Philosophy of Law Spring 2018

Austin’s Command Theory


We went over Austin’s theory that laws are commands. We noted several problems with the theory. None of these are necessarily fatal to Austin’s theory, but Hart will argue that his theory has better ways of addressing most of them. The problems are:

  1. Not all laws fit the model of commands; some appear to be more like what we called instructions.

  2. Having an obligation is not generally thought to be the same thing as being threatened with a sanction.

  3. We recognize new sovereigns even before the bulk of the population has developed habits of obeying them.

There is one last issue that will bedevil both Austin and Hart.

  1. Judges often decide cases based on what is called “customary law.” This does not seem to come from the commands of a sovereign.


Austin thought in trees. He liked to begin with a root concept and then subdivide it into branches until he had identified the concept he wished to analyze. So, for example, there are two branches off of the root concept of “expressions of desire:” requests and commands. While they are grammatically similar, they are different because commands come with sanctions. Laws will be one of the sub-branches of commands.

Commands, according to Austin, always involve three things (Austin [1832] 1955, 17):

  1. a desire concerning someone’s behavior
  2. an expression of that desire
  3. a sanction, threatened harm for non-compliance

The first two elements are common between requests and commands while the last one distinguishes commands from requests.

I asked about an expression that looks like a command: “plug in the toaster.” Of course, that could be a command. But it could also be what Nico called an instruction; if you look at the manual for my toaster, for instance, you will see this listed as the second step for making toast (after “take the toaster out of the box”; I am not making this up).

Instructions look like commands, but they do not meet two of Austin’s criteria: they do not necessarily express the speaker’s desires about what the person receiving the instruction is going to do and there is no sanction attached.

The author of the instruction manual need not care whether you ever make toast. He or she is just telling you what to do if you want to do so. By contrast, when I command my son “turn the lights back on!” I really want him to turn on the lights; there is no implied “if you want me to be able to see what I am doing” clause.

The other thing missing is a threatened sanction. When a speaker issues a threat, what the speaker means is usually that he or she will do something bad to you if you do X. Here, the author of the instruction manual is just telling you that you won’t get toast if you don’t follow the instructions. The author isn’t going to come to your house and take the bread away.

Why does this matter? Well, if parts of the law are more like instructions than they are like commands, the command theory will look like at best an incomplete theory of what laws are.


For Austin, the terms “command,” “sanction,” and “duty” (or “obligation”) are all defined in terms of one another. To receive a command is equivalent to being threatened with a sanction and being threatened with a sanction is equivalent to having a duty (Austin [1832] 1955, 14–18). This, according to Austin, is why people are obliged to obey the law.

We talked quickly about how this diverges from our ordinary use of the term “duty.”

For instance, if I owe a debt to Helena, the fact that she can’t threaten me with a sanction does not make it the case that my debts are cancelled. I still have a duty to pay even if she can’t threaten me with harm. This suggests that being threatened with a sanction is a not a necessary condition of having a duty; you can have a duty even if you are not threatened.

Conversely, being threatened with a sanction is not a sufficient condition of having a duty. In the “your money or your life” situation, you are certainly being threatened. But you would not do anything wrong if you gave the gunman the slip and refused to pay.

These points seem true about moral duties. Are they also true about legal duties? We will have to talk about that. If the two cases are quite different, then Austin may be on solid ground. If not, he’s not.


A sovereign, according to Austin, is someone that the “bulk” of the population is in the habit of obeying while not being in the habit of obeying anyone else (Austin [1832] 1955, 193–94).

So what happens in a transition from, say, one king to the next? The king is dead, they say, long live the king. How did the new guy get to be the king? The bulk of the population isn’t in the habit of obeying him yet.

We said that the population is in the habit of obeying the office. That sounds right. Austin doesn’t say a lot about what an office is or how we could identify it. Maybe he could while keeping to the main ideas of his theory, however.

Customary law

When judges say things like “like cases should be decided alike” or “no one should profit from their own wrongdoing,” they are sometimes not drawing on anything that could be traced to a sovereign’s command. Instead, they are referring to what is sometimes called “customary law.”

Is customary law part of the law or not?

Austin could say “my theory says it has to come from the sovereign’s commands, this doesn’t, so it’s not law.”

But he doesn’t say that. He says that the customary law is “judge-made law” (Austin [1832] 1955, 32). How do judges make law? They get the authority to do so from the sovereign.

We’re going to talk about whether that makes sense, especially when we read Dworkin.

What are we doing?

Camille asked a good question. Is a theory of law supposed to be true to the way the law actually is or is it supposed to propose an ideal of how the law ought to be?

If it’s the former, then any bit of the law that doesn’t fit into Austin’s theory is bad news for Austin. But if it’s the latter, then he would be free to say that the bits that don’t fit should be excluded.

I think the answer is “both.” The reason why we want to have one of these theories is that we want to better understand what the law is. If we knew it all perfectly, we wouldn’t need a theory. So the theories have to explain the practices that we have. But, at the same time, it wouldn’t be surprising if they showed us that there were some things that we would be better off dropping, if only in the name of clarity and consistency. That’s what Austin says we should do with international law, for instance: admit it’s not really law.

Main ideas

Here is a list of key terms in Austin’s theory that you should be able to explain after today’s class.

  1. Command
  2. Duty (or obligation)
  3. Sovereign

You should also know the problems Austin faces on each point.


Austin, John. (1832) 1955. The Province of Jurisprudence Determined. London: Weidenfeld and Nicolson.


There was a handout for this class: 02.Austin.handout.pdf