J. GOODMAN 23
imagined, hence what is to be true in a fiction, when certain
circumstances obtain. The principles that generate fictional
truths appeal to the germane conventions (conversational, be-
havioral, etc.) present in the population of people engaged with
the story. So, for example, it may be prescribed that one imag-
ine that there exists a pipe-smoking detective named ‘Holmes’
whenever one reads a sentence in A Study in Scarlet that ap-
parently quantifies over such a detective (just as it may be pre-
scribed that one imagine that there exists a chocolate pie
whenever one sees a round clump of mud).
So, any proposition p that is rightly prescribed to be imag-
ined by the conventions present in the population engaged with
a literary fiction F is something that is true in F. Some of these
propositions will be directly constitutive of F (e.g., the ones
expressed by sentences originally tokened by Sir Conan Doyle),
but not all of them. Some of the propositions that a population
is prescribed to make-believe are propositions that are culled
from the background. One rightly imagines that London is the
most populous city in England when reading A Study in Scarlet
even if “London is the most populous city in England” is no-
where tokened in any story-copy. In virtue of its being imported
in this way into the story, this proposition is just as true in the
fiction as a ny proposition tha t is dire c tly c onstitutive of the story .
Of course, not all propositions in the background of a story
ought to be imagined by those engaged with the story. The
propositions that are not prescribed to be imagined are those
that the conventions in place make irrelevant to the subject
matter of the story; if there is no convention that generates the
approp riate imagin ative acts i n the po pulat ion engaged with th e
story regarding some background proposition p, then p is not
true in the story. For example, one ought not imagine that
Holmes’ pipe was manufactured by a wizard when reading A
Study in Scarlet. The relevant proposition is utterly irrelevant to
the subject matter of the story. Thus, the proposition is rightly
not imported into the story, is thus not indirectly constitutive of
the fiction, and is thus not true in A Study in Scarlet5.
So, it is via the appropriate conventions present in a popula-
tion that propositions in the background for a fiction F become
true in F. We should also note that the members of the back-
ground that are eligible for importation need not be true propo-
sitions; some false propositions may rightly be imported into
some stories6. What’s central to the pretense account is simply
whether the person engaged with the fiction is indeed pre-
scribed to imagine p by the appropriate conventions of the
game of which that fiction is a prop. A person may be engaged
with a fiction F and mistakenly think that p ought to be im-
ported, but let us say that when p is rightly imported by some-
one into F given the germane co nventions in place, that p erson
is compet ently engaged with F.
Premise 2 is obviously the crux of the above argument, and
its force can now be made clear. It’s possible that there be dis-
tinct people, each of whom are competent readers of some fic-
tion F, yet some proposition p is prescribed by F given the
relevant conventions in the population to which the first person
belongs, while it is simultaneously not the case that p is pre-
scribed by F given the relevant conventions in the population to
which the second person belongs. On Walton’s account, it thus
follows that p will be both true in F and it will not be the case
that p is true in F7. (I of course will not be arguing for the claim
here that there are no fictions in which p is both true and not
true; of course there are “fictional worlds” that internally in-
volve contradiction in this way. But if there is any case in
which ‘in fiction F, p’ is true yet it’s also not the case that “in F,
p” is true, then we are left with a “real world” contradiction.
And dialetheist8 views notwithstanding, this seems to me be to
be an intolerable breach of lo gic.)
Consider the following short work of fiction. Call it Earth
Story.
“Once upon a time on planet Earth, the very planet we live
on, there were intelligent non-human creatures that walked,
talked and behaved very much like we behave. The end”.
Suppose Alex is part of a population of people here on Earth
and Alex engages himself with Earth Story. And suppose fur-
ther that everyone in Alex’s population has reason to believe
that Earth is the shape of a flat disc roughly 10,000 miles in
diameter resting on the back of a turtle. The above copy of
Earth Story never includ es any sent ence token s about the shape
of the Earth, nor do any of its sentence tokens entail anything
about its shape. Part of what’s true in Earth Story, though, by
virtue of what’s directly constitutive, is the proposition that
intelligent non-humans once lived on Earth. Given the relevant
conventions present in Alex’s population, Alex is competently
engaged with Earth Story; he thus rightly imports into the story
the background proposition that intelligent non-humans once
lived on a disc that is roughly 10,000 miles in diameter resting
on the back of a turtle. So, the proposition that intelligent non-
humans on ce walked o n a disc t hat rests on t he back o f a turtle,
e.g., will be true in Earth Story .
5The issues surrounding which propositions are relevant or irrelevant to a
story and the mechanisms by which a proposition gets rightly imported o
rightly fails to get imported are ones that are incredibly complex. At the very
least, Gricean conversational maxims and other subtle conventions of com-
munication and behavior must be appealed to in order to do these issues
ustice. However, I do not think these matters are hopelessly complex, and
they are ones that, in principle, could be dealt with coherently and com-
pletely. For my purposes here, however, all we will need is an example
whereby—no matter what the specifics are surrounding the germane con-
ventions in the relevant populations—Walton’s theory entails that some par-
ticul ar proposition bo t h s ho uld a n d s ho uld no t be im po r t e d in to some s to r y.
6Philos ophers ofte n speak of background facts th at get import ed into stori es.
However, as (Lewis, 1983: p. 274) points out, the proposition that there
exists an animal that breathes fire, e.g., may rightly be imported into a typi-
cal story about dragons. Lewis would call this a case of “inter-fictional
carry-over” of a proposition into a story rather than a case of importation o
a proposition from the background. However, the false proposition that
something breathes fire, I assert, may be true in a fiction F even when that
proposition is not imported from any other fiction G (or even any genre o
fiction). A population may (wrongly) believe in fire-
reathing animals for
scientific reasons. What seems to matter is merely whether or not the proposi-
tion in question is relevant in the right sort of way to the subject matter of F.
Now consid er Boris. He en gages h imself with Earth Story as
well, yet Boris is part of a population who (reasonably) all be-
lieve that Earth is an oblate spheroid. Given the relevant con-
ventions present in Boris’s population, Boris is competently
engaged with Earth Story; he thus rightly imports into the story
the background proposition that intelligent non-humans once
lived on an oblate spheroid. Further, it will not be prescribed by
any convention in Boris’s population that Boris imagine that
intelligent non-humans once lived on a flat and finite disc
roughly 10,000 miles in diameter resting on the back of a turtle.
So, the proposition that intelligent non-humans once walked on
7(Lewis, 1983: p. 272) addr esses thi s sort of problem for hi s own view, but
manages to cut off this worry by rigidly tying what’s imported into a story to
the prevailing beliefs of the community in which the story originated.
8See, e.g., (Priest, 1987) .