Enzymes stitch together non-natural DNA [Getting closer to lab-made life]

Nature Magazine ^ | 24 February 2004 | PHILIP BALL

[VadeRetro]

When your species is superbly adaptive...

"...adapted" was intended.

[PatrickHenry]

Sniveling about "just so" stories is a signature of those blockhead creationists you don't want to be confused with.

Typical example: Creationist claims "There's no way X could have evolved!" In response, the evolutionist presents a possible scenario (as in the evolution of our eyes). "Aha!" exclaims the creationist, "you've giving me a just so story!" Then follows the customary outpouring of lame objections: you have no witnesses, you haven't done it in the lab, Piltdown Man, etc.

Lost in the noise of the creationist's response is the fact that his original claim ("There's no way X could have evolved!") has been refuted. There is a way. Maybe not the way, or the only way, but the point remains that because there is a way, the creationist claim has been demolished.

[The_Reader_David]

The phrase 'just so' stories was also used by Steven J. Gould about the same phenomena: he has a wonderful story about how important the spandrels around the dome of St. Mark's Cathedral are to illustrate his point.

I'll give you a better reply later, but I have to run just now.

[The_Reader_David]

Of course, you will note I have never made the claim that something couldn't have evolved. My argument is that the theory is unfalsifiable because it can provide an account of how any conceivable observed biological variation evolved.

[VadeRetro]

The phrase 'just so' stories was also used by Steven J. Gould ...

Still doing a wonderful imitation of one of those people you don't want to be mistaken for.

[PatrickHenry]

My argument is that the theory is unfalsifiable because it can provide an account of how any conceivable observed biological variation evolved.

Your argument isn't very persuasive. One can very easly conceive of animals that don't fit the observed patterns that lie within the theory. For example:

According to the theory of evolution, that creature couldn't have evolved on Earth. The feathered wings, the mammalian body, and the six-limbed structure are from such widly divergent lines of descent that there's no way they could all be combined in one animal.

Very conveniently for the theory, however, every species that actually does get observed actually fits within the theory. This is not, as you imagine, a weakness of the theory, which somehow renders it unfalsifiable. Rather, it demonstrates that the theory has not been falsified -- although a Pegasus fossil would certainly be a falsification. The theory places severe limits on what is possible, and as a result it accurately predicts the kinds of animals that could exist. It's not surprising, therefore, that existing species are explainable in accordance with the theory.

[The_Reader_David]

Let me be very clear. I have no doubt that a properly scientific theory in the Popperian sense to account for biological diversity can eventually be found. Unfortunately, I believe the polemical defense of Darwinism (of which your sneering use of 'lawyer' is an example) is an impediment to the process. I would expect a good falsifiable 'theory of fitness' which takes into account both ecological and organismic characteristics to replace the tautological natural selection as a central component of this theory.

It is quite possible that the theory will subsist nicely as an explanatory theory with variation included as an unexamined stochastic process (Black-Scholes does well enough for futures markets). Even so, I would like to see the hypothesis that genetic variation is random subjected to tests.

My objection to the current state of evolutionary biology, particularly as it is represented polemically by folks like you is threefold:

First, 'natural selection' is tautological to a degree that even the definitions of classical physics are not.

Second, and I had hoped this wasn't true, but the fact you, who seem quite well-versed, keep giving me guff and bravado when I ask, the hypothesis that genetic variation is random appears to be untested.

And third, that the sociological pressure of defending science against religious obscurantism has caused the Darwinism to become a closed, unfalsifiable system of thought, more akin to an ideology than a scientific theory. Your responses to my posts confirm me in this view: if you can blithly accept variation as a response to the environment as an 'adaptation', and brush aside the possiblity that reverse-transciptase could introduce Lamarkian elements into the process as not falsifying the system, then you can go on spinning 'gaseous Vulcans' to account for the precession of Mercury forever, or until your subject gets an Einstein.

[VadeRetro]

First, 'natural selection' is tautological to a degree that even the definitions of classical physics are not.

You have repeated this (several times now) after I have formulated why I do not accept it. In an intelligent dialogue, when I object to something you would respond to my objection.

Environmental pressures change. "Natural selection" is not different in any important ways from the human-driven selection that has produced so many varieties of our domestic animals. You can make an apparent tautology there, too. "Humans allow to breed those which they allow to breed." You can do that with almost anything which you want to describe as corrupt.

Newton's F = ma has been attacked as logically rotten because mass appears as a property of things by which they resist acceleration by a force. (m = F/a, that is.) But force is what it takes to move a mass, so mass is in the definition of force.

Newton didn't provide a non-force definition of mass or a non-mass definition of force. (Either or both would have been hard to do.) But he wasn't wrong, at least not for that reason. And not for 200 years. He was making useful predictions no one had made before. That doesn't go away because some religious objector being dragged kicking and screaming through the thing finds a term he can label as a tautology.

While you can "define" natural selection so that it superficially looks tautological (but it's just a shorthand), the term describes pressures upon a population which drive it toward characteristics not initially much in evidence. It is useful to have a generic term for the pressures leading to differential reproductive success. I do not see a logically corrupting influence. You have not justified your endless repetition of the "It's a tautology!" creationist mantra.

Second, and I had hoped this wasn't true, but the fact you, who seem quite well-versed, keep giving me guff and bravado when I ask, the hypothesis that genetic variation is random appears to be untested.

Long ago, you had passed me a Greek Orthodox paper which hinted that you were among the people of faith who could accept findings from science without feeling challenged to contest the results on religious grounds. You apparently still claim such, but have not helped your case by borrowing arguments and tactics from the ICR crowd whose specifics you do not endorse. A bad penny is a bad penny, and a bad argument is a bad argument, no matter from whom you got it. When you keep thumping the same sour notes, I wonder when you're going to come in with "There are no transitional forms! Piltdown Man was a fake! Nebraska Man was a pig's tooth!"

Now to your questions on "random mutation." (Which is not to be confused with "natural selection. Have you decided which one is bothering you most?)

I am simply unaware of studies concerning whether cosmic rays are predisposed to strike in one place versus another. Some areas of the genomes of some species are known to be highly conserved and to have mechanisms to detect and repair mutations. Mutations to the involved areas when they occur tend to be fatal. IOW, such mechanisms are clearly adaptive.

I am unaware of any studies which would purport to check whether God is running around causing the initial mutations (the copy errors, the particle strikes) to be non-random. The output is non-stochastic because 1) there are conserved regions and, 2) natural selection starts with whether a mutated cell remains viable and fertile.

I suppose it is theoretically possible that an adaptation is flat-out unevolvable, but I repeat that negatives are hard to prove. That antievolutionists are always pointing to something, saying it's complicated, and saying "... Therefore it did not evolve" does not help anyone insisting upon falsifications in precisely that area, especially if that someone is showing creationist levels of perseveration in bad argument.

And third, that the sociological pressure of defending science against religious obscurantism has caused the Darwinism to become a closed, unfalsifiable system of thought...

You keep saying it. I've given you counterexamples and linked dozens of others, all of which you have ignored except to complain of their length. Evolution predicted in advance that whales with legs must once have existed, that feathered dinosaurs must once have existed, that creatures with funny double-jointed jaws must once have existed (and that our remote ancestors were among them). No other theory was saying such things and evolution could have been wrong to do so. It wasn't.

[VadeRetro]

I'll risk anticipating a point here. If you weren't going to raise it, sorry!

However, perhaps you're simply more familiar with the arguments from a certain camp and less so with mainstream biology. Mere complexity, even irreducible complexity, presents no hurdle to evolution. I realize that one hears something different from the prestigious Discovery Institute, but those are the facts as real science understands them.

That's why I'm not impressed when people try to falsify evolution by pointing out clearly adaptive features, however complex. That's just the wrong place to look. That falsifications do not spring readily to hand in a specific area does not mean they do not exist. (Especially given that some large number have been pointed out already but ignored because there are too many of them for handy reading.)

[Doctor Stochastic]

My argument is that the theory is unfalsifiable because it can provide an account of how any conceivable observed biological variation evolved.

If Piltdown Man had been an actual rather than designed creature, that would have gone far towards falsifying evolution. So would the discovery of a naturally-occuring three-legged one-wheel giraffe.

Likewise if cladistics were to give a lawn rather than a tree (based on Linnaeus's observations or on modern genetic typing), this would refute evolutionary theory.

