without claiming that they are eternally actual.”Coombs (1993: 450) quotes Punch as
holding that the solution lies in positing“a certain diminished being, so to speak, an
intermediate being between beings of reason and actual being without qualification.”
It is not plausible to identify possibilia with beings of reason. However, we can
develop a view inspired by Punch’s remarks. Let us distinguish between two versions
of the most extreme kind of modal realism. Both versions agree that concrete possible
worlds other than the actual one exist, but one version demotes the mode of being of
non-actual concrete possible worlds to being-by-courtesy, whereas the other grants
them full reality. The former view holds that possible beings enjoy a“diminished”
kind of being. Note that the question of whether concrete possibilia are mere beings
by courtesy is not settled by any of the arguments Lewis (1986) gives for the existence
of concrete possible worlds. (In section 2.5.4 I discussed a kind of modal realism in
which the possible and the actual have different modes of being, but did not take the
further step of arguing that the possible have an inferior mode of being.)
Fictional entities are beings by courtesy. In fact, they are excellent examples of
beings of reason in the medieval sense. As I see things, all beings of reason are beings
by courtesy, but the converse doesn’t hold. Many scholastics mistakenly held
that almost nothings are beings of reason, but they were not mistaken in thinking
that there were beings of reason. Chimeras and other imaginary objects, creatures
of myth andfiction, and characters of dreams all are excellent candidates for
being less than fully real as well as being mind-dependent in exactly the way the
scholastics thought.^39
We can entertain a further expansion of being-by-courtesy to include merely
intentional objects of all varieties, even those for which it is metaphysically impos-
sible that they be actual. A kind of qualified Meinongianism in which the merely
intentional enjoy being-by-courtesy might prove to be a defensible position.^40 As
noted in section 1.5.2, I don’t think that this is a view that the historical Meinong
could accept, since by my lights he held that the outermost quantifier—the one that
ranges over absolutely everything there is—is a perfectly natural expression, whereas
on the view discussed here, it is not. On my preferred interpretation, Meinong also
makes use of two other perfectly natural semantically primitive quantifiers, a sub-
sistence quantifier that ranges over obtaining abstracta and an existence quantifier
(^39) Sol Kim has suggested to me that the mode of being of holes and the mode of being offictional
entities should not be conflated, since holes are spatiotemporally located, causally potent, and mind-
independent, whereasfictional entities are none of these. Moreover, these differences could be explained by
the difference in mode of being. Intuitively this is right. Fictional entitiesexist instories, dreams, etc. This
suggests a fourth model forfictional characters, one that combines grades of being with orders of being. On
this fourth model, the mode of being offictional characters is both relative and non-fundamental. Like
attributes,fictional charactersexist insomething else, but unlike attributes, their relative mode of being is
not particularly natural. If we want to distinguish the mode of being of holes, shadows, etc. from that of
fictional characters, this is one way to do this—holes have a not particularly natural yet absolute mode of
being, while 40 fictional entities have a not particularly natural and relative mode of being.
Berto (2013: 74–6) alludes to but rejects something like the position developed here.