Well, this is really a challenge for skeptics of a Necessary Being (a necessarily existing causally powerful entity), but I wanted a catchy title, and as a matter of sociological observation, atheists are typically (though not always) skeptics of a Necessary Being.
The challenge is this: come up with a general, non-ad hoc causal (or explanatory) principle that's evidently more plausible than any of the ones that imply that there is a Necessary Being.
It is widely agreed (by the experts) that each of these principles implies the existence of a Necessary Being:
1. Contingency implies explicability.
2. The existence of any contingent xs has a cause.
3. Every contingent fact has an explanation.
Some lesser known routes to a Necessary Being make use of any of these:
4. Every explicable contingent fact has an explanation.
5. Any possible beginning is explicable.
6. Every wholly contingent fact has a cause.
7. Any possible contingent arrangement is causable.
Each principle above seems to be supported by a vast array of instances, and we've never observed any counter-examples (though it is controversial whether or not q.m. and/or the "taxicab objection" might imply counter-examples to (3)). The principles are also relatively simple and so don't seem to have low prior probabilities. The challenge for a skeptic of a Necessary Being is to present a causal principle that accounts for all known (or apparant) cases of causation (explanation, causability, or explicability) and that is at least as simple or non-ad hoc as any causal or explanatory principle that implies a Necessary Being. Good luck.
The challenge is also open to believers in a Necessary Being.
The challenge is this: come up with a general, non-ad hoc causal (or explanatory) principle that's evidently more plausible than any of the ones that imply that there is a Necessary Being.
Maybe the atheist cannot meet the challenge, but I'm not sure the theist can avoid brute contingent fact, either. The PSR's that are strong enough to remove brutality also remove contingency. Those PSR's weak enough to preserve contingency also preserve brutality. Let's preserve contingency. There are worlds W0 and W1 that God might have actualized. It is true at W0 that reasons R prevailed over reasons R' to actualize W0. But it is true at W1 that reasons R' prevailed over reasons R to actualize W1. What explains why R prevail over R' sometimes and R' prevail over R sometimes? In the end, when the reasons are all in, you will have brute facts about why reasons R prevail against R' in some cases and not others.
Now let W0 be actual. The contingent states of affairs that obtain are such that (of course) they might have been different, and the reasons R that they obtain are such that it is at least to some degree brutal that they prevailed over the reasons R' to actualize W1. In short, God could have had the very same set of reasons and actualized W1 instead. So it is hard to see how theism avoids brutality in explanation of contingent fact, where atheism does not. The only question is in where things get brutal.
Mike,
You present a standard objection to (3). If the objection is sound, then, as you suggest, it could still be the case that the atheist cannot meet the challenge (because of the other principles that seem to imply a Necessary Being).
I myself am not persuaded by your objection (though I used to be). For I'm not convinced that there cannot be an explanation of why R prevailed over R'. To my lights, one possible explanation would be this:
(P) God actualized W0 while having reason R to do so, and necessarily, if God actualized W0 while having reason R to do so, then R prevailed over R'.
Of course, (P) is contingent and so will need an explanation. To explain (P), I think it suffices to explain its contingent conjuncts. Only the first conjunct is contingent (by S4). And only the first "part" of the first conjunct is contingent, the part that says that God actualized W0. That part is explained by God's having reason R. (So God's having reason R--and necessarily so--explains why God actualized W0 while having reason R.)
There's no circularity here because (P) doesn't explain why God has R; rather, it explains why God's having R explains God's actualizing W0 (that is, why R prevailed).
We might go on to ask why (P) explains why (R) prevailed, since that, too, will be contingent. To my lights, one possible explanation would be this:
(P*) God actualized W0 while having reason R to do so, and necessarily, if God actualized W0 while having reason R to do so, then (P) explains why (R) prevailed.
From here, you see the pattern.
For I'm not convinced that there cannot be an explanation of why R prevailed over R'. To my lights, one possible explanation would be this:
(P) God actualized W0 while having reason R to do so, and necessarily, if God actualized W0 while having reason R to do so, then R prevailed over R'.
That's no explanation! (P) tells me nothing that I don't already know. I know that R prevailed at W0. What I don't know is WHY it prevailed. It would also be good to know what R' did not prevail, where R' is my equally good reason not to actualize W0.
It's amazing what God can get away with by way of sufficient explanation. Suppose you held out in from of me two intrinsically indiscernible pennies, one in your left hand and one in your right. I choose the one in your right hand. I admit to having reasons R to choose right and R' to choose left, and sometimes R prevails and sometimes R' prevails. But I do have this to offer by way of explaining why I chose right. Here we go:
I chose the penny on the right, and the explanation is (P).
(P) I chose right while having reason R to do so, and necessarily, if I chose right while having reason R to do so, then R prevailed over R'.
If I actually offered that to you as the explanation for why I chose right, you'd start giggling.
Mike,
The issues here are subtle. I was thinking of "R prevailed" as short-hand for "God's having R explained why God chose to actualize W0." Now in your explanation of why "I chose right", you cited a proposition that contains "I chose right", which (I think) makes the explanation circular. But in my explanation for why "R prevailed", I didn't cite a proposition that contains "R prevailed." So, I don't think your parody undermines my proposal.
I didn't understand your "that's no explanation!" argument. Once I come to know (P), I no longer wonder why R prevailed. It satisfies the why question for me anyway.
Perhaps you and I have a different proposition in mind by R prevails. At any rate, I'm not aware of any contingent proposition for which I can tell there could be no explanation. I'm open to new data.
Think of it this way. The explanation I gave was an instance of the following general form:
[X is the case, and necessarily if X is the case, then Y is the case]
explains
[Y is the case]
And surely that form of explanation doesn't inspire giggles.
The issues are subtle, no doubt. You're explaining (1), right?
1. God chose to actualize W0
and you believe that you can do that by appeal to (P),
P.God actualized W0 while having reason R to do so, and necessarily, if God actualized W0 while having reason R to do so, then R prevailed over R'.
But (P) contains (a) the fact that God actualized W0 for reason R, which I already know, and that (b)necessarily, if God actualized W0 while having reason R to do so, then R prevailed over R', which is a trivial truth. If God actualized W0 for reasons R and not reasons R', then obviously R prevailed. How does that explain anything?
In my case I want to explain why I chose right.
1'. Almeida chose right.
My explanation:
P1. Almeida chose right while having reason R to do so and necessarily, if Almeida chose right while having reason R to do so then R prevailed over R'.
That's perfectly analogous. And so is P2. They're all instances of MP.
P2. I chose right while having reason R to do so, and necessarily, if I chose right while having reason R to do so, then R prevailed over R'.
But these do not explain anything. I can't figure out why you think they do. The sort of case we're discussing is analogous to Buridan Ass cases. God has as good a reason to actualize W0 as he does to actualize W1. But Buridan's Ass cases are not so easily solved, assuming a rational Ass. Suppose the Ass chose the left bale. Does P3 explain anything at all?
P.The Ass chose the left bale while having reason R to do so, and necessarily, if the Ass chose the left bale while having reason R to do so, then R prevailed over R'.
If he chose the left bale, and the reasons in R and R' are the only sources of explanation, as we have stipulated, then he did so without explanation. The other side of explanation is prediction, and we have no reason to predict his going left. What you are claiming is that IF the Ass did so, then there exists an explanation. But that's not necessary, given the way the case is devised. It is rather the case that IF he did so, then he did not do so on the basis of the reasons available for that choice.
