Wednesday, April 06, 2011

How a theist defeated an atheist

Via Cosmic Variance

If you have 2 hours and 20 minutes, you may try to watch a debate about "Whether there is evidence of God". Theologian William Lane Craig represents the believers; cosmologist Lawrence Krauss is the atheist.

The first video in the playlist (with 11 items) contains nothing else than music. Go to 2:10 of the second video.

As Sean Carroll admits, the theologian won the debate when it comes to the actual formulations, compactness of the arguments, and rhetorical skills in general. That's why atheist pundits are collectively scared of Craig. In comparison, Krauss seemed like the guy who had no idea what he's talking about and why he's talking about it. Compared to Craig, Krauss is an impossibly lousy, stuttering speaker. He didn't really convey the key points, so even if you a priori strongly agree with Krauss's position, you will find it very strange to declare him a winner.

Needless to say, I chose my title because Craig was voted to be the winner - a fact that could reflect a largely religious audience in North Carolina (NCSU) but it didn't have to be the actual explanation. Don't get me wrong - Krauss has also made points that made me laugh and that were appropriate. I liked his comment how easy it would be for God to make miracles that prove His existence. It was nice to see Krauss as he defends that the sum of positive integers is -1/12, and the maths is only consistent in this way.

I would probably agree with Krauss' comments about the surprises of quantum mechanics - although his presentation wasn't really clear. But I don't understand the relevance of these remarks for the question about the "evidence for God" and I agree with Craig that the Universe is ultimately free of logical contradictions so the mysterious talk about quantum mechanics can't really suppress the power of logic (well, QM denies realism, but that's more than just "pure logic"). In this context, Krauss just emitted some fog.

Although I would have to count myself as an atheist, Craig has made some important points I agree with - but also many points I disagree with. Craig is right that one must fully respect mathematical logic, even when it comes to metaphysical debates such as this one, and he has perhaps even caught Krauss as offering some logical fallacies.

I also agree with Craig that if the conditional probability of an observed fact is predicted to be higher with God than without God, the observation is evidence for God (replace God by anything else, if you wish). However, many of his pieces of evidence of this kind are either fishy or damn weak when it comes to the number of units of evidence. ;-) Also, Craig may be overlooking lots of comparably or stronger pieces of evidence against God. ;-)

I don't really believe that one can prove or disprove the notion of God in His or Her or Its most general sense - and I wouldn't participate in a debate whose topic were this ill-defined. But many related statements that are much more accurate may be addressed by arguments based on the existing empirical evidence - and logic itself.

Craig has made some incorrect and some "at least" unjustified assumptions, however. For example, he seems to believe that the notion of an "infinity" is automatically a symptom of inconsistency, even if mathematics may build axiomatic frameworks that avoid such an inconsistency.

This is a very subtle assumption because he implicitly believes that the "real world" and "physics" have higher standards of what "consistency" implies. Well, in general, I even agree that this is the case but I am not sure whether it's the case in the particular situations that have been discussed.

For example, there was an important question whether the Universe could possibly last eternally - from "t" equals minus infinity - just like physicists thought until the early 20th century. This is a subtle question and the answer is not as clear as some people on both sides could think.

First, Craig is surely wrong that one couldn't even consider a spacetime that existed in the infinite past - just because in such a Universe, we couldn't count the "time since the beginning" as a finite number. The reason why this constraint doesn't exist is that there doesn't have to be any "time since the beginning": the Universe may be parameterized by a time coordinate that may be real - both positive and negative - and arbitrarily large negative values of "t" may still exist.

There is certainly no contradiction at the level of spacetime geometry! Craig and other theologians - and not only theologians - are just mixing their prejudices about the reality with actual proofs of inconsistency - which they don't possess. In some sense, their perception that there is an inconsistency follows from their emotional problem to work with negative numbers. When it comes to the real and important questions, it seems to me that they think that coordinates such as "t" always have to be positive.

On the other hand, a time coordinate that could have been continued to the infinite past is a potential source of inconsistencies; it becomes clear if you look at some detailed laws of physics. In the 19th century, the Universe would be thought of as being infinite both spatially and temporarily and there is no real problem with this assumption.

However, if you were imagining that the Universe is spatially finite - the volume is finite much like the volume of the currently visible Universe - a temporally infinite Universe would be a problem for physics. It would be a problem because of the second law of thermodynamics.

The entropy of the visible Universe today is something like 10^{100} "nats" and it is never decreasing. In fact, it has to substantially increase all the time. When you retrodict the past, you will find out that the entropy of the portion of the Universe that evolved to the currently visible Universe was much smaller a fraction of a second after the Big Bang. When you look at an even earlier moment, you will find out that the entropy was de facto zero. Because quantum mechanics prohibits negative values of the entropy - the entropy is the logarithm of the number of microstates which can't be e.g. 0.0000001 - there couldn't have been any "past" before the Big Bang.

An alternative scenario would be a quasi-equilibrium - the state of the Universe that could sustain a nearly constant, very low entropy for eons, but suddenly decided to "melt" and increase the entropy again. Such a quasi-equilibrium state is arguably impossible in physics, especially because its entropy should have been much lower than what it became later which contradicts the usual states of equilibrium.

Of course, there is one more realistic alternative: the eternal inflation. In eternal inflation, you may extrapolate time to minus infinity again - and there's no contradiction with the second law because every time you get to the beginning at which the entropy of your Universe is nearly zero, you will figure out that the Universe you knew is just a tiny bubble in a much bigger parent Universe whose entropy was, is, and will be much larger. So there's new room to reduce this entropy in the parent Universe, and so on.

