The Necessary Existent (and a refutation of Toasterism)

0 views
Skip to first unread message

Joseph Geloso

<josephfg@hotmail.com>
unread,
Sep 14, 2006, 2:58:02 AM9/14/06
to Atheism vs Christianity
Reposted from

http://groups.google.com/group/alt.religion.christian.roman-catholic/msg/b8bb4de5412e9cdd

Ted King wrote:
> In article <1157244974.934902.196...@e3g2000cwe.googlegroups.com>,
> "Joseph Geloso" <josep...@hotmail.com> wrote:

> > Ted King wrote:

> [snip]

> > I have now realized that to define God as Necessary Existent without
> > qualification is to define Him into existence, since that which
> > exists necessarily, necessarily exists.

> > So I have altered my definition accordingly, so as to be clear:

> > The concept referred to by the name "God" is the concept of a
> > Cosmologically Necessary Existent.

> > Thus, I am asserting neither logical nor ontological necessity.

> The "not logical necessity" part I can understand, but the notion does
> seem to be one of ontological necessity. At least it seems to me to be
> that it is if we consider "ontological necessity" to mean "an
> unconditioned existence".

That is a good definition, as far as I can see, the only remaining
question being, does one exist? Since we have defined the thing in
terms of its ontology, it seems we cannot then invoke ontological
necessity as any proof of its existence, because then we are guilty of
defining into existence, again. Cosmological necessity means that the
thing must exist _Given the existence of the cosmos_, which is the
qualification that saves us from defining it into existence. Now if, as
a result of cosmological necessity, a thing must also have ontological
necessity, then that follows. But since it was derived from
cosmological necessity _which must first be demonstrated_, it is not
"by definition."

There didn't need to be anything at all, but since there is something,
consequently we can infer that there must always have been something,
because it is incomprehensible to us (and thus not meaningfully
talkable-about by us) that something could come from nothing.

> > > [snip formalization]

> > > > In English:

> > > > 1. Ideaspace is properly dichotomized into possiblities and
> > > > impossibilities.
> > > > 2. All possibilities are ideas. (1)
> > > > 3. All impossibilities are ideas. (1)
> > > > 4. Possibilities are properly dichotomized into existents and
> > > > nonexistents, but not all nonexistents are possibilities.
> > > > 5. All existents are possibilities. (4)

> > > In the first version of your argument, you stated, "So of all
> > > possibilities, some are instantiated in reality, and these I call
> > > existents, while others are not thus instantiated, and these I call
> > > nonexistents." I think it is important that you keep that clarification
> > > in your argument, otherwise it is not clear by premises 1-5 whether or
> > > not "existent" refers to the ideas of ideaspace or to the "actual-space"
> > > referents of the ideas. If "existence" were taken to be about the ideas
> > > themselves, the argument would imply a self-contradiction. So, 5.,
> > > expanded to show the clarification, could be stated as: All referents
> > > that are actual-space existents are referents of ideas that are
> > > possibilities.

> > > > 6. Existents are properly dichotomized into contingents and
> > > > noncontingents, but not all contingents are existents.

> > > I believe you mean by the term "contingent" here what I call "causal
> > > contingency"; that is, where X is the sole or partial cause of Y, Y is
> > > causally contingent on X. This is in contrast to logical contingency;
> > > e.g., that the consequent of a conditional is logically contingent on
> > > the antecedent. Logical contingencies may refer to causal contingencies
> > > but not necessarily. There are logical contingencies that have nothing
> > > to do with causal contingencies.

> > I do not mean to limit contingency to causality. For example, there are
> > whole systems of causality (e.g. acorn -> oak -> acorn. . .) that are
> > contingent upon conditions (e.g. sun, water, air, Earth) not
> > necessarily their direct causes. (The sun, the Earth, the wind, and the
> > rain did not cause the generations of oaks.)

> It isn't clear to me how it is that pointing out that there are systems
> of causality explains what it means for you to say that you don't limit
> contingency to causality.

