Realism in physics depends upon realism in
mathematics. Any complex description of the world will call upon the latter.
But mathematics itself is not descriptive. It doesn’t tell us about some
special class of objects lurking in or behind the fabric of the world. And if
we have no reason to abandon realism in mathematics merely because it is not
descriptive, then we have no reason to abandon realism in ethics on the grounds
that a good deal of it is also not descriptive. (To say that it is wrong to
do x does not tell us about any particular things that exist.) For Putnam,
we need not appeal in either case to special accounts of what there is
in order to underpin truth claims. We can have ethics without ontology.

What is explicitly assumed by Putnam is that it
is essentially the same cluster of anti-realist arguments which resurface in
area after area of philosophy, and that a response to anti-realism in one
domain gives us good grounds for its rejection in the others. This is Putnam’s
big strategic assumption. However, by dispensing with ontology in order to
defend realism in ethics, Putnam distances himself from a core feature of
contemporary analytic philosophy, its determination to treat questions of
ontology as a significant feature of ethics. His rejection of just such an
assumption, might seem to place him in proximity with what he calls
"chatter about ‘postmodernism’" which also seeks to remove appeal to
objects that sit beyond discourse, and which underpin its truth-claims.

Thence the second part of the book which moves
on from the question of ontology and tries to distance his position from
postmodernism by setting out a narrative of intellectual progress. Putnam
writes of three enlightenments: the ancient Platonic enlightenment, the
familiar 17th and 18th century Enlightenment; and a third pragmatist enlightenment
exemplified by John Dewey. The latter is hoped for, still indecisively taking
place and features a more fallibilistic humility and an anti-metaphysical
approach. Excessive metaphysical and ontological commitments are to be pared
away but commitment to progress is to be retained. This anti-postmodern
narrative of enlightenment is supplemented by a final chapter appealing to the
familiar charge that while some uses of language are indeterminate, every
utterance does not put itself into an abyss of possible different
interpretations in the way that postmodernists have suggested. (A view is
cautiously attributed to enthusiasts for Derrida more than the man himself.)

Be that as it may, what is most original here
is the attempt not just to pronounce ‘an obituary’ on ontology but to spell out
the case of its demise in some detail. Other moral realists (such as Iris
Murdoch, who is always an important background influence on Putnam) have
flirted with a similar position, rejecting appeal to any special class of moral
properties, but Putnam’s is probably the best thought-out version of how to
advance in this direction.

In brief, his argument takes the form of a
criticism of Quine’s concept of ‘ontological commitment’ in the paper "On
What there Is". Quine claimed that by quantifying we commit ourselves to
the existence of the range of objects that we are quantifying over. For
example, in quantifying over numbers, by saying ‘There are numbers
greater than a million’ we commit ourselves to the existence of numbers. In response
Quine anticipated the charge that this is just a manner of speaking.
However, unless some other replacement way of speaking is given, such a
response amounts to unsubstantiated hand-waving. If, instead, one does provide
a replacement (for example replacing talk of numbers with talk of sets) then it
will, in turn, be committed to the existence of the range of objects over which
it, in a less offhand manner, now quantifies. (Following Quine, we may
call these objects them ‘abstract entities’.) 
Given that mathematics carried such commitment, Quine accordingly but
embraced a reluctant Platonism of a sort that Putnam views as suspect and
highly uneconomical.

Against it, Putnam points out that there are
alternative ways of formulating mathematics. Numbers can be identified with sets
(Quine’s move) but they can also be identified with functions. These are
equivalent optional languages, equivalent in the sense that they allow us to do
the same sorts of things. So what attitude should we take towards them? Which
one is supposed to get the ontology right? This is a question over which Quine
has vacillated. Put in these terms, unless we start off as Platonists,
it is difficult to see that there is in fact an issue to resolve. Quine’s
problem becomes even greater once we take into account more recent work in
mathematics which entirely avoids quantifying over abstract entities and
instead formalizes in modal logical language. Quine rejected the latter, in
part, because it made ontological commitments unclear even though it allows for
everything else we want a formalization of mathematics to do. Ontological
commitment appeared initially as a characteristic feature of mathematics but
subsequently is seems to be an imposed normative requirement, a tail that
starts to wag the dog.

Outside of mathematics, similar considerations
apply, albeit with an important modification. Quine held that quantifications
of everyday messy language had no ontological significance. (A move that looks
suspiciously like the hand-waving he rejected for other domains of discourse.)
Only our first-class conceptual system, science, (or at least some properly formalized
version of it) really commits us to a view of what there is. But here
again, we can point out the possibility of alternative equivalent formulations
(equivalent, again, in terms of what they will allow us to do) but not in terms
of Quineian ontological commitment.

Putnam’s argument tries ‘to inoculate some
readers’ against thinking that the resulting disputes do any useful work, that
there is anything to be settled between such equivalent formulations. There is
no single bounded and limited sense of ‘exists’ that is at issue in such
disputes. Ontology in that sense ‘has long since outlived its
usefulness’. Perhaps Putnam is right about this in the case of mathematics.
Indeed, it is a domain within which I am inclined to think that Putnam must
be right on pain of an odd sort of Platonism. However, it does not strike me as
obvious that all of the different domains that he wants to hold together can be
held together, and that ontology can accordingly be banished from ethics by
analogy with its redundancy in mathematics.

That is to say, there may be a role that a
certain kind of ontological ambiguity (and disputes about ontology) can and
does play within ethics that it does not play in other domains of discourse.
(Although it would take some work to show just what this role happened to be.)
My concern, my reservation about what is otherwise an appealing argument, is
that any such ambiguity will be banished from Putnam’s approach, every bit as
much as it was previously banished by Quine. Both cherish a certain kind of
hostility towards vagueness, a preference for the simplicity which
Putnam has elsewhere recognized as a normative value. But it is not clear that the
same kind of simplicity that is to be desired in mathematics, is
transferable into the domain of the ethical, although some further argument not
presented might be given to show that this is indeed the case.

©
2006 Tony Milligan

Tony Milligan completed his doctorate on Iris
Murdoch at Glasgow University where he currently tutors in philosophy. He also
teaches philosophy with the Lifelong Learning Centre at the University of
Strathclyde.

Categories: Philosophical, Ethics