that is not present in obvious ways. That does not prevent large numbers of theists--almost all, really--reporting that they have a more or less strong, personal relationship with God."

That is a new trend. For most of history, people didn't think of religion as having a relationship with Jesus.

"Right, we can tell a story that links openness to a god of love. But I've never found the stories convincing. There are other stories we can tell that link hiddenness to the god of love. The call to intimacy with god, for instance, might demand a slow approach to god. I don't find that wildly implausible. So, what I'm underscoring is that the logical link between openness and the god of love isn't obvious. We need some story about how one demands the other. But there are no convincing stories to tell here."

A God of love? Isn't that Venus?

Mike, you say: "Change the description to an utterly reclusive quasi-dragon. A quasi-dragon that's so reclusive that it has never been observed. Is such a being possible? Yes. Would we see it, were it actual? No. Can I reasonably believe that such a being doesn't exist? I say yes."

-There's tons of evidence that a being like that doesn't exist! See my comments above about why we have reason to think there are no dragons. In order to account for why such beings don't exist, we don't have to posit an a priori presumption against them.

Mike:

Is it simplicity considerations that drive your intuitions?

While positing mammals with prehensile ears would not modify us to go beyond the explanatory schemata we have (we can easily give an evolutionary story, etc.), to posit reclusive dragons would. And we want to keep the number of explanatory schemata low.

But that won't apply to the case of a necessarily existing God if one already believes in some necessary beings, like numbers. For the same explanatory scheme would apply (exists because exists necessarily, say).

I find myself drawn to the following intuition: The world contains many extraordinary things. Platypuses, quasars, black holes, Osedax worms, etc. It would be surprising if we were done.

So, we should expect if there were extraordinary things we do not have any evidence for.

But it does not, of course, follow from this that for any particular extraordinary kind of thing, we should expect it to exist. But the consideration might weaken our scepticism.

I meant that we should expect that there are extraordinary things.

Maybe one could argue for your view as follows. The number of existing kinds of extraordinary things is likely to be small. (I don't know how one argues for this. Maybe this is a consequence of the orderliness of the world--if there is too much of the extraordinary, the world is not orderly?) But the number of kinds of extraordinary things is infinite. So for any particular kind of extraordinary thing, the probability of that kind of thing existing, absent further evidence, is very small.

But I think the argument that the number of existing kinds of extraordinary things is small is weak. We have little reason to deny that the world is infinite. And if it's infinite, it would be likely that there would be infinitely many extraordinary things.

OK, so these animals are supposed to be such that they're so reclusive, or good at hiding, or whatever, that we would expect there to be no evidence at all that they existed, if they existed. I think we do have evidence that there are no animals like that. We no animals must eat. So the animals must consume stuff in their environment. In doing so, they'll leave traces behind, which is how we could get evidence of them. On the other hand, describe say a fish that only lives in very deep waters, where we essentially never go. And describe that fish in such a way that, given what we know about fish and those locales, it's plausible that a fish like that would be well adapted to and could have evolved to live in such a location. Go ahead and make it an extraordinary fish, but make sure you satisfy the condition in the previous sentence (otherwise we will have evidence against it). Now, does our lack of evidence that it exists give us reason to think it doesn't. I feel no inclination to say yes.

The aliens on Alpha Centauri is a better example. Go ahead and make your aliens as weird as you want. But we do know a few things about living things. So don't give them properties that given this knowledge, it's improbable that beings like that exist. Plausibly, if there were such aliens, there would be no evidence that they existed. Now, should we be inclined to think they don't exist, just given that there's no evidence they do? I say no -- the lack of evidence gives us no reason at all to think they don't exist. And I'll say that even if you make them really weird. If anything, I'd kind of expect them to be weird/extraordinary.

Hello. Most of this isn't going to be tremendously relevant, and I apologize for that, but I had some PoR-related questions that I've had a hard time tracking down answers to. I did philosophy as an undergrad, and I emailed an old professor of mine, but it's not his field and he wasn't able to help with most of it.

Briefly, is there good work on whether or not it makes sense to say that "reasonable people can disagree" (on religion in particular, but also more generally)? As well, is there much focus in modern PoR on defending particular religious beliefs? I remember reading a lot of older stuff as an undergrad, but it was mostly pro-/anti- generic religion. I think the distinction between a god of love and a god of justice is the kind of thing I'm interested in here. I'm also curious as to the demographics of PoR departments in particular and philosophy departments in general - I gather that PoR skews theist while philosophy departments skew atheist, but is there anything more specific floating around (I've been completely unable to find this from APA surveys and the like)? I don't mean to take up much of anyone's time - a pointer to an author or a good search term to look through journals with is all I really need. Thanks for any help with any of this.

To the topic: my (unqualified) take on the Alpha Centauri aliens is that, if we have some significant amount of uncertainty as to whether or not they exist, don't we have to conclude that some hidden aliens are actually watching us? If we can specify aliens from AC as things that we can't rule out, then we can also specify aliens from other planets as things we can't rule out. We can come up with a list of thousands of stars that reclusive peeping aliens might come from. And if we can't write off any single one, and if the likelihood of one is more or less independent of the likelihood of others, then doesn't it become probable that there's at least one actually here? More generally, I think the problem is that, as long as we wouldn't expect evidence of a thing's existence, we can get ridiculously specific (different colors of aliens from AC, aliens from AC with different hats, etc).

I should clarify that I'm operating under some possibly naive epistemological assumptions - I read an inability to presume something's nonexistence as saying that we assign enough of a likelihood to it existing that we can't believe that it doesn't. And if this is some finite likelihood that it exists, then, eventually, it becomes very probable that one of a large collection of such things does exist. But the specificity allowed seems to require that we say that pretty much every category of thing that wouldn't leave evidence if it existed does actually exist.

It seems to me that the likelihood of something existing for which we have no evidence is inversely proportional to the number of attributes we assign to that entity. For instance, consider the proposition that there is an as yet undiscovered species of mammal living in the jungles of Borneo. This proposition is certain plausible, perhaps evenly likely to be true. Now, suppose that I contend that this mammal is a flightless bat that forages for insects with its vestigial wings; that it is black with two white stripes down its back; and that the female of the species is more than four times larger than the male. None of these proposed attributes seem impossible or contradictory to what we know about existing mammals, but obviously the more such specific attributes we assign to this creature the less likely it is that such a thing exists.

For this reason I would have to agree at least with Sobel's reasoning if not with the details of his argument. The likelihood for the existence of a God with minimally defined attributes (say, an eternally existing being who is the cause of the universe coming into existence) should be considered greater than the likelihood that the God of the bible, with all his proposed attributes, exists. But I'm not sure I'd agree with where we should draw the line as regards to burden of proof. I also do not agree that we have no evidence for the minimal definition of God; rather, I would say the existence of the universe is in itself evidence.

Mike, thanks very much. That paper was very helpful.

Guess I misunderstood. I didn't realize the Alpha Centaurians were supposed to be living here on earth. Now that I understand that, I'll change my response. We have evidence there aren't any here. We realize how difficult it would be to travel that far and then arrive without anyone noticing. This decreases the probability that they're here.

"You don't want it to be in general true that the hypothesis that X exists is less probable than the hypothesis Y exists simply because X has more properties than Y."

I think you misunderstood what I was saying. I was only making the rather obvious point that the more specifically we try to describe the unknown, the more likely in general we are to be wrong about it. It seems indisputable that there are things in the universe that exist that we have no evidence for. It is contestable but still accepted by many that some of what exists is in the supernatural realm (defined here as something other than the matter and energy contained within space-time). But the more we try to pin down the possible nature of these things (without the evidence to back it up), the more likely we are to be wrong. This seems to me to be Sobel's point. He seems to think that the existence of some kind of deity, even without evidence, is likely enough that the burden of proof is on those who deny the concept. But once we try to pin down the nature of the deity (e.g. as a god of love), the burden of proof shifts, in his view, because we have gotten too specific in our description. As I said, I don't agree with the details of his argument, but I do agree with the logic.