It's somewhat embarassing that this topic continues to waste storage space, but the linked article from the OP did give me some insight into why it's such a hot button issue. Actually, the linked article didn't do that, but the journal article to which it referred did do that.
The article studied evolution of vulval traits in a family of nematodes. This choice of species of study thus guaranteed that laymen will continue to make fun of scientists for a long time to come.
From studying variations among similar species in conjunction with developmental biology, they were able to infer patterns of evolution. For example, they were able to infer that in some cases, the vulva of an ancestral nematode was found mid-body, while the vulva of the moder nematode was found posterior.
The interesting question was how did that evolution occur. If one were to follow the movement of vulvae (Sorry if I got that plural wrong. I never studied Latin. It just looks right.) across the generations from mid-body to posterior, would it ever change course. Would it go a little bit back, then more back, and still more back, until it arrived at its final posterior destination? Or would it go a little bit back, and a little bit back, and then reverse direction just a bit, but then resume its rearward trajectory, perhaps reversing course a few times in the process?
This study concluded that the first pattern was the case. Once those nematode vulva started moving backwards, they never reversed course.
They summarized their study thusly:
"In conclusion, our study shows that evolution of vulva
development is strongly biased and only few aspects
are likely to change in an unconstrained stochastic fashion.
Our phylogeny for rhabditid species, including the
model systems C. elegans and P. pacificus, provides a
foundation for evolutionary analyses of other characters
as well. If the patterns that we observed in the vulva
system are found more generally, then most of developmental
system drift is driven by deterministic and not
stochastic processes."
The magazine article from the OP then summarized the summary by saying
"Evolution not random"
Which proves, beyond all doubt, that one sentence summaries of extremely complex explanations are inevitably so oversimplified that they contain almost no information.
The real scientific controversy involved in this paper has nothing to do with whether evolution "is random" or "is deterministic". Both summaries are too simplified to contain any information that is of any use. The real controversy concerned something called "developmental systems drift", and whether or not phenotypical changes occur because species' characteristics wander around in phenotype space in a manner described by a possibly biased drunken walk, until they settle into something quite by accident, but which survives because it happens to work, or do they tend to start in a direction and continue in that direction because going in that direction makes things just a little bit better, until the trait can no longer be further optimized. This paper concludes that DSD usually behaves in the latter manner.
None of which actually means that evolution "is random" or "is not random".
