Since the time of the Dover trial in 2005, I’ve made a hobby of debating Intelligent Design proponents on the Web, chiefly at the pro-ID website Uncommon Descent. During that time I’ve seen ID proponents make certain mistakes again and again. This is the first of a series of posts in which (as time permits) I’ll point out these common mistakes and the misconceptions that lie behind them.
I encourage IDers to read these posts and, if they disagree, to comment here at TSZ. Unfortunately, dissenters at Uncommon Descent are typically banned or have their comments censored, all for the ‘crime’ of criticizing ID or defending evolution effectively. Most commenters at TSZ, including our blog host Elizabeth Liddle and I, have been banned from UD. Far better to have the discussion here at TSZ where free and open debate is encouraged and comments are not censored.
The first misconception I’ll tackle is a big one: it’s the idea that the evidence for common descent is not a serious threat to ID. As it turns out, ID is not just threatened by the evidence for common descent — it’s literally trillions of times worse than unguided evolution at explaining the evidence. No exaggeration. If you’re skeptical, read on and I’ll explain.
Common Descent and ID
The ‘Big Tent’ of the ID movement shelters two groups. The ‘creationists’ believe that the ‘kinds’ of life were created separately, as the Biblical account suggests, and these folks therefore deny common descent. The ‘common descent IDers’ accept common descent but argue that natural processes, unassisted by intelligence, cannot account for the complexity and diversity of life we see on earth today. They therefore believe that evolution was guided by an Intelligence that either actively intervened at critical moments, or else influenced evolution via information that was ‘front-loaded’ into the genome at an earlier time.
Creationists see common descent as a direct threat. If modern lifeforms descended from a single common ancestor, as evolutionary biologists believe, then creationism is false. Creationists fight back in two ways. Some creationists argue that the evidence for common descent is poor, or that the methods used by evolutionary biologists to reconstruct the tree of life are unreliable. Other creationists concede that the evidence for common descent is solid, but they argue that it can be explained equally well by a hypothesis of common design — the idea that the Creator reused certain design motifs when creating different organisms. Any similarities between created ‘kinds’ are thus explained not by common descent, but by design reuse, or ‘common design’.
The ‘common descent IDers’ do not see common descent as a threat. They accept it, because they see it as being compatible with guided evolution. And while they agree with biologists that unguided evolution can account for small-scale changes in organisms, they deny that it is powerful enough to explain macroevolutionary change, as revealed by the large-scale structure of the tree of life. Thus guided evolution is necessary, in their view. Since common descent IDers accept the reality of common descent, you might be surprised that the evidence for common descent is a problem for them, but it is — and it’s a serious one. Read on for details.
The Problem(s) for ID
I’ve mentioned three groups of IDers so far: 1) creationists who dispute the evidence for common descent; 2) creationists who accept the evidence for common descent, but believe that it can be equally well explained by the hypothesis of common design; and 3) IDers who accept common descent but believe that unguided evolution can’t account for macroevolutionary change. Let’s look at these groups in turn, and at why the the evidence for common descent is a serious problem for each of them.
The creationists who dispute the evidence for common descent face a daunting task, simply because the evidence is so massive and so persuasive. I can do no better than to point readers to Douglas Theobald’s magnificent 29+ Evidences for Macroevolution for a summary of all the distinct lines of evidence that converge in support of the hypothesis of common descent. Because Theobald does such a thorough and convincing job, there’s no need for me to rehash the evidence here. If any IDers wish to challenge the evidence, or the methodologies used to interpret it, I encourage them to leave comments. The good news is that we have Joe Felsenstein as a commenter here at TSZ. Joe literally wrote the book on inferring phylogenies from the data, so if he is willing to respond to objections and questions from IDers, we’re in good shape.
I have yet to encounter a creationist who both understood the evidence and was able to cast serious doubt on common descent. Usually the objections are raised by those who do not fully understand the evidence and the arguments for common descent. For this reason, I emphasize the importance of reading Theobald’s essay. Think of it this way: if you’re a creationist who participates in Internet discussions, the points raised by Theobald are bound to come up in debate. You might as well know your enemy, the better to argue against him or her. And if you’re open-minded, who knows? You might actually find yourself persuaded by the evidence.
The evidence also presents a problem for our second group of creationists, but for a different reason. These are the folks who accept the evidence for common descent, but argue that it supports the hypothesis of common design equally well. In other words, they claim that separate creation by a Creator who reuses designs would produce the same pattern of evidence that we actually see in nature, and that common design is therefore on an equal footing with common descent. This is completely wrong. The options open to a Creator are enormous. Only a minuscule fraction of them give rise to an objective nested hierarchy of the kind that we see in nature. In the face of this fact, the only way for a creationist to argue for common design is to stipulate that the Creator must have chosen one of these scant few possibilities out of the (literally) trillions available. In other words, to make their case, they have to assume that the Creator either chose (or was somehow forced) to make it appear that common descent is true, even though there were trillions of ways to avoid this. Besides being theologically problematic for most creationists (since it implies either deception or impotence on the part of the Creator), this is a completely arbitrary assumption, introduced only to force common design to match the evidence. There’s no independent reason for the assumption. Common descent requires no such arbitrary assumptions. It matches the evidence without them, and is therefore a superior explanation. And because gradual common descent predicts a nested hierarchy of the kind we actually observe in nature, out of the trillions of alternatives available to a ‘common designer’, it is literally not just millions, or billions, but trillions of times better at explaining the evidence.
