More Recent Comments

Thursday, October 11, 2007

The Difference Between Fish and Humans

 
Here's a Press Release from BBSRC [The difference between fish and humans: scientists answer century old developmental question]. What's BBSRC, you might ask? Here's the answer,

The Biotechnology and Biological Sciences Research Council is one of seven Research Councils sponsored through the UK Government’s Office of Science and Innovation and invests around £336 million per annum in the biosciences.

BBSRC sustains a world-class bioscience research community for the UK. Its mission is to fund internationally competitive research, provide training in the biosciences, foster opportunities for knowledge transfer and innovation and promote interaction with the public and other stakeholders on issues of scientific interest.
Pretty impressive, eh? With a mission like that you'd expect a world-class press release, right? Let's see how they do ...
Embryologists at UCL (University College London) have helped solve an evolutionary riddle that has been puzzling scientists for over a century. They have identified a key mechanism in the initial stages of an embryo’s development that helps differentiate more highly evolved species, including humans, from less evolved species, such as fish. The findings of the research, funded by the Biotechnology and Biological Research Council (BBSRC), were published online today by the journal Nature.
Undergraduate students in my university are taught that humans and fish share a common ancestor. They have both evolved for the same length of time from that common ancestor, therefore you cannot say that one is more evolved and one is less evolved. People who say that are demonstrating a fundamental misunderstanding of evolution.
Early on in development, the mass of undifferentiated cells that make up the embryo must take the first steps in deciding how to arrange themselves into component parts to eventually go on to form a fully developed body. This is a process known as ‘gastrulation’. During this stage, the cells group into three layers, the first is the ‘ectoderm’ which then in turn generates the ‘mesoderm’ and ‘endoderm’ layers. In higher vertebrates, such as mammals and birds, the mesoderm and endoderm are generated from an axis running through the centre of the embryo. However, in lower vertebrates, such as amphibians and fish, the two layers are generated around the edge of the embryo.
Using the terms "higher vertebrates" and "lower vertebrates" is just as bad as using "more evolved" and "less evolved."
Scientists have been speculating for over a century on the difference between the embryonic development of higher vertebrates and lower vertebrates, to help answer how the simple cell structure of an embryo goes on to form the various highly complex bodies of different species. Research leader Prof Claudio Stern explains: “This is a significant find as it is a clear difference between the embryonic development of more advanced species and less advanced species. It suggests that higher vertebrates must have developed this mechanism later on in the history of animal evolution.
Scientists who use terms like "more advanced" and "less advanced" to distinguish modern species are demonstrating their ignorance. They should not be publishing papers about evolution.

The article can be found on the Nature website [Voiculescu et al., 2007)]. Some of the authors are well-known experts with impeccable reputations in the field of development biology (e.g., Lewis Wolpert). Thus, it is surprising that the press release is so bad. What does the paper actually say?

The paper describes the process of gastrulation and formation of the primitive streak in chicken embryos. It present results that support the involvement of a particular signalling pathway in this process. The data is supported by supplemental movies [Stage XIII].

The authors go on to compare the chicken pattern of cell movement and differentiation to that in Xenopus, an amphibian. Birds and mammals are amniotes and amphibians and fish are not (= anamniotes). The paper ends with a single paragraph that mentions evolutionary implications.
We propose that local intercalation in the epiblast is responsible for positioning and shaping the primitive streak and can also explain the polonaise movements without the need for long-range gradients. Convergent extension of the axial mesoderm and neural plate in anamniotes is almost certainly conserved in amniotes, but our study reveals an additional, much earlier (pre-gastrula) cell intercalation, required for morphogenesis of the primitive streak independently of mesendoderm specification. This is apparently unique to amniotes and provides a possible answer to the classical question of how evolution converted the equatorial blastopore or shield of Anamnia into the radially oriented primitive streak of amniotes.
There's nothing wrong with this. It does not claim that amniotes are "higher" than fish or amphibians and it does not claim that fish have stopped evolving.

The disconnect between the emphasis in the paper and in the press release is disconcerting. The differences in the language used to describe evolution is very troubling. The wording of the press release perpetuates the false concept of a "ladder of life" and that's not a way to advance the education of the general public. The fact that the press release quotes the senior author making the same conceptual error suggests that the errors in the press release may not be the fault of science journalists in the media department—although in an ideal world they should have been able to correct the scientists!

Voiculescu, O., Bertocchini, F., Wolpert, L., Keller, R.E. and Stern, C.D. (2007) The amniote primitive streak is defined by epithelial cell intercalation before gastrulation. Nature (advance online pubication: doi:10.1038/nature06211).

37 comments :

Anonymous said...

Isn't it true though, that some existing species show less variation from their ancestors than other species? (i.e. sharks). In that case, couldn't they be called "less evolved"? After all, in a more general sense, evolution means "change over time".

Unknown said...

They might be similar purely anatomically, but are you sure they are so similar in their physiology and biochemistry or other areas?

Anonymous said...

Ugh...I hate that "higher" and "lower" stuff too. And "less-evolved" is also pretty bad.

The common ancestor of mammals and zebrafish would have been a bony fish, back before the ray-finned and lobe-finned versions split. I don't know that there's much we could say for sure about the genetic control of development in that long-gone fish, except that it used all the genes that are regarded as homologous in modern zebrafish and, say, lungfish. Still, in terms of the observable processes of early development, you'd have to bet that it was a lot more similar to zebrafish than to chickens.

Eamon Knight said...

Grrr.
I think any science journalist (or worse yet, any scientist) who spews this higher/lower, more/less evolved crap, should be forced to write out "There is no Scale Of Nature" ten thousand times, while under that spell the evil headmistress casts on Harry Potter during the most recent movie.

Anonymous said...

I noticed that too when I read the press release. I agree, it's incorrect to use a term like "evolved," which is easily misinterpreted in even a release intended for the general public.

"you cannot say that one is more evolved and one is less evolved."

Within a larger group (for example, vertebrates), it often is obvious which sub-groups are more primitive - more like the common ancestor - and which are more derived, even if all groups cannot be ranked on a unilinear scale.

"They have both evolved for the same length of time from that common ancestor"

The same is true of, say, a squid and a species of bacteria. But which is more derived? More complex? Is anyone tempted to think of the squid as more "evolved"? I know I am.

Sure, something can be simpler and more derived. For example, an arthropod that is a worm-like parasite, or those transmissible dog tumors. Hence "devolved" - another value-laden term.

I am not an orthogenecist (though I am pro-gradist), but have a certain sympathy for progressionist language. And I'm even more tolerant of terms like "higher-lower" and "advanced." When used informally, of course.

Tupaia

Anonymous said...

I think we lose nothing in (desired) precision of meaning but lose much in (undesired) subconscious connotation by replacing "primitive" and "advanced" with "ancestral" and "derived," respectively. "Higher" and "lower" we don't need at all.

Anonymous said...

Come to think of it, don't differential rates of morphological evolution, as well as the concept of stasis and the informal notion of a "fossil species" suggest that some species are more "evolved" than others?

Tupaia

Torbjörn Larsson said...

So fish may be more ancestral than humans, more fishy. But they evolved their "fishiness" for a longer time so they could with some stretch be considered "more evolved" too...

I think that is proof enough that it is the question itself that is fishy!

Anonymous said...

Sven DiMilo makes a good good against my position. Some terms are simply too loaded with historical and ideological baggage.

Tupaia

Anonymous said...

Ack - I meant to write "a good case." (What happened?)

Tupaia

Anonymous said...

Anonymous said:

Isn't it true though, that some existing species show less variation from their ancestors than other species? (i.e. sharks).

Yes.

In that case, couldn't they be called "less evolved"?

No.

After all, in a more general sense, evolution means "change over time".

Evolution can be defined as "change over time" but this is genetic change over time - not all genetic change equates to phenotypic change.

The problem is that people are equating degree of phenotypic variation with amount of evolution, but the amount of phenotypic variation from a common ancestor does not equate with the degree or amount of evolution.

Humans and modern fish share a common ancestor, and have experienced the same amount of evolution since that common ancestor, it's just that, in the case of humans, that evolution has resulted in greater phenotypic variation from the common ancestor that it has in modern fish.

The problem of equating amount of evolution with phenotypic variation is exacerbated by the common use of terms such as "advanced v. primitive" and "higher v. lower", with the connotation that "advanced" and "higher" equalling more evolution, and "lower" and "primitive" equalling less evolution.

Better terms are "derived" and "ancestral" which reflect the true situation - that of greater or lesser phenotypic variation from a common ancestor.

Here's an example, based loosely on Zou Hongyan & Niswander L. (1996) Requirement for BMP signaling in interdigital apoptosis and scale formation. Science, 272: 738-741.

Say we have a common ancestral group of 'birds' with, not feathers, but avian scales – the scaley-baldy group. Now in one group there appears a point mutation, a single amino acid substitution affecting Bone Morphogenetic Protein (BMP) that changes Lysene at the 231 position to Argenine. In the second group, a point mutation changes Lysine at the 23 position to Argenine in BMP.

The mutation in group 1 blocks the binding of BMP to the receptor site. The result is that avian scale sites now produce feather-like structures and not scales – the feather duster condition. In group 2 the mutation does not affect the binding of BMP and so the group retains the scaley-baldy condition. If we were then to compare the two groups we would say that the feather duster group are more 'advanced' and (because we prefer our birds with feathers) 'higher' (and even more complex) than the scaley-baldy group, which, because they continue to be similar to the ancestral condition, we would class as "primitive' or 'lower'.

However, in terms of evolution, both groups have undergone exactly the same amount of evolution (1 more Lysine in the BMP).

Anonymous said...

In that case, couldn't they be called "less evolved"?

No.

However, in terms of evolution, both groups have undergone exactly the same amount of evolution (1 more Lysine in the BMP).


Does that mean that a current shark species (or a horseshoe crab) has the same genotypic "distance" compared to it's ancestor of 100 million years ago, as we do to our ancestors? If so, then we are equally "evolved". If not...

Anonymous said...

Anonymous asked:

Does that mean that a current shark species (or a horseshoe crab) has the same genotypic "distance" compared to it's ancestor of 100 million years ago, as we do to our ancestors? If so, then we are equally "evolved". If not...

The modern shark and humans have the same "genotypic" distance from the last common ancestor of sharks and humans, so if you went back 100 million years along the shark lineage and the human lineage, and measured the difference between the 100 million year ancestor in each lineage and it's modern decendent, then yes the difference would be similar in both lineages.

Anonymous said...

then yes the difference would be similar in both lineages.

You may be right, but intuitively it seems hard to believe. Sharks of 200 million years ago were phenotypically very similar to today's sharks. But our ancestors of 200 million years ago were...?

Anonymous said...

You people seem to assume evolution rates are constant over time, which is very likely not true.

Even ignoring punctuated equilibrium etc
which is a little controversal:

A simple non controversal example would be comparing a species with long reproduction/life time over one that has a very fast cycle.

One day fruit flies clearly evolve faster than humans.

And the reason humanity wasn't very successfull against AIDS was that
HIV is evolving very quickly. Compare that to the success of the mammuts or Neanderthals against humanity.

A. Vargas said...

there are no primitive groups; only primitive traits.

Anonymous said...

Sharks of 200 million years ago were phenotypically very similar to today's sharks.

Actually, the vast majority of extant sharks result from a radiation over the past 50-100 my, similarly timed to the radiation of mammals; most of the shark fauna of 200 mya would appear very bizarre today (not that there aren't some pretty bizarre extant sharks!).

And I agree that, to the extent that "more or less evolved" means anything, it means that the rate of evolution is correlated to the average generation time of a lineage.

Diane said...

This is a topic I struggle with when teaching undergrads. "More and less evolved" just makes too much intuitive sense to them, and the "Homo sapiens wins!!" mentality is reinforced in so many ways that it's hard to shake.

What is the best way to describe what makes a living fossil species so special among species? Or is the answer, as I suspect it might be, "Nothing," or perhaps "Media hype"?

Anonymous said...

Anonymous wrote:

Sharks of 200 million years ago were phenotypically very similar to today's sharks. But our ancestors of 200 million years ago were...?

. . . probably nested in the Mammaliaformes, (for example Hadrocodium). This is a group transitional from the cynodonts (mammal-like reptiles) to mammals.

Hadrocodium was a small shrew-like animal with most of the typical mammal characters.