Something that is contingent relies on necessary conditions (relative
necessity). Causality, on the other hand, means sufficient conditions.
I think the set of sufficient conditions (cause) is the superset of all
necessary conditions, or any set of necessary conditions is a subset of
the set of sufficient conditions.

Thus: each of {Sun}, {Water}, {Air}, {Earth}, {Viable Acorn}, {Proper
Planting}, {Safety} and {Time} is a necessary condition for {Oak Tree}.
The set {Sun, Water, Air, Earth, Viable Acorn, Proper Planting, Safety,
Time} consitutes the sufficient condition (cause) for {Oak Tree}.
(There may be other necessities that I have failed to account for, but
this list will do for the sake of discussion.) Take away any one of the
relative necessities, and there can be no oak tree. That is what it
meant by saying the oak tree is contingent upon each of its relative
necessities. But Water, by itself, cannot produce an oak tree, so it
cannot be its cause. Really, only the whole set of necessary conditions
constitutes the cause.

So: causality is a specialized case of contingency. Every instance of
causality is an instance of contingency, but not every instance of
contingency is an instance of causality. Consequently, everything we
understand about contingency necessarily applies to every causality,
but there may be some further understanding of causality that would not
necessarily apply to every contingency.

> > > > 7. All noncontingents are existents.

> > > IOW, if an idea refers to something that is not causally contingent, its
> > > referent exists in actual-space. I think this statement needs to be
> > > buttressed with another statement that says that for a referent of an
> > > idea to be causal, it must exist in actual-space.

> > > > 8. There exists at least one noncontingent. (The set of noncontingents
> > > > is nonempty.)
> > > > 9. Nonexistents are properly dichotomized into impossibilities and
> > > > nonexistent contingents.
> > > > 10. Contingents comprise the existent and nonexistent contingents, and
> > > > may thus be taken as a single set.
> > > > 11. Ideaspace may thus be triply divided (trichotomized?) into
> > > > impossibilities, contingents, and noncontingents. (1, 4, 6, 9, 10)
> > > > 12. God is an idea, and God cannot be a contingent. (Definition)

> > > This is the statement that causes me the most trouble. It led me to
> > > think that you meant "existent" to refer to the ideas of ideaspace
> > > rather than to the referents of the ideas. In the process of analyzing
> > > the argument with that understanding of existent (ideas were to be taken
> > > as existents), it became evident that the argument would imply a
> > > self-contradiction. It wasn't until I reread your first version of the
> > > argument that I realized that that must not be what you intended by your
> > > use of "existent" here.

> > I had thought, but perhaps rashly, that my initial starting point in
> > Ideaspace would make it clear that I was categorizing ideas, not
> > realities, and that therefore the designations "existent" and
> > "nonexistent" referred to the existential reality or unreality,
> > respectively, of those ideas' referents ,

> Your point below about "referents" predicating words, not ideas is, of
> course, correct. As I said in my first response to the posting of the
> first version of your argument, my philosophic skills are really rusty
> (and I will admit, were never as fully developed as I would have liked).
> I got sloppy when I accessed my moldy memory banks and chose the term,
> "referent" - I should have been more certain it was appropriate before I
> used it. Anyway, I'm certainly open to suggestions for alternative
> terminology or phrasing. It certainly seems as though you are working
> with some kind of duality here (that Sphere understandably has
> reservations about given his ontological inclinations) and that there is
> some relationship between some ideas and "actual" phenomena that somehow
> correspond with some ideas. So it seems like there needs to be some kind
> of vocabulary to distinguish the ideas from the actual phenomena and
> also to describe the relationship between the two. I suppose it could
> very well be that the terminology you have already employed should have
> been sufficient and that it is just that I'm too dense to parse it.

