On the Circularity of the Argument from Intelligent Design

There is a lot of debate in the comments to recent posts about whether the argument from ID is circular.  I thought it would be worth calling this out as a separate item.

I plead that participants in this discussion (whether they comment here or on UD):

  • make a real effort to stick to Lizzie’s principles (and her personal example) of respect for opposing viewpoints and politeness
  • confine the discussion to this specific point (there is plenty of opportunity to discuss other points elsewhere and there is the sandbox)

What follows has been covered a thousand times. I simple repeat it in as rigorous a manner as I can to provide a basis for the ensuing discussion (if any!)

First, a couple of definitions.

A) For the purposes this discussion I will use “natural” to mean “has no element of design”. I do not mean to imply anything about materialism versus supernatural or such like. It is just an abbreviation for “not-designed”.

B) X is a “good explanation” for Y if and only if:

i) We have good reason to suppose X exists

ii) The probability of Y given X is reasonably high (say 0.1 or higher). There may of course be better explanations for Y where the   probability is even higher.

Note that X may include design or be natural.

As I understand it, a common form of the ID argument is:

1) Identify some characteristic of outcomes such as CSI, FSCI or dFSCI. I will use dFSCI as an example in what follows but the point applies equally to the others.

2) Note that in all cases where an outcome has dFSCI, and a good explanation of the outcome is known, then the good explanation includes design and there is no good natural explanation.

3) Conclude there is a strong empirical relationship between dFSCI and design.

4) Note that living things include many examples of dFSCI.

5) Infer that there is a very strong case that living things are also designed.

This argument can be attacked from many angles but I want to concentrate on the circularity issue. The key point being that it is part of the definition of dFSCI (and the other measures) that there is no good natural explanation.

It follows that if a good natural explanation is identified then that outcome no longer has dFSCI.  So it is true by definition that all outcomes with dFSCI fall into two categories:

  • A good explanation has been identified and it is design
  • No good explanation has yet been identified

Note that it was not necessary to do any empirical observation to prove this. It must always be the case from the definition of dFSCI that whenever a good explanation is identified it includes design.

I appreciate that as it stands this argument does not do justice to the ID position. If dFSCI was simply a synonym for “no  good natural explanation” then the case for circularity would be obviously true. But is incorporates other features (as do its cousins CSI and FSCI). So for example dFSCI incorporates attributes such as digital, functional and not compressible – while CSI (in its most recent definition) includes the attribute compressible. So if we describe any of the measures as a set of features {F} plus the condition that if a good natural explanation is discovered then measure no longer applies – then it is possible to recast the ID argument this way:

“For all outcomes where {F} is observed then when a good  explanation is identified it turns out to be designed and there is no good natural explanation. Many aspects of life have {F}.  Therefore, there is good reason to suppose that design will be a good explanation and there will be no good natural explanation.”

The problem here is that while CSI, FSCI and dFSCI all agree on the “no good natural explanation” clause they differ widely on {F}. For Dembski’s CSI {F} is essentially equivalent to compressible (he refers to it as “simple” but defines “simple” mathematically in terms of easily compressible). While for FSCI {F} includes “has a function” and in some descriptions “not compressible”. dFSCI adds the additional property of being digital to FSCI.

By themselves both compressible and non-compressible phenomena clearly can have both natural and designed explanations.  The structure of a crystal is highly compressible. CSI has no other relevant property and the case for circularity seems to be made at this point. But  FSCI and dFSCI  add the condition of being functional which perhaps makes all the difference.  However, the word “functional” also introduces a risk of circularity.  “Functional” usually means “has a purpose” which implies a purpose which implies a mind.  In archaeology an artefact is functional if it can be seen to fulfil some past person’s purpose – even if that purpose is artistic. So if something has the attribute of being functional it follows by definition that a mind was involved. This means that by definition it is extremely likely, if not certain, that it was designed (of course, it is possible that it may have a good natural explanation and by coincidence also happen to fulfil someone’s purpose). To declare something to be functional is to declare it is engaged with a purpose and a mind – no empirical research is required to establish that a mind is involved with a functional thing in this sense.

But there remains a way of trying to steer FSCI and dFSCI away from circularity. When the term FSCI is applied to living things it appears a rather different meaning of “functional” is being used.  There is no mind whose purpose is being fulfilled. It simply means the object (protein, gene or whatever) has a role in keeping the organism alive. Much as greenhouse gasses have a role in keeping the earth’s surface temperature at around 30 degrees. In this case of course “functional” does not imply the involvement of a mind. But then there are plenty of examples of functional phenomena in this sense which have good natural explanations.

The argument to circularity is more complicated than it may appear and deserves careful analysis rather than vitriol – but if studied in detail it is compelling.

171 thoughts on “On the Circularity of the Argument from Intelligent Design

  1. Slightly tangential, but related to the topic of circularity.

    Gödel’s incompleteness results have to be among the most abused in all of mathematics and science. Check out this quote from an opinion piece in the Asia Times:

    The most devastating refutation of determinism came from Austrian mathematician Kurt Goedel, whose famous incompleteness theorems of 1931 proved that no mathematical system could prove all of its own assumptions.

  2. Thanks for the laugh.

    Yes, I have seen lots of crazy ideas that are said to be proved by Gödel’s incompleteness theorem.  I guess when people don’t understand it, they allow their imaginations to run wild.

    For myself, I see Gödel’s theorem as a very technical result in mathematical logic with no relevance to the real world.

  3. There’s a famous mathematical story. When Gödel’s theorem appeared, the famous English mathematician G.H. Hardy said that of course we should have known this — we know that when one has inconsistent axioms we can prove anything true or anything false. Including proving from them that the axioms are consistent!  So a proof of consistency is not valid.

    He was then asked “do you mean that if I grant you that 2+2=5 that you can prove that McTaggart is the Pope? Without hesitation he said of course: “2+2=5 but we also know that 2+2=4. So 5=4. Subtracting 3 from each side, 2=1. Now McTaggart and the Pope are two, therefore McTaggart and the Pope are one.”

    The funny thing is, that is not just a word game, it is a legitimate proof. 

  4. Keiths

    By the way, it’s also still circular: Today at 12:15 pm, gpuccio thinks this gene could not have evolved. Why? Because it exhibits dFSCIgpuccio@12:15 pm. Why does he think it exhibits dFSCIgpuccio@12:15 pm? Because today at 12:15 pm, gpuccio thinks this gene could not have evolved.

    Yes, that’s a sort of circularity but it is still true that it may exhibit dFSCIgpuccio@12:15 pm and not be designed. So in that sense it is not circular. 

  5. Toronto: ID can’t be refuted because it has made no claims of its own.

    You could change all that Mung by showing how to change a single “bit” in the DNA of a living creature.”

    Mung: “We’ve been doing it for decades now. “

    No, we’ve been modifying the DNA of cells that later become living creatures.

    My whole point in specifying living creatures is that waiting for organisms to get to adulthood before the “designer” gets feedback on the prototype “design” change, forces the “designer” to have to deal with the same problems evolution does, i.e. “Let’s try this and see if it works”.

    “Design” has two ways out though.

    1) Simply show the mechanism by which the designer can foresee future needs and effects on a population and their competitor populations, and this removes trial and error.

    2) Instead of changing the population, change the environment to match the abilities of the population taking into account competitor populations.

    Either requires the designer to have the abilities of the god in Genesis.


  6. The difference between a circular argument for ID and a non-circular argument for ID:


    1. X can’t be produced by RV or a known necessity mechanism.
    2. Therefore X exhibits dFSCI.
    3. Therefore X can’t be produced by RV or a known necessity mechanism.


    1. If you remove a part from X, its function ceases.
    2. Therefore X is irreducibly complex.
    3. Therefore X could not have evolved.

    The second argument isn’t circular, but it’s incorrect. Step 3 does not follow from step 2.

    What IDers need is an argument that is both correct and non-circular. Good luck to them.

  7. And don’t forget the next step in both arguments:

    4. Therefore we can make the default assumption that X is designed! 

    So correct, non-circular and not relying on Sherlock Holmes.

  8. Gilbert Harman

    “The inference to the best explanation” corresponds approximately to what others have called “abduction,” the method of hypothesis,” “hypothetic inference,” “the method of elimination,” “eliminative induction,” and “theoretical inference.” I prefer my own terminology because I believe that it avoids most of the misleading suggestions of the alternative terminologies. In making this inference one infers, from the fact that a certain hypothesis would explain the evidence, to the truth of that hypothesis. In general, there will be several hypotheses which might explain the evidence, so one must be able to reject all such alternative hypotheses before one is warranted in making the inference. Thus one infers, from the premise that a given hypothesis would provide a “better” explanation for the evidence than would any other hypothesis, to the conclusion that the given hypothesis is true.

    In relation to ID, what does the hypothesis “therefore design” actually explain? There is no design hypothesis to infer. Someone among ID theorists should perhaps address this omission.


    Having an ID hypothesis that was some kind of alternative explanation for the observed pattern of diversity in past and present life would then enable an inference to the best explanation to be drawn. Omitting this vital step reminds me of another argument that glosses over important details.

