From 'Is' to 'Ought'? Help for Hodges

Our  expat friend, Seoul man, and professor of English, Jeff Hodges, has been puzzling over whether an 'ought' statement can be validly derived from an 'is' statement.  Here is his example, put in my own way:

1. Democratic regimes contribute more to human flourishing than do non-democratic ones.

Therefore

2. If we want to maximize human flourishing, then we ought to support democratic regimes.

(1) purports to state what is the case.  In this sense, it is a factual claim.  On this use of 'factual,' a factual claim need not be true. ('I live in New Mexico' is false but factual as opposed to normative.)  Factual claims on this use of 'factual' are opposed to claims as to what one ought to do or ought not to do, or what ought to be, or ought not to be, or what is better or worse or what is more valuable or less valuable.

It is worth noting that both (1) and (2) are in the indicative mood.  Thus we ought to distinguish (2) from the hypothetical (as opposed to categorical) imperative

2*.  To maximize human flourishing, support democratic regimes!

One difference is that while it makes sense to inquire whether (2) is true or false, it makes no sense to inquire  whether (2*) is either true or false.  It follows that our question is not whether an imperative can be validly inferred from an indicative.

Let us also note that (2) is a conditional.  It is a compound statement  consisting of two simple component  statements,  an antecedent (protasis) and and a consequent (apodosis).  To assert a conditional is not to assert either its antecedent or its consequent. It is to assert a connection between the two.  For example, if I assert that if the light is on, then current is flowing through the filament, I do not thereby assert that the light is on, or that current is flowing throught the filament; what I assert is a connection between the two, in this case a causal linkage.

Given this fact about conditionals, I do not consider Jeff's example to show that one can validily derive an 'ought' from an 'is,' a normative statement from a factual statement.  Both (1) and (2) are nonnormative statements.  The first is obviously nonnormative.  But the second is as well despite the fact the 'ought' occurs within it.  For all it asserts -- or, to be precise, all a person asserts who assertively utters a token of the sentence in question -- is a connection between two propositions, a connection that it nonnormative.

We could of course detach the consequent of (2) thusly:

1. Democratic regimes contribute more to human flourishing than do non-democratic ones.

2a. We want to promote human flourishing

Therefore

2c. We ought to support democratic regimes.

(2c) is unabashedly normative.  But it does not follow from the premises which are both of them nonnormative.

So Jeff has not given a counterexample to what philosophers claim when they claim that an 'ought' cannot be derived from an 'is.'

But I will irenically add that there is nothing wrong with Jeff's original argument.  It is just that it is not an example of the derivation of a normative statement from a nonnormative one.  It is an example of how a statement containing the word 'ought' can be validily derived from a statement not containing the word 'ought.'  If this is all that Jeff means to show, then he deserves the coveted MavPhil imprimatur and nihil obstat. 

Crucial here is the fact that not every statement containing 'ought' is a normative statement.  Besides (2), there is this example:  'I just replaced the battery, so my car ought to start.'  This is not a statement about  what anyone ought to do, or even about what ought to be; it is a prediction.  One could just as well say, 'I just replaced the battery in my car, so it is highly likely that the car will start.'

And now it occurs to me that 'ought' can be paraphrased away, salva significatione,  even in the case of (2).  Try this:

2p.  If we want to maximize human flourishing, then it is necessary that we support democratic regimes.

Related:  The Ought-to-Be, the Ought-to-Do, and the Aporetics of "Be Ye Perfect"

Related articles

Asserting and Arguing: Analysis of an Example and Response to Novak

 

Comments

[Horace Jeffery Hodges]

Thanks, Bill. I'll have to give this some thought.

Jeffery Hodges

* * *

[Miloš]

In my memory left example from Alasdair MacIntyre's After Virtue:

1. He is naval captain

2. He ought to do what naval captain ought to do.

In this book MacIntyre give few such examples and notice that in case of functional terms such derivation is quite simple (which is one of center themes of MacIntyre's ethics).

[Bill Vallicella]

Milos,

I haven't looked at MacIntyre's book in a long time, but the example you give is intriguing.

Call the naval commander 'Tom.' Qua naval commander, Tom has certain duties. These are tasks he ought to perform. But the 'ought' is not categorical, but conditional: it is conditional upon Tom's being a naval commander. There is no necessity that Tom be a naval commander

As I understand the nonderivability of 'ought' from 'is,' however, the claim is that categorical normative statements cannot be derived from factual statements.

Example:

1. Almost all people want the respect of (some) others and will receive respect if they show respect.

Therefore

2. We ought to show others respect.

How does that follow?

[Peter Lupu]

I think Bill has nailed the correct response in both cases. Bill's response to MacIntyre's example cited by Milos is also on the mark. So I guess I do not have much to add. My question is: what should we conclude from the fact that normative statements cannot be validly inferred from descriptive ones? Is this enough to refute naturalism about morality?

Consider this example: theoretical terms (and the laws within which they are couched) cannot be reduced to observational terms. Thus, a physical law-like statement involving theoretical terms cannot be validly inferred from a collection of observation statements. Does this show that physical laws are not naturalistic?

[Bill Vallicella]

Thanks for the comment, Peter.

I'd have to think about it, but it seems that one who thinks that norms can be inferred from facts commits what Moore calls the naturalistic fallacy. Said fallacy is in the close vicinity of Moore's Open Question argument. Off the top of my head, the nonderivability of norms from facts would seem to refute naturalism about morality.

The point you make about physical law statements refutes a regularity theory of laws, and thus perhaps empiricism about laws, but not naturalism about laws. But let's not get involved in this mess.

We ought to stick to putative CEs to the nonderivability of norms from facts. There is Searle's famous argument from 1964, and the MacIntyre examples.

