Wednesday, May 12, 2010

Sam Harris, Sean Carroll, and Deriving Ought from Is

Last month I wrote a series of articles concerning Sam Harris' speech at TED on the possibility of a science of morality.

(See Sam Harris' Presentation Science Can Answer Moral Questions)

I was not the only one to offer criticism of Sam Harris' comments. So, to compliment the points that I raised, I would like to say some things about the points that others have raised. Specifically, I would like to address the points raised by Sean Carroll.

(See Sean Carroll The Moral Equivalent of the Parallel Postulate.)

Harris asserted that the concern of morality is with the well-being of conscious creatures. Furthermore, we can tell, just by looking at the world, that there is obviously greater well-being in some parts of the world than others. He takes the well-being of conscious creatures to be an objective fact that we can study and come up with scientific truths about. These are not just scientific truths about how to obtain the well-being of conscious creatures, but scientific truths about the well-being of conscious creatures itself.

I would like to start by giving my take on Sam Harris' argument. As I see it, Harris' argument is much like the following:

Yes, we can have a science of physical matter. After all, every object in the universe is made up of a mixture of the four elements - earth, air, fire, and water. These elements are made up of atoms, which are extremely small pieces of an element which themselves have no parts. Now, we can have an objective science of the atoms of the four elements. Therefore, we can have a science of physical matter.

In other words, what Harris got right was his claim that we can have a science of morality. What he got wrong was his specific claims about the nature of the subject that we are studying scientifically.

The reason that this is important is because some of the criticisms that Harris has had to put up with commit a clear fallacy. They have attacked the specifics of Sam Harris' theory and then asserted that this disproves his claim that we can have a science of morality.

This is as fallacious as taking objections to the theory of matter included in my analogy to Harris's argument and taking that as proof that we cannot, in fact, have a science of physical matter. "Your theory fails, so a theory of physical matter is not possible."

It does not fail, so long as there are other theories that are out there. What the person raising this objection needs is not to find problems with Harris' specific theory, but some reason to believe that no theory is possible.

Carroll appeals to David Hume to provide that argument.

There is an old saying going back to David Hume that says that you cannot derive an ought from an is. And Hume was right! You can't derive an ought from an is! Yet people keep on trying.

Prove it.

Show me an argument that proves that you cannot derive an 'ought' from an 'is'. Hume gives us no argument. Hume gives us a fallacy. He gives as an "argument from ignorance". He says, I don't know how to do it; therefore, it cannot be done." Yet, it is a bit like saying that I cannot figure out the volume of an irregular solid; therefore, it cannot be done. Or, I cannot calculate the orbital mechanics necessary to get a spaceship from the Earth to the Moon; therefore, it cannot be done.

Here is Hume's often quoted paragraph:

In every system of morality, which I have hitherto met with, I have always remark'd, that the author proceeds for some time in the ordinary ways of reasoning, and establishes the being of a God, or makes observations concerning human affairs; when all of a sudden I am surpriz'd to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible; but is however, of the last consequence. For as this ought, or ought not, expresses some new relation or affirmation, 'tis necessary that it shou'd be observ'd and explain'd; and at the same time that a reason should be given; for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it.

There is nothing in here but an argument from ignorance.

It is as if we have introduced a new principle of logic that says:

• X is inconceivable to Hume.

• Therefore, not-X

Since this ill-fated paragraph was written, people have used it as an excuse for closing their minds to any discussion of real-world 'ought'. As soon as somebody starts talking about 'ought' as something in the real world, they shut their minds, shout "You can't do that", and point to Hume's argument from ignorance as proof.

It is as bad as any religion.

If there is an actual proof - something other than Hume's fallacy - I would like to know what it is.

There is an argument against accepting Hume's claim or, more precisely, concluding that what Hume finds to be inconceivable can somehow be done.

It has to do with all of the questions that surround the alternative.

If you believe that 'ought' cannot be derived from 'is', then I have a question for you. Do 'ought' statements have any relevance in the world of physical matter?

Is an 'ought' relationship - whatever it is - ever in any way a part of the explanation for, for example, the movement of atoms through space and time? Has it, for example, ever influenced a human action or been a part of explaining why a person did something or refrained from doing something else?

If the answer is "yes" then you are telling me that something that is completely divorced from what "is" - that is completely separate from it - has the power to influence the movement of matter through space-time.

How does that happen?

Now that you have told me what an ought-relationship is not - that it is not derived from 'is' - can you please tell me what this 'ought-relationship' is? And, in giving me this explanation, include an account of how 'ought' has the power to effect the movement of objects in the real world?

In fact, if we divorce 'ought' from 'is', can we even make sense of the question: "What 'is' an ought relationship?"

This is how Hume's so-called law shuts down the brain. If we accept it, not only can we not get answers to any questions, we cannot even form questions to ask. We have placed 'ought' outside of all intelligible conversation.

If the answer is "no", then my question to you is, "Why are we still talking about it?" If ought-relationships have no impact on the real world - if it has no explanation to offer as to why matter moves in a particular way - if it has no influence on any intentional action at any time - then why speak about it at all as if it has real-world relevance?

In any discussion of any real-world issue, anybody who brings up an ought-relationship as if it is relevant to the discussion can be hereby dismissed as speaking nonsense. It is as senseless as bringing up God or ghosts or any other imaginary entity that, because it is imaginary, can never have relevance to that which is real

So, this is my answer to those who say, "And Hume was right!"

If Hume was right, then explain to me how this 'ought-relationship' you are talking about has relevance in the real world? How can an 'ought' distinct and separate from 'is' have any relevance to what is?

Or, in other words, if you cannot derive 'ought' from 'is', then is it not also the case that you cannot derive 'is' from 'ought'?

Or, my common way of expressing this problem:

There is no mutually exclusive is-ought distinction. There is an is-is not distinction. Ought either needs to find a home in the world of what 'is', or we need to put 'ought' in the realm of what is not

Having said this, Harris is still wrong when he says that morality has to do exclusively with the well-being of conscious creatures. He is going to struggle to defend his claim that moral ought resides as a property in the well-being of conscious creatures - because it does not. However, Harris is wrong as a matter of fact, not because of the impossibility of deriving fact from value.

For the record, I hold that the 'ought' relationship is a relationship between states of affairs and desires, and that 'moral ought' is a relationship between malleable desires other desires. A virtue is a malleable desire that tends to fulfill other desires, while a 'vice' (using the term in its classical sense) is a malleable desire that tends to thwart other desires.

However, it doesn't follow from the fact that Harris made a mistake about what 'ought' is that 'ought' cannot be derived from 'is'. You might as well say that because the ancient Greeks were wrong about atoms that there can be no science of chemistry.

We need to find a theory that actually works - even if it turns out to be one that Hume could not have conceived.


Anonymous said...

I hope Sam Harris reads this.

Unknown said...

Critical dualism is the view that there is no way to derive moral principles from matters of fact. The last time I checked there is no morality to a fact. It is what it is. In the traditional language of moral philosophy this dualism is often called “the is/ought problem”, as though philosophers have to find some way to get over it or solve it.

The problem arises because philosophers want to find some way to justify moral principles and the best way would be to find some way to derive moral principles from some set of facts,thereby establishing an objective morality, unless it is believed that they can be handed down from some supernatural authority.

Almost everyone who has dealt with this issue has tried to answer it by reference to human nature or to the nature of ‘the good’. But each of these forks leads to a dead end: the first because anything that anyone can conceivabely do can be attributed to human nature and the second because there is the question of defining ‘the good’.

The same logical structure underlies the is/ought problem and the problem of induction which looms large in the philosophy of science. In each case the hope is to derive general principles (natural laws in the case of science, moral rules in the case of moral philosophy) from statements of fact. Deriving the general from the specific is inductive reasoning which will prove nothing. In each case the problem arises from the desire for justification based on facts and in each case the problem is insoluble in principle. The way forward is to aim to establish critical preferences for scientific theories or moral proposals based on their capacity to solve the problem that they are supposed to solve, and to stand up to criticism.

When you say, "What the person raising this objection needs is not to find problems with Harris' specific theory, but some reason to believe that no theory is possible.", I would say that it's far more important to look for the problems in a specific theory which would falsify it. I don't need a reason to believe that something doesn't exist. I need a reason to believe that it does.

josef said...

Alonzo, I read Hume differently. I agree with you that the argument Sean Carrol reads from Hume is an argument from ignorance.

But I think Hume himself wasn't making any assertion. Hume was just noting that moral philosophers move from 'is' to 'ought' without explaining how they did it. I think it is perfectly right to require moral philosophers to talk about how they cross "the divide".

I think Carrol's reading is a misreading, and a tragic one.

josef said...
This comment has been removed by the author.
Unknown said...

Hi Alonzo,

Your post was really helpful here, where we've been having a belated discussion about the Harris video across three blogs. Feel free to contribute.

Ophelia said...

Like josef says, you're reading Hume all wrong. You've taken his criticism of those who jump from is to ought with nary an explanation as a statement that it is impossible to go from is to ought under any circumstances.

That is not what he is saying. Try and find it anywhere in his writing, trust me you will fail. So your criticism of 'his' position is really a criticism of a position that he did not hold, nor espouse. It may be relevant to what other people say, but is irrelevant to what he had to say.

Alonzo Fyfe said...

Josef, Ophelia

I agree that another interpretation is possible, and I have defended it. However, in the course of a blog post I do not have the space to enter into that controversy.

I think that Hume's use of the phrase, "for what seems altogether inconceivable," is sufficient to support the interpretation that he considers the leap to the impossible. Though, as I said above, his only argument is an appeal to ignorance.

I also hold that he has a false assumption in stating that bridging this gap involves a leap from one type of relation to another "which are entirely different from it." The way this gap is bridged is by pointing out that "ought" is not, in fact, "entirely different" from "is" but is a subset of "is".

Since Hume does not correctly identify the problem it would be difficult to argue that he had a solution in mind.

Ultimately, he offers a burden of proof claim (the burden of proof is on those who make the leap), with a second claim that this burden of proof challenge will never be met.

Yet, in the end, I am going to repeat that there isn't space in the course of a blog post to debate this issue in detail. It is ultimately a red herring. The subject here is whether there is any reason other than Hume's claim that he finds it inconceivable that prevents us from deriving 'ought' from 'is', noting that what Hume could have conceived is a poor argument.

truth machine said...

Prove that you cannot create leprechauns from nationalism, or fashion hope from water. If you cannot, that does not imply that an assertion that these things cannot be done is argumentum ad ignorantiam. And if you can prove it, then I suggest that a similar proof can be applied to the impossibility of deriving an ought from an is. (Hint: Gilbert Ryle.)

David said...

Hume's point is simply this...moral imperatives cannot be derived from experience. for example (and in brief), if we observe Smith hitting Jones, we know that (1) Smith is hitting Jones. from this, however, we cannot infer that (2) Smith ought not hit Jones unless we add (3) if A is hitting B, then A ought not hit B (or the logical equivalent), but (3) is not a matter of observation; i.e., it is not know from experience (if we videotape Smith hitting Jones we will fail to capture the imperative in question, even if we review the video frame by frame).

Hume doesn't commit a fallacy. rather, he demands rational from those who issue moral imperatives and upon whom the burden of proof properly lies.

as for Sam Harris, and inasmuch as i agree with him on other points, his dismissal of Hume on this matter is at best philosophically naive.

Alonzo Fyfe said...


Many things are not "a matter of observation". We can watch a compass needle point north but, we can look frame by frame of the compass needle turning north and not see the magnetic fields causing the event.

And morality, I argue, is a relational party. You can take a picture of Jim standing alone and not observe the relationship that Jim is taller than Sandy. You need to observe Sandy at the same time.

Since (as I hold) the wrongness of Smith hitting Jones is in the relationship between the desires that motivate Smith to hit Jones (which we cannot see - regardless of how many frames we observe) and other desires (which are not even in the picture), Hume is looking in the wrong spot.

Yet, Hume commits the fallacy of ignorance. "Since I cannot see the wrongness in this picture, it does not exist". Or, "Since I cannot see the tallerness of Jim in this picture that shows only Jim, he must not be taller than Sandy."