X is intelligently designed means ...

William A. Dembski (bill@desiderius.com)
Sat, 3 Apr 1999 10:48:16 -0600

Howard Van Till consistently urges that design theorists are being less
than ingenuous when they explain some object, event, or structure as the
result of intelligent design. Accordingly, Howard wants us first to state
precisely what we mean by "intelligent design."

As he writes:

>I presume that this term refers to some category of action performed by
>some type of agent. So, perhaps the answer to the question could be put in
>the followiing form:
>To be (or have been) 'intelligently designed' is to be (or have been)
>_______________________________ [decribe the category of action here]
>by ____________________________________ [identify type of agent here].

Let me fill in the blanks: To be (or have been) 'intelligently designed' is
to be (or have been) **designed** by an **intelligent agent**.

Now I'm sure this answer will not satisfy Howard. But it points up that
design theorists are unwilling to cash out intelligent agency in anything
other than intelligent agency. Intelligent agency is a sui generis notion,
and it does not admit reduction to some other category of causation. It
does not require a naturalistic causal story of how an intelligent agent
effected some designed system. Nor does it require some account of
"extra-natural assembly" for systems where no natural precursors are

To which Howard responds with his "gzorply muffnordled" objection:

>Suppose that a person were to say to you, "I have empirical evidence that
>>species X or biotic subsystem Y was gzorply muffnordled." I presume that
>your >first response would (rightly) be, "What does it mean to be (or have
>been) >'gzorply muffnordled'? Until you define that term, I have no idea
>how to >evaluate your claim."

Does Howard mean to suggest that "intelligent design" is as
incomprehensible a notion as "gzorply muffnordled"? "Gzorply muffnordled"
is a nonsense phrase with no history of usage and no semantic content.
"Intelligent design," on the other hand, has a perfectly well understood
common usage.

Okay, but this isn't enough on which to build a science. Granted. And
indeed, design theorists do not build their science simply by arbitrarily
assigning the phrase "intelligently designed" to some things and not to
others. The point is to take an intuitively well-understood
notion--intelligent design--and discipline its use so that it can be
assigned reliably to cases where intelligent causation is in evidence. The
means for doing this is to establish a criterion for intelligently caused
objects. Behe has his criterion of "irreducible complexity." Marcel
Schützenberger had one of "functional complexity." I've got my own of
"specified complexity" or "specified small probability." What's more, I
make an explicit argument for why specified small probability is a reliable
marker for intelligent causation (section 2.4, _The Design Inference_,
Cambridge, 1998). This is enough on which to get a science going.

But Howard still wants a definition. The very word "definition" means to
put limits on something. I've had discussions about defining "intelligent
design" in the past with Howard. What design theorists offer is an
intuitively well-understood notion of "intelligent design" together with
bridge concepts and criteria (like specified complexity) that render
intelligent design empirically detectable and scientifically tractable.
This strikes me as no different from Isaac Newton taking an intuitively
well-understood notion of "force" and giving it a mathematical formulation.

Howard and I have been through all this ground before. He wants a
definition of "intelligent design" before he is even willing to consider
the possibility of specified complexity as a criterion for detecting
design. What's more, he wants in the definition some acknowledgment of the
mode of assembly (and in the case of biological systems, an acknowledgment
of extra-natural assembly). But the mode of assembly is a separate question
from design. What is the "mode of assembly" by which I'm typing this text
at my computer? Yes, my fingers are typing away. But what is the mode by
which my mind is putting together sentences that convey meaning? The
cognitive neurosciences haven't a clue. For most instances of intelligent
design we don't have any account of the mode of assembly.

Definition in general is highly overrated. With only a finite number of
words in our vocabularies definitions must in the end always be circular.
Although I'm no disciple of Wittgenstein, he was correct to stress that we
understand the meaning of words by their use, and not by giving abstract
definitions. Howard's call to define "intelligent design" seems to me not a
plea for clarity, but an attempt to straitjacket intelligent design so that
it is sure to remain outside science and also to preserve his theology of a
fully-gifted creation.

Let me therefore turn the problem around. Howard, presumably you agree that
a pocket watch is designed? How would you fill in the blanks in your scheme
in the case of a pocket watch:

>To be (or have been) 'intelligently designed' is to be (or have been)
>_______________________________ [decribe the category of action here]
>by ____________________________________ [identify type of agent here].

How would you fill in the blanks in case the object in question is a long
sequence of prime numbers sent as a radio signal from an extraterrestrial
source? Presumably you agree that the cosmos itself is designed. How would
you fill in the blanks in case the object in question is the world itself?
Finally, what is the difference in your answers to these questions and in
the case where the object in question is one of Michael Behe's irreducibly
complexity biochemical systems?

In brief, please fill in Y and Z in

For X to be "intelligently designed" is to be Y'ed by a Z

where X is respectively: (1) a pocket watch; (2) an extraterrestrially
generated sequence of prime numbers sent as a radio transmission; (3) the
world; (4) an irreducibly complex biochemical system.

Bill Dembski