Hume treats promises like he did property. First, he argues that the obligation to keep promises cannot be natural. Second, he tries to explain how the practice of promising could arise out of a convention.
At the end of the section (3.2.5 pars. 13-15), he adds two arguments for the conclusion that the obligation to do what you promise is not natural. While this was a poor organizational choice, one of the arguments is quite good.
Two arguments for the conclusion that the obligation to do what you promise is not natural: (a) promises are unintelligible prior to human conventions and (b) even if they were intelligible, they would not create obligations prior to human conventions (3.2.5 pars. 2-5). (I am going to say I think this is just one argument, for what it’s worth.)
The circularity argument again and some remarks about the difference between natural and artificial virtues. (3.2.5 pars. 6-7)
Hume’s positive account: how the convention establishing promises is established and sustained (3.2.5 pars. 8-12)
Third argument for the conclusion that the obligation to do what you promise is not natural: the rules governing the relationship between what a person says and what that person intends in using words like “I promise …” are too inconsistent to be natural. Only a human convention could look like this. (3.2.5 pars. 13-14)
Fourth argument for the conclusion that the obligation to do what you promise is not natural: the rules governing promises made under duress are too inconsistent to be natural. Only a human convention could look like this. (3.2.5 par. 15)
Note that many of Hume’s arguments have a common form. It goes like this.
If the obligation to do what you promise was natural, it would be like this (where “this” varies from argument to argument).
The obligation to do what you promise is not like that.
Therefore, the obligation to do what you promise is not natural.
When you make a promise, you create an obligation, namely, the obligation to do what you promised to do. If you promise to meet me for lunch, presto chango, you have created an obligation to meet me for lunch. The moral universe has changed with one little sentence! How does that work?
Hume says that he has two arguments for the conclusion that creating an obligation with a promise is not a natural phenomenon (3.2.5 pars. 4-5). As far as I can tell, they are basically the same. I am going to break the argument into two parts.
This is the first part.
This is the second part.
Hume says his two arguments lead to two different conclusions: that promises are naturally “unintelligible” (par. 4) and that even if this were not so, promises could not create obligations (par. 5). The reasoning in both paragraphs looks the same to me.
Hume follows this by repeating the argument he made earlier about how the motive of duty involves a kind of circular thinking: what makes an action virtuous is that it is performed for a virtuous motive and what makes a motive virtuous is that it involves “regard to the virtue of the action”. (3.2.5 par. 6-7)
I think paragraph six repays careful study because of what it says about the difference between the natural and artificial virtues. Maybe we will take another crack at the circularity argument too.
So far, what we have is this. Promises and promissory obligations cannot be created by anything we are naturally capable of thinking or doing. But there is obviously such a thing as promises and promissory obligations! The reason that they exist is that we created them in, yes, a convention.
Hume proceeds as he did with property. We have an interest in being able to make promises. Your crops need to be harvested today, mine will have to be harvested tomorrow, and both jobs require two people. You want me to help you today and I want you to help me tomorrow. I am willing to help you only if you will help me and vice versa. So far, this sounds like the rowboat again. Only there is a difference. We want to cooperate by doing things at different times whereas in the rowboat, the two people have to row at the same time. That is a problem. If I help you today, what assurance do I have that you will help me tomorrow? Promises are supposed to supply the assurance.
But how do they do that? If I look at your motivations, I will see that you do not have any reason to help me once I have helped you. You do not just help strangers. While you would help someone close to you, such as, say, a cousin, you are not close to me.
The convention goes like this. When someone uses special words, like “I promise to do X” everyone else will do two things. First, they will trust that the speaker will, in fact, do X and second, they will not trust the speaker in the future if he does not do X. When a group of people coordinate their behavior in those two ways, then they have a convention about promising.
Before the convention existed, no one had any reason to do X after saying “I promise to do X.” Once the convention exists, then we all have a strong incentive to do X when we promise to do X.
Suppose you say “if you help me today, I promise I will help you tomorrow” and I know that it is conventional in our society to distrust people who do not do what they promise, then I will be assured that you will follow through. Even though you are the one with the obligation, you benefit: you got me to help harvest your crops.
What is your motivation for following through? Hume describes two stages of the convention. At first, your motivation will be self-interest: you do not want everyone to distrust you. That will only take us so far. In a larger society, most people will not know if you go back on your word. So self-interest will not give you a reliable motivation for keeping your promises. Instead, the three forces of sympathy with the public interest, education, and “the artifices of politicians” fill the void. Roughly, either as a rule of thumb, internalizing the opinions of others, or having it drilled into you by your parents, you will be motivated by a thought like “it’s my duty” or “I have to because I promised.” (3.2.5 pars. 11-12)
At this point, I think it is worth writing down the ways that Hume thinks the sense of duty is not natural. So we will do that in class.
At the end of the section, Hume makes two more arguments for the conclusion that promises are an artificial creation. Both of them involve arguments to the effect that the rules governing promises cannot be characterized by describing the natural elements of a promise. (Sorry, I can’t think of a better way of putting it; hopefully we can improve on that sentence in class.)
The first argument is about the relationship between a person’s intentions and that person’s utterance of words like “I promise to X.” The words are necessary: you can’t make a promise just by thinking it, you have to say something that others hear. But the words alone are insufficient. I might be joking: “if you tell him that, I promise to eat my socks!” I use the words, but I do not have the right intention, so there is no obligation to eat my socks. I never really meant it. Other times, the fact that I did not mean what I said does not cancel my words. Suppose I mean to deceive you by making a promise that I have no intentions of keeping. That is still a promise. Here, the words do count even though I have no intention of doing what I said. (3.2.5 par. 13)
If promising were natural, Hume asserts, there would be a consistent alignment between intentions and words. Since that is not so, he concludes that promising is not a natural phenomenon.
Human conventions, by contrast, can set up whatever rules about the relationship between intentions and words that we want.
The second argument concerns duress or coercion (3.2.5 par. 15). I am going to leave this one as an exercise for you. Despite being stuck at the very end, it is one of his better points, in my opinion.