Tuesday, December 04, 2007

Paul Davies: "Too Busy" Writing Crappy Op-Eds to Justify Claims

Paul Davies, the British physicist and popularizer of science, wrote an astonishingly silly op-ed in the New York Times recently, in which he equates science and religion because both are based on "faith". It was a pleasure to see Davies' ideas completely shredded by Lawrence Krauss, Sean Carroll, and P. Z. Myers.

This isn't the first time Davies has said silly things. In The Fifth Miracle, for example, he attributes the ideas of algorithmic information theory to Gregory Chaitin, despite the fact that the Soviet probabilist Andrei Kolmogorov came up with them earlier (and despite the fact that nearly everyone calls the field "Kolmogorov complexity"). He also demonstrates his misunderstanding of Kolmogorov complexity when he says "Ordinary laws just transform input data into output data. They can shuffle information about but they can't create it." Of course, this is false. Take, for example, the transformation that maps a string x to the string xx. Then it is an elementary exercise in algorithmic information theory that the information (in the Kolmogorov sense) of xx is greater than that in x infinitely often. So, in fact, it is quite possible for "ordinary laws" to create information, in the Kolmogorov sense.

Also in the The Fifth Miracle, Davies makes the claim that quantum algorithms can make the solution of the traveling salesman problem "tractable" - a misconception so common that Scott Aaronson has resorted to debunking it in the masthead of his blog. It seems that when Davies pontificates about issues involving computational complexity and information theory, he cannot be relied upon.

Several years ago, Davies told me he would correct these mistakes (to his credit). But he's also been quoted as claiming, in Larry Witham's book By Design, that "Dembski's attempt to quantify design, or provide mathematical criteria for design, is extremely useful. I'm concerned that the suspicion of a hidden agenda is going to prevent that sort of work from receiving the recognition it deserves. Strictly speaking, you see, science should be judged purely on the science and not on the scientist." No surprise, Dembski flogs this quote whenever possible.

Two years ago, I asked Davies to justify his claims about Dembski. How, precisely, are Dembski's bogus claims "extremely useful"? Where have they been used? What about all the mathematical criticism of Dembski's work? Davies refused to justify his remarks, saying he was "too busy" to address them.

Now I see why he's "too busy". He's too busy writing silly op-eds for the New York Times.


Anonymous said...

Steven Novella has also replied to Davies' silly comments on his blog and his podcast: http://www.theness.com/neurologicablog/index.php?p=56

Bayesian Bouffant, FCD said...

I guess paying someone a million plus bucks for saying silly things encourages them to say more silly things, and causes them to think that what they say is worth something.

Erdos56 said...

What is the protocol on developments that have a non-contemporaneous but parallel development? Did Greg Chaitin reinvent Kolmogorov independently or was he informed by some translations of K? I tend to use Chaitin-Kolmogorov just out of a lack of knowledge about whether there is sufficient evidence to doubt that Greg added to algorithmic probability in a nontrivial manner.

If nothing else, I have to applaud Greg for bridging the accessibility gap. Paul Davies I offer no such leniency.

Jeffrey Shallit said...


There's no hard and fast rule about how credit gets assigned. Although Thue definitely discovered the sequence now known as the Thue-Morse sequence in 1912 or earlier, Morse gets part of the credit despite having made an independent discovery in 1921. Why? Partly because Morse wrote in English in a more accessible publication. Sometimes Prouhet gets the credit, despite the fact that he did not explicitly give the sequence; one has to "read between the lines" in an 1851 paper of his.

Similarly, the ideas of Kolmogorov complexity were independently discovered by Kolmogorov, Chaitin, and Solomonoff. There is a detailed discussion of this in the book of Li and Vitányi.