Thursday, 6 September 2012

Planting a Demigod

In this article the namesake of the WLC series, William Lane Craig, is only obliquely involved because he seems to have "borrowed" more than one idea from Alvin Plantinga without attribution.  However, the core concepts of the WLC series (the misuse of logic and conflation) are very much involved.  Plantinga’s speciality, however, is the misuse of probability.

In a debate with Lawrence Krauss, Craig indicates that he is setting out to show that the likelihood of god (given specific evidence and background evidence) is greater than 50%, implying that when any statement is more than 50% probable then it is somehow equivalent to a truth statement.  (Note that he immediately says that he will leave it up to the listener to determine the actual probability involved.)

More than 50% likelihood is sufficient to prove a thing to be true?  Really?

If Craig seriously believes the argument he is presenting here, an argument which is taken wholesale from Plantinga, we should all breathe a sigh of relief that the two of them are Christian apologists and not working in the area of medicine, or financial management, or road crossing.

Let’s look at how scientists manage likelihood.

Recently there was a flurry of excitement because physicists had announced that they had found evidence for the Higgs boson.  The excitement was not so much that they had something that they considered to be cast iron "end of argument" proof, but that they had a five sigma confidence level (equating to only slightly higher than a 99.9999% likelihood).  Earlier that month, July 2012, they had an even lower four sigma confidence level (equating to about a measly 99.9967% likelihood) which was naturally not enough to go public with, so they kept working releasing the five sigma data on 31 July.

A 50% likelihood is equivalent to a zero sigma confidence level.  In other words, no confidence whatsoever.

Let’s look at it another way.  Scientists, when reporting that a scientific claim is probably true, mean there is less than one chance in a million that the data they are reporting is bad.  Theists (of the Craig and Plantinga variety), when reporting that a religious claim is true, mean there is less than one chance in two (not two million, just two) that the data they are reporting is bad.

I’m not totally certain that Craig and Plantinga use this low standard in all parts of their lives but if they did, imagine having Dr Craig and Dr Plantinga as surgeons offering advice as to whether to go ahead with a procedure.

You: “Is it safe?”

Dr C and Dr P: “Yes, certainly!”

You: “So what are the chances that I might die from this procedure?”

Dr C and Dr P: “No more than 50%.”

You: “50%!!!!”

Dr C and Dr P: “Of course, either you’ll live or you’ll die, so 50-50.”

I strongly suggest seeking confirmation from your surgeon – or investment manager, or school crossing attendant – as to what qualifies as a certainty.  If they say “anything over a 50% likelihood”, get yourself a new surgeon – or investment manager.  If it’s a school crossing attendant, make sure that they are obliged to be first to walk out across the road after declaring that it is “likely” that it is clear of traffic.

That brings us nicely to evolution (via the natural selection of school crossing attendants).  Plantinga’s take on evolution will be the topic of a later article.


For those of you wondering about the title, a demigod is a half-god.  If you only have a 50% confidence in your god …

In another sense, Jesus was a demigod too, if genetically related to Mary.  Although, the child of a demigod might be better described as a quadgod.

1 comment:

  1. That's really funny. I watched that video a while ago and I do remember the 50% thing. I didn't think about it too terribly hard at the time, but the way you lay it out here really demonstrates how ridiculous a statement it is.


Feel free to comment, but play nicely!

Sadly, the unremitting attention of a spambot means you may have to verify your humanity.