Similarly for God. If God actualized W0, then he did not do so on the basis of the reasons available for that choice. So, (P) should be replaced with P4.
P4. P.God actualized W0 while having reason R to do so, and necessarily, if God actualized W0 while having reason R to do so, then R prevailed over R' OR God did not choose to actualize W0 on the basis of the reasons available for that choice.
Mike, I don't think Josh is presenting a Buridan's Ass case. Remember that the reasons for the ass choosing one bale or the other are equally rational. Thus the ass cannot choose, and starves as a result. In Josh's case, however, God DOES choose--viz., P over P'-- suggesting that P is the more rational choice. Having said this, I think you are correct when you state that Joshua's argument has not told us why. I wish to know this as well.
Let me formulate an argument for why I think P does not offer an explanation for actualizing W0.
1. R is the complete set of reasons for actualizing W0. Assume
2. R' is the complete set of reasons for not actualizing W0. Assume
3. The reasons in R are no better (and no worse) than the reasons in R' (I'll write, R ~ R', for that). Assume
4. [](If R ~ R' and God is rational, then R cannot prevail over R for God and R cannot prevail R' for God). From essential rationality.
5. [](If R ~ R' and God actualizes W0, then he does not do so because R prevails over R'). From (4)
6. [](If R ~ R' and God actualizes W0, then the reasons in R & R' do not explain why he actualizes W0). From (5)
7. [](If R ~ R' and God actualizes W0, then there is no explanation for why he does so). From (1), (2), (3), (6)
Mike,
Thanks for spelling all that out. I must apologize because it's now apparent that I didn't realize what it was you were suggesting couldn't be explained. I thought it was
(2) God's having R explained why God chose to actualize W0
But you apparently thought I was using (P) to explains this:
(1) God chose to actualize W0
I agree with you that (P) is NOT a good explanation of (1). :)
I believe a good explanation of (1) is this:
(3) God had R (where R is a good reason to actualize W0)
_______
Perhaps what you are really looking for is a contrastive explanation of
(4) God chose to actualize W0 over W1, despite having reasons for both.
There I go with Pruss' solution... (expressed on his blog and partially in his recent Blackwell article)
As for your argument above, premise (4) doesn't strike my mind as evidently true.
Josh:
Have you seen Graham Oppy's suggestions along these lines, as well as Rob Koons' list of proposals (and why they either yield the existence of a necessary being or don't work)?
For your list:
8. Every contingent arrangement much smaller than the whole universe (i.e., the whole contingent arrangement of reality) has a cause. (That's a fun one. Take a finite universe. Surely that's possible. Divide it up into a finite number of pieces much smaller than itself. Use the principle to conclude that each has a cause. On pain of causal circularity, this cause must be outside the universe, and hence must be non-contingent.)
Mike:
I agree with Josh that (4) does not appear right. Rational agents can be moved by incommensurable considerations, which then prevail.
Josh:
This might work, too:
9. Every fundamental particle has a cause.
(Consider a world whose contingent part is constituted by a finite number of fundamental particles, and then use anti-circularity. Wouldn't it be really weird, after all, if necessarily any world constituted by fundamental particles had infinitely many of them?)
I need 8 and 9 to be necessary principles.
Mike:
It might help to place Josh's arguments about free will if you think of Kane's account of free will. Kane thinks that if in a SFA you choose A, that's explained by the motivational structures that favor A, and if you choose B, that's explained by the motivational structures that favor B.
Ok, so (4) is supposed to be the problem.
4. [](If R ~ R' and God is rational, then R cannot prevail over R for God and R cannot prevail R' for God).
I'd love to know why (4) is false, if it is. But so far I haven't been given a reason for this. The reason cannot be Kanean, since we are clearly not talking about explanation in the context of indetermnism. I agree that in indeterministic contexts we can speak coherently about the role of chancy causation in an explanation of behavior. But God's action is not in an indeterministic context. The reason cannot be the incommensurablity of reasons, since it is not true that the reasons God has for actualizing our world are incommnensurate with the reasons he has for actualizing any other possible world. They might be incommensurate with the reasons for actualizing some other worlds, but not with the reasons for actualizing any other world.
So, now, choose a world W0 that God failed to actualize, but could have. To make it easier, choose a W0 in such a way that the reason R for actualizing our world are no better or worse than the reasons R' for actualizing W0 (matters are actually much worse since there are worlds God might have actualized for the actualization of which he has much worse reasons. It is still true that those bad reasons might have prevailed. But leave those tough cases aside.).
Now R is not incommensurate with R', R and R' are all of the reasons God has for actualizing our world and for actualizing W0, and R ~ R'. Given these conditions, give me a counterexample to (4). And please don't say that God actualized @ and R is a good reason for actualizing @. God's reasons include BOTH R and R'. And those together make God's reasons to actualize @ neutral. On balance, his reasons would explain, at most, his indifference to actualizing @, it would not explain why he did actualize @.
The reason cannot be the incommensurablity of reasons, since it is not true that the reasons God has for actualizing our world are incommnensurate with the reasons he has for actualizing any other possible world. They might be incommensurate with the reasons for actualizing some other worlds, but not with the reasons for actualizing any other world.
I'm probably just missing something, but it's unclear to me why you think this is true.
"Contingency" as a concept depends on necessity. Therefore it already contains its opposite, and hence using it in an argument against necessity simply begs the question.
Mike,
One reason to reject (4) is PSR. :) I see no reason to accept it.
I don't want to stray too far from the main point of this post. PSR is the most controversial of the principles I mentioned, and if it's false, the open challenge still stands.
Alex, yes, I'm aware of them. Nice proposals; you got me thinking about (8) from your presentations at St. Thomas....
"The challenge is this: come up with a general, non-ad hoc causal (or explanatory) principle that's evidently more plausible than any of the ones that imply that there is a Necessary Being."
The universe (or, whatever entity whose existence you are trying to explain with a Necessary Being) has necessary existence of the same sort that a Necessary Being has.
Exactly as plausible as any Necessary Being, and exactly one assumption more parsimonious.
What do I get for winning the challenge?
I'm probably just missing something, but it's unclear to me why you think this is true.
Brandon,
It might be the way I phrased it. The question concerns why God actualized this world rather than any other actualizable world. I think it is extremely unlikely that, for any other possible world W that God might have actualized instead, the reasons R for actualizing @ are incommensurable with the reasons R' for actualizing W. So, for instance, take a world W' that differs from ours with respect to a single particle of matter. The reasons for actualizing @ are not commensurable with the reasons for actualizing W'? Hard to believe.
Josh,
Sorry. I didn't know I was straying off topic. I was trying to show that theists are in the same boat as atheists on this score.
Teapot,
Strictly speaking the challenge is for skeptics of a Necessary Being. Yes, one can be an atheist who thinks there is a Necessary Being (such as the universe). Still, few atheists would say that our universe is a Necessary Being, for they would grant the metaphysical possibility of a different universe (e.g., one with one fewer protons or one with slightly greater mass). Wouldn't you?
Teapot,
There are lots of other (candidates for) necessary beings, such as sets, properties, numbers, propositions, worlds, essences, etc. It might even be the case that we (finite creatures) necessarily exist. But Josh pretty clearly has in mind by 'necessary being' a perfect sort of being that can do certain sorts of explanatory work.
Mike, by 'Necessary Being', I mean only what I said: a necessarily existing causally powerful entity. Perfection isn't built into its meaning.
I see. And I'm guessing you mean by 'skeptics of a necessary being' skeptics of there existing any necessary being whatsoever fitting the description. Otherwise, we're all skeptics of a necessary being. Suppose I'm not such a skeptic, but I understand by 'necessary being' a world-bound individual. Do the principles entail that not only every contingent being requires some explanation or other, but also every non-contingent world-bound individual requires some explanation or other?
Mike,
Some of the principles I listed don't entail that every contingent being requires some explanation (even if they do entail that there is a necessarily existing entity with causal capacities). If someone thinks that some world-bound (concrete) individuals are necessarily existent, then she isn't a skeptic of a Necessary Being. But I don't know anyone who thinks that.
The world-bound individuals needn't be concrete. I mean, you can think of worlds along the lines of Plantinga and still be a counterpart theorist. But what I'm puzzling over is how a world-bound individual might serve the causal purposes required by those principles while a contingent being cannot. That's tricky. There is an interesting sense in which any world-bound individual would be regarded as (certainly by Plantingans) contingent beings.
Mike, perhaps your point is that any Necessary Being that might be implied by any of the causal principles listed would, for various reasons, contain certain perfections not included in my definition of a 'Necessary Being'. If that's your point, I accept and applaud it. :) That's where "stage II" comes in--to draw that out. (I realize that this was your point to Teapot; I only wanted to make sure it was clear that this step isn't merely a matter of definition.)
I think that's likely to be true, Josh. But my point in the comment above is just to air my own puzzlment about how well a necessary being of a certain sort would serve the purposes of the principles you cite. Maybe there's no genuine worry about world-bound, necessary beings serving the purpose. I just don't know.
Mike:
"But God's action is not in an indeterministic context."
I'd like to hear why you think it isn't. Obviously, it's not in an external indeterministic context, but neither is the indeterminism that Kane is talking about external.
Mike,
One reason to deny (4) is this:
1. Possibly, someone is responsible for freely performing an action for reasons R, where she could have instead performed a different action for equally good reasons.
2. Necessarily, if someone is responsible for a free action, then that action is not unexplained (else it would be random).
3. Therefore, possibly, one's freely (and responsibly) performing an action for reasons R, where one could have instead performed a different action for equally good reasons, is explained.
And what else would explain a person's free action other than the reasons she has for performing it?
Obviously, it's not in an external indeterministic context, but neither is the indeterminism that Kane is talking about external
What does this mean? What is the difference between an internal and external indetermnistic context? As I read Kane (and I'd be happy to ask him about it) his indeterminism just is the indeterminism that the physicists are talking about. What other indeterminism is there?
One reason to deny (4) is this: Possibly, someone is responsible for freely performing an action for reasons R, where she could have instead performed a different action for equally good reasons.
This is where hyperrational free agents, unlike the rest of us, can be impeded by their own rationality. If a hyperrational agent faces a choice between A and ~A, and his reasons R for A and R' for ~A are such that R ~ R', then he cannot make a choice. He cannot choose because he only does what he has most reason to do. He does not choose arbitrarily when reasons wash out. Now maybe God is choosing arbitrarily when reasons wash out. If so, then his choice is, to that extent, brutal.
There is also the indeterminism of libertarian free choices, whether or not these supervene on quantum of physics.
There is also the indeterminism of libertarian free choices, whether or not these supervene on quantum of physics.
The indeterminsim has the same source. Kane is simply trying to make the physicist's indeterminism compatible with free choice. This is part of what self-formation is supposed to do.
Mike:
1. Why think that the only two options are that God is hyperrational (in the sense that he cannot choose between balanced or incommensurable reasons) or that his choices are reasonless? Why not allow that God might have the same kind of freedom as we do, which lets us choose on the basis of reasons even when the reasons are balanced or incommensurable, as in Robert Kane, Randolph Clarke, etc.?
2. But suppose the above are the only two options. As long as you admit that the kind of freedom I am talking about--the kind where there is no determination of outcome and the actions are nonetheless explained by reasons--is logically possible for a creature, we can have the following somewhat heterodox model. (I am not proposing the model as true, but only as sufficient to show a way to reconcile the PSR with the negation of fatalism.)
Say that a collection C of propositions is a "core" provided that (a) God can strongly actualize all the propositions in C, (b) C is maximal with respect to (a), and (c) for every world w in which all the propositions in C hold, there is a set (perhaps empty) D(w) of propositions of the form <x libertarian-freely and not hyperrationally chose option A in circumstances C for reasons R&rt; where x is a creature, such that for every contingent proposition p true at w that is a member neither of C nor of D(w), there is a subset C' of C and a subset D' of D(w) such that p is both entailed and explained-at-w by the conjunction of the propositions in C' and D'.
Then the theory is this. Each core C has a value V(C). This value may be the expected utility if one thinks rationality is tied to expected utility maximization, but it may also be calculated with a somewhat different measure if one thinks, for instance, that there is some risk minimization going on. So far this is fairly orthodox, and may be even true.
The heterodox part (which I am only introducing because you were balking at what I think is the correct story, namely 1) is that necessarily God chooses a core C with the highest V(C) and strongly actualizes every proposition in C. Call such a core "optimal". If there is no such core (or if there are no cores), God is not capable of doing anything.
Suppose that there is a core C with the highest V(C). Moreover, suppose that it is a necessary truth that C is an optimal core. In assuming this, I am assuming Molinism is false. Molinism, anyway, appears inconsistent with PSR (the counterfactuals of freedom have no explanation--David Manley observed this in conversation with Plantinga and me once).
I now claim the following follow:
1. Probably, modal fatalism is false.
2. The PSR is true at every possible world.
For it follows from the earlier assumptions that if C is the optimal core, then every proposition in C is a necessary truth, since necessarily God strongly actualizes it. Now, (1) holds because, probably, the optimal core will include some proposition that entails that there is a free non-hyperrational creature that chooses on reasons. If there is some such proposition, then there will be a world compatible with C at which the creature chooses one way and a world compatible with C at which it chooses in another way.
What about (2)? Well, that's easy. Let w be any possible world. Since C is true at every possible world, C is true at w. Let p be a contingent proposition. Then p is not a member of C, since all the members of C are necessary. Either p is a member of D(w) or it is outside of both C and D(w). If it is outside of both C and D(w), there is a subset C' of C and D' of D(w) such that the conjunction of the propositions in C' and D' explain p, by condition (c). If it is a member of D(w), then p is of the form <x libertarian freely and non-hyperrationally chooses A for reasons R in circumstances C&rt; and hence is explained by x's having reasons R and being in circumstances C. Thus, every contingent proposition true at w has an explanation.
Now maybe God is choosing arbitrarily when reasons wash out.
When God chooses "arbitrarily", is God responsible for his choice? Do you think that a being can be responsible for a choice that occurs without any explanation?
I suppose it seems to me perfectly rational (in whatever sense a perfect being would be perfectly rational) for a being to chose A for reasons R while having equally good reasons to chose ~A. For example, God might see that creating X is just as good as creating a particular intrinsic duplicate of X, but it seems to me that it would still be perfectly rational for God to go ahead and create X, and that if He does, his action would be explained by his reasons for doing so.