Effectively, eternal inflation restores the 19th century possibility that the Universe may last eternally. Of course, the multiverse as imagined by eternal inflation is nothing like "Einstein's static Universe" or its 19th century predecessors - simply because it is not static. However, all eternal models of the Universe may lack an explanatory power which may fail to be a "sharp and inevitable" inconsistency but it could still be a problem for a physical theory.

(By the way, in the case of cyclic models of cosmology, Craig is completely right that one may only add new steps but he can't eliminate the need for the existence of a "beginning".)

At the end, I think that all such things must be studied by the methods of physics - or cosmology - and Krauss still arguably knows more about the stuff than Craig does, despite Krauss' weak rhetorical skills and his somewhat self-evident atheist prejudices. There will always be a difference between the different cultures - between theologians and philosophers on one side; and physicists and cosmologists on the other side.

There are aspects in which my respect to the theological and philosophical communities is nearly zero or negative; and there are aspects in which it is significantly positive. Those people tend to be educated in humanities and they may be great speakers - as Craig's example shows. On the other hand, they often use lots of verbal constructs that are meaningless and that only "look" clever, and they often ignore or deny the genuinely important arguments that may be derived from physics and its empirically rooted logic.

I was reminded about the world of the "wise men" of philosophy when I was giving the talk show at the Prague Jewish Community a week ago. Although the host told me that only very old Holocaust survivors would attend, there were also at least two young students of philosophy at the Charles University, among other young participants.

There were many points, questions, and exchanges about cosmology, string theory, and its testability - all the things you may imagine in such contexts. But the philosophers have asked me a somewhat unusual question - whether string theory was "normative or descriptive".

Of course, I was lost so I said that I had no clue (much like I said when I was asked about a giant laser that is going to be built somewhere in Czechia - I didn't follow that story). The last time when I heard about the "normative vs descriptive" dichotomy was about 20 years ago, during a philosophy (Citizens' Teachings, "Občanská nauka") class at the high school. It has never made much sense to me.

According to some explanations, it looks like "normative" is about "how things should be" and "descriptive" is about "how they are". Obviously, all science is "descriptive" in this sense - science never defines the right moral values, among similar things.

However, there was a sign that the "descriptive" meant an effective theory while "normative" was an accurate model identified with the reality. As long as string theory is the accurate theory of everything, it is no longer "effective" and could be even counted as "normative".

So even after a week, I still don't know what the right answer is to the question. I don't really understand why one would ask these confusing questions with somewhat ill-defined words. There exist similar questions that physicists could ask and that could actually be addressed. For some reason, philosophers don't seem to prefer questions that are more likely to produce valuable and/or decisive answers. They don't seem to care whether their questions may become a starting point for some progress.

Also, I was surprised to see that the philosophers must have thought that whether a theory was normative or descriptive was a question that scientists in general and philosophers in particular would be solving (almost) all the time. I forgot to emphasize to them that it wasn't the case. Most physicists would never use the term "normative theory" in their whole life.

This not-so-subtle fact is obviously being hidden from the students of philosophy. The students of philosophy are apparently never honestly told by their instructors that scientists consider speculations about "normative theories" to be ill-defined and meaningless verbal masturbations. So they don't know and they must be surprised if they meet scientists who are not obsessed by the normative vs descriptive dichotomy.

Back to Craig and Krauss. I found some of the arguments by Craig creative and witty. For example, Krauss would say that there's no "objective morality" but still, science depends on "honesty, open-mindedness" and other moral values. Craig pointed out the contradiction; I agree, there is a contradiction although it obviously exists only because we're combining different aspects of morality - and their different purposes - under the same umbrella.

I also agree with Craig that in order to claim that Jesus's resurrection is an extraordinary claim (one that requires extraordinary evidence), one needs to presume that God doesn't exist. After all, the religious viewpoint is not that the resurrection was a natural process - which would be highly unlikely to follow the Biblical scenario, indeed. Instead, the resurrection was done by God.

So instead of having a "purely empirical" evidence against God and/or resurrection, an atheist scientist has to apply a kind of Occam's razor here. The assumption is that the natural laws don't have to be supplemented by an extra deity - because there hasn't ever been a good observed reason to be sure that it is needed. Well, I agree that the opinion that such extra miracles don't exist is a belief, much like the religious beliefs. Still, any statement that the miracles occur at least "somewhat regularly" with a frequency higher than "f_0" may be empirically falsified, so if you measure the might of God by "f_0", science can make God arbitrarily tiny. ;-)

Of course, if one assumes that the miracles are restricted to special, irregular events (such as the resurrection of Jesus) that don't include a sizable fraction of cold scientific experiments (where the miracles never show up in a statistically significant way), be my guest. One can't really falsify this theory - this belief.


  1. An agnostic would argue thus: "Any evidence of God's existence would change religious matters from a question of faith to a statement of fact, destroying the very purpose of faith itself. Thus God can provide no direct irrefutable evidence of himself if he wants to preserve people's faith in him and he can only speak indirectly through signs and prophets which all can be only understood through your faith, not reason. Trying to use reason to understand God or to ask about existence of God is futile as he is by his very nature unknowable. The idea of God (in Plato's sence) clearly exists as we talk about it. That is as far as reason can take you. Past that point you either believe or not. The choice is yours, a matter of free will."

  2. In this debate with Harris

    Craig does not come out nearly as well as with the one you present.
    Harris's views are extremely interesting in themselves.