At William James exhibit in Houghton Library (Harvard), there was a test from 'Philosophy 9' (academic year 1904-1905), one of the questions read:
How far can any state of facts experienced be taken either as proving "design," or as being incompatible therewith?....
This immediately grabbed my attention, as the context was presumably the so-called 'problem of design' that faces evolutionary theorists... and Nick Thompson and I have been wrestling with how to get 'design' back into evolutionary theory (he's been wrestling with it for 40 years, me only for the last 10)... and as I have been trying to convince Nick that his approach derives strongly out of James's late work.
While the test was staring up at me, I realized that nothing I had read of James dealt with the issue explicitly. How was James trying to deal with the problem of design? What had he told his students? What might his students have replied? This lead to some digging.
I soon found out that parts of James Philosophy 9 lecture notes are published, though for the following year. Here are the relevant bits:
a) objective. "Design."*
b) Subjective: "Personal experience"
Criticize 2a. Any state of things may be designed. Question of "what is character of world?"
*Any kind of world after the fact may have been designed—by that kind of designer namely. It is just like the lion's den Absolute. The really important thing is the pragmatic thing, what is the kind of world before us? If its character turns out empirically "divine," it makes no difference whether it have a designer or not. We can lead an adequate life in its presence. Cf. Perry's A to P. Chap. IIII
Also, on the facing page, James wrote--
Do we live in presence of a world that is divine in Character? If we do, the question of its having a 'designer' is a merely formal one. The essential practical thing is the Character. That lets us be religious. Obviously its empirical character is the stumbling block, *to remove for which all the theologies are bro't in.
Also relevant is some material from James's 'Philosophy D: General Problems of Philosophy'. In his '06–'07 notes, he has a section on 'Critique of the Teleological Argument':
Observe that the point here is not to prove that, some end being assumed, the physical means that led to it were well adapted. Of course they were, or they would not have led to it. The point is, rather, to decide what or which end is the really purposed end. What has nature really aimed at with all her machinery? We can only guess at that by ascertaining what she has produced. Then, no matter what she may have produced, the means were fitted to that production; and the argument from design would apply, whether the product were good or bad. The recent Mont Pelée eruption required all previous history to produce that exact combination of ruined houses, human and animal corpses, sunken ships, ashes, etc., in just that configuration of positions. If God aimed at just that result, the means by which it was led up to through the centuries were exquisitely designed. And so of any state of things whatever, either in Nature or in History, which we find actually realized: the parts of it must form some kind of a resultant, and must appear 'fitted' to each other to that effect. The cosmic machine may have been designed for just that result.
Our inquiry thus takes the same turn here which it took on pp. 16–17 [ed., 396.27–397.39] of the syllabus. What we are practically interested in is not whether design exists in the world, but what the design is. What is the world's character? its outcome?
The reason why the argument from design has been so popular is that certain parts of the world are so good in our human sight, and the means of their production so apt, that we infer a humane and intelligent contriver for them. Then, generalizing our conclusion, we are reassured as to the total process, even if we can't well understand it. In some vague way, this must be God's world, we say. His proved presence in any part of it is a guarantee of its whole character.