(Cross-posted to my own blog.)
Sam Cole, one of the students in my upper level metaphysics class, wrote an interesting paper (I am writing this with his permission) where he argued that if we do not accept the Principle of Sufficient Reason (PSR), then the following question will be unanswerable:
- Under what circumstances should we accept a given explanatory hypothesis instead of the hypothesis that the phenomenon in question simply has no explanation?
I think this is a really neat question. We have some idea of the sorts of criteria we employ in choosing between alternate explanatory hypotheses: simplicity, prior probability (perhaps I repeat myself), etc. But if we do not accept the PSR, then the no-explanation hypothesis is going to be, presumably, always available. On what grounds do we judge between our best explanatory hypothesis and the no-explanation hypothesis?
It is tempting to say: If the best explanatory hypothesis is pretty good, then we go for it. But the evaluation of the quality of hypotheses seems to be innately comparative. So this "pretty good" does not seem like it should be absolute. But if it is relative, then what is it relative to? If it is relative to other explanatory hypotheses, then its being "pretty good" seems irrelevant when comparing it against the no-explanation hypothesis. The hypothesis that JFK was shot by a bunch of gorillas escaped from the zoo is pretty good as compared to the hypothesis that JFK was killed by a rifle-toting clam, but that is irrelevant when we compare the gorilla hypothesis to the Oswald hypothesis. So what we need to know is whether the explanatory hypothesis is "pretty good" as compared to the no-explanation hypothesis. But we have no criteria for that sort of comparison!
Another tempting suggestion is this: Whenever any narrowly logically coherent explanation has been offered (asking for more than that may run into Kripkean problems), we should reject the no-explanation hypothesis. This is a more promising answer to (1). Note, however, that an opponent of the PSR who takes this route cannot oppose the use of the PSR in the Cosmological Argument. For in the context of the Cosmological Argument, the PSR is employed to claim the existence of explanations for phenomena for which narrowly logically coherent explanations--namely, theistic ones--have indeed been offered.
1. Under what circumstances should we accept a given explanatory hypothesis instead of the hypothesis that the phenomenon in question simply has no explanation?
I'm asking this my friends who reject the PSR every time they say they don't believe it's true. But maybe a Boltzmann-Brain-style argument might be fun as well.
I think this is a much-neglected topic. If we assume that everyone has to live with some bruteness (Hudson is persuasive that our choices are bruteness, necessitarianism or Lewisian modal realism), where does it start? I guess I'm unmoved by what look like skeptical worries about comparative hypotheses. The problems you cite would just never arise, would they? That is, we are far too adept at advancing hypotheses to find ourselves with such bad (albeit, comparatively good) conjectures. But then I don't follow this claim,
If it is relative to other explanatory hypotheses, then its being "pretty good" seems irrelevant when comparing it against the no-explanation hypothesis.
Why is it irrelevant? Effectively, this says that any explanation with some credibility beats the there's-no-explanation hypothesis. I find that plausible, but that might be the rationalist in me.
I'm with Mike on this one. Time for some serious epistemology...
Here is a proposal, and I don't know how exactly your criticisms apply:
We should prefer an explanatory hypothesis over the hypothesis that the phenomenon is unexplained where the explanatory hypothesis is a simpler stopping point of explanation or fits in better with our background knowledge.
That is, we use the very criteria we use in deciding between rival explanatory hypothesis here.
Mike:
If an explanation with any credibility beats the no-explanation hypothesis, then we can still run a cosmological argument. So that's a pretty happy outcome for me. :-)
I agree: we're very adept at advancing explanations. In fact, one might conjecture that for any contingent state of affairs an explanation could be advanced that has some credibility.
T:
I think this is a promising proposal. I guess the way I'd run it would be rather like what Lewis says about laws: go for the optimal balance between ontological simplicity and explanatory power. We get lots of ontological simplicity by supposing that one is the only thing that exists, and that none of one's mental states have any causes. But we don't get any explanatory power. I guess how well this proposal works depends on whether there is any good measure of that balance. I have some doubts.
Alex, you write,
If an explanation with any credibility beats the no-explanation hypothesis, then we can still run a cosmological argument. So that's a pretty happy outcome for me. :-)
In fact, there is an interesting argument for a de facto weak PSR here that does not deny the possibility of brute facts. Let a de facto weak PSR be state that as a matter of contingent fact, every event has an explanation with some credibility. Roughed out,
1. Possibly, there are brute facts.
2. Necessarily, any explanatory hypothesis H for event e with some credibility is rationally preferable the no-explanation hypothesis H' for e.
3. For any actual event e, there is some explanatory hypothesis H for e with some credibility.
4. Every actual event e has an explanation with some credibility.
5. Therefore, PSR is at least contingently true.
This is a fascinating topic. I have a bad habit of teasing family members and friends when they ask something like, "Why did that happen?" I say, "perhaps there is simply no explanation whatsoever." My statement always strikes people as absurd. I wonder why it wouldn't be similarly absurd to say that a happening of all contingent things has no explanation.
I wonder why it wouldn't be similarly absurd to say that a happening of all contingent things has no explanation.
One worry is that the conjunction C of all contingent things might not be a happening. But that's a small point. Absurdity is a pretty strong charge. But, as I say above, it does seem more reasonable (to me anyway) to prefer an explanation H for C that has some credibility (i.e., some non-infinitesimal positive probability) than the "non-explanation" hypothesis. This entails that it's always rational to believe that the no-explanation hypothesis has near-zero credibility. I think that's right; I'm not sure how controversial it is.
By 'a happening of all contingent things' I had in mind the property being contingent coming to be exemplified. If that's possible, and if there would be an explanation of that happening, then...
No, I think I know what you mean. Of course, I happen to instantiate 'being contingent' and I'm pretty sure I'm not a happening. It just doesn't feel like I'm a happening. Maybe Whitehead would tell me I'm a happening. That would make Whitehead a happened, I guess? . . .:)
Alex,
I have a question about the conculsion of the cosmological argument that's based on the PSR.
When we ask "Why is there something rather than nothing?" and answer with "Something exists necessarily", that might work fine. Then we argue that the universe is contingent - and that it's explained by a necessary deity. One might argue "We have no reason for believing that the universe
is not contingent.", but the nonbeliever could simply reply "Ok, but saying the universe is necessary takes on less explanation than what you believe. We would have to ask: Why is there God rather than nothing?". The answer is obvious: because is God is a necessary being.
And now, you have to run an ontological argument to support your cosmological argument.
So isn't it the case that the cosmological argument alone is quite useless (unless, like me, you are convinced that it's completely obvious that the universe is contingent)?
While I do believe that the cosmological argument from contingency and the modal ontological argument taken together can be quite powerful, one should always keep in mind what you really have to achieve.
But here, you find a new problem for the LCA. Every objection towards the ontological argument could also be used against the cosmological argument based on the PSR.
One of the more interesting is that one might say "No being can exist necessarily."
How do you solve this problem? I just did a search for ontological arguments and found this entry from you:
http://prosblogion.ektopos.com/archives/2007/11/a-variant-on-th.html
I like the elegance of the argument and I would say that Axioms 1,2 and 4 are both obviously true and true by definition. The trouble is Axiom 3. Is the Lemmas demonstrate, all you would have to show to defeat the objection that there can't be necessary concrete objects, is that "Necessary existence is a positive property" or "a perfection".
Now, I believe Norman Malcolm had the following argument:
(1) It's possible that a perfect being exists.
(2) If a perfect being exists, it could not have failed to exist (Anselm's principle)
(1) was considered to be begging the question. But now it's (2).
And it seems to me that in order for this cosmological argument to be succesful, you would also have to defend (2).
So, how do you think that can be achieved? By arguing something like "omnipotence/omnipresence means acting in every possible world"?
. . . I would say that Axioms 1,2 and 4 are both obviously true and true by definition
Axiom 4 is certainly not obviously true (and less so, definitionally). Would that it were! In fact, it begs the question against those advancing incompossibility arguments: indeed, this is the very focus of the debate on ontological arguments.
Mike,
I'm puzzled how one would object to Axiom 4. Here's why I believe it's obvious and true by definition:
Axiom 4: God-likeness is a positive property.
Definition 3. God-likeness is the property of having all strongly positive properties.
Axiom 4*: The property of having all strongly positive properties is a positive property.
Maydole argues in Philo 2003 that something is supreme (God-like) "if and only if it is necessarily greater than everything else by virtue of its having some set of perfections or great-making properties K."
The intersection of the extensions of the members of K should be the extension of a perfection. Therefore, the extension of Supremity is identical with the intersection of the extensions of the members of K. Hence, Supremity (God-likeness) is a Perfection.
Now, you brought up the question of compatibility.
But that doesn't bother me, because of Axiom 1 and 2.
What do you think about the following argument:
(1) If two properties are incompatible, one of them is the negation of the other or entails the negation of the other.
(2) A positive property is not the negation of a positive property (Axiom 1)
(3) A positive property does not entail the negation of a positive property (Axiom 1 and 2)
(4) Therefore, there are no two positive properties that are incompatible.
One might of course question which properties really are positive. But the argument can easily establish the existence of a god-like being. It is then up to us to show that this being is the God of theism.
You seem to be less concerned about Axiom 3. Can you recommend a paper defending it?
The problem with this argument is that it pretends like a "no-explanation hypothesis" is a vague and unrealistic thing, when in fact they are quite common in science.
To take the obvious example from quantum physics, The no explanation hypothesis for the state of Schrodinger's cat, is obviously a better explanation than the explanatory hypotheses that angels are holding a gun to the cats head and will decide whether or not to kill it based on the if it rains or not in Bali.
Under what circumstances should we accept a given explanatory hypothesis instead of the hypothesis that the phenomenon in question simply has no explanation?
So to answer the question we should accept a no-explanation hypothesis when in it is logically and empirically proper the context of a holistic theory, such as uncertainty principle in quantum physics.
The argument is simple. Given axioms 1 and 2 every possible property is possibly exemplified. So conceding axiom 4 amounts to conceding that the divine attributes are compossible. But their compossibility is exactly what's in question. That's exactly what atheists deny. In this particular case, they would simply deny that the conjunction of such properties is a positive property.
One might of course question which properties really are positive. But the argument can easily establish the existence of a god-like being.
No, this is not easily established by any argument. It depends entirely on what sorts of properties are compossible.
So conceding axiom 4 amounts to conceding that the divine attributes are compossible. But their compossibility is exactly what's in question. That's exactly what atheists deny. In this particular case, they would simply deny that the conjunction of such properties is a positive property.
Wouldn't it be sufficient to demonstrate that all divine attributes are positive?
No, this is not easily established by any argument. It depends entirely on what sorts of properties are compossible.
You didn't seem to disagree with my argument for all positive properties being compossible. And as I pointed out, Axiom 4 can be re-formulated as saying "The conjunction of positive properties is always positive" (which would include all strongly positive properties), which is definitely not begging the question.
So, given Axiom 3, a being that has all strongly-positive properties should exist. What is at question is "What are those strongly positive properties?"
Pumbelo:
There are three types of "objections to the ontological argument": (1) objections to the argument itself, and (2) objections to the argument's conclusion. If the cosmological argument has the same conclusion as the ontological, then type 2 objections apply. But type 2 objections to the existence of a necessary being are not very convincing. The most famous is a paper by Findlay (I've heard that even he eventually stopped being convinced of that paper). Anyway, so the Leibnizian cosmological argument does not require an ontological argument--it simply entails the necessary-being part of the conclusion of the ontological argument.
afu:
It seems that a reasonable thing to say in cases like the quantum ones is that the phenomenon has no explanation within the bounds of the theory. But quantum theory is quite compatible with non-physical determination.
You didn't seem to disagree with my argument for all positive properties being compossible.
What's to disagree with? Given the assumption of ax. 1 and ax. 2 this obviously follows. Alex shows this is detail in lemma 1 (or maybe 2). But it doesn't show anything interesting. What we're interested in is the compossibility of the traditional attributes of God. If you have an argument that the traditional attributes are compossible, I'd be happy to hear it.
Mike,
as I said before: It would be sufficient to show that every single one of them is positive, because if they are positive, they are compossible.
But in the context of the argument, we don't have to show that they are positive. We are not using the God of monotheism to define what properties are strongly positive, we use strongly positive properties to define God.
Definition 3 states "God-likeness means having all strongly positive properties". So every property of a God-like being is positive. Their conjunction is possible and positive itself. So God-likeness is possibly exemplified. And by Axiom 3, "God" exists.
The problems arise here:
1) Is necessary existence really positive?
2) What God are we talking about? And this is how your objection should be formulated.
You are absolutely correct when you say: "We have to show that thy are compossible."
But if there is even 1 strongly positive property (different from necessary existence), the argument actually works (assuming that something can exist necessarily, 1 property shouldn't face any compossibility-problems).
Just use a revised version of it:
Axiom 1. If a property is positive, its negation is not positive.
Axiom 2. If P is a positive property, and P entails Q, then Q is a positive property.
Axiom 3. Necessary existence is a positive property.
Axiom 4. The property of having all strongly positive properties. is a positive property.
Axiom 5. Z is a strongly positive property.
Therefore, Z is necessarily exemplified.
I'm especially surprised that this comes from you, since I'm actually borrowing from your "Irrefutable Ontological Argument"!
As you said: "I think once we have some necessarily existing being with these properties, we are well on our way to establishing that there exists a religiously important being."
And actually, all we have to do in order to show that a religiously important being exists is that some of the significant properties of the monotheistic God (power, knowledge, goodness) are compossible or strongly positive.
If I would insert "maximal power" for Z in Axiom 5, how strongly does it beg the question?
Pumblo, you write,
And actually, all we have to do in order to show that a religiously important being exists is that some of the significant properties of the monotheistic God (power, knowledge, goodness) are compossible or strongly positive..
What do you mean "all we have to do"? No small task! Indeed, this is the hard part. The rest is detail! This is why I offer an argument for why we might expect there to be such a set of interesting compossible properties. I don't stipulate that there are some that are interesting and I don't define 'positive property' in a way that guarantees that they are possibly instantiated. This is just misleading, since (for all we know) the positive properties might turn out to be a pretty nasty set of properties.
So we have to offer an argument for why there is an interesting set fo compossible properties. I offer a probabilistic argument. I don't know any other way to go on this question.
Mike,
you might be correct that showing "love", "wisdom", "power", "goodness" are positive properties can be difficult (actually, I would use "perfection" instead of positive).
You said that atheists might deny compossibility of love/wisdom/power/goodness with necessary existence. But how many atheists are going to deny:
(1) "Love" is a perfection.
(1*) "Wisdom" is a perfection.
(1**) "Power" is a perfection.
(1***) "Goodness" is a perfection.
I do believe that they are quite modest. And even if you are correct in saying "Proving any of those is hard", how many atheists will say that all of those examples I offered are implausible?
I believe that there are very few if them, which is why I believe Axiom 3 is more problematic.
I guess you have in mind essential perfect love, essential perfect wisdom and so on. If showing that any one of these is a perfection (in some sense) is distinct from showing that it's a positive property, then I predict you'll get some agreement. I also predict you won't get any agreement where it really matters: i.e., where agreeing that any one of these properties has a property (positivity, for instance) that entails that it's compossible with other perfections.
Mike,
maybe one should simply run a hybrid-argument, simply borrowing from Alex and Maydole.
Definition 1: A being is supreme if no being could possibly greater.
Definition 2: A perfection is a property greater to have than not.
(1) If a property is a perfection, it's negation is not a percection (Premise)
(2) If a property X is a perfection and X entails Y, Y is a perfection (Premise)
(3) Being supreme is a perfection (Premise)
(4) Necessary existence is a perfection (Premise)
(5) A supreme being necessarily exists (from A-D)
Maydole uses almost the same argument In Philo 2003 and the Blackwell Companion ot Natural Theology, except that he doesn't use (D) but the Barcan Formula. Since I have problems with the Barcan Formula, my main trouble is (4).
re premise (2)-every perfection entails being self-identical, being green or non-green and being colored if red. are these perfections?
Pumb,
I'm repeating myself, but (1), (2) and (3) entail that the perfections are compossible. Two possibilities:
a. If by perfections you mean the traditional attributes of God, then these premises beg the question.
b. If by perfections you mean anything that satisfies (1)-(3), then there's no reason to believe that this proof shows that the traditional perfect being exists.
It looks like you mean (a), given your definitions. Given that, the atheist will just deny that perfections satisfy your initial premises/assumptions.
overseas,
yes, all necessary conditions are greater to have than not.
Mike,
I think you mean that (1) and (2) entail that all perfections are compossible.
I think it's a mix between (a) and (b)
By "a perfection", I mean "a property that it greater to have than not" (this is how Maydole formulates it in hs article on the Modal Perfection Argument in Philo). And I do believe this satisfies (1)-(4).
You are quite right that I believe all the traditional attributes of the monotheistic God are perfections. But the definition of "being supreme" in this argument is not "being the traditional monotheistic God", but "having maximal greatness". How we interpret the argument rests on our perception of "maximal greatness".
You bring up compatibility again.
- First, let me give you a primitve argument for why we should think God's traditional attributes are perfections:
(1) The concept of God is made by humans.
(2) If the concept of God is made by humans, all of God's attributes were chosen by humans.
(4) If all of God's attributes were chosen by humans, they would have chosen attributes, they would have chosen attributes that were considered to be great-making.
(5) The human perception of what a great-making property is is quite reliable.
(6) We have good reasons for thinking God's properties are perfections (from 1-5)
- Second, not all non-theists will claim that God's attributes are incompatible. So some people might actually be persuaded by this type of argument.
- Third, how many people are going to deny /how implausible is it that power, wisdom and love are greater to have than not? The argument is not intented as a full proof, but rather to justify belief in a supreme being. And to me, it just makes sense to interpret a supreme being as some sort of God.
- Finally, I don't think incompatibility arguments are really convincing. Divine attributes can be reformulated/limited and those arguments almost never do any better than "Hey Flanders, can Jesus microwave a Burrito so hot that he can't eat it?"
And I'm having the impression that the overwhelming majority of arguments against the existence of God that are discussed in contemporary philosophy-journals are arguments from evil. Quentin Smith's cosmological arguments are a rare exception, but that's basically it.
Alex, how do you think about this:
"Explanations only hold within spacetime."
(1) The concept of God is made by humans.
(2) If the concept of God is made by humans, all of God's attributes were chosen by humans.
No theist can take the position that the concept of God is "made" or that the attributes of God were "chosen". We discover what properties God has essentially, we don't invent or make up what God is essentially. When we say (at least when I say) that God is omnipotent, I'm not talking about a concept. I'm talking about God. When I insist that it's part of the concept of God that he is essentially perfectly good, I'm saying that it is impossible that God (not his concept) should fail to be essentially perfectly good. This is the long way of saying that I deny both (1) and (2) in this argument.
I was wondering if you could use some sort of argument similar to Plantinga's EAAN against people who deny the PSR. I recall that in your Blackwell article you mentioned a possible argument that could be charged against people who believe naturalistic evolution, yet deny the PSR. If it's true one cannot differintiate between the no-explanation hypothesis for the origin of man, and other hypotheses, what would the probability of us having reliable cognitive facutlies be? If it's inscrutable whether there is an explanation for our existence or not, wouldn't the probability of us having reliable cognitive faculties be inscrutable. I think it would be really neat if someone could develop an argument that PSR-deniers have a defeater for their cognite faculties being reliable.