The Myth of Evolutionary Ascent

Many physicists whine about the public's grotesque misunderstanding of basic concepts like centripetal force and electromagnetism. Some of those very physicists often like to consider biology to be a simple subject, delivering profound lines like "people still study evolution???". Of course, how can anyone have any problems understanding something that barely uses any formulas! Of all sciences, biology uses the smallest portion of the Greek alphabet, and hasn't even moved on to Hebrew yet. How can there be much work in a field unless Norse runes must be invoked to complete a PhD thesis without assigning multiple meanings to symbols? Life is a giant differential equation, and unless you're cracking your skull open over one you're not doing it right! Quantify and model everything!!one

But seriously, how can any one misunderstand something as simple as evolution? Let's go on a bit of a rant adventure and find out!

The Ladder to Apocalypse
There is a classical image frequently used to represent evolution. This appears on the cover of some editions of The Origin:

We really like this picture. It's polite to our fragile feelings. It dares not offend (too much) our sanctum of superiority. Fine, the abyss between Man and Animal may not be so sharp after all - we may actually be related to the beasts. Fine, we'll even let Man arise from Animal, and none other than the coarse graceless ape. But at least we can still keep our final tatter of self-importance: for while but a small chapter in the story of life, this story was written for us. We are the ultimate Species, the crown of the tree of life. How flattering seems the depiction of progress, this procession of life from the lowly ape to the fully-formed proudly-standing masterpiece of evolutionary craftsmanship!

This view is reflected in the vernacular use of 'evolution'; exposed blatantly in cases like the Russian phrase "through hard work monkeys lost their tail and became man" (although that may be just my family, who knows...) and more the familiar Japanese phenomena like this:

You may laugh, but this is a rather accurate representation of the public's (and some biologists') understanding of evolution. One may recall the oft-recited progression of life from bacteria to amoebae to sponges, fish, monkeys, and us. While preparing a talk for some compsci students, I realised Toxoplasma may have a slightly different opinion: (and almost got an aneurysm making this)

Since our good friend Toxo actually managed to parasitise most mammals and birds, it clearly must preside over the crown of our lineage. Opisthokonts(well, unikonts) are but a basal lineage to Chromalveolata, of which Toxo is a proud empress. All hail T.gondii, the fierce goddess of the crown eukaryotes!

But seriously, does anyone else get a bit tired of constantly hearing about how the stupid lowly amoebae somehow congregated together and became multicellular and wise and awsome? For the record, the sister group to animals is choanoflagellates, who are not amoebae! Our amoeboid cell types arose secondarily, since after all there isn't that much of a problem in doing both... (see Naegleria, which can switch between amoeboid and flagellate forms in <120min...) What that image does represent is an approximation of our lineage's particular path, outlined in blue here:

Homonids being basal to chimps, of course. We stopped evolving, they moved on. So how does it feel to be a basal lineage anyway? Still 'primitive'?

Furthermore, the generation span in the animal lineage tends to be much longer, and among the larger animals we have rather insanely long time periods between heritable genetic modifications (and thus material for selection to play with). This means bacteria, some of which can replicate multiple times a day, until present day have passed through many orders of magnitude more 'versions' of themselves than animals have. In some ways, one can argue bacteria are more advanced than large metazoa. And this view is quite objective, if we only use the number of generations to go by.

Now that we've established that prokaryotes rule the world, we could simply concede defeat and go home. However, there still remains a nagging thorn in the side: complexity. If we define complexity strictly as the number of components involved in a given system, one must agree metazoa do have far more components than their bacterial counterparts. That does not make us superior; however, it does raise some further points and misconceptions about directionality of evolution.

The evolutionary 'ladder' may be a valid model for one thing: the history of a single lineage, with height representing nothing more than simply the time axis. Complexity has nothing to do with it. Nor does that ladder reach anywhere but the inevitable demise of our Earth at the end of biological time.

Evolutionary Directionality
Evolutionary directionality is a myth, a massive misunderstanding of evolutionary processes. I like to call it 'evolutionary creationism', for it eerily mimics the fairy tale put forth by the 'Intelligent' Design movement, albeit devoid of a few supernatural elements. Often, 'evolutionary creationism' and religiosity can be seen to go hand-in-hand, eg. Simon Conway Morris*. Even some brilliant, quite rational thinkers like Dennett have fallen for the appeal of anthropocentrism, emphasising the rise of human 'consciousness' in the discussion of memetics and devoting an entire chapter to the importance of the intentional stance in evolutionary thinking (1995 Darwin's Dangerous Idea). Consciousness is a topic for another day, but intentionality must be used with great care.

Perhaps in some cases it may be useful to assume nature strives towards perfection and survival in order to deduce the 'function' of a certain trait (as in Dennett 1995), but even if this may help, it must be kept in mind that the concept of 'function' itself is an artefact of the human mind. We say the bird's wings are 'for' flying, but this is just a shorthand way of saying "the bird would not be capable of flying without this feature, therefore it is essential for that activity. Furthermore, it seems that flying would suffer the most in the absence of this feature, thus we shall denote this structure's purpose as 'for flying'."

This can get murky upon deeper examination. What if these structures have multiple functions? Wings help maintain balance while walking; thus we can add that an additional 'function' of the wings is exactly that. However, if you chop those wings off, the bird's circulatory system would fail. Is a function of wings to keep blood from spilling out? That sounds a bit off. Ok, let's say it never had them in the first place. Then how about escaping predators? The bird would likely fail miserably without its wings. So predator avoidance a function of the bird's wing? How about foraging? Finding mates? If the wings of male specimen have some decorative pattern, we have no problems stating they serve also 'to' attract mates; however, a bird devoid of an elaborate winged mating ritual must also have them to mate eventually! But in the latter case, we hesitate before stating mating as a 'function' of wings. Ultimately, the function of every trait is to perpetrate itself, otherwise it wouldn't be there to speak of. But this definition is even more useless.

Hopefully the above discussion seems blatantly pointless, because it is. We use whatever definition of function we find must convenient at the moment, and that's ok. Because there is no such thing as function! We made it all up! It helps us model the world around us, but just like hunches and stereotypes, it has no place in scientific explanation!

If there is no function, and therefore no purpose, how can evolution have any aim? How can there truly be any progress? We must realise that while we desperately strive to seek patterns, to find meaning in life, and therefore science, we cannot project the same features onto the world around us. (I'd go as far as arguing that our own intentionality is an epiphenomenon, an illusion, but save that for a[n even] more philosophical post...)

Life just is. Life doesn't even strive to replicate, strictly speaking. It's just that if it fails, it ceases to exist. This thereby generates an illusion of a goal (to perpetrate genes, for example). In a way, it's a ratchet - you can go in one direction, but you cannot go back (and continue existing). If you have a chemically-successful body plan capable of replication, and you maintain it, you keep replicating. If you fail, you're gone. Entire lineages can eventually fail. The only reason I'm sitting here blabbering on a blog is that a tiny fraction of lineages did manage to somehow make it this far. They didn't wish for this, it just happened.

Perhaps some other lineage could have invented computers with internet and blogs and all the rest, and the story of evolution would have a slightly different twist to it. But this wouldn't change the nature of evolution itself! It's still an aimless process, an extension of chemistry, which in turn is but a subset of the laws of physics. A subset detailing what happens if you apply certain constants and laws of the universe to certain chance conditions, of which we just happen to be part of. Of all things, I sure hope no-one would arge the universe has any intentions!


It takes discipline to avoid invoking direction and intentionality in evolutionary discussion. Our brains are adapted to a certain type of environment, and combined with a vast plethora of non-adaptive 'spandrels'**, have certain innate ways of modeling the world around us that may be less than scientific. Thus we have devised the formal Scientific Method to compensate for these biases. But it's natural to assume the intentional stance if you're a social creature. As Dennett explains, you don't make predictions about someone else's behaviour by analysing their neurology; you assume they aim for certain things, and base your predictions on that. Lacking the capacity for intentionality results in severe psychological and social problems.

Even despite my ranting here, I have, do and will make mistakes of this nature in biology. For example, in a post about Tetrahymena nearly a year ago I went on and on about how wonderful nuclear dimorphism (germline vs. somatic nucleus) is in terms of "protecting germline DNA from abuse by transcription". That was completely and utterly wrong. Upon seeing this strange phenomenon, I naturally assumed an adaptationist approach to the whole thing and blurted out this insane hypothesis. Perhaps the next section may help see how my reasoning was embarrassingly wrong in that post.

Rosie Redfield put it nicely: "[about constructive neutral evolution] Shouldn't that be our null hypothesis?" You see, we should only invoke adaptationism when simpler explanations (physics, biochemistry, molecular genetics) fail.

The Rise of Non-Adaptive Complexity
Ratchets
So if evolution lacks direction, how do we get such wonderful complex creations? Shouldn't everything stay at the unicellular, prokaryotic level? Well, in a way, it does. The vast majority of life on earth, by far, is unicellular. Does this mean unicellularity is better? Or does it mean very few organisms happen to 'discover' the rarer ways of being?

