Is and Ought
Robert Bass
I’m reasonably confident that nobody ever has or ever will
derive an ‘ought’ from an ‘is.’ More precisely, my claim is
that substantive ought-statements can’t be validly derived from any
set of is-statements that do not themselves presuppose substantive
ought-statements.
The two qualifications there are crucial. First, the ought-statement must be substantive. It’s easy to get an ought-statement from an is-statement if you don’t insist on that. For example, from “John is a sea-captain� it follows that “John has whatever obligations sea-captains have� – which of course is consistent with saying that there aren’t any obligations that sea-captains have. Or, from “people sometimes drink tea� it follows that “Either people sometimes drink tea or all red-heads ought to be shot� – which doesn’t imply anything about whether red-heads should be shot.
Second, it’s important that the is-statements don’t themselves presuppose substantive ought-statements. It’s relatively easy to build some judgment about what ought to be done into the definition of a term the reference of which is fixed by descriptive criteria – and this is not uncommon in ordinary language. So, for example, “that’s stealing� might be taken both to have its correct application settled by descriptive criteria and also to imply that the action described is wrong. The problem with that is that if the reference of “stealing� really is fixed by descriptive criteria, we can just substitute an invented term, “schmealing� say, that uses exactly the same descriptive criteria, but without carrying any moral meaning, and ask why we should think that schmealing is wrong.
In any case, here’s a short argument that no substantive ought-statement can validly be derived from any set of is-statements that do not presuppose substantive ought-statements. Suppose you have a set of is-statements for premises and you validly derive a substantive ought-statement – let’s say that Mike ought not to kill Mark. If the derivation is valid, that means that the denial of the conclusion, “it’s not true that Mike ought not to kill Mark� has to contradict something in the premises. That’s just a basic point of logic. But if the denial of the conclusion does contradict the premises, that means that “Mike ought not to kill Mark� must somehow be part of the premises. But what that means is that the premises were not, as advertised, free from any presupposition of substantive ought-statements.
Note that this is a completely general argument. It doesn’t matter what the premises are. It just points out that if the derivation is valid and the conclusion is a substantive ought-statement, then a denial of that conclusion – a denial of some particular substantive ought-statement – has to contradict the premises, and so, one way or another, the substantive ought-statement must be there in the premises. If you don’t have a substantive ought-statement, either explicit or presupposed, somehow in the premises, you can’t non-fallaciously end up with one in the conclusion.
Now, there are lots of sophisticated attempts to get around that kind of consideration, but I don’t think any of them are successful. Explaining how to derive an ought-statement from some set of is-statements is less obviously wrong but on the same order as explaining how to draw round squares. Either the argument will turn out to be straight-forwardly fallacious, turning on some equivocation or contradiction in the premises or else there will be some claim taken to be reason-giving at least implicit in the premises and relied upon in the course of the argument.
Before closing, though, please note that I have not said anything
to the effect that we can’t justify positions in ethics, can’t
have good reasons for preferring one to others or anything of that
kind. When people are faced with the claim that an “ought� can’t
be derived from an “is,� they often seem to go in one of three
directions. Some seem to accept it and draw the conclusion that
“oughts� can’t be justified at all. Some seem to give up on
reasoning in ethics, in effect giving up, not on their views
about what ought to be done, but on providing any reasoned defense
for their views. Still others find both of those unacceptable and, in
effect, stamp their feet and insist that yes, one can somehow
derive an “ought� from an “is.� All these reactions seem to
me to terminate the discussion just where the interesting questions
start.
Rob
---
Rob Bass
rhbass@gmail.com
oocities.com/amosapient
A great many people think they are thinking when they are merely
rearranging their prejudices.
– William James