Be sure to check out Goodness Over God, the counter-apologetics podcast hosted by myself and philosopher Michael Long!

Tuesday, May 1, 2012

More on Rasmussen's New Argument

Recall that Joshua Rasmussen in his "New Argument for a Necessary Being" (2011), argues that
(1) Normally, for any intrinsic property p that (i) can begin to be exemplified and (ii) can be exemplified by something that has a cause, there can be a cause of p's beginning to be exemplified. (p1)
When I expressed concerns with his published defense of (1), he (first privately, and later publicly) offered the following supplement (my paraphrase): Consider mundane intrinsic properties of the form being an apple, or being aluminum, etc., which can begin to be exemplified. Clearly such properties possibly have a cause for their exemplification, and so inductively we infer (1).

On the face of it, the argument appears structurally satisfactory. In order to affirm a statement of the form (*) "normally, for all x, if x is F then x is G," he appeals to the fact that in all observed cases, of which there are many, we find that if x is F then x is G. Meanwhile, we have no independent reason to suppose that (*) is false. Divorced of content, then, the inference to (*) appears appropriate. So what has gone wrong?

The answer, I think, lies in the fact that, at least in my own intuitive judgment, his argument fails to cut up the world in a natural way. To use Rasmussen's term, it seems gerrymandered in the sense that his choices for properties F and G are too unusual.

In contrast, the way I interpret the observations to which he appeals is that physical objects like chairs, apples, computers, etc., have physical causes which bring them into existence. Indeed for any such type, we can imagine a first instance of this type of object, and these will, like their successor instances, have physical causes bring about their existence. This is the backdrop against which Rasmussen wants to infer (1). According to him, a property like being of type A is intrinsic, and a cause of the existence of a first object of type A, at least in each of these cases which we have considered, also doubles as a cause of the property being of type A beginning to be exemplified. In this way, he extracts from a natural telling of the story the artificial elements which he needs to construct his inductive argument. In so doing, I find that he leaves behind any persuasive force the argument might otherwise have had, at least for me.

This all reminds me of Goodman's paradox. To recap the famous example, Goodman proposes that all emeralds observed up to time t happen to be green, but that all other emeralds (notably those observed after time t) happen to be blue. This induces the predicate "x is grue," where grue denotes the property of either being green and being observed before t, or else being blue and not being observed before t. For a suitable choice of t, we find that all observed (so far) emeralds are grue. Inductively we infer that all emeralds are grue, which is to say that we should expect that any emerald discovered after t will be blue.

Clearly, though, this is an unnatural way to cut up the world. Though we might not be able to explain why apart from appealing to intuition, emeralds ought not be classed as grue or not grue, but rather as green or not green. Once we leave behind the natural generalization of our observations, the inductive argument loses all its force, even though it has appropriate formal structure. In the same way, Rasmussen's argument does not seem to me the least bit compelling, despite its otherwise satisfactory structure.

Now, he might object that I have not properly characterized the situation in question, that is, the situation of having intrinsic properties of the form being of type A caused to begin to be exemplified. In particular, I have assumed that if there is a first object of type A, and that object has a cause B of its existence, then B caused the property being of type A to begin to be exemplified. Perhaps this is not always the case, though. Consider the property of being a rock with seven points, i.e. a rock having a mostly round shape interrupted by seven sharp points. The first object bearing this property undoubtedly had a cause of its existence, but the property itself was not caused to begin to be exemplified in the sense that the cause of the existence of the rock was perhaps causally indifferent to whether or not it should have seven points. In that case we would have to re-cast the general story, but not in any way that I find would improve our confidence in Rasmussen's inference. For the only means I have of interpreting such properties to have a cause of their beginning of exemplification is to imagine them to have resulted from a physical process which tends to produce new properties of that sort. So for instance a cause of the property being an apple beginning to exemplified might consist in the running of a physical system (natural selection) which tends to produce new types of things.

This is a more complicated story, but the result is the same: it seems unnatural and counter-intuitive to make the leap from physical systems generating diversity to the conclusion that the beginning of the exemplification of an abstract property like being a contingent concrete particular possibly has some kind of inscrutable cause. Structuring it as an inductive argument doesn't improve our situation any more than it did for blue emeralds.

However the above objection is based primarily on intuition. I have made some pretty bold assertions as to what is natural versus what is not, and although I'm confident in these assertions, I don't have much support for them other than intuition. In effect, I'm gambling that most others share my intuition in this matter, and are able to see Rasmussen's argument as I do---gerrymandered to the point of being too unnatural to have any persuasive value. Perhaps I am mistaken about that, however; maybe some folks have differing intuitions which favor Rasmussen's argument. If so, I still have to go with my own best judgment, and this leads me to reject the inference as unwarranted.

+ + +

Note: I wrote the above over a month ago, before speaking with Rasmussen about it directly on the Goodness Over God podcast last March. Unfortunately the conversation splintered a bit with other topics, and we didn't spend a lot of time discussing this particular objection. Perhaps we can hash it out another time. Meanwhile, enjoy the post!

No comments: