Evolution drives bigger animals nonrandomly

Here is a hot new evolution result in AAAS Science Mag:

In today's world, many animal species are large, with even larger species only recently extinct, but the first animals to evolve were tiny. Was this increase in size due to active selection or to some more random process? Heim et al. test the classic hypothesis known as Cope's rule, which posits that there is selection for increasing body size. They analyzed a data set that spans over 500 million years and includes more than 17,000 marine animal species. In support of Cope's rule, body volumes have increased by over five orders of magnitude since the first animals evolved. Furthermore, modeling suggests that such a massive increase could not have emerged from a random process.

It was reasonable until that last sentence. What kind of modeling could possibly show that something is not from a random process?

A random process for size changes would not necessarily mean that animals are equally getting larger and smaller. A random process could make 50 % of the species get larger, 30 % stay the same, and 20% get smaller. Or any other pattern could be a random process.

The rest of the article is behind a paywall, so I don't know how it makes the argument. One of the co-authors is from a math department. He should know better.

A lot of evolutionists are preoccupied with questions of randomness. I say that Random chance does not cause anything.

Here is the CS Monitor explanation:

While the overall increase in marine animal size is pretty much indisputable, some scientists argue that size is not a matter of “active selection,” but a result of random, non-selective mutations – an concept known as neutral drift. In other words, neutral drift could cause some lineages to grow in size, but only by chance – that doesn’t necessarily mean evolution “favors” size. The neutral drift argument is supported by evidence from bird and insect populations, who have not grown in size as Cope’s rule postulates. ...

“One question that has intrigued biologists ever since Darwin's time is whether evolution is largely random or whether it is predictable and directional,” Payne says. “Stephen Jay Gould famously asked whether the world would look similar if we could rewind the tape of life to the Cambrian and then let it run again. He spend much of the second half of his career arguing that a lot of evolution is less directional than we we tend to believe – that people often see patterns where there are none.”

“In this case, it appears that there really is a pattern. As we come to better understand the underlying causes, it will be easier for us to predict whether or not we should expect it in terrestrial systems as well – or even on other planets, if we were to find life on them.”

Gould dismissed Cope's rule as a "psychological artifact."

I don't know how it can be a psychological artifact. Isn't it obvious that dinosaurs were huge, and that it took millions of years of evolution to get that large? On the other hand, there are physical limits to insect size.

All this talk of rewinding the tape of life is meaningless. Would that include a meteor to wipe out the dinosaurs? Or was that too random for the tape?

I submitted a FQXi essay on randomness and related topics. I will post a link when it is online.