Here is my attempt to explain Sober’s reason for thinking …

Comment on Elliot Sober: Just Don’t Call the Designer “God” by Brad.

Here is my attempt to explain Sober’s reason for thinking there is a difference between design inference in ordinary affairs and design inference in the biological case.

Sober formulates the design argument as a likelihood argument. The premises are that Pr(O │ Chance) is very low and Pr(O │ Design) is higher, where O refers to some observed feature of the biological world. The conclusion is that O supports Design over Chance. Note that here I am summarising his arguments on the biological design argument, not the cosmic design argument, which brings in independent issues. For those who are interested in the details, there is a more much more detailed development of Sober’s argument concerning the biological case in Sober (2008).

The first thing to appreciate about the likelihood formulation is that the conclusion of the argument is not that design is more likely than chance, but rather more weakly that certain observations O provide evidence for the hypothesis of design over the hypothesis of chance. As he writes:

[L]ikelihood arguments have rather modest pretensions. They don’t tell you which hypotheses to believe; in fact, they don’t even tell you which hypotheses are probably
true. Rather, they evaluate how the observations at hand discriminate among the hypotheses under consideration.

The second thing to appreciate is that the argument provides a comparative evaluation of two hypotheses. These hypotheses may not be exhaustive, so it does not itself ground the conclusion that the favoured hypotheses is better supported by the evidence than all possible rivals. (Incidentally, this makes for a difference between the likelihood version of the argument and the version given by Dembski (1998). For more on why Sober’s formulation should be preferred, see Fitelson, Stephens and Sober (1999)).

Note also that this entails that design can not be supported merely by showing phenomena to have low probability conditional on the evolutionary hypothesis. Rather, it must be shown that those phenomena have higher probability conditional on the design hypothesis.

Sober’s objection to the argument is deceptively simple. It is that we have no independent evidence for Pr(O │ Design), and so cannot justify the claim that Pr(O │ Design) > Pr(O │ Chance). This objection can be subdivided into two points.

First, in order to estimate the likelihood of some observation conditional on design, we need to know both the goals and abilities of the designer. For ordinary design inferences, we do this by inductive inference from cases with which we are familiar. Sober’s first point is that this ordinary form of inductive inference is unavailable for biological design arguments, since we do not have any instances of agents with abilities of the required magnitude. Since we have no instances, we simply have no idea what their goals would be. That is, we have no independent evidence of creatures existing before the biological systems in question, and capable of engineering those systems. So we have no idea what they would be likely to build if they existed. So we have no way of estimating Pr(O │ Design). Moreover, there is an additional obstacle to this form of inductive inference if we attempt to infer to a supernatural designer or designers. For in the case of natural designers we at least have the possibility of appealing to some general naturalistic constraints on their goals and abilities, while in the case of a supernatural designer we have no such constraints.

Second, while it is easy to arbitrarily define a class of design hypotheses such that the probability of the observations conditional on that class is high, it is just as easy to arbitrarily define a class of non-design hypotheses such that the probability of the observations conditional on that class is high. So some independent motivation is required to justify the particular hypotheses invoked in the argument. But in the case of the biological design argument, we have no such independent evidence. So again, we have no way of estimating Pr(O │ Design). Here I think it is helpful to consider the possibility of overlooked naturalistic explanations for the observations in question. Just as for all we know there are possible designers who would build what we see and possible designers who would not, so there are natural processes that would result in what we see and those that would not. Absent independent evidence for the existence of any of these, we are in no position to comparatively assess their probabilities against each other.

Finally, note that this argument is independent of anything to do with evolutionary theory. If it works, it tells us that design is not even supported over chance for the observations in biology. (Of course, evolutionary theory is dramatically supported over chance, but that is neither here nor there with respect to the arguments above).

Let me apply the argument to some salient cases.

1a. We find a watch in a field. We have independent evidence of the existence of designers with the goals and abilities requisite for building watches, and thereby know that the likelihood of the watch conditional on the existence of those designers is vastly higher than the probability of the watch conditional on chance. So the observation supports design.

1b. We come across a Shakespearean sonnet, or find the pansies in a yard spelling the phrase, “Good Morning. We hope you have a great day!”. The inference is identical to 1a.

2. We observe that all bio-molecules have the same handedness. We have independent evidence that there are a variety of natural processes capable of producing this, though we do not know exactly which is most likely to have independently occurred or exactly what the probabilities are of their producing the chirality we see. Still, this is enough to show that the disjunction of this group of hypotheses confers a likelihood on chirality greater than chance. We have no such independent evidence for the goals and abilities of the designer on the hypothesis that chirality was designed. So the hypothesis that there is some natural explanation is supported over the hypothesis that the handedness is the product of design, and also over the hypothesis of chance.

3. We find smooth and symmetrical granite blocks in a place where it is extremely unlikely that they were produced by humans. All of the natural processes we know confer very low likelihoods on this and we have no idea whether there are other natural processes capable of producing them, though we also have no reason to think the hypotheses we are aware of exhaust the possibilities. We also have no independent evidence of the goals and abilities of hypothetical designers of these blocks. In sum, we are in a position where we are unable to assign determinate probabilities to any hypothesis that gives a non-infinitesimal likelihood. So we rightly suspend judgement on what the evidence supports. (I can imagine someone who independently assigns a non-infinitesimal probability to the existence of other life forms, and who also thinks there are general natural constraints on the sorts of goals and abilities such life forms would have. Such a person would be willing to suppose that the observation supports design. But note that the independent evidence is crucial here—there is no automatic inference to design absent this evidence).

4. We observe the total biological evidence. Evolutionary theory is well supported against chance, and design fails to confer likelihoods for the reasons that are now familier.

Now, reading between the lines of some earlier discussions, I suspect Sean ought to deny the likelihood formulation of the argument and support something more like what Dembski says. This is suggested by his repeated insistence that it just obvious that there are certain features that if observed support design, independently of any knowledge of the nature of the designer. I hope that Sober’s discussion makes it clear that this is far from obvious. If this is the locus of disagreement however, to properly address it would in turn require arguing over Fitelson, Stephens and Sober (1999), who I regard as having decisively discredited Demsbki’s formulation. I’m not sure I have the patience to go through all that, given the effort it’s taken to get on the same page here. But perhaps I am mistaken, and Sean disputes somethin else in what Sober says. If so, let’s hear it.


Dembski, William A. 1998. The Design Inference: Eliminating Chance through Small Probabilities, Cambridge University Press, Cambridge.

Fitelson, Branden, Christopher Stephens, and Elliott Sober. 1999. “How Not to Detect Design”, in Philosophy of Science, Vol. 66, No. 3, September 1999, pp. 472–488. Reprinted in in Pennock, Robert T (Ed). 2001. Intelligent Design Creationism and Its Critics: Philosophical, Theological and Scientific Perspectives, MIT Press, Cambridge MA, pp. 597–616.

Sober, Elliott. 2008. “Intelligent Design”, in Evidence and Evolution: The Logic Behind the Science, Cambridge University Press, Cambridge, pp. 109–188.

Brad Also Commented

Elliot Sober: Just Don’t Call the Designer “God”
Dear all,

I hope to be able to reply within the week—real life is getting in the way.


Elliot Sober: Just Don’t Call the Designer “God”

One caveat , if I may: there is no need to turn the argument into an attack on personalities. This might suggest an egotistical bias that I’m sure neither of you intend.

Do you have in mind the argument I’ve been running about why I do not take Sean’s argument about protein evolution seriously? I don’t intend that to be an ad hominem attack on Sean. Rather, I have been trying to explain why, even if his argument is perfectly good, none of us can be justified in believing that it is any good until it is at least published in a peer reviewed scientific journal. This, I believe, should be all of our attitudes on any scientific topic we are not experts on and are not confident on regarding the expertise of others.

Brad, I don’t know your faith or non faith background, but I think it would be useful for all of us to know in order to assess any bias of lack thereof.

I was raised a Seventh Day Adventist but am now an atheist in the sense that I do not regard belief in anything supernatural as reasonable. I have many friends and family members who remain Adventists. Is that sufficient?

Can chance be part of design?

Not sure exactly what you mean, but I can infer design from Pollock paintings. Indeed, I can infer Pollock.

