> My acquaintance had no source for this, and a web search
> shows it quoted in a number of places - never with context,
> but when it is sourced it is given as "in a speech at
> Washington University in St. Louis", with a date of 1997.
> The closest thing I can find that can be confirmed is this
> from Dawkin's review of "Full House" at
... and thanks to Nomen Publicus for pointing me to this link:
.. which enabled me to discover that Washington University Archives
still had a tape of the talk, in their "Assembly Series Recordings
Collection." Furthermore, they very kindly digitized it and gave me
access, so I can now give the precise context of the quote, about half
an hour into the talk (entitled "Is Evolution Progressive?"):
"The vertebrate eye must have evolved progressively for the following
reason: ancient ancestors, the first ancestor to have any sort of eye at
all, must have had a very, very, simple eye containing only a few
features good for seeing. We don’t need evidence for this. It has to be
true. Because the alternative, an initially complex eye, starting off
complicated, starting off with many features good for seeing, all there
at once, pitches us right back to Hoyle country, and the sheer
unscalable cliff of Mount Improbable."
In other words, it does not refer to evolution generally, but to a very
specific point in vertebrate evolution. Nor is it a resort to faith;
it's a logical deduction from the nature of evolution (granting that
evolution is backed by a great deal of evidence, which he's entitled to
assume in a talk on what evolution is like, as opposed to why it's true).
Interestingly the talk shares a chunk of material with the review I
mentioned in my original post. But there, the sentence looks like this:
"We don’t need evidence for this (although it is nice that it is there)."
Not surprising, since Dawkins wrote a chapter of Climbing Mount
Improbable on the evolution of the eye, he knows a bit more about the
evidence than the arm-chair deduction above.
... in case any of you were wondering.