Compatibilism about practical reasons and the propositional account

Let’s call a theory of practical reasons a compatibilist view just in case it holds that there are two different types of practical reasons: motivating and normative reasons. Eric Wiland (here) argues that compatibilism cannot account for all of the following constraints:

(C1) The theory [of practical reasons] should display how reasons
explain the actions they are reasons for.

(C2) The theory should display how reasons justify the actions that they are reasons for.

(C3) The theory should display how reasons that justify action are also the sort of thing to explain action, and vice versa.

To be clear, Wiland thinks that explanatory reasons are the same as motivating reasons. Below we’ll see that this is wrong, but that isn’t that important.

I think that a version of compatibilism–the propositional account–can easily account for (C1)-(C3). The propositional account holds that both motivating reasons and normative reasons are constituted by propositions. Motivating reasons and normative reasons play different functional roles, however. Motivating reasons motivate and normative reasons justify. There are at least two important factors that determine whether something is a motivating reason. First, in order for a proposition to be a motivating reason R for agent A to Φ, R must be able to sensibly explain A’s Φing if A Φs. So, in order for the proposition that the drink is a cosmopolitan to be a motivating reason for Eve to drink from the glass, that the drink is a cosmopolitan must be a sensible explanation of Eve’s drinking if she drinks. But this isn’t enough. For suppose Eve’s twin sister Yvette unintentionally drinks from glasses with red liquids in them every time they are placed in front of her. In Yvette’s case, one could give a sensible explanation of her drinking by citing the proposition that the drink is a cosmopolitan, but that proposition wouldn’t be her motivating reason because she didn’t act for any reason. Thus, the second important factor is the proposition must be the thing (or one of the things) that moves the agent to act; it must be the reason the agent acts for. Yvette doesn’t act because of the proposition that the drink is a cosmopolitan, but Eve does. Thus, in order for proposition to be a motivating reason R for A to Φ, it must be (1) that citing R will be a sensible explanation of A’s Φing and (2) that R moved A to Φ.

On the propositional account, normative reasons are also propositions. However, they play a different functional role than motivating reasons. They justify intentions (or actions). Thus, there are different identity conditions for normative reasons than motivating reasons. First, in order for a proposition P to be a normative reason to Φ, P must ‘count in favor’ of Φing. Second, P must be true.

It’s important to notice that normative reasons can at the same time be motivating reasons. Take Eve’s case as an example. Given certain background conditions, she is sufficiently justified in drinking the cosmopolitan. Thus, that the drink is a cosmopolitan is a sufficiently weighty normative reason. Moreover, it is also her motivating reason to drink. It’s also important to note that it’s plausible that one can have motivating reasons that aren’t normative reasons, and vice versa. Joe’s motivating reason for intending to kill his wife might be that he ought to kill his wife. Plausibly, though, that proposition is false, and thus cannot be a normative reason. Moreover, Yvette might have the same normative reason as her sister to drink from the glass. That is, she might be in an identical situation with respect to normative reasons as her sister and thus have a normative reason to drink the cosmopolitan. Nonetheless, she doesn’t have a motivating reason to drink because she unintentionally drinks.

Now back to Wiland. He argues that compatibilist theories have two fatal flaws. First, he argues that the bifurcation of reasons into normative and motivating is ad hoc. Second, he argues that on compatibilist theories no one can ever act for the right reason. This is because the reasons that move people to act are motivating reasons, and the reasons that make an act the right one are normative reasons. But since they aren’t the same thing, it follows that no one can ever do the right thing for the right reason.

In pushing both objections Wiland assumes that the compatibilist must claim that motivating and normative reasons are ontologically different–i.e. constituted by different things. But this is just false. The propositional account holds that both types of reasons are constituted by propositions. Nonetheless, they are distinct kinds of reasons because they both play a different functional role. This confusion leads Wiland down the wrong path.

First, (C1)-(C3) implicitly assume that something plays both functional roles. Indeed, (C1)-(C3) force theories of practical reasons to account for the different functional roles. Moreover, there doesn’t seem to be a single concept that can play both roles. This is because of Joe’s case and Yvette’s case. In Joe’s case, Joe has a motivating reason but not a normative one. In Yvette’s case, Yvette has a normative reason but not a motivating one. Thus, it seems we must posit two concepts to get the extension right. This is precisely what the propositional account does. Thus, I fail to see why the bifurcation is ad hoc.

Second, agents can do the right thing for the right reason on the propositional account. This is because both motivating reasons and normative reasons are constituted by propositions. Moreover, nothing about the nature of normative reasons prevents them from simultaneously being motivating reasons. In other words, it’s easily possible–indeed it happens daily–that proposition P can (1) be true, (2) count in favor of Φing, (3) be a sensible explanation of A’s Φing and (4) be the thing that gets A to Φ.

~ by Errol Lord on March 23, 2008.