Refuting Michael Behe's "Irreducible Complexity" with Roman Arches

Michael Behe claims that many biological systems could not have evolved because they are irreducibly complex:

By irreducibly complex I mean a single system composed of several well-matched, interacting parts that contribute to the basic function, wherein the removal of any one of the parts causes the system to effectively cease functioning. Darwin's Black Box, p.39
Wooden Roman Arch
This wooden arch comes from Edmund Scientific.
Roman Arch in downtown Vienna
This stone arch is in the Roman
ruins in front of Franz Joseph
Palace in Austria, Vienna.
(click link for full view.)
By this definition, a Roman Arch is impossible to build, because it would always fall down before being completed.

Yet it is clear that such arches could form in nature without an intelligent agent. For example, a landslide could put a pile of rocks onto a stream bed. The stream may then erode the pile to leave an arch.

Natural Stone Arch
There are spectacular large arches carved out of stone
in the South west of the United States at Arches National Park.
(google: Arches National Park)

Likewise, "Biochemical scaffolding" could have once existed that is now gone. One might imagine that the scaffolding does not have an advantage and therefore the scaffolding must be externally directed, but that would miss the point. In a biological system many variations can appear that have no significance, such as a duplicate gene, but sometimes the variations become advantageous. Eventually they may become required and at this point they would no longer be "scaffolding".

The Ev paper shows that the genomes of the evolved creatures have exactly the properties required by Behe to be "irreducibly complex" and yet they evolve from scratch. This demolishes Behe's "irreducible complexity" argument.

The kind of arch does not matter! Some people think that just because Roman Archs are made by a people, that this disrupts the argument. However, a simple counter example is to notice that there are arches in nature too, such as found at the Arches National Park in Moab, Utah.

Scaffolds do not imply goals! Is the sea a goal of a river? No. To form an arch is a straightforward physical process in which layers of material are formed, raised up, and then undercut from the side to remove the "scaffold". Alternatively, a U shaped rock could be turned up by, for example a landslide. There is no outside goal being met, and if the system fails to make an arch there is no consequence.


I sent an email to Behe to ask him about this issue:

From toms Thu Aug  5 01:38:27 1999
Subject: question about irreducible complexity
Content-Length: 1757

To: Michael Behe

Dear Mike:

I just read your interesting and well written book "Darwin's Black Box".
Thank you for sending me a copy.

I have a question about your thesis and I would like to post our
correspondence on my web site.  May I have permission to do so?

The question is whether you would consider a genetic control system to be
irreducibly complex.  Your definition is:

  By irreducibly complex I mean a single system composed of several
  well-matched, interacting parts that contribute to the basic function,
  wherein the removal of any one of the parts causes the system to
  effectively cease functioning. (p.39)

In each genetic control system there is a recognizer (protein or RNA/protein
such as the ribosome or spliceosome) that recognizes and binds to specific
places on DNA or RNA, and in doing so controls some function such as the
expression of a set of genes.  The recognizer is well-matched to the binding
sites, and they interact to perform the basic function.  Both contribute to
the control.  If the recognizer didn't exist or was not functional, the
control would not work.  If the binding sites had the wrong sequence or were
in the wrong places, they would not work.  As I understand your concept, this
meets your criteria for an irreducibly complex system, since lack of any
component leads to failure of the system.

Since each depends on the existence of the other for function, isn't the
entire system of recognizer and binding sites "irreducibly complex"?

Thank you,


  Dr. Thomas D. Schneider
  National Cancer Institute
  Laboratory of Experimental and Computational Biology
  Frederick, Maryland  21702-1201
  permanent email:

Michael Behe refused to allow me to post his response. In December of 1999 I sent him a Roman Arch.

2006 May 18: I have written a Summary of Behe's response.


color bar Small icon for Theory of Molecular Machines: physics,
chemistry, biology, molecular biology, evolutionary theory,
genetic engineering, sequence logos, information theory,
electrical engineering, thermodynamics, statistical
mechanics, hypersphere packing, gumball machines, Maxwell's
Daemon, limits of computers

Schneider Lab

origin: 1999 Aug 05
updated: 2013 May 01

color bar

National Cancer Institute    National Institutes of Health    Health and Human Services    USA Gov - Official Web Portal    Viewing Files    Accessibility