The answer probably involves a little bit of both. Despite fiercely disagreeing with Dennett above, I do like the 'design space' concept he tends to often use. Essentially, it's as if evolution is blindly stumbling about this design space, exploring some of the possibilities available at any given situation. Some of those options are successful, at least for some time, and lead to further possibilites. Some options are dead ends, and lead to extinction. Some options are simply never touched. Sections of this design space have repeating motifs that manifest as evolutionary convergence, for example the camera eye of mammals, cephalopods and Warnowiid dinoflagellates(Leander 2008 Trends Ecol Evol); or the morphologies of a tapeworm and Haplozoon, a similar but microbial intestinal parasite (Leander 2008 JEM).

Now how can aimlessly wandering about the corridors of design space lead to such stunning, and recurring, complexity?

The traditional explanation is positive selection. Unlike negative selection, responsible for killing things off for developing, losing or modifying certain features, positive selection 'encourages' certain traits by killing off the failing competition. While negative selection kills off the rabbits that are too slow to avoid the wolf, positive selection kills off rabbits that can't eat fast enough to compete with their rivals, to use a very crude example. This can explain some features, but opponents, as irrational as they are, do have some valid reason for being confused how this leads to the formation of new features.

Luckily, we have a powerful tool on our side. New features don't have to be selected for to arise. As long as they don't inhibit the survival of the carrier, they are free to explore design space at their whim. An additional feature is ratchets - sometimes, either due to negative selection or simply physical characteristics, once something is done, it cannot be undone. (Eg. you go to 4chan.org (don't!), and see things you don't really happen to savour. This is a psychological ratchet, or in the words of 4chan speak: "What has been seen cannot be unseen")

A less abstract example: Dinoflagellates like to do everything the weird way. One such oddity is the need to attach a special DNA sequence, called a splice leader(SL), onto every single mRNA transcript made in the nucleus. Without the SL, the transcript is degraded before the gene can be translated. So far so good. Now, sometimes a reverse transcriptase, which can convert RNA to DNA and insert it back to the genome, happens to do exactly that to an SL-mRNA transcript. The result is a duplication of the gene, although already containing a SL sequence. However, after transcription of this SL-containing gene, a new SL must still be reattached, otherwise the mRNA transcript degrades (the key is that the SL also contains the 5' cap, which protects the transcript from degradation. Simply containing the SL sequence is not enough).

So now we have a situation like this: 5' cap-SL-SL'-gene. Since SL' (the old splice leader) is no longer necessary, and therefore no longer conserved, it can degrade as much as it likes. In fact, so can the entire SL'-gene sequence, considering the old gene still exists. Gene duplication is a very unstable event, since either gene (but not both!) can go to hell and the organism would not feel a thing. Thus, let's assume an equal chance that either gene goes away. If the new SL'-gene sequence mutates and disappears into genomic background noise, we would never even know it happened. However, what if the old gene vanishes, leaving behind the new SL'-gene?

We get strange genes with relict degraded splice leader sequences (Slamovits & Keeling 2008 Curr Biol)! Sometimes you can even have a gene with two or more SL sequences, one more degraded than the other. Ie. SL"-SL'-gene. Because the new gene cannot easily lose the relict SL sequence, this is a perfect example of a ratchet - the genome can acquire splice-leader containing genes, but cannot lose them. One should expect that over time, the genome becomes full of genes with SL garbage at the 5' end.

There is nothing adaptive about this situation, even though arguably the complexity is increased. But this alone doesn't yet explain how seemingly functional complexity happens. What we need is an addictive personality.

Constructive Neutral Evolution
Let's say you're going about your business, and something wonderful comes up that makes your life easier. For example, here you are, happily wandering about, and suddenly come across a Blackberry or similar device. Prior to this find, you managed to organise your life quite fine - you kept your adresses neatly organised on a paper notebook, you could sit in public transit without any electronic paraphernalia, you only emailed from work and home. And the thought to complain never crossed your mind. So you find this nifty thing, and after some initial grumbling about its excessiveness and lack of real necessity, you give it a try.

Fast forward a few months later, you lose it. Suddenly, you realise how much your dependence upon this device has grown, and wonder how you've managed without for decades before. This device becomes a necessary part of your daily life, and you've even forgotten how to use a paper adressbook.

Initially, the find was neutral; you managed quite well without the device. However, after time, your other functions (eg. paper-based organisation) degraded due to excess capacity, and the electronic device became fixed and necessary for function. You have experienced a gain in complexity, although arguably the initial adaptive value of this device was minimal. (sorry, I don't worship technology as much as the media does) Of course, after you became stuck with the device, perhaps you found additional new uses for it, some of which were actually adaptive.

This type of process is not always so positive. Many patients of psychological illnesses, such as depression, become hooked on drugs and later become incapable of being happy without them. Worse yet is when a 'normal' person comes across an addictive drug they have the misfortune to enjoy. Initially, their brain could function fine on its own (one-component system). They try the drug, and enjoy it. For some time, they can quit and still revert back to the one-component state. They are in a transitional two-component situation. However, time goes by and the brain chemistry changes (degrades). Now, the drug becomes fixed in a two component system. The complexity has increased, but one probably cannot really call this adaptive or beneficial in any way.

Evolution has a rather addictive personality, largely due to an inherent system of ratchets at many points. The exploration of design space is a tricky thing, with many traps and hidden moving walkways along the way; much like some airports (eg. Heathrow).

This brings us to constructive neutral evolution, as explained in the fascinating Stoltzfus 1999 Trends Ecol Evol article (free access), which shifts too many paradigms to be well-accepted. (A follow-up is in the process of being written, however, according to reliable sources...)

The basic concept is that we have a single component system, say enzyme U that catalyses the change of molecular A to molecule B. Let's say this pathway is very important, and the organism dies without it. Say we have a mutation X that completely inhibits U's function. You would never see this mutation in action as it's lethal to its carrier.

Now, assume another protein, V, which is involved in a whole other pathway somewhere, just happens to stabilise U a little. This enhancement is insufficient to be selected for via positive evolution, and its absence reverts us back to the initial state. However, just by chance, V happens to compensate for mutation X in U! Now, if X never happens, this convenient little detail would probably never surface, and thus remain silent. We can say there is excess capacity present in this state. However, say mutation X does happen, as it's bound to eventually...

...we now have a two component system requiring both U and V for the organism's survival! The complexity has increased, although the original single-component state has done quite well on its own. Now there are two elements for evolution to tinker with, and new design space has suddenly become open for further exploration. Perhaps now some of this new design space has adaptive venues to pursue, although that is still not necessary for further gain in complexity. Hopefully this figure makes the process a little clearer:

(based on Stoltzfus 1999; any errors in this depiction are my responsibility alone)

When I first heard of this, it seemed perfectly logical and obvious, and some part of me was wondering "well, what's new here?". Then it sunk in how outright wrong my prior understanding of evolution was! I too was obsessed with hyperadaptationism, not even bothering to think that some complexity may in fact arise without any adaptive reason! My friend overheard some classmates being pissed off by this idea - they thought it was simply ridiculous! Of course, they've been taught a certain way of evolutionary thinking in all the other classes, and even wrote exams on it. Now some guy from an obscure field dares to suggest they've been taught wrong!

If undergrads feel like they must resist this idea, then what can we expect of people who've gone through entire careers thinking a certain way? Much like the tremendous psychological difficulties in abandoning one's religious views from childhood, the resistance (masked by apparent apathy) towards constructive neutral evolution plagues the bulk of evolutionary biology as a field. Thus, we will still find hyperadaptationist explanations in evolutionary research, the media and public science writing for many years to come. A wonderful example of how ideas generally become accepted not by winning over the opponents, but an entire generation growing up exposed to them! (more space for memetics-based explanations? Oh, speaking of which...! /unsubtle hint)

So are there any real physical examples that can be explained by this view?

A fairly straightforward one would be of a biochemical nature: the story of CYT18 and mitochondrial self-splicing introns of Neurospora crassa (fungus) . (Collins and Lambowitz 1985 J Mol Biol; example discussed in Stoltzfus 1999)

While we generally think of introns as requiring spliceosomal complexes to be excised (and allow the gene to be translated properly), there is a class of introns which splice out by themselves, due to the secondary RNA structure of the transcript (the assumed initial state in intron evolution, by the way). CYT18 is a mitochondrial tyrosyl-tRNA synthetase; ie it's involved in making the tRNA containing tyrosine, for protein synthesis. This is to stress it initially had nothing to do with introns. The mitochondrion also had plenty of vital genes with self-splicing introns, polite enough to remove themselves prior to translation. In doing so, the RNA structure must form a double-stranded neck. Mutations in that region result in incomplete splicing or absence thereof altogether, thereby leading to death.

CYT18 is necessary for some of those introns to splice out. One could postulate plenty of adaptive 'regulatory' reasons for this, but this turns out to be quite unparsimmonious (and therefore much less likely) when compared to the following explanation.

CYT18 initially just happened to have an affinity for the intron RNA 'neck' structure, and in doing so stabilised it a bit. Now, at that point it still wasn't necessary, nor did it actually make any difference in terms of selection (the intron was spliced out with or without). However, N.crassa is among the minority in requiring CYT18 for proper intron splicing; other fungi don't need it. This suggests an accident happened.

The 'accident' was likely a mutation in the neck region that would have otherwise been lethal, but was stabilised by CYT18 in this case. This fixed the role of CYT18 in mitochondrial intron splicing in affected lineages. That simple. No adaptive explanations required, for the N.crassa system is unlikely to be better than the regular self-splicing state. This process was constructive in that complexity has been generated; and neutral in that positive selection hadn't participated. And yeah, it's still evolution.

Interestingly, this is likely how all spliceosomes came to be! Mitochondria come from bacteria, and bacteria generally are not supposed to have introns (fuzzy territory but let's leave it at that for today). Mitochondria are not supposed to have introns, and most of introns present are self-splicing (possibly coming from transposons...as a transposon, you can spread more if you politely remove yourself from interfering with the host's life). However, dependencies on random proteins like CYT18 would eventually result in arguably unnecessarily ridiculously complex structures like spliceosomes.

There's more applications of constructive neutral evolution (probably more than enough stuff for today!), and after it sinks in you become to see it more and more.

Again, as Rosie pointed out: it should be our null hypothesis.


Hopefully, I have managed to at least convince you that evolution never aspired to 'create' humanity; has neither aim, nor desire, nor direction; evolution does not strive towards complexity or progress; and that complexity can arise on its own without being called for by adaptive processes. There is no such thing as evolutionary ascent, and can we please stop desperately clinging to the last surviving strands of creationism?

*SCM visited our university a while ago, and my friends and I went to one of his talks. His talk was pretty good, although towards the end his arguments clearly began to crumble: the obsession with convergence was still tolerable, but the final note he ended on was simply insane. He actually suggested outside 'forces' guiding evolutionary processes! (I likely missed his precise argument as it was steeped in rather convoluted philosophy-speak and had little to do with actual science; but the gist of it was such.) That day, I just happened to have the Stoltzfus 1999article on me, for some inexplicable reason. So we proceeded to the front after he finished. I mentioned being quite impressed with his awareness of Paulinella (obscure Rhizarian that seems to have undergone an independent primary endosymbiosis of cyanobacteria! (=plastid)), and to give him credit, the guy does know a lot of stuff. So I brought up a rather opposite perspective on evolution to his, and showed him the paper. Interestingly, his response was along the lines of: "While this may or may not be valid, I don't care about how evolution got here; what I'm more interested about, is where evolution is going."

Now one must wonder - is it even possible, or wise of us to try to figure out the future of evolution? He does have a point there - evolutionary biology is a historical science, focusing on what happened and the mechanisms behind it, rather than predicting where things will go. He's fascinated by the more philosophical aspect. Of course, in doing so, he screams directionality. Perhaps this is his way of reconciling his religion with science; while admirable in his attempt, it seems to be driving him literally crazy. Conway Morris' religiosity has forced him to go through some intense mental gymnastics, providing yet another illustration of the disease-like properties of religion...

**While the spandrel concept is useful in biology, it may be interesting to note that the real (architectural) spandrels are not really spandrels in Gould's sense. (Dennett 1995)

References

Dennett, D.C. (1995). Darwin's Dangerous Idea

Collins RA, & Lambowitz AM (1985). RNA splicing in Neurospora mitochondria. Defective splicing of mitochondrial mRNA precursors in the nuclear mutant cyt18-1. Journal of molecular biology, 184 (3), 413-28 PMID: 2413216

LEANDER, B. (2008). Different modes of convergent evolution reflect phylogenetic distances: a reply to Arendt and Reznick Trends in Ecology & Evolution, 23 (9), 481-482 DOI: 10.1016/j.tree.2008.04.012

LEANDER, B. (2008). A Hierarchical View of Convergent Evolution in Microbial Eukaryotes The Journal of Eukaryotic Microbiology, 55 (2), 59-68 DOI: 10.1111/j.1550-7408.2008.00308.x

SLAMOVITS, C., & KEELING, P. (2008). Widespread recycling of processed cDNAs in dinoflagellates Current Biology, 18 (13) DOI: 10.1016/j.cub.2008.04.054

Stoltzfus, A. (1999). On the Possibility of Constructive Neutral Evolution Journal of Molecular Evolution, 49 (2), 169-181 DOI: 10.1007/PL00006540