1. The Problems of Modal Realism
& Possible Alternatives
Robert Morien
990-41-5960
Philosophy 685 – Modality
The University of Wisconsin-Milwaukee
2. I. Introduction
Historically, modal realism; or possible world semantics was developed as a mechanism
to aid with correctly interpreting the epistemic modals necessarily, possibly, could,
should, probably, etc. when embedded in sentences. These epistemic modals are required
in order to speak coherently about the things that surround us. For example, the use of
could is used to mark a difference between such things as talking donkeys which could
exist, and a married bachelor which could not exist. Furthermore, epistemic modals are
used when making assertions about moral actions. Morality then allows for such claims
like that should have been done in order to prevent that, or action should not precede
thought.
However, disagreement over the correct interpretation of these modal notions has led to
the formation of several possible worlds formulations. Some of these include modal
realism, modal fictionalism, ersatzism, modal antirealism, etc. These disagreements have
led to an ambiguity in epistemic modal usage that is present in the philosophical
literature.
It is the purpose of this paper to critically examine how the different theories of modal
realism, and fictionalism, suggest the expression of modal claims through the concept of
possible worlds. After providing a detailed explanation about the version of modal
realism advanced by David Lewis, I will present three of the common objections against
his view along with how he responds to these objections. After hopefully accomplishing
these things, I will then attempt to show how Gideon Rosen’s description of fictionalism
attempts to provide a better alternative for proposing modal claims. To conclude I will
provide my own assessment as to which theory gives the strongest compelling evidence
to advance their claim.
II. The Thesis of Plurality of Worlds
In the first chapter of David Lewis’ book, On the Plurality of Worlds, Lewis promotes
modal realism, or a thesis of plurality of worlds. This view about the plurality of worlds
resembles the different possible ways a world might have been, and one of these
possibilities is the actual world; that is, the world we are a part of. Possible worlds are as
concrete and real as the actual world. They are separate entities with no overlap, isolated
from any external spatio-temporal property, such that the consequences that result from
one world cannot affect the outcome from any other world. This is the framework Lewis
provides to advance his thesis of modal realism, the framework which ought to deliver
coherent accounts about modal notions.
Lewis tries to sell his version of modal realism by drawing a parallel between the
plurality of sets in mathematical space with the plurality of worlds in logic space. He
claims that it is from this harmony between sets and worlds that allows for an analog
between the fruitfulness of set theory and the theoretical utility of modal realism. It is
due to this fruitfulness found in mathematical set theory that we should accept the theory
3. of possible worlds as being a beneficial mechanism from which an abundance of modal
claims can be produced.
To illustrate how Lewis’ framework for possible worlds yields modal claims, consider
the following example. Hubert Humphrey might have been elected president if and only
if there exists a world W (concrete, discrete, spatio-temporally isolated, etc.) such that at
W, Hubert Humphrey wins the presidential election. So, although in the world we
inhabit, Hubert Humphrey never became president, we can nevertheless suppose,
consider, imagine, entertain, etc. a Presidential Humphrean world.
Alternatively, we can also imagine worlds in which something does not exist. In the case
of Adolph Hitler then, we know that he did exist in the actual world, that he did possess
shortcomings in moral hygiene, and that he did oversee inhumane mistreatment over
millions of people. Analogous to the Hubert Humphrey example given above, we can
imagine a world in which Adolph Hitler might not have existed if and only if there exists
a world W (concrete, discrete, spatio-temporally isolated, etc.) such that at W, Adolph
Hitler does not exist, i.e. Adolph Hitler never existed in the possibly dictatorless German
world of the 30’s and 40’s.
So, possible worlds can work in both ways; there can exist a possible world in which
donkeys can talk and there can exist a possible world in which Germany has always been
under democratic rule.
III. Cases against Possibilia
Several objections have been introduced that try to discredit Lewis’ explanation for
possible worlds and his claim that possibilia generate modal claims. Of the numerous
objections proposed, I will only consider three which I hope will suffice to provide
enough evidence to support one of the theories mentioned above. The three objections I
will consider include:
1) The Problem of “Transworld Identity”
2) The Epistemic Objection
3) The Incredulous Stare
The Problem of Transworld Identity
Consider the following claim:
(a) Hubert Humphrey might have been elected president of the United States.
This certainly seems like a possibility that could have been true, had he not lost to
Richard Nixon in the 1968 presidential election. There is not a decree which
required the presidential election to turn out the way it did. It could be the case
that Mr. Humphrey could have been elected president of the United States of
4. America. Therefore, according to possible world semantics, (a) is supposed to
correspond to:
(b) There is a possible world in which Hubert Humphrey became president of the
United States.
This also seems like a plausible claim. We can certainly entertain the thought of
the possibility of Hubert Humphrey winning the presidential election (although
not in the actual world) in which he enjoyed a life of political delegation, and
policy making.
The question arises, however, from the properties described by Lewis about
modal realism; if possible worlds are concrete, distinct objects with no overlap,
how does a spatio-temporally isolated world allow for the existence of the same
entity in more than one world? Given the framework provided by Lewis, it seems
irrational that an entity can exist in more than one world since the existence of
multiple entities in different worlds conjures up thoughts of a causal necessitation
that must exist between the entities.
Lewis suggests two solutions to this dilemma. His first suggestion is that there is
a world where somehow Hubert Humphrey satisfies the winning of the
presidential election in absentia. In other words, he need not inhabit a world in
order to satisfy formulas such as ‘Humphrey is president’ there. Lewis’ second
suggestion involves a counterpart theory. In modal terms, let the diamond (in
modality the diamond symbol represents possibly) quantify over other-worldly
counterparts of this-worldly individuals. Through vicarious satisfaction he
possibly wins the election if and only if at some world W he has a counterpart
who wins.
Our free will enables us to imagine, suppose, entertain, etc. possible entities
which do not exist in the actual world. We can, therefore, imagine such things as
a world where Hubert Humphrey inhabits the white house, or a world in which
donkeys can talk and the swans are blue, or whatever. This freedom to imagine
compels me to agree with Lewis’ counterpart theory in the fact that orating
donkeys and swans that are blue might have exist if there is a world where in fact
they do exist.
The Epistemic Objection
After seeing Lewis’ reply to the transworld identity problem let’s see how he
responds to the epistemic objection. The epistemic objection is analogous to the
problem of transworld identity in that; given the fact that possible worlds are
concrete, distinct objects, having no spatio-temporal entities which can penetrate
into another possible world, nor affect the outcome in that world, how can we
know facts about that world? In other words, how can our minds transcend into
another possible world and know facts that occur there.
5. The spatio-temporal separation of each possible world implies that we cannot see,
hear, touch, smell, taste, or feel anything in that world. Nor can we get in a hover
craft and fly over to a possible world to find out facts about that world since
observation, measurement, identification, etc. are all empirical, spatio-temporal
properties which must transcend from this world to that in violation of Lewis’
characterization for the properties of possible worlds.
Lewis responds to this objection by saying that our minds operate under a
recombination principle. There is an infinite number of possible combinations
that we can imagine which would sooner or later provide us with the possibility
that the fact we are considering obtains in a particular possible world.
Mathematics helps Lewis out here. We are not causally connected to objects
which are known a priori such as the integers, yet we can come to have
knowledge of them through a systematic set of rules, i.e. axioms, theorems, and
the like which the integers must follow. Similarly, in modal realism, we come to
have knowledge about things in other worlds by means of systematic rules our
imaginations are governed by. In particular, the systematic set of rules which
govern our imaginations in this construction is the recombination principle. Since
our observations confirm or deny these a priori imaginings, the same will hold
true for knowledge acquisition about objects in other worlds.
I side with Lewis here since, if our minds are infinite (there are no restrictions in
place to limit where our imaginations can take us) as they are commonly thought
to be, then we surely should be able to combine an infinite amount of possibilities
until we reach the right one in a particular possible world. In other words, I can
conceive of a way in which donkeys can talk and swans are blue. I imagine a
world in which donkeys have been domesticated to the point that they begin to
take after their masters. So much so, that they have developed the habit of talking
back whenever they are used to haul heavy loads, something which has irritated
the donkeys for thousands of years. Furthermore, in a different possible world, a
mutation has occurred in the genes of the swans. This mutation has caused a
change in the natural white pigment to a color of blue. The validity of Lewis’
recombination principle is evident here from the ease from which these possible
worlds can be imagined as well as the infinite nature of mental recombination.
The Incredulous Stare
Some might say that the viewpoint of possible worlds just plainly seems
implausible, contrary to common sense opinion. Possible worlds, under Lewis’
framework, would allow for the construction of any artifact through the
recombination principle or any other mental permutational device. Therefore,
negating the fact that there exist countless talking donkeys, yet hold that there is
still a possibility that we may discover a land someday inhabited by these talking
donkeys seems to be a bit bizarre. The reason for this peculiarity is that an
6. orating ass seems to go against common sense opinion. In particular, neither a
priori nor a posteriori knowledge has ever revealed anything nonhuman that
possessed the capability of oration.
As Stalnaker states,
… for this to be convincing, the shift from ‘ways things might have been’ to ‘possible
worlds’ must be an innocent terminological substitution, and I do not believe that, as
Lewis develops the concept of a possible world, it is (Stalnaker [1984]: 26).
In other words, an ‘innocent terminological substitution’ in the possible worlds
framework might be something like making the assertion that Hubert Humphrey
might have been president, therefore, there is a possible world in which President
Humphrey reads his newspaper in front of the fire in the white house. This does
appear to be a case of justifiable language exchange since this claim seems to be
coherent. On the other hand, making the assertion that there might have been
talking donkeys, therefore, there exists a possible world in which donkeys talk
back to their masters after being loaded to the max does not appear to be a
justifiable terminological exchange. This assertion appears to be incoherent since
donkeys that talk lie outside of our normal common sense opinion. This anomaly
is what is commonly referred to as
… the incredulous stare. (Rosen [1990]: 329)
Lewis can reply to this objection by saying that the incredulous stare isn’t much
of an objection since the mere suggestion of credulity is not an argument.
I can think of several cases, however, where we do form beliefs that seem to go
against common sense opinion. The reason for siding with these alternative
beliefs which contradict our common sense notions is a conviction that they will
provide us with a better theory of the way the world actually is. In abandoning
determinism, the position of a particle can be better described through the wave
function generated by the time dependent Schrödinger equation. This was
demonstrated by the double slit experiment in which a stream of particles reach a
target in varying concentrations at widely separated points creating a diffraction
pattern of unpredictable locale. This was disanalogous with the locational
predictability of a stream of bullets fired from a gun which hit its target at
precisely the same spot every time. This indeterminism in a particles position
during a double slit experiment indicates that the position variable is a wave-like
object. Further evidence is provided by the de Broglie wavelength which states
that a particles momentum is inversely proportional to its wavelength. Just as the
position of a wave is indeterminate due to the span of its wavelength, so too is a
particles position due to its behavior in tests such as the double slit experiment.
Just as this experiment provided us with an improvement in our understanding of
objects on the quantum scale and, therefore, an improvement in our common
sense beliefs, so too does modal realism. I, therefore, side with Lewis here again
7. since it seems to me that the mere belief that there are no talking donkeys is not
reason enough to abandon the notion of modal realism.
III. Does Fictionalism Provide a Better Alternative?
In Gideon Rosen’s paper, Modal Fictionalism, Rosen proposes a method to employ
modal realism without adhering to any of its propositional content. According to this
new theory, one can utter a proposition without commitment to a belief in its claim
regardless of the truth value. This is what is commonly done in works of fiction where
assertions are made based on the context the facts of that work are found in. Take the
garden variety claim, “There is a brilliant detective at 221b Baker Street.” If interpreted
as an existential claim, this assertion is false. If, however, this claim is interpreted in
reference to a particular fictional context, such as, “In Sherlock Holmes stories, there is a
brilliant detective at 221b Baker Street,” the assertion is true. In this case, there is usually
a mutual agreement between conversants that the antecedent, “In Sherlock Holmes
stories,” refers to a work of fiction rather than to the residency of British detectives.
According to Rosen, the above Holmes example can be carried over into possible world
semantics. In regards to talking donkeys then, the literal interpretation of the claim,
“there exists a world in which donkeys talk,” would yield a false truth value assignment.
However, when interpreted in the context of how the modal realists view the framework
of possible worlds, we get, “according to the hypothesis of modal realism, there is a
world in which donkeys talk,” which is clearly true since this is indeed the claim which
modal realists advance. By claiming that possible worlds are exactly the kinds of things
the modal realists say they are, i.e. concrete, discrete, and as real as the actual world,
fictionalists are allowed to utter propositions which are guaranteed to be true without ever
needing to believe in any of its contents.
…one characteristic feature of fictions is that we may legitimately have opinions about
their contents without believing that what they say is true, or that the objects they purport
to describe exist (Rosen[1990]: 331).
This allows for the advancement of philosophical views through the conceptualization of
possible worlds without needing to hold any belief at all in possibilia. With this
construction then, fictionalists are able to overcome the incredulous stare since they are
not really abandoning common sense. They are just shifting the responsibility of
interpretation to the modal realists since it is according to them that there is a possibility
of a world in which donkeys talk back to their masters.
Fictionalists also seem to get out of the problem of transworld identity as well as the
epistemic objection by applying the same defense mechanism, “according to the modal
realists about possible worlds, …” when making any philosophical assertion.
Fictionalists maintain, therefore, that modal realism is not true. Possible worlds is only
what David Lewis says it is, concrete, discrete, spatio-temporally isolated, etc. and is by
no means to be taken as factual.
8. Based on the fictionalists assertions, I do not believe that they advance a persuasive
argument as to why we should adhere to their views. By saying that a theory is false and
using that claim to advance your own ideals seems to be a bit hypocritical.
IV. So, Which View Should we Accept Anyways?
After all of these considerations about the modal realist’s and the fictionalist’s claims
about epistemic modal output through the use of possible worlds semantics, David
Lewis’ argument appears to hold more weight over the fictionalist’s account. Lewis
provides us with the properties these possible worlds possess as well as the modalized
output these possible worlds provide us with. The content of his views seem to be
carefully thought out in great detail in order to provide us with a framework which is
efficiently employable in order to obtain modal propositional output.
The fictionalists on the other hand do not state clearly what it is we can hope to gain by
claiming that modal realism in fact false. They simply claim that asserting possible
world claims within the fiction of modal realism, fictionalists are allowed to utter
propositions which are guaranteed to be true without ever needing to believe in any of its
contents. Though optimistic, this really does not sound very convincing since the use of
falsehood claims (as proposed about modal realism by the fictionalists) to advance your
own claims just sounds like a bad idea.
The mere ease of use of the recombination principle along with the hope that employment
of the possible worlds framework will provide us with a better theory of the way the
world actually is makes modal realism the more appealing choice.