[Miloš]

Thank you Dr Vallicella for your comment. I will try to reconstruct MacIntyre's approach as I understand it. MacIntyre, as I understand his work (and some other philosophers, Peter Geach and G. E. M. Anscombe from example) try to reestablish Aristotelian approach to ethical theory (you certainly know this but I must make some introduction :).

Hume said here is not valid derivation from factual to normative statement. But MacIntyre and some other thinkers show that there can be but under some conditions. Those conditions are that man actually always is in some specific function (there is no some abstracted person as Tom, but Tom in some specific social funcion). But if we accept wider Humian and Kantian conceptual network such derivation become much harder even impossible. Maybe my reading of these writers are is on wrong trail, but I have made mistake in good faith.

I believe that Ms. Anscombe first in analytic philosophy (I've got good link about this but unfortunately I lost it.

[P. S. I must add one personal note, it can be published or not. I am not philosopher but I really enjoy in your blog last few months, practically I read few posts every day. I feel gratitude and respect for your efforts.]

[Bill Vallicella]

Milos,

You are right: if we add the Aristotelian background, then the derivation becomes possible. See my latest post where I explain this as best I can. I want to thank you for referring me to MacIntyre. I had his book in my library for 30 years or so but only today really studied parts of it.

Thank you for reading my blog. I'm glad you find it helpful. Where are you located?

English is a very difficult language. "I read a few posts every day" not: "I read few posts every day."

Tom has few books: He doesn't have many.

Tom has quite a few books: He has many books.

What a crazy language!

[Horace Jeffery Hodges]

Speaking of English as a crazy language . . . the English subjunctive can be hard to pin down. Take this example:

"2. If we want to maximize human flourishing, then we ought to support democratic regimes."

Bill, you say that "2" is in the indicative mood. I don't quite see what you mean. The "if-clause" is indicative, but the "then-clause" looks subjunctive to me. This might be easier to see if we replace "ought" with "should":

"2. If we want to maximize human flourishing, then we should support democratic regimes."

The word "should" is here the subjunctive of "shall," isn't it? It expresses a situation that does not exist and of which there remains some doubt as to whether or not it will exist, I think. Compare:

We shall support democratic regimes.

We should support democratic regimes.

The first is indicative and expresses certainty that a situation (i.e., support of democratic regimes) will exist. The second is subjunctive and allows doubt that a situation (i.e., support of democratic regimes) will exist.

I don't know that this affects your argument, but it's a point that needs some clarification, at least for me.

Jeffery Hodges

* * *

[Bill Vallicella]

Indicative conditionals:

If Mary didn't cook the dinner, Tom did.

If Oswald didn't kill Kennedy, someone else did.

If we want to maximize human flourishing, then we should support democratic regimes.

Subjunctive (counterfactual) conditionals:

If Mary hadn't cooked the dinner, Tom would have.

If Oswald hadn't killed Kennedy, someone else would have.

If we had wanted to maximize human flourishing, then we ought to have supported democratic regimes.

>>It is controversial how best to classify conditionals. According to some theorists, the forward-looking "indicatives" (those with a "will" in the main clause) belong with the "subjunctives" (those with a "would" in the main clause), and not with the other "indicatives". (See Gibbard (1981, pp. 222-6), Dudman (1984, 1988), Bennett (1988). Bennett (1995) changed his mind. Jackson (1990) defends the traditional view.) The easy transition from typical "wills" to "woulds" is indeed a datum to be explained. Still, straightforward statements about the past, present or future, to which a conditional clause is attached — the traditional class of indicative conditionals — do (in my view) constitute a single semantic kind. The theories to be discussed do not fare better or worse when restricted to a particular subspecies.<< SEP entry on conditionals

But none of this is necessary, Jeff, to understand the point I am making against you.

[Miloš]

You're welcome, I am happy to help in this interesting discussion. Thank you for correct mistake in my use of English. I am from Belgrade (Serbia).

[Miloš]

In one earlier comment I mention that Anscombe in analytic philosophy first to undermine view that there is no correct derivation from factual to normative statements. This few lines from article Language, Truth and Logos (http://standpointmag.co.uk/node/681/full) can be of some interest:

One corrosive source of indifference about such distinctions is an old dogma, perhaps falsely ascribed to David Hume: that one cannot derive an "ought" from an "is". Which can suggest the following: if anyone offers an inference to a moral conclusion from agreed facts, either the conclusion doesn't follow, or, if it does follow, the premises must state not just facts but also ethical presumptions as contentious as the conclusion. Anscombe undermined this assumption in an early essay, On Brute Facts, which was as simple as succinct. She imagined the following sequence of descriptions: "The grocer had the potatoes brought to my house and left there"; "He supplied me with potatoes"; "I owe him such-and-such a sum of money" - to which John Searle was explicitly to add, "I ought to pay him such-and-such a sum of money." Anscombe notes that none of these descriptions entails its successor: at any point, for all that is stated, something untoward might exclude the ordinary inference. And yet in normal circumstances (which couldn't be spelled out exhaustively) the inference goes through.

I believe that Hilary Putnam also wrote on this subject but I have not read those works.

[Horace Jeffery Hodges]

But consider these:

If we want to maximize human flourishing, then we shall support democratic regimes.

If we want to maximize human flourishing, then we should support democratic regimes.

In the first example, "shall" allows for no uncertainty, given the "if-clause," and both clauses are indicative.

In the second example, "should" allows for some uncertainty, even given the "if-clause," and the "if-clause" is indicative, whereas the "then-clause" is subjunctive.

I just want to clarify this point completely -- I still have to think about your more central argument.

Jeffery Hodges

* * *