4.5 Restricting Recombination
I’ll now explore how someone could use the meta-ontology defended here to
articulate and justify restrictions on the principle of recombination. A friend of the
meta-ontology needn’t follow me here, and there might be other ways of appealing to
modes of being when formulating and justifying various principles of recombination.
But it is useful to see whether a way is possible.
Consider this powerful principle of recombination: every logically consistent
sentence wholly composed of perfectly natural expressions is possibly true. (It is
much like the principle of recombination roughly stated in section 4.2.) This power-
ful principle seems to quickly lead to trouble. On the assumption that“is a member
of,”“is a material object,”and“something”are perfectly natural expressions, this
principle delivers the verdict that it is possible that some material object has a member.
This seems absurd: it is a“category mistake.”The principle must be restricted in some
way to rule this out. The hard task is to justify some particular way of ruling this out.^43
What should we think of“category mistakes”? Gilbert Ryle (1949, 1971) intro-
duced the technical term“category mistake.”What did he mean by it? Some of
Ryle’s examples suggest the following explication: one commits a category mistake
when one believes that a term functions as a referring expression (when in fact it
does not) and uses that term accordingly. Suppose you and your friend, Dr. McX,
attend a football game. You comment on how impressed you are with the team’s
spirit. Dr. McX sees the team, sees the team’s coach, but is unable to see the
team’s spirit. He then concludes that the team’s spirit must be an entity that he
can’t see (perhaps the team’s spirit is an immaterial substance) but nonetheless
causally interacts with the team. After all, when the team’s spirit is strong, the team
plays well, and when the team does poorly, the team’s spirit suffers. Dr. McX has
committed a paradigmatic Rylean category mistake.^44 And it also seems that Dr.
McX’s mistake stems from (if it is not identical with) his using a non-referring
expression as if it were a referring expression. (We use expressions like“the team’s
spirit”to say something about the behavior of the team, but we do not use the
expression as a referring expression. There is no intention to refer to any entity at all
with that term.)
However, although this explication of“category mistake”is serviceable for certain
purposes, this doesn’t seem to be what Ryle meant.^45 Recall that Ryle held that
(^43) The principle of recombination formulated above is very powerful, and might generate bothde dicto
andde remodal possibilities, depending on what is in the metaphysically perfect language. We might wish
to focus on a principle of recombination that yields onlyde dictopossibilities. A weaker principle of
recombination, one that yields onlyde dictopossibilities, is this: every logically consistent sentence in
which no proper name or other rigid designator appears that is wholly composed of perfectly natural
expressions is possibly true. 44
45 Compare with Ryle (1949: 17).
As Magidor (2013: 9–10) notes, it is difficult to understand what Ryle means by“category mistake”
and also what exactly the category mistake committed by Cartesian Dualism consists in.