Alex,
We might be able to explain contingent counterfactuals.
For example, let C be the counterfactual that if Alex were in the garden, etc., he would have freely refrained from eating the fruit. If that's true, perhaps it could be indeterministically explained by C*: if Alex were in the garden, etc., he would have had some reasons to be faithful to God's instructions and so to refrain from eating the fruit. And C* might be a necessary truth--or else ultimately explained by a necessary truth.
1. Why think that the only two options are that God is hyperrational (in the sense that he cannot choose between balanced or incommensurable reasons) or that his choices are reasonless? Why not allow that God might have the same kind of freedom as we do, which lets us choose on the basis of reasons even when the reasons are balanced or incommensurable, as in Robert Kane, Randolph Clarke, etc.?
I don't think I made that dichotomy. It is either true or false (let's say) that God is hyperrational. If he is not hyperrational, then he is closer to us in rationality than I would expect, but ok. When we are faced with a decision between alternatives and our reasons for each alternative washes out, then we can choose arbitrarily. We can flip a coin, say, or just choose (why bother with a coin?). But when asked, why did you go left at the fork, given all the reasons you had? I have to say that the reasons I had, taken all together, did not give me the direction "go left". But I went left. Same with God. And that leaves going left finally unexplained.
I'm not sure I'm following entirely your reconciliation argument above. The basic idea, I take it, is to let a certain set of states of affairs be strongly actualizable, and also necessarily strongly actualized, and allow worlds to differ with respect to contingent states of affairs that free creatures strongly actualize or that occur as a result of chancy events in nature. What I don't see is how the states of affairs that are the result of chancy events are entailed by some C' and D'.
Josh:
Clever. It won't work as well, though, if we have a causal PSR, or if we think that contingent states of affairs have causal explanations, will it?
Mike:
So you dispute the possibility of explanation even in our own case. I didn't realize that. I thought you were willing to grant that Kane-style explanations worked for non-hyperrational beings. Do you in general deny the possibility of probabilistic explanations?
Molinism, anyway, appears inconsistent with PSR (the counterfactuals of freedom have no explanation--David Manley observed this in conversation with Plantinga and me once).
I've never understood the claim that CCF's have no explanation or that they are brute. I know that's the standard description. Here's why. CCF's describe contingent properties of uninstantiated individual essences. But the contingent properties of individual essences are all explained. Either, God causes an individual essence to have property P or the essence causes itself to have property P. God can cause the individual essence to be such that, were it instantiated in T, it would be six feet tall. But God cannot cause the essence to be such that, were it instantiated in T, it would freely perform A. But that property does not come out of the blue. The essence causes itself to have that property by freely choosing to act that way in those circumstances. That's all the CCF comes to.
So you dispute the possibility of explanation even in our own case. I didn't realize that. I thought you were willing to grant that Kane-style explanations worked for non-hyperrational beings. Do you in general deny the possibility of probabilistic explanations?
In our case, there are de facto reasons for going left that are not de jure reasons for going left. Maybe this is where our disagreement is. I don't count any old de facto explanation for someone's going left as a genuine explanation. So, maybe someone flips a coin with the outcome 'go left if it falls heads', and it falls heads. De facto, the explanation for his going left is that the coin fell heads. But the coin's falling heads is not a reason to go left and so not a de jure explanation for going left, and so not a genuine explanation for the choice. So, you can have an explanation for choices that are, at the same time, arbitrary.
Mike, I do not understand. What explanatory work does the Necessary Being who is the subject of this post do? Help me to understand what it explains. Oh, but also make sure to include what it explains that is not explained by the principle I offered. I am particularly interested in an explanation of what is more parsimonious about the Necessary Being who is the subject of this post.
I am asserting that the universe is a necessary being. This does a lot of explanatory work: it explains why there is a cosmos, for example. It even does so in a way that is one entity more parsimonious that this strange hypothesis called theism which, so far as I can tell, explains nothing whatsoever. I've heard it said that theism explains why there is a universe, but I have never heard a good explanation for why the Necessary Being offered by theism would create a universe, or how it does it, so I must disdain this hypothesis until someone explains it to me. Perhaps I am just a slow learner.
I didn't read all this carefully, but I skimmed through, and I didn't see any mention of what seems to me to be the main resistance to this nowadays, which is just to go Humean about causation. If there's nothing explaining causation because it's not metaphysically real in any important sense but is just whatever regularities there happen to be (as surprising as it might be that there are as many as there are), then there need be no principle explaining causation (but I think the unwelcome result for Humeans is that they have no account of why such surprising degree of regularities occurs, at least for those resisting Lewis' claim that every world exists).
Teapot,
Here are seven reasons one might doubt the universe is a necessary being:
1. [weakest reason] The 'universe' doesn't refer to a thing; rather, it refers to a bunch of things, none of which are necessary.
2. Our best science suggest that the universe had a beginning, but a Necessary Being cannot have a beginning.
3. Possibly, our universe had a beginning, but a Necessary Being cannot have a beginning.
4. There could have been just 26 electrons (by modal continuity) in which case our universe wouldn't have existed.
5. The properties of our universe (be it a single cosmos or a multiverse) is finely tuned (has properties within a very narrow range) to make biological complexity physically possible, but the physical possibility of biological complexity isn't a matter of necessity: e.g., there could have been a barren world--one in which matter and energy are too disperse or too dense to allow for the possibility of biological complexity. Thus, there could have been a different universe instead. Thus, ours isn't necessary.
6. Particles in our universe are spatially related to each other, and their exact spatial relationships isn't a matter of necessity (any two particles might have been slightly farther apart--given standard views of quantum mechanics). So, the totality of all spatial relationships (throughout time) is a contingent matter. If every contingent arrangement has an explanation, then this spatial arrangement has an explanation. But the particles themselves cannot explain the totality of all the spatial arrangements they've ever been in (because they are already spatially arranged at every moment of their existence). Therefore, something outside our universe is required to help explain the total spatial arrangement of the particles (for example, by explaining their initial distribution).
7. The gravitational constant is essential to our universe (which is why it's constant). It's value =
6.67300 × 10-11 m3 kg-1 s-2. But by "modal continuity", it's broadly logically possible for there to have been a universe whose gravitational was instead 6.67307 × 10-11 m3 kg-1 s-2. (I'm assuming the commonsense view that not all possibilities are actual of necessity.)
I am asserting that the universe is a necessary being. This does a lot of explanatory work: it explains why there is a cosmos, for example
Teapot, the challenge is to come up with a hypothesis that is at least as plausible. The hypothesis that the universe necessarily exists hasn't been seriously advanced since roughly 1920 or so. The universe had a beginning and it will have an ending, so it is not a necessarly existing being.
The gravitational constant is essential to our universe (which is why it's constant). It's value =
6.67300 × 10-11 m3 kg-1 s-2. But by "modal continuity", it's broadly logically possible for there to have been a universe whose gravitational was instead 6.67307 × 10-11 m3 kg-1 s-2.
It's a delicate point, but I'm not sure that our universe has any non-trivial essential properties. This is part of what makes a modal reply to Teapot difficult. You can believe that the universe necessarily exists and believe that it does not otherwise have any essential properties. In different worlds, the same universe might be wildly different from the way it actually is. Might the universe have had 26 electrons? I'm not sure, but I wouldn't want to deny it on the basis of, say, mereological essentialism.
