Comments on Ken Miller's Reply to My Essays

Michael J. Behe

Discovery Institute
January 8, 2001

Kenneth Miller, Brown University Professor of Biology and author of Finding Darwin's God, has posted a response to my essays: Overall I'm satisfied with his reply because, although he continues to defend his position, from the substance of his writing I think it should be plain to most open-minded readers that he is struggling to fend off examples that weigh heavily against Darwinism. So, for the most part, I am content to let the exchange end here. I simply urge all who are interested to read my essays as well as his response and come to their own conclusions. I do, however, want to make a few additional comments, in just two areas, to keep the issues in focus.

On the "acid test"

Professor Miller claims that the careful work of University of Rochester biologist Professor Barry Hall is an experimental demonstration of the ability of Darwinian evolution to produce an irreducibly complex biochemical system. (Barry Hall himself never made such a claim.) I disagree. The fact that the artificial chemical inducer IPTG was added to the lactose-utilizing system effectively mitigated its irreducibility, turning the system into one that could be improved a step at a time. In his recent essay Miller wrote:

"Does Barry Hall's ebg system fit the definition of irreducible complexity? Absolutely. The three parts of the evolved system are: (1) A lactose-sensitive ebg repressor protein that controls expression of the galactosidase enzyme; (2) The ebg galactosidase enzyme; (3) The enzyme reaction that induces the lac permease. Unless all three are in place, the system does not function, which is, of course, the key element of an irreducibly complex system."

Miller's claim is incorrect because in the presence of IPTG the three features he lists are not all needed. In the presence of IPTG, the "enzyme reaction that induces the lac permease" is not required because IPTG itself induces the lac permease. Thus in the presence of IPTG the system is not irreducibly complex. And, as I wrote in my original essay, Barry Hall clearly noted that in the absence of IPTG--when the system actually is irreducibly complex--no viable mutants have been found in his 25 years of investigation.

The inclusion of IPTG was the result of the decision of an intelligent agent (Barry Hall) to deliberately alleviate the irreducibility of the system. In the absence of that intelligent action, Darwinian processes alone were ineffective. That is exactly what intelligent design theorists would expect.

Miller also writes, "the ebg gene is actually only 34% homologous to the gene whose activity it replaces (meaning that about 2/3 of the protein is quite different from the galactosidase gene whose function it replaces)". Yet he knows as well as I do that 34% general sequence homology makes it virtually certain that the three-dimensional structures of the two enzymes are essentially identical. And since the active sites (the business end) of the enzymes are much more similar (they are identical in 13 of 15 residues), the ebg enzyme is pretty much a spare copy of the lac enzyme. Thus it seems to me that the taking over of lac galactosidase function by ebg hardly even rises to the level of microevolution.

What is actually surprising--even to a design theorist such as myself--is Barry Hall's finding that no enzyme other than ebg could fill in for the missing lac galactosidase. I would have expected otherwise. Perhaps even changes we would consider to be "microevolution" are often times beyond the reach of Darwinian processes. Perhaps even I give natural selection too much credit.

On the mousetrap example

Professor Miller writes, "MacDonald's drawings address Behe's contention that 'all components have to be in place before any mice are caught.' They don't, of course, because there is more than one way to construct a mousetrap from mechanical parts."

But they do--all of the parts of my mousetrap do indeed have to be in place for it to function. I noted clearly in my book, and in my mousetrap essay on this website, that I am very much aware "there is more than one way to construct a mousetrap from mechanical parts." Nonetheless, if you just take away pieces from the trap I pictured--and don't manipulate it further--the trap doesn't work. What Miller actually means is that if you take away some components and then go on to, say, twist a couple of metal pieces in just the right way and add a few staples in the correct positions, you can construct a new kind of working trap, which may superficially resemble the starting trap. That, however, is intelligent design. Neither Miller nor anyone else has shown that the mousetrap I pictured in my book can be constructed by a series of small changes, one at a time, as Darwinian evolution would have to do. The important take-home lesson is that even things that look superficially similar, such as the series of traps Miller showed, may not be able to be transformed into each other through a Darwinian process.

Towards the end of his essay Miller writes, "If simpler versions of this mechanical device [the mousetrap--mjb] can be shown to work, then simpler versions of biochemical machines could work as well . . . and this means that complex biochemical machines could indeed have had functional precursors."

However, consider the general statement, "System A looks like system B but cannot be transformed into system B by a series of small, Darwinian-like steps." Miller explicitly agrees that is true for the mousetrap series.(1) Yet in the statement above he seems to think that the very fact it is true for mousetraps somehow shows it is not true for biochemical systems.

That reasoning isn't completely backwards, but it's pretty close. If the simpler mousetraps can't be transformed into the similar-looking-but-more-complex ones by something analogous to a Darwinian process--which, again, Miller freely admits in his essay--then the mousetrap series gives us no reason to think that postulated simpler biochemical systems could have been transformed by natural selection into the complex systems we see today. On the contrary, the mousetrap series gives us a prima facie case to think they couldn't.


(1) In his essay Miller wrote the following:

"Behe argues that MacDonald's four simpler mousetraps do not present a good model of a 'Darwinian process.' Even the simplest mousetrap, Behe argues, requires 'the involvement of intelligence,' and the 'involvement of intelligence at any point in a scenario is fatal.'

"I agree. And if I or MacDonald or any one else had presented the simpler mousetraps as examples of an evolutionary transition, Behe would be right."