Critically though, we have the same basic body plan as the first mammals - a head at one end, a tail at the other end, and a backbone and 4 limbs in the middle. Ribs, lungs, heart, liver kidneys, and brain were all present in the first mammals. We may look different, but we possess no novel structures that were not present, or that are not derived from structures present, in the first mammals. Our outward differences have been achieved by tinkering around the edges.

Intuitively it may look like we are significantly different from our 200 million year ancestors compared with the shark lineage, but actually when you look in detail there isn't very much difference at all.

Anonymous said...

Anonymous wrote:

You people seem to assume evolution rates are constant over time, which is very likely not true.

Over reasonable time scales, the rate of evolution is probably constant. Do you know of any mechanisms that would speed up the rate of evolution over reasonable time frames, in one group over any others? The reproduction time frame isn't one.

A simple non controversal example would be comparing a species with long reproduction/life time over one that has a very fast cycle.

One day fruit flies clearly evolve faster than humans.


. . . and bacteria can reproduce in a few hours. A simple calculation shows than in only a very short period of time we should be knee deep in bacteria, yet we are not. If fruit flies reproduced every hour we should also be knee deep in fruit flies, but we aren't. In reality fruit flies have an annual breeding cycle, like most organisms (I think). Over evolutionary time scales, the reproductive cycles generally even out (e.g. given the shark and human example, since the last common ancestor of the two, the participants in each the two lineages have probably had life/reproductive cycles of similar length.) But even if they didn't there are other things need to be taken into consideration, such as population size, survival rates of offspring, geographic distribution.

Fruit flies are being artificially influenced, and represent an extraordinarily small population, and is not representative of what is happening in the wild population. One could argue that humans are also being artificially influenced, with people surviving to reproduce, thanks to our science, that would normally not survive.

And the reason humanity wasn't very successful against AIDS was that
HIV is evolving very quickly. Compare that to the success of the mammuts or Neanderthals against humanity.


I think HIV is also under artificial selection. What's happening is that our treatment system is artificially selecting from normal variation in the population. We are not influencing or speeding up the mechanism of evolution in HIV, our treatment system artificially selects variants caused by mutation, by destroying the competition. (a simplistic explanation).

Anonymous said...

Diane asked:

What is the best way to describe what makes a living fossil species so special among species?

'Living fossils' provide important phyiological, anatomical and environmental information that can help us 'flesh out' our understanding of the fossil form.

Imagine what information we could get from a living dinosaur about the lives of fossil dinosaurs - thermoregulation, growth rates, running speeds, gestation times, time from laying to hatching, behaviour such as looking after young, feeding, mating, cooperation, etc.

Larry Moran said...

anonymous says,

You people seem to assume evolution rates are constant over time, which is very likely not true.

But empirically it is true. That's the basis of what we call the "molecular clock." If you plot the changes in nucleotide or amino acid sequences for a bunch of species (including sharks) then all the species will have changed by about the same amount from their common ancestor. (No more than two fold difference for a particular protein.)

It was one of the prediction of neutral theory that the rate of evolution would be close to the rate of mutation and it turns out to be true to a remarkable degree. Since the rate of mutation is mostly due to the rate of misincorporation by DNA polymerase, it is pretty much invariant—all species have the same DNA polymerase. Thus, the rate of mutation doesn't change significantly either.

You can't stop evolution. You may be able to slow down or speed up natural selection but that's not the same thing.

Anonymous said...

"Imagine what information we could get from a living"^ nonavian ^"dinosaur about the lives of fossil dinosaurs..."

TheBrummell said...

Since the rate of mutation is mostly due to the rate of misincorporation by DNA polymerase, it is pretty much invariant—all species have the same DNA polymerase. Thus, the rate of mutation doesn't change significantly either.

I think this point needs some elaboration in the context of the wide variation in eukaryotic body sizes and generation times. Yes, some organisms experience a great many life cycles in the time of one life cycle of another organism. The examples presented here of Drosophila melanogaster and Homo sapiens will do nicely.

Notice, also, that an individual human has many more cells than an individual D. melanogaster. Given that the majority of mutations are the result of errors during DNA replication, it is perhaps not surprising that total per-year mutation rates are fairly close to equal between species.

The relevant number of cell generations per organism generation is that of the germ line - how many cells replicate to go from gamete to gamete? In large-bodied, long-lived species such as H. sapiens, that number of cell generations is much larger than the corresponding number of cell generations in D. melanogaster and other small-bodied, short-lived species. So the opportunities for replication errors to arise in a developing human are more numerous than in a developing "fruit fly".

Mutation rate differences between species are probably much more influenced by things like effective population size and per-cell metabolic rate than by body size or generation time.

Larry Moran said...

thebrummel says,

The relevant number of cell generations per organism generation is that of the germ line - how many cells replicate to go from gamete to gamete? In large-bodied, long-lived species such as H. sapiens, that number of cell generations is much larger than the corresponding number of cell generations in D. melanogaster and other small-bodied, short-lived species. So the opportunities for replication errors to arise in a developing human are more numerous than in a developing "fruit fly".

I described the overall mutation rate per generation in humans in "Mutation Rates." The number is not likely to be much different in fruit flies.

Mutation rate differences between species are probably much more influenced by things like effective population size and per-cell metabolic rate than by body size or generation time.

It's a complicated issue. There's also the problem of fixation that you allude to here. Mutation rate and fixation rate are two different things even though they are identical under random genetic drift. I don't know of anyone who has a firm handle on all of the factors.

I think it would help if we all agree to use "mutation rate" as the rate of mutation per nucleotide per replication. Then we can use "overall mutation rate per generation" to refer to something else. We get into trouble when people start using the same terms to mean two different things.

Finally, not everyone agrees that there's a correlation between mutation rates and metabolic activity. I know you come from a lab that holds to that rather fanciful notion but that doesn't make it right. :-)

Timothy V Reeves said...

“Advanced species and less advanced species”??? I see no problem with this terminology. Larry, you might like to explain to this member of the public why I have got this so badly wrong. This is my take on it:

If the evolution of biological structures is abstracted as a series of assembly steps arrived at via stochastic paths (entailing the traversal of at least some blind alleys), then it seems clear to me that some structures require more steps in their construction than others.

This is fairly obvious (to me at least) when a stage of evolution depends on organic precursors or biological antecedents being in place, thus implying that there are variations in the number of construction steps required by different biological structures. For example, when a structure has a nested “components of components” organization then the number of construction steps will increase as nesting level increases. Moreover, some structures would appear to present a much ‘harder problem’ for evolution than others: e.g. Meta problems like evolving ‘evolvability’ or evolving heuristics capable of creating heuristics. (That latter ‘problem’ might have been ‘solved’ in humans).

Frankly I don’t see why Prof Stern’s references to ‘more advanced’ and ‘less advanced’ species warrants the rather strong comment that he shouldn’t be publishing papers about evolution. My lay interpretation of Stern is that on the basis of the foregoing considerations ‘advancement’ refers to the natural sequencing of species defined in terms of construction steps. My perception is that the ‘ladder of life’, if based on construction steps, is a theoretically robust concept. However, as with cladistic ‘hierarchies’or ‘trees’, in spite of this robustness of concept, it may be difficult to practically place species in their correct topological relations.

The above, of course, doesn’t necessarily give license to valued judgments about ‘superiority’ of structure, but it is very difficult to discuss the subject in neutral language as naturally enough human language is laden with the noise of association and values: hence terms above like: ‘hierarchies’, ‘trees’, ‘ladder of life’ ‘more advanced’, ‘harder problem’, ‘solved’ etc
.

Fred Hoyle was frank, I believe, about his ‘perfect cosmological principle’ being motivated by the atheism of his youth. He sought to impose on the universe a perfect symmetry in time and space. To him big bang was anathema because it posited a very special event on the universe, an event giving it an asymmetry in time, a center, a place of deep mystery and significance. Much better to dilute that significance over an infinite and homogeneous universe, thus having the effect of suppressing questions about its nature and helping to deny theists a foothold with an glaring ‘gap’ in science. However, with the advent of multiverse versions of big-bang Hoyle’s super-copernicanism is back on the agenda.


Be that is it may, attempts to impose an analogous symmetry on the complexity space of biological structures are doomed to failure. In complexity space everywhere does not look the same. Also, there is a time asymmetry in evolution; the process doesn’t start with complex structures – complex structures are going to appear later rather than sooner. The principle of mediocrity can’t be applied to biological structures. Scientifically speaking organisms might be all equal – but some are more equal than others. In any case I’m feeling particularly equal at the moment as I sure I’ve got this one right!

But if I have got it wrong, Larry you have only got yourself to blame. I’ve been reading your blog for several few weeks now and clearly I still haven’t understood this evolution business! Call yourself an educator? Perhaps I ought to complain to the Her Majesty!! Sir Larry? No chance! You’ve got some explaining to do!

Anonymous said...

Timothy V Reeves wrote:

“Advanced species and less advanced species”??? I see no problem with this terminology. Larry, you might like to explain to this member of the public why I have got this so badly wrong. This is my take on it:

If the evolution of biological structures is abstracted as a series of assembly steps arrived at via stochastic paths (entailing the traversal of at least some blind alleys), then it seems clear to me that some structures require more steps in their construction than others.


You appear to be saying that more complex pathways are more evolved than less complex pathways. If my reading is correct then there are two problems. One is that this assumes there is some directionality to evolution, that it is striving for complexity. The other is that the position fails to take into account instances were evolution results in a reduction of components to produce the same structure. I.e. according to the view you appear to be suggesting, an organism that evolved a more efficient pathway to produce a structure, leading to less steps in the development process, would result in devolution, or a retrograde step back towards a less advanced species.

In other words the most convoluted, tortuous, complex process is more advanced than a simpler process that achieves the same result using fewer steps. There are many complex (i.e. many component) processes in biology, but many are needlessly complex - a biochemist could devise a simpler process, using less steps, to produce the same outcome. These processes are not complex because there is an advantage to them being complex, or that evolution is striving for complex solutions, but because evolution can only work by co-opting the processes and components available to it.

In manufacturing, the more 'advanced' process is one which uses the least number of steps to produce the desired result.

You appear to be equating evolution with complexity. This is problematic because we have no good definition of what constitutes "complex". While intuitively we mammals appear to be more 'complex' than single celled organisms, are mammals more 'complex' than any other vertebrates?

Frankly I don’t see why Prof Stern’s references to ‘more advanced’ and ‘less advanced’ species warrants the rather strong comment that he shouldn’t be publishing papers about evolution.

Larry was commenting on the language used in the press release associated with the paper, not on the language used in the paper itself.

My lay interpretation of Stern is that on the basis of the foregoing considerations ‘advancement’ refers to the natural sequencing of species defined in terms of construction steps. My perception is that the ‘ladder of life’, if based on construction steps, is a theoretically robust concept.

Species and groups of species do not 'advance'. As a group they evolve, expand into new areas of morphospace, contract from other areas. The 'natural sequence' or lineage is the result of the somewhat random evolutionary 'walk' that started at from A and ended at form D. The sequence A-B-C-D did not occur from a striving for complexity. We could not say at form A that D would be the result, nor that it would be achieved via B-C. In fact the true sequence would be more like A-P-H-V-D. And D does not necessarily have to be more complex than A.

There is no "ladder of life". It's not just an incorrect analogy, it’s a bad one. It is incorrect because it does not remotely reflect the actual situation. And it's bad because it entrenches the very value judgements you clam it doesn't -

The above, of course, doesn’t necessarily give license to valued judgments about ‘superiority’ of structure, but it is very difficult to discuss the subject in neutral language as naturally enough human language is laden with the noise of association and values: hence terms above like: ‘hierarchies’, ‘trees’, ‘ladder of life’ ‘more advanced’, ‘harder problem’, ‘solved’ etc

Hence terms like "upward", "above", "advanced" "higher", 'lower", "below", "downwards". It also perpetuates judgements about purpose and direction in evolution.

Scientifically speaking organisms might be all equal – but some are more equal than others. In any case I’m feeling particularly equal at the moment as I sure I’ve got this one right!

This is the problem. What do you mean by "more equal"? More numerous? Can live in more environments? Can reproduce more rapidly? Can exists in the absence of light? In all of these, bacteria have it all over vertebrates. Clearly 'complexity' doesn't help us there.

Arguments about ladders and complexity usually have more to do with entrenching the view that humans are special, or more equal, than they do of accurately reflecting evolution.

But if I have got it wrong, Larry you have only got yourself to blame. I’ve been reading your blog for several few weeks now and clearly I still haven’t understood this evolution business! Call yourself an educator? Perhaps I ought to complain to the Her Majesty!! Sir Larry? No chance! You’ve got some explaining to do!

It's a bit harsh to say that after 'several few weeks' of reading a blog entitled "Strolling with a Sceptical Biochemist", it's the author's fault for any misunderstanding of evolution you may have?

Timothy V Reeves said...

Thanks for the reply Chris. Quoting myself:

It seems clear to me that some structures REQUIRE more steps in their construction than others.

Sorry, but I realize that this statement doesn’t make the emphasis clear: I’m talking about the minimum path by which a stochastic process could be expected to arrive at a structure: hence the word ‘require’. Actually, if a process did arrive at a structure by the absolutely minimum path it is unlikely to be stochastic because such processes do not work at maximum efficiency as construction algorithms. Hence more relevant is the expectation value of the number of construction steps given the random walk nature of the process. This expectation value would be a mathematical property intrinsic to the structure itself and would not necessarily be equal to the length of the actual route by which the structure was arrived at; but I expect a correlation between the magnitude of this mathematical quantity and the actual number of construction steps

One point you have put your finger and which I should have taken cognizance of is the distinction between function and structure. A structure’s complexity might, as you suggest, be over kill given the function it serves. Hence, we really need to talk about ‘function structure’ rather ‘than actual structure’.

Although I’m not an expert in the subject of computation (actually I’m not an expert at anything apart from being augmentative) I think I grasp the subject well enough to apply it to evolution. For some biological structures the expectation number of construction steps is going to be greater than for other structures, just as some computations are necessarily longer than others. I think this is fairly obviously for nested structures like organisms that are communities of cells; in this case the construction entails both cell construction and the construction of a cell community.

So, in short I believe we can lay biological structures out on a scale of construction complexity. (emotively ‘The ladder of life’!) This construction complexity lurks in the biological background just as do cladograms, although in both cases it may be practically difficult to correctly relate organisms in these respective spaces. In any case I think it is intuitively clear that bacteria requires less construction steps than say an animal like a cat. In that sense a cat is ‘more equal’ than bacteria – it is very probably going to appear much later in the evolutionary processes than bacteria because of a construction step disparity. As I said, evolution is asymmetrical in time, just as is big bang theory. The universe is thermodynamically and dynamically in disequilibrium. It is also in morphological disequlibrium - we believe slots in morphospace have been slowly populated over the course of evolution, and there was a time when they were not populated on Earth.

In the light of the above considerations I still can’t see what is wrong with Prof Stern using the terms ‘less advanced’ and ‘more advanced’ (which the report quoted by Larry claims he used), if those terms are meant in the technical sense I am groping toward above.


I like your the term ‘morphospace’; in my imagination I picture the metaphor of a ‘cloud’ of probability probing that space and ‘finding’ the ‘quasi-equilibrium’ states of survival solutions. Is this process directed? Yes and no. ‘No’ because, given our physical regime, standard ‘balanced’ statistics applies to the dynamics of the probability ‘cloud’; that is, there is (I guess) no special weighting on the construction paths consistent with the physical regime. ‘Yes’, because evolution is sensitive to the exact details of the physical regime – the latter must effectively cut down the size of morphospace to a sufficiently small size for the probabilistic probings of evolution to work in realistic time. Moreover, certain physical regimes don’t work at all, like for example if the universes were just composed of electrons. So our physical regime is special – special enough to make evolution work like a computer suitably programmed to run genetic algorithms.

This now leads us into multiverse theory: multiverse theory attempts to reinstate ‘mediocrity’ by merging the apparently special conditions of our physical regime into an enormous background of randomness. Our physical regime is then no longer special but is seen as statistically expected given a sufficiently large multiverse. Hence my comments about Hoyle. (Interestingly Hoyle turned to a kind of eccentric ‘scientific’ theism later in life)

Some things scare theists and some things scare atheists. Many theists are scared by evolution – it cuts across their theology. But many atheists are scared by ‘special conditions’ because it is a barrier behind which theists may claim God is hiding. Hence atheists seek to neutralize the question begging effect of special conditions with a generalized principle of mediocrity (or super-copernicanism).

I am myself a theist. I try hard not to allow myself to become emotionally uptight about either the theist of atheist options and I continue to probe them both, but for a variety of reasons I won’t bore you with here, theism now has the ascendancy in my mind.

Don’t take all that stuff I said about ‘Sir Larry’ , ‘Her Majesty’, and Larry being a bad educator too seriously!! It’s just a wind up! I have a very abstracted view of evolution, so abstracted that really I need to get up to speed on the details. Hence I read Larry’s very educational blog in order to do this. As for his grumpy atheism, I like it: it’s like having hot mustard in a ham sandwich! I love a bit of contention!

Anonymous said...

Timothy V Reeves said...

So, in short I believe we can lay biological structures out on a scale of construction complexity. (emotively ‘The ladder of life’!) This construction complexity lurks in the biological background just as do cladograms, although in both cases it may be practically difficult to correctly relate organisms in these respective spaces. In any case I think it is intuitively clear that bacteria requires less construction steps than say an animal like a cat. In that sense a cat is ‘more equal’ than bacteria – it is very probably going to appear much later in the evolutionary processes than bacteria because of a construction step disparity.

The issue is that you are equating 'complexity' with 'advanced' and, by inference, that more complex organisms are somehow 'better', 'more evolved', 'more equal, than less advanced forms. The problem is that the term "advanced" has inescapable connotations of 'better', 'moving forward', 'moving up', and other positive inferences compared with 'less advanced'. These connotations cannot be divorced for the term and so inevitably lead to reinforcement of the ladder, or tree, or directional, or 'more complex is best' view of evolution.

More importantly, apart from the fact that your definition of 'advanced' is somewhat circular, 'advanced forms are complex forms that are more complex than less complex forms', what information does the term 'advanced forms' give us? The answer is, not much.

Advanced forms are more complex than less advanced forms. So what? Is this significant in evolutionary terms? The answer is no.

Your average bacterium might say, "you keep your advanced, convoluted, and extraordinarily resource intensive, constructional complexity, and we'll keep our less complex, less advanced constructional complexity, and we'll just continue to be the most successful group in the history of life on Earth.