What about our third subset of IDers — those who accept the truth of common descent but believe that intelligent guidance is necessary to explain macroevolution? The evidence is a problem for them, too, despite the fact that they accept common descent. The following asymmetry explains why: the discovery of an objective nested hierarchy implies common descent, but the converse is not true; common descent does not imply that we will be able to discover an objective nested hierarchy. There are many choices available to a Designer who guides evolution. Only a tiny fraction of them lead to a inferable, objective nested hierarchy. The Designer would have to restrict himself to gradual changes and predominantly vertical inheritance of features in order to leave behind evidence of the kind we see.
In other words, our ‘common descent IDers’ face a dilemma like the one faced by the creationists. They can force guided evolution to match the evidence, but only by making a completely arbitrary assumption about the behavior of the Designer. They must stipulate, for no particular reason, that the Designer restricts himself to a tiny subset of the available options, and that this subset just happens to be the subset that creates a recoverable, objective, nested hierarchy of the kind that is predicted by unguided evolution. Unguided evolution doesn’t require any such arbitrary assumptions. It matches the evidence without them, and is therefore a superior explanation. And because unguided evolution predicts a nested hierarchy of the kind we actually observe in nature, out of the trillions of alternatives available to a Designer who guides evolution, it is literally trillions of times better than ID at explaining the evidence.
One final point. Most IDers concede that if the evidence supports unguided evolution, then there is no scientific reason to invoke a Creator or Designer. (It’s Occam’s Razor — why posit a superfluous Creator/Designer if the evidence can be explained without one?) It is therefore not enough for ID to succeed at explaining the evidence (which it fails to do, for the reasons given above); it’s also essential for unguided evolution to fail at explaining the evidence.
This is a big problem for IDers. They concede that unguided evolution can bring about microevolutionary changes, but they claim that it cannot be responsible for macroevolutionary changes. Yet they give no plausible reasons why microevolutionary changes, accumulating over a long period of time, should fail to produce macroevolutionary changes. All they can assert is that somehow there is a barrier that prevents microevolution from accumulating and turning into macroevolution.
Having invented a barrier, they must invent a Designer to surmount it. And having invented a Designer, they must arbitrarily constrain his behavior (as explained above) to match the data. Three wild, unsupported assumptions: 1) that a barrier exists; 2) that a Designer exists; and 3) that the Designer always acts in ways that mimic evolution. (We often hear that evolution is a designer mimic, so it’s amusing to ponder a Designer who is an evolution mimic.) Unguided evolution requires no such wild assumptions in order to explain the data. Since it doesn’t require these arbitrary assumptions, it is superior to ID as an explanation.
Here’s an analogy that may help. Imagine you live during the time of Newton. You hear that he’s got this crazy idea that gravity, the force that makes things fall on earth, is also responsible for the orbits of the moon around the earth and of the earth and the other planets around the sun. You scoff, because you’re convinced that there is an invisible, undetected barrier around the earth, outside of which gravity cannot operate. Because of this barrier, you are convinced of the need for angels to explain why the moon and the planets follow the paths they do. If they weren’t pushed by angels, they would go in straight lines. And because the moon and planets follow the paths they do, which are the same paths predicted by Newton on the basis of gravity, you assume that the angels always choose those paths, even though there are trillions of other paths available to them.
Instead of extrapolating from earthly gravity to cosmic gravity, you assume there is a mysterious barrier. Because of the barrier, you invent angels. And once you invent angels, you have to restrict their behavior so that planetary paths match what would have been produced by gravity. Your angels end up being gravity mimics. Laughable, isn’t it?
Yet the ‘logic’ of ID is exactly the same. Instead of extrapolating from microevolution to macroevolution, IDers assume that there is a mysterious barrier that prevents unguided macroevolution from happening. Then they invent a Designer to leap across the barrier. Then they restrict the Designer’s behavior to match the evidence, which just happens to be what we would expect to see if unguided macroevolution were operating. The Designer ends up being an unguided evolution mimic.
The problem is stark. ID is trillions of times worse than unguided evolution at explaining the evidence, and the only way to achieve parity is to tack wild and unsupported assumptions onto it.
If you are still an IDer after reading, understanding, and digesting all of this, then it is safe to say that you are an IDer despite the evidence, not because of it. Your position is a matter of faith and is therefore a religious stance, not a scientific one.
We’re okay using it in its broad sense. The way we understand the claim is that the evidence for common descent largely precludes design, that there is something about the nested hierarchy that renders any god-hypotheses superfluous.
We don’t see that, and neither did Darwin. Darwin knew that without a natural mechanism to explain adaptation, he couldn’t explain the evidence without positing some mysterious force, an élan évolutif. Common descent is interesting, but from whence the drive to adapt and to form complex integrated wholes?
Sure. That’s what happens when you throw out most of the evidence. A lot of things are possible.
(By the way, it might be possible to look very closely at the pattern of descent and find statistical evidence supporting unguided evolution, but the claim seems to be that the very existence of the nested hierarchy precludes guided evolution. In many ways, natural selection does guide the pattern of descent, and that’s why it can be difficult to distinguish natural guidance from artifice.)
Surely when keiths refers to “unguided evolution” (non-telic) and compares it to intelligently guided evolution as a likely producer of an objective hierarchy, he’s assuming known processes like natural selection and drift as part of the package.
On perverse designers, why wouldn’t the guiders be included in the description? Why restrict oneself to guiding when new organisms could be created from scratch? Why choose the path of common descent with a recognisable objective nested hierarchy when so many other options are available? Why bother guiding things towards near inevitable periodic extinction events? Or are those events part of the guidance which is meant to appear like non-guidance? If an omphalistic god is perverse, why aren’t other designers who seem to actively attempt to create illusions?
If the guiders are guiding towards complex function, why guide both sides of what we call the evolutionary arms race?
Add in all the evidence we’ve got, and I think you’ll find that the hypotheses you suggested remain logically possible. I was called away in the middle of making a brief post above, so we crossed. I suggested that keiths was assuming natural selection and any other known natural processes when he refers to “unguided evolution”.
Don’t think so. From above,
–
All reasonable questions, but ones we can’t answer without additional evidence. We can posit all sorts of causes.
Natural selection explains the unexplained élan évolutif. While compatible in the strictest sense, design becomes a superfluous entity for scientific purposes.
(Actually, it may be possible to look very closely at the pattern of descent and find statistical evidence supporting unguided evolution, but the claim seems to be that the very existence of the nested hierarchy precludes guided evolution. In many ways, natural selection does guide the pattern of descent, and that’s why it can be a difficult question in isolation.)
Weird… Your comment got flagged by the spam filter, but I have no idea why.
But the nested hierarchy is not of those things. The nested hierarchy is a natural consequence of bifurcating descent with modification.
keiths:
Zachriel:
Well, ID proponents tell us that we can’t make any assumptions about the Designer, so according to them, perverse designers are not ruled out. (Of course, ID proponents violate their own rule about not making assumptions when they assume that the Designer would not leave lots of ‘junk DNA’ in the genome.) Anyway, I think the designer of the giraffe’s recurrent laryngeal nerve would certainly qualify as ‘perverse’.
That aside, the more important point is that even non-perverse designers, including those who guide evolution via common descent, can act in a way that does not allow us to infer a single objective nested hierarchy.
For example, consider a designer who is guiding evolution. Let’s say the unfolding tree currently has fifty ‘twigs’. The designer decides to introduce a complex new trait into half of the twigs, scattered randomly across the tree. If you were trying to infer a nested hierarchy from the evidence of the twigs, you wouldn’t be able to make sense of the data. That’s because the method you use to infer the hierarchy assumes that complex traits don’t magically appear in different twigs at the same time. It tries to find a tree where the complex trait arises once and then gets passed to the descendants. Now, if this kind of thing is happening with many different traits at many different levels of the tree, it will be impossible to reconstruct a single objective nested hierarchy from the data. Instead, you’ll get a bunch of wildly incongruent hierarchies that are subjective because they depend on the order in which you consider the various traits.
This is crucial, so I’ll emphasize it yet again: An objective nested hierarchy implies common descent, but common descent does not necessarily lead to an objective nested hierarchy. Common descent via design can lead to many incongruent subjective nested hierarchies, because design is not limited to gradual changes and predominantly vertical inheritance. Unguided evolution is generally limited to gradual changes and predominantly vertical inheritance, so it predicts a single objective nested hierarchy.
Unguided evolution successfully predicts, out of trillions of possibilities, that we will find an objective nested hierarchy. ID, including guided evolution via common descent, makes no such prediction — unless you make the arbitrary and unjustified assumption that the designer acts in a way that mimics unguided evolution.
Unguided evolution absolutely crushes ID as a theory, and that’s when we limit our attention to the evidence of the objective nested hierarchy. If we include the rest of the evidence, it becomes even more lopsided.
Mung,
I see that you’re making the same mistake about my weed patch analogy that you made when we were talking about streambeds.
I haven’t claimed that designed weed patches are impossible. What I have claimed is that when we come across a weed patch, we should infer that it was created by natural causes unless we have reasons to think otherwise.
Kairosfocus agrees:
I also didn’t claim that designed streambeds were impossible. What I did claim is that when we come across a realistic streambed, unguided natural causes are a better explanation than design unless we have strong reasons to think that the streambed is artificial.
Similarly, a designer who mimics unguided evolution is not impossible. But why assume 1) that there is a designer, and 2) that the designer mimics unguided evolution and produces an objective nested hierarchy, when we have no justification for either of these assumptions? And why make crazy assumptions when unguided evolution fits the evidence trillions of times better without the need for unjustified assumptions?
It’s called convergence.
In any case, it’s easy to imagine a designer that is not compatible with the nested hierarchy, which you have done; but your claim is that no designer (Omphalos excepted) is compatible with a nested hierarchy.
Zachriel,
Not if exactly the same complex trait arises simultaneously in 25 different lineages, and especially not if the trait is underwritten by long genes that are identical, nucleotide by nucleotide, in all the lineages. A designer guiding evolution could achieve this, but unguided evolution could not. If it happened in unguided evolution, it wouldn’t be called convergence — it would be called a miracle.
Not only that. The vast majority of design schemes are not compatible with an objective nested hierarchy, and only a tiny sliver are. The question is, why assume that the designer mimics unguided evolution by picking a design scheme from that tiny sliver? What is the independent justification for such a claim? Without that independent justification, any rational person would choose unguided evolution as the better explanation over ID. It explains more without requiring ad hoc, unjustified assumptions.
It’s analogous to my weed patch example. You can force-fit a design hypothesis if you assume that 1) there is a weed patch designer who 2) just happened to choose a design scheme that made the patch appear natural, but what is the basis for those assumptions? The natural explanation is far better, and any rational person would prefer it over the ‘designed weed patch’ hypothesis.
Most forms of Omphalos don’t qualify. Only the ones that lead to an objective nested hierarchy fit the bill.
Only a tiny subset of designers are compatible with the objective nested hierarchy. Why assume that the designer falls into this one-out-of-trillions subset? Unless and until someone comes up with an independent justification for the assumption that the designer chose a scheme that mimics unguided evolution, then unguided evolution is the better theory, hands down.
Or maybe just flight in insects, birds, reptiles and mammals, and plants. Showing that some designers are incompatible doesn’t mean that all designers are incompatible.
The vast majority of theories of gravity are not compatible with the evidence, and only a tiny sliver are.
Your question assumes your conclusion. Why do organisms mimic machines?
“Unguided evolution” is not an explanation. You have not explained the history of vastly increasing complexity. You may suspect a natural cause, but suspicions are not evidence. Organisms are highly integrated machines, so it is not unreasonable to conjecture an intelligent cause.
Sorry, just don’t see it. Without an explanation of adaptation, the nested hierarchy merely shows common descent. It’s a tree, and you can’t tell whether it is a tree growing in a field, or one that has been pruned and tied in a Japanese garden.
http://etc.usf.edu/clippix/pix/small-manicured-tree-at-the-morikami-japanese-garden_medium.jpg
keiths:
Zachriel:
Those didn’t happen simultaneously or with an identical genetic basis, so they can plausibly be explained as examples of convergence. My hypothetical example is too improbable to be explained by convergence.
Of course it doesn’t. The problem isn’t that every version of the designer hypothesis contradicts the evidence, it’s that too many ad hoc assumptions are required to create a version of the designer hypothesis that fits the evidence.
keiths:
Zachriel:
Yes, and within that sliver are general relativity and the hypothesis that angels are pushing the planets around. They both fit the evidence, but rational people prefer GR hands down. Why? Because it fits the evidence without needing the ad hoc assumptions that the angel hypothesis requires.
You don’t think that Darwinian evolution — random variation with natural selection — explains adaptation? That’s a surprise! Tell me more.
No, the fact that we infer a single objective nested hierarchy implies common descent of a particular kind, with gradual change and primarily vertical inheritance. This is exactly what we expect from unguided evolution, but it’s not what we expect from ID — unless we add additional ad hoc, unjustified assumptions to the designer hypothesis.
‘Compatibility’ of the nested hierarchy does rather depend on the ‘flavour’ of ID. Assuming no outright intent to deceive the Designer’s subsequent Creation, is the Common Descent evidence compatible with:
1) Separate Creation of each species? No.
2) Separate Creation of higher taxonomic ‘type-species’ followed by limited Darwinian evolution? Locally, Yes, globally, No.
3) Creation of a single ancestral replicator and subsequent mutational tinkering? Yes, provided that tinkering is limited in scope – gross tinkering scrambles the signal.
4) Creation of a single ancestral replicator and subsequent ‘intelligent selection’? Yes.
The last is perhaps the most interesting. Really, the Designer is simply adding himself to the many selective agents in operation on populations. Some organisms have more offspring because the designer(s) wants them to. This is indistinguishable from Natural Selection – volition, or any other aspect of causality, is completely absent from the raw fact of offspring production. Indeed, because NS is a statistical tendency, causality would be hard to pin down even with close observation of the entities in action.
But … what is the Designer selecting ‘for’? If he is favouring traits that NS would otherwise demote (if NS would promote them anyway, he would not need to do anything), he is pushing against a ‘force’ . As soon as he stops pushing, RM/NS would come into operation and demote them again. ie, if intelligence guides them to an adaptive peak, it has to keep acting to keep them on it. Unless it is simply pushing them across ‘seas of detriment’ between peaks, knowing that NS is available to keep them there once the traverse has been accomplished. But, of course, there is not simply one population in the world. ‘Pushing’ one population has an inevitable effect on another; a never-ending task.
And the necessity of this process globally is predicated largely on the ‘space of fit organisms’ being perceived as mostly disconnected. If you are ‘aiming for humans’, of course, it does not really matter how connected the overall ‘space of fit organisms’ is – you are trying to get to a particular location in it, and prevent moves in the wrong direction. You’d have to know the way, and you’d probably want some say over the mutations too.
So the Nested hierarchy per se is not incompatible with ID, if the ID performs mutation and selection – it is merely another mutagen, and another selective agent. But because these processes leave little record of cause, there is little prospect of detecting such a designer. For the rest – the majority, I suspect – if they deny Common Descent, then what they may wish to call “ID” is simply Creationism, protests noted.
keiths:
I think I was right somewhere above when I suggested that keiths is assuming known processes like natural selection when he uses the phrase “unguided evolution”. In other words, we should read “unguided evolution according to the modern theory”. Also, importantly, I think keiths is claiming that “unguided evolution” is a far better explanation of the evidence we have for common descent on this planet, rather than just the fact of common descent in any hypothetical biosphere.
Let’s have a look at an I.D. hypothesis of guided evolution including common descent, and see if anyone can make predictions from it.
“The species we see around us today descended from an original single common ancestral species by a process of intelligently guided evolution.”
Unless someone places further constraints on the intelligent designers, I can’t see any reason why we should necessarily be able to see any evidence for common descent at all under that hypothesis. As the designers could vary the genetic code, then vary their variations and so on, we could see hundreds of apparently unrelated codes. And as they could do HGT at a far faster rate than it appears to happen in prokaryotes, and at the same rate in eukaryotes, there’s absolutely no reason an observer would ever perceive anything resembling a tree. Then we can consider keiths example of instant parallel (not just convergent) evolution of one feature in many different species.
But of course, they’ve stayed within or very close to the parameters of unguided evolutionary possibility. They would, wouldn’t they. They remind me of the Cottingley fairies, who suddenly got shy when lots of people came looking for them with cameras. Guided evolution is a shy designers hypothesis.
Can anyone get any interesting predictions from the above hypothesis?
It sounds like PSI research. The more you do the more you reduce the “results” to within the margin of error and the less of an effect you see.
So what appears to be designed will always upon closer inspection reveal itself not to be designed. Even if it was, the “shy designer” will never reveal it’s hand.
By the way, the data example, which in the format for the program Pars in PHYLIP would be:
8 16
1 ,,,,,,,,,,,,,,,,
2 ,,,,,,,,,,,,D,,,
3 ,,,,,,,,,,I,,,,,
4 ,,,,,,,R,,I,,,,,
5 ,,,,,,,,,,,,,,,G
6 ,,,,,O,,,,,,,,,G
7 ,,,B,,,,,,,,,,,G
8 ,,EB,,,,,,,,,,,G
is compatible with the tree you give, but the inference assumes that the comma is the ancestral state in each character.
Otherwise you get an unrooted tree.
And the elusiveness of quantum phenomena, or Douglas Adams’s “SEP effect”, which ID-ers could equally accuse ‘Darwinists’ of suffering from. There is a theological curiosity about it, though – there is an assumed reward for faith. Being too obvious about your existence would run counter to that. It appears only selectively obvious. Everything and anything is evidence of it to some; nothing to others, with degrees in between.
A Zen Buddhist who thinks that life strives towards enlightenment wouldn’t consider her view ad hoc. Indeed, the view predates modern science, but is consistent with common descent. In any case, any reasonable conjecture is going to be modified to fit the data. Without evidence, your conjecture of undirected evolution is just fill in the gaps too.
It doesn’t just fit the evidence, but explains and predicts it. Angels become superfluous. But you can’t explain adaptation from the nested hierarchy. You have to posit some unobserved élan évolutif.
That’s a different question. You said Intelligent Design was incompatible with common descent. When you consider the complete theory, including natural selection, then a designer becomes superfluous to explaining the main contours of evolutionary history.
http://www.youtube.com/watch?v=4UxBjmuxh2Y
That’s exactly what we see here:
http://etc.usf.edu/clippix/pix/small-manicured-tree-at-the-morikami-japanese-garden_medium.jpg
That’s right.
Perhaps difficult, but not impossible. We can observe natural selection, and the results of artificial selection. We have theories of how organisms would adapt historically. While a designer could be hidden within this history, at some point, that designer, too, absent positive evidence, becomes superfluous.
(You seem to have touched on this in the remainder of your comment.)
With the right theology (the intelligent guider wants us to believe through faith alone) the absence of positive evidence can become positive evidence.
However, our friends in the I.D. camp are definitely from the evidentialist school of theology, and whatever one might say about them, they’re not putting forward the idea of teaching fidiesm in the science classes. That would be an uncontroversial course popular with lazy kids, as it would inevitably stop before it starts.
Allan Miller:
Agreed. Those are ruled out by the evidence.
Yes, but a “no gross tinkering” rule severely constrains the designer. Human designers violate it all the time. They are not compelled, either by desire or by limitation, to restrict themselves to the sorts of changes that are produced by unguided evolution. If human designers can violate the “no gross tinkering” rule, why can’t (or won’t) the big-D Designer? Unless ID proponents supply an independent justification of this ad hoc assumption, then this version of their theory cannot complete with unguided evolution.
I would say no. A Designer employing ‘intelligent selection’ — artificially preserving some mutants and killing off others — could do the following, for example:
This is not something that unguided evolution could do, but ‘intelligent selection’ certainly could. Intelligent selection could even do this on an entire group of genes.
So #4 is also not compatible with the evidence for common descent unless we make further assumptions about the Designer.
To summarize, #1 and #2 are ruled out by the evidence, and #3 and #4 require ad hoc assumptions, rendering them inferior to unguided evolution as an explanation of the evidence for common descent.
Allan,
I’ve heard Christians express it this way: If God made himself too obvious, then we would be compelled to believe in him, which would deprive us of our free will.
There are many problems with this argument, but the most obvious is that it underestimates the ability of humans to ignore compelling evidence and continue believing what they want to believe (c.f. ID proponents).
dr who:
Yes, absolutely. I’m certainly not defending some vague pre-Darwinian idea of unguided evolution. I’m comparing modern evolutionary theory to ID as an explanation of the evidence for common descent.
I chose the phrase ‘unguided evolution’ rather than ‘evolution’ for this discussion simply to avoid Joe-like ‘ID is not anti-evolution’ objections.
Zachriel:
The view that life strives towards enlightenment doesn’t predict an objective nested hierarchy. To make it compatible with the evidence for common descent, we must add the ad hoc assumption that life can only strive for enlightenment via gradual change and primarily vertical inheritance. Unguided evolution is a better theory.
Sure, but the fewer ad hoc assumptions, the better. That’s why we prefer modern meteorology to the Rain Fairy hypothesis.
keiths:
Zachriel:
We’re discussing the evidence for common descent, not the growth patterns of actual botanical trees.
That seems to put you and Zachriel on the same page.
We’ll see. 🙂
Again, “unguided evolution” is too vague to be an explanation. Artificial selection leaves a nested hierarchy too. Natural selection is not part of your original post.
Some of the time you seemingly argue the conclusion evolution is ‘unguided’ is intrinsic to common descent alone. Feel free to clarify your position. We may be arguing past each other.
Common Descent is essential to evolutionary theory, but not complete in-and-of-itself. It leaves a significant explanatory gap. The Theory of Natural Selection is a robust theory. We’ve said this several times, even pointing to Darwin’s own belief that without an explanation of adaptation, he didn’t have a workable theory.
It isn’t being proposed to explain the nested hierarchy, but to explain complex adaptation. Without a mechanism of adaptation, organisms would meander about the fitness floor. Without adaptation, you may have a tree, but not the one we observe. There is something missing, indeed, something so important that saying evolution is ‘unguided’ itself seems misleading.
You said, “the fact that we infer a single objective nested hierarchy implies common descent of a particular kind, with gradual change and primarily vertical inheritance.” Just like a tree. If you were a leaf, you would observe the tree. You could suppose unguided growth, or could conjecture guided growth. The nested hierarchy alone is not sufficient to make that determination.
http://etc.usf.edu/clippix/pix/small-manicured-tree-at-the-morikami-japanese-garden_medium.jpg
Allan:
Where would the bottle-neck be? At around 106 cells? It would certainly be that high.
As to the mutation rate, it is estimated that humans produce about 100-200 mutations per replication? This corresponds to a mutation rate of around 10-7. I can’t see where I would be that far off, maybe an order of magnitude. This would mean 400 million years instead of 4 billion years for something similar to develop.
This is exactly why I have no faith whatsoever in Darwinism (neo-Darwinism).
I can no longer keep track of anything I’ve written, or what people have left for me. So, I just give up. I can’t spend all kinds of time just hunting down the latest response.
Sorry.
True, but it is a necessary inference given the retention of a signal of common descent. ‘Gross tinkering’ is the equivalent of rampant mutation or LGT from an unidentifiable source. Though, strictly speaking, there are such discontinuities, and I suppose an IDist could hope that at least one of them is Designer-caused. I guess it only takes one. On the evidence, the Designer is constrained by Limited Gross Tinkering (LGT!), rather than none.
I don’t see how that could operate in practice, with an absentee Designer. Mutagens can’t distinguish the ‘repressor’ from the gene itself. They are as likely to disable it as strip away its silencing, since it would be subject to the neutral mutation rate until it became selectively advantageous. All theories requiring genes to be ‘kept for a rainy day’, front-loading etc, suffer from this drawback. The Designer could keep them in a lead-lined cage, allow mutagens privileged access to certain parts of the genome but not others, or continue acting to kill intragene mutants … Without an effect on offspring numbers, this would not be ‘intelligent selection’, but ‘intelligent mutation’. If the Designer is capable of building such a complex, it doesn’t really need to rely on Intelligent Selection of random mutations to expose the thing!
I’d agree. I also left out #5 Common Design which, as far as I am concerned, is about as ad hoc as they come. Similarity in every last nook and cranny, even in dispensible sequence, is due to a Design need to make them in exactly the same way, in exactly the same order on the chromosome, looking exactly as if they had been copied from a common sequence? Except, of course, where they need to be different, and that’s Design too. The more they need to be different, the more different they are – brilliant!
PaV:
That problem is easily solved if you simply ask people to post their replies to you at the end of the thread, rather than on sub-threads. And if you do the same, of course.
Then each time you visit you can just find your last comment and read forward from there.
Pav,
You do some back-of-a-fag-packet calculations based on a bacterial monoculture. You can’t see where you might be far off. You think that you have thereby spotted something about human evolution that many thousands of PhD’s, Professors and Nobel laureates have somehow missed? As I said, and you ignored, you have omitted the very important contribution to rate-of-evolution from recombination. And as I didn’t say, there are numerous other rate-affecting distinctions you might like to consider. What exactly is it that you think we’d wait 400 million years for humans to develop – citrate metabolism? What about the Lenski apparatus as a model justifies your sweeping extrapolation to human evolution?
