Tuesday, October 17, 2017

The argument from DNA

Here.  Oh, and can we skip the "Flew didn't write his book" discussion? This is an argument, so focus on that, not the personalities.


Jimmy S. M. said...

the article is titled:

Science gives ample reason to believe in God. Why is DNA important?"

Putting aside Hal's objection, which I agree with, the article also contains this quote:

"Perry Marshall, an information specialist, comments on the implications of this. "There has never existed a computer program that wasn't designed...[whether it is] a code, or a program, or a message given through a language, there is always an intelligent mind behind it.""

100% of intelligent code generating minds that we've observed are dependent on a physical brain. So how does this argument get you to a god? (Mormons excluded)

jdhuey said...

As far as I can see, there is no "argument" presented; only the assertion that "God did it".

I noticed that there was no mention of the 4 billion years of evolution by natural selection that might have had something to do with DNA.

World of Facts said...

All good points in the comments here... I would summarize the argument as one from ignorance: the author simply doesn't see how this could be the result of natural processes.
(How can one explain this sophisticated messaging, coding, residing in our cells?)

Unknown said...

This argument is legit trash. It says all "code" comes from a designer. But let's be more specific. The "designers" referred to in this premise all share the following qualities: they are spatial, temporal, physical, their consciousness is in some way dependent on a physical brain, their actions are mediated by intermediary causes, and their designs are fashioned out of preexisting matter. It is precisely because of these qualities that designers are causally adequate to explain information. Take the example of a computer programmer producing information in the form of code. The explanation is not simply the fact that he formed the intention to write code in his mind; rather, in order to explain the code, the programmer must physically carry out his intention through the use of a complex nervous system which sends signals from the brain to his fingers, which then physically tap on a keyboard, which is physically connected to the computer through a complex electrical system. This chain of events is what allows us to draw a connection between the cause and the effect. His ability to spatially, temporally, and physically interact with the world through the use of his body is what allows him to produce information. If he could not use his body--if he were paralyzed, for instance--then he wouldn’t be able to do this. His mind alone isn’t causally adequate to get the job done.

Yet God is defined as nonspatial, nontemporal, nonphysical, immaterial, capable of thought without a mind, capable of fulfilling intentions without operating through intermediary causes, and fashions his designs out of matter he created ex nihilo. This couldn’t be any more different than the “designers” of ALL code. All the factors that make design a causally adequate explanation are absent in God’s case.

Steve Lovell said...

I don't have the detailed scientific knowledge required to really defend this argument, but it seems not nearly so bad an argument as people are here suggesting. Admittedly, it's not a very persuasive presentation, but I don't think we are it's intended audience.

I think the rough structure of the argument is really:

1) DNA contains large volumes of information.
2) The probabilistic resources of the natural world, conceived in naturalistic terms, are insufficient to explain the origin of this information.
3) Therefore to explain the information we need to go beyond those resources.
4) One such explanation would be a "divine mind".

One might quibble about premise 1, but surely the main debate should be about premise 2. And depending on how one argues for 2, it certainly could be an argument from ignorance or a "god of the gaps" argument, but I don't think that's inevitable.

Kevin said...

I don't find DNA to be very useful in a God debate, either for or against, because the central question isn't "Is there a conceivable natural process that can give rise to DNA", because natural processes exist either way. Rather, the question for me is "Is something like DNA more likely in a created or a just-because universe?". The answer is painfully imbalanced to me in God's favor, but ultimately to answer this question would require a side by side comparison of at least one created universe and one just-because universe, which is impossible. And others can comfortably disagree with my assessment and there isn't really any way for one to refute the other.

Unknown said...


4 is not a conclusion of that argument. 3 follows logically from the previous premises, but 4 is just an added premise in need of justification. The article's justification is based on an analogy to human designers. But as I argue in my previous comment, the factors that make human designers "casually adequate" explanations of code are entirely absent in the case of a divine mind. So the argument equivocates on the word "designer" or "mind"

World of Facts said...

Regarding information in DNA, this id already a non-starter. DNA doesn't contain information in the same way a piece of paper with a long string of letter does.

The letters on the paper could be interchanged, say all 'a' with 'A' And, though it would look weird to us, the meAning would be the sAme. But with DNA, because the molecule drives chemical reaction and protein folding, the physical shape and arrangement are not independant of the medium.

The letters on a piece of paper tell us nothing about the paper itself; so-called letters of thr DNA code tell us nothing but something about the molecule and host themselves.

As to your point Legion, it sounded interesting for a split second, but it's really just pushing the same question back further in time. You're asking how could thr natural process od DNA evolution come to be in a universe that is just so, not designed? Exactly the same argument, and no justification as to why we should even entertain the idea of such designer, and that's where Ron's points are relevant.

Steve Lovell said...


Interpreted as an argument from analogy, your points against the argument are as old (and solid) as the hills. But I don't think it's best understood that way. The examples are mostly given to illustrate and not as part of an inductive base for an argument from analogy.

I completely agree of course that my (4) doesn't follow logically from the rest of the argument. However, it's not obvious what the other alternatives might be if the phenomenon to be explained really does outstrip the probabilistic resources of a naturalistic universe.

In fact, you'll find most versions of (modern) Intelligent Design actually say that an appropriate "natural" mind could also explain the phenomena ... and surely you need to find something more wrong with the argument than that it could only legitimately allow us to infer the existence of alien overlords or a Mormon style god somewhere in deep space?

If not, do you believe in such alien overlords?

World of Facts said...

"it's not obvious what the other alternatives might be if the phenomenon to be explained really does outstrip the probabilistic resources of a naturalistic universe."

Isn't that an example of argument from ignorance?

Unknown said...


Let's say that the probability of DNA assembling naturally is very low, but at what point does P(DNA|naturalsim) become unacceptable as to be "outstripping"? Dembski says it's 1 in 10^150. But let's not get confused about this. Even if P(DNA|naturalism)=1 in 10^150, that doesn't mean naturalism only has a probability of naturalism is only 1 in 10^150. That would be to confuse P(A|B) with P(B|A). Even if P(DNA|naturalism) is absurdly low, this doesn't preclude P(naturalism|DNA) from being very high, because P(design) and P(DNA|design) may be absurdly low as well, or perhaps inscrutable. There's no avoiding Bayes theorem here. Someone may object to using Bayes theorem to test metaphysical hypotheses like naturalism and theism, but then they have no business trying to rule out naturalism in the basis of "outstripped probabilistic resources"

Jimmy S. M. said...

@Legion: god doesn't need anything like DNA, natural selection, a multi-billion year program of predation, death and extinction. The system is monstrously inefficient for a being capable of anything.

Steve Lovell said...


You'll get no argument from me on your points about Bayes. Still, it seems pretty obvious to me that these considerations point in a particular direction. It's not decisive, perhaps (although that will depend on assessment, if possible, of the other probabilities involved), but neither is insignificant. As Victor often comments, to claim as some people do, that there is no evidence for Theism rather than that there is insufficient evidence seems flatly mistaken.


That part of the argument would indeed be an argument from ignorance (at least as I've expressed it). But the first part, 1 though 3, needn't be.


I'm no biochemist, so I don't know enough details to give chapter and verse on proper examples, but hopefully I can get across the concept I'm reaching for ...

Scientists sometimes use the example of an "infinite monkey cage" wherein an infinite number of monkeys tapping randomly at keyboards would eventually produce the complete works of Shakespeare by mere chance. But how many monkeys are really available? And how long would they really need?

Now replace Shakespeare with some basic building block of life, some X, which cannot be explained by evolution (but is assumed by evolution). Replace the monkeys and their typing with all the particles in the known universe, their typing speed by known rates at which it is possible for these particles to combine, re-combine ... and give them the life time of the universe in which to cycle through combinations. At what point would it become even remotely probable for these combinations to create X? Where X is whatever basic structure is required for evolution of any type we'd recognise to take hold.

The naturalist typically multiplies his probabilistic resources to be as large as possible. (The "many worlds" response to design arguments is one way to do this.) The idea in these arguments is that our knowledge about the universe puts real limits on just how big these probabilistic resources can be, and that we can also calculate how likely or unlikely our X
(or something sufficiently like it) might be to occur in such a way. If the probabilities suggest X (or something sufficiently like it) would be highly unlikely to occur in any universe unless that universe was many, many times the age of our own, then that phenomenon outstrips our probabilistic resources.

So what is this X? And how unlikely is it? There you need a real expert, and it isn't me. So from a scientific perspective, I don't know how persuasive the following example is and how complex the case really is or what other mechanisms of explanation are really available, but I find it a helpful "illustration":

The way that DNA works, it requires a piece of "machinery" to read the "code" and use that code as instructions to build something. I'm sure these things have proper names that I've long since forgotten. That piece of machinery is itself pretty complex and is in fact "coded for" in the DNA which it decodes. As such, to a complete amateur like me, it appears that DNA and the machine which decodes it have to arrive on the scene together for either to be of any use. I'm pretty sure there is lots that biologists say about this and how they get over the obvious difficulty that this creates. The important thing is the idea ... something complex which evolution itself cannot explain because evolution assumes it's existence and where we have some hope of calculating how likely it is to appear "by chance".

More could be said, but I don't think I'm the best person to say it. William Dembski edited a collection of papers on Intelligent Design called "Mere Creation", and his chapter in there is good on this.

Kevin said...

Jimmy: "god doesn't need anything like DNA, natural selection, a multi-billion year program of predation, death and extinction."

1. Neither you nor I know what God "needs" without first addressing what God's purpose in creation would be in the first place.

2. Worst case, it's still much more likely than a just-because universe based on everything I'm aware of. I've never heard a convincing argument to the contrary anyway. Again, impossible to prove either way.

Jimmy: "The system is monstrously inefficient for a being capable of anything."

Efficiency is a concept used to measure a system's effectiveness against a purpose or goal. What is God's purpose that the system is inefficient?

World of Facts said...

Steve said:
"The way that DNA works, it requires a piece of "machinery" to read the "code" and use that code as instructions to build something."
No, that's not how it works. It is a good analogy to simplify the description of the process, but nothing actually 'reads' the code nor 'use' the code. The "code" itself reacts following Chemistry laws to form proteins, cell machinery, organs, organisms... the analogy breaks down right where you're trying to use it to make an argument in favor of design.

David Brightly said...

Some links: Ribosome, RNA world.

Steve Lovell said...


Thank you, those are helpful links.

I've stumbled across the "RNA World" hypothesis before. I'd like to read more about it and perhaps will when time allows.


Based on the first link from David I think my illustration holds up pretty well. However, as I originally said, it was merely an illustration. I was using it to show how the argument might work and I hedged it in all manner of qualifications to make it clear that I wasn't especially wedded to it. If I thought of what I'd written as "the whole argument" then David's second link would be a serious challenge. I don't think many ID theorists would use my example as their X.


I totally agree about the "problems with monkeys". They become a much better explanation when any success they achieve can be preserved and built upon by later stages in the process. Evolution gives a means for that to happen. The ID theorists point, or attempt to point, to cases where no such means are available.

The difficulty is, I believe, that the more basic and therefore more probable Xs are unstable in any environment which they are remotely likely to have occured in.

I again admit that I have no specialist knowledge here. This is the sort of thing I've read from others ... How good it is as science I simply cannot tell you.

Unknown said...


I haven’t read Mere Creation, but I’m familiar with Demsbki’s books “The Design Inference” and “No Free Lunch.” (He specifically addresses the bacterial flagellum, not DNA, but my criticism of his approach is equally relevant to this argument from DNA). Let’s say we can use his universal probability bound of 1 in 10^150 to definitively rule out “chance” as an explanation. But Dembski alternates between different definitions of “chance” when it suits him. Sometimes he defines “chance” as simply “everything other than design.” But when he wants to say that the probability of “chance” producing a biological structure is below 1 in 10^150, he adopts a definition a pure randomness, an explicitly not Darwinian approach. He uses this definition because it allows him to apply a uniform probability distribution and perform a simple probability calculation. But no biologist proposes that complex biological structures came about by “random chance.” Dembski is therefore completely wrong when he writes that his calculations “render natural selection utterly implausible as a mechanism for generating the flagellum and structures like it.” (Design Inference,p. 164). Dembski is aware that biologists have proposed non-random Darwinian explanations for the flagellum, but he does not consider them to be “live possibilities” because “the scientific literature shows a complete absence of concrete, causally detailed proposals for how [they] might happen.” No Free Lunch, pp 255-56. He seems unconcerned that ID lacks the ability to provide such proposals, even in principle (given that the ID designer is a generic hypothesis that someone, somewhere, somehow did something for some unknown reason to create life). Dembski is making demands from evolutionary theory that he does not make from ID. Dembski also dismisses the relevant nonrandom “chance” hypotheses as “too implausible” without ever calculating their probabilities . No Free Lunch, pp254-256. But Dembski himself says that hypotheses cannot be dismissed just because they seem implausible. See Design Inference, p. 228 (“There is a calculation to be performed. Do the calculation. Take the numbers seriously. See if the underlying probabilities really are small enough to yield design.”)

Basically, anytime I hear an ID advocate say that the universe’s probabilistic resources have been outstripped, whenever I look into it I always find that they aren’t saying that “non design” doesn’t have the probabilistic resources, but rather only that “random chance” doesn’t have the resources. But random chance is only a small subset of the “nondesign” category that no one ever took seriously to begin with. So P(DNA|nondesign)=P(random chance|nondesign)xP(DNA|random chance)+P(nonrandom “chance”|nondesign)xP(DNA|nonrandom chance). Even if we can say P(DNA|nonrandom chance).is low, we cannot say that P(DNA|nondesign) is also low because P(DNA|nonrandom “chance”) is essentially inscrutable at this point

Steve Lovell said...

Thanks Ron,

I don't have time for a detailed response, although I probably don't have much to say in any case as I think that I pretty much agree with you here.

It's difficult to assess the probabilities on anything other than a "pure chance" model, and clearly the probabilities associated with the other forms of explanation are also relevant. And those certainly are difficult to evaluate. The particularly important part is that the "nonrandom non-design" explanations may include elements of interaction between chance and necessity, just as Darwinism does.

I understand your point about ID advocates not requiring the same standards of their explanations as they do of others, but I can't help feeling that there may be something legitimate about that. They are offering a quite different kind of explanation.

If at the scene of a death we have the options of attributing that death either to deliberate "design" or to a tragic "undesigned" accident ... I'm not sure we'd require quite the same things of both positions before we started to take them seriously.

That'll have to do for now. Not that I'm too sure I have anything further to add.