Mike,
Those are good points. But there is this consideration. If someone thinks there is such a thing as the universe, presumably she thinks there are such things as galaxies, too. Now if there could be just 26 electrons, then it seems that there could also be just 1 of the galaxies that there are. But now consider that the universe cannot be identical to one of its proper parts (by the necessity of identity); therefore, the universe cannot be identical to any galaxy it contains. Therefore, if there were just 1 of the galaxies that there are, then our universe wouldn't exist.
Also, I think it's plausible that the if any fundamental particles could be absent from existence, then all could be, in which case there would be no material things and hence no universe. (E.g.: if there could be just 26 electrons, surely there could be just 25, or just 24, or... just 1, or 0!)
If someone thinks there is such a thing as the universe, presumably she thinks there are such things as galaxies, too. Now if there could be just 26 electrons, then it seems that there could also be just 1 of the galaxies that there are. But now consider that the universe cannot be identical to one of its proper parts (by the necessity of identity); therefore, the universe cannot be identical to any galaxy it contains.
Well, sure. I agree. But I don't think we need to violate LL to get this outcome. The idea is that it is logically possible (obviously, not physically possible) for the universe to evolve in such a way that there remains nothing but that single galaxy. Right. Given that possibility a story has to be told to avoid threatening inconsistency that you point up. Which story? Well, you know how it goes from here. Either you take some sort of presentist view where the universe is identical to everything there currently is, or you take some 4Dist route, where the current part or stage of the universe should not be confused with the big 4-dimensional thing that is the universe, or you index the properties of the universe to times and preserve consistency that way.
So now take that possible world in which (say) the current stage of the universe is one galaxy. We are talking about the same universe (same past stages), but an odd evolution to a single galaxy.
Josh:
"But now consider that the universe cannot be identical to one of its proper parts (by the necessity of identity); therefore, the universe cannot be identical to any galaxy it contains. Therefore, if there were just 1 of the galaxies that there are, then our universe wouldn't exist."
This reminds me of my argument against heads. Suppose Tibbles is wholly material. Call Tibbles' head "Ti". Tibbles could survive the loss of everything but Ti. If Tibbles lost everything but Ti, then Ti would not be a proper part of Tibbles, since Tibbles would have no larger parts, but it would not be identical with Tibbles, which is absurd.
The argument pulls in the direction of nihilism about non-mereological parts. (And since there are no mereological sums, nihilism about parts.)
The best answer is to deny weak supplementation, and say that you can have an improper part that has nothing outside of it (in the whole). Then, you say that after the operation, Tibbles has Ti as a part with nothing outside of it (in the whole), and Tibbles is wholly constituted by Ti, but Ti is still a proper part of Tibbles.
The same goes over for your universe case. Your opponent should say that the galaxy is a proper part of the universe that wholly constitutes the universe.
Alex,
I've thought of that, actually. It helps to distinguish between two categories of concrete things--arrangements and substances. If the universe is an arrangement, then on my account of arrangements, it would fail to exist were even one proton eliminated from it. So, let's think of it as a substance. Same for galaxies. Now I'm uneasy about applying your suggestion to wholly material substances, for it would seem to have the implication that there could be distinct material substances that occupy the exact same space. Most people I've surveyed (mostly freshman) find that counter-intuitive, as do I. Now I do not think of organs of organisms as substances in their own right. Rather, they are arrangements whose existence depends upon the existence of the organism itself (I say). So, I suppose an organ could come to occupy the same space as the organism of which it is an organ but only because the organ isn't itself a substance.
Of course, my opponent could say that no galaxies are substances. But if she says that, then I don't see how she could get away with saying that the universe is a substance.
Mike,
Some valuable points. But to be clear, I don't think that the universe could evolve in such a way that it still exists while there remains nothing but that single galaxy. For if it could, then the universe could become identical to one of its proper parts, which is impossible.
I'm not sure I understood two of your "stories." I don't know what it means to say that the universe is identical to everything. If you meant it's identical to the sum of everything, I don't see how this is supposed to escape the problem. (Perhaps you had mereological essentialism in mind plus the thought that when there's nothing but the single galaxy, our universe no longer exists...)
I also don't understand how temporal indexing is supposed to help. If x is a proper part of y at t, then there's no time at which x is identical to y. So, if at any time nothing but x exists, then y doesn't exist at that time. So, if at any time nothing but a certain galaxy exists, then our universe doesn't exist at that time.
I understand the 4Dist route, but I don't think it comes into play when we consider the possibility of a worm universe U that's just like our worm universe except that there's a time at which it has one fewer protons. U would seem to be a proper part of our worm universe, so if U could exist without being a proper part of any bigger universe, then our universe isn't necessary.
Some valuable points. But to be clear, I don't think that the universe could evolve in such a way that it still exists while there remains nothing but that single galaxy. For if it could, then the universe could become identical to one of its proper parts, which is impossible.
As far as I know, the problem under discussion is the violation of LL and how to avoid it. I've been offering ways of doing so. So, when someone says the universe might (metaphysically 'might') consist in just one galaxy, what he means is that the universe might evolve in such a way that there is one galaxy. That's perfectly possible. It does not entail the whole universe is identical to a part of the universe or whatnot. There is a world in which a single 3 dimensional galaxy is the current stage in the 4-dimensional object that is the universe. What people point to in that world and discuss as 'the universe' is just a 3D stage of the universe; we are doing the same here. So when you point to the 'universe' and say that it couldn't evolve into a single galaxy, there is a sense in which what you say is true. Certainly this part of the universe (the one you're pointing to and calling 'the universe') could not evolve into a single galaxy. But the universe could. Just to be clear, the universe could in the 4D sense that the current 3D stage of U might be nothing more than a single galaxy.
I don't know what it means to say that the universe is identical to everything. If you meant it's identical to the sum of everything, I don't see how this is supposed to escape the problem.
The universe is the whole 4 dimensional object, the sum of its stages/parts. I've lost track of what you are calling 'the problem'. I thought the problem was the avoidence of a violation of LL.
So, if at any time nothing but x exists, then y doesn't exist at that time. So, if at any time nothing but a certain galaxy exists, then our universe doesn't exist at that time.
Let me give an example. Suppose you're a 3Der and want to explain the endurance of the universe. The universe at one time has property P (some number of water molecules) and at another time ~P. Oh no, a violation of LL! No enduring object U can be both P and ~P. To preserve consistency, some have argued that U is not both P and ~P. Rather, U is P-at-t and ~P-at-t'. These properties (relations, really) are such that a single enduring object can exemplify both consistently; the initial properties did not have that property. There are other metaphysical worries about going in this direction, but LL is not one of them. And I've been taking the violation of LL as the problem to be avoided.
The other 3D alternative I mentioned does not relativize properties to times, but rejects the idea that the current 3D object--the universe with property ~P--is identical to any past object exemplifying P. That too avoids the LL problems.
Josh:
1. "Now I'm uneasy about applying your suggestion to wholly material substances, for it would seem to have the implication that there could be distinct material substances that occupy the exact same space."
I don't know for sure if elementary particles are material substances, but they seem pretty good candidates. But it is an important part of the physics that bosons can have the same quantum state, and hence the same (fuzzy, but fuzzy in the same way) position. Moreover, it seems pretty clearly possible that there be ghosts, and ghosts can occupy the same space (think of the ghost as made of a lightly glowing interpenetrable body, that makes you feel cold when it occupies the same space as you do).
