Saturday, January 09, 2016

Reasons: Motivating, Explanatory, and Normative

This post concerns the fourth article in an anthology that discusses John Mackie’s Error Theory. (Joyce, Richard and Kirchin, Simon (eds), Philosophical Studies Series, World Without Values: Essays on John Mackie’s Moral Error Theory.

In “Mackie’s Internalisms”, Walter Sinnott-Armstrong does me a great favor by spelling out a number of distinctions regarding reasons, motivations, and their place in moral claims. For a person who is not greatly familiar with this territory, a survey of distinctions is quite useful. Furthermore, it provides a person such as myself with a way to place my own beliefs on this landscape.

The first distinction is between motivations and reasons – with the latter further distinguished between explanatory reasons and normative (justifying) reasons.

Distinguishing Types of Reasons

A normative reason is a type of reason that justifies an action. Question: “Why did you shoot him?” Answer: “It was self-defense.” The person who offers a reason like this is not concerned with explaining his actions - or, at least, not solely concerned with explaining his actions. He is concerned with showing that it was a legitimate action - an action that ought to have been performed.

Note that an agent can have a normative reason to perform an action he has no interest in performing. An employer can have a good normative reason to fire an employee (i.e., he has caught the employee steeling from the company), even though the agent has no interest in doing so (he hates confrontation). In fact, the justificatory reason is applicable even though the boss absolutely refuses to terminate the employee. Here, a normative reason means that, if the agent were to perform the action, he would be justified in doing so.

Explanatory reasons, in contrast, are used to explain existing actions. “Why did he sneak out of work a half hour early?” Answer: “He was angry at his boss and simply wanted to be someplace else.” This does not justify sneaking out of work early, but it may explain his actions.

Explanatory reasons are necessarily linked to motivating reasons. In fact, the quest for an explanatory reason for an intentional act is a quest to discover the motivation that caused the act.

However, the set of motivating reasons may be larger than the set of explanatory reasons. A person with a fear of flying may have a motivating reason not to fly to Perth, Australia for a job interview. However, the prospect of getting the job could, at the same time, provide motivating reasons to fly to Perth. Regardless of the option that this agent chooses, one set of motivating reasons explains the decision, while the decision is made "in spite of" the other set of motivating reasons.

When a person is asked to explain her actions, she will sometimes do so by offering a belief, rather than a desire. She can say that the reason that she turned left on Maple Street was because she knew - or, at least, believed - that there was a bakery at the end of Maple Street. However, this belief invites us to fill in the blanks regarding motivation. Why did she want to get to a bakery? To pick up a cake for somebody's birthday? To grab some bagels for her coworkers? We are still looking for the reason why the belief was important.

"Ought" and Normative Reasons

I follow Mackie in thinking that an "ought" statement is a search for reasons to act.

If a person says, “You ought to do X,” a natural response would be to ask, “Why?” This "Why?" can only be answered by providing an end-reason for intentional action that exists - or the combined weight of several reasons (such as all of the desires of the agent who is being given advice).

Now, we get to ask, "What kind of reasons are "ought" questions looking for?"

Actually, it seems that this is going to depend on the type of "ought" one is looking at. If this is a practical "ought", then the relevant reasons would be the motivating reasons of the person who is being advised. "If you are going to rob that convenience store, you ought to wear a mask." This "ought" would be grounded on the motivating reasons that the agent has not to be identified and punished as a robber.

On the other hand, moral "ought" appears to be looking for normative reasons - reasons that would justify the action.

Let's remember that this discussion is taking place in the context of J.L. Mackie's Error Theory and Walter Sinnott-Armstrong's analysis of that theory. Mackie is claiming that "objective values do not exist" and Sinnott-Armstrong is trying to figure out what, exactly, is Mackie claiming not to exist. Whatever they are, one of Mackie's reasons for claiming that they do not exist is that they are "queer".

What would be "queer" according to Sinnott-Armstrong are motivating reasons that are not linked to desires - that are built into states of affairs themselves such that merely being aware of them would motivate the agent to act in a particular way. When Mackie claims that objective values do not exist, he should be understood as claiming that motivating reasons independent of desires or interests do not exist.

Reasons that Exist vs. Reasons an Agent Has

I want to add to this discussion another distinction - a distinction between the reasons that an agent has and reasons that exist.

When we say that an agent "ought to do X", we are often not clear as to whether we are saying that the agent "has a reason to do X" or whether "there exists reasons for agent to do X".

I hold that desires provide the only end-reasons for intentional action that exist. Beliefs and desires explain intentional action - where desires identify the goal while beliefs select the means for reaching that goal. That is to say, desires provide the motivation for intentional action. Here, then, we relate desires to explanatory and motivating reasons.

The desires/reasons that an agent has is a very small portion of the desires/reasons that exist. The latter includes every other desire that every other creature has.

A desire only motivates the agent who has it. All of the other desires that exist - all of the other reasons that exist - are reasons for those other agents to act, but neither explanatory nor motivating reasons for the agent in question.

Let me just toss out a suggestion here for "normative reasons".

There are, of course, two widely recognized families of normative reasons. There's the practical "ought" and the moral "ought". The practical "ought" looks at the desires that the agent has. Moral "ought", I would argue, looks at reasons that exist. That is to say, it considers the desires of people other than the agent.

It is not the case that every reason that exists counts as a justification for some action. However, every reason that exists has its voice counted.

Instead, we look at these reasons/desires that exist and we measure the number and strength of their connections to other reasons/desires that exist. We see that some desires tend to fulfill other desires, while some desires tend to thwart other desires.

In the case of moral ought, "normative reasons" looks for the desires that tend to fulfill other desires.

So, for example, we discover that a desire to keep promises is a desire that tends to fulfill other desires. This is why we promote a desire to keep promises by rewarding and praising those who keep promises, and punishing and condemning those who break promises. These tools help to promote in people generally an interest in keeping promises.

"Because I promised to do so" then refers to a reason that people generally have reason to promote using rewards and praise. It is a "good reason".

This account would be consistent with the observation that a normative reason can be applicable even though the agent has no motivating reason to do what she ought to do. The fact that an agent has no motivation to keep a particular promise, and no intention to do so, "because you promised" remains a reason that people generally have reason to promote by praising and rewarding those who act in ways consistent with this interest, and punish or condemn those who violate this interest. It's truth - and its relevance - does not depend on the beliefs or the desires of the person being praised or condemned.


As I said, Sinnott-Armstrong's article is great for making a number of important distinctions. The distinction between explanatory, motivating, and normative reasons was the first of many. I will discuss other distinctions in future posts.

No comments: