Gradualism is NOT Back, but they are Trying

 "First they ignore you, then they laugh at you, then they fight you, then you win."- Mohatma Ghandi

It looks like we are on stage three now. I say this because of a study that was recently concocted which purported to show that darwinian gradualism could still explain the biological record despite the appearance of a vast number of sudden changes that seem to fly in the face of any gradualistic answers. The study was entitled "General statistical model shows that macroevolutionary patterns and processes are consistent with Darwinian gradualism"by M. Pagel et al. Here is an article, optimistically entitled "Gradual Evolution is back", on the study. 

Over my years of attempted dialogue the narrative has changed. Early on, the narrative was that there was no such thing as "macroevolution", only "evolution". A long series of gradual minor changes becomes a big change over time. You still hear that now and then from those whose knowledge isn't current. But lately the focus has shifted as it has become obvious that there are very many large and in evolutionary terms very fast changes in living things at certain times. These periods of rapid change are usually followed by long periods of relative stasis. 

A handful of scientists in prior decades attempted to float the idea of massive sudden changes in organisms via natural means. The proponents called it "punctuated equlibrium", the critics "the Hopeful Monster Theory". Either way, it looked too much like miracles for the dedicated philosophical naturalists who have control of our science institutions, and it has fallen into disrupute. After all, there are no known evolutionary mechanisms anywhere in the animal kingdom that have any reasonable chance of producing such changes at the rates and volumes observed in the biological record. 

What they needed then, was a way to show how known evolutionary methods which are inherently gradulaistic could produced big changes quickly in some cases. Rapid gradualism seems like a contradiction in terms, but that is what they needed to preserve a naturalistic explanation for the earth's biosphere. With this study, they claim they have found it. I guess many will accept their claim. Just don't look too closely at the details. This study will only convince those who want to be convinced. Of course science has a higher standard than that. Even propaganda is designed to seem superficially plausible to those who want to be convinced. 

Please forgive me here as I break into a few paragraphs to explain where I am coming from. What I have to say may not fit any of your existing categories, and thus further explanation will often be essential to understanding my point. Are these mechanisms enough to provide a reasonable explanation for the history of life on earth?  I say "no", though do I acknowledge that nature has significant power to shape living things. In other words, "evolution happens." Indeed as I have explained before, many if not most defintions of evolution do not exclude creationism. "Change over time" says nothing about whether the change came from naturalistic causes only, or whether God acted to cause the largest changes and nature only filled in the gaps.

At this point I expect two tribalistic reactions, one from creationists and the other from naturalists. The naturalist will cry "God of the gaps!". Here is why this accusation falls flat: I am not saying that God stumbled into a universe complete with laws and decided to intervene in it. I am saying that He created a universe including the laws but is Himself above them. He is God of both the gaps and everything inbetween them. And the universe is a test, a test for beings like us. And to be suitable for the test it has to be like us- we can't really do God's will without God's help, and neither can creation. I could go on about this but the bottom line is that most creationists have it very wrong concerning what scripture says about the initial conditions of the universe. 

The tribal creationist might recoil at my "mixing" of evolution and creation. To them I would suggest a closer look at the actual text of Genesis chapter one that they think they are defending. God commanded the waters and the earth to bring forth living creatures. That was the original command. Subsequent to that, God intervened Himself and formed "kinds" of various broad categories. That doesn't change the initial command. It just means that what I said in the paragraph above is true. The earth can't do God's will without God's help. He did the hard stuff, but the earth itself filled out the details in accordance with God's command. So evolution and creation are working hand in hand. This doesn't apply to humanity of course, but that is for another post. 

It is my position that what science is discovering about nature confirms at least that aspect of the text of the creation account in Genesis. There is constant graudual change which is driven by nature combined with (in the past) a large number of greater changes which happened quickly and seem beyond any reasonable natural explanation available to us. The Pagel paper attempts to fill this gap with a model (they call it the "Fabric model") which they offer to show how gradualism can still be the answer, even if it often wasn't gradual in effect.

To even attempt this, they must cede some ground. They freely acknowledge the distinction between "macroevolution" and "microevolution", They then divise two categories of observed change. One they call "evolvability", which is how well a population of organisms can adapt to their environment. You might think that is all the positive power that evolution has, but they describe another component which they call "directional". This they vaguely describe as one which changes which environment is optimal for the population of organisms. The two types interact, but they say there is no correlation between them as far magnitude is concerned. For example, an organism which doesn't have much room left to further adapt to its environment is just as able to have a directional change as one with a lot of "evolvability" room in its present environment. Thus every "evolutionary dead end" is still a possible source of great change, right up until extinction. That's gotta help the odds some, right?                          

The heart of their study is a computer model in which they use these two methods working together to get big changes in a small amount of time with gradualistic processes. But how flexible do they have to make the scenario in order to do this?  In their introduction they are vague enough, but admit, "Our methodology does not constrain the number, position or trajectories of effects,...". 

So unlike the real world where there are physical and chemical contraints on basically everything, in the model, none of these things were contrained. And just in case you missed the hint that this risks becoming a sophisticated GIGO exercise, they add "By comparison to parametric models that fit one or a small number of processes throughout a tree, the Fabric model is typically rich in parameters, and this requires care in fitting and interpretation..."

Did they just say their algos basically allow them to make up different rules as they go along? I think so. They are saying in effect that a process that happens over here need not be happening throughout the tree or even anywhere else in the tree. And the parameters of those processes are varied. It really sounds like they are saying that the mathematical equivilent of miracles are built into their computer model, but let's not call it that! That is what they are trying to not say. 

Maybe that is why they follow their statement up with "But the Fabric model’s approach avoids the potential for one or a handful of effects in a tree to give the impression that the parametric model’s form is true everywhere in the tree when it might not even be true of any part of the tree."

Got it? Things can happen from time to time in some part of the tree we build that are not true in any part of the tree when you look at it from the perspective of fixed laws! That sounds like a fancy way of saying in their model, miracles can happen. OK then, but why are you using the model to defend a naturlistic view of evolution?

"The average correlation in the posterior sample between the observed and predicted data at the tips of the tree (obtainable from Eqs. (4) or (5)) is 0.90 ± 0.03, 95% range 0.82-0.95; we do not explore the predictive capabilities of the model further in this paper." - so one not-great fit even after all the leeway they gave themselves, and then they quit looking. It doesn't sound like they thought their results were exactly bursting with exploratory power. More like just enough so that they could spread the word that they have "shown" that known evolutionary processes can explain it all!

On the positive side, they used the term "macroevolution" freely. A refreshing change from the conversations I've had with ardent naturalists who insist that there is no such thing, only evolution. Even though that is basically what these researchers are trying to prove too, they admit that the best way to get a fit is to 1) not have fixed rules, allow some things to happen that can't ordinarily happen and 2) view the system in terms of two distinct kinds of changes, directional changes and evolvability changes. Both must occur for macroevolution to occur in their view.  Microevolution can happen with either one, but mostly the latter.

I say that because they write....

Directional and evolvability changes are substantial and distinct

Comparison of the five models’ marginal likelihoods (Table 1) illustrates one of the central points of this article: that a statistical description of macroevolution must account for the substantial and distinct contributions of directional (β) and evolvability (υ) changes: modelling one of these processes at the expense of the other or linking them a priori, risks missing important elements of the macroevolutionary picture.

They based their findings on looking at only one very narrow slice of nature- changes in body size in mammals. So they were not even talking about the creation of new orders of animals but rather the changes in body size in the orders and sub-orders of a particular class. Even at this level, changes that were "large and abrubt" frequently occurred, or as the heading of the next section put it.

"Directional changes are large and possibly abrupt"

And in this section they note, " the net amount of directional change in a branch appears not to be limited in any general way by the length of that branch.". Just calling it "directional change" doesn't make a large and abrupt change any more likely. It still looks like a miracle and it still contravenes what most of them have been telling us for decades about how evolution was supposed to work.

Zooming out to the big picture, if there is no correlation between directional changes and evolvability changes, and having no correlation is necessary to their model, then all they are doing is "spreading out" the miracle, not eliminating it. That's because you would need a lot of graudal directional changes in a row, in the same direction, intersperesed in the evolvability changes to get the big and abrupt directional changes observed. 

So instead of rolling pair of die with a billion faces and rolling snake eyes when needed, you have a thousand rolls of dice with a million faces and each of them come up snake eyes when needed, moving you in the same direction each time. The results are still unreasonably improbable even if you break the odds down into a series of somewhat more probably but still unlikely events. Maybe no one directional change is unreasonably improbable in itself, but stack them all together and it becomes so. A population of organisms may vary rarely have a successful mutation producing a "directional change", but to assume that the same population can reach fixation on dozens of such events which happen over a brief period of time so that a "large and abrupt" direction change occurs is still as indistinguishable from a miracle as the original Hopeful Monster theory. 

Ironically, what they are doing sounds a lot like the language I used five years ago to describe how creationism and evolution are not inherently contradictory. I wrote "Take the gap between a fish and an amphibian. What if over the course of thirty or forty generations God acted to put just enough changes in each generation that they would still be able to be birthed and bred by natural means but each generation would also be further toward the amphibian end of things? This so that even though no amphibian was created out of thin air, or clay, one still had a very different creature though only forty generations removed from the fish. That result would be due to genetic engineering moving things a bit further along each generation. That is “descent with modification” but the modification that matters is via genetic engineering. So is that evolution, special creation and intelligent design all rolled up into one?" 

Okay so maybe they are saying 4000 or more generations instead of 40, but does that really matter? And it is not because they know it took that long, it is just because they have a hard time believing what they have seen. In a sense they are doing what creationists have been accused of, "Argument from Incredulity". They find it increduous that God could intervene in nature and so they must propose models like this one instead, which might make creationists incredulous. The issue isn't who is arguing from incredulity, but whose incredulity is the most reasonable.

Given all this, I didn't feel the need to dive into the details of this study as I sometimes do with others. Scientists have every right to keep looking for natural explanations, but it is pretty clear to me the science media is way over-selling this. The researchers were honest about their goals- they are trying to retain gradualism (thus keeping the door open for naturalism) in the face of massive evidence that these changes were very often not gradual, especially the largest ones. But the result sounds a lot like what I as a particular form of creationist was proposing five years ago! Perhaps we are finally reaching a place where creationists can find out they still have something important to learn about the bible, and evolutionists can learn whether they are really committed to science or their real committment is to philosophical naturalism. 

**************************************

My book on early Genesis is mostly not about this, but something far more wonderful- a proof that early Genesis was pointing to Christ, thus proving the validity of both scripture and the Christian faith. Amazingly, doing so resolves the supposed conflicts between scripture and science merely as a by-product of viewing the text through the lens of Christ. 

  You Tube Channel