In Sir P.F. Strawson’s brilliant 1949 paper ‘Ethical Intuitionism’,
I came across a short and seemingly powerful argument for the buck-passing
accounts of value and other ethical words that I haven’t seen elsewhere. Even
if I am a buck-passer, I do think that this argument is too good to be true.
So, what I want to do is to first give the argument in a full quote, then make
three observations about it, and finally sketch out the ways in which opponents
of the buck-passing view could reply to this argument.


Strawson’s paper consists of a dialogue between a defender
of intuitionism, North, and his opponent, West. The argument for the
buck-passing account is made by West during his arguments against intuitionism.
Here it is:

       “Looking for a logical nexus where there was none to be
found, you overlooked the logical relations of the ethical words among
themselves. And so you forgot what has often been pointed out: that for every
expression containing the words “right” and “good,” used in their ethical senses,
it is always possible to find an expression with the same meaning, but
containing, instead of these, the word “ought.” The equivalences are various,
and the variations subtle; but they are always to be found. For one to say, for
example, “I know where the good lies, I know what the right course is; but I
don’t know the end I ought to aim at,
the course I ought to follow” would
be self-contradictory. “Right”-sentences, “good”-sentences are shorthand for
“ought”-sentences. And this is enough in itself to explode the myth of
unanalysable characteristics designated by the indefinable predicates, “right”
and “good.” For “ought” is a relational
word; whereas “right” and “good” are predicative.
The simplest sentences containing “ought” are syntactically more complicated
than the simples sentences containing “right” or “good.” And, hence, since the
equivalences of meaning hold, the various ethical usages of “right” and “good” are all definable: variously definable
in terms of “ought.””

          Observation 1: It is interesting to see that in 1949 West is
able to present the buck-passing account as a generally accepted majority view
which North has merely forgotten about. I wonder when this view exactly went
away so that it was able to be rediscovered later.

         Observation 2: It is also interesting that Strawson doesn’t
think that the words ‘right’ and
‘good’ are analysable but rather all the sentences that contain them or the
uses of these words in these sentences. He seems to allow that the analyses in
terms of ‘ought’ can vary from sentences to sentence just as long as there is
one. This is an interesting version of the network analyses of a given domain.

        Observation 3: This also strikes as a powerful argument. Not
only does it defend the necessary co-extensiveness of value/right and oughts
but also the idea that it is conceptually contradictory to deny this. This
would rule out various modern views according to which value and ought are
distinct but still co-extensive due to perhaps the fact that same things make
things valuable and create oughts, or the fact that values explain oughts.

        Now to the responses. First, the opponent of the
buck-passing view could say that the argument is neutral between value/right
being analysable in terms of ought and ought being analysable in terms of
value/right. The contradiction would be created due to either one of these views being true. Perhaps the idea of simplicity vs. complexity in the end points towards
the buck-passing view though – the analyses must go from simple to complex.

       Second, the opponent could say what Frankena said to Moore. She could claim
that the argument is question-begging. The alleged contradiction is a contradiction only on the condition that the buck-passing view is true. So, West must be assuming the
truth of this view in making an argument for it which obviously is not justified.

        There is a response North could make to this. He could say
that all he is referring to is his pretheoretical semantic intuition about the contradiction
as a competent speaker, and the fact that most speakers under careful
deliberation would have the same intuition. He could then claim that truth of
the buck-passing analyses is the best explanation for why speakers have this
intuition.  At this point, the opponent
of the view could challenge whether competent speakers generally think that the
relevant contradictions exist. This would be an interesting debate.

        Third, the opponents of the buck-passing view could claim
that the relevant contradiction does not suffice to show that the buck-passing
analyses are valid. They could hold that ‘good’ and ‘right’ have a simple
meaning that cannot be decomposed. However, they could argue (in the style of
Barbara Partee) that there is a meaning-postulate attached to the simple
meaning according to which good things ought to be pursued, and so on. And,
they could claim that possessing the concept of ‘good’ requires conformity to
the meaning postulates in the use of the concept. This would be an alternative
explanation for the contradictions of the argument that did not require the
truth of the buck-passing view.

4 Replies to “Strawson’s Argument for Buck-Passing

  1. Your Observation 2 seems to be two observations, or anyway it has two independent parts. There is the idea that the sentences but not the words, have ‘good’-free equivalents; and then there is the idea that the analysis could differ from case to case. The second just seems to be the idea that ‘good’ could have a number of different senses. The first is a little harder to pin down, but it is in keeping with Strawson’s general approach (at least at the time); think of “Truth”, e.g., where there is no equivalent offered for the truth predicate, but a general recipe for forming truth-predicate-free equivalents. (Unsuccessful, of course.)

  2. First, I would deny the premise. “Fertilizer is good for the magnolias” is not equivalent to any ought-claim. Nothing follows about what I or the magnolias ought to do.
    One might reply to this that what does follow is a conditional ought-claim, e.g. that “if one wants nice magnolias, one ought to fertilize them”. But then I will say that (a) this claim is not equivalent, in that one can imagine cases where “If one wants nice magnolias, one ought to F magnolias” is true but “F is good for magnolias” is false. (E.g. imagine that a powerful fairy has promised to grow my magnolias if I dance around them.) And (b) even leaving those other cases aside, in this case it seems clear to me that the ought-claim is conceptually dependent on the good-claim rather than the other way around.

  3. Jamie,
    that’s a good point. I didn’t think he was thinking that the difference analyses reveal different senses of the word. I was thinking that he would think that it is somehow built into the one meaning of the term ‘good’ that we can on different occasions analyses the sentences in which it appears with sentences only using ought. I didn’t think that the implication of this would be that the fact that the ought sentences differ would be evidence that ‘good’ has been used in different sense. So, maybe the idea is some sort of contextualism on the level of word-meaning that doesn’t resolve into ambiguity.
    Heath,
    I’m not immediately convinced by the counter-example. Maybe he could say something like ‘if you were looking after magnolias, then ceteris paribus you ought to F magnolias’. I think some hedged principle like this will get the extension right. What will be more problematic for the argument is the idea that it will turn out to be *contradictory* to deny this.

Leave a Reply

Your email address will not be published. Required fields are marked *