Sunday, April 13, 2014

On cosmological arguments

A redated post. 

In making sense of questions concerning cosmological arguments, I think perhaps an important place to begin is to think through what kind of necessity can be attributed to the physical universe. The physical universe, at least if it is beginningless, can be considered to be factually necessary, and perhaps that concept needs to be clarified. I’d like to see a detailed definition of factual necessity.
Once we get this, we then have to ask if we have good reason to think that a universe that possesses factual necessity is unexplained in some important way that could be overcome by accepting theism. Is there a version of the Principle of Sufficient Reason that is rational to accept that gives us reason to think that a theistic explanation is nevertheless needed?
And then we have to ask what reason we have to suppose that the universe had a beginning. If the universe had a temporal beginning, how does that change the situation?


Jason Pratt said...

Maybe this would be a good place to suggest an archetypical naturalistic ontology argument.

I find, on reflection, that the argument (which is only seconardily atheistic in this case) runs like this:

1.) agree that infinite regression is not only conceptually worthless and a practical impossibility for application, but that even trying to state it requires tacitly presuming a single-IF (Independent Fact) reality is true instead.

2.) agree that the proposition of multiple-limited IFs (e.g.God/Nature cosmological dualism, or the cosmological tritheism proposed by some Mormons) either tacitly presumes a single-IF reality as a common overarching factor (which, incidentally, is why other Mormons treat the three Gods as existing within an already existant reality, even if that reality is technically supernatural to our own--"as man is, so God was", etc.); or else
the proposition makes absolutely no effectual difference to claiming one IF (Nature) exists instead (because of a total lack of interlocking or communion between the IF systems).

3.) either we should say Nature, then, is the IF; or we should say that Nature as a system is dependent for its existence upon something substantially different from it. (If substantially similar, it wouldn't really be anything other than a larger scope of Nature.) i.e. we should believe either naturalism or supernaturalism to be

It is important to note that the naturalist and the supernaturalist (regardless of theism/atheism differences) should be in agreement up to here.

At this point, the argument necessarily turns to the question, 'should we accept supernaturalism instead of naturalism?'

4.) if a ground cannot be supplied for properly preferring supernaturalism to naturalism, then we should accept naturalism instead, on the ground that needlessly multiplying hypotheses should be avoided.

From this point on, the naturalist has the default position (which is fair enough), and the burden of argument is on the supernaturalist to come up with a reason to accept supernaturalism instead of naturalism. Consequently, from this point on the naturalist's arguments will be directed toward countering supernaturalistic propositions, if possible; or accepting supernaturalism instead, if a proposition cannot be effectively countered _and_ if the naturalist accepts the importance of accepting the supernaturalist argument. (This is
because it's theoretically possible to come up with a venially trivial argument for supernaturalism that cannot be effectively countered.)

The naturalist and the supernaturalist are both at this point considering either an uncaused IF, or a self-causing IF. (If the proposed IF was caused by something other than itself, it wouldn't be an IF, and we would be committing a category error to treat it as an IF.) The options may be divided out as follows:

Opt1.) Only nature exists, and its existence is uncaused.

Opt2.) Only nature exists, and it continually causes its own existence.

(Note: there are technical reasons for why I am not including as an ostensible option 'X exists as the IF and began to be caused once upon a time but now exists uncaused.' Put shortly, this proposition would tacitly build in a larger system referent upon which X depends, thus self-refuting the proposition element 'X exists as the IF'.)

Opt3.) Nature exists dependently on another system, the existence of which is uncaused.

Opt4.) Nature exists dependently on another system, which continually causes its own existence.

The two non-causation options may be classed together under the heading of privative aseity. The two self-causation may be classed together under the heading of positive aseity.

Note that arguments for or against either privative or positive aseity should therefore be independent of (and perhaps logically prior to) arguments and/or conclusions in favor of naturalism or supernaturalism per se.


mattghg said...


Just trying to get my head round some issues. Does your 'uncaused' map neatly onto Victor's 'factually necessary', do you think? If so (or maybe even if not, I'm not sure :s) is an entity which 'continually causes its own existence' necessary in some stronger (logical?) sense? Also,

if a ground cannot be supplied for properly preferring supernaturalism to naturalism, then we should accept naturalism instead, on the ground that needlessly multiplying hypotheses should be avoided

OK, but surely our grounds for rejecting naturalism could have nothing to do with cosmological arguments? We might reject naturalism because of e.g. qualia or moral duties or Plantinga's EAAN.

Jason Pratt said...

Matt asks: {{Does your 'uncaused' map neatly onto Victor's 'factually necessary', do you think?}}

I think either uncaused or self-causing does. Nature might be factually necessary in a supernaturalistic reality, for subordinate purposes (i.e. in order to accomplish some purpose), but it couldn’t be ontologically necessary in a supernaturalistic reality.

{{is an entity which 'continually causes its own existence' necessary in some stronger (logical?) sense?}}

Strictly comparing the two, I would say yes, but it has to do with a lack of logical closure; i.e. I think there are formal grounds for preferring positive aseity as a belief. Aside from that, and apart from the question of theism (I find positive aseity more plausible given that I should believe theism to be true), the two options may be equally viable. I would like to see more work done on this, in the field.

{{OK, but surely our grounds for rejecting naturalism could have nothing to do with cosmological arguments? We might reject naturalism because of e.g. qualia or moral duties or Plantinga's EAAN.}}

First, despite the term he’s using, Plantinga’s EAAN has (I want to emphasize this) NOTHING to do with naturalism per se. This can be demonstrated from an analysis of his article on defeaters “Naturalism Defeated?”: after his first paragraph, where he defines naturalism ontologically (and properly so, I think, as far as he goes), he treats ‘naturalism’ as though it means ‘atheism’. (Lewis does the same thing in MaPS: after chapter 2, which really is more-or-less about naturalism per se, he’s talking about atheism, not naturalism.)

Consequently, I wouldn’t call the EAAN into play against naturalism anyway. To do so is a category error.

Naturalism, as I presented it above in the archetypal argument, has nothing to do with atheism per se--the claims are quite completely distinct. It has everything to do with a claim of ontological status for a system, namely the system of evident Nature. Consequently, unless an argument has some kind of ontological reference in it, I doubt it could be levelled against naturalism per se.

The question of supernaturalism/naturalism (and in fact other options, too, such as infinite regression or limited-plural IFs, mentioned in previous steps--no rigorous naturalist will accept those any more than a rigorous supernaturalist would), is altogether one of ontology: it’s a question of independence or dependence for systems.

The route I myself take, after the argument given above (which incompletely summarizes about 275 pages of argument, incidentally), leads me first through the atheism/theism question; after which I consider whether I myself am the sentient IF which I should believe exists. Since I am evidently not an IF at all, but am a dependently existent entity instead (pretty easy to discover {g}), this discovers a God/not-God distinction; and supernaturalism of one sort or another follows from there. A comparison of system properties of Nature, with my evident dependence on one hand and God’s active rationality on the other hand, sorts out which kind of supernaturalism I should believe--not a staggeringly original result, to say the least. {g}

There may be other ways to go about this (I mean reasoning out supernaturalism vs. naturalism), but as I said I’m doubtful it can be done without first discovering an ontological distinction. It will be noted that my application of the CosA is limited to inferring a single IF and eliminating infinite regression (plus either formally or practically eliminating limited-multiple IF proposals), which is why a naturalist (atheist or otherwise) should in principle be willing to go along with it.


Jason Pratt said...

eYeesh, "seconardily"... I wish the comments on blooger had an edit button... (only caught it on a recent re-read. Wonder why it didn't show up in spell-checker...?)