Let me give you a hypothetical. Let’s say the three of us climbed into a life supporting bubble ship and were hooked up to an immortality machine, designed by man. The ship has the ability to travel to infinite, different universes and survive the physical characteristics of same (through wormholes of some other Star Trek like fictive concoction) Moreover we are aware each time we enter a new universe( the NEW UNIVERSE light comes on in the bubble and a toy monkey comes out and bangs a drum). Through our clear bubble we are able to observe the characteristics of each universe. Each one is different, but ours is the only one with organic life.

Here’s my question:

1. Would your ideas on ID, or God, be any different than they are know? If so, how?

Yes, I would believe that we were Gods ourselves—there are even three of us! Actually, in so far as I understand the idea, for me the answer is no.

Elliot Sober: Just Don’t Call the Designer “God”
@Sean Pitman:

I had assumed, perhaps wrongly, that you support Sober’s conclusion that SETI is a valid science and that the invocation of ID hypotheses can be quite reasonable short of identifying the intelligent agent as God.

Yes, I don’t think there is anything intrinsically unreasonable unscientific about invoking design hypotheses. The details depend on the context, as I have tried to show. However, I do disagree with SETI if that project is understood as founded on the belief that finding the signals they are looking for would in itself provide sufficient evidence to prefer the design hypothesis over the hypothesis that the signal was due to some unknown natural process. On this point, I am curious what your judgement is on the black hole example.

Are you saying it is possible to have a reasonable belief in a particular hypothesis without evidence?
Isn’t that what science is all about? – appropriate background information?

No, the reverse—it is possible to have evidence for a hypothesis without that hypothesis being reasonable to believe. We all have beliefs for which there is evidence both for and against, and we all have evidence too flimsy to found beliefs on. For example, if I buy a ticket in the lottery, my having the ticket provides some evidence that I will win. But not enough to believe that I will.

The “mass opinion” I’m talking about here is the mass opinion of scientists who are in the know. Given such scenarios as I’ve described scientists all over the world would herald the discovery of ETI – without any knowledge of motive or anything else about the ETI other than the overwhelmingly obvious conclusion the ETI is intelligent.

So to be clear, you are claiming that scientists in general believe that discovery of the signals SETI is looking for would in itself provide sufficient evidence to prefer the design hypothesis over the hypothesis that the signal was due to some unknown natural process? If so, do you have any evidence for this claim? Notice that I agree with you that we would all be very surprised and excited to find those signals, but merely disagree that this is because design would be obviously better supported that an unknown natural hypothesis (rather, I claim it is because design is thought to be obviously better supported than chance).

I don’t see how. For me your two statements here are quite inconsistent. How can you say, on the one hand, that my granite cubes would be obvious artifacts to the vast majority of scientists in the world, even if discovered on Mars, yet, on the other hand, claim that this hypothesis is not in any way superior to the claim that such cubes are just as likely to be the result of some heretofore unknown natural process? That argument simply doesn’t make any sense to me. If two hypothesis are in fact equally likely or unlikely, then, as you originally pointed out, the only reasonable conclusion or belief regarding which one is actually correct is, “I don’t know. Therefore, I ‘suspend judgment’ until more information is available.” It seems almost like you’re trying to backtrack away from your “suspend judgment” argument. Am I wrong?

Yes. I have never said that anyone would conclude that the observations obviously supported design over all other possibilities. You need to distinguish between the design hypothesis (D), the chance hypothesis (C) and the unknown natural process hypothesis (N). Call our observations O. I have claimed that there are two reasonable positions that may be taken (this may have been confusing earlier, for which I apologise). The first is that it is indeterminate what the relations are between p(O | C), p(O | D) and p (O | N) and so we should suspend judgement on where the evidence leads. The second is that p(O | C) < p(O | D) and p(O | C) < p (O | N) but it is indeterminate what the relations are between p(O | D) and p(O | N), and so we should believe that the observations are due to either design or some unknown natural process but should suspend judgement on which.

Indeed. However, consider that this same narrow band signal was received with mathematical tag added such as the first 50 terms of pi or the Fibonacci series. Such a tag, as Shostak pointed out, would be very hard even for a black hole to explain. Therefore, it would be much easier to support the ID-only hypothesis in such a situation.

This brings in quite different and complex considerations, as I am sure you are aware. One thing I have liked about your argument so far is that it hasn’t depended on relying on this sort of information transmission. I would prefer to leave discussion of the case where there is information until we have sorted out where we stand on the narrow-band case, if you do not mind.

This assumption is irrelevant to the question of ID because the actual origin of the ETI need not be known before the ETI can be detected as intelligent. The ETI could have been eternal for all we know yet we could still detect certain types of artifacts of the ETI as true artifacts.

You are still using the language of detection, as if there is some automatic inference that takes us from a certain kind of evidence straight to outright belief, without the interference of background knowledge. I have been emphasising that this is not the case, and that the way in which scientists justify SETI in part depends on their belief that life could arise elsewhere, which in turn reflects their belief that it arose evolutionarily. There is a reason they aim their detection equipment at planetary systems rather than at random. Do you think this is arbitrary?

Yes, very funny…Humor aside, however, this is an entirely different argument. You are now arguing that a mindless natural mechanism is actually known that can produced all of the features of bio-systems, thereby falsifying the ID-only hypothesis. This is Sober’s real basis for distinguishing between SETI scientists and IDists, like myself, who propose the obvious need for ID behind certain biological features.

No, Sober agrees with me that design wouldn’t be supported even if we weren’t aware of the evolutionary hypothesis.

What is interesting here is that this argument is quite different from your original statement were you said that you were “ambivalent” about the Darwinian mechanism of RM/NS. Yet, is this not the very same mechanism promoted by your referenced author above?

This legendary textbook does of course describe natural selection, along other mechanisms of evolutionary change. You have been misunderstanding my remark on natural selection, however. What I said was: “Regarding the relative importance of natural selection, I remain neutral”. I said this to try to sideline the peripheral issue of the extent to which the biological world consists of adaptations, like I have tried to sideline the peripheral issue of whether punctuated equilibrium is correct or not in my exchanges with our friend Roger.

We could get even more obvious here, for illustration, and imagine a situation where our Mars rover came across a polished granite monolith with the following inscription, “Welcome Earth Rover.We have been expecting you. After many years of secretly studying your human culture and various civilizations, we concluded that your technology would allow you to make it to Mars around this time and have decided to leave you a little message of welcome to the Universe.”Would such a message, written in English as well as many of the other main languages of the peoples of the Earth, be enough to convince you of deliberate intelligent non-human activity?

Yes—because in this case I think we can have independent evidence to believe that all natural processes, even presently unconceived ones, would confer a low probability on this event.

The biological case is supposed to be even more conclusive, far more so on a similar basis of logic, than is the SETI argument or those arguments I’ve presented above for polished granite cubes with geometric etchings or mathematical sequences carved on each face.

Do you know some biological facts I don’t? Take me to the Bible verses etched in DNA!

Given enough background information, this statement is not true. SETI scientists can immediately infer design, very rationally, when/if they discover the types of radio signals they are looking for without further ado.This is not to say that further ado is not helpful.It is.Science always looks for additional information regardless of how solidly a hypothesis/theory seems to be confirmed by past experience.However, given a great deal of past experience that is already in hand, the SETI hypothesis seems to be, even in your own estimation (and even more so in Sober’s estimation), built on solid evidence and rational thinking.

So tell me, what are the grounds for believing that p(O | D) > p (O | N) in this case? You’ve said a number of times that it is obvious. It is far from obvious to me.

Every time a design hypothesis has been proposed in biology or physics for a phenomenon outside the then-current scope of explanation, which has later been subject to independent confirmation, it has been false. See for example Kepler on lunar craters, Newton on planetary orbits, Arbuthnot on male to female birth ratios, Paley on the eye, Behe on everything, and (coming to a science journal near you) EducateTruthers on protein handedness.

Without any idea as to the probable limits of what non-deliberate natural forces are likely able to achieve you would not be able to determine that the watch is or is not a likely artifact vs. the amorphous stones scattered round about it. In order to make any rational judgment regarding the artificial nature of the watch you must have prior experience with both the potential and limits of at least human-level ID as well as with non-deliberate natural processes.

This is a good point. That is, for outright belief in the watch case we need background evidence that the overall probability of natural processes producing these things is low. But we have such evidence for watches produced on earth. I claim we do not have this evidence for the SETI case, since we are dealing with remote and exotic physical situations.

So, where do we disagree? You think SETI signal detection would license outright belief, I disagree. You think SETI signal detection would differentially support design over unknown natural hypotheses in advance of further investigation, I disagree. You think that if we didn’t know anything about evolution then the biology case would obviously be exactly analogous to the SETI case, I disagree. And the most important disagreement: you think in our current situation, knowing all of modern evolutionary theory, the design hypothesis stands to biology as it stands to SETI signals. I couldn’t disagree more, and it is astounding to me that you are willing to believe this based on one argument you have formulated in a field on which you are no expert, without having it vetted by experts. By the way, did I tell you about the time I squared the circle?

Recent Comments by Brad

GC Votes to Revise SDA Fundamental #6 on Creation
@Roger Seheult:

Please show me where Sean Pitman claimed to be an “expert” in proteins.

Sean believes he has an argument concerning protein evolution that if correct would show that most of evolutionary theory is false. I have observed Sean to be otherwise rational, and so infer that he believes he has expert knowledge on this matter. But perhaps we can ask him. Sean, do you take yourself to have expert knowledge on protein evolution?

Sean has been working in this topic for literally years. It’s his Hobby. He’s talked about and read about it and is even writing a book about it. To put it another way, he knows more about this topic then I know about Nantes.

Thanks for telling me you meant literal years. I’m not sure who the joke’s on, but this did make me laugh. If you like, you can add to my mathematical example that I have been working on topology for years, it is my hobby, I’ve talked about and read about it and am writing a book about it. Also, I tell you that I know more about topology than you know about Nantes. Should you believe I have disproved the Poincaré conjecture?

GC Votes to Revise SDA Fundamental #6 on Creation
@Roger Seheult:

Brad said:“But in Sean’s case I have every reason to think he is self deluded, since: i) the subject is (much, much) more difficult to master without training; and ii) I have the opinion of an expert…” So if you’re too good at what you don’t have a degree in then you must be self deluded? And obviously if someone can find someone who disagress with you somewhere in this vast WWW then that seals it. Am I getting it now Brad?

No, you are not getting it. Maybe an example will help.

Suppose I tell you that I have discovered a flaw in Grigori Perelman’s proof of the Poincaré conjecture. I have studied mathematics as part of my graduate training, but it is not my central area of expertise, and I have never published anything in a mathematics journal. I publish my attempted explanation of the error on my webpage. The one expert in topology who has read my webpage tells me that I demonstrate a profound ignorance of topology and need to get out and talk to some real mathematicians. Should you believe that I have discovered a flaw?

GC Votes to Revise SDA Fundamental #6 on Creation
@Roger Seheult:

PS. For what it is worth, here is the most recent survey on the level of scientific consensus on anthropogenic climate change.

William R. L. Anderegg, James W. Prall, Jacob Harold, and Stephen H. Schneider, “Expert credibility in climate change”, in PNAS, Vol. 107, No. 27, 6 July 2010, pp. 12107-12109. URI:


Although preliminary estimates from published literature and expert surveys suggest striking agreement among climate scientists on the tenets of anthropogenic climate change (ACC), the American public expresses substantial doubt about both the anthropogenic cause and the level of scientific agreement underpinning ACC. A broad analysis of the climate scientist community itself, the distribution of credibility of dissenting researchers relative to agreeing researchers, and the level of agreement among top climate experts has not been conducted and would inform future ACC discussions. Here, we use an extensive dataset of 1,372 climate researchers and their publication and citation data to show that (i) 97–98% of the climate researchers most actively publishing in the field surveyed here support the tenets of ACC outlined by the Intergovernmental Panel on Climate Change, and (ii) the relative climate expertise and scientific prominence of the researchers unconvinced of ACC are substantially below that of the convinced researchers.

GC Votes to Revise SDA Fundamental #6 on Creation
@Roger Seheult:

Come on! So when evolutionary scientists slash literally millions of years of evolution from there uniformatarian Darwinian influenced time lines because of punctuated equilibrium theory – its only a “dispute” but when another scientist wishes to doubt that a T-rex lived 65 million years ago because of RED BLOOD CELLS found intact in the fossilized bone, well then… they must be crazy, or “infants, monkeys, or others not in a position to properly assess it.” Of course it’s always easy to claim that there must be contamination, er… from somewhere…er….each time they do the experiment. Yes…Amazing.

You’re all over the place here. Slow down and read what I have said more carefully. I simply pointed out that the predictions I had in mind are independent of the issue of punctuated equilibrium (which concerns the tempo of evolutionary change, not the common ancestry hypothesis or the hypothesis of that natural selection is a mode of speciation). I have not made any remarks about red blood cells and T-Rex. Finally, my remark about infants and monkeys concerned my comments on expertise, which are irrelevant to this part of the conversation, so far as I can tell. If you want to rant and rave, go ahead—just don’t pretend you’re replying to any argument I’ve made.

Bottom line, we have every right in the Seventh Day Adventist Church to refine what we believe based on Biblical Truth and attacks made on her just as much as you do to go on believing in some evolution theory (which ever one you choose to believe in).

Did I deny it?

GC Votes to Revise SDA Fundamental #6 on Creation
@Roger Seheult:

Brad, may I submit that you are no more an expert here than anyone else.

This is amusing. I never mentioned my own expertise until you asked me about it, and it was clear from my exchange with Sean that I do not consider myself an expert on the topic I was urging him to publish on. If I did, I would have engaged the arguments myself rather than referred them to others. Moreover, part of the point of remaining anonymous is to level the playing field and not rely on my own credentials. Some other people on here, you will have noticed, seem quite proud of the letters that can be placed after their names.

You have been purposely guarded about your identity which tells me volumes.

What does it tell you?

If you do not have the courage to stand up for your beliefs in a public forum then your beliefs are not worth believing and your posts will be viewed as such.

I choose to remain anonymous on this particular forum for personal reasons that I do not need to reveal to you. An argument is an argument is an argument—if you want to ignore mine, then so be it. Let me quote Sean from another thread, who said it quite nicely (please ignore his quotation marks around “rational”):

Your goal should be to speak to those who have not yet thought about these issues and have not made up their minds to the point of being so biased that they cannot see “rational” arguments for what they really are for themselves (without seeing if others of high academic rank are convinced first). There are many “lurkers” who frequent this forum, usually over 2000 per day, – not all of whom are completely blinded by their religious background and many of whom, I know for a fact, would preferentially favor your perspective if they were ever exposed to it.

You then write:

I have a post doc degree in medicine with a subspeciality in internal medicine and a sub-subspeciality in pulmonary diseases totaling 7 years after my doctorate. I published a article in an European peer reviewed journal on Hypoxic pulmonary vasoconstriction and it’s response to endothelin recpetor antagonists at high altitude from a randomized controlled study that we performed. Based on your criteria I would be an expert in this field. While the process of publishing a scientific paper is eye-opening and lets you see the bias that exists especially when peer review journals distribute submissions to the same experts in the field that you are competing with, I would say that I know more about the City of Nantes, France where my ancestors were from and the many cathedrales and churches that they built in the 1700s and 1800s than I do about the different meachanisms of hypoxia at altitude or inert gas technique employed to determine such stuff (others know far more) even though I have no degree in history nor degree in architecture. Bottom line – get new criteria

I fear you have lost the thread of the conversation. The question is not whether it is possible to have expertise in a field in which you have not studied or published. Of course it is. The question concerns the conditions under which a non-expert bystander can determine the expertise of someone who claims to be one. This is the question my criteria are designed to address. If you tell me a lot of things about Nantes, I have no way of knowing your accuracy if: i) I can not confirm it myself; and ii) no-one I independently know to have such expertise can confirm it; and iii) I have no prima facie reason to take your word for it. Likewise for Sean on protein evolution. Now of course, in the Nantes case I don’t have any reason to think you are lying or self-deluded, so I am prepared to just take your word for it. But in Sean’s case I have every reason to think he is self deluded, since: i) the subject is (much, much) more difficult to master without training; and ii) I have the opinion of an expert who after reading his work wrote:

I am probably going to write far too much but if you want the conclusion, it is that Sean Pitman is completely and utterly wrong in everything he says in his comments and displays a great ignorance of proteins and their structure and function.


I hope the above short essay on protein structure and function is useful even to Sean Pitman who needs to stop being obsessed with computer-based numerology and do some reading and talk to some practical protein scientists.

From David Dryden of the University of Edinburgh. See: