Monday, 13 February 2012

William Lane Craig's Premises and the Kalam Cosmological Argument

William Lane Craig seeks to put forward deductive arguments for the existence of God.  

If Craig gets his logic right he will have produced a valid deductive argument. If an argument is valid then if the premises are true the conclusion is true.  Validity is not enough though,  Pooh's argument here is valid:
"And if anyone knows anything about anything," said Bear to himself "it's Owl who knows something about something," he said, "or my name's not Winnie the Pooh," he said. "Which it is," he added. "So there you are." (Milne, 1977)
We can only be sure of the truth of the conclusion if the argument is valid and we are sure of the truth of the premises.  Now Craig cannot, none of us can, establish the truth of his premises beyond doubt.  What he seeks to do is adopt premises that are “more plausible than their negations” (2010

This is problematical.  Accepting that the premise adopted is more plausible than its negation may lead the reader, or debate attendee, to believe that the premise adopted is the most plausible premise.  Were the only options to be that premise and its negation this would be the case.  Take, however the Kalam Cosmological Argument:
1. Everything that begins to exist has a cause.2. The universe began to exist.3. Therefore, the universe has a cause(Craig 2010)
Other premises

There are a number of alternative first premises other than "not everything that begins to exist has a cause", each with good arguments for its adoption:

Alternate A: Not everything has a cause
Argument for: We might consider that "everything has a cause".  However causes happen before effects, meaning that each effect has a predecessor.  As each cause is "something" and requires it's own cause, a cause which also requires a cause further back.  The regress, if everything had a cause, would never terminate.  There are good arguments for this being impossible: meaning that as there is something not everything had a cause. (Note, for later, that if an effect can precede a cause or be simultaneous with its cause there is no obstacle to everything having a cause")

Alternate B: Everything we can understand has a cause
Argument for: Though we speculate about entities we don't understand we shouldn't introduce that speculation into a deductive argument as, well, we don't understand them.  We can confidently say that anything comprehensible has a cause as causality is part of the way that we understand things.

Alternate C: For any X : if a Plank second of time occurred before X's existence then X had a cause.
Argument for: Again, causes happen before effects.  It appears that a Plank second is the shortest period of time any cause can be before its effect.  For any effect, therefore, there has to be at least one Plank second before it for its cause to be in.  

There we have three alternatives to the first premise given by Craig.  

Whilst it is difficult (probably impossible) to quantify plausibility, each of the alternatives appears to be at least as plausible as the premise given by Craig.  It would be difficult to argue that together they are not considerably more plausible than the premise given by Craig.  To argue that they were not more plausible would require arguing that:
"Alternate A or Alternate B or Alternate C" is more plausible than "Craig's first premise" 
Now "Alternate A or Alternate B or Alternate C" fails to establish the conclusion "the universe had a cause", whilst Craig's first premise does.  But  "Alternate A or Alternate B or Alternate C" is more plausible than Craig's first premise.  Limit our choice to these alternatives and the following is true:

The most plausible conclusion is that Craig's conclusion has not been established

What Craig is arguing for.

Supposing we come to the debate holding Alternate C.  What must Craig show in order to establish his contention that the universe had a cause?  Some of our position (Alternate C) and Craig's overlap, its just that that part of Craig's position fails to establish his conclusion.  What Craig needs to convince us of is the difference between his first premise and ours.  What is that difference? We already agree that all stars, all planets, all living things, all atoms and sub atomic particles, all light and, if they exist, all dark matter and dark energy have causes.  We accept that everything arising since the first Plank second of the universe's existence  had a cause.  

Our dispute is about the first Plank second of the universe.

And what is the first Plank second of the universe's existence?  Is it anything more than the beginning of the universe?  Is there anything else to attribute a cause to other than "the universe"?  In order to persuade us to accept his premise over ours does Craig not have to persuade us that the universe had a cause?  In other words, to establish that the universe has a cause he first has to establish that the universe has a cause in order to establish his premise that everything that began to exist had a cause. 

And what does he have to do to establish this?  He has to persuade us that the universe had a cause even though there was no time for the cause to be in.  If he succeeds in doing that he must either allow simultaneous causes (which allows us to curtail the infinite regress mentioned in Alternate A, which allows "everything has a cause") or exempt the universe from this rule.  If we can exempt the universe from the rule to have a prior cause we can just as well exempt the universe from having a cause.  


It is the dichotomy created by the "more plausible than its negation" that allows other, more plausible, alternatives to be ignored and Craig's need to argue for a cause of the universe tout court to remain hidden. It's the dichotomy which allows a startlingly implausible argument to look plausible.

Craig, W. L. (2010, February 4). Five Arguments for God. Retrieved February 8, 2012, from The Gospel Coalition:

Milne, A. A. (1977). Winnie the Pooh. New York: Dell Pub. Co.


Mike D said...

Interesting angle. Craig has on several occasions defended the "simultaneous causality" argument.

Tony Lloyd said...

I found an example of Craig's arguing that causes can be simultaneous with effects:

Seems strange to me:

1) lack of simultaneous causes means we can allow "everything that exists has a cause". If we're allowing "everything exists has a cause" then if God is per definito uncaused, then God does not exist.

2) the persuasive force of his argument depends on us thinking that there is no time before the universe for a cause of the universe to happen: so we have to look to a timeless God to cause it. If we allow simultaneous causation we can posit quite mundane causes for the unverse (the anti-matter caused the matter and the matter caused the anti-matter)

Thomas Larsen said...

"Alternate A: Not everything has a cause"

Craig grants this, like most philosophers who have studied the KCA: permanent, necessary things require no causes. (God, on the theistic view, would be such a thing.)

"Alternate B: Everything we can understand has a cause"

Isn't this kinda special pleading, though? I mean, the whole project of science is based on the concept of things (even things that we really don't understand) having causes. There doesn't seem to be any reason to adopt "Everything we can understand has a cause" over "Everything that begins to exist has a cause."

Moreover, everything we can understand has a cause? Really? This would lead to interesting arguments like the following:

(1) Everything we can understand has a cause.

(2) God does not have a cause.

(3) Therefore, we can not understand God.

(That's right, take that, problem of evil!) ;-)

"Alternate C: For any X : if a Plank second of time occurred before X's existence then X had a cause."

This seems pretty arbitrary, and not broad enough. Suppose a person puts his hands on a car and pushes it. The car's movement is (essentially) simultaneous with the person's push, but the push is the logically prior cause of the car's movement.

Now, I don't use the KCA, mainly because I'm not so sure whether the A-Theory understanding of time that it requires is correct.

Tony Lloyd said...

Hi Thomas

On Alternate A, the question would be if it were accepted then why not choose this as the first premise? (One answer would be that it does not establish the conclusion sought!)

I don't see that Alternate B is special pleading: nothing is being claimed about non-understood entities. It's surely not special pleading to shrug ones shoulders and say "I don't know".

Yes, it would mean that we can't understand God. Physicists don't understand the first Plank second of the universe but it doesn't stop them drawing negative conclusions about what it wasn't (it wasn't, for example, a universe of whipped cream). I don't understand why anyone would support Liverpool Football Club, but they do and it's evil: the problem of evil is back on! (Or add "we can understand God" to your argument and conclude that God does not exist).

I chose the Plank second because I'm told that that is the shortest period of time. This is the option if we think that causality is prior in time (yes the car's movement is near enough simultaneous: a Plank second is near enough no time at all!) We could just say "anything with a certain (non-specified) time before it has a cause"

But whatever you think of the alternates, Alternate B and Alternate C are entailed by Craig's first premise. Craig accepts them and the question I asked of Alternate A goes for Alternate B and Alternate C: why not adopt them as the first premise?

Tony Lloyd said...

Actually, Craig doesn't accept Alternate B/Alternate B isn't entailed by Craig's first premise.

Thomas Larsen said...

Out of interest, (why) do you think premise (1) of the KCA is false, or more probably false than true?

Tony Lloyd said...

I don't have a problem with Craig's first premise per se, I have a problem with his methodology.

There is a whole slew of premises for any given argument which are plausible. The methodology of adopting one that is more plausible than its negation fails to cut down the number of options, and totally fails to identify one premise.

It leaves you with only marginally smaller slew of premises some of which conclude the way desired, some of which don't and free choice between them.

The premise is one chosen from a large number of "live" options because it is liked.

Thomas Larsen said...

But if a premise in an argument is altered, a new argument is created, surely?

I don't see why the process of premise-selection should matter much to someone merely trying to evaluate a particular argument. If you accept both premises of the KCA, you're rationally committed to the conclusion.

Do you see any particular problem with simultaneous causation? I don't see how the possibility of simultaneous causation would affect arguments against an infinite regress of causes.

Tony Lloyd said...

I don’t accept Craig’s first premise. I don’t have a problem with it, but then I don’t have a problem with other premises that reach different conclusions to Craig’s. Why should I accept Craig’s premise over other premises I also don’t have a problem with? Craig doesn't have a decent answer to that. His formulation of “more plausible than its negation” fails to give a reason to adopt it over all the other descriptions of the scope of causality that are more plausible than their negations.

The problem with simultaneous causes is that they open up the possibility of symmetric causes: A causes B and B causes A.

If A causes B and B causes A then, for a universe of A and B:

1. Everything has a cause and,
2. Nothing distinct from the universe caused the universe

With simultaneous causality the conclusion of the KCA would still stand: the universe would have a cause. But we wouldn’t have to hypothesise something outside of the universe to be that cause. The conclusion would stand, but the whole point would disappear.

Jonathan MS Pearce said...

Hi there

Thought you might be interested in a set of objections to the KCA. I am writing a paper on the KCA and WLC and am starting a series of blog posts on these objections. You might like to check them out:

Sorry for randomness