Is natural selection a driverless car? Then where is the software?

Much of the literature about Darwinian evolution uses variations on the term “driver.” Something is “driving” evolutionary change. Who is it? What is it? 

• The University of Otago says “environmental change is driving rapid evolutionary shifts in New Zealand’s native species.”
• The University of Sydney says a geological cycle “drives biodiversity.”
• The origin of sex chromosomes was driven by the “recombination landscape,” says Current Biology.
• “Positive selection” could be driving the evolution of a minimal cell, says Nature. That’s like saying selection drives selection.
• PNAS says that “the neural crest was a crucial evolutionary innovation driving the origin and diversification of dermal armor” in the first vertebrates.
• Two evolutionists in Science Advances say that “the teeth of fish and marine reptiles could represent an important driver for greater shell resilience” in the fractal-like, intricate morphologies in the shells of ammonites, but admit that “the driver behind their iterative evolution has remained unanswered since the first hypotheses introduced in the early 1800s.”

Are any of these things the primary drivers of evolution? Or like a row of taxis, can Darwin take any driver available? Does Darwin get into the nearest driverless car to get where he wants to go? If so, how does he give directions?

Fuel and Ignition

Impersonal forces can drive things. Gravity drives the planets in their orbits. The sun’s energy drives the weather. Wind drives the formation of sand dunes. In evolutionary biology, though, something must drive the formation of new functional information-rich structures: eyes, wings, and brains that did not exist before. Evolutionists call these structures “innovations” or “novel structures.” What drives innovations uphill against the drag of entropy? The answer often given is: “selective pressure.”

Some authors writing in Current Biology looked for a “common selective pressure” that drove the evolution of leg-like appendages in a group of benthic-dwelling fish called sea robins. Whatever this pressure is, it seems remarkably focused on helping the fish.

Remarkably, legs appear to have evolved at least three separate times within the Scorpaenoidea superfamily, suggesting a common selective pressure may underlie this adaptation. [Emphasis added.]

Is selective pressure, then, what drives evolution? It might be, but they don’t know.

A longstanding goal in biology is to understand the origins of novel traits. Sea robins represent a particularly clear example of an organism that has specialized for an environmental niche by evolving new organs….

Sea robins are therefore ideally positioned to become a powerful comparative model for illuminating general principles that drive the evolution of biological novelty.

An evolutionist at Columbia University claims that selective pressure makes us fat. It’s a strange kind of pressure, with no units or equations. 

By consensus, though, the “general principle” that drives evolution seems to be “selective pressure.” Maybe this force is like the crankshaft that pushes the piston down in the engine, compressing the gas. We need a spark.

At Victoria University of Wellington, evolutionary biologists proposed a candidate for ignition.  “Warming waters spark ‘evolution at super speed’ in marine sponges,” they announced. Darwin’s limo can go from 0 to 60mph in record time!

Marine heatwaves caused by the warming climate are capable of sparking rapid changes in sea sponges with scientists describing the pace of change as “evolution at super speed”.

Having an engine and a spark plug is fine, but the vehicle still needs a driver who knows where to go.

A Model Driverless Car for Darwin

One of the most focused papers recently that tried to identify Darwin’s driver was by Charles Kocher and Ken Dill in PNAS, “Darwinian evolution as a dynamical principle.” Dill was the one who gave an excellent TEDx talk about molecular machines being true machines, not just a metaphor. Maybe he and Kocher can take the metaphor of “driver” out of evolution and turn it into an empirical, measurable force. “The driver,” in their model, “is external resources out of equilibrium.”

Darwinian evolution (DE) — biology’s powerful process of adaptation — is remarkably different from other known dynamical processes. It is antithermodynamic, driving away from equilibrium; it has persisted for 3.5 billion years; and its target, fitness, can seem like “Just So” stories. For insights, we make a computational model. 

Smarting from DE’s just-so story reputation, they promise a new model that makes it scientific. In their description, evolution begins somewhat like a rollercoaster ride, with hills and valleys and ratchets forcing it upwards. But is this reality or an amusement park?

In the Darwinian Evolution Machine (DEM) model, resource-driven duplication and competition operate inside a cycle of search/compete/choose. We find the following: 1) DE requires multiorganism coexistence for its long-term persistence and ability to cross fitness valleys. 2) DE is driven by resource dynamics, like booms and busts, not just by mutational change. And, 3) fitness ratcheting requires a mechanistic separation between variation and selection steps, perhaps explaining biology’s use of separate polymers, DNA and proteins.

This “unrelenting drive for biological adaptation on earth” pushes organisms up fitness peaks, they claim, without mind or intelligence. The driver is resourceful, innovative, and powerful! Interestingly, they make DE synonymous with SOF: survival of the fittest. They trot out several icons of evolution, state the Central Dogma, and heroically affirm neo-Darwinism. But then, they look under the hood and behind the wheel.

Also referred to here as Survival of the Fittest (SOF), DE is among the most resourceful, innovative, and powerful drivers in the earth’s balances of energy, matter, water, and food. Ever since the work of Charles Darwin and Herbert Spencer in the 1860’s, SOF has been invoked in narratives about biological adaptations, such as the shapes of the beaks of Darwin’s finches, the color patterning on the wings of England’s peppered moths, and many others. Its operational basics are well known: Genes encode proteins encode traits; variations are explored by mutational change; organisms compete for finite resources; and natural selection retains those genes/proteins that give the species greater fitness for its environment.

But as a matter of basic principle, DE is not fully understood. What is the force? Why is there any force at all?

A driver needs to put his foot on the pedal and know where to go. Think of our rollercoaster metaphor, but now, let’s take the steel track and ratchets away, leaving the cars without aim on the peaks and valleys. In place of a metal car, let’s put a cell on the ground, or a pebble for that matter. What will drive it up to the peak, transforming it into an evolutionary biologist who can think about where he came from?

Living systems do not drive toward equilibrium. Rather, they tend toward adaptation, selfishness, and intricacy, powered by persistent inflows, in a process that has not died for billions of years. The driver of evolution is not the Second Law of Thermodynamics and is not a tendency toward equilibrium. Fitnesses in biology are unlike energies in physics. While material equilibria are tendencies toward endstates, fitnesses are tendencies across uncountably many different molecules, mechanisms of action, and degrees of freedom, toward opportunistic advantages across uncountably many environmental situations. Here, we frame these questions in terms of evolution as a machine-like cycle.

It’s quite a rollercoaster. They watch it go across hills and valleys, with organisms being pushed and shoved by mindless forces looking for “opportunistic advantages,” searching out the reward of “fitness” (which could be anything other than extinction) as if on a treasure hunt. Adding to the confusion, their scenario of “unruly and fluctuating resources” is not static: the hills and valleys are moving up and down in a continuous slow earthquake. Why wouldn’t the objects on the landscape fall into the low points? Where are the ratchets? They mix metaphors again:

We use the metaphor of golfers on a mountainside to express how fitness valleys are crossed. Because evolution entails a driven nonequilibrium system with persistent resource availability, it is much like aimless golferslocated broadly across the landscape driving golf balls uphill and downhill. The ratcheting upward on a fitness landscape happens only because those points are selected for after the fact, not targeted a priori. Since the DEM is not winner-takes-all, there are many such golfers sampling the space simultaneously. There’s power in numbers.

Driver’s Test

Kocher and Dill’s model looks impressive, with equations and all, but it has fatal flaws as I see it. 

1. It is anthropomorphic. They impose their human ideas of what is “good” on a mindless, valueless landscape. Fitness is “good”; extinction is “bad.” Why? Who cares? Why wouldn’t a material organism take the easy way out, follow entropy, and die? Who is there to shed a tear? What is this “drive to persist” they assume exists in mindless machines? If billions of blind, robotic golfers hit billions of balls simultaneously, some will land higher than others. But if nobody cares where the balls end up, won’t they roll back down when the environment changes again? To be consistent, Darwinians must work extra hard to resist the temptation to humanize material nature with terms like “opportunistic advantages” that organisms supposedly pursue, and “competition for resources” they portray as if they were watching sports.
2. It is question-begging. By starting and ending with DE, they assume what they need to prove: that beneficial mutations occur, and that they will be selected. “Darwinian Evolution (DE) is the unrelenting drive for biological adaptation on earth,” they say. But that is the question! “Extinction can be avoided by innovations that arise in coexisting variants,” they say at one point. They “arise”? That sounds like a miracle. Their conclusion assumes “evolution’s power of innovation,” treating it like Dawkins’s blind watchmaker, another personification fallacy that uses anthropomorphic metaphors. They assume random mutations will ratchet up to a functional adaptation. They assume a blind selector will select. They gloss over the origin of life, which cannot appeal to natural selection, and by so doing, they assume a can opener.
3. It is vacuous. By equating DE with SOF, fitness becomes a function of survival, and survival a measure of fitness. The survivors survive because they are fit. How do you know they are fit? They survived. Nowhere in this paper do they give an example of an actual complex innovation that arrived stepwise by mutation and selection.
4. It has no source of information. It’s the same old neo-Darwinism with a royal robe on. Random mutation and selection-without-a-selector does all the work. But the information content in a human compared to a bacterium had to come from somewhere. 
5. It is a posteriori. Looking out over a complex functioning biosphere of animals and plants that non-Darwinians consider designed, they force it into their materialistic worldview, speculating that blind processes might have driven beautifully adapted organisms to their current fitness peaks. That’s a non-sequitur. Why not accept the explanation that they look well adapted because they were designed that way? 

Everyone Agrees

Complex adaptations do exist. But rather than attributing their success to blind drivers, why not use abductive inference by drawing on our uniform experience with machines exhibiting complex adaptations? The Mars rovers are well adapted to the harsh Martian environment, in which they move, work, and persist because they were designed for it. Should Kocher and Dill waste time concocting models of how rovers “emerged” and “competed” for resources? The only thing rovers don’t do is lay eggs and grow baby rovers, but a Martian landscape with a growing population of rovers subject to random variations would not help their model. The rovers would still rust out and die. Nobody would care except the rover drivers back at JPL trying to keep them going.

A better model is the engineering model. Mars rovers are adapted to Martian environment because they were designed for it. They can avoid risks and adjust to changes because they have sensors and algorithms that their designers foresaw and planned for. At CELS this year, biologists and engineers had lively discussions exploring this model and comparing its fit to the observations. The question-begging, vacuous, anthropomorphic DEM has had its day. Kocher and Dill gave it a valiant last hurrah. But self-driving cars that make it to a destination are, like organisms, “deeply designed and purposeful, active and engaged in their environments, and capable of adapting within their operating parameters,” as Eric Anderson put it. Blind drivers who don’t care about destinations end up in demolition derbies with no fans to cheer them on.