The latest issue of Philosophical Studies includes an article by Bernard Berofsky on global control (it also includes an article from fellow Gardener Kevin Timpe). It is good to see this subject, which fascinates me, getting more attention. But I found Berofsky's article to be frustrating reading. My own feeling is that other compatibilists like Fischer and Mele and (or semicompatibilists and agnostics, as they prefer to be called) have wrestled with incompatibilist (or non-realist) global control arguments more, and given them more credit. But others may, of course, decide for themselves and I would be very interested to know what other commentators think.
Let me offer some quick notes on the article:
1. Berofsky cites Mele's 1995 book and the arguments there. But he does not seem to mention Mele's latest book, and this strikes me as unfortunate, because that book includes a discussion of the Zygote Argument which (as the last Garden Reading Group showed) would exactly on point and fitting in Berofsky's own article. It might also have been good to mention other work on global control scenarios (by Michael McKenna, Gary Watson, etc.).
2. Berofsky calls Double a compatibilist and cites his 1991 book. But, although Double seems to have moved closed to compatibilism, after abandoning his more non-realist position in The Non-Reality of Free Will and taking up his subjectivist view in Metaphilosophy and Free Will, I'm not sure he counts as a compatibilist. I'm the first to admit, however, that I don't entirely understand his view.
3. Perhaps Berofsky clarifies this later in the article, in a way that I will understand after a second reading, but I cannot tell whether he is adopting a soft or hard compatibilist line. First he says:
"Have we not in effect answered Kane since, of course, there is that fundamental difference between a world governed by a controller and one that is not? To be sure, this response does require the repudiation of the simple idea that constraint and coercion are the only ways to undermine freedom. All right, let us then add CNC to the list. Even if we cannot say precisely what it is that bothers us about CNC, it is necessarily a form of agential causation and determination is not. Is that not enough of a defense for the compatibilist? I hope to show in this paper that it is. I hope to show that a distaste for global controllers is no reason to be turned off by determinism itself."
This would seem to make Berofsky a soft compatibilist, and the first prominent one, to my knowledge, in contemporary philosophy. But later he says:
"Why should we find the creation of a completely unfree 1 year old by a global controller whose omniscience permits her to see exactly where this child is headed more disturbing than the natural development in a deterministic world of a completely unfree 1 year old unless again the worry concerns agential causation? The issue between the combatants would then concern the possibility of freedom emerging from unfreedom and this battle has to be fought anyway. If compatibilists cannot show how freedom can arise out of unfreedom in a deterministic world, their position is doomed independently of the global control matter."
And this makes him sound like a hard compatibilist, as are Frankfurt and Fischer and Mele (according to his compatibilist view). Moreover, he doesn't seem to give enough credit to the global control argument. His last sentence there seems to amount to "the Zygote Argument is redundant because of the Basic Argument", and that may be true (although I think arguments like the Zygote and four-case argument have their own special role to play, due to their unique rhetorical force), but I don't think this establishes Berofsky's conclusion that, if there is symmetry between some globally-controlled agents and natural agents, we should infer freedom from the latter onto the former, and not un-freedom from the former onto the latter.
As Berofsky notes, and as I've noted before (in other words), the issue can be framed thus:
"Kane: There is a perfect parallel between a global control-world and a deterministic world. Since there is no freedom in a global control world, there is no freedom in a deterministic world.
Double: There is a perfect parallel between a global control-world and a deterministic world. Since there is freedom in a deterministic world, there is freedom in a global-control world."
Berofsky wants to defend Double against Kane here, but how? In the context of what Berofsky calls power-incompatibilism, I sympathize with Berofsky. But this leaves another worry, about source incompatibilism (and it is not at all clear to me that the two kinds of incompatibilism can be so separated), and Berofsky addresses it as follows:
"Second, as I said above, all these scary stories about global controllers molding little ones to their whims are really irrelevant. Our world already is like that! What matters is not the accepted fact that we are pawns at the start of our lives and for a short while thereafter. It is rather what we do later once we become rational, reflective, and informed. CC says that, even in a deterministic world, we can sometimes alter the input and change our lives in rational ways. So these worries about ‘our source’ are much overblown."
This argument seems to leave the reader with the same dilemma we started with: do we infer freedom from natural agents to globally controlled ones, or do we infer un-freedom from globally controlled agents onto natural ones? At best, I see no reason for the reader to adopt one over the other. Sure, Berofsky states that "[w]hat matters is not the accepted fact that we are pawns at the start of our lives and for a short while thereafter." But this is a conclusion without much of an argument. For one, the use of "for a short while thereafter" is unjustified. By hypothesis, a global controller might design an agent's entire life, from the cradle to the grave. It is not as if, as Berofsky hints, the cognitive capacities that young children develop somehow render them immune to such design or control. No, the development of these capacities, and how they are used, might have been woven into the very fabric of the designer's story from day one. Thus, in at least one important sense, we might be "pawns at the start of our lives" and forever after.
There is much more to say about this topic, but I will leave it at this. I'm very interested to hear what other Gardeners think of these arguments and Berofsky's new article.