Which of the following statements seems more accurate to you?:

Manufacturer1: My process results in exactly the same outcome as my competitor, but my process contains more steps, is therefore more complex, and so more advanced.

Manufacturer2: My process results in exactly the same outcome as my competitor, but my process contains less steps, is therefore less complex, and so more advanced.

Organisms are all looking to produce the same outcome in evolution terms. Some organisms attempt it through more complex pathways than others, but that doesn't make them more advanced. Why not say that, given the same outcome, the organism that achieves it in the most simple way is the most advanced?

I like your the term ‘morphospace’; in my imagination I picture the metaphor of a ‘cloud’ of probability probing that space and ‘finding’ the ‘quasi-equilibrium’ states of survival solutions. Is this process directed? Yes and no. ‘No’ because, given our physical regime, standard ‘balanced’ statistics applies to the dynamics of the probability ‘cloud’; that is, there is (I guess) no special weighting on the construction paths consistent with the physical regime. ‘Yes’, because evolution is sensitive to the exact details of the physical regime – the latter must effectively cut down the size of morphospace to a sufficiently small size for the probabilistic probings of evolution to work in realistic time.

I didn't invent the term, it's been in use for a while now. But morphospace is a much better way of visualising evolution than a ladder or tree.

Imagine a cube filled with water. This represents the maximum extent of morphospace. Now, black ink is injected into the cube. The black ink represents the region(s) of morphospace occupied by organisms. The movement of the ink during and after injection is a much better description of evolution than a ladder or tree. There is no directionality, no forwards or upwards (and consequently no backwards and downwards). Groups are free to expand and contract along any axis x, y, or z.

Just because one set of coordinates within the ink corresponds to greater complexity, does not make that area 'better' or more advanced, in terms of evolution, than any other coordinates.

Also there is no process of direction as we would normally understand the term. Groups entering regions of morphospace may find that this blocks their access to certain other areas of morphospace. There are constraints, but no directions to a definite end result (turning left on axis x does not mean that you will inevitably become a reptile – there is no card that says "go to reptilia, go directly to reptilia, do not pass mammalia, do not collect fur!)

Timothy V Reeves said...

Thanks again Chris for the reply and for grappling thoughtfully with my comments. It’s bedtime here I’m afraid! I’ll be back!

Timothy V Reeves said...

I’ll accept that I should drop all these provocative terms like ‘ladder of life’, ‘more equal’, ‘advanced’, which as you suggested are over laden with connotations of ‘better’. Your average human being might well reply to your average bacterium with this:

“Stuff all this business about being the most successful group in the history of life; you’re so boring! All you do is multiply and you all look the same. You live to reproduce and reproduce to live. In any case what does ‘most successful group’ mean? Rapidity of reproduction? ability to colonize the greatest number of environmental niches? robustness? Greatest biomass? Rapidity of evolution? Genus Longevity? You’re just playing the numbers game, just more of the same; you know nothing about quality of existence - feelings, thoughts, reason, creativity, language, love, etc etc.”

This argument is really only rightly taken up in a philosophy publication, and in any case such publications are going to be biased, because most of them have been written by humans for humans. I have yet to locate any philosophical publications written by bacteria for bacteria.

However what the above argument does tell us is that evolutionary outcomes are qualitatively very diverse. Moreover, value laden terms apart they are also quantitatively very diverse and there is a hard mathematical parameter that allows us get one handle on this diversity: namely, the minimum number of construction steps required to generate a structure, a quantity which is a function of structure. Your metaphor of the manufacturer raises a valid issue, but in setting the ‘outcome’ as a constant, rather than a variable, it misses the question of what effect varying the outcome has on the number of steps needed to generate it. The ‘outcome’ will have a structure, and therefore we can ask the question about the minimum number of construction steps required to generate that outcome. This minimum number of construction steps will be a mathematical quantity intrinsic to the outcome itself, and will vary from outcome to outcome. In the case of manufacturer 1 he might be generating the outcome inefficiently, but what we do know is that the manufacturer 2 will NOT be generating the outcome in less steps than the mathematically determined minimum value for that outcome.

It is clear to me that outcomes in evolution are not the same (although in scientific papers we, of course, should resist evaluating those outcomes and stick to description). Some ‘outcomes’ are intrinsically more complex than others in the sense that they demand a greater minimum number of construction steps True, this mathematical quantity is a relatively gross concept and misses a lot about complexity we may be interested in, but nevertheless it is very informative.

You comment as follows:

Advanced forms are more complex than less advanced forms. So what? Is this significant in evolutionary terms? The answer is no.

The answer is, I believe, yes and no: No because there need be no morphotropic directionality in the ‘random walk’ process of evolution and ‘yes’ because of the following:

Although your morphospace cube captures some of the features of evolutionary diffusion it suggests it has a homogenous ‘ everywhere looks the same’ quality. That this is not true can be intuitively appreciated if we ask this question: At what point should we introduce the liquid? At the point that represents a cat? At the point representing a bacterium with a ‘engineered’ looking flagellum? At the point representing a prototype replicator? At a point representing a prebiotic precursor? The whole point of evolution is that it attempts to explain a fundamental asymmetry in nature: that is, if we introduce the fluid at the ‘simpler’ points, in time complex outcomes develop. Those more complex outcomes will have a higher minimum number of construction steps and so are less probable and will most likely appear sooner rather than later. Above all, structures with a high number of construction steps number will not appear instantaneously ‘IDiot-wise’ from who knows where. The diffusing fluid may not be morphotropic, but the process of evolution has direction, an arrow of time, just as does big bang theory. As I said before, the universe is morphologically in disequilibrium as it is dynamically and thermodynamically.

It’s easy to glibly talk about evolution: lots of assumptions have to be made about the speed of diffusion and the physical regime which has to cut out lots of avenues for an indifferently diffusing fluid if it is to eventually arrive at cats and dogs and their masters in realistic times. We are assuming that these are tuned to allow evolution to have the efficacy to reach complex organized structures in a mere 10^9 years.


BTW: What’s your view on theism? As I said, I am a theist myself.

Anonymous said...

Terrence Deacon: "Biologically, we are just another ape. Mentally, we are a new phylum of organisms."

http://www.washingtonpost.com/wp-srv/style/longterm/books/chap1/symbolicspecies.htm

Tupaia

Anonymous said...

Your metaphor of the manufacturer raises a valid issue, but in setting the ‘outcome’ as a constant, rather than a variable, it misses the question of what effect varying the outcome has on the number of steps needed to generate it.

But in terms of evolution there is only one outcome - successful representation in the next generation. Fail that and it's the end of the line, game over. Everything that doesn't directly lead to that outcome is window dressing.

Don't get me wrong, I am not denigrating the achievements of humans. Being able to extract meaning from daubing pigment on chapel ceilings in Rome is amazing, but then so to is Staphylococcus aureus. In an environment that would have either eradicated Homo sapiens, or at least bombed them back to the stone age, Staphylococcus aureus has survived and is now thriving.

There are many pathways to successful representation in the next generation, some are simple, some are complex. Can we extract meaningful information regarding evolution from a quantitative measure of complexity? Maybe we can, but certainly not an indication of successful representation in the next generation.

A bacteria may respond to your response thus:
"you can keep your quality of existence - feelings, thoughts, reason, creativity, language, love, etc, as they are irrelevant distractions to the only outcome in town, and we'll still be here when you are a ghostly outlier on a contracting axis of evolution."

So I would rewrite your paragraph here:

It is clear to me that outcomes in evolution are not the same (although in scientific papers we, of course, should resist evaluating those outcomes and stick to description). Some ‘outcomes’ are intrinsically more complex than others in the sense that they demand a greater minimum number of construction steps True, this mathematical quantity is a relatively gross concept and misses a lot about complexity we may be interested in, but nevertheless it is very informative. as:

"It is clear to me that while the outcome in evolution is the same, the pathways to achieving that outcome are not the same (although in scientific papers we, of course, should resist evaluating those pathways and stick to description). Some ‘pathways’ are intrinsically more complex than others in the sense that they demand a greater minimum number of construction steps. True, this mathematical quantity is a relatively gross concept and misses a lot about complexity we may be interested in, but nevertheless it is very informative."

Although your morphospace cube captures some of the features of evolutionary diffusion it suggests it has a homogenous ‘everywhere looks the same’ quality. That this is not true can be intuitively appreciated if we ask this question: At what point should we introduce the liquid?

The morphospace analogy is not perfect, but it is better than the usual tree approach. In fact, one of its strengths is precisely the point you raise. At what point should we introduce the liquid? The answer is - it doesn't matter. It could be in the middle towards the top, the bottom or the sides. Evolution expands outwards from the starting point. With trees or ladders you are forced to start at the bottom and then upward is the only direction open to you. With the morphospace example, unless you start ridiculously close to a corner evolution is free to expand in three dimensions. It doesn't matter if you start in the centre (0,0,0 using cube-centred coordinates) or anywhere else (0,246,10987604539).

It’s easy to glibly talk about evolution: lots of assumptions have to be made about the speed of diffusion and the physical regime which has to cut out lots of avenues for an indifferently diffusing fluid if it is to eventually arrive at cats and dogs and their masters in realistic times. We are assuming that these are tuned to allow evolution to have the efficacy to reach complex organized structures in a mere 10^9 years.

I would argue that there is no tuning involved. Conditions on Earth allow life to exist. If we were to look at the situation 1 billion years ago, would we remark upon the efficacy of evolution to produce complexity? Nope. There would be very little. However, in the past billion years there have been no fundamental changes in speed of diffusion and the physical regime and yet complexity has evolved. It is conceivable that is we were to run the Gouldian tape again over the history of life on earth that, in some cases no complexity would evolve at all, or that it may take much longer.

However, given the conditions on Earth, and the mechanisms of evolution, my gut feeling is that some form of complexity would evolve, but whether that outcome has been tuned into the system, and whether it would result in organisms that daub paint on chapel ceilings in Rome and domesticate carnivorous mammals, is another question entirely . . . :-)

BTW: What’s your view on theism? As I said, I am a theist myself.

I'm an atheist

Timothy V Reeves said...

OK Chris, let’s see if we can take this a bit further.

I can’t argue that an aptitude for quasi-eternal self-perpetuation (via replication) isn’t critical, but we just can’t ignore the degree of complexity that a genus maintains by successfully pitching it into the next round of replication. True, complexity is a vague notion, but we have a sufficient qualitative grasp of it, and in these days of computational theory some semblance of a quantitative grasp, to understand that an organism which is clearly perpetuating a much larger burden of complexity than bacteria, is doing something far more significant than the latter. In the self-perpetuation stakes what and how much is being perpetuated is at least as an important an issue as the perpetuation powers of bacteria measured simplistically and quantitatively in terms of longevity of genus and niche occupation. Let’s face it, ‘organisation’ as a physical phenomenon is at least, if not more interesting than energy, temperature, pressure, gravitation etc.

As you say the ‘higher’ organisms (‘higher’ in terms of complexity) may not be as robust as a ‘simple’ staphylococcus aureus (which I had to look up), but then look at what the ‘higher’ organisms are perpetuating – a whole new ball game in terms of organization – something far from trivial. That organization has one very fascinating character – it is self-referencing – if an organism is highly complex, it must have the necessary complexity to carry and perpetuate that complexity; in other words it is complex because it is complex! What we then have is a non-linear feedback effect, (perhaps the sort of thing that Sir Paul Nurse was talking about). This concept is likely to knock the socks off anyone who is stuck with a linear conception of organization. Don’t be distracted by the indifference of the ‘diffusing fluid’ – self-sustaining organisation has fascinating intrinsic properties no matter how it is arrived at.


On the introduction of the fluid: Yes, accepted, an organism could conceivably de-evolve because the fluid is not morphotropic (would flightless birds fit into this category?) However, we have to say this: there is a basic asymmetry in evolution demanded by the conjunction of the starting conditions, the mathematical properties of stepwise construction and an ‘isotropically diffusing fluid’.
In fact I thought that this asymmetry is where the explanatory power of evolution resides. It explains why an Achaean Earth had very little going for it complexity-wise and why we now see complex organisms. I think that Steven Jay Gould captures this well in his graph showing numbers of organisms against complexity. The ‘fluid’ of evolution diffuses and away from the wall of minimum complexity, just as gases released from volcanic eruptions of a primeval planet diffuse to reach a maximum height. True, some gas molecules, in their random meanderings, will move downwards, but there is a bulk upwards migration of probability populating the levels available, given the physical constraints that ‘tune’ the probabilities. (I have to be careful and say that it is the probability that ‘migrates’ because it is possible that the probabilities don’t come up trumps and nothing actually evolves). In terms of evolution the phrase that captures the analogous movement in morphospace is ‘The cosmos is in morphological disequilibrium’. But now we hit another potential non-linearity; the organisms ‘stumbled’ upon, might be so subtle that they start to warp and bias the flow of the diffusing fluid. For me this disequilibrium and non-linearity is the trigger for conjectures about theism, but I won’t go into that.
And this brings us on to ‘tuning’. This, I believe, is a conclusion that cannot be avoided. A gas of ‘model’ electrons will evolve nothing: the physical regime has to have the ability to reach and lock in self-sustaining organization should it come about. A tin of paper clips sitting on a table will do nothing. Now open the tin and open up each paper clip, then close the tin and shake it up: in a short while you end up with networks of paper clips. Clearly the physical regime impinges upon the reachability and fixability of a self-sustaining complexity. You, might as an atheist, however, object to the word ‘tuning’ as it hints at somebody out there tinkering around with the physical system. Nevertheless, mathematical facts suggest that not every physical system has the probability to evolve anything. From your angle the efficacy of our physical regime to ‘compute’ organisms is simply a brute fact, just as for me theism is a brute fact – well actually let me concede a ‘brute conjecture’.
We (and apes, cats, dogs) are special - more special than bacteria. Why? Because they cut across that bland ‘more of the same’ type mathematics. They represent discontinuities and asymmetries. And discontinuities and the heterogeneity of asymmetry interest us because we are complex creatures and complex self-sustaining creatures are curious and alert to discontinuities and changes in their environment. Homogeneity is boring and lacks significance; discontinuities are interesting and stimulating. Science is human because the focus our science on is based on valued judgments about what we should study. Accordingly human science identities a discontinuity between self-sustaining high organisation and homogenous matter as something of great interest and value. I suggest that one of the underlying motivating factors behind multiverse theory is that it attempts to remove the mysterious discontinuities inherent in our cosmos by blending them into a bland ‘more of the same’ type background. This is why Hoyle so hated big bang, but the irony was that in time he discovered a weird kind of theism!

Anonymous said...

.. to understand that an organism which is clearly perpetuating a much larger burden of complexity than bacteria, is doing something far more significant than the latter.

Is it? What is the "significance" of complexity in evolution?
Complexity is not a 'goal' of evolution, it is a mechanism to achieve the 'goal' of evolution. We can celebrate complexity and the fact that such complex mechanisms are the by-product of a natural process, but by the same token we should also celebrate 'simplicity' because that also delivers on the 'goal'.


In the self-perpetuation stakes what and how much is being perpetuated is at least as an important an issue as the perpetuation powers of bacteria measured simplistically and quantitatively in terms of longevity of genus and niche occupation. Let’s face it, ‘organisation’ as a physical phenomenon is at least, if not more interesting than energy, temperature, pressure, gravitation etc.

Complexity of form does not translate to complexity of the genetic information passed on. Many 'simpler' organisms have far larger genomes than humans (for example). Nor is there any evidence that humans have significantly more genes to pass on than less complex forms. Clearly it is not how much is being perpetuated that is important, but what is done with it subsequently. Indeed may of the key genes associated with developing complexity are homologous to genes in insects and other "simpler' forms. So if what is being passed on does not correlate with complexity, how much is being perpetuated is not important.

I agree, organisation is important, it is interesting, and is worthy of study. But let's not assume that it has greater stature in evolution. It's more important to us, let's not make the mistake of therefore assuming that it is more important to evolution.

"On the introduction of the fluid: Yes, accepted, an organism could conceivably de-evolve because the fluid is not morphotropic (would flightless birds fit into this category?)"

This is the problem inherent in the directional view of evolution. Organisms do not "de-evolve" But, using the directionality view, any evolution that appears to repeat ancestral conditions is viewed as a retreat, a backward step. However, the ratites (the group of flightless birds such as emus) have not de-evolved but have entered an area of morphospace in which they are superficially like their ancestors. In fact their development could almost be considered more complex because they start in the same way as flying birds but changes are then made that override the normal development, or shut down 'normal' development before it has run its full course. Note they have not 'lost' the ability to develop into flying birds, but that process has been interupted.

The ‘fluid’ of evolution diffuses and away from the wall of minimum complexity, just as gases released from volcanic eruptions of a primeval planet diffuse to reach a maximum height.

You continue to use directional metaphors. There is no "wall of minimum complexity" and there is no movement "away" The point of 'minimum complexity' is somewhere in morphospace away from any confining walls. Movement is outward not up. Also you appear to be equating evolution with complexity.

Evolution has resulted in complexity, but most evolution has not produced complexity. The vast majority of evolution has not resulted in complexity. The directional view of evolution tends to emphasis complexity as the main result of evolution. It isn't. Modern bacteria have undergone the same amount of evolution from their ancestors in the last 500 million years as we have from our first chordate ancestors. We have focussed on the increase in complexity because its result contains us, and that has influenced our view of evolution – that complexity is more important, that complexity is the main result of evolution, that complexity is the only result of evolution, that evolution can be measured as the rise of complexity, etc. The result is that, evolution that has not resulted in complexity, has been ignored, or waved away as not being as important. The tree of life metaphor is a classic example of this. It represent evolution solely as the rise of complexity, whereas, that rise represents only a small part of evolution. In the morphospace metaphor, the rise of complexity represent only a small area of morphospace, only a small fraction of the amount of evolution activity that has, and is, taking place.

By the way, as this thread has now cycled off the front page of Larry's blog, I suspect no one else is reading these exchanges. So if you want to move on, that's fine.

Timothy V Reeves said...

Fine, let's call it a day on that one Chris. Thanks for giving the issue some serious consideration.

Anonymous said...

The correct terms are plesiomorphic and apomorphic. If you use them no one will know what you are talking about.

If rates of evolution were uniform within a group, then all the branches of the tree (proportional to number of differences) would all be the same, wouldn't they. This is sometimes not the case.