Does Science Consider Design Without Considering a Designer?

From email concerning James Tour and Joshua Swamidass: A Fiery Debate on the Origin of Life.

I was listening to the discussion that you and Dr Tour had on “Capturing Christianity” channel about the origin of life. I myself am not a biologist or chemist; I am completing MSc in Energy Systems at Oxford. I am a Christian, and I find arguments for intelligent design (that the coded information in the cell points to some activity by an intelligent agent) to be the most reasonable from what I’ve read and heard on the origin of life. In the video, you stated that, among other things, there should be a clear idea of who the designer is and what he can or cannot do for ID to be a proper scientific argument. To be honest, I didn’t agree at all, so I wanted to understand your thoughts better. For example, the SETI program looks for signs of intelligent life in the universe, and receiving some digital signal with coded information would be an evidence for that. Would you say that we must know who sent the signal and what the sender can and cannot do for us to infer that there is intelligence behind it (as a reasonable scientific proposition)? Or say if we went to Mars and found some form of a machine there (e.g. a computer, but much more sophisticated than ours, so that we couldn’t even build a copy of it), would you say that we must know who built that computer and what they are capable of in order to infer that some intelligence must’ve designed it.

To be clear, if I understand correctly, Stephen Meyer doesn’t even claim that all of the cell is best explained by intelligent activity; he proposes that ID adds another analytical tool (insofar as informational content is concerned), because the only cause known to produce coded information is intelligence. As an analogy, if there was a single chemical reaction known to produce some feature of cell, it would be scientifically reasonable to infer that that reaction too place in abiogenesis, alongside with everything else. It seems to me that the reason ID hypothesis doesn’t state who the designer is is precisely for it to be a scientific hypothesis, because scientifically speaking we can only infer that some intelligence did it, just as we would if we receive a radio signal (like in the movie “Contact”) or discover a machine on another planet. Just as many scientists believe that solely chemical processes must’ve been responsible for abiogenesis without necessarily being able to specify which processes exactly, so do I (admittedly non-scientist) believe that we can construct a scientific argument for ID without being able (within the strict limits of biochemical science) to identify who the designer exactly is, for that would be a philosophical (in my case, theological) inference. I look forward to hearing your thoughts on this subject, and wish your family stays safe during the current crisis.

I’m not going to answer this completely now, but I’ll start. There seems to be some real confusion about how science detects design. Case in point is SETI.

In both these cases, scientists would be forming and refining (informally or formally) a model of the designer. A great example is how scientists ruled out human-design of the COVID-19 virus, with a particular model of design: COVID-19 genome and design detection - #18 by swamidass. In the case of SETI, in almost all cases they start with a model of how aliens would produce a signal. When putative signals have been found, they look for other evidence of the designer to confirm, which once again depends on a model of the designer.

SETI, also, has the distinction of never touting a signal as evidence of intelligent aliens. So their filter has been very successful. In contrast, ID has touted several signals that fell apart on scrutiny.

There is also a misleading analogy being used here, regarding a “machine on another planet.” It is notable that they did not say “life on another planet.” We can already see that this changes the character of the question entirely. The fact of the matter is that, even if life is in fact designed, it doesn’t have the same signature of design as would machinery.

Moreover, if we did find machinery on another planet, we would immediately consider different models of a designer. On top of the list would be other people on earth who sent it there, and that’s an example of a model of the designer. Until there was confirmation of an extraterrestrial designer by other lines of evidence, it is highly unlikely we would conclude it was an intelligent alien.

I could go on, but it is important to keep in mind that when Meyers draws analogies to these different things, these analogies break down. Where and how spectacularly the break down is important. SETI is a good case study, in fact, because the methodology used by SETI is inconsistent with the methodology used by ID. I agree with Meyers, however, that science can and should consider design. In fact, it already does. The innovation of ID, however, is to consider design without considering a designer, but that doesn’t work in science. That’s the problem in the end.

There could be a way forward. Some IDists have proposed models of designer, and I have endeavored to engage these models with rigor: Winston Ewert: The Dependency Graph of Life.

7 Likes

3 posts were split to a new topic: Is SETI Science?

Dear Dr Swamidass,

Thank you for your reply! The way I used the analogy with a machine may have been misleading, as I didn’t intent to use a machine as an analogy to DNA, only to argue that we can infer intelligence (whether from machine design or digital code design) without knowing who that intelligence is. You say that we would immediately consider different models of a designer if we find a machine; but that underscores my point that we would believe intelligence to be have taken part, and even if we couldn’t decide if its humans or aliens who did that, we would simply keep that second question in the open. Do you believe that identifying that there is a designer and building a model of who that designer is must go together, i.e. that the first step is not valid on its own unless we can perform the second one? (Incidentally, I slightly disagree with you in that that machine, if it were shown to far exceed the capacity of human intelligence, could on its own be considered evidence for aliens, even if we wouldn’t find anything else).

It seems to me the real question then is, while DNA and a machine are not identical, is coded information (a key feature of DNA) as strong an indicator of intelligence as a machine, or at least strong enough to infer intelligence the way we would from a machine. In regards to signals from aliens, you’re right to point out that it doesn’t automatically tout intelligence, and I did a mistake of not specifying my point: I meant a signal of coded information, not simply, say, a radio wave that could be emitted by a satellite or a natural cause. I wanted to argue that, just as the only known cause of machines is intelligence, so the only known cause of coded information is intelligence. If SETI has received coded information from the space and showed it to be caused by natural phenomena, then my argument fails.

I very much appreciate that you take time to respond to my questions/thoughts, and I will look further into what you pointed out about SETI programs and designer models.

Can you explain why?

Because for both of them, as far as I know, there is only one known cause - intelligence. Even more, I think there’s strong theoretical argument against natural processes being able to generate coded information. Information is inversely related to uncertainty (you can use Shannon’s information theory, where the amount of information generated equals log(1/p), with p being probability of a particular outcome. As an analogy, if you know that tomorrow will be sunny and I tell you that it will be sunny, you gain zero information. But if you didn’t know if it will be sunny or windy (there was uncertainty), then you would gain information. This is at the core of information theory. Polanyi argued that you cannot reduce information (such as DNA) to chemical and biological processes, because they can at best produce predetermined, repetitive order.

Here is an illustration: imagine you are a prisoner, and you are allowed to write a letter to your friends. But there’s a condition: every time you write an “A”, you must follow with “B”, and every time you write “B”, you must follow with “C”, and etc. Given those conditions, you wouldn’t be able to transmit any information, because your code would be predetermined, you have no freedom of placing the letters differently. In a similar manner, if the sequence of nucleotides in DNA/RNA were determined by chemical bonds/laws, it would be a repetitive sequence, capable of carrying very little information. The natural processes (as described by natural laws) operate by a deterministic logic “if A, then B”, while the whole point of coded information (whether in DNA, or human language, or computer code) is that there are many possible sequences, and deliberately few have to be picked to transfer information.

Polanyi writes “Suppose that the actual structure of the DNA molecule were due to the fact that the bindings of its bases were much stronger than the bindings would be for any other distribution of bases, then such a DNA molecule would have no information content. Its code-like character would be effaced by an overwhelming redundancy… Whatever may be the origin of a DNA configuration, it can function as a code only if its order is not due to the forces of potential energy. It must be as physically indeterminate as the sequence of words is on a printed page.” (Life’s Irreducible Structure, Polanyi, 1968).

So to sum up, the fact that the only known cause of coded information in our experience is intelligence, and that there’s what I see as a good theoretical argument against natural processes generating information, I think coded information is a strong reason to infer intelligence. The alternative would be to propose that information in the DNA got put together by chance, by from what I know that has been mostly abandoned by origin of life researchers, as the probabilities are astronomically low.

Regards :slight_smile:

@Armen_Danielyan, we are very informed about ID arguments. Just reiterating them isn’t going to get you very far. I’d rather we didn’t. Instead our goal is to seek understanding. In this case, you might benefit from learning why we reject ID arguments.

In this case, key premises are trivially false.

That isn’t a fact. It is a false. Very commonly apparently non-intelligent causes produce coded information. One such example is the evolution of cancer: Cancer and Evolution

If that premise is false, and it is, nothing else follows. Perhaps life is designed, but we can know with certainty that this argument (and any argument dependent on that premise) is an invalid argument for design.

5 Likes

I haven’t seen any “argument for ID” with any strength yet. But I do agree, unequivocally, that design detection is different from designer detection. In practice, as @swamidass argues, design detection is always linked to known designers and types of designers. (At least that’s my read of his comments.) But in principle, we could devise a definition of a set of features that represents design, without even hinting at the nature (or existence) of the designer. My position is that design is a thing, that it exists and requires explanation, and also that its existence does not always imply a designer (and that this too requires explanation).

Of course, it’s easy to define design as something that comes from a designer, but I don’t think that’s interesting at all. Design without a designer is really interesting.

Perhaps, but I can’t conceive of how this could be done without embedding implicit assumption of what type of designer’s design and the type of design you are looking for. Certainly one could massage one’s language to avoid explicit mention of a designer, but design implies designer, and and particularity to design implies particularity to the designer.

One might as well talk about a definition of “husband” that doesn’t imply a spouse.

2 Likes

You do you, but I’m not granting that definition of design.

This is not true. There are many natural processes known to encode information without the use of external intelligence. Tree rings encode information about local environmental conditions in ring widths. Spectral lines in starlight encode information about the elements in the star. In the same way arrangements of base pairs in genomes encode information about the local environment the creature lives. The only codes requiring intelligence are ones where arbitrary symbols are used as agreed upon abstractions to pass messages between a sender and receiver. Examples are Morse code and computer code. DNA does not meet such criteria. There is no intelligence-requiring abstraction in the DNA → amino acid process anywhere. There is no sender, no receiver, no message passed. Life is an extremely complex but entirely natural self-sustaining chemical reaction.

2 Likes

(Also a question for @sfmatheson)
When you use the word “design”, how much does it overlap with “purpose/teleology/function”? Are the two identical?

1 Like

Remember the Information to do this was already there. The tree was designed…

Good question. I’m referring to the thing we see when we look at ATP synthase or halteres or lenses in eyes. I suppose an intersection of purpose and function, that moves me to say “I see design in this.” Different from when I look at a mountain and say “that’s beautiful,” even though both can be believed to be creations of a deity. I’m talking about whatever an ID person is talking about when they geek out about suckers on lizard feet or whatever is coolest about bladderworts.

Teleology is a bit more slippery for me but I’m currently reading a fascinating if somewhat maddening defense (it seems) of teleology by David Haig. (This book.) My hesitations are primarily due to what sounds like “looking ahead” by evolution but I think Haig is going to attempt to dismantle that response.

1 Like

Thank you for your responses.

My impression was that cancer is a result of mutations in an already existing DNA code, not an example of creating new information from raw materials. I’ll research on it further. As for “seeking understanding”, I was simply replying to Rumraket Mikkel, who asked “can you explain why?”, so I wasn’t reiterating them for you, sir, I’m sure you are aware of the arguments. Surely me replying to someone who asked for further explanation doesn’t entail “just reiterating”, but rather “seeking understanding” with each other. Perhaps that was just a misunderstanding between us.

I wish you best,
Armen Danielian

1 Like

Dear Timothy, I do not believe that any of those is an example of coded information in the way that DNA is. Tree rings or spectral lines are simply environmental features from which we can deduce information about their past. DNA information, on the other hand, is an actual encoded information that carries instructions for cell construction and operation, like a computer code. A computer software operates on the basis of a code, and the analogy is not about sending or receive information between intelligent beings via a computer code, but about a computer operating on the basis of a binary code. At least, everything I’ve read from biologists (of whichever view) tells me that DNA is a code required to perform function within a cell. Tree rings are simply a feature of the tree that remain as an aftereffect of cyclical (repetitive) environmental changes, whereas DNA is much more like a computer code, because the nucleotides have to be arranged in a specific sequence to enable function.

No, this is incorrect. DNA is not like a computer code. Computer code is sometimes used as an analogy for DNA but the operations of the two are fundamentally different. Once again there is no intelligence-required abstraction in DNA like there is in computer code. DNA doesn’t “carry instructions”. That’s merely another analogy with human devices. DNA simply follows the laws of chemistry and physics. Intelligent Design advocates have exploited the layman’s misunderstanding of biology and genetics to push the DNA=computer code false equivalence.

1 Like

First of all I think it’s important we get the terminology right. You speak of “coded information in DNA”, by which someone who knows genetics would think you are referring to the genetic code, as in the system of translation whereby DNA sequence is translated into amino acid sequence in the process of protein biosynthesis. But later in your post you make it clear you’re actually referring to functional DNA sequences(as in merely the sequence of bases in a string of DNA), not the genetic code.

A second problem is DNA is generally not thought to have come with the origin of life(though some do postulate this, it is a minority view afaik), as there is some evidence that indicates it was preceded by the related genetic polymer RNA. But in any case that still leaves the question open of what caused any putative functional RNA sequence to first originate. Good question, we don’t know.

A problem I have with your inductive generalization here is it ignores important evidence, which is that it’s human beings that have made all known sequences of information for which we know the origin. But for DNA(or RNA), we not only don’t know the origin, we know it couldn’t have been made by humans. Even if you postulate another intelligent designer than humans, just hypothetically, we still only know of intelligent designers that take the form of cellular life. So we’re still left wondering how that life would come into existence.

A further problem is there is another perfectly good candidate explanation for sequence information, which is evolution by natural selection. After all, evolution is a natural process, essentially just physics and chemistry, occurring in large populations of organisms, which today creates novel genetic sequence information.

That seems to me nothing but an assertion, flatly contradicted by what we know from evolutionary biology. Novel functional DNA or RNA sequences can and do evolve by mutation and natural selection, which at a basic level is just chemical and physical processes giving rise to the sequence information in the DNA or RNA polymer.

So that is a natural physical and chemical process giving rise to sequence information. Granted it happens to already living, cellular organisms, raising the question how the first life came into existence(which we don’t know), but the fact remains that we do then have an example of a mindless natural process creating sequence information.

A fourth issue is, as far as I know, all known intelligent designers are physical cellular organisms that evolved by natural processes. Saying there was an intelligent designer that made life is sort of to put the cart before the horse.

Sure but, they aren’t. There is no such rule that says if one DNA base reads as G, the next one must read as C or whatever. DNA sequences can and do demonstrably mutate. G can be substituted for C, or T, or A. There is no chemical or physical laws that constrain or demand what a particular DNA sequence has to be. The only thing that constrains this is natural selection. But here it is the phenotypic effect of the DNA sequence on the reproductive success of the carrier orgaism, that determines whether it is favored or disfavored (or neutral) by natural selection. Not some sort of chemical bond.

It is exactly this mostly unconstrained malleability of DNA sequences, that they can mutate, and that the mutations have phenotypic effects, that is to say it’s potential for step-wise, heritable, incremental change, that makes evolution possible. There is just no in-principle barrier to any particular sequence of DNA evolving. AGCTATAC can change into AGCTCTAC, which can change into AACTCTAC, which can change into AACTCTAT, etc. There just is no physical law that says AGC must be followed by T.

Since researchers generally don’t think life began with DNA, I think we can just substitute in RNA here for your argument instead.

But in this case you are mistaken. There are many researchers who posit a role for, at least initially randomly generated RNA sequences, with useful functions at the origin of life. It just isn’t clear at all that the probabilities would be astronomically low to an extend that would make this problematic, as it isn’t really known what useful functions the first RNA molecules could potentially have had, as this would require one to know the possible cellular or environmental context in which they might have evolved. You might be tempted to think about large genes found in extant life, many dusins if not hundreds or even thousands of nucleotides long, as being required to have funcitonal RNA or DNA molecules. Researchers generally don’t think such molecules are what life began with(among other reasons because, as you say, that would potentially make the odds of their spontaneous emergence extremely low).

But much, much smaller, functional RNA sequences are known. Just to give an example of a very small but functional RNA molecule, there is a 5-base long functional ribozyme known.

2 Likes

This has always seemed to be a problematic argument for ID. It would imply that if a designer created a simple RNA replicator that the natural process of evolution could take over from there and produce all of the biodiversity we see today.

In my time discussing these topics for over a decade, understanding is difficult because the two sides are approaching the question from very different positions. Supporters of ID tend start with a conclusion that coded information must be from an intelligence, and their definitions and applications are solely serving that conclusion. Essentially, ID supporters define coded information as “whatever we decide evolution can’t do”. Skeptics of ID start from first principles and ask how we can define coded information at its most basic. I think this is what causes a lot of misunderstanding and friction.

To what extent do you think Behe has done this?

If you are claiming that DNA contains information, which is reasonable, then the above statement is trivially wrong. Information in DNA is not caused by intelligence. Actually, there are countless forms of information that do not require intelligence. Why did my vet ask for a stool sample when I took my dog in for a check up? Because it contained information. No intelligent force put it there. Or do you disagree?

1 Like