Thursday, August 18, 2005

"God does not exist" is nonsensical

People who claim the god of the Bible exists (or does not exist) are linguistically confused. The verb to exist means to have the potential to have its properties observed. My teddy bear exists because it has the potential to have its properties observed. It is meaningless to speak of "spirit of teddy bear" which has no properties of teddy. Things cannot be said to exist in the absence of their properties.

Yet, the common definition of god is an entity whose properties can never be observed. So, to say that god exists is to claim that the entity whose properties can never be observed has the potential to have its properties observed, which is clearly self-contradictory. Notice that inserting a "not" in front of this proposition doesn't make it any more sensible.

Now, if one's god does have properties that can be observed, then presumably one's god theory is falsifiable. For, to be observable, the properties of god must be distinguishable from the properties of things which are not god. If god is more than just a conjunction of observations, then it is a scientific theory about a sequence of observations, and therefore, it has to be falsifiable. Of course, no one I know believes in a falsifiable god.

5 comments:

Robin Zebrowski said...

Does it change anything if the immediate object is forever unobservable but the consequences are observable? Like say, perhaps I can never ever observe greenhouse gases because of the type of body I have or technological limits we have (this isn't true, just the first example that came to mind), but we can observe the results of those gases. This is how so much science starts, actually. By observing results of something and assuming something else must be there, even if we can't see it (this is how Uranus was discovered I think). Anyway, I think most people who believe in god claim that the results are verifiable even if the immediate object can never be observed.

Of course, I'm just playing devil's advocate.

Doctor Logic said...

Suppose you have a set of observations. Each one can be represented by a vector of values in multi-dimensional space, where the dimensions are spacial, temporal, electrical, thermal, and so on through all empirical quantities. For example, you could have a vector {Colonel Mustard, Drawing Room, 5pm, 72 degrees F, pressure 29.92 in of Mercury,...}.

To claim that a finite set of observations is evidence of something other than the observations themselves, you must claim that there is a correlation between the observations and that this correlation has implications for other observations one might make. You can think of a theory as a choice of a curve in this space of observations that passes through the existing observation points (or within some error bars of it).

For any finite set of observations, there is an infinite number of curves that are consistent with the observations. A valid theory is a description of a curve that is adequate to test with additional observation points. If your theory makes no claims beyond the observations themselves, then it is equivalent to them. If it is not equivalent to the observations, then it must make a prediction.

All theories have invariants. For example, energy is an invariant in physics, as is angular momentum. Without invariants it is impossible to have a predictive theory because there would be no rule to constrain a curve in observation space. A theory of existence is a proposition that an invariant can be measured.

Teddy has several invariants including presence in a locale (Earth, North American continent) at a time period (late 20th, early 21st century), mass, specific heat, low-radioactivity, furriness, approximate morphology, stoicism, and so on. There may be an evil Teddy in the Crab Nebula, but he would be distinguishable by, among other things, being in the Crab Nebula.

Unlike Teddy, god has no invariants, no properties that constrain a curve through the space of observations. Because there are no constraints, every observation is consistent with god.

So, to refine my definition, "X exists" means that "X has a measurable invariant over some domain". If god had an invariant, one could make a prediction about observations, and falsify her existence. By definition god is not falsifiable, so there are no invariants. Hence "God exists" is equivalent to "that which can have no invariants will have measurable invariants over some domain."

Actually, now that I think about it, isn't my proposition looking like

(A intersection B = empty) and (A intersection B = non-empty)

which is actually false, not non-sensical? I still claim that the statement "God exists" is non-sensical, but for other reasons.

Rest assured, if the proposition were not meaningless, it would be false.

Dale Carrico said...

I find it interesting that "atheist" is widely defined as precisely the claim you rightly point out is nonsensical -- namely, a denial of the existence of (a) god. But etymologically, of course, "atheist" is simply a- theist, "without god."

I think of myself as atheist, but I mean by the term simply to point out that I "do without god" personally.

Of course the nonsensical character of either the claim that god exists or doesn't exist is perfectly characteristic of all manner of theological discourse -- to say "god exists" (or not) always means something on the order of "god exists, but in a way of existing that isn't like the way common or garden variety things exist otherwise at all but which I'll call existing nonetheless."

This is a tired dance move, indeed, by now, re-enacting for one thing exactly the gesture of the descriptive omni-predicates by means of which god is presumably characterized: omniscience, or a knowing that in being all-knowing isn't like any knowing we know of, omnipotence, or power that in being all powerful squares with no phenomenological experience of power available to us, or omnibenevolence, or a good that calls good things that by any standard are not at all good, but which somehow are kinds of knowing, power, or goodness after all.

Or, hell, why get fancy, it's like a kid who grappling after godhood decides god is an old man with a long grey beard in a big stone chair, only, you know, god, so I guess not so much like any old man who has ever existed after all...

Catechresis can never accommodate the supernatural, and so there are always only disanalogies... I quite understand that this opens the door for some powerful creativity and meaningfulness for some people. But for me... I do fine without god. "Atheist."

Tom FitzGerald said...

Refreshingly positivist.

But I like to think of God as existing in the sense that Hamlet exists in the play: as something in which disbelief is susepended as part of a cultural practice, the practices being theater and religion. So "God exists" has meaning within the religious drama, but not in the scientifically falsifiable world of fact.

"Hamlet exists" isn't a falsifiable claim of fact, it's just part of the fun.

leibniz said...

Hi Dr. Logic,
I've made these points already at Tu Quoque, but I'll repeat them here for the benefit of your readers.

1. To define 'to exist' as 'to have the potential to have its properties observed' is to rule the existence of entities like God out from the beginning by definition. What is to prevent the theist from simply rejecting this definition?

2. In your second paragraph, you seem to be arguing that 'God exists' is incoherent or self-contradictory. But I thought your view was that 'God exists' is meaningless or void of semantic content. If so, then it is empty, not incoherent or self-contradictory.

3. ...no one I know believes in a falsifiable god.

Perhaps few believe that 'God exists' is empirically falsifiable, but there are some who believe that 'God exists' can be falsified conceptually, by showing that the concept of God is implicitly incoherent. This provides us with at least one way of showing that 'God exists', if false, is falsifiable.