Cosmological arguments in general appeal to some version of the Principle of Sufficient Reason. Rowe’s version of PSR is the following:
PSR. There must be an explanation for the existence of any being whatsoever, and there must be an explanation for every positive fact.
Now the cosmological argument is supposed to provide us with good reason for believing that there is a necessarily existing being whose existence—as what Rowe calls a ‘self-existent’ being— is explained by it’s necessity and which is the ultimate explanation for contingent events, states of affairs, etc. in the world. Peter Forrest’s ‘anthropic theism’ has me wondering why PSR requires that the ultimate explanation be a necessarily existing being. It would be interesting if there were another way to go. Peter does not consider the question, but why couldn’t PSR be satisfied by appeal to the intrinsic value of God, rather than the necessary existence of God?
So, my question concerns whether P is true.
P. If X exists and X is perfectly valuable, then X’s existence is sufficiently explained by the fact that X is perfectly valuable.
I do not mean to suggest that X could not fail to exist. It could, for sure, since X is a contingent being of a special sort. But were X to fail to exist, it would be less than perfectly valuable. How much value could a non-existent thing have? So, if you point to X, so to speak, and ask why X exists, I could completely answer your question by pointing out that X is perfectly valuable and noting P1.
P1. Necessarily, for all x, if x does not exist, then x is not perfectly valuable.
If X is perfectly valuable, then by universal instantiation on P1 and one application modus tollens on the strict conditional, we conclude that X must (wide scope) exist. Caution: P1 does not entail that there exist any perfectly valuable things at all. I’m happy to agree that in some worlds there are none (well, happy to agree for the sake of this discussion). But if we suppose that’s right, then it seems like we could have a cosmological argument whose ultimate explanation is a contingently existing, perfectly valuable thing. We could then explain the existence of other created objects in terms of moral value. In Forrest’s God without the Supernatural (sort of misleading title, as it happens), he argues that God creates the universe for moral reasons: because he is aware of the value of such a creation or as a spontaneous manifestation of God’s joy (cf. 45-56). These are steps in the direction of what might be called an axiological cosmological argument.
I suppose questions about why X is perfectly valuable in some worlds but not others would come up. It seems to that perhaps if X is perfectly valuable in any world, then X is perfectly valuable in every world. And, thus, the explanation would appeal to a logically necessary being, i.e., a being that exists in every possible world, instead of merely a quasi-logically necessary being, i.e., a beings who exists of logical necessity in only some possible worlds.
The notion of a quasi-logically necessary being seems to me confused. In any case, I don't appeal to it. I do not claim that X has the property of being perfectly valuable contingently. I claim that X exists contingently. So, there is no need to explain why X has the property of being perfectly valuable. But even if it did have that property contingently, (i) cosmological arguments needn't aspire to explain every contingent property that every thing instantiates, since some extremely improbable facts are simply the result of chance processes and (ii) we should expect lots of dissimilar contingent facts are compatible with divine purposes. It would not have affected God's purposes at all, I expect, had he created me a millimeter taller. So I might have had a lots of properties that I do not instantiate, and that is not incompatible with PSR.
I don't want to derail the conversation before it even really gets going, but I'd like to explain briefly what I have in mind by quasi-logically necessary being. In logically necessary being, its a property the being has that explains why it exists in the possible worlds that it does. But what if we get Copernican and switch it? What if we think of a property that the possible world has that explains why a particular being exists in it. In other words, suppose possible world W has property P such that P logically entails the existence of being X in W. So, I suppose I was (mis)reading your argument to suggest that possible world W has the property of having X be perfectly valuable to it and that was the explanation of why X exists in W.
If what I've said will derail the conversation that you wanted to have then please feel free to ignore it and pursue more fruitful tracks.
So, I suppose I was (mis)reading your argument to suggest that possible world W has the property of having X be perfectly valuable to it and that was the explanation of why X exists in W.
What explains X existence is the fact that X instantiates F and having F entails existing. So, I don't appeal to properties of worlds in my explanation for why X exists.
John Leslie has an interesting argument to the effect that the existence of the universe is explained by it being better that the universe exist than that it not exist. There are necessary moral truths and these truths can have explanatory power (So says Leslie). Is this the same sort of thing.
