Bulletin of the Santayana Society, No. 16 Fall 1998

              Caresses or Insults: A Note
On Santayana's Metaethics
It is generally agreed that Santayana offered a non-cognitive analysis of moral
judgements. His statements that "Moral terms are caresses or insults and describe
nothing" (dl 38-39), and "The cry ... How good! may be sincere, and it may be
applauded, but it is never true" (RB 479), express the two central tenets of traditional
non-copitivism. First, our moral "beliefs" are not intentional states about some matter
of fact; rather, they are intuitions expressive of the life of the psyche. Second, our moral
judgements are not truth-functional. For Santayana the idea of a true moral judgement
is unmeaning. Instead of being an attempt to describe the world, Santayana seemed to
suggest, similar to Stevenson, that moral expressions have persuasive force insofar as
they are "an overflow of the physical basis of thought" and "an audible gesture" (lr5
181).
The point I wish to make in this note is that it is simply not obvious why
Santayana offered a non-cognitive analysis of moral judgments. That he did so is, I
think, clear enough. Yet his naturalistic account of valuation would seem to support a
number of alternative analyses of moral judgments and current trends in metaethics. In
order to motivate a presentation of these alternatives I will use as a foil what is today the
standard and seemingly ubiquitous objection to non-cognitivism.
Non-cognitivism has few adherents and many critics. And, modern critics are
fairly animated in their discontent. Michael Smith, for example, states that non-
cognitivism is "outlandish," while Crispin Wright asserts that non-cognitivism allows
for a "grotesque lapse of radonality.,n The problem, more specifically, which both these
critics find is that non-cognitivism cannot plausibly explain the disciplined nature of
moral discourse. It is said that our moral talk exhibits "all the overt syntactic trappings
of assertion — negation, the conditional construction, embedding within propositional
attitudes, hypothesis and inference and so on" (Wright 3). Since non-cognitivism
appears unable to account for the surface grammar of our moral claims and cannot make
sense of the way we engage in moral arguments, it is dismissed from the outset.
One may object at this point that Santayana would have been unmoved by such
a criticism. He paid virtually no attention to problems in the philosophy of language and
eschewed the idea that by studying language we gain insight into the structure of
anything else. I think this a fair criticism and grant that, to a certain extent, I am
examining an aspect of his philosophy within a setting he would have found alien. Yet
in some ways this objection is beside the point. For even if we grant that Santayana
would not put much stock in contemporary metaethics with its linguistic bias, the
argument against non-cognitivism does raise certain questions about Santayana's reasons
for accepting non-cognitivism; questions which can be framed and discussed within the
context of his own system.
For example, one way to account for the propositional surface of moral discourse,
but still preserve the naturalism associated with non-cognitivism, is to hold that moral
judgements are actually truth-functional but false. I refer to Mackie's projective error-
theory. Mackie's theory has it that moral judgments make the claim to objectivity, but
1 Michael Smith, "Internalism's Wheel," and Crispin Wright, "Truth In Ethics," in Truth In
Ethics, Brad Hooker (ed.), Blackwell Publishers: Oxford, 1996, p.70 and p.4, respectively.