It does not surprise me that you have no ‘faith’ in Darwinism, if that is the way you reason.
Yet your view of the Tree of Life requires Darwinism, or something very like it. You hypothesise taxonomic ‘type-species’, then a bunch of natural mechanisms generating variations on that theme. This takes place in substantially less than 400 million years. So what, other than ‘Darwinism’, do you propose for all that variation? According to your tagline, every day is a Bad Day for it! Darwinian evolution, you have just calculated, is impossible in long-generation species. Yet … given you on the one side, and institutions stuffed with competitive Profs on the other, each more than happy to bash the other if he says something silly … they aren’t infallible, but if you were an objective observer and a betting man, who would you back to have missed something?
Zachriel:
I thought it was obvious that by ‘unguided evolution’ I meant evolution as understood in modern evolutionary theory, particularly since we’ve been talking about genetics and mutations. Also, I don’t see how you could conclude that I was talking about something other than MET given statements like the following (from the OP):
As I affirmed to dr who:
Zachriel:
My argument hinges on the distinction between ‘subjective’ and ‘objective’ nested hierarchies. Because this distinction is so important, let me quote Theobald at length:
My contention is that unguided evolution, proceeding via gradual change and predominantly vertical inheritance, predicts an objective nested hierarchy, but that ID (including common descent via guided evolution) does not — unless we add arbitrary, unjustified, ad hoc assumptions to the ID hypothesis. Unguided evolution, as understood by modern evolutionary theory, fits the evidence of common descent far better — literally trillions of times better — than ID does.
keiths:
Allan:
Necessary only if you’re trying to force-fit a designer hypothesis to the data. Why bother, when unguided evolution does the job without the need for force-fitting?
Yes, which is why I’ve stressed that an objective nested hierarchy is recoverable only in cases where change is gradual enough and inheritance is primarily vertical.
keiths, describing a task that guided evolution could accomplish but unguided evolution could not:
Allan:
We were talking about intelligent selection, not an absentee Designer. You wrote:
My scenario is quite possible if the Designer employs the kind of intelligent selection you described in that quote. Unguided evolution could never pull it off.
keiths:
Allan:
Yes, separate creation with common design is the ad hockiest of them all.
It’s amazing that these folks can suggest it with a straight face. They’re willing to make their Designer (and themselves) look ridiculous in order to salvage their belief in the literal truth of the Bible. It’s perilously close to bibliolatry.
Mmmm … now we’re talking ad hoc! I was thinking more of an ‘intelligent selector’ that could only see phenotype, not one that could view genotype directly to make it a selectable phenotype in itself.
So … I, the Designer, have the capacity to generate a two-part gene: a control region and the gene itself. But I since seem to have lost my capacity to manipulate genomes directly, and am powerless to prevent mutations … but wait, I can use those mutations to my advantage! I can still sequence genomes, and use that sequence as selectable phenotype to preserve only those mutations that nibble away at the control region … ! Good job I have excellent foresight, and know my limitations. Is there an Upper Probability Bound on the amount of ad hoc a theory can contain?
The Rain Fairy says no.
PaV, don’t blame you. Nested comments don’t work well on blogs. It’s too hard to keep track of all the sub-threads. Except for asides, we just start a new comment at the bottom of the page.
Okay, but then you say this:
That statement seems to read, due to the emphasis on “gradual change and predominantly vertical inheritance”, that the evidence for common descent alone is sufficient to render design superfluous Only by reading elsewhere do we see that you mean modern evolutionary theory in toto.
We are largely in agreement, then. Modern evolutionary theory renders a designer extraneous to explaining the broad contours of evolutionary history.
The other question was interesting, though. The answer to that might be found in the statistical pattern of the nested hierarchy, but it would hardly be trivial.
Zachriel,
Just to be absolutely clear, here’s what I’m claiming:
1) If you place modern evolution theory (aka ‘unguided evolution’) side by side with ID (including the forms of ID that accept common descent), and
2) ask which theory best fits the evidence for common descent, then
3) you’ll find that unguided evolution fits the evidence for common descent far better — literally trillions of times better — than ID does.
My point in starting this thread was to address a myth. According to the myth, the evidence for common descent is a problem only for creationists. IDers who accept common descent via intelligently guided evolution are off the hook.
It’s just not true. Unguided evolution fits the CD evidence overwhelmingly better than ID does, even when we consider the hypothesis of intelligently guided evolution.
Unguided evolution is the better theory, and the race is not even close.
I haven’t said much because I agree pretty much with everything that’s been said. I also believe it’s a good thing to hash out miscommunications when people basically agree on everything except terminology.
My own take has, for several years, been that design is impossible. I don’t know if it’s intrinsically impossible, like faster than light travel, or just impossible via any foreseeable technology. I noticed someone said we had hit a brick wall with biological design. At any rate, emergence is a tough nut to crack, and complexity is as tough or tougher. it looks to me like you would need to be omniscient.
I also believe that gpuccio’s intelligent selection is a dead end. you would still have to be omniscient to know what to select for. The Lenski experiment demonstrates that critical mutations that enable new function sometimes have no discernible effects.
I don’t think design per se is impossible, but multiply-interactive design, where every form that emerges is just what the doctor ordered, is completely fanciful. It’s the same difficulty that one would face if one decided to go into the weather-control business. Which butterfly’s wing do you need to flap just so in order to generate a hurricane a couple of months hence? Well … all of them. And everything else. Yet if you don’t act everywhere at once, will air movement just stop? Or won’t there still be weather? Just not the weather you want.
You can’t turn off the ‘natural’ process that occurs even when there is no selection. Stochasticity will promote a random genome to ultimate common ancestry, if there is no designer or other Selector around favouring a particular one. Or so the mathematics of sampling would argue. Having a sample which is not a distorted representation of the broader population is statistically much less likely than having one that is, and the distortion is cumulative.
So if you have a long-term goal, you have to eliminate chaos. Which means that nothing, but nothing, happens unless you will it. It is not enough to keep flapping that butterfly’s wing and hope that there isn’t something going on elsewhere that will screw your plans. You have to flap them all. You want a particular recombinant gene, you have to ensure that two individuals pass through 20 years of their life on a romantic collision course. You have to ensure that the fused gametes have the right sample of the diploid gene set. Then, a few years later on still, you have to ensure that meiosis joins the required pairing, which avoids the polar body fate in females or the massive genetic surplus in both sexes, survives, fuses, survives some more … imagine trying to organise things such that this process progresses in one particular direction for 3-4 billion years, against the buffeting of random chance throughout. Or trying to control it all, and eliminate the ‘random’. Wouldn’t you then keep interfering with your own plans? We might imagine omnipotence/omniscience as a solution, but it’s hard to envisage an interactive Creation being able to throw up absolutely any composite result the Creator might wish – especially if you stuck ‘free will’ in the mix! Most designers are forced to make compromises, deal with circumstances beyond their control, and witness unintended consequences.
You are focusing on implementation, which is certainly a major hurdle, but I think design itself is a hard problem, not solvable by computation. I’ll believe otherwise when I see a proof of concept experiment that doesn’t involve selection.
I don’t think you can consider design without considering the implementation, other than in the most ‘magical’ form of Creationism (even then, I find myself wondering how it appears mechanistically – do zebras pop into existence audibly, or do they get ‘drawn’ from the hooves up, or begin as non-gamete zygotes in a virtual womb … don’t these details bother people? ;))
The implementation issue, and chaos, become important if one is looking for an anticipatory Designer whose ultimate goal is humans, but one isn’t a YEC. The Designer invents some fundamental particles which form a range of atoms which form a range of molecules by ‘natural’ interaction. Among the many molecular configurations is a self-replicator, and you somehow know that self-replication is the vital capacity that allows, via selection, the generation of complex phenotypes. Now what? “One-step” design would be a hard problem, true. But this method of exponential, interactive chaos is mega-hard, as a means of achieving a pre-specified outcome rather than the generality of ‘adapted descendants’.
Not sure where you get “literally trillions”. That sounds like the odds of universal monophyly.
In any case, absent positive evidence, the designer has to be put into the gaps. Darwin theory is sufficiently robust that he fairly well closed the door on any direct role for a designer. It’s still possible that a designer could manipulate the trajectory of evolution, but there is no need for that hypothesis to explain the evidence we have. That leaves unevidenced Providence.
Of course, there’s a gap at the origin of life, which is still sufficient to allow for a designer, though there is significant evidence that this also has a natural cause.
I don’t think there’s any shortage of weaknesses in the ID conjecture.
Enough for you and for me. The only version that make sense is poof. Perhaps with a thunderclap, depending on the size of the object poofed into existence.
GP and some theistic evolutionists seem to think that the designer has some non-material consciousness that interacts with matter at the quantum level, so genomes could be altered at or immediately prior to conception.
But GP is a thoroughgoing dualist who believes our minds are non-material and interact directly with our brains.
Zachriel:
‘Trillions’ is actually conservative. The real number is many orders of magnitude higher, but that seemed like overkill, so I left it at trillions.
It comes from the fact that multiple lines of evidence establish the consensus phylogenetic tree shown in Theobald’s Figure 1 to an accuracy of 1 in 1038. Under a design hypothesis we have absolutely no reason to expect this spectacular degree of convergence, unless (you guessed it) we make the ad hoc and unwarranted assumption that the Designer just happens to operate in a manner that matches unguided evolution.
See Consilience of Independent Phylogenies in Theobald for an explanation of the 1038 figure.
That’s the odds of monophyly, and Theobald’s results apply regardless of natural selection or artificial selection or no selection.
Zachriel:
No, it’s the odds of getting that particular nested hierarchy by chance using a method that generates a best-fit tree assuming monophyly. The fact that multiple lines of evidence converge on that hierarchy is a stunning confirmation of the theory of evolution.
You may be thinking of a different argument of Theobald’s in which he concludes:
Zachriel:
Well, the ‘no selection’ hypothesis is clearly falsified, because a) we see selection happening and b) we don’t see the unimaginably baroque variety of life we’d expect if every mutant were equally fit.
Artificial selection matches the evidence only if you make unwarranted assumptions about the desires and/or limitations of the Designer. Otherwise, artificial selection is compatible with scenarios such the “repressor shadow” example I sketched out earlier.
Unguided evolution fits the evidence trillions of times better. No need for ad hoc assumptions.