Hi Gordon,
I'm only vaguely familiar with Leslie's cosmological argument, but I think he rejects the existence of a theistic/personal God. The argument in the post is for a personal, contingent God. But there is the similarity that both arguments appeal to moral value in the explanation of what exists.
Mike,
Clearly I am missing something. I would endorse
P2. Necessarily, if x does not exist, then x is not even a little bit valuable
on the basis that if something does not exist, it has no (non-modal) properties at all. Likewise,
P3. Necessarily, if x does not exist, then x is not green.
Now, by instantiating these principles for x=my corroded copper penny, we get
P2*. Necessarily, if my corroded copper penny does not exist, then it is not even a little bit valuable.
P3*. Necessarily, if my corroded copper penny does not exist, then it is not green.
Those are true. And we can add the facts that my corroded copper penny is both green and a little bit valuable. But from these contingent facts, with P2* and P3*, nothing necessary follows. What follows is “my corroded copper penny exists.” And while this is a proof of my penny’s existence, I would not think it is much of an explanation of my penny’s existence.
What am I missing?
Hi Heath,
That's a nice point. What distiguishes your case from the one I'm advancing is that, I claim, existence is part of the concept of being perfectly valuable. Existence is not part of the concept of being copper. The fact that X instantiates a property that a priori entails existence is the source of the explanation. Does that help?
Mike,
I was missing the idea that perfect value included existence as part of the concept. OK, let it be so. But now I have a different objection.
Borrowing a page from PvI, I define an “exmount” as an existing mountain. If something does not exist, it is not an exmount, and we know this apriori because existence is part of the concept of being an exmount. Now suppose you ask me to explain the existence of Mt. Everest and I reply that it is an exmount, as both of us can clearly see. Are you satisfied with my explanation?
Heath,
Good. First, I hope PvI footnoted Rowe, since exmounts are just Rowe's magicans with another name! I'd say two things: first, I'm not sure that there is a concept of mountain that entails existence. It can look like a concept is coherent, and so such that it's possibly instantiated, and not be so. Perhaps the most famous is (again Rowe's) concept of being in less than perfect company, L. Let something instantiate L just in case every existing thing suffers from some flaw. Is the concept coherent? It depends on whether something like the Anselmian God exists. If so, the concept expressed by L is incoherent: nothing could instantiate it. It looks like L is perfectly coherent, but it might not be. Similarly for exmounts.
But set that aside. Second I'd say that if there is such a concept and X instantiates it, then I'd be satisfied with the explanation that X exists because it is an exmount and being an exmount apriori entails existing. I'd explain my hesitance in accepting this, all in, by my uncertainty about the coherence of the concept.
Mike:
I don't see a reason to believe P unless one has reason to accept some sort of deep metaphysical principle that makes things that are perfectly good exist. This principle could be necessitarian or probabilistic. If the former, then we end up with the necessary existence of something perfectly good (but this is a bit weaker than in the standard cosmological argument, because it is not shown that in every world it is the same perfectly good being). If the latter, we have a weird view on which a principle leads to things existing with different probabilities.
I don't see a reason to believe P unless one has reason to accept some sort of deep metaphysical principle that makes things that are perfectly good exist.
Alex,
That's seems strange. Would the world be perfectly good in virtue of a perfectly good, non-existing Meinongian object? (P) does not commit us to the position that the conceivability of a perfectly valuable object entails that the object exists any more than the conceivability of a Magican entails that there are existing magicians. It commits us to the claim that the property of a perfectly valuable thing includes (if you'll let me talk this way) the property of existing.
I'm actually after something fairly simple here. It is apriori that the property red includes (in some sense) the property of being colored. This does not entail that anything is red. Similarly, it is apriori that the proeprty of being perfectly valuable includes the property of existing. This does not mean that anything is perfectly valuable.
Mike,
This is interesting.
One worry (which perhaps has already been expressed): can there be a perfectly valuable, contingent thing? If existence is a value contributing property, wouldn't necessary existence also be a value-contributing property? (Isn't necessary existence equally part of the concept of perfect value?)
Another worry: the fact that something is perfectly valuable doesn't seem to me to explain why that thing exists unless perhaps an additional fact is included in the explanation. For example:
(1) x exists because x is perfectly valuable and it is necessary that there be a perfectly valuable being.
Or:
(2) x exists because x is perfectly valuable and it is probable that there be a perfectly valuable being.
But then see Alex's point above.
Now you offer the following principle to motivate the idea that one may explain a thing's existence by virtue of the fact that it is perfectly valuable:
P1. Necessarily, for all x, if x does not exist, then x is not perfectly valuable.
I'm not sure I'm understanding this. Consider a parallel:
P2: Necessarily, for all x, if x does not exist, then x is not a Jeep.
Surely, I cannot explain why my Jeep exists by virtue of the fact that it is a Jeep and citing P2... I know that's not what you are trying to say.
Hi Josh,
Your point about the jeep was raised by Heath at October 30, 2009 7:11 AM. I tried to answer in the subsequent comment. I'm not sure why I'm having such a difficult time making this point clear. I tried doing so in response to Alex above urging,
(P) does not commit us to the position that the conceivability of a perfectly valuable object entails that the object exists any more than the conceivability of a Magican entails that there are existing magicians. It commits us to the claim that the property of a perfectly valuable thing includes (if you'll let me talk this way) the property of existing.
I added, in the next comment,
It is apriori that the property red includes (in some sense) the property of being colored. This does not entail that anything is red. Similarly, it is apriori that the proeprty of being perfectly valuable includes the property of existing. This does not mean that anything is perfectly valuable.
But I guess it is still unclear. Maybe this will help. If you asked me to explain why object X has the property of being colored, I could reasonably answer, because it is red, and the property of being red entails (includes) the property of being colored. On analogy, if you asked me to explain why X exists I could reasonably answer that X has the property of being perfectly valuable and the property of being perfectly valuable entails the property of existing.
Perhaps I'm not understanding what you mean by 'x includes y'. My understanding is that every property conceptually includes the property of existing: every property is such that as a matter of necessity, if something has it then that something also exists (and so has the property of existing). Since I take it that you don't think that, you must have a different concept in mind here. Can anything be said to illuminate the concept you have in mind--for example a formal definition of 'x includes y'?
Wait: maybe I'm seeing what you are trying to say. In my own terms, I guess you have in mind something like this
'x includes y' def= 'nec, anyone who fully grasps x thereby grasps y'.
Or:
'x includes y' def= 'y is a constituent part of x'.
For example, being red and round includes being round. And perhaps we can explain why something is red by the fact that it has a property that includes being red--e.g., being red and round. In some cases this explanation may not seem satisfactory, but I'd have to think more about the case of perfect value...
Thanks for your patience. :)
But given my analysis above, for every x, we could explain it's existence by the fact that it has the property of being x and existing in combination with the fact that being x and existing includes the property of existing.
So, I'm still not sure how to think of 'x includes y' to get the results you are after.
Sorry for so many posts. :)
But given my analysis above, for every x, we could explain it's existence by the fact that it has the property of being x and existing in combination with the fact that being x and existing includes the property of existing.
I agree this point is a bit subtle. Maybe it would be useful to say, as I do somewhere above, that no Meinongian, non-existing object (supposing we could make sense of this) could have the property of being perfectly valuable.
But let me put it another way. It is true, as you say--Heath, too--that nothing could instantiate any property with also instantiating the property of existing. Agreed. But my claim is about the concepts expressed by these properties. I say that you can't understand the concept of perfect goodness without understanding the concept of existence. On the other hand, you can understand the concept of redness without understanding the concept of existence. I agree that all of this is a bit woolly, and probably should be spelled out in terms of synonymy. Perhaps it is also useful to note that (as it happens) the distinction is similar to one Rowe makes between magicians and magicans.
Mike,
Can we understand the concept of an existing Jeep without understanding the concept of existence?
If not, then why can't we explain why a certain Jeep exists by the fact that it has the property of being an existing Jeep?
The general principle here is that we explain why x is F by virtue of the fact that x is G and G includes F.
I take it that this is the general principle you are relying on when you say that the existence of a perfectly valuable thing can be explained by the complex fact that (i) it is perfectly valuable and (ii) being perfectly valuable includes existence. Yes?
Can we understand the concept of an existing Jeep without understanding the concept of existence?
Now this is an interesting question. Is there a concept of a jeep* such that I know apriori that jeep*'s includes the concept of existing? I honestly don't know that there is such a concept. I admit that it looks like there is such a concept, but that doesn't mean much. It looks like there is a coherent concept of being in less-than-perfect company. But there might be no world at all in which anything instantiates that property. In jeep*'s, is the relation between being a jeep and existing analogous to the relation between being red and being colored? Or is it more like the relation between (being red & being round) and being round? It looks ot me like the latter, but I couldn't swear to it. If so, then you'r explaining the existence of jeep*'s by appeal to the existence of jeep*'s, since being a jeep* is simply having two properties and two concepts, not one, one of which is the one we are trying to explain. But I don't deny that gerrymandered properties (if in fact they are properties) create a concern.
The concern you raise here sounds vaguely like the objection that all deductive arguments are question begging. Interesting, though I think we know they aren't all question-begging.
Dear Mike:
Consider the following argument:
- Necessarily, everything exists.
- Necessarily, if everything exists, then everything falls under the concept of existence.
- Necessarily, if everything falls under the concept of existence, then everything that falls under any concept falls under the concept of existence.
- Therefore: Necessarily, everything that falls under any concept falls under the concept of existence.
This argument appears plausible to me. And it also seems to me that its premises and conclusion are apriori if they are knowable. But if 4 is apriori, then 4's instances appear apriori too, including:Hey Douglas,
No, that's a nice argument. I don't deny your conclusion, but I don't think it presents a counterexample ot the argument. As I suggested to Josh, it is not true that everything that is entailed by P is explained by P. So, suppose you take as paradigmatic of explanations of something X that P entails X. It is pretty familiar to take entailment to subsume explanation. But not every entailment explains. For instance, P entails P, but P does not explain P. Further (P & Q) entails P, but (P & Q) does not explain P. On the other hand, the fact that something is red does explain why it is colored.
With respect to the particulars of your argument, I guess I'd ask whether, if everything exists, pink elephants exist. If premise (1) is understood standardly as everything is self-identical, we might worry whether things that are not in the domain of discourse are self-identical.
Thanks Mike,
Now I understand why my conclusion is not a counterexample to the idea that perfect value would, if something had it, explain its existence. Explanation and entailment are distinct phenomena, and apriori entailment does not suffice for explanation.
I am not confident that you will like my answer to your question. I wasn't expecting you to ask me anything!
I would deny that if everything exists, then pink elephants exist (that is, I would deny that if everything exists, then something is a pink elephant). I must deny this because I accept its antecedent and deny its consequent. But I accept that every pink elephant exists. I accept that everything is such that if it is a pink elephant, then there is something that it is. But I also accept that every pink elephant is diverse from everything! More carefully, I accept that everything is such that if it is a pink elephant, then there is nothing that it is. Surprisingly, if there are no pink elephants, then (I believe) every A-proposition with pink elephants as its subject is true!
Note: my answer is not the only possible one. David Lewis could accept my argument, and also accept your conditional that if everything exists, then pink elephants exist. Lewis would only have to deny that everything actually exists (on one of its readings), and 1 can be true even if, as Lewis thought, possible things are some of everything. The question that really bothers me is how to avoid Lewis' conclusion given I have not restricted admissible concepts in my second and third premises.
That last question doesn't bother me anymore.
But I also accept that every pink elephant is diverse from everything! Yes, this is the problem, if there is one. If pink elephants are diverse from everything, then they are diverse from themselves. So, the consequent. Though, you want to say that they are diverse from everyhting, but not diverse from themselves (since they're not included in everything). So, you wind up saying that pink elephants are not diverse from themselves, and not identical to themselves, either. So, ~(pelelephants = pelephants) and ~~(pelephants = pelephants). That's not good. Do have another way you want to put it?
It is pretty familiar to take entailment to subsume explanation. But not every entailment explains. For instance, P entails P, but P does not explain P. Further (P & Q) entails P, but (P & Q) does not explain P.
Mike, I'm glad this point came out in discussion. My original "exmount" objection was that nothing is easier than to come up with a concept that entails existence a priori: you just conjoin any old concept with the concept of existence, and you have a conjunctive concept which entails existence. But the objection was that this explains nothing about _why_ something that falls under this conjunctive concept exists. And I think Douglas' "existing jeep" concept made the same point.
Back to the original post: your P1 is logically equivalent to the claim "Perfect value entails existence." But if not every entailment explains, then all by itself this provides no support at all to your P. What we need is an account of why something's being perfectly valuable explains its existence, in a way that goes beyond merely noting the entailment.
P1 is logically equivalent to the claim "Perfect value entails existence." But if not every entailment explains, then all by itself this provides no support at all to your P. What we need is an account of why something's being perfectly valuable explains its existence, in a way that goes beyond merely noting the entailment.
I agree that it would be nice to have something like a general account of why some apriori entailments explain and other don't. All I have are examples. So, as I mentioned, it does seem like a proper explanation for why object X is colored that object X is red. That is an apriori entailment that explains. I assimilated the concept of a perfectly valuable being to that one. I want to distinguish those from cases in which we have gerryrigged a conjunctive property/concept (P & Q) in which one of the properties P is used to explain P. We agree that no explanation occurs there are all. I don't know off hand what the principled difference is between these kinds of cases.
Hi Mike
I appreciate your comments. I now see that I stated my view in a needlessly paradoxical way. I was thinking that the statement `Every pink elephant exists’ has the following truth conditions:
Since the predicate ‘either it is not a pink elephant or it is identical to something’ is true of everything the statement above is true. Indeed, it expresses a logical truth.
The statement ‘Every pink elephant is diverse from everything’ appears to have the following truth conditions:
There are no pink elephants. Thus, the predicate ‘either it is not a pink elephant or it is not identical to anything’ is true of everything, since ‘it is not a pink elephant’ is true of everything.
I mostly agree with you Mike. I don’t for a moment think that some pink elephant is diverse from itself. On the other hand, I don’t think some pink elephant is itself either. Since there are no pink elephants the question of whether some pink elephant is itself has a false presupposition: that there is a pink elephant.
Mike:
This is a very clever argument that I am only now fully appreciating when I see that it's hard to argue against.
Objection 1: Say that x is quasi-perfect iff it has all the perfections other than existence--roughly, it is such that if it is such that were it to exist, it would be perfect.
Then, x is perfect iff x is quasi-perfect and x exists. But if so, then to explain why x exists in terms of x's being perfect would be circular.
Probably you can say that perfection is a unity, and is not just the conjunction of quasi-perfection and existence.
Objection 2: Dilemma: Either non-existents do or do not have properties.
Horn 1: Non-existents don't have any properties. Existence is explanatorily prior to the possession of any properties. If so, then one cannot explain existence in terms of the possession of properties on pain of circularity.
Horn 2: Non-existents do have properties. This is going to be tough. Probably we don't want to attribute properties to necessary non-existents, or we get serious logical problems. Being a square circle entails, by explosion, being existing. So if objects can be given incompossible properties, they must exist, and hence there are actual contradictions.
So, we only attribute properties to possible non-existents. But that seems to mean that we attribute properties to non-existents in the worlds in which they exist. Thus, maybe, Sherlock Holmes in w1 is a detective, in w2 is a nurse and in w3 is both. To say that Sherlock Holmes is a detective simpliciter is to say something false.
So when we say that x is perfect, we mean that it is perfect in some world w1. But that only requires x to exist in w1, not in ours.
Dear all:
I like Alex's objection 2, Horn1. Here's how I would develop Horn1.
Necessarily, everything exists.
Necessarily, if everything exists, then everything that has any properties exists.
So: Necessarily, anything that has any property exists.
Therefore:
Necessarily, anything that does not exist has no properties.
Alex's comments suggest that if this conclusion is right, then existence is explanatorily prior to the having of any properties. Perhaps this is where Mike should poke.
I exist. Why? Because I have the property of having parents, I have the property of having a body that is maintaining chemical homeostasis, I have the property of breathing oxygen in the air, and so it goes. Necessarily anything that has properties exists, but we can sometimes explain something's existence by citing certain of its properties.
Perhaps Mike can argue that being perfectly valuable works (somewhat and somehow) similarly to chemical homeostasis, and the rest of the properties without which we would not be. Mike's driving thought, as I see it, is that we stand to gain some understanding of why a thing exists once we learn that it is perfectly valuable. Mike doesn't want to say that being perfectly valuable causally explains anything's existence, but perhaps that doesn't matter.