Friday, September 09, 2016

Robert Marks: Two Years Later, Still No Answer

One characteristic of creationists is their unwillingness to follow the usual academic norms. To name just a few things:
  1. While a tiny fraction of them publish papers in legitimate peer-reviewed academic journals, they typically do not publish their creationist views or evidence for creationism and intelligent design in such journals. Instead, they invent their own bogus journals, which then struggle to stay afloat for lack of acceptable submissions. Do any of you remember Origins and Design? I think it died around 2000. Do you remember Progress in Complexity, Information, and Design? It died around 2005. Now the intelligent design crowd has Bio-Complexity, but it has published only one paper so far in all of 2016. It, too, is headed for death.
  2. They typically do not present their creationist views at legitimate peer-reviewed conferences. The few exceptions seem to be closed, invitation-only conferences devoted only to creationism or intelligent design. You do not see, for example, William Dembski (the supposed "Isaac Newton of information theory") presenting his work at the top information theory conferences, such as the IEEE International Symposium on Information Theory.
  3. They inflate their credentials.
  4. They hold meetings at universities by renting space and then suggest or imply that the university somehow sponsored their meeting. The 2011 "biological information" meeting at Cornell is an obvious example.
  5. They are prone to making public claims that they are not willing to justify.
The illustrious Robert J. Marks II, professor at Baylor University, is an example of this last characteristic. Back in 2014, he made the following claim: "we all agree that a picture of Mount Rushmore with the busts of four US Presidents contains more information than a picture of Mount Fuji". I wanted to see the details of the calculation justifying this claim, so I asked Professor Marks to supply it. He did not reply.

Nor did he reply when I asked three months later.

Nor did he reply when I asked six months later.

Nor did he reply when I asked a year later.

It's now been two years. Academics are busy people, but this is pretty silly. Who thinks the illustrious Professor Marks will ever show me a calculation justifying his claim?


Mark said...

Would it be fair to say that you could make the comparison by taking a digital file of each photo, compressing it, and comparing by how much each file could compress? Seems simple and objective.

Jeffrey Shallit said...

That would be one possible measure. However, I suspect that doing so would depend very much on exactly what position one chose to take the picture from. So Marks' claim is suspect right away just for that.

Pseudonym said...

You'd have to agree on a compression algorithm, and therein lies one of the trickier problems. Modern image compression algorithms are optimised for perception, and faces are one of those things that humans can reconstruct from very little information (sometimes randomness; see pareidolia).

Algorithms which are optimised for face images are a current active area of research, given their importance (e.g. for real-time video). Hell, you could easily imagine an algorithm which "compressed" a portrait by substituting a caricature.

Mark said...

I guess my suggestion was simplistic, owing to my naivete in technical matters regarding image compression. But I think the original argument is easily misinterpreted--the comparison is between photographs, not cultures. A photo of Mt. Fuji taken and processed by a skilled, artistic photographer may contain more information that a snapshot of Mt. Rushmore taken by a tourist; photos taken at other wavelengths may show surface-temperature or moisture content distributions, etc. The argument was made, I think, by someone considering the subjects of the photos rather than the photos themselves.

Tom English said...


A passage from something I've had on the back burner for a long time:

That is, the specified complexity of the full-resolution image, measured in the context of Wikipedia articles relevant to the "intelligent design" (ID) movement, is upward of 22 million bits. In practical terms, the image is large when stored in the PNG format. But a short computer program can convert the articles into the image. Subtract the number of bits in the program from the number of bits in the PNG image, and you have measured, according to the gurus of ID, the meaning in the image.

What the short program does is to calculate the bitwise XOR of two images in the given context. Look closely at the elliptical region below the setting sun, and you'll see a ghost and a cone.