On the other hand, nothing can refute Cretionism because Creationism (and it's doppleganger, Intelligent Design) make no testable predictions.

[The_Reader_David]

I think perhaps an analogy to the creation/evolution debate as actually framed in public discourse might clarify my position.

A room, to which we have access only once daily, contains a large table say about 10 meters square. On it are a large number of plastic cubes. From where these are most readily observed, only the tops can be seen. On these are numbers between 1 and 6. The patterns of numbers change from day to day, though mostly remain the same. Some very beautiful patterns (to a mathematician's mind) are observed, both statically and dynamically, other regions of the table seem to have random bunches of numbers, or random progressions, though these are rare and infrequent.

A dispute breaks out as to the cause of these patterns, and two schools of thought arise: the placers and the tablists. The placers hold that some intelligent being accesses the room when we are not looking and places cubes with numbers in the beautiful patterns or successions of patterns, and that the seemingly random parts are not random, but contain patterns we cannot discern. The tablists, on the other hand, hold that the cubes are, in fact, dice, and the something about the table and the nearby dice causes some of them to be re-rolled from time to time, while other properties of the dice and regions of the table tend to favor certain patterns remaining longer, or certain progressions of patterns, by causing dice with certain outcomes not to re-roll in certain locations.

The tablists, quite sensibly, hold that the placers' explanation is non-scientific, since it cannot be falsified: the intelligent being could decide to place the cubes in any pattern or non-pattern. The tablists even manage to set up a camera in the room, and observe some of the cubes changing orientation so a new number shows in a way consistent with them being dice. The tablists loudly proclaim that their view is scientific!

And it is fruitful in terms of the useful explanations it yields: The collection of cubes almost entirely containing 1's in the far north-east of the table arose because conditions there favored many numbers of the same sort remaining, and a roll early in the history of the table happened to come up mostly 1's. The same explanation accounts for the predominance of 5's in the center of the table. Condition at the south-east corner were different, and favored a placements of even numbers nearest to odd numbers, so that an approximate checkerboard pattern of parities is found there. Conditions in the south central table favored dice remaining unrolled when they were to the left of a die with a smaller number, thus many sequences increasing from right to left are found there. Conditions in the small area where the random numbers change often are very unsettled, all numbers seem equally well suited to that portion of the table, and the harsh conditions there cause a very high re-rolling rate. (It is hotter there, and some suspect a thermal effect on the re-rolling rate--others think the system there shows evidence that chaotic dynamics will be needed to explain the changes in the patterns over time--still others that the conditions in nearby regions which favor incompatible patterns coexist there.)

The theory predicts change over time as the condition of the table changes, and the record of observations of the table show this to be the case. Hypotheses abound about an event about 20 years ago when with the exception of a few ordered patterns near the corners of the table, all of the patterns were radically different a two weeks later (records back then were only kept of biweekly observations), but all are quite consistent with the theory.

A critic arrives, and listens to the dispute for a while. He decides the placers really are non-scientific, and many of them are idiots--they seem to attach almost mystical significance to some of the patterns of numbers, and insist these are completely stable over time. But he has gnawing doubts about how scientific the tablists are:

The notion that conditions in a part of the table (including the presence of un-re-rolled dice) 'favor' the pattern seen there seems tautologous. The tablists seem very clever at coming up with a kind of condition to explain any pattern or progression of patterns ever seen anywhere on the table, and he strongly suspects they have become so clever that they can explain any conceivable pattern or lack of pattern in terms of the local conditions on the table and random dice rolls, and thus have made their theory as unfalsifiable as that of their opponents. He also wonders whether the tablists have ever actually tested their hypothesis that the changes in individual dice (he suspects they are right about that aspect of the system on the table) really are random re-rolls.

When he raises his doubts, the tablists all raise hue and cry "Placer!" "Only a placer would object that our theory is unscientific!" "What are your objections?" When he states them again, they address only the part of what he says, which sounds to them most like a placer speech they heard once, and ignore the rest.

The critic points out that he happily admits that tablism is a philosophy which can lead to the framing of properly scientific hypotheses about the table, as for example the competing pair of hypotheses which were once held about the north-east corner--that conditions favored 1's, vs. conditions favoring like dice and a random preponderance of 1's (the former was falsified when it was discovered that conditions in the center where 5's predominate were in fact identical in all measurable respects to those in the north-east). He nonetheless holds that tablism is not itself a scientific theory, but a metaphysical position concerning the table and the cubes on it.

This, nonetheless is not good enough for the tablists, they insist he admit that their position is scientific! That it is falsifiable! That it has survived every falsification test! Anything else proves he's a placer! They point to predictions correctly made by some of the properly scientific theories framed by using the tablist philosophy, linking webpages . They point to observations which could have falsified the most well-developed scientific theory they have framed using the tablist philosophy as proof that the philosophy itself is scientific. They eventually admit that they have never heard of their colleagues applying a test for randomness to the re-rolls.

He insists that what he really wants is a good falsifiable theory of conditions on the table and their dynamical laws, as well as a dynamical theory about the changes in individual dice (though he's settle for some stochastic elements in this), but they still shout "Placer!".

The critic leaves in disgust and goes back to proving theorems about mathematical structures related to spin-foam quantum gravity.

[The_Reader_David]

Incidentally, I won't leave in disgust. I composed that little piece while off-line. I'll give you a decent reply to your two most recent posts sometime this weekend, or Monday at the latest (especial thanks to VR for answering my questions directly, I intend to give a considered response).

[Doctor Stochastic]

I'm not sure I understand your point. Your example has nothing to do with evolutionary theory. It lacks the two main points of evolutionary namely, reproduction and mutation.

[The_Reader_David]

Perhaps you and DoctorStochastic can dispute whether or not a tripedal-unicyclic giraffe would falsify evolutionary theory. DS seems to think the configuration is unevolvable, while you, VR, don't seem to think one can convincingly argue that an adaption is unevolvable. (Now isn't that getting dangerously close to admitting that the theory is unfalsifable?) However, since I'm here staking out what could perhaps be described as a hyper-Popperian position rather than a creationist position, I would be thrilled to have a convincing argument that some hypothetical adaptation is unevolvable. Such an argument would convince me that the theory was falsifiable--there would be a hypothetical observation it could not account for!!!

My suspicion is that in point of fact tripedal-unicyclic would not be considered unevolvable. An account of how the one wheel provided advantages when fleeing predators in hilly terrain, and of how three legs are all that is needed to stand stably while browsing for food in tree-tops would be given. Accounts of how intermediate forms could have conveyed slight survival advantages would be given. The beast would be held up as an example of how evolutionary theory can account for completely novel adaptations.

Now, DS's 'lawn' vs. 'tree' gave me some cause for thought. Certainly natural selection + random variation + common ancestry predicts a tree structure. And a tree structure would seem to be a falsifiable outcome. The trouble is, as I see it, on the one hand common ancestry alone predicts a tree structure (and as I have already pointed out, I regard common ancestry as a properly scientific hypothesis which has survived lots of falsification attempts), and on the other, trees may like beauty be in the eye of the beholder: classification by multiple characteristics tends to produce trees when there are logical dependencies among the characteristics.

This then leads to a question in the philosophy of science: does one really have a falsifiable theory if one begins with a falsifiable theory and appends an unfalsifiable theory to it? I would argue not, or at least one only really has the falsifiable theory one started with.

Consider two theories of biological diversity, both of which feature common descent: common descent + natural selection + random mutation, and the theory that the gods from time to time bless or curse offspring of living beings--with divine benefactions or malefactions being visited upon the being and all its progeny. Both predict a cladistic tree, both predict commonalities in all living organisms and that similiarities will decrease as one gets further from a common ancestor. But the later is plainly obtained by appending an unfalsifable subtheory onto the falsifiable theory of common descent. Even so, its adherents, equipped with a mythology which told something about the gods' likes and dislikes, would be able produce further predictions from the theory, not immediate from common descent, and point to these as proofs of its scientific nature.

Unfortunately, I must go for the day. I have some paperwork I need to get out. I hope to address the rest of your post tomorrow.

Oh, to DS: reread the table fable, but think of configurations of numbers as the organisms rather than the numbers themselves.

[VadeRetro]

You repeat yourself at length, but you simply repeat. I've explained what I think about what and where the falsifications are.

[RightWhale]

It's probably begging the question to draw a line when the domain might be a continuous spectrum of possibilities. We might one day want to consider all matter as living--all the way from hydrogen atoms to galaxies. Some people already do that, but they are hard to understand sometimes.

[Doctor Stochastic]

DS seems to think the configuration is unevolvable,...

No, it isn't the evolvability or unevolvablity of the giraffe that is in question. It is the lack of suitable ancestors.

[VadeRetro]

That was an overly short critique for a long post, so I'll give you some more. You still haven't committed to what sort of alternate theory is being neglected in favor of "random variation and natural selection," or even whether the randomness of variation or the naturalness of selection is the sticking point for you. Are the interventions of the gods in one area, the other, or both? How are we supposed to be detecting the action of these gods? What makes the "curse of the gods" theory scientific?

How does "curse of the gods" theory explain ring species in birds and salamanders?

[The_Reader_David]

Do try to continue a civil discussion and actually read my posts to address their content rather than some preconceived idea of what a critic of Darwinism would say.

You seem to have overlooked the end of my post in which I promised a more thorough answer. It would have been courteous to wait, rather than 'giv[ing]. . .some more'. Still, I will answer your 'more' first:

You miss the point of introducing the "curse of the gods 'theory'": it was not introduced as a serious attempt to explain biological diversity (hence your jibe about ring species is utterly irrelevant) but to clarify a point about the nature of falsifiable theories by illustrating how an unfalsifable theory can be appended onto a falsifiable one, the composite inheriting the falsifiability of the properly scientific subtheory, and then moreover dressed up with additional (readily modifiable) assumptions to 'make more predictions'. The point is that both Darwinism and the "curse of the gods 'theory'" inherit falsifiablity from their common element: the theory of common descent, which I have already told you a regard as a sound scientific theory which has survived many falsification tests.

I am also puzzled why you insist that I must specify one thing which bothers me. I have made plain what my objections to Darwinism are: first that natural selection (or environmental pressure, if you prefer) is not a mechanism at all, but a tautological definition which can be given any non-tautological content a biologist sees fit to fill in to offer an explanation of any observed change; second that the one falsifiable element in the usual description of the theory--random variation--has not, in fact, been subjected to any serious falsification tests (something you seem to admit); and finally that a combination of cleverness in inventing 'selection pressures' to explain any observation together with sociological effects of defending Darwinism against religiously motivated obscurantists (which make the definition of theory itself maleable) has led to a situation in which Popper's original view that Darwinism is a 'metaphysical research programme' which can generate real, falsifiable scientific theories, but is not itself one, is in fact correct.

Now as to the rest of the long post you faulted me for only beginning to critique:

First, I am puzzled why you assume cosmic ray strikes are the only issue. We are both aware that there are mutagenic chemical compounds. We are both aware that reverse-transciptase enzymes exist. You even point to strongly conserved portions of some genomes--an observation which I was hitherto unaware, and which plainly tells against the randomness of variations. Of course, you explain this as an adaptation, just as you explained the reverse 'rolling the dice more often' as an adaptation.

And pushing selection pressure down to the cellular level doesn't really do anything but insist that I push my critique down to the cellular level, which I am glad to do.

Finally, as regards the tautological nature of selection pressure, your analogy with F = ma isn't really a good one. The fact that force, which is also related to energy expenditure via W = Fd (for simplcity I remain in the 1-dimensional picture with forces constant over time), and then linearly to the second time derivative of a body's position is hardly tautologous, and subject to test (one could imagine a dynamics in which while W = Fd, there is a non-linear relation between force and acceleration, or a world where Aristotle's physic correctly predicted motion).

Natural selection (or environmental pressure), however, really comes down to 'survivors survive'. The 'Humans allow to breed those which they allow to breed,' tautology likewise does not contribute to a scientific theory of the observed outcomes of animal husbandry, as without a theory about what humans want in their domesticated animals it provides no testable predictions (beyond those built in by common descent).

Replacing 'environmental pressure' with an actual testable theory of what is well-adapted for each ecological niche and a theory of what ecological niches will exist under what conditions would give a properly scientific theory of natural biological, just as replacing the 'humans allow to breed. . .' tautology with a testable theory of what humans want in their domesticated animals would yield a scientific theory of the outcomes of human animal husbandry.

[The_Reader_David]

make that '. . .natural biological diversity,. . .' rather than '. . .natural biological,. . .' late in my last post.