  9. Mung 192. I just noticed this:

    To your OP On the Circularity of the Argument from Intelligent Design I posted somewhat in the area of 15 responses. I saw a total of two replies from you.

    Is that how you deal with refutations of your arguments? Just ignore them?

    Mung I count 15 comments from you last night on the The TSZ and Jerad Thread alone. Only a few of them were for me. How on earth I can be expected to pick out the comments for me from all the others?  Do you think I read every comment on the thread?  I look out for those from Gpuccio as he is intelligent and I am in the middle of a dialogue with him.  I reply to others if I notice them.  If you really want me to respond then post here.

    PS I have noticed your comment 202 and will respond separately.

  10. Mung 202

    You write:

    First you must decide on a function.

    Then you must find a string that performs the function.

    Then you must find an algorithm to create a string that performs the function.

    Then you must solve the search for a search problem.

    Then you must establish that the entire process is deterministic.

    This sort of answers my question but I am not at all sure that Gpuccio would be happy with your answer! I am not sure what you mean by the “search for a search” problem.  If you are referring to the LCI this is highly controversial and I do not accept it at all. But anyway what would be the problem in practice?

  11. Out of the “infinity” of worldviews, there are at least two general ones. The view that complex phenomena are the work of capricious (read intelligent) daemons or spirits, and the view that phenomena are the working out of regular processes.

    As many of us have pointed out, there are thousands of phenomena that have come under the umbrella of “regular phenomena.” 

    There are no instances of explanations going the other way.

    When an argument depends on gaps in knowledge, it is just a matter of time before it becomes obsolete. I have followed the evolution  debate since about 1956 and have watched hundreds of arguments fall by the wayside.

  12. http://www.herbert-armstrong.org/Plain%20Truth%201960s/Plain%20Truth%201969%20%28Prelim%20No%2005%29%20May.pdf

    Warning: this is an 18 megabyte pdf file.

    Here’s an example of an obsolete claim (my bolding). I encountered this while I was in college, back in 1969. It was recognizable as rubbish then, even more so today. But that didn’t stop a glossy magazine from making the claim. Just as the Discovery Institute continues shamelessly to publish rubbish regarding the Second Law.

    SCIENTISTS are still in a state of shock after having discovered the basic blueprint of life.

    They call it DNA – short for the almost unpronounceable word deoxyribonucleic acid. DNA is a genetic code, like a master computer or file. It is in the nucleus of every living cell, programmed instructions, for example, make a cow reproduce a calf, not some other kind of animal, or make a liver cell reproduce a liver cell and not a heart cell.

    This genetic recipe is so complex that the entire DNA system within a human contains as much information as several encyclopedia sets, or one hundred large dictionaries. DNA – with incredible accuracy – will direct the reproduction of a colt from a horse. It will split any living cell into exact twins. This DNA could be likened to a master stencil grinding out endless copies of itself.

    But-and here is the phenomenal KEY that allows mind-defying variety within a set kind. When a new life is engendered, two DNA codes or stencils are used one each contained in the father’s and mother’s chromosomes. DNA reproduces a limitless amount of variety, all molded by a similar template.

    This variety, though extraordinary in scope, is limited by the fact that parents are of a similar kind. As a result, variety within a species can reach astronomical proportions. (For example, note the number of varieties among moths and butterflies.) But, because DNA reproduces itself exactly, kind reproduces kind. Evolution cannot occur!

  13. To recap:

    The argument was never demonstrated to be circular.

    The argument that it is circular fails.


  14. Could you stick a little mustard on that burger, Mung? The thread’s been dead for a while. It would help if you could summarise the argument on circularity as you perceive it, and explain why you think it fails. So far, you’ve just said “This Is So”.

  15. It would help if you could summarise the argument on circularity as you perceive it, and explain why you think it fails

    It would help if you could summarize the argument on circularity as you perceive it, and explain why you think it succeeds.

  16. Sorry, the thread’s been moribund even longer now. You wish to kick the corpse: provide some substance. 

    Otherwise it’s just a pointless assertion-fest. To borrow your game:

    To recap:

    The argument was demonstrated to be circular.

    The argument that it is circular wins.


  17. The claim that the argument was demonstrated to be circular cannot be sustained from either evidence or argument.


  18. er … The claim that the argument was not demonstrated to be circular and can be sustained from either evidence or argument has not been sustained by either evidence or argument. Gee, this is fun!

  19. The claim that the argument was not demonstrated to be circular and can be sustained …

    You blinked. So you lose.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.