Revisiting Nelson’s Recycled Dysteleology Rant

As noted on Panda’s Thumb, Paul Nelson is giving yet another talk on the argument from dysteleology. This was one of the first things I ever heard Nelson talking about, way back in 1997 at the “Naturalism, Theism, and the Scientific Enterprise” conference.

And I had a response to Nelson at the time. I’ll append it here.

My response to Paul Nelson got sent to a couple of online fora, one of which Nelson participated in. I didn’t realize that Nelson actually managed to get his paper published in a philosophy journal, or I’d have submitted a version of the response there, too. Maybe I’ll have an opportunity shortly to respond to Nelson’s rehashed version of his presentation.

Posted to and the NTSE listserv.

A response to Paul Nelson’s paper from the NTSE conference, “Jettison the arguments, or the rule? The place of Darwinian theological themata in evolutionary reasoning.”

I agree that the argument from imperfection, as used by Gould, does not represent “scientific reasoning”. However, I don’t feel that Gould advanced these arguments as scientific reasoning.

Creation ex nihilo as a conjecture exists. Arguing against this conjecture will be an exercise in metaphysics. The argument from imperfection is such an exercise when it references postulated attributes of a deity.

Must all scientists recuse themselves from commentary upon an existing theological conjecture? I don’t think that either an unreserved “yes” or “no” is an appropriate response. Yes, scientists should refrain from making theological arguments which are deceptively packaged as if they were scientific. No, scientists are free to engage in theological conjecture that is labelled as such.

I also don’t agree with the necessity of Paul’s formulation of the premises. Let me try a different set:

P1. Organic design p does not accord with the known principles of good engineering design.

P2. Organic design p shows a function different from that known in similar structure p’ in other species.


C1. Organic design p is consistent with a historical process that adapts existing structures to new functions.

C2. From the attributes of organic design p, we can infer that a creator of organic design p is not an optimizing engineer.

Note that I have off-loaded the theological stuff into the conclusions, not the premises. Note that C1 is not dependent upon theological themata. If no further conclusions were derived, the argument as it is composed to that point is eminently scientific in character.

C2 makes a theological statement. At the NTSE conference, we heard from Michael Corey on how we can infer attributes of a creator from aspects of the creation. Either we junk the rule of inferring attributes of a creator from the creation, or we must admit these disparate inferences as starting from an equal premise.

I found Nelson’s defense of a role for a creator to be interesting. Basically, given the facts of the matter, one can conceive of attributes of a creator which remain consistent with the known data. Nelson’s example of Mill’s conception of a possible creator makes this point. Thus, Gould’s argument fails according to Nelson because Gould cannot exclude *every* creator concept. If this isn’t the classic “god of the gaps” type of apologetic, what is it? I can certainly see that Gould’s theological basis is not well developed, but it seems to me that it does the job of establishing C2. While there are some who will prefer a Millsian concept of a creator, I think that in the USA the most common conception (not necessarily the most astute conception) of a creator has been and is that of an optimizing engineer. Gnosticism or Zoroastrianism also accords with the facts as seen, yet I see no widespread move in theological circles to adopt either of those, nor any detectable political movement to have their origins accounts taught as if science in science classrooms.

Nelson’s discussion of perfection and imperfection raises many important points, but failed to address the main issue: the comparative method allows for identification of suboptimal design. The instantiation of components achieving particular functions gives us a basis for comparing the systems which provide for similar functionality. In those cases where such systems meet an engineering criterion of modularity, it is eminently reasonable to ask why module p was employed in one instance, but kludgy module p’ was employed in another. We don’t have to be able to identify the optimal in order to identify what is suboptimal.

Nelson argues from a mathematical viewpoint that we can’t use the suboptimality argument against a theological notion of a “reasonable” creator. He gives an equation for illustration:

ObD / OptD = DesShort

ObD is Observed Design
OptD is Optimal Design
and DesShort is the Design Shortfall

Because we can’t obtain OptD, we must forego use of this manner of argumentation, according to Nelson.

However, I can derive a different equation that demonstrates the possible utility of the comparative method:

ObD_a / ObD_b = DAR

ObD_a is the figure of merit for Observed Design “a”
ObD_b is the figure of merit for Observed Design “b”
and DAR is the Design Astuteness Ratio

This is a more appropriate metric to critique. There are no unknowns hiding here, and no necessity for finding or even worrying about an “optimal” design. That both ObD_a and ObD_b are or may be suboptimal does not detract from the utility of the comparison.

If the argument from imperfection is to be used as a scientific argument, then it must, as Nelson points out, drop the theology. This can be done without jettisoning the entire argument as unworkable. The result will be an argument for the historicity of adaptations without reference to the creation ex nihilo conjecture or such a creator.

Wesley R. Elsberry

Falconer. Interdisciplinary researcher: biology and computer science. Data scientist in real estate and econometrics. Blogger. Speaker. Photographer. Husband. Christian. Activist.

3 thoughts on “Revisiting Nelson’s Recycled Dysteleology Rant

  • 2008/05/23 at 10:38 am

    Looks solid to me.

    I thought one could push multiple designers (MDs) into the conclusions, but that will have to be a separate argument. Say, as several evolutionary mechanisms can be observably separated, the designs will most likely be made up by MDs with different characteristics. The problem being “functions” (so drift is out, for example, while I guess endosymbiosis would be in).

    If Nelson publishes, I assume you could include a reference to his earlier work in a published reply to take care of it this time around.

  • 2008/05/24 at 6:35 am

    Torbjörn Larsson, OM Says:

    “I thought one could push multiple designers (MDs) into the conclusions, but that will have to be a separate argument.”

    Long before I read RBH’s MD hypothesis, I thought (as personal speculation, with no pretense that the argument was scientific) that life looks “designed by committee.” I guess many others did the same.

    While it is important to address the flaws in design arguments, we should always keep in mind that they divert attention away from the whats, whens and hows of design actuation. Thats a useful tactic for anti-evolution activists because YEC and OEC (several versions, some of which include common descent) can’t all be right. And there’s no hope that rank and file anti-evolutionists will ever come to a consensus on one.

  • 2008/06/03 at 5:16 pm

    You mention that you have pushed all the theological stuff to the conclusions. But I would suggest that C2 requires the premise that there is a creator.

    As to Gould, I suggest that he was making a retort. Much like so called proofs of god’s existance are attempted, I suggest Gould did not intend his agrument to be solid. But to point out that you cannot make this a logical proof either way.

Comments are closed.