Creation Myths: A chat with Michael Behe on Irreducible Complexity

Posting this because I think many people here would be interested, and possible want to discuss after: Tomorrow (that’s Monday, 3/22/21) at 1pm Eastern (EDT, GMT-4), I’m hosting Michael Behe for a chat on irreducible complexity.

So…we’re gonna see how this goes. Would love to hear y’all’s thoughts after.


This is great. I’m glad he was willing to appear on your show.

Have you ever had him on the forum before?

Not on the forum, but I’ve done several events with him. We are on good terms now, but it was a bit touch and go when our review of Darwin Devolves came out…

1 Like

This should be fun.

This was pretty good particularly near the end.

I think you made some good points with your concluding questions Dan. Particularly the one about his “purposeful arrangement of parts” statement which he basically revealed with his answer, is totally subjective, and that he started trying to run for the probability blather as soon as he was put on the spot to defend the subjectivity of the statement. Or “muh Rushmore”.

I also found it noteworthy that he says something like the CCC (Chloroquine Complexity Cluster) mutations is “the edge” of evolution. It evolved in a few decades. Think about that. Behe is saying another 40 million, or a billion years of Plasmodium evolution would fail to find mutant combinations resulting in anything more complex that that double mutation. That flat out doesn’t make logical sense.

Why can’t more complex results with probabilities a many times lower than the CCC mutations, evolve in an amount of time that take many times longer? Is it not obvious to everyone that we can know from reason alone that of course they can?

Behe is effectively saying that if it can’t evolve in a few decades or centuries, then it can’t evolve at all.


The applicability of the CCC mutations to evolutionary pathways in general is also questionable. There may be just a handful of mutations that would allow for chloroquine resistance, but how does that apply to more general adaptations in many other lineages?


What’s you summary @dsterncardinale ?

Behe was very cordial, but…it’s been 25 years since Black Box, and “irreducible complexity” is, still, not a well-fleshed-out idea.

I believe I asked for a quantifiable standard for “this thing is designed” five or six times, but did not get one, and just as many times he used a phrase like “in my mind” or equivalent.

The takeaway, I think, is that this is a completely subjective, and therefore unfalsifiable idea. Not that this is revelatory. I’m just honestly surprised in spite of myself that there isn’t an answer to “but how do you quantify that distinction?”

I do plan to have him on again, though probably with a different focus. It’d be pointless to rehash this same discussion.

And this is a very minor point, but jumped out because a few weeks ago, Dr. James Carter made exactly the same mistake: Where did creationists get the idea that the Cit+ phenotype was due to a promotor duplication rather than a CitT duplication? Because that seems to be A Thing.


Why do you focus on unfalsifibility? Science doesn’t work by falsifiability…

1 Like

It’s just one item on the “is this a robust, well-thought-out idea?” checklist. Is there a way to show the idea is wrong? Is there a way to quantify the concept? It’s been clear for…years, decades…that the answers are “no” for IC, and given however many opportunities there were to explain why that impression is wrong, it doesn’t seem like it is.


Although you did not press the point, “purposefully arranged parts” is the acme of Behe’s stance today. This amounts to little more than “it looks that way to Behe”. I don’t know about falsifiable, but this standard is scientifically bankrupt.


True enough, but we must distinguish science as practiced from the necessary requirements of hypotheses. What makes for a good scientific hypothesis? It is true that scientists do not, in practice, work by trying to falsify their hypotheses with experiments and observations, and then modify or come up with new ones in a sort of iterative process.

But a scientific hypothesis definitely should be capable of being tested. It must say something about the data such that the data can fail to fit the hypothesis.

There must be some way to compare the predictions of hypothesis to the data, and evaluate how well it conforms to the data by some sort of objectively quantifiable criterion.

It looks designed to me because look at Mt. Rushmore is no such thing.

1 Like

I think a better way to pose the question is:

How have you tried to falsify this hypothesis?

Science is characterized by diligent effort to falsify one’s hypotheses. That is true whether the hypothesis is currently unfalsified or not, whether it’s unfalsifiable or not.


In what way to do you mean this? If any possible observation would support a hypothesis then it isn’t a hypothesis. Some sort of negative control needs to be part of any scientific methodology, and that has everything to do with falsifiability. No hypothesis is complete without the null hypothesis.

1 Like

This is what I mean: Does Science Work by Falsifiability?

There is a distinction between saying:

  1. The hypothesis is not falsifiable so it is not scientific.

  2. The scientist is diligently seeking to falsify the hypothesis.

#1 is not a good description of how science works, and that is what Popper meant by falsifiability. In contrast, #2 could be a correct (partial) description of good science. As for negative controls, ID could choose to use negative controls if they wanted, e.g.



The fact that they don’t usually specify negative controls (in biology) isn’t intrinsic to their hypothesis, and just speaks to a deficiency in their methodology.

I would consider #2 a work in progress, and that is certainly science. Every hypothesis starts as an idea that is not immediately falsifiable.

“Deficiency in their methodology” is more generous than “a blindspot for their own biases”. I do understand the psycohological draw of possible positive results and why negative controls can sometimes be ignored.


I’m confused now, because didn’t you just state above that “Science doesn’t work by falsifiability…” ?

So what exactly are you saying? How does science work, in your view? What is science characterized by? How does science as practiced in reality by scientists, differ from how science is described or defined in textbooks or by philosophers of science?

It’s not clear which, if any of these questions you are addressing. Can you clarify?


“Unfalsifiable” is a adjective applied to hypothesis, an term that doesn’t make much sense when you carefully think about it. 1. It is very hard to distinguish unfalsifiable hypothesis from unfalsified hypothesis. 2. What is falsifiable changes over time as technology improves. 3. There are several large classes of true statements are technically unfalsifiable, but that does not render them false, and science actually depends on some of these statements.

In contrast “work to falsify” is a verb used to describe the effort of scientists, and this ends up being a better description. This relocates the discussion from the internal logic of hypotheses to the actions of the scientist. In this way it is more personal, but also more verifiable and objective than “unfalsifiability”.

I was just discussing that “unfalsifiability” is not a very good place to focus, and “working to falsify” is a better place. That’s it.

Nothing I’m saying here is new. As I understand it, this is old hat to philosophers of science. The first link I provided gave several references.

IMO the deficiency in their methodology is that they pathologically avoid even articulating a testable ID hypothesis, presumably to avoid testing one.