> not to the ideatic existence
> > of the ideas themselves. It was assumed from the start that every
> > possible thing we can talk about is represented ideatically in
> > Ideaspace, so I did not think it at all necessary to treat of any ideas
> > that were not in it. I think it is plain that such treatment is
> > logically impossible. (By the mere fact of naming them we would be
> > bringing them into Ideaspace, if only as notions.)

> > > If God is an idea and existents are the actual-space referents of some
> > > ideas, then it would seem to follow that God, as an idea and not an
> > > actual-space referent of a God idea, could not be an existent.

> > I don't follow you. Do you mean to say that if the idea of God is a
> > fiction that God is not an existent? That would be true.

> I was thinking of "existence" as an aspect(?) of the actual "thing"
> rather than as an aspect of the idea of the thing. Maybe part of the
> reason I think of it that way is because if we think of an idea as a
> concept developed by the mind, then it seems as though we can conceive
> of a thing as being an existent even though it does not actually exist -
> and if we did conceive of the thing (have the idea of it) as existing,
> and if aspect "existent" is related to the idea rather than the thing
> itself, wouldn't we be in the position of saying that an assertion that
> the thing exists is correct, even though it doesn't actually exist
> because we conceive that the thing exists and existence relates to the
> conception and not the thing itself? That's really hard to express. I
> wouldn't be surprised if it didn't make any sense to you at all. I do
> appreciate, though, that there are the problems associated with this
> treatment of "existence" as though it is a predicate, so, anyway...

> ...what I was thinking is that "existence" is an aspect of the actual
> thing and not an aspect of the idea of the thing, thus if God is an idea
> it could not have the aspect "existence" (since ideas don't have the
> aspect "existence", what the ideas are about do). But I see now you are
> thinking of this in a different way. When you say an idea is an
> existent, what you mean is that what the idea is about exists in
> actuality. From that vantage point there isn't the problem I thought
> there was.

> > > One way
> > > to restate this so as to avoid that problem is: by definition, the idea
> > > of God has an actual-space referent that is not causally contingent on
> > > any other actual-space referent of an idea.

> > No, there again that would be attempting to define God into existence.

> > > But you can see that within
> > > the definition it defines God as an idea that has an actual-space
> > > referent that is an existent (essentially "defining God into
> > > existence").

> > Just so.

> > > So, of course, the conclusion of the argument will be that
> > > if the God idea is not an impossible idea then it is an idea whose
> > > referent is an existent.

> > Well, that is the conclusion of the orgininal argument. . .

> > > I anticipate that what you would say is that you do not intend to state
> > > that, by definition, the God idea has an actual-space referent.

> > Right.

> > > But as
> > > I've shown above, if "God" is only an idea, it can't be an existent.

> > Key word being "only." An idea in Ideaspace is classified as an
> > existent if it has a real (i.e. existentially real) referent.

> > Does this create too much confusion? Perhaps it would be better if we
> > avoided classifying ideas as existents at all.

> > To be clear: An idea is of an existent whose referent is a real
> > existent.

> > Then if "God" is an idea AND God exists, then the idea "God" is of an
> > existent.

> Okay.

> > > So
> > > God must be the referent of an idea and this referent is either an
> > > actual-space referent or not.

> > I looked up "referent" and I found that it is usually predicated of
> > words, not of ideas. But as far as I know, words are simply vehicles
> > for the communication of ideas.

> My apologies for my sloppiness.

No, it was my extension of the term. I originally began speaking of
ideas and their referents when I began talking about Ideaspace. There
is some precedence for that, as was pointed out to me, in - I think -
Frege.

But if words are never more than vehicles for ideas (whether, upon
further reflection, they are communicated or simply thought) then it
seems acceptable to use 'referent' in the way you and I have been using
it. It is intuitive to me, which is why I began using it in the first
place, and apparently at least you, and Frege, also thought so.

> > > I think the proper way to state this is to
> > > use a conditional rather than say, "God is..." But a conditional is not
> > > a definition and a conditional will not establish God as a causal
> > > non-contingent as is required for the argument to be valid. It appears
> > > that you either have to define God in a way that essentially says he
> > > exists, or state, via a conditional, that God, as an idea referent,
> > > having the actual-space condition of causal non-contingency is logically
> > > contingent on his existence - neither of which appears to get the job
> > > done satisfactorily to be used to make a cogent argument for God being
> > > an existent.

> > The concept of God is the concept of a Cosmologically Necessary
> > Existent.

> > > > 13. Therefore, God is either impossible or noncontingent. (1, 11, 12)
> > > > 14. Thus, unless God is impossible, God exists. (7)

> > > > The whole argument depends on 8, which states essentially that there
> > > > must be at least one necessary existent.

> > > I disagree. I think the cogency or non-cogency of the argument hinges
> > > more on 12. than 8. And I am far from convinced that 12. can be
> > > rectified or rationalized in a way to make the argument cogent. But
> > > maybe it can be. :-)

> > Replace 12 with:

> > "The concept of God is the concept of a Cosmologically Necessary
> > Existent."

> > Does that work?

> I don't think you can escape modal analysis if you use the modal terms
> "necessary", "possible" and "contingent". At the very least I think you
> need to make clear just what those terms mean. Do you intend them to be
> used as they are often used in "possible worlds" talk in philosophy? :

> http://en.wikipedia.org/wiki/Possible_worlds

I am highly suspicious of any concept that requires "possible worlds"
to discuss. What is the existential status of "possible worlds?"

Anyway, it seems to me the qualification of "necessary" by
"cosmologically" makes it clear what is being asserted.

I don't fully understand modal logic, and from what I can glean, there
are a whole family of modal logics, and it gets confusing. The original
investigation that led to my initial post in this thread was into
Goedel's Modal Ontological Argument, and Platinga's treatment of the
same using "possible worlds."

The original formulation of my understanding of it used possible
worlds, and it involved this step (or these steps):

1. If God is possible, then in some possible world, God exists.
2. In any world where God exists, God necessarily exists (by
definition).
3. If God necessarily exists in any possible world, then God
necessarily exists in all possible worlds.
4. Thus, God necessarily exists, in this world.

Well, I tried that out on one of my atheist friends, and he said, "what
the hell is a 'possible world?' You can't use "possible worlds" to
prove anything _in this world_."

I reflected upon that, and I admitted that it seemed to me he was
right, and I told him I would get back to him because my argument
needed work. Then I was able to reformulate it without invoking
possible worlds.

I have since grown suspect of possible worlds scenarios in general,
because how do we know they are even possible? They are purely
hypothetical, which means they exist in Ideaspace without any real
referent. Reality cannot contradict itself, but Ideaspace certainly
can. It may turn out that this is the only possible world.

Or maybe, all I have done is to subsume all "possible worlds" under
Ideaspace. But if that is what I have done, I like it better, because
Ideaspace is something with which we are already intimately familiar,
whereas "possible worlds" is a foreign, and to my mind, strained,
concept. (I mean strained as in stressed, not as in sieved.)

> I don't think now that I can give a thoughtful consideration to the
> replacement for 12. without having that question answered.

If modal logics are entirely dependent on possible worlds scenarios,
then modal logic itself might be suspect. Modal logic is a recent
development, thought it has its roots further back. I understand: ~ L ~
(x) -> M (x) and ~ M ~ (x) -> L (x), and that seems sound, but I do not
necessarily agree with Thomas Malthus's contention that we cannot even
use the English words "possibly," "necessarily," and "contingently"
unless we employ Modal Logic of some form or other, _especially_ if all
Modal Logics depend substantially upon these ontologically suspect
"possible worlds."

I'm not sure if I've answered you or not.

> But assuming for the time being that there are no problems with 12.,
> then I would agree that 8. is a central issue. But I also think 7. is
> not robust enough to stand on its own to be acceptable. I think there
> needs to be some justification for the assertion that all
> non-contingents are existents - at least it isn't so obviously
> self-evident that it doesn't need some explicit treatment to show why it
> should be accepted.

O.K. I will give that some thought.

> Getting back to 8., I haven't seen your argument for this assertion and
> I can't find it on any of the posts still on my ISP's servers (anyway,
> the ones I can access). I think there are at least three approaches to
> dealing with that assertion. One is to prove that there cannot be any
> non-contingents. That seems to be a very difficult, if not impossible
> job. A second approach would be to show that no one has convincingly
> demonstrated that there is at least one non-contingent so without that
> demonstration there is no reason a skeptic should be expected to accept
> that it is true. And a third approach would be to show that it is
> possible that are many, if not an infinite number, of things that are or
> could be non-contingent and therefore proving one particular thing is
> not contingent is not necessarily remarkable.

Let us define - with the caveat that we do not assert existence by mere
definition - God as the set G of all noncontingent entities. The only
way God could then fail to exist is if there are no noncontingent
entities, i.e. every entity in the Universe is contingent. Let us
define U as the set of all existents. If we assert G = U, then we have
asserted pantheism. If we assert, on the other hand, G !<= U (G is not
a subset of U), then we have asserted universal contingency (all
existents are contingents) and true atheism (there is no God, i.e.
there is no Necessary Existent). Between the extremes of pantheism (God
is everything) and atheism (God is nothing) there is every flavor of
Deism and Theism (God is something.) Or, G <= U. (G is a subset of U.)

I do not say that I have derived a disproof of pantheism, only of pure
atheism. Since G = U -> G <= U, pantheism is one possible flavor of
Theism or Deism. Atheism, on the other hand, as G !<= U, is incoherent
if there must be at least one Necessary Existent.

And it seems there must be, given the consideration of the following,
which I will copy and paste from my other post(s):

http://groups.google.com/group/alt.religion.christian.roman-catholic/msg/377d59644b5f6499

Lion Of Judah wrote:

. . .

> But why can't there be an infinite regress in a chain of
> contingencies?

Contingency is relative to necessity. Thus, for an oak tree to grow, a
viable acorn is a relative necessity. The acorn itself is contingent,
of course, on a previous oak, which itself grew from an acorn, and so
on. This is an example of cyclic causality, and of course the whole
system is contingent upon certain necessities as well, like DNA, and
Earth.

Now of contingencies, I find two types at least, one being causality,
and another being the contingecy of any cyclic or recurrent causality
upon a larger system, which has been named hierarchical contingency.

Just as causes are logically prior to their effects, so are relative
necessities logically prior to their contingents. Thus in any search
for relative necessities, I may stop at any point that I am satisfied,
and idenitify that condition as the prime necessity. Prime necessity
here is the equivalent of first cause in a discussion of causality. If,
in the case of the oak trees, I am satisfied with the explanation, "the
ecosystem of the Earth," then I may assign to that the position of
prime necessity relative to the existence of oak. Without the Earth, or
at least some place like it, no oak could ever have come about.

However, if I then ask, "how did the Earth come about?," then obviously
I can no longer assign prime necessity to the Earth's ecosystem - I
must move it back, perhaps to the formation of planets in whatever ways
they were formed. In every case I must either be satisfied for my
purposes of requiring an explanation or I must continue the search, and
in every case, when my search is ended, I can assign to the foundation
of all these contingencies the position of prime necessity.

Now I must ask, what can we mean by an infinite chain of contingencies?
If by it we mean a chain of contingencies supported by no prime
necessity, then I submit, none of them could ever have come about. By
definition, the whole chain from beginning to end depends upon its
first link, and if there is no first link, then there can be no chain.
There cannot be the whole series of contingencies which depends upon a
nonexistent prime necessity.

On the other hand, if by an infinite chain of contingencies we mean one
where the prime necessity is infinitely far back, then, I submit, there
is a misunderstanding of what is meant by "infinite." 'Infinity' is not
the name of any number, no matter how large. If the prime necessity in
our chain of contingencies is a something definite, then it can be
counted. We have so far been calling the prime necessity prime, meaning
first, but we can just as easily call it the ultimate necessity,
meaning, last, and thus we have reversed the order of our naming of
contingencies. Then we can begin with the contingency we have at hand,
whatever it may be, and assign to that the first position in the chain.
Then whatever it is contingent upon receives the designation of second,
and then third and so on, for each link in the chain. Since each
contingency is something definite, consequently it can be counted, and
so we may, at least in principle, assign an ordinal number to each one.
But then, which ordinal shall we assign to the ultimate necessity? What
is the number of links in that infinite chain? If you answer,
"infinity," I must simply remind you that 'infinity' does not name a
number. I think I am right to insist that the phrase, "an infinite
number," is incoherent. But what else could we mean by "infinite
regress in a chain of contingencies" apart from "a chain of definite
contingencies composed of an infinite number of links?"

http://groups.google.com/group/alt.religion.christian.roman-catholic/msg/dce80023153d7fc7

Thomas Robert Malthus wrote:

. . .

> Sophistry. You have in essence said "infinite regression cannot
> be the first cause". While this is true it avoids the fact that
> when "infinite regression" is used in answer to the first-cause
> question it is used to *negate* the question, not to answer it.

What an interesting technique! I was not aware that that avenue was
open to us. So if anyone presents me with a question I have no answer
to, I can just "negate the question," and be done with it? Or is that a
privelege you extend only to yourself?

> In fact it also negates the definition of causality that you
> assert when you say "by definition, the whole chain from
> beginning to end depends upon its first link". The ordinary
> definition of causality is that things come into being in
> response to prior events. That definition in no way requires
> that the chain of prior events be finite. The dependency upon
> the first link that you assert only exists when the first
> link is presumed.
> In short, you have not reasoned to a first cause; you have
> merely tried to (self-referentially) define one into being.

Nonsense. My argument above is based on the requirement of a prime
necessity in any chain of contingencies, which requirement is built
right in to the definition of contingency. Every contingent existence
is supported by some relative necessity, and if we are satisfied at any
given point and do not wish to pursue the matter further, then our
arbitrary stopping point can be assigned the designation, prime. If we
go farther back, then we move that designation farther back. In no case
is it possible that a chain of contingencies arises unsupported; if it
were to do so, then it could no longer meaningfully be called a chain
of contingencies, since, resting as it did on no necessity whatsoever,
but existing, the whole chain itself would have to be considered
necessary.

As you yourself said with regard to causality, "things come into being
in response to prior events." A chain of occurences is a set of
discrete occurrences, and if the relationship of each member of that
set is as a contingency, then it must be contingent on either some
other member of the set or on some occurence outside of the set. By
definition of logical priority, which must exist in every relationship
of a relative necessity to its contingent, there must be at least one
member of any set of contingencies only whose logically prior relative
necessity is outside the set.

(end quoted material)

> There is also the issue of how to take it that the physical universe as
> a whole may be non-contingent. Now you could define as a god any
> non-contingent, so that saying that the physical universe is
> non-contingent implies that a god exists. But that is really just
> definitional sophistry.

No - it is pantheism, by definition.

> Most of us already except that the physical
> universe exists and tagging the label "God" onto it doesn't mean much of
> anything. Most people would accept my claim that my toaster exists. I
> could define "God" as being my toaster. Therefore if you accept that my
> toaster exists, you must accept that God exists. That doesn't amount to
> anything meaningful.

Because you cannot demonstrate the absolute necessity of your toaster.
And, you could apply a similar line of reasoning to the objects of the
world. I think it is plain (unless you want to follow Sphere, who might
just be a nihilist) that "the Universe taken as a whole" does not name
just one thing, and that a collective such as the Universe is requires
consideration of its composite parts, some of which may be contingent
upon others - which places us squarely in the situation addressed in
what I pasted from my other posts, above.

Tagging the label "God" onto the Universe as a whole is the definition
of pantheism. On the other hand, tagging the label "God" onto Necessary
Existence might simply be the definition of "God." It is the working
definition for the purposes of this thread.

If there is no necessary existent at all, then by definition there is
no God. But if the cosmological argument obtains, the existence of God
so defined is proven by it. That is the point of making that argument.

"Necessary Existent" necessarily exists in Ideaspace, right along with
Sphere's "little green men who don't exist," and also, "little green
men who _do_ exist." Ideaspace contains both realities (ideas whose
referents are real) and fictions (ideas whose referents are not real).
How do we further classify these three ideatic existences?

1. "Necessary Existent" refers to a reality IFF at least one entity
exists necesasarily, i.e. noncontingently.
2. "Little green men who do not exist" is an Honest Fiction. The little
green guys don't exist, but at least they acknowledge their
nonexistence.
3. "Little green men who _do_ exist" is a Dishonest Fiction. They do
not exist, but the idea in Ideaspace asserts their existence, which is
a lie.

> I believe I recall you mentioning in a post that if atheists were to
> accept that the physical universe as a whole is a non-contingent that
> that actually makes them pantheists and not atheists. But let's look at
> the definition of "pantheism":

> 1. the doctrine that God is the transcendent reality of which the
> material universe and human beings are only manifestations: it involves
> a denial of God's personality and expresses a tendency to identify God
> and nature.
> 2. any religious belief or philosophical doctrine that identifies God
> with the universe.

> Accepting that the physical universe as a whole is a non-contingent does
> not imply acceptance of 1. And 2. is kind of mushy-headed. If one
> believed that the physical universe as a whole is a non-contingent and
> someone said that God is defined as that and therefore the physical
> universe is God, that wouldn't carry any more import than me defining my
> toaster as God and if you believe my toaster exists you are some kind of
> theist. These are just examples from the class of arguments that diddle
> with the notion that atheism is the position that no gods exist and then
> go about defining things as gods that are not normally considered to be
> gods. Of course what atheists almost always are referring to when they
> use the term "god" is a supernatural (and also could be transcendent
> from the physical world) agent that is not bound by natural laws.

That is fine. My argument has not yet approached even asserting such
qualities of God. So far, the definition of God I have advanced for the
purposes of this thread perfectly fits Pure Materialism (The only
necessary existence is Material Energy, from which all life including
human life has evolved, and which could simply oscillate over eons of
time in a cycle of Big Bang, Expansion, Contraction, Big Crunch, Big
Bang. . .)

> I will admit I am going from a memory of what you said about pantheism,
> though, and obviously my memory has been shown to too often be faulty,
> so if I mischaracterized your point, I apologize.

When I say Pure Atheism is incoherent, there is even a caveat in that:
Pure Atheism is incoherent given the validity of our reason including
the principle of contginency and necessity. There may be other
considerations that would invalidate Pure Atheism even without that
principle. But I do not think I could argue with a nihilist, because
where would be the starting point? How would you argue with someone who
does not accept the validity of logic or reason? Pure Nihilism would
reject these along with everything else, and Pure Nihilism is thus a
form of Pure Atheism - I don't know if there are other forms of it.

Pure Materialism, on the other hand, is a form of Pantheism, and it is
the form defined in what you have given as definition 2 above. Pure
Materialism (which can also be called Pure Physicalism) is
Impersonalist Pantheism. And most of those who call themselves atheists
are really, it seems to me, Impersonalist Pantheists.

As I have stated previously, I'm not clear yet on just how far Natural
Theology is capable of proceeding. But I believe its first step is to
acknowledge the fundamental impossibility of there being no Necessary
Existent, and even though we are including so-called atheists among
those who accept that, I have no problem with that. I mean to include
everybody, insofar as that is possible.

Your "Toasterism" is incoherent on the grounds that your toaster did
not always exist.

Reply all
Reply to author
Forward
0 new messages