Where I think one would be on solider ground (pun not intended, but also not prevented) would be with the weaker principle that that the same exact same matter could not wholly constitute more than one wholly material substance at the same spacetime location. And that's all you need for your argument.
2. "Now I do not think of organs of organisms as substances in their own right. Rather, they are arrangements whose existence depends upon the existence of the organism itself (I say)."
But then I don't have the same brain that I had yesterday. That seems only slightly less counterintuitive than my view that my brain doesn't exist (actually, that might not be counterintuitive to you at all, depending on what you think of my arguments). :-)
3. It seems wrong to take a galaxy to be an arrangement. Then we live in a different galaxy every moment.
4-5. You might respond to 2 and 3 that galaxies and organs are four-dimensional arrangements. But then today you can affect what galaxies and brains existed yesterday (e.g., by destroying a particle). And that doesn't seem right.
6. How about drawing the distinction between organs and substances differently. In addition to substances and arrangements, there are pseudosubstances (if you forgive the Greco-Latin). Pseudosubstances have their basic parts inessentially, but they essentially depend on a substance for their existence and some of their individuation. Then brains are pseudosubstances.
7. Here's an interesting thing. Suppose that x is an mereological sum of particles, and some of x's particles wholly constitute a material substance y. Then it seems that y is a part of x. But typically y is not an essential part of x. For typically we could move x's particles in such a way that x ceases to exist, but y surely continues to exist (if we have a heap of particles, and some of the particles make up a flea, the heap survives the death of the flea). So either (a) material substances can't be parts of mereological sums or (b) every particle in a material substance can be a part of a mereological sum without the material substance being a part of the sum. Since the same applies to arrangements with liberal enough relations, (a) will be a problem for you if galaxies are arrangements. I guess you could take (b), but I am not sure you'd be happy with it.
Mike,
We may be on different tracks. I gave an argument for the conclusion that if there could be just 26 electrons (or a single galaxy), then our universe isn't necessarily existent. My argument didn't make use of LL, and I don't see how any of your stories reveals a problem with my argument (but perhaps that wasn't your goal).
Alex,
Those are all excellent points. I shall remark on two of them.
Re 2. Well, I didn't say that my organs are arrangements of material parts. :) On some days I'm tempted to think of my organs as arrangements consisting of me bearing exemplification to certain functional properties--but that makes me "part" of my organs... I'm open to the "pseudo-substance" idea.
Re 3: I'm tempted to translate "galaxy" talk into talk of things arranged galaxy wise... (this relates to point 7). I know there are issues here. But none seem as troubling to me as the idea that there could be two wholly material substances that occupy the same space(and are made of the same matter).
Josh, you wrote this,
4. There could have been just 26 electrons (by modal continuity) in which case our universe wouldn't have existed.
Now there is a sense in which (4) is true and a sense in which it is false. The only sense (other than an appeal to mereological essentialism) in which it is true appeals to a violation of LL. Our universe could not have been just 26 electrons because the assumption entails a violation of the indiscernibility of identicals. a 26 electron universe has properties that our universe lacks, so they cannot be identical.
There is also a sense in which (4) is false. This is what I've been providing for you. Are we still on different pages? :)
Mike, here's perhaps a clearer formulation of the argument:
1 Suppose there is such a thing as the physical universe, U, and that there there could be just 26 electrons.
2. Then (plausibly), there could be just the Miley Way, M.
3. M is not identical with U (as it's a proper part).
4. Therefore, nessarily M is not identical with U (necessity of identity).
5. Therefore, possibly, U doesn't exist (from (2) and (4))
The 4-Dist should interpret the 'is' and 'be' as tenseless.
Alex suggested denying (2) on the grounds that if everything outside M were deleted, perhaps M would still be a proper part of U. (This led to further discussion.)
Mike, I posted before seeing your last post, but perhaps you could clarify which premise above you are challenging (if any). I also invite Teapot to consider the argument above.
The argument you offer cannot be one you believe is sound. If you did, then you'd have to say that the chair in your living room is not identical to the chair that was there yesterday, since it has lost a few atoms since yesterday, and your car is not identical to the one you bought yesterday and the house is not identical, and the lamp, pencils are not, and the apple banana, pear, are not and so on and on. Just utterly implausible. What the argument you offer does is present a problem to be solved. The solution is one of the one's I've offered a few times already.
Well, Mike, it looks like I have two beliefs you think I cannot have: (i) there are no brute contingent facts, and (ii) the argument I gave above is sound. :) Though I'm much more confident about (ii) than I am about (i).
Your objection to the argument seems to be that a parallel argument gives us the counter-intuitive conclusion that my laptop isn't the same one I used yesterday. Well, I have two hypotheses about the identity of laptops: (a) to say that my laptop is the same one I used yesterday is loose speak for practical use, but strictly speaking the laptop doesn't maintain strict numeric identity through loss of parts (because it isn't strictly speaking a substance), or (b) when someone says 'this is the same laptop I used yesterday' he ultimately expresses the same proposition as I do when I say 'this exemplifies the same type [e.g., being my laptop] as what I used yesterday exemplified', where there can be ambiguity as to how specific the type in question is. The disjunction of (a) and (b) doesn't strike me as too unlikely. Or, at least the disjunction doesn't seem any more unlikely than does the denial of any one of the premises in the argument I gave above.
The puzzle, as I see it, is to see how to deal with the results of parallel arguments to the sound argument I gave. I don't believe any of your stories provide a way of escape from my argument.
I'm curious, which premise of my argument do you not accept?
Of course, there is also (c): "laptop" talk is loose speak for talk of particles arranged laptop-wise (though this can be subsumed under (a)).
I disagree with a few things you've said, but I hope I'm not sounding disagreeable. About the view you've just sketched, I don't believe you. That is, I don't believe that you believe this. Suppose you purchase your laptop at t. The following day at t' you say, "strictly, this laptop is not identical to the one I purchased yesterday". Since I know what your metaphysical views are, I rightly conclude that the laptop on your desk is not the one you legitimately acquired, since you didn't purchase it and it wasn't given to you and you did not otherwise legitimately acquire it. All you did was purchase another laptop that 'exemplifies the same type'. Indeed the laptop on your desk belongs to no one legitimately. Since the laptop belongs to no one, I'm free to take it for myself. And that's what I do, of course with your blessing. Since, consistent person that you are, you agree that it just isn't yours, and anyone who claims its yours is speaking with the vulgar.
If you persist in taking this view, I'll be over tomorrow.
As I said about your other argument, it presents a problem to be solved. I have offered you several pretty standard solutions. One of those solutions (the presentist solution) rejects (2). I don't find that any less plausible than a solution on which I may legitimately take your laptop for myself :).
Mike:
I don't think ownership point is decisive. When I own a cow and a bull, I also own the calves. And this is not limited to cattle, I think. If I own a bit of forest and a big machine that automatically harvests the trees and transforms them into notebooks, I also own the notebooks. Likewise, if one takes the Chisholm line that the laptop yesterday and the laptop today are strictly speaking distinct, there is still the fact that the laptop yesterday produced the laptop today.
I thought you were friendly to the view that there can't be vague existence. But if there is strictly laptop identity over time, there can surely be vague existence of a laptop. Here's a scenario. At t1, replace 1% of the laptop with new components. At t2, replace 2% of the laptop with new components. At t3, replace 3% of the laptop with new components. Go on until t100, where you replace 100% of the laptop with new components. Clearly, the replacement at t100 produced a new laptop, if it makes sense to talk of strict identity of laptops. But, plausibly, so did the replacement at t99. On the other hand, the replacement at t1 didn't produce a new laptop. So either it is vague which replacement produced a new laptop, in which case there is vague existence, or else we have the weirdness that replacing, say, 49% of the laptop doesn't make a new laptop, but replacing 50% does.
Josh:
If you like the idea that our bodily parts not arrangements of material parts, you might want to take Rob Koons' line on laptops. He thinks they aren't material objects. Rather, they are individual social practices.
likewise, if one takes the Chisholm line that the laptop yesterday and the laptop today are strictly speaking distinct, there is still the fact that the laptop yesterday produced the laptop today
This analogy is not good. My cow produces heat, but I don't own the heat. My dog produces sounds, but I dont own the sounds. The laptop does not 'produce' anything in the way that a cow produces a calve. The laptop does not even become something else. The laptop ceases to exist altogether.
But if there is strictly laptop identity over time
Yikes, I've never said there can be. I said you can avoid the inconsistency Josh is generating in lots of ways. One 3Dist way is the presentist way on which Josh's (2) is false. There is, incidentally, a 'vague existence' I don't worry about, and that is an epistemological one. But you're right that I deny that things can exist vaguely in any interesting sense.
Mike,
I wouldn't analyze ownership in terms of strict identity for the reason you drew out. (Though I wouldn't mind it if you still came over :)).
Above you suggest that you never said there can be strict laptop identity over time. Does that mean you think there isn't strict laptop identity, and that per your comments on ownership you would permit me to take the one on your desk???
I'm sorry for being slow to understand, but I presently fail to see how a "presentist strategy" escapes (2). Could you elaborate? (Notice (2) doesn't say there could become just M.)
Alex,
Koons' line reminds me of my (b); I'll have to look into that, thanks.
Mike:
I don't own the heat or the sounds, because heat and sounds are not the sort of thing our society allows to be owned.
For a closer analogy, when my cow dies, I also own its corpse. The cow produces the corpse, in much the way that on the Chisholmian view, the earlier laptop produces the later laptop. In each case, the earlier item causes the later, and ceases to exist precisely when it succeeds.
Hello Mike, what evidence do you advance for the hypothesis that the universe had a beginning? I am quite curious, but I have no patience for conflation between evidence that space inflated outwards, distributing a singularity quite widely at an accelerating rate, for "the universe had a beginning." This would require you to subscribe to the unscientific belief that, prior to the Big Bang, there were no events anywhere, as if there is some certainty or even some evidence that said singularity was prior to everything and subsequent to nothing.
Along these lines, please provide references to evidence that the universe end that does not similarly betray a casual disregard for the findings of science.
For a closer analogy, when my cow dies, I also own its corpse. The cow produces the corpse, in much the way that on the Chisholmian view, the earlier laptop produces the later laptop.
When your cow dies, I don't think it ceases to exist. And when your plant dies, it does not cease ot exist. It ceases to be alive. Your cow is a material thing. It still exists, it is just no longer alive. That is, it's current stage is non-living matter.
On Josh's view, strictly speaking, my laptop vanishes at the instant the new laptop 'acquires' (and even that's not quite right, since the property is not acquired) a new property. But it does not cause the new laptop to acquire a new property. I might be the one who causes that.
Hello Mike, what evidence do you advance for the hypothesis that the universe had a beginning?
Right. I'm taking the standard scientific position that the universe began at the big bang. This does not entail that there was nothing before the big bang. It entails only that the universe did not exist before the big bang.
Mike:
I think that a dead cow or dead plant is only a cow or plant by analogy, as Aristotle would say.
Anyway, suppose I vaporize a plant with a powerful laser, or squash it into a paste by using a ten ton press. I then own the resulting gas or paste. And surely then the plant does cease to exist then, no? Or do you say that the plant exists as long as its elementary particles do?
Anyway, suppose I vaporize a plant with a powerful laser, or squash it into a paste by using a ten ton press. I then own the resulting gas or paste. And surely then the plant does cease to exist then, no? Or do you say that the plant exists as long as its elementary particles do?
The objection I was pressing was that the laptop today (on Josh's account) is not the one he purchased yesterday, and so not the one he owns. Suppose the cow does not survive its death. Still, what remains is something you purchased. You purchased the body of the cow too (assuming that is not identical to the cow). But Josh cannot argue similarly since what remains the next day is not a part of the laptop he purchased. Suppose a particle or two is exchanged in the laptop. Nothing remains of his laptop. All the parts that are present are parts of a different laptop. So there is nothing remaining of the object he purchased.
Now this is genuinely confusing me, Mike. You say that the universe began at the big bang, but that there could have been something prior to the big bang. What do you mean by "the universe?"
What do you mean by "the universe?"
I mean the spatiotemporally extended individual in which you and I and lots of other beings are universe-mates. There was no spatiotemporal things prior to the big bang.
Teapot,
If one defines 'universe' as the sum total of all reality (as some Internet atheists do), then if one says that the universe exists necessarily, one isn't in disagreement with a theist who thinks that a necessarily existing God is an essential component of reality. This is one reason to prefer the definition of universe Mike recommends (above).
But then our best science suggests that the universe had a beginning and so isn't necessary.
Joshua,
"2. Our best science suggest that the universe had a beginning, but a Necessary Being cannot have a beginning.
3. Possibly, our universe had a beginning, but a Necessary Being cannot have a beginning."
Both of these are false. Our best science suggests that our universe has a beginning. For convenience, let's suppose that a universe has a beginning just in case there is a first time-slice. I'd have to say something about what that means, but I think it's intuitive enough. There's no conceptual problem with the claim that that first time-slice exists necessarily, along with all the rest of them.
C, Thanks for those remarks. I suspect we have something different in mind by "exists necessarily." What I mean by 'x exists necessarily' is this: no matter what situation were to obtain, x would exist (either timelessly or presently). The first time-slice doesn't presently exist, and nor does it exist outside of time. So it isn't "necessary" in the sense I have in mind. So, the challenge is to come up with a reasonable causal principle that doesn't imply a necessary being (in this nuanced sense of necessary) but is simpler than any of the causal principles that do.
Josh:
Do you think your definition of a necessary being is plausible for a non-presentist?
Intuitively, that x is a necessary being shouldn't entail that either x is aspatial or x is spatially omnipresent. But then why should it imply that x is atemporal or temporally omnipresent? Maybe this is another one of those differences between space and time that are hard for me to understand (but may be real nonetheless).
There's no conceptual problem with the claim that that first time-slice exists necessarily, along with all the rest of them.
It sounds intelligible, but it turns out trivial. What is the trick to making objects with a beginning into necessarily existing objects? Easy: Assume that the actual world is the only possible world. If the actual world is the only possible world, then my pencil is a necessarily existing object, and my pencil has a beginning. What's even more interesting is that my pencil is a necessarily existing being and it will cease to exist! But this trivializes the notion of a necessarily existing being. We've simply collapsed the rather significant modal distinction between the possible and the necessary. Obviously, there is no interesting sense in which my pencil is a necessarily existing being, even if ours is the only possible world!
For Mike and Joshua,
Thanks for your remarks. I'm afraid that both of you are simply begging the question against the philosopher who holds that the physical universe exists necessarily. First, my claim was about time-slices and their sum, the universe, not objects like pencils. For objects like pencils, it makes more sense to require that it exists necessarily only if it exists at all times in every world. For universes and time-slices, that requirement is less obvious.
Josh says,
"What I mean by 'x exists necessarily' is this: no matter what situation were to obtain, x would exist (either timelessly or presently). The first time-slice doesn't presently exist, and nor does it exist outside of time. So it isn't "necessary" in the sense I have in mind."
I think there is an equivocation here between "outside of time" and "timelessly." Let's suppose for argument's sake that the four-dimensional universe exists necessarily. By your conception of necessity, we would have to say that it exists timelessly because it surely doesn't exist presently - the entire manifold isn't a part of 2010, but rather all times are parts of the manifold. So now we are four-dimensionalists who espouse eternalism, as most all four-dimensionalists do. But if the manifold exists timelessly, why can't we say, in some sense, that some of its parts do too? Given our assumption, I think time-slices would exist as timelessly as their sum would. After all, there need not be a hyper-time in which those time-slices themselves change. That's the sense in which time-slices would be timeless. The claim that a time-slice exists timelessly is compatible with the claim that it exists at a time. The timelessness claim is just a consequence of the conjunction of four-dimensionalism and eternalism. And the claim that a time-slice exists at a time, in the mouth of the four-dimensionalist, is just a trivial consequence of four-dimensionalism.
The philosopher who wants to block these sorts of cosmological argument probably has the following conception of necessity in mind: The universe (or any of its times) exists necessarily just in case it exists in every possible world. I don't see why one must also accept the further requirement that for any thing at a time, it exists necessarily just in case it exists at all times in every world. That just begs the question against the four-dimensionalist-eternalist in our current context.
So the main point is this. A four-dimensionalist - eternalist may claim that there is a first time-slice (whatever that means). That time-slice exists in every possible world. The physical universe of which it is a part exists in every possible world. So she may claim that there is a necessary being, the universe, and that it has a beginning.
And the claim that a time-slice exists at a time, in the mouth of the four-dimensionalist, is just a trivial consequence of four-dimensionalism.
How is that trivial? Let S and S' be distinct time-slices of universe U. S exists at t. S' exists at t' S is not S'. So there is a time, t', at which S' exists and S does not. So S is not timeless.
It's just a mistake, as far as I can tell, to reason from properties of the whole 4D object to properties of it's parts. Sure the whole 4D object is timeless. But how do you arrive at the parts being timeless? This is like saying the whole 4D banana is multicolored--in fact it is green in some places and yellow in others and black in still others--therefore there is some sense in which the parts are multicolored. That's obviously a bad inference. What makes you think the inference from the timelessness of the whole to the timelessness of the parts is any better?
Alex,
That's a good point; perhaps I should have said "exists" either in the tenseless sense or in the tensed sense (which for a presentist may come to the same thing, but not for a non-presentist).
c,
Those are excellent remarks. There's a lot to be said, but perhaps the simplest thing to say is just this: let's stipulate that if x is essentially temporal and has existed for just a finite extent of time, then it isn't necessary. Then, we can agree that the universe isn't "necessary" (in this stipulated sense). The challenge then is to come up with a causal principle that's at least as simple (non-ad hoc, etc.) as the ones listed, which, if interpreted in terms of this stipulated definition of "necessary", imply that there is a necessary being that isn't essentially temporal or has existed for an infinite extent of time.
Hi Joshua,
I'm not sure the stipulation helps any. The 4-dist I have in mind doesn't think that the universe itself has existed for a finite amount of time. The universe, i.e., the manifold, is eternal. And so the universe doesn't thereby count as contingent, according to your stipulation. She will, on the other hand, permit the claim that the universe has existed for a finite amount of time if that just means that there is a first or a last time-slice within the manifold. But then there's no reason to adopt your stipulation about necessity when reasoning about whether something "contingent" must or can be caused.
That time-slice exists in every possible world. The physical universe of which it is a part exists in every possible world. So she may claim that there is a necessary being, the universe, and that it has a beginning.
Its a strange conclusion since, not only is the universe a necessarily existing thing, but the first stage of the universe is also a necessarily existing thing. Presumably there are parts of objects existing in the first time slice and those parts of objects are necessarily existing things, too. But in what sense is an instantaneous part of an object a necessarily existing thing? That can happen, as far as I can tell, only if our world is the only possible world.
c, I'm with Mike. But per the other strategy of simply redefining "necessary existence", I took it as part of the definition that if there were a first slice, then it has existed for a finite extent of time. So, let me just stipulate that if x has a first slice then it isn't necessary. Then we can consider any number of causal principles that imply a necessary being. Perhaps they are less modest than before, but the challenge is still open to come up with one that's more modest, less ad hoc, accounts for all known causation, and that doesn't imply a necessary being (in this stipulated sense). No one has yet to even attempt giving a better causal principle than any of the ones listed...
Take for example the principle that any beginning (first time slice of x) can have an explanation. That principle is quite simple, accounts for a wide variety of cases, has no known counter-examples, and yet if there can be a first timeslice of non-necessary beings (beings that can have a first timeslice in some world), then by the argument given at the link, there is a necessary being that cannot have a first time-slice.
AHOY PHILOSOPHERS! THREE ALTERNATIVES
1) God is by definition eternal and infinite and creative.
2) What is the alternative? A cosmos that is also eternal and infinite in all directions. Not that our cosmos is equal to THE cosmos as a whole. Our time-space bubble could be one of an infinite number arising and declining throughout an infinite and eternal cosmic matrix.
One could believe either because in a truly infinite and eternal flux of matter and energy, evolutions of all sorts have the possibility of occurring, perhaps the inevitability of occurring, in that infinite and eternal flux. So "creativity" of a pre-programmed and directed sort is not necessary.
3) One other alternative to keep in mind is this, from Robert Anton Wilson:
What if everything's an infinite feedback loop in a cosmos of cosmoses that is a form of highly distributed intelligence (no hierarchical personal or conscious deity at the "top") that sometimes manifests itself in the evolution of life and sentient beings in the tiniest portions of the whole--on the thin atmospheric surfaces of a few fragile lifeboats bobbing in space--and those beings are part of a dance of equilibrium between life and death, since they all eventually die, along with planets and galaxies, and so forth and so on?
_______________________________
Here are some suspicions about the cosmos as entertained by the philosopher/novelist, Robert Anton Wilson:
I don't believe anything, but I have many suspicions.
I strongly suspect that a world "external to," or at least independent of, my senses exists in some sense.
I also suspect that this world shows signs of intelligent design, and I suspect that such intelligence acts via feedback from all parts to all parts and without centralized sovereignity, like Internet; and that it does not function hierarchically, in the style an Oriental despotism, an American corporation or Christian theology.
I somewhat suspect that Theism and Atheism both fail to account for such decentralized intelligence, rich in circular-causal feedback.
These suspicions have grown over 72 years, but as a rather slow and stupid fellow I do not have the chutzpah to proclaim any of them as certitudes. Give me another 72 years and maybe I'll arrive at firmer conclusions.
http://www.rawilson.com/prethought.shtml