Let me restate my position.  One recognizes "design" not by
identifying "irreducibly complexity" or "specified complexity," but by
recognizing similarities to things known to be designed, in structure,
composition and methods of construction, and purpose.  To take Paley's
famous watch example, he could tell that the watch has gears and
springs, because he recognized them as members of known classes of
manufactured items.  He recognized that it told time, because he
already had the concept of telling time.  Whether he understood either
how these gears and springs told time, or how they were manufactured,
is another question.  At the low extreme of complexity, one recognizes
the crudest stone tools of early hominids because they show the sorts
of chips we recognize as the results of human manipulation.  Design is
recognized by analogy with the work of known and observed designers.

This applies, of course, to SETI as well -- the search for
extraterrestrial intelligence depends crucially on the assumption that
ETIs would design in similar ways and for similar purposes to those of
humans.  One could, I suppose, hypothesize a Designer of radically
different capabilities, methods, and goals; if one had a sufficiently
detailed hypothesis, one could predict what sort of results one should
expect of that design.  That is, one must *recognize* the
specifications of the complexity.  IDers argue, on the one hand, that
living things are obviously designed for their functions.  But when
examples of seemingly bad or just eccentric design (what Designer
would use one basic wing design for all birds, flying or nonflying,
and another for all bats, of all sizes?) are adduced, they retort that
we can't know the purposes of the Designer.  Well, if we can't know
them, we can't very well marvel at how wonderfully the design
accomplishes them, can we?

Now, ID proponents argue that SCI can be recognized because no natural
mechanism can produce it, and intelligence can, even if we can't be
sure what exactly the specification is.  But even to the extent that
currently known regularities of nature, operating alone or in
combination in currently known ways, can't explain a phenomenon, all
that shows is that some currently unknown mechanism (whether employed
by an intelligent Agent, or purely nonteleological) produced it.
Without an exhaustive knowledge of all nonteleological regularities of
nature, and all their possible combinations, we can't rule out the
possibility of unknown, natural, unintelligent causes.  Nor, of
course, can we rule out intelligent causes of sorts (e.g.
intelligences no more interested in our morals, welfare, or worship
than we would be in that of bacteria in a petri dish) that would not
greatly interest most ID supporters.

*shrug*  Take the IDers' own word for it.  In Phillip Johnson's
_Darwin on Trial_ , Behe's _Darwin's Black Box_, and quite a few other
books, the author deals with some variant of the "panda's thumb"
argument that the sort of design we see in living things is *not* the
sort of design we would expect from any observed sort of intelligent
designer.  The response is invariably that this is a theological, not
scientific, position -- that we aren't entitled to any assumptions
about how the Designer would work.  But if we aren't entitled to any
assumptions about how the Designer would work, we surely can't make
any predictions about what design will and will not look like.
Therefore we can't tell design from the results of unknown, but
unintelligent, causes -- or, indeed, from the results of known
unintelligent causes (maybe the Designer crafts each snowflake
individually and intelligently -- how would we ever know otherwise?).
Very unlike that, indeed.  ID does not seize on newly identified
mechanisms with which to explain this or that aspect of design.  Its
flaws do not include finding one purpose or technique for design, and
using it to explain the bacterial flagellum, while seizing on a
different sort of design for a completely different purpose to explain
the immune system.  It does not seek mechanisms or explanations for
anything at all, or make predictions detailed enough that it needs to
rescue them with _ad hoc_ explanations.  Rather, it simply argues that
this, and that, and some other thing can't be explained in perfect
detail by current models, so "theDesignerdidit" (in some unspecified
manner, at some unspecified time, for some unspecified purpose) is
somehow a